1

CHAPTER 1 1
2
Introduction, Chromatography Theory, and Instrument Calibration 3
4
1.1 Introduction 5
6
Analytical chemists have few tools as powerful as chromatography to 7
measure distinct analytes in complex samples. The power of chromatography 8
comes from its ability to separate a mixture of compounds, or analytes, and 9
determine their respective identity (chemical structure) and concentration. 10
Chromatography can be divided into three basic types that include gas, liquid, 11
and supercritical fluid chromatography. Liquid chromatography can further be 12
divided into ion exchange, separations based on size, and even extended to gel- 13
based electrophoretic techniques. This book will provide a basic introduction to 14
different types of liquid and gas chromatography. The relationship between each 15
type of chromatography is illustrated in Figure 1.1. 16
17
18
19
Figure 1.1. Categories of Chromatography and Their Relationship to Each 20
Other. 21
22
2
In general, each type of chromatography is comprised of two distinct 1
steps: chromatography (or separation of individual compounds in distinct elution 2
bands) and identification (detection of each elution band). Gas chromatography 3
is the process of taking a sample and injecting it into the instrument, turning the 4
solvent and analytes into gaseous form, and separating the mixture of 5
compounds into individual peaks (and preferably individual compounds). Liquid 6
chromatography completes the same process except the separations occur in a 7
liquid phase. Individual band or peaks exit the column and identification occurs 8
by a relatively universal detector. One particularly common detector for both gas 9
and liquid chromatography is mass spectrometry (MS) which transforms each 10
analyte from a chemically neutral species into a positive cation, usually breaking 11
various bonds in the process. Detecting the mass of the individual pieces 12
(referred to as fragments) allows for conclusive identification of the chemical 13
structure of the analyte. Principles of gas chromatography (GC) will be covered 14
in Chapter 2, liquid chromatography (LC) in Chapter 3, capillary electrophoresis 15
(CE) in Chapter 4 and mass spectrometry (MS) in Chapter 5. 16
17
In mass spectrometry, the combination of compound separation and ion 18
fragment identification (the subject of Chapter 6) yields an extremely powerful 19
analysis that is said to be confirmatory. Confirmatory analysis means the analyst 20
is absolutely sure of the identity of the analyte. In contrast, many other individual 21
techniques and detectors are only suggestive, meaning the analyst thinks they 22
know the identity of an analyte. This is especially true with most universal GC 23
and LC detectors since these detectors respond similarly to many compounds. 24
The only identifying factor in these chromatographic systems is their elution time 25
from the column. In order to obtain confirmatory analysis the sample would need 26
to analyzed by at least two or more techniques (for example, different separation 27
columns) that yield the same results. Mass spectrometry and nuclear magnetic 28
resonance (NMR) are two confirmatory techniques in chemistry. 29
30
At this point, it is important to understand the different applications GC-MS 31
and LC-MS offer for two different types of chemists, analytical and synthetic 32
organic chemists. Organic chemists attempt to create a desired chemical 33
structure by transforming functional groups and intentionally breaking or creating 34
bonds; in their resulting identification procedures they already have a relatively 35
good idea of the chemical structure. To characterize the resulting product the 36
chemist will use Infrared Spectroscopy (IR) to observe functional groups, Mass 37
Spectrometry (MS) to obtain the compounds molecular weight, and Nuclear 38
Magnetic Resonance (NMR) spectroscopy to determine the molecular structure. 39
Information from all three techniques is used to conclusively identify the 40
synthesized product. 41
42
Analytical chemists are forced to approach identification in a different way, 43
because they have no a priori knowledge of the chemical structure and because 44
the analyte is usually present at low concentrations where IR and NMR are 45
inaccurate. Often, analysis is performed to look for a desired compound by 46
3
comparing the sample analysis to that of a known (reference) compound. The 1
reference is used to identify the unknown compound by matching retention time 2
(in chromatography) and ion fragmentation pattern (in mass spectrometry). With 3
todays computer mass spectral libraries that contain ion fractionation patterns for 4
numerous chemicals, the analyst has the option of not using a reference 5
standard. This is especially valuable if a reference compound is not available or 6
is expensive. In some cases, especially with low analyte concentration, this 7
approach may only result in a tentative identification. 8
9
This book will focus on GC-MS and LC-MS applications from an analytical 10
chemistry perspective even though many synthetic chemists will also find much 11
of this information useful for their applications. 12
13
1.2 Chromatographic Theory 14
15
All chromatographic systems have a mobile phase that transports the 16
analytes through the column and a stationary phase coated onto the column or 17
on the resin beads in the column. The stationary phase loosely interacts with 18
each analyte based on its chemical structure, resulting in the separation of each 19
analyte as a function of time spent in the separation column. The less analytes 20
interact with the stationary phase, the faster they are transported through the 21
system. The reverse is true for less mobile analytes that have stronger 22
interactions. Thus, the many analytes in a sample are identified by retention time 23
in the system for a given set of conditions. In GC, these conditions include the 24
gas (mobile phase) pressure, flow rate, linear velocity, and temperature of the 25
separation column. In HPLC, the mobile phase (liquid) pressure, flow rate, linear 26
velocity, and the polarity of the mobile phase all affect a compounds retention 27
time. An illustration of retention time is shown in Figure 1.2. The equation at the 28
top of the figure will be discussed later during our mathematic development of 29
chromatography theory. 30
31
32
4
Figure 1.2. Identification of Analytes by Retention Time. 1
2
In the above figure, the minimum time that a non-retained chemical 3
species will remain in the system is t
M
. All compounds will reside in the injector, 4
column, and detector for at least this long. Any affinity for the stationary phase 5
results in the compound being retained in the column causing it to elute from the 6
column at a time greater than t
M
. This is represented by the two larger peaks 7
that appear to the right in Figure 1.2, with retention times t
RA
and t
RB
. Compound 8
B has more affinity for the stationary phase than compound A because it exited 9
the column last. A net retention (t
RA
and t
RB
) time can be calculated by 10
subtracting the retention time of the mobile phase(t
M
) from the peaks retention 11
time (t
RA
and t
RB
). 12
13
Figure 1.2 also illustrates how peak shape is related to retention time. 14
The widening of peak B is caused by longitudinal diffusion (diffusion of the 15
analyte as the peak moves down the length of the column). This relationship is 16
usually the reason why integration by area, and not height, is utilized. However, 17
compounds eluting at similar retention times will have near identical peak shapes 18
and widths. 19
20
A summary of these concepts and data handling techniques is shown in 21
Animation 1.1. Click on the figure to start the animation. 22
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
61
63
65
Animation 1.1. Baseline Resolution. 66
67
5
Chromatographic columns adhere by the old adage like dissolves like to 1
achieve the separation of a complex mixture of chemicals. Columns are coated 2
with a variety of stationary phases or chemical coatings on the column wall in 3
capillary columns or on the inert column packing in packed columns. When 4
selecting a columns stationary phase, it is important to select a phase 5
possessing similar intermolecular bonding forces to those characteristic of the 6
analyte. For example, for the separation of a series of alcohols, the stationary 7
should be able to undergo hydrogen bonding with the alcohols. When attempting 8
to separate a mixture of non-polar chemicals such as aliphatic or aromatic 9
hydrocarbons, the column phase should be non-polar (interacting with the 10
analyte via van der Waals forces). Selection of a similar phase with similar 11
intermolecular forces will allow more interaction between the separation column 12
and structurally similar analytes and increase their retention time in the column. 13
This results in a better separation of structurally similar analytes. Specific 14
stationary phases for GC and HPLC will be discussed later in Chapter 2 and 3, 15
respectively. 16
17
Derivation of Governing Equations: The development of chromatography 18
theory is a long established science and almost all instrumental texts give nearly 19
exactly the same set of symbols, equations, and derivations. The derivation 20
below follows the same trends that can be found in early texts such as Karger et 21
al. (1973) and Willard et al. (1981), as well as the most recent text by Skoog et 22
al. (2007). The reader should keep two points in mind as they read the following 23
discussion. First, the derived equations establish a relatively simple 24
mathematical basis for the interactions of an analyte between the mobile phase 25
(gas or liquid) and the stationary phase (the coating on a column wall or resin 26
bead). Second, while each equation serves a purpose individually, the relatively 27
long derivation that follows has the ultimate goal of yielding an equation that 28
describes a way to optimize the chromatographic conditions in order to yield 29
maximum separation of a complex mixture of analytes. 30
31
To begin, we need to develop several equations relating the movement of 32
a solute through a system to properties of the column, properties of the solute(s) 33
of interest, and mobile phase flow rates. These equations will allow us to predict 34
(1) how long the analyte (the solute) will be in the system (retention time), (2) 35
how well multiple analytes will be separated, (3) what system parameters can be 36
changed to enhance separation of similar analytes. The first parameters to be 37
mathematically defined are flow rate (F) and retention time (t
m
). Note that F has 38
units of cubic volume per time. Retention behavior reflects the distribution of a 39
solute between the mobile and stationary phases. We can easily calculate the 40
volume of stationary phase. In order to calculate the mobile phase flow rate 41
needed to move a solute through the system we must first calculate the flow rate. 42
6
1
2
In the equations above, r
c
is the internal column radius, d
c
is the internal column 3
diameter, L is the total length of the column, t
m
is the retention time of a non- 4
retained analyte (one which does not have any interaction with the stationary 5
phase). Porosity (e) for solid spheres (the ratio of the volume of empty pore 6
space to total particle volume) ranges from 0.34 to 0.45, for porous materials 7
ranges from 0.70 to 0.90, and for capillary columns is 1.00. The average linear 8
velocity is represented by u-bar. 9
10
The most common parameter measured or reported in chromatography is 11
the retention time of particular analytes. For a non-retained analyte, we can use 12
the retention time (t
M
) to calculate the volume of mobile phase that was needed 13
to carry the analyte through the system. This quantity is designated as V
m
, 14
15

!
V
M
= t
M
F Eqn 1.1
mL min. mL/min
16
17
and is called the dead volume. For a retained solute, we calculate the volume of 18
mobile phase needed to move the analyte through the system by 19
20

!
V
R
= t
R
F Eqn 1.2
mL min. mL/min
21
22
where t
R
is the retention time of the analyte. 23
24
In actual practice, the analyst does not calculate the volume of the 25
column, but measures the flow rate and the retention time of non-retained and 26
retained analytes. When this is done, note that the retention time not only is the 27
transport time through the detector, but also includes the time spent in the 28
injector! Therefore 29
30

!
V
M
= V
column
+ V
injector
+ V
detector
31
32
The net volume of mobile phase (V
R
) required to move a retained analyte 33
through the system is 34

!
F = ( " r
c
2
) # (L/t
m
)
F = " (d
c
/2)
2
# (L/t
m
)
F =
" d
c
4
$
%
&
'
(
) # (L/t
m
)
where
" d
c
4
$
%
&
'
(
) = cross sectional area of column
# = porosity of column packing
(L/t
m
) = average linear velocity of mobile phase
7
1

!
V
R
'
= V
R
- V
M
Eqn 1.3 2
3
where V
R
is the volume for the retained analyte and V
M
is the volume for a 4
nonretained (mobile) analyte. 5
6
This can be expanded to 7
8

!
t
R
'
F = t
R
F - t
M
F 9
10
and dividing by F, yields 11
12

!
t
R
'
= t
R
- t
M
Eqn 1.4 13
14
Equation 1.4 is important since it gives the net time required to move a 15
retained analyte through the system (Illustrated in Figure 1.2, above) 16
17
Note, for gas chromatography (as opposed to liquid chromatography), the 18
analyst has to be concerned with the compressibility of the gas (mobile phase), 19
which is done by using a compressibility factor, j 20
21
where P
i
is the gas pressure at the inlet of the column and P
o
is the gas pressure 22
at the outlet. The net retention volume (V
N
) is 23
24

!
V
n
= j V
R
'
Eqn 1.5 25
26
The next concept that must be developed is the partition coefficient (K) 27
which describes the spatial distribution of the analyte molecules between the 28
mobile and stationary phases. When an analyte enters the column, it immediately 29
distributes itself between the stationary and mobile phases. To understand this 30
process, the reader needs to look at an instant in time without any flow of the 31
mobile phase. In this snap-shot of time one can calculate the concentration of 32
the analyte in each phase. The ratio of these concentrations is called the 33
equilibrium partition coefficient, 34
35

!
K = C
s
/ C
M
Eqn 1.6 36
37
where C
s
is the analyte concentration in the solid phase and C
M
is the solute 38
concentration in the mobile phase. If the chromatography system is used over 39
8
analyte concentration ranges where the K relationship holds true, then this 1
coefficient governs the distribution of analyte anywhere in the system. For 2
example, a K equal to 1.00 means that the analyte is equally distributed between 3
the mobile and stationary phases. The analyte is actually spread over a zone of 4
the column (discussed later) and the magnitude of K determines the migration 5
rate (and t
R
) for each analyte (since K describes the interaction with the 6
stationary phase). 7
8
Equation 1.3 (V
R
=V
R
- V
M
) relates the mobile phase volume of a non- 9
retained analyte to the volume required to move a retained analyte through the 10
column. K can also be used to describe this difference. As an analyte peak exits 11
the end of the column, half of the analyte is in the mobile phase and half is in the 12
stationary phase. Thus, by definition 13
14

!
V
R
C
M
= V
M
C
M
+ V
S
C
S
Eqn 1.7 15
16
Rearranging and dividing by C
M
yields 17
18

!
V
R
= V
M
+ K V
S
or V
R
- V
M
= K V
S
Eqn 1.8 19
20
Now three ways to quantify the net movement of a retained analyte in the column 21
have been derived, Equations 1.2, 1.4, and 1.8. 22
23
Now we need to develop the solute partition coefficient ratio, k (also 24
knows as the capacity factor), which relates the equilibrium distribution coefficient 25
(K) of an analyte within the column to the thermodynamic properties of the 26
column (and to temperature in GC and mobile phase composition in LC, 27
discussed later). For the entire column, we calculate the ratio of total analyte 28
mass in the stationary phase (C
S
V
S
) as compared to the total mass in the mobile 29
phase (C
M
V
M
), or 30
31
32
where V
S
/V
M
is sometimes referred to as b, the volumetric phase ratio. 33
34
Stated in more practical terms, k is the additional time (or volume) a 35
analyte band takes to elute as compared to an unretained analyte divided by the 36
elution time (or volume) of an unretained band, or 37
38
rearranged, gives 39
40
41
42
43

!
k' =
C
s
V
s
C
m
V
m
= K
V
s
V
m
Eqn 1.9
9
1
3
5
7
8
9
where is the linear gas velocity and the parameters in Equation 1.10 were 10
defined earlier. So, the retention time of an analyte is related to the partition ratio 11
(k). Optimal k values range from ~1 to ~5 in traditional packed column 12
chromatography, but the analysts can use higher values in capillary column 13
chromatography. 14
15
Multiple Analytes: The previous discussions and derivations were 16
concerned with only one analyte and its migration through a chromatographic 17
system. Now we need to describe the relative migration rates of analytes in the 18
column; this is referred to as the selectivity factor, !. Notice Figure 1.2 above 19
had two analytes in the sample and the goal of chromatography is to separate 20
chemically similar compounds. This is possible when their distribution 21
coefficients (Ks) are different. We define the selectivity factor as 22
23
24
where subscripts A and B represent the values for two different analytes and 25
solute B is more strongly retained. By this definition, ! is always greater than 1. 26
Also, if one works through the math, you will note that 27
28
29
The relative retention time, !, depends on two conditions: (1) the nature 30
of the stationary phase, and (2) the column temperature in GC or the solvent 31
gradient in LC. With respect to these, the analyst should always first try to select 32
a stationary phase that has significantly different K values for the analytes. If the 33
compounds still give similar retention times, you can adjust the column 34
temperature ramp in GC or the solvent gradient in LC; this is the general elution 35
problem that will be discussed later. 36
37
Appropriate values of ! should range from 1.05 to 2.0 but capillary column 38
systems may have greater values. 39
40

!
" =
K
B
K
A
=
k'
B
k'
A
Eqn 1.12

!
" =
V'
B
V'
A
=
t
R,B
- t
m
t
R,A
- t
m
=
t'
R,B
t'
R,A
Eqn 1.13

!
t
r
= t
m
(1 + k' ) =
L
u
(1 + k' ) Eqn 1.11

!
k' =
t
r
- t
m
t
m
=
V
s
- V
m
V
m
Eqn 1.10
10
Now, we finally reach one of our goals of these derivations, an equation 1
that combines the system conditions to define analyte separation in terms of 2
column properties such as column efficiency (H) and the number of separation 3
units (plates, N) in the column (both of these terms will be defined later). As 4
analyte peaks are transported through a column, an individual molecule will 5
undergo many thousands of transfers between each phase. As a result, packets 6
of analytes and the resulting chromatographic peaks will broaden due to physical 7
processes discussed later. This broadening may interfere with resolution (the 8
complete separation of adjacent peaks) if their K (or k) values are close (this will 9
result in an a value close to 1.0). Thus, the analyst needs a way to quantify a 10
columns ability to separate these adjacent peaks. 11
12
First, we will start off with an individual peak and develop a concept called 13
the theoretical plate height, H, which is related to the width of a solute peak at the 14
detector. Referring to Figure 1.2, one can see that chromatographic peaks are 15
Gaussian in shape, can be described by 16
17
18
where H is the theoretical plate height (related to the width of a peak as it travels 19
through the column), " is one standard deviation of the bell-shaped peak, and L 20
is the column length. Equation 1.14 is a basic statistical way of using standard 21
deviation to mathematically describe a bell-shaped peak. One standard 22
deviation on each side of the peak contains ~68% of the peak area and it is 23
useful to define the band broadening in terms of the variance, "
2
(the square of 24
the standard deviation, "). Chromatographers use two standard deviations that 25
are measured in time units (t) based on the base-line width of the peak, such that 26
27
28
Here, L is given in cm and t
R
in seconds. Note in Figure 1.2, that the 29
triangulation techniques for determining the base width in time units (t) results in 30
96% if the area or 2 standard deviations, or 31
32
33
34
Substitution of Equation 1.16 into Equation 1.14, yields 35
36

!
" =
#
L
t
R
Eqn 1.15

!
H =
"
2
L
Eqn 1.14

!
W = 4" = 4
#
L
t
R
or # =
WL
4t
R
Eqn 1.16

!
H =
LW
2
16t
2
R
Eqn 1.17
11
1
H is always given in units of distance and is a measure of the efficiency of 2
the column and the dispersion of a solute in the column. Thus, the lower the H 3
value the better the column in terms of separations (one wants the analyte peak 4
to be as compact as possible with respect to time or distance in the column). 5
Column efficiency is often stated as the number of theoretical plates in a column 6
of known length, or 7
8
9
This concept of H, theoretical plates comes from the petroleum distillation 10
industry as explained in Animation 1.2 below. Click on the Figure to play the 11
animation. 12
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
Animation 1.2 Origin of H and the Theoretical Plate Height Unit 60
61
To summarize Animation 1.2 with respect to gas and liquid 62
chromatography, a theoretical plate is the distance in a column needed to 63
achieve baseline separation; the number of theoretical plates is a way of 64
quantifying how well a column will perform. 65
66

!
N =
L
H
= 16
t
R
W
"
#
$
%
&
'
2
Eqn 1.18
12
We now have the basic set of equations for describing analyte movement 1
in chromatography but it still needs to be expanded to more practical applications 2
where two or more analytes are separated. Such an example is illustrated in 3
Animation 1.3 for a packed column. 4
5
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
50
51
Animation 1.3 Separation of Two Analytes by Column Chromatography. 52
53
Separation of two chemically-similar analytes is characterized 54
mathematically by resolution (R
s
), the difference in retention times of these 55
analytes. This equation, shown earlier in Figure 1.2, is 56
57

!
R
s
=
2(t
R' B
- t
R' A
)
W
A
+ W
B
Eqn 1.19 58
59
where t
RB
is the corrected retention time of peak B, t
RA
is the corrected retention 60
time of peak A, W
A
is the peak width of peak A in time units and W
B
is the width 61
of peak B. Since W
A
=W
B
=W, Equation 1.19 reduces to 62
63
64
65

!
Rs =
t
R' B
- t
R' A
W
Eqn 1.20
13
Equation 1.18 expressed W in terms of N and t
R
, and substitution of Equation 1
1.17 into Equation 1.18, yields 2
3
4
Recall from Equation 1.10, that 5
6
substitution into Equation 1.21 and upon rearrangement, yields 7
8
9
Recall that we are trying to develop an equation that relates resolution to 10
respective peak separations and although k values do this, it is more useful to 11
express the equation in terms of !, where ! =k
B
/k
A
. Substitution of ! into 12
Equation 1.22, with rearrangement, yields 13
14
16
18
20
or the analyst can determine the number of plates required for a given 21
separation: 22
24
26
28
30
Thus, the number of plates present in a column can be determined by direct 31
inspection of a chromatogram, where R
s
is determined from Equation 1.20, k
A
32
and k
B
are determined using Equation 1.10, and ! is determined using Equation 33
1.12. 34
35
36

!
R
s
=
t
R' B
- t
R' A
t
R' B
"
#
$
%
&
'
N
4
Eqn 1.21

!
R
s
=
k'
B
- k'
A
1 + k'
B
"
#
$
%
&
'
N
4
Eqn 1.22

!
R
s
=
t
R,B
- t
R,A
W
Eqn 1.20

!
k =
t
r
- t
m
t
m
=
V
s
- V
m
V
m
Eqn 1.10

!
" =
K
B
K
A
=
k'
B
k'
A
Eqn 1.12

!
R
s
=
" - 1
"

k'
B
1+ k'
B

N
4
Eqn 1.23

!
N = 16 R
s
2

" - 1
"
#
$
%
&
'
(
2

k'
B
1 + k'
B
#
$
%
&
'
(
2
Eqn 1.24
14
Another use of Equation 1.23 is that it can be used to explain improved 1
separation with temperature programming of the column in GC and gradient 2
programming in HPLC. Recall that poorer separation will result as peaks 3
broaden as they stay for extended times in the column and several factors 4
contribute to this process. N can be changed by changing the length of the 5
column to increase resolution but this will further increase band broadening. H 6
can be decreased by altering the mobile phase flow rate, the particle size of the 7
packing, the mobile phase viscosity (and thus the diffusion coefficients), and the 8
thickness of the stationary phase film. 9
10
To better understand the application of the equations derived above a 11
useful exercise is to calculate all of the column quantification parameters for a 12
specific analysis. The chromatogram below (Figure 1.3) was obtained from a 13
capillary column GC with a flame ionization detector. Separations of 14
hydrocarbons commonly found in auto petroleum were made on a 30-meter long, 15
0.52-mm diameter DB-1 capillary column. Table 1.1 contains the output from a 16
typical integrator. 17
18
19
20
Figure 1.3 Integrator Output for the Separation of Hydrocarbons by Capillary 21
Column GC-FID. 22
23
15
Table 1.1 Integrator Output for the Chromatogram shown in Figure 1.3. 1
2
Analyte Retention Time
(min)
Area Peak Width at the
Base (in units of
minutes)
Solvent (t
M
) 1.782 NA NA
Benzene 4.938 598833 0.099
Iso-octane 6.505 523678 0.122
n-Heptane 6.956 482864 0.100
Toluene 9.256 598289 0.092
Ethyl Benzene 13.359 510009 0.090
o-Xylene 13.724 618229 0.087
m-Xylene 14.662 623621 0.088
3
Example 1.1 4
Calculate k, !, R
s
, H, and N for any two adjacent compounds in Table 1.1. 5
6
Solution: 7
Using peaks eluting at 13.724 and 13.359 minutes the following values 8
were obtained. 9
10
11
12
Problem 1.1 13
Figure 1.4 and Table 1.2 contain data from an HPLC analysis of four s-Triazines 14
(common herbicides). Calculate k, !, R
s
, H, and N for any two adjacent 15
compounds. Compare and contrast the results for the resin packed HPLC 16
column to those of the capillary column in the GC example given above. 17
18
16
1
2
Figure 1.4 HPLC Chromatogram of Four Triazines. The analytical column was 3
an 10.0 cm C-18 stainless steel column with 2 m resin beads. 4
5
6
Table 1.2 Integrator Output for the HPLC Chromatogram shown in Figure 1.4. 7
8
Analyte Retention Time
(min)
Area Peak Width at the
Base (in units of
minutes)
Solvent (t
M
) 1.301 NA NA
Peak 1 2.328 1753345 0.191
Peak 2 2.922 1521755 0.206
Peak 3 3.679 1505381 0.206
Peak 4 4.559 1476639 0.198
9
10
1.3 Optimization of Chromatographic Conditions 11
12
Now we will review and summarize this lengthy derivation and these 13
complicated concepts. Optimization of the conditions of the chromatography 14
system (mobile phase flow rate, stationary phase selection, and column 15
temperature or solvent gradient) are performed to achieve base-line resolution 16
for the most difficult separation in the entire analysis (two adjacent peaks). This 17
process results in symmetrically-shaped peaks that the computer can integrate to 18
obtain a peak (analyte) area or peak height. A series of known reference 19
standards are used to generate a linear calibration line (correlating peak area or 20
17
height to analyte concentration) for each compound. This line, in turn, is used to 1
estimate the concentration of analyte in unknown samples based on peak area or 2
height. 3
4
An instruments resolution can be altered by changing the theoretical plate 5
height and the number of theoretical plates in a column. The plate height, as 6
explained in the animation below, is the distance a compound must travel in a 7
column needed to separation two similar analytes. The number of theoretical 8
plates in a column is a normalized measure of how well a column will separate 9
similar analytes. 10
11
Now it is necessary to extend the concept of theoretical plate height (H) a 12
bit further to understand its use in chromatography. Since gas and liquid 13
chromatography are dynamic systems (mobile flow through the column), it is 14
necessary to relate a fixed length of the column (the theoretical plate height) to 15
flow rate in the column. Flow rate is measured in terms of linear velocity, or how 16
many centimeters a mobile analyte or carrier gas will travel in a given time 17
(cm/s). The optimization of the relationship between H and linear velocity (), 18
referred to as a van Deemter plot, is illustrated in Figure 1.5 for gas 19
chromatography. 20
21
22
Figure 1.5. A Theoretical van Deemter Plot for a Capillary Column showing the 23
Relationship between Theoretical Plate Height and Linear Velocity. 24
25
18
It is desirable to have the smallest plate height possible, so the maximum 1
number of plates can be contained in a column of a given length. Three factors 2
contribute to the effective plate height, H, in the separation column. The first is 3
the longitudinal diffusion, B (represented by the blue line in Figure 1.5) of the 4
analytes that is directly related to the time an analyte spends in the column. 5
When the linear velocity () is high, the analyte will only spend a short time in the 6
column and the resulting plate height will be small. As linear velocity slows, more 7
longitudinal diffusion will cause more peak broadening resulting in less 8
resolution. The second factor is the multi-flow path affect represented by the red 9
line in Figure 1.5. This was a factor in packed columns but has been effectively 10
eliminated when open tubular columns (capillary columns) became the industry 11
standard. Third are the limitations of mass transfer between and within the gas 12
and stationary phases, C
u
(the yellow line in Figure 3) defined by 13
14
15
16
where is the mobile phase linear velocity, D
s
and D
m
are diffusion coefficients in 17
the stationary and mobile phases respectively, d
f
and d
p
are the diameter of the 18
packing particles and the thickness of liquid coating on the stationary phase 19
particles respectively, k is the unitless retention or capacity factor, and f(k) and 20
f(k) are mathematical functions of k. 21
22
If the linear velocity of the mobile phase is too high, the entire packet of a 23
given analyte will not have time to completely transfer between the mobile and 24
stationary phase or have time to completely move throughout a given phase 25
(phases are coated on the column walls and therefore have a finite thickness). 26
This lack of complete equilibrium of the analyte molecules will result in peak 27
broadening for each peak or skewing of the Gaussian shape. This, in turn, will 28
increase H and decrease resolution. 29
30
The green line in Figure 1.5 represents the van Deemter curve, the 31
combined result of the three individual phenomena. Since the optimum operating 32
conditions has the smallest plate height; the flow rate of the GC should be set to 33
the minimum of the van Deemter curve. For gas chromatography this occurs 34
around a linear velocity of 15 to 20 cm/s. However, in older systems, as the oven 35
and column were temperature programmed, the velocity of the gas changed 36
19
which in turn changed the mobile phase flow rate and the linear velocity. This 1
has been overcome in modern systems with mass flow regulators, instead of 2
pressure regulators, that hold the linear velocity constant. 3
4
These concepts are reviewed in Animation 1.4. Click the figure to start the 5
animation. 6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
49
Animation 1.4 Construction of a van Deemter Curve for an HPLC System 50
51
Now that the theoretical basis for understanding chromatographic 52
separation has been established, it is necessary to extend these ideas one step 53
further. Remember, the power of chromatography is the separation of complex 54
mixtures of chemicals; not just for two chemicals as illustrated previously. In 55
most cases separating mixtures of many compounds is required. This requires 56
that the resolution, R
s
, be constantly optimized by maintaining H at its minimum 57
value in the van Deemter curve. 58
59
This optimization is accomplished by systematically altering the column 60
temperature in GC or the solvent composition in HPLC. Analytes in the 61
separation column spend their time either dissolved in the stationary phase or 62
vaporized in the mobile phase. When analytes are in the stationary phase they 63
are not moving through the system and are present in a narrow band in the 64
length of the column or resin coating. As the oven temperature is increased, 65
each unique analyte has a point where it enters the mobile phase and starts to 66
move down the column. In GC, analytes with low boiling points will move down 67
20
the column at lower temperatures, exit the system, and be quantified. As the 1
temperature is slowly increased, more and more analytes (with higher boiling 2
points) likewise exit the system. In reverse-phase HPLC, analytes with more 3
polarity will travel fastest and less polar analytes will begin to move as the 4
polarity of the mobile phase is decreased. Thus, the true power of GC separation 5
is achieved by increasing the oven/column temperature (referred to as ramping) 6
while in LC the separation power is in gradient programming (composition of the 7
mobile phase). This is the general elution problem that is solved by optimizing 8
the mobile phase, linear velocity, and the type of stationary phase. As noted, 9
temperature programming is used to achieve separation of large numbers of 10
analytes in GC. An example of the effects of temperature programming on 11
resolution is illustrated in Figure 1.6. In Figure 1.6a, a low isothermal 12
temperature is used to separate a mixture of six analytes with limited success as 13
some peaks contain more than one analyte. A higher isothermal temperature, 14
shown in Figure 1.6b, is more successful for analytes with higher boiling points 15
but causes a loss of resolution for peaks that were resolved at the lower 16
isothermal temperature. The temperature program used to produce Figure 1.6c 17
achieves adequate separation and good peak shape for a complex solution. 18
19
21
1
2
Figure 1.6. Temperature Programming: The Solution to the General Elution 3
Problem for GC Applications. 4
22
1
1.4 Calibration of an Instrument/Detector 2
3
We now have a basis for understanding separation science with respect to 4
chromatography. All chromatography systems rely on these principles. But how 5
does the analyst relate instrument output to analyte concentration in a sample? 6
Instruments yield signals (also referred to as responses) that are specific to the 7
type of detector being used. Most GC detectors result in electrical currents while 8
most LC detectors yield absorbance values. MS units can be attached to both 9
GC and LC systems and yield counts of ions per time. But before actual samples 10
are analyzed each instrument detector must be calibrated. Two common forms 11
of calibration are internal and external calibration. 12
13
Detector response yields two useful means of quantification in 14
chromatography: peak area and peak height. In the old days these 15
measurements were made manually; a strip chart recording was obtained by 16
passing a strip of paper consisting of uniform weight past a pen that moved 17
relative to the detector signal. The shape of the peak was drawn on the paper 18
and the peak height was measured with a ruler or the peak area was measured 19
either by triangulation or by actually weighing a cutout of the paper containing the 20
peak! Fortunately for us, these archaic methods are no longer required. The 21
major disadvantage of these techniques is that the range of detector responses 22
was limited by the height of the paper. Today, peak area and height 23
measurements are calculated by electronic integrators or computers, and most 24
systems are automated such that peak area/height are directly correlated 25
between standards and samples. Most systems use peak area to generate 26
calibration lines, which are usually linear relationships between the detector 27
response and the concentration or mass of analyte injected into the instrument. 28
Such a plot is shown for an external calibration method in Figure 1.7. 29
30
23
1
2
Figure 1.7 External Calibration of Benzene on a Capillary Column GC. 3
4
5
A summary of integration concepts is illustrated in Animation 1.5. Click on the 6
figure to begin the animation. 7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
24
1
2
3
4
Animation 1.5 Integration of Chromatographic Peaks. 5
6
After an instrument has been calibrated, a sample extract is analyzed 7
under the same conditions as the standards. The calculated area for the sample 8
is then analyzed by a linear regression of the standard line and a mass or 9
concentration of the analyte in the sample is calculated. Usually a dilution factor 10
adjustment is made and the concentration of analyte in the original sample is 11
then calculated. 12
13
A special type of additional calibration is used in capillary column gas 14
chromatography because of analyte losses during sample injection and due to 15
the possibility of inconsistent injections when manual injections are preformed. 16
This method is referred to as an internal standard where every sample and 17
standard injected into the instrument contains an identical concentration of a 18
compound with similar chemical structure to the analyte but one that has a 19
unique retention time in the column. The instrument is set to measure a constant 20
concentration (and therefore measured area) of the internal standard and adjusts 21
all injections to that constant value. For example, if a sample is found to only 22
contain 90 percent of the internal standard, then it is assumed that 10 percent of 23
the injection was lost and all analyte concentrations are increased by 10 percent. 24
Similarly adjusts can be made of over injecting a sample. 25
26
The next chapters of this book will focus on the components of GC, LC, 27
CE and MS with an additional chapter on interpretation of MS fragmentation 28
patterns. Both GC and LC rely on the chromatography theory discussed in this 29
chapter, but CE requires a different derivation that will discussed in Chapter 4. 30
All instruments rely on some form of calibration if quantitative results are 31
required. 32
33
1.5 Questions 34
35
1. List the three basic types of chromatography. What are the subcategories of 36
each type? 37
38
2. What does confirmatory analysis mean with respect to chromatography? 39
What two ways can it be accomplished? 40
41
3. What are mobile and stationary phases in chromatography? 42
43
4. With respect to GC and LC, what types of gradients are used to improve 44
analyte separation? 45
46
25
5. Hand draw a chromatography with a solvent peak and two analytes. Label 1
components of the diagrams with respect to retention time. 2
3
6. Explain the old adage like dissolves like with respect to chromatography. 4
5
7. What possible intermolecular forces can be involved in like dissolves like? 6
7
8. How are time, gas/liquid volume, and flow rate related in GC and LC? 8
9
9. Why is retention time so important in chromatography? 10
11
10. Explain the concept of capacity factor, k. Is k a factor for one analyte or 12
many analytes? 13
14
11. How is the capacity factor related to retention time of an analyte? 15
16
12. What are the acceptable range for k for traditional chromatography? 17
18
13. How are the capacity factor and the selectivity factor mathematically related? 19
20
14. What is the value/purpose of using the selectivity factor? 21
22
15. What are acceptable values for selectivity factors? 23
24
16. Explain the theoretical plate height, H. What is the origin of H? 25
26
17. How is the total number of plates in a column related to the length of the 27
column and H? 28
29
18. Why is proper column packing so important in packed-column GC and LC? 30
31
19. Explain resolution with respect to chromatography. What is the 32
mathematical relationship between resolution and retention time and peak width? 33
34
20. Explain each component in the governing equation for resolution, equation 35
1.23. 36
37
21. How is equation 1.23 used to improve separation of chemically similar 38
analytes in GC and LC? 39
40
22. Using the data in Table 1.1 and Example 1.1, calculate k, !, R
s
, H, and N for 41
iso-octane and n-heptane. 42
43
23. Why do capillary columns provide higher resolution as compared to packed 44
columns? 45
46
26
24. Draw a van Deemter curve for a GC analysis and explain each factor that 1
contributes to H. 2
3
25. Use figures and words to explain the General Elution Problem. How is this 4
solved in GC and LC? 5
6
26. Why is instrument calibration so important in chromatography? 7
8
27. What is the difference in internal and external calibration? 9
10
11
1.6 References 12
13
Karger, B.L., L.R, Synder, C. Horvath. 1973. An Introduction to Separation 14
Science. 1
st
Edition, J ohn Wiley and Sons, New York, USA 15
16
Skoog, D.A., F.J . Holler, and S.R. Crouch. 2007. Principles of Instrumental 17
Analysis. 6
th
Edition. Thomson Publishing USA 18
19
Willard, H.W., L.L. Merritt, J r., J .A. Dean, F.A. Seattle, J r. 1981. Instrumental 20
Methods of Analysis, 6
th
Edition. Wadsworth Publishing Company, Belmont, CA 21
USA 22
1
Chapter 2 1
2
Basic Gas Chromatography 3
4
2.1 Introduction and History 5
6
This chapter will focus on the components and operation of basic gas 7
chromatography (GC). The general field of chromatography dates back to 8
1903 with the work of Russian scientist Mikhail Tswett, who separated plant 9
pigments using liquid chromatography. Fritz Prior, as part of his graduate 10
work, developed solid-state gas chromatography in 1947. Modern gas 11
chromatography is generally considered to have been invented by Martin 12
and J ames in 1952. A review of the history of gas chromatography can be 13
found in Bartle and Myers (2002). Since 1952, gas chromatography has 14
advanced from using solid spheres to act as the stationary phase (gas-solid 15
chromatography) to using liquid coated resins as the stationary phase, and 16
finally to using covalently-bonded stationary phases attached to wall of a 17
capillary column (gas-liquid chromatography). Components of the actual 18
chromatograph have also advanced from many manual parts such as rotary 19
gas flow regulators being updated to electronic flow or mass flow 20
controllers, resin packed column have been replaced with fused silica 21
capillary columns, and manual injection and control of the instrument has 22
been replaced with automated injection and computer control. Most notably 23
is the diversity of detectors utilized today with GC, especially the ability to 24
connect capillary column GCs with mass spectrometers for confirmatory 25
analysis. Additionally, in the past, analyses of a set of samples could take 26
days to complete and required the constant attention of an analyst, but today 27
with the help of computers, a set of samples and standards can be started on 28
the instrument and the scientist can return later with all of the samples 29
analyzed and the data processed. Typical automatic sampler units can hold 30
up to 100 samples. These improvements have greatly increased the 31
capabilities of laboratories and advanced scientific endeavors but in many 32
cases have decreased the analysts knowledge of the chromatographic 33
system. But such is the price of advancement and economics. In this 34
chapter, we will discuss the types of samples and analytes that can be 35
analyzed by GC, the components of the GC and their operation, the variety 36
of detectors available today for use with GC, and examples of specific 37
analyses. 38
39
2.2 Types of Samples and Sample Introduction 40
2
1
A basic GC of reasonable quality costs from 30,000 to 50,000 US 2
dollars today depending on the detectors that are purchased with the GC, 3
although more inexpensive models can be purchased for limited routine 4
analysis. A capillary column gas chromatography-mass spectrometer (GC- 5
MS, quadrupole) will cost slightly less than $100,000. With this relatively 6
high price tag, students sometimes trust the results as unquestionably 7
accurate. Reality could not be farther from this belief. Every step, including 8
extraction of the analytes from the sample matrix, conducting serial 9
dilutions, injection into the GC, and identification of the fragmentation 10
pattern in MS is prone to errors. From experience, sample extraction can be 11
the most difficult and is the source of considerable error. Samples come in a 12
variety of forms: gaseous, liquid, solid, and biological. In order for the 13
reader to fully appreciate sample preparation for GC analysis the following 14
discussion will present several sample collection, extraction and sample 15
preparation techniques. 16
17
2.2.1 Gaseous Samples 18
Gaseous samples are the easiest samples to analyze. For on-site 19
analysis, a gaseous sample can simply be drawn into a syringe and the 20
sample injected into a sampling valve/loop. Sampling loops are necessary in 21
GC analysis in order to inject a consistent volume of a compressible sample. 22
When a gas sample is taken at atmospheric pressure and injected into a GC 23
inlet, the pressure in the GC will compress the gas in the syringe and not 24
allow all of the sample volume to be injected. A sampling valve and loop 25
consist of a four- or six-port valve that allows the sample to be injected into 26
a fixed-volume (loop of tubing that is at atmospheric pressure. A valve is 27
then turned that transfers all of the gas contained in the sample loop into the 28
GC injection port. For field gaseous samples that need to be transported to 29
the laboratory for analysis, a variety of sampling containers are available 30
including Teflon bags and metal cylinders (referred to as bombs) that can be 31
filled with the sample gas. It should be noted that when these containers are 32
analyzed that they be adjusted back to their field temperature in order to 33
avoid condensation of some gaseous analytes to liquids; this is especially 34
true when industry smoke stack or process gases are being sampled and 35
analyzed. Another possibility for sampling gaseous analytes is a resin tube. 36
To collect a sample a known volume of gas is passed through a glass or 37
metal tube containing a resin that has a strong affinity for the analytes. The 38
analytes adsorb to the resin and after a sufficient volume of gas has passed 39
through the system, each end of the resin tube is capped and transported 40
3
back to the lab. In the lab, the resin is extracted with a solvent specific to the 1
analysis and the solvent/analyte solution is injected into the GC. A 2
relatively simple calculation yields the concentration of each analyte in the 3
original gas volume. The obvious advantage of this method is concentration 4
of the gaseous analytes and the improvement of detection limits, as opposed 5
to analyzing the gas by direct injection. The resin tube method is commonly 6
used in the monitoring of solvents in the work place where an industry 7
worker will wear a portable personal pump that takes in atmospheric gases at 8
the same rate as a human would breath under working conditions. At the 9
end of the day, the tube is extracted and analyzed to determine if the worker 10
was exposed to chemicals in excess of workplace limits according to 11
Occupational Safety and Health Association standards. 12
13
2.2.2 Liquid Samples 14
Liquid samples are the next easiest to analyze by GC since they are 15
already in an injectable matrix. Samples from organic synthesis procedures 16
usually have products (analytes) present at high concentrations and are 17
analyzed by direct injection. Unfortunately, relatively few products fit the 18
requirements of GCthat analytes be volatile and thermally stableso most 19
products are analyzed by HPLC, the subject of Chapter 3. Analytes in 20
aqueous samples are also easy to analyze by GC. Some GC detectors and 21
columns allow the direct injection of aqueous samples if the concentration of 22
the analyte is sufficiently high. The aqueous sample is frequently extracted 23
into an organic solvent using a standard separatory funnel when there is a 24
low concentration of analyte is present in sample or when water could harm 25
the GC column or detector. Usually the aqueous sample is extracted three 26
times with a relatively small volume of organic solvent, the organic extracts 27
are combined, the volume is reduced by evaporation, and the resulting 28
organic extract is injected into the GC. One disadvantage of the organic 29
extraction is the need to purchase expensive and very pure organic solvents 30
(typically priced at approximately $150 for four liters) and the expensive 31
disposal costs of the resulting organic waste solvents. A more automated 32
version of the separatory funnel is a liquid-liquid extractor, but the glassware 33
is relatively expensive and typical extraction times run from 8 to 24 hours. 34
In this extraction setup, the organic solvent is boiled, condensed, and passed 35
through a water vessel where the non-volatile hydrophobic analytes partition 36
into the organic solvent that is constantly recycled into the boiling vessel. 37
The recycled solvent is re-evaporated again, leaving the analytes in the 38
boiling flask, and passes again into the condenser and water column. Figure 39
2.1 shows a typical liquid-liquid extraction system. 40
4
1
2
3
4
5
Figure 2.1 A Liquid-Liquid Extractor for Extracting Analytes from Water 6
Samples. Reprinted with permission from VWR Scientific Products 7
International, Chemglass Life Sciences, and Corning. 8
9
A relatively easy way to avoid the need for expensive glassware is to 10
use resin packs (SPE; solid phase extraction) that are available from a 11
variety of vendors. In this technique, the water sample is passed through a 12
resin pack (a solvent-resistant tube usually one to a few centimeters in 13
diameter and slightly taller in height). Again, the resin has a high affinity 14
for the analytes. After the passage of the aqueous sample through the 15
packet, the resin is dried by passing ultra-pure gas through it and the 16
adsorbed analytes are removed by passing a small volume of organic solvent 17
(usually a few mL) through the packet. The solvent volume is adjusted to a 18
known volume and injected into the GC. 19
5
1
2
3
Figure 2.2 Three Resin Packets for Extracting Analytes from Aqueous 4
Samples. 5
6
An even more novel way of extracting analytes from water samples is 7
to use Solid Phase Micro Extractors (SPMEs) that consist of a syringe 8
containing a fused silica capillary fiber coated with a chromatography 9
stationary phase with a high affinity for the analytes of interest. The fiber is 10
housed in a metal needle where it can be extended for collecting analytes or 11
for desorption in a GC injection inlet. The SPME needle and fiber are passed 12
through a septum in the sample bottle, either into the gaseous space above 13
the water or directly into the water, the fiber is exposed through the end of 14
the needle and allowed to equilibrate (adsorb the analytes) for typically 10 to 15
30 minutes while the sample is mixed with a stir bar. After this time most or 16
all of the analytes are transferred to the SPME fiber. The fiber is drawn into 17
the metal needle; the needle is withdrawn from the sample bottle and placed 18
directly into the GC injector. The advantages of this technique are (1) no 19
need for expensive organic extraction solvents, (2) relatively rapid analysis, 20
(3) possibly improved extraction recovery, and (4) significant concentration 21
of the analytes and improvement of detection limits (up to 10 000 to 1 000 22
000 fold concentrations). Fibers can be reused from 50 to 100 times. The 23
minor disadvantage of the SPME technique is the cost of the apparatus 24
(approximately $600 for three fibers and a holder/injector). 25
26
6
1
2
Figure 2.3 A SPME Device with the Microfiber Exposed (middle item). 3
Extra needles are available (top item) since the injector syringe can be 4
reused indefinitely. Needles can be reused from 50 to 100 times depending 5
on the composition of the sample. The bottom item is the needle protection 6
guard and GC injection guide. 7
8
Volatile analytes present an additional problem since considerable 9
quantities of the analyte can be lost during sample preparation. Analyses of 10
volatile analytes are best preformed with some type of commercial head- 11
space analyzer or purge and trap device where the actual water sample 12
(with no gaseous headspace) is attached to a sample processing unit, a gas is 13
used to transfer the volatile analytes to a resin trap or directly into the GC 14
injector, and after the required purge time the transferred analytes are 15
analyzed by GC. 16
17
2.2.3 Soil/Sediment Samples 18
Soil and sediment samples present considerable difficulties in sample 19
preparation since the analytes must be extracted and transferred to a liquid 20
phase before introduction into the GC. Early techniques focused on simply 21
washing the air-dried solid matrix with organic solvent but these methods 22
proved to yield low extraction efficiencies (analyte recoveries were 23
considerably less than 100 percent). The gold standard for the extraction of 24
analytes from soil and sediment matrices is the Soxhlet extraction technique. 25
The Soxhlet is a glass distillation setup that repeatedly passes pure solvent 26
through the soil/sediment matrix over a period of 24 to 48 hours. After this 27
time, the solvent is collected and the volume is reduced and analyzed by GC. 28
7
Laboratory studies have recovered approximately 100 percent of analytes 1
with this method but Soxhlet glassware is expensive (each setup costs at 2
least $300), it uses expensive organic solvents (approximately $150 per four 3
liters), and is very labor and time intensive. Alternatives to the Soxhlet 4
technique include relatively rapid sonication procedures and automated 5
heated solvent extraction systems. 6
7
8
9
10
Figure 2.4 Soxhlet Extraction Glassware. Reprinted with permission from 11
VWR Scientific Products International, Chemglass Life Sciences, and 12
Corning. 13
8
1
2.2.4 Biological Samples 2
Biological tissue samples are undoubtedly the most difficult to extract 3
and analyze. During the extraction process, the analytes need to be 4
effectively transferred from the outside and inside of cellular matter to the 5
solvent phase. The approaches used are as diverse as the high number of 6
sample tissue types. Common approaches include (1) drying the tissue, 7
followed by grinding, and Soxhlet extraction and (2) a combination of 8
grinding and sonication, followed by liquid extraction. Whichever method 9
is used, extensive sample cleanup (the removal of interfering substances and 10
analytes) is necessary since the analyst should not inject non-volatile 11
biological material into a GC. 12
13
An additional point should be made here. Gas chromatography is 14
only used for analytes with boiling points below approximately 300 C and 15
this limits the utility of GC analysis for both the organic and analytical 16
chemist (HPLC was developed for most other non-volatile compounds). 17
However, some analytes can be reacted with derivatizing agents to remove 18
functional groups that tend to make them nonvolatile. A common 19
derivatizing agent (also referred to as a silylating agent) is N,O-bis 20
(trimethylsily) acetamide which converts groups such as -OH, -COOH, - 21
NH
2
, =NH, and -SH to a -O-Si(CH
3
)
3
group that renders the compound 22
volatile. It should be noted that derivatizing agents are very hazardous and 23
usually carcinogenic. 24
25
2.2.5 Analyte Recovery 26
Now that we have presented some of the common extraction 27
techniques, another problem must be pointed out. How is it possible to 28
know all of the analytes were extracted from the sample (i.e. water, urine, 29
soil, fish)? This question becomes more difficult to answer as the sample 30
matrix becomes more complex. For example, how does the chemist 31
quantitatively recover all of the analyte from lake sediment or from food 32
items? These sample matrixes can have analytes contained within every 33
clay particle or biological cell and require the development and testing of 34
rigorous extraction procedures. Fortunately, many of these procedures have 35
been developed and are published by governmental agencies, industry, or 36
research scientists. As a result, incorporation of these procedures into the 37
laboratory is relatively easy. As an aid to determining how well your 38
extraction procedure works, relatively expensive reference samples that 39
contain a known amount of analytes can be obtained for a variety of sample 40
9
matrixes (i.e. fish, sediment, and manufactured goods). A procedure can be 1
validated if the results from your method are statistically equivalent to the 2
known concentration. For many procedures, it is not necessary to have a 3
high recovery (i.e. 98%) but it is necessary to have a known and consistent 4
recovery, even if it is low. 5
6
In addition to the potential human errors present in an analysis, 7
instrument detectors can also contain errors due to non-optimum 8
instrumental setting, out-of-date tuning or calibration, and when peaks elute 9
from the column with more than one analyte or in mass spectrometry when 10
more than one reference spectra is identified in the computer search library. 11
This latter situation is common with low concentrations of analytes. 12
13
Now that the basic problems and common errors associated with 14
sampling handling and instrumentation have been identified, we will move 15
on to distinctions between gas and liquid chromatography. Gas and liquid 16
chromatography were originally developed due to the existence of two basic 17
different types of analytes: (1) those that are thermally stable (do not 18
degrade at temperatures up to 300 C) and are volatile at relatively low 19
temperatures (below 300 C), and (2) for analytes that are not volatile and/or 20
thermally degrade at temperature above room temperature. GC is used for 21
thermally stable and volatile chemicals while HPLC is used for both non- 22
volatile compounds and ones that degrade at high temperatures. Recent 23
advances in the stationary phases on separation columns and mobile phase 24
selection (solvent gradient in HPLC) allow many analytes that were 25
exclusively analyzed by GC to be analyzed by HPLC. For example, GC was 26
the exclusive technique for analyzing mixtures of volatile organic solvents. 27
Yet today, by changing HPLC to a reverse phased system (where the 28
separation column is the nonpolar phase and the solvent is the more polar 29
phase) it can now analyze components of organic solvents. HPLC will be 30
discussed in depth in the next chapter. 31
32
GC analysis can also have special concerns. Impurities introduced 33
during sample preparation can result in contamination that may interfere 34
with the analysis of a desired analyte or introduce additional peaks into the 35
chromatogram (the output of a chromatograph). A notable case is a class of 36
compounds known as phthalates that are found in plastics that interfere with 37
the analysis of chlorinated pesticides such as DDT and PCBs in GC analysis 38
with an electron capture detector (ECD). Even with detection by mass 39
spectrometry, the analysis may conclude that these compounds were present 40
10
in the original samples when in fact it they are laboratory contaminants. As 1
a result, contact with plastics must be avoided regardless of the detector that 2
is used. It is also important to purchase GC grade solvents (at over $150 per 3
four liters) that are certified to contain an extremely low amount of 4
impurities when trace analyses are being conducted. 5
6
Some functional groups of analytes, such as in the analysis of 7
Bisphenol A, a known endocrine disruptor present in some plastic bottles, 8
may react with or irreversibly adsorb to the glass surfaces in the GC injector 9
liner and result in the analyst reporting the absence of Bisphenol A in a 10
sample when in fact it was present but lost during the analysis. This can be 11
overcome by deactivating the surfaces with a silanization agent that coats the 12
glass with a non-reactive trimethylsilane group, and allows the analyte(s) to 13
pass through the system to the detector. What and when to worry about 14
these problems, and many others, come with experience and knowledge of 15
the literature. 16
17
2.3 The Gas Chromatograph 18
19
The main purpose of chromatography is to separate a complex 20
mixture of compounds into discrete chromatographic peaks containing only 21
one analyte. Todays capillary column chromatographic systems are ideal 22
for this task and interface well with detection by mass spectrometry due to 23
the low volume of carrier gas used in capillary columns (1 to 5 mL/min as 24
opposed to 60-100 mL/min in packed column GC used prior to the 1980s). 25
Figure 2.5 below, illustrates the major components of a modern capillary 26
column gas chromatograph mass spectrometry (GC-MS) system. 27
28
11
1
Figure 2.5. A GC-MS System 2
3
2.3.1 Carrier Gases: The first important component is the carrier gas 4
or mobile phase. For a basic GC system, extremely pure helium is usually 5
used, and hydrogen is less frequently encountered due to its explosive 6
nature. Helium is used due to its inertness, non-reactive nature, and the 7
shape of its van Deemter curve that allows for a relatively wide range of 8
optimum mobile phase linear velocities. The common grade of helium used 9
is referred to as five-nine gas, meaning that it is 99.999% pure. But this 10
level of purity is still not sufficient for most systems when trace (parts per 11
million or parts per billion) analyses are being conducted. Before entering 12
the GC, the 2500 psi (18 000 kPa) pressure in the gas cylinder is reduced to 13
approximately 60 psi (400 kPa) with a two-stage regulator before entry into 14
the GC. But first, the He gas is passed through at least one resin trap to 15
further remove hydrocarbons, oxygen, trace analytes, and/or water vapor 16
12
that could interfere with analysis, degrade the column or interfere with the 1
detector. 2
3
2.3.2 Injectors: After passing through the purification traps, helium 4
enters the injector where it acts as the mobile phase and helps push the 5
analytes through the analytical (separation) column. A variety of injectors 6
are used in GC, but this text is concerned with the most common, a split- 7
splitless injector. This type of injector can be used in two modes. For 8
solutions containing extremely concentrated levels of analytes (in the parts 9
per thousand or percent level as is encountered in synthesis operations), the 10
injector is operated in the split mode. In this mode only a small fraction of 11
the 0.2 to 1 L of solution injected actually enters the separation column and 12
the majority of the sample is vented to the atmosphere. The high 13
concentration of analytes in the solvent allows for adequate identification 14
and quantification. For solutions containing lower levels of analytes (parts 15
per million and parts per billion), the injector is operated in a dual or 16
splitless-split mode. Upon injection of a sample, the injector is operated in a 17
splitless mode where all of the injected volume is being pushed onto the 18
column. But if this mode of operation is allowed to continue throughout the 19
chromatographic run, the peaks will be non-symmetrical (they will tail or be 20
skewed) which will interfere with peak integration because of a continual 21
addition of solvent molecules entering from the injection port. To avoid this 22
problem, the split mode is switched on approximately 30-60 seconds after 23
injection. This splitless-split mode allows the majority of the sample to 24
load onto the column while clearing out the remainder of the sample to 25
allow for a clean, well-shaped chromatographic peak. A typical split- 26
splitless is shown in Animation 2.1. 27
28
29
30
31
32
33
34
35
36
37
38
39
40
13
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Animation 2.1 Illustration of the Spiltless-Split Mode of Injection. 19
20
21
A few more points need to be made about the injector inlet. The 22
interface between the atmosphere and the injector is separated by a septum 23
such as the ones shown in Figure 2.6. Septa are manufactured out of various 24
materials, all of which must be inert to leaching organic constituents into the 25
GC or are coated on the GC side of the septum with Teflon. Septa are 26
inexpensive and are routinely replaced, usually after 30 to 50 injections. 27
28
14
1
2
Figure 2.6 GC Septa. Note the puncture holes in the top left septum. The 3
lower left septum contains a Teflon coating (yellow). The septa shown here 4
are about 1 cm in diameter. 5
6
Samples are injected through the septa and enter a glass liner in the 7
injection port. The purpose of the glass inserts (liners) is to avoid exposure 8
of the analytes to reactive hot metal surfaces such as those contained in an 9
unlined injector. Inserts come in a variety of forms. All liners contain a 10
hole in the top to allow entry of the injection needle, a wider middle space 11
for the expansion of liquid solvents into the vapor phase, and a hole in the 12
bottom for insertion of the capillary column. Glass wool is usually present 13
in the glass insert to keep pieces of the septum from blocking the inlet of the 14
capillary column and to trap non-volatile components of the sample. Figure 15
2.7 shows two common injector inserts, one with the glass wool in the 16
middle and one with the glass wool at the end. 17
18
15
1
2
Figure 2.7 Injection Glass Inserts. The insert on the right shows the o-ring 3
that seals the injection chamber and forces carrier gas through the column. 4
5
As noted earlier, samples are typically introduced into the GC with a 6
glass syringe with a metal needle. Samples can be injected manually or with 7
an automatic sampler. Standard 10-L syringes are shown in Figure 2.8. 8
9
10
11
Figure 2.8 GC Injection Syringes. 12
16
1
All connections in GCs, from the carrier gas cylinder to the detector 2
are made with Swedge Lock (a.k.a. Swagelok) fittings that seal the 3
connections at high gas pressure. These fittings consist of a threaded nut, 4
back ferrule, and front ferrule, all placed around a piece of tubing (refer to 5
Figure 2.9). Fittings come in Teflon, stainless steel, and copper and in a 6
variety of sizes ranging from smaller sizes for capillary columns as small as 7
0.2 mm to 6.0 mm packed columns. Ferrules are also available in graphite 8
and in a variety of advanced materials such as Vespel, a composite of 9
graphite and ceramic. A gas-tight fitting is achieved by tightening the nut 10
and compressing the ferrule around the metal tube. A selection of ferrules 11
and how they fit around a piece of tubing is shown in Figure 2.9. 12
13
14
15
Figure 2.9 Swedge Lock Fittings. 16
17
17
1
2
Figure 2.10 Several Types of Ferrules. 3
4
In addition to the split-splitless injector, other types of injector 5
systems are available including an on-column injector and a cryogenic 6
focusing injector. In the past, when packed columns were used, most 7
injections were made directly onto a section of the column that did not 8
contain any stationary phase resin. This concept has been extended to 9
capillary column technology by using a wide-bore column (typically 0.5 mm 10
in diameter or greater). A syringe with a very narrow capillary column 11
needle (0.2 mm in diameter) is placed through a special port at the head of 12
the injector and liquid samples are placed directly onto the column. The 13
needle is withdrawn, the injector port is sealed and the chromatographic run 14
is started. On-column injection avoids exposing the analytes to any reactive 15
surfaces. 16
17
Highly volatile analytes, normally not separated by standard GC 18
conditions can be analyzed using a cryogenic focusing injector. Here, gas or 19
liquid samples are injected through a septum, but the bottom of the injector 20
contains a cryogenic fluid (liquid N
2
) around the head of the capillary 21
column. The cryogenic fluid cools the column and causes the analytes to 22
condense at the head of the column. After injection, the cryogenic fluid is 23
removed, the column oven slowly heated, and the volatile analytes are 24
analyzed. 25
26
18
With todays extremely small injection volumes (0.2 to 1 L), 1
reproducibility of sample injection can be a problem. It is necessary to 2
inject exactly the same volume of sample (to within three significant figures) 3
to avoid introducing considerable error into the results. Two solutions have 4
been devised for this problem: (1) automatic samplers/injectors and (2) 5
internal standards. Mechanical automatic samplers can accurately and 6
consistently reproduce small volume injections and save considerable labor 7
costs (and time if you happen to be a graduate student or on a slim budget). 8
A typical automatic sampler today can hold up to 100 samples and be 9
programmed to run the samples in any order. This is more convenient than 10
manually injecting a sample and waiting for the GC run to end, which can 11
range anywhere from 5 minutes to hours. The automatic sampler allows the 12
user to simply return hours to days later to find the samples analyzed and the 13
data stored and ready for processing on the computer-controlled station. 14
Automatic samplers are common features today even in graduate schools 15
since they run hour after hour, day after day, with minimal oversight or 16
maintenance. The typical cost of an automatic sampler is only $10,000 and 17
they rapidly pay for their self in high sample volume work environments. 18
19
The second option for overcoming injection errors, an internal 20
standard, is also common in capillary GC and is usually used in conjunction 21
with automatic samplers. Internal standards were discussed in section 2.1 22
but will be repeated here due to its importance. An internal standard is a 23
chemical (analyte) that is not originally present in any sample. Equal 24
amounts of the internal standard are added to every sample and reference 25
standard. The computer program in the control station can then be 26
programmed to correct for injection errors. One way to use this technique is 27
to average the peak area for the internal standard that is measured in every 28
external standard and compare the area to that observed for each sample. If 29
the internal standard measured in a sample falls below or above this average 30
due to a poor injection, the computer will automatically adjust the areas of 31
every peak in the chromatogram accordingly. By using a combination of 32
automatic injectors and internal standards, highly accurate and consistent 33
results can be obtained. 34
35
2.3.3 Columns/Ovens: Separation columns are the heart of the GC 36
and are housed in a temperature controlled and temperature programmable 37
oven that can control temperatures to within 0.5 C. Considerable advances 38
were made in gas chromatography in the 1980s, especially with regard to 39
columns. Prior to the advent of capillary columns, chromatographic systems 40
19
used packed columns. Packed columns are 1/8 to 1/4 inch metal, Teflon, or 1
glass tubes filled with an inert resin coated with the stationary phase. Early 2
stationary phases were highly viscous, non-volatile liquids that interact with 3
the analytes to achieve separation or were molecular sieves that separated 4
the analytes by molecular size and diffusion. In the early 1980s, packed 5
columns were mostly replaced with capillary columns that are open tubular 6
columns (internal diameters in the tenths of millimeters) with the stationary 7
phase placed on the column walls. Initially, stationary phases were simply 8
applied to the walls as non-volatile liquids, however today most phases are 9
covalently bonded to the fused silica wall which yields more thermal 10
stability. Capillary columns have dramatically more theoretical plates than 11
packed columns and greatly improve resolution. The reader should recall 12
this relationship from Example 1.1 in section 1.2. For example, capillary 13
columns can have as many as 1000 times more theoretical plates as 14
compared to a packed column. Figure 2.11 shows several GC columns. 15
16
17
18
19
20
Figure 2.11 A 1/4-inch Glass Packed GC Column, a 1/8-inch Stainless Steel 21
GC, and a Fused Silica Capillary Column (from left to right). 22
23
20
1
2
Figure 2.12 Resin for GC Packed Column. 3
4
5
A summary of the most common stationary phases is given in Table 1.1. 6
The selection of a particular phase depends on the intermolecular 7
interactions expected for the analyte of interest. As discussed in Chapter 1, 8
the selection of the phase follows the adage like dissolves like or in this 9
case like stationary phases attach to like chemicals. The uses of each resin 10
are also shown in Table 1.1. Additional stationary phases are available for a 11
variety of analyte applications. Even some chiral compounds can be 12
separated with specialty stationary phases. 13
14
Table 1.1. Common Stationary Phases and Their Primary Use 15
STATIONARY PHASE APPLICATIONS
Polydimethyl siloxane
(Trade names are DB-1, HP-1, OV-
1, and SE-30)
This is a general purpose nonpolar
phase for separating hydrocarbons,
polynuclear aromatics, nonpolar
drugs, chlorinated pesticides, and
PCBs
Poly(phenylmethyldimethyl)
siloxane (5-10% phenyl)
(Trade names are DB-5, HP-5)
Still mostly nonpolar but with some
polarity. Used to separate fatty acid
methyl esters, alkaloids, drugs, and
halogenated chemicals
Poly(phenylmethyl) siloxane (50%
phenyl)
(Trade names are DB-15 and OV-17)
Slightly more polar. Used to separate
more polar drugs, pesticides, and
glycols
21
Poly(tritluoropropyldimethyl)
siloxane
(Trade names are DB-210 and OV-
210
More polar. Used to separate
chlorinated aromatics nitroaromatics,
and alkyl-substituted benzenes
Polyethylene glycol
(Trade names are DB-WAX and
Carbowax)
The glycol functional group makes
this phase considerably polar. Used
to separate free acid, alcohols, ethers,
essential oils, and glycols
Poly(dicyanoallyldimethyl) siloxane
(Trade names are DB-1701 and OV-
275)
The most polar phase shown here.
Used to separate polyunsaturated
fatty acids, free acids, and alcohols
1
Today most columns are fused silica capillary columns with typical 2
internal diameters ranging from 0.25 to 0.53 mm. Column lengths range 3
from 5 to 100 meters. As noted in Chapter 1, the longer the column, the 4
more theoretical plates it will contain and therefore long columns are 5
capable of separating almost any mixture of compounds. Film thicknesses 6
range from 0.25 to 3.00 m with thicker films usually providing more 7
resolution (and longer analysis times). Cross-linking of films is also 8
common and provides more thermal stability and less column bleed of the 9
stationary phase. Lower column bleed provides for a more stable detector 10
baseline and these columns are preferred in mass spectrometer applications. 11
12
2.3.4 Detectors: While the focus of this book is utilizing mass 13
spectrometry (Chapter 4) as the detector, it is informative to note that a 14
variety of detection systems are available for GC. The most common and 15
commercially available ones are listed in the table below with information 16
on their detection limits and analytes of interest. 17
18
Table 2.2 Commercially Available GC Detectors. 19
20
Detector General Type Analytes it is
used to measure
Typical
Detection
Limits
Flame Ionization
Detector (View the
FID Animation
below)
Selective Any chemical
that will burn in
a H
2
/O
2
flame
parts per million
Thermoconductivity Universal Any chemical parts per
22
Detector (View the
Thermocon-
ductivity Animation
below)
with a thermal
conductivity
(~specific heat)
different from
He
thousand or
hundred
Electron Capture
(View the ECD
Animation below)
Selective Electrophores
such as
halogenated
hydrocarbons
parts per billion
or less
Flame Photometric Specific P and S
containing
compounds
parts per million
or less
Fourier Transform
Infra-Red
Specific Chemicals with
specific
molecular
vibrations
parts per
thousand or
hundred
Mass Spectrometry Universal Any chemical
species
parts per million
or less
1
Three of the most common GC detectors will be discussed in detail 2
here, while types of mass spectrometers will be presented in Chapter 4. One 3
of the earliest GC detectors was the thermal conductivity detector (TCD). 4
The basis for this detector is that most analytes have a thermal conductivity 5
lower then that of helium. Helium is used as the carrier gas and as it passes 6
a wire with a current applied to it the wire heats up via electrical resistance. 7
Helium molecules remove the maximum amount of heat and the wire 8
reaches thermal equilibrium and a constant current reading. As an analyte 9
with a different thermal conductivity enters the detector, the wire heats up 10
with increasing electrical resistance and the measured current decreases. 11
Since the analytes pass through the detector as a chromatographic plug a 12
bell-shaped current reading results known as a chromatographic peak. After 13
the analyte has passed through the detector, the current returns to the original 14
baseline reading as the helium re-cools the wire. Usually two matched 15
columns and detectors are used, where only He is passed through one setup 16
and samples are injected into the other. While this detector responds to any 17
chemical with a thermal conductivity different than helium (which includes 18
almost every other compound), these detectors suffer from relatively poor 19
detection limits (parts per thousand to parts per hundred). Animation 2.2 20
illustrates the operation of a TCD. 21
22
23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Animation 2.2 Illustration of a Thermal Conductivity Detector. 23
24
The most common detector in gas chromatography is the Flame 25
Ionization Detector (FID). This detector is based on the fact that most 26
chemicals will burn in an H
2
-air flame and current can be passed through the 27
path of ions produced in the flame. Helium is again used as the carrier gas 28
and analytes are injected in the standard split-splitless injector. As 29
individual packets of analytes are separated in the column and enter the 30
detector they burn in the flame. As illustrated below in Animation 2.3, a 31
potential is placed across the flame jet and an electron collector plate is 32
placed above the flame. As ions are produced, electrons are passed through 33
the ion cloud, and a current is measured that is proportional to the mass of 34
analytes produced in the flame. The FID is also considered a universal 35
detector, although not every chemical will burn in an H
2
-air flame. FID are 36
relatively sensitive with detection limits of 1 ppm for most chemicals. 37
38
39
40
24
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Animation 2.3 Illustration of a Flame Ionization Detector. 21
22
The electron capture detector (ECD) is perhaps the most sensitive 23
detector for a GC and was developed primarily to detect chlorinated 24
hydrocarbons in the environment. It relies on the electrophilic nature of 25
halogens contained in an organic chemical, but can be used to detect other 26
electrophilic-contained elements such as oxygen. The detector is a sealed 27
unit and contains a radioactive isotope of nickel,
63
Ni. This isotope gives off 28
a steady supply of beta particles that are essentially high speed electrons. 29
These high-speed electrons collide with trace amounts of methane carrier 30
gas that enters the column after the column effluent and produces slower 31
speed electrons (thermal electrons). These thermal electrons are captured by 32
the anode in the middle of the detector and provide a constant current in the 33
absence of any electrophilic analytes. As electrophilic analytes enter the 34
detector they attract the thermal electrons and carry them out of the detector. 35
This removal process results in a lowering of the current measured by the 36
detector and the change in current is measured as an inverse 37
chromatographic peak. ECDs are extremely sensitive and yield detection 38
limits of pg or sub-parts per billion concentrations in the injection solvent. 39
The operation of an ECD is illustrated in Animation 2.4. 40
25
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Animation 2.4 Illustration of an Electron Capture Detector. 23
24
2.4 Advanced GC Systems 25
26
Most commercial GC systems come with ports for at least two 27
injectors, and therefore two columns and two detectors. This allows for 28
versatility and some inventive operational designs. With a special ferrule, 29
two columns with different stationary phases can be inserted into one 30
injection port allowing two analyses per injection and confirmatory analysis. 31
32
Confirmatory analysis, with respect to chromatography, is usually 33
restricted to mass spectrometry detection, but when a sample is analyzed on 34
two columns with different stationary phases, the likelihood of two different 35
compounds yielding the same retention time on both columns and the same 36
response on identical detectors is highly unlikely. Thus, dual column 37
analysis usually produces confirmatory identification. 38
39
26
More enhanced arrangements include detectors aligned in series 1
where the effluent of one detector is passed into another detector. This 2
arrangement does not necessarily provide confirmatory analysis, but does 3
allow considerably more information to be collected about the analyte. Note 4
that the first detector cannot be a destructive detector since the chemical 5
integrity of the analyte must be intact for the operation of the second 6
detector. 7
8
2.5 Applications 9
10
An almost endless variety of chromatographic separations are 11
achievable today due to the diversity of analytical columns. Major column 12
manufacturers and distributors provide very useful Internet sites that contain 13
chromatograms for common analytes that can be used to help select the 14
appropriate column for your needs. In addition, technical help is available 15
from professional chromatographers at these companies for more complex 16
separations. The chromatograms below were selected from the hundreds 17
available from Agilent technologies at http://www.chem.agilent.com/en- 18
us/Search/Library/Pages/ChromatogramSearch.aspx. There many excellent 19
additional examples from chemical and chromatography suppliers. 20
21
One of the most common published lists of GC applications is for the 22
analysis of environmental pollutants. The chromatographs below are for a 23
variety of chemicals, analytical columns, and detectors. As you review these 24
note the correlation between the intermolecular forces available to the 25
analytes and the stationary phases used to separate them. Also note the GC 26
detectors used for each type of analyte. 27
28
27
1
2
Figure 2.13 Analysis of Chlorinated Pesticides by GC-ECD using a Ultra 2 3
Column (5 percent cross-linked phenyl methyl silicone). Source: Copyright 4
2006 Agilent Technologies, Inc. Reproduced with Permission. 5
6
7
28
1
2
Figure 2.14 Analysis of Semivolatiles by GC-MS using HP-5 column. Note 3
the large number compounds that can be separated in one GC analysis. 4
Source: Copyright 2006 Agilent Technologies, Inc. Reproduced with 5
Permission. 6
7
The pharmaceutical industry also heavily uses GC and HPLC to 8
determine the purity of reagents, the identity of synthesis products, and the 9
identity of medicines and illicit drugs. A few examples are shown below. 10
11
29
1
2
Figure 2.15 Analysis of Anticonvulsants by GC-FID using an HP-1 column. 3
Source: Copyright 2006 Agilent Technologies, Inc. Reproduced with 4
Permission. 5
6
30
1
2
Figure 2.16 Analysis of Alkaloids and Barbiturates by GC-FID with an 3
Ultra 2 column. Source: Copyright 2006 Agilent Technologies, Inc. 4
Reproduced with Permission. 5
6
GC can also be used to determine the identity of natural products 7
containing complex mixtures of similar compounds. For example, the 8
geographic source of crude oil or natural gas can be determined by the 9
fingerprint, or relative distribution of major and trace compounds in each 10
oil. Natural produce oils, such as food products or fragrances, can be 11
31
identified by GC-FID or GC-MS. A few examples of the separation of these 1
complex mixtures are shown below. 2
3
4
5
Figure 2.17 Analysis of Natural Gas by GC-TCD using a HP-PLOT Q 6
column. Source: Copyright 2006 Agilent Technologies, Inc. Reproduced 7
with Permission. 8
9
32
1
2
3
Figure 2.18 Analysis of Peppermint Oil by GC-FID using an HP-INNOWax 4
column. Source: Copyright 2006 Agilent Technologies, Inc. Reproduced 5
with Permission. 6
7
8
33
The purity of solutions, from relatively pure solvent such as xylene, to 1
liquors such as scotch can be also be determined by GC. Two examples are 2
shown below. 3
4
5
6
Figure 2.19 Analysis of p-Xylene by GC-FID using a HP-INNOWax 7
column. Source: Copyright 2006 Agilent Technologies, Inc. Reproduced 8
with Permission. 9
10
34
1
2
3
Figure 2.20 Analysis of Minor Components of Scotch by GC-FID using a 4
HP-101 Column. Source: Copyright 2006 Agilent Technologies, Inc. 5
Reproduced with Permission. 6
7
2.6 Summary 8
9
This chapter focused on the use of gas chromatography. A variety of 10
separation columns and detectors allow the analysis of a diverse set of 11
35
chemical structures and the separation of complex mixtures of chemicals. 1
Advances in technology have increased the utility of GC analysis and 2
automated instrument controls have greatly decreased the cost of analysis by 3
decreasing labor costs. GC allows the relatively rapid analysis of analytes 4
present in high concentrations, such as in product quality assurance/quality 5
control and in product identification, as well as in the analysis of trace 6
analysis such as the identity of pollutants in environmental media and 7
confirmation of medicines or illicit drugs in human samples. As noted 8
several times in this chapter, GC is used to analyte volatile, thermally stable 9
compounds, and in general cannot be used to analyze for the more extensive 10
and diverse compounds found in biological systems. These compounds are 11
typically analyzed for by HPLC, the subject of the next chapter. Later, in 12
Chapter 5, GC, HPLC and CE will be coupled with MS, the ultimate 13
detector since it allows immediate confirmatory identification. Identification 14
by fragmentation pattern in GC analysis is the subject of Chapter 6. 15
16
2.7 Questions 17
18
1. In approximately what decade was chromatography first discovered? 19
20
2. What is the purchased price of a basic GC? 21
22
3. What are the four basic types of sample matrices? 23
24
4. List and describe the common ways gaseous samples are collected for GC 25
analysis. 26
27
5. List and describe the common ways liquid (aqueous) samples are prepared 28
for GC analysis. 29
30
6. List and describe the common ways soil and sediments samples are extracted 31
and prepared for GC analysis. 32
33
7. How are biological samples prepared for GC analysis? 34
35
8. Why are analysts concerned with extraction efficiency when they prepare 36
samples for chromatographic analysis? 37
38
9. What chemical characteristics must a chemical have in order to be analyzed 39
by GC? 40
41
10. How are derivatizing agents used in GC analysis? 42
43
36
11. Draw a diagram showing all of the components of a basic GC. 1
2
12. How are the mobile phase flow rates different between packed column and 3
capillary column GC? 4
5
13. What is the most common carrier gas used in capillary column GC? 6
7
14. Draw and explain how a split-splitless injector works. Why do we use a 8
combination of splitless and split injection? When would you advantageous or 9
necessary to use a total splitless injection? 10
11
15. Why is it important to use a Teflon coated septum in some GC analyses? 12
13
16. What is the purpose of the glass liner in the GC injector? 14
15
17. What are the typical sizes (diameters and column lengths) of fused silica 16
capillary columns? 17
18
18. What are typical injection volumes for capillary column analysis? 19
20
19. Contrast the cost of an automatic sampler with the advantages of using one. 21
22
20. List the six common types of stationary phases used in GC and describe 23
what types of analytes they can be used to analyze. 24
25
21. What are the advantages (and disadvantage) of cross linking the stationary 26
phase coating on a capillary column. 27
28
22. List five common GC detectors, give their acronym, list the types of 29
chemicals they are commonly used to detect, and their detection limits. 30
31
23. Explain, with drawings, how a thermo-conductivity detector (TCD) works. 32
33
24. Explain, with drawings, how a flame ionization detector (FID) works. 34
35
25. Explain, with drawings, how an electron capture detector (ECD) works. 36
37
26. What is meant by confirmatory analysis in chromatography. 38
39
27. For each of the chromatograms shown in Section 2.5, identify the 40
intermolecular force involved in the separation of each class of analytes. 41
42
28. Select a compound that can be analyzed by GC (relatively volatile and 43
thermally stable) and use the Internet to find what GC column and temperature 44
program is used in its analysis. 45
46
37
29. Say that you are analyzing a mixture of compounds by GC and that you are 1
having trouble achieving separation of some of them (they co-elute or appear as 2
a shoulder peak). What three major things can you change about your GC to 3
possible improve separation? 4
5
2.8 References: 6
7
Several instrument and column manufacturers were consulted while 8
researching this chapter. Manufacturers have excellent web sites for 9
researching their products, talented sales staff, and very helpful technical 10
assistance. 11
12
Bartle K.D. and P. Myers. History of gas chromatography, Trends in 13
Analytical Chemistry, Volume 21, Number 9, 10 September 2002, pp. 547- 14
557 15
16
17
18
19
"
Chapter 3 "
Basic High Performance Liquid Chromatography #
$
3.1 Introduction and History %
In Chapter 2, Michael Tswett (1872-1920) was credited as the father of &
chromatography due to his 1903 separation of the green-leaf pigments into '
bands of colors, a demonstration of liquid chromatography. While similar work (
was being conducted in the petroleum industry, Tswett is credited with coining )
the term chromatography. Despite Tswetts results, chromatography did not *
develop quickly. The next major developments were the use of thin-layer "+
chromatography (TLC) in 1937-38 and the use of paper chromatography in the ""
mid-1940s, but thin layer chromatography quickly won popularity. Thin layer "#
chromatography was originally developed by Nikolai Izmailov (1907-1961) and "$
his graduate student Maria Shraiber (1904-1992) for pharmaceutical "%
preparations. Early TLC was conducted with microscope slides that were coated "&
with suspensions of calcium, magnesium, and aluminum oxides. As used today, "'
a small spot of solution was placed on one end of the slide, the slide was dipped "(
into a solvent, and the analytes migrated at different rates through the oxide ")
coatings where they were later detected (today by a UV lamp or chemical stain). "*
TLC advanced slowly during the next few years but a major advancement #+
was made in 1956 by Egon Stahl (1924-1986) when he attempted to standardize #"
the preparation of the sorbents used to make the plates. While these advances ##
and others such as forced flow TLC, significantly matured TLC into an accepted #$
(and reproducible) practice; it was still only a qualitative technique, at best. #%
However, Izmailov and Shraibers spot chromatography, commonly known today #&
as TLC, is the workhorse of undergraduate organic synthesis labs where #'
synthesis reactions are conducted and the resulting products are selected for #(
using common glass open columns filled with silica gel (refer to Figure 3.1). #)
#
Eluent from these columns is collected in fractions that are then run by TLC to "
identify which column fraction contains the desired product. #
$
Figure 3.1 An Atmospheric Pressure Open Column Chromatographic Column. %
Fractions are collected in the test tubes and later ran by TLC to determine the &
purity of the fraction and the presence of synthesis products. '
$
Since the development of TLC, liquid chromatography needed a technique "
comparable to gas chromatography where a complex mixture of analytes could #
be quantitatively separated and identified. Several attempts were made to $
pressurize the relatively large glass preparatory column (shown above) with little %
success due to the fragile nature of the column. High-pressure liquid &
chromatography (HPLC) was later developed to meet this goal in the 1970s. The '
pressure was first delivered by a large syringe, but this approach limited the (
volume of solvent that would be passed through a column and therefore limited )
the analysis time. Syringes were later replaced by a single reciprocating pump *
but these delivery systems experienced flow surges, between strokes of the "+
single piston, interfering with stable detector baselines. The placement of two ""
reciprocating pump, operating opposite to each other with respect to flow, greatly "#
minimized the flow fluctuations which were later removed completely with a pulse "$
damper. This form of chromatography is referred to as HPLC, where the HP "%
stands for high performance or high pressure. Some jokingly refer to the HP as "&
meaning high priced since it replaced TLC plates, that only cost pennies, with "'
$20 000 to $30 000 instruments. The inflation of the cost of a LC analysis is "(
even greater when an HPLC-MS is considered, a minimum of $150 000. But ")
regardless, HPLC-MS is considered the technique of choice for isolating a "*
synthetic product and is widely utilized in most synthesis laboratories. #+
Chromatography, as noted in Chapter 1, is divided into gas, liquid and #"
supercritical fluid techniques. Liquid chromatography can be divided up into a ##
relatively large collection of techniques. Those mentioned above include #$
atmospheric or low-pressure open column chromatography and thin layer #%
chromatography. Pressurized liquid chromatography can be divided into ion #&
exchange, exclusion, partition, and liquid-solid chromatography as summarized #'
in Animation 3.1. View this animation by double clicking on the figure below. #(
#)
#*
%
#
$
Animation 3.1. A Discussion of the Various Types of Chromatography. %
&
As noted in Animation 3.1, a variety of separation techniques are available '
with high pressure liquid chromatography. Most relevant to this chapter is (
partition chromatography, although a few others will be discussed in later )
sections. The most important point here is to distinguish between normal and *
reverse phase chromatography. HPLC was first developed using normal phase "+
conditions (NP-HPLC), that followed the logic of atmospheric open-column ""
chromatography, where the stationary phase acted as the polar phase and the "#
mobile phase was non-polar, specifically an organic solvent. Normal phase "$
HPLC focused on the separation of analytes that were readily soluble in non- "%
&
polar solvents but had slight affinities for the polar stationary phase. However, "
NP-HPLC required the use of relatively expensive and large volumes of organic #
solvents that also led to high waste disposal costs. NP-HPLC was effectively $
replaced by reverse phase HPLC (RP-HPLC) that operates with a non-polar %
stationary phase and an aqueous, moderately polar mobile phase. Gradient &
programming, an additional development, changes the composition of the mobile '
phase during a chromatographic run which greatly enhanced the utility of RP- (
HPLC. In RP-HPLC, the gradient is initially more polar (i.e. water) and as the )
chromatographic run progresses, more and more less-polar solvent is added to *
the mixture (i.e. methanol or acetonitrile) to end the gradient program with pure "+
organic solvent. As noted in Animation 3.2, the retention order of analytes in RP- ""
HPLC is opposite NP-HPLC. For example, the first analyte that eludes on a RP "#
system would elude last is separated by NP-HPLC. "$
"%
"&
"'
"(
")
"*
#+
#"
##
#$
#%
#&
'
#
$
Animation 3.2 Demonstration of Normal and Reverse Phase HPLC. %
Additionally, the pH of the polar solvent in RP-HPLC can play an important &
role in optimizing analyte separations. Ionic or ionizable analytes that would not '
normally be separated on a RP (non-polar) analytical column can also be (
analyzed by ion-pair chromatography. In this type of chromatography, the ionic )
analyte is bound to another ion (usually a large organic counter-ion such as *
quaternary ammonium or alkyl sulfonate) to form a neutral pair that has selective "+
affinity for the non-polar stationary phase. Even many chiral compounds can be ""
separated on special chiral stationary phases or by the additional of chiral "#
resolving agents that selectively bind to one of the enantiomers. Additional "$
modifications to an HPLC system such as ultra-high pressure LC, ion exchange, "%
and supercritical fluid chromatography will be discussed in Sections 3.5 and 3.6. "&
(
One final point should be made with respect to HPLC. Chemists use "
HPLC for two completely different purposes. Organic chemists, especially in the #
pharmaceutical industry, use large-scale systems, referred to as preparatory $
HPLC or flash chromatography, to recover relatively large-scale milligram %
quantities of their products. The main differences of a preparatory HPLC, as &
opposed to an analytical HPLC, are the pump flow rates and the size of the '
columns. In contrast, analytical chemists use HPLC to separate and identify (
nanogram or smaller quantities of analytes. Which ever practice is needed, the )
overall chromatography is the same. *
3.2 Types of analytes, samples and sample introduction "+
Gas chromatography is somewhat limited in that the analyte has to be ""
volatile below 300 C and thermally stable with molecular weights less than 1000 "#
Daltons. HPLC greatly expands the range of possible analytes and with reverse "$
phase HPLC and can include many of the same analytes as GC (but with less "%
resolution as compared to capillary column GC). Analysis of compounds "&
commonly used in HPLC increases the analyte range of molecular weights to just "'
less than one million Daltons. Many chemicals, such as petroleum "(
hydrocarbons, solvents, illicit drugs and environmental chemicals, can be ")
analyzed by both GC and HPLC. But with the ability of HPLC to analyze non- "*
volatile chemicals many bio-molecules can be analyzed, including sugars, amino #+
acids, proteins, and a large variety of other non-volatile compounds. #"
Samples analyzed by HPLC are always in liquid form, in either aqueous or ##
organic solvents. Little to no sample preparation is needed, except that it is good #$
practice to filter all samples through a 0.2 m low-volume cartridge filter prior or #%
during injection. Filtration of samples and solvents avoids the buildup and #&
eventual clogging of the in-line filter or guard column. Samples containing high #'
concentrations of analytes may need to be diluted in order to avoid overwhelming #(
the capacity of the stationary phase and remaining in the linear range of the #)
detector (governed by Beers law in UV-Vis applications). Solvent exchange of #*
the samples may be necessary depending on the solvent gradient conditions $+
)
needed for separation. Samples containing compounds with significantly "
different chemical structures may interfere with the detector of the analyte(s) and #
may require removal with a micro-column clean up (such as silica gel or alumina) $
or may require the use of a specialty clean up cartridge (such as those used in %
GC for concentration of the analytes from aqueous samples described in Section &
2.2). Samples containing relatively low concentrations of analytes may need to '
be concentrated using one of these same micro-resin columns. (
3.3 The Liquid Chromatograph )
3.3.1 Overview: In recent decades the modern HPLC system has *
become increasing complex. Automatic injection systems have all but replaced "+
manual injections, manual six-port values are now pressure actuated, the ""
instrument is usually computer controlled, and data are collected and processed "#
via a specified computer method. It should come as no surprise that the cost of a "$
basic HPLC system (without mass spectrometry detection) has risen from $10 "%
000 a few decades ago to nearly $40 000 today. In this section, each component "&
of an HPLC will be presented. Figure 3.2 illustrates an overview of a modern "'
HPLC-MS system. "(
*
"
Figure 3.2 An Overview of a Modern HPLC-MS System. #
3.3.2 Actuator Gas: Pressurized gas is required to degas solvents and to $
drive many of the mechanical functions of modern HPLCs. Solvents used as the %
mobile phase contain trace concentrations (ppm) of atmospheric gases (N
2
and &
O
2
) which create bubbles in the pump (during the low pressure intake of '
solvents) causing pumping problems. Dissolved gases can also evolve in the low (
pressure detectors in ultrahigh pressure LC, ion exchange, and supercritical fluid )
chromatography giving the appearance of a chromatographic peak depending on *
the specific detector being used. Pressurized gas is also needed to turn the six- "+
port injection valve during automatic injection mode. ""
3.3.3 Solvents and Solvent Preparation: Solvent purity requirements "#
depend on the type of samples being analyzed. All solvents used as the mobile "$
"+
phase must be filtered through 0.2 m filters prior to entry to an HPLC system in "
order to avoid the scratching of the pump pistons by particles and to avoid #
clogging the in-line filter or guard column. Organic chemists are usually not $
concerned with the purity of solvents since their analytes are present in high %
concentrations and their samples usually contains many side-products and &
solvents. Hence, relatively low purity, and therefore inexpensive, solvents can be '
used. In contrast, the analytical chemist must purchase HPLC grade solvents (
that are considerable more expensive. For example a 20-L metal can of ACS )
grade methanol costs less than $70, while a 4-L glass bottle of HPLC UV-Vis *
grade bottle costs $150. Many solvent gradients call for acetonitrile that in recent "+
years (2009-2009) has risen considerable in costs due to a decrease in ""
production, from $100 in 2005 to $250 in 2009 (if supplies allow you to even "#
order it). The increase in cost has tempted many chromatographers to switch "$
their acetonitrile solvent gradients to one based on methanol while maintaining "%
the same polarity index (a measure of the polarity of the solvent mixture that is "&
used to optimize analyte separation). However, as noted by William Campbell in "'
a recent Supelco Analytical Note (volume 27.1, page 13), while analyte retention "(
orders may (or may not) remain the same, the relative retention times and ")
resolution can significantly change, in some cases where analyte separations are "*
no longer achievable. Solvents used for elution are contained in glass reservoirs #+
(usually one liter in capacity) and connected to HPLC with Teflon tubing. The #"
solvent bottle is purged with He gas to remove dissolved gases, mentioned ##
above, and kept under a He atmosphere during HPLC operation. #$
Tubing and Connections: Different components of an HPLC system #%
operate under different pressures ranging from atmospheric to 4 500 psi (3.10 x #&
10
7
Pa) in standard HPLC and to 20 000 psi (1.38 x 10
8
Pa) in ultra high pressure #'
liquid chromatography. Systems are currently under development that will #(
operate at 100 000 psi. While GC fittings are usually standardized to Swagelok #)
fittings, many HPLC manufactures have established their own type of high #*
pressure fittings. All operate on the same ferrule-nut configuration as in GC but $+
the shape of the ferrule and receiving system is different. A few of the more $"
""
common fitting are shown in Figure 3.3. Tubing sizes are much smaller in HPLC "
as compared to GC. Common tubing materials are stainless steel and PEEK #
(polyaryletheretherketone), a very strong organic polymer. $
%
Figure 3.3 A Selection of HPLC Tubing and Connectors. &
'
3.3.4 Proportioning Valve: Up to four separate elution solvents enter the (
HPLC first through a proportioning valve that adjusts the flow of each solvent to a )
predetermined amount. A proportioning valve is shown in Figure 3.4. Prior to *
the development of the proportioning valve, a separate pumping system was "+
needed for each solvent. Since the pumping system (Figure 3.4) is typically the ""
most expensive component of an HPLC system, the proportioning valves help "#
reduce the cost of an HPLC system. "$
"%
"#
"
Figure 3.4 A Dual Piston Reciprocating Pump (left) and a Four-Way #
Proportioning Valve (right). $
The relatively inexpensive proportioning valve allows the mixing of up to %
four solvents which enter a single pumping system. The mobile phase can be &
delivered as a constant composition of solvents (isocratic) or as a changing '
composition (gradient programming). In gradient programming, the solvent (
composition changes from a more polar solvent mixture to a more non-polar )
organic composition. Isocratic operations are used for relatively simple *
separations, while gradient programming is used for complex separations. In the "+
past, the relative flow rates of multiple pumps were used to control the mobile ""
phase composition. Today the less expensive proportioning valves adjust the "#
composition. The need for gradient programming is illustrated in Animation 3.2 "$
and is similar to temperature programming in GC and is again referred to as the "%
general elution problem. Click on the figure below to start the animation. "&
"$
"
Animation 3.2 The Solution to the General Elution Problem in HPLC. #
3.3.5 Pump: As noted earlier, HPLC pumping systems have evolved $
more than any other component of LC. Gravity-fed open column systems were %
later pressurized with gas, which was replaced by syringe pumps in early HPLC, &
then by single piston pumps, and today with specially engineered dual '
reciprocating piston pumps. Todays pumps provide constant pressure and (
mobile phase flow by alternating the pumping actions between two pumping )
systems. When either pump is in full stroke, constant flow is easily provided but *
as each pump reaches the end of a stroke flow rates may be altered. This is "+
overcome in some systems by an oval-shaped caming device that speeds the ""
pumping rate at the end of each piston stroke. At the end of the stroke, one "#
pump speeds up the flow of mobile phase as the other pump head decreases the "$
pumping rate. This combined action provides constant flow rates in modern "%
HPLC systems. While the simultaneous operation of two pumps and their "&
associated pistons, cams, and check valves are difficult to describe in words, an "'
"%
animation makes the process easy to understand. Such an animation is shown "
in Animation 3.3. The animation can be viewed by clicking on the figure below. #
$
Animation 3.3 Illustration of a dual reciprocating piston HPLC pump. %
&
Pulse Damper: Early HPLC pumps produced pressure, and therefore flow '
pulses, this was especially true with single piston pumps which required a pulse (
damper to be installed down-line from the pump. Todays systems, with )
advanced pump design, produce little pressure fluctuations and either do not *
require a pulse damper or electronically compensate for the very small pressure "+
fluctuations. ""
3.3.6 Six-Port Injector: High pressure systems require a special type of "#
sample injection since injections with standard syringes are not possible. Four-, "$
"&
six-, and eight-port valves are used for high pressure systems. These systems "
are equipped with a fixed volume loop of tubing that serves as the sample loop #
that is loaded with a standard blunt-tipped syringe. These multiple port systems $
allow un-interrupted flow to the column during the loading of a sample on a %
sample loop and during injection of a sample under high pressure. As with a &
dual reciprocating HPLC pump, it is easier to show how a six-port valves works '
through an animation. Animation 3.4 shows the alignment of the valves during (
the loading of a sample onto a sample loop, switching of the values, and injection )
of the contents of the sample loop onto an HPLC column. View the animation by *
clicking on the figure below. "+
"#
Animation 3.4 Operation of a six-port sampling/injection valve for an HPLC. "$
"%
The sampling loop of a six-port valve can be loaded by a syringe or by "&
pumping liquid sample into the valve by an automatic sampler. Sample loop "'
"'
sizes range from 5 - 100 L for analytical-scale HPLC to milliliter volumes in "
preparatory-scale HPLC. The advantage of a syringe (manual) injection is that #
less sample volume is required (since filling of the tubing in an automatic sampler $
is not required). The obvious advantage of an automatic sampler is that %
numerous samples can be automatically analyzed by a computer-controlled &
system. '
In-Line Filter: Although samples and solvents are filtered prior to injection (
or use with an HPLC, an in-line filter is usually present immediately after the )
injection valve. This filter removes any remaining particles in the mobile phase *
that may clog the guard or analytical column. In-line filters are one of the most "+
commonly maintained items for HPLC systems. ""
3.3.7 Columns: "#
Guard Column: The next component of an HPLC is the guard column. "$
Guard columns are miniature versions of the analytical (separation) column and "%
they contain the same stationary phase. The purpose of the guard column is to "&
adsorb any permanently adsorbing chemicals that could destroy the more "'
expensive analytical column. Guard columns typically cost one-fourth or less of "(
the cost of an analytical column. ")
Analytical Column: The heart of an HPLC system is the analytical column. "*
This is where the mixture of chemicals injected in the system is separated into #+
individual analytes that appear as peaks in the chromatogram. Columns are #"
available in a variety of diameters and lengths ranging from large preparatory ##
columns (20 - 50 mm in diameter by 50 - 250 mm in length), to analytical #$
columns (typically 4.5 mm in diameter by 12 - 25 mm in length), to narrow bore #%
analytical columns for improved performance and MS applications (1-2 mm in #&
diameter by 10 cm in length), to capillary columns for MS detectors (from 0.075 - #'
0.1 mm in diameter). Larger columns require a larger mobile phase volume to #(
push the analytes through the system. A collection of HPLC columns is shown in #)
Figure 3.5. #*
"(
"
Figure 3.5 Various Sizes of HPLC Analytical and Quard Columns. Source: #
Analytical Sales and Services, Inc. Reprinted with permission from Analytical $
Sales and Services. %
&
With the exception of relatively new capillary columns, all HPLC columns '
are packed with small resin beads (typically 2-5 m in diameter or less) that (
contain a coating of stationary phase. As in GC, stationary phases are selected )
based on the expected intermolecular forces available to the analyte for bonding. *
The adage like dissolves like is also appropriate for liquid chromatography "+
applications. Separations in HPLC are considerable more complicated then in ""
GC. In GC, the chromatographer is only concerned with the stationary phase "#
and the temperature program. In HPLC, one must also be concerned with the "$
polarity of the mobile phase. Adequate analyte separations are only achieved "%
with an appropriate match of gradient polarity and stationary phase interactions. "&
In more advanced applications, the pH and ionic composition of the mobile phase "'
")
must also be controlled. The most common stationary phases used in reverse- "
phase HPLC include alkylamine, octodecyl (C18), octyl (C8), butyl (C4), #
cyanopropyl (CN), and methyl functional groups. Additional and custom-made $
stationary phases can be ordered from many suppliers. %
Early chromatography columns that were open to the atmosphere were &
easy to pack. The chromatographer filled a glass column (1 to 5 cm in diameter) '
with solvent, slowly poured in silica gel with stirring to remove air pockets, and (
applied the solution to be separated to the top of the column. These types of )
columns yielded adequate separation for organic chemists attempting to isolate *
their synthesis products but have extremely limited use for analytical chemists "+
who require more theoretical plates in the column and on-line detection of ""
column effluents. Today, analytical chromatographers almost exclusively "#
purchase their separation columns from manufactures because of the need for "$
perfectly packed columns with no void spaces. The emphasis here is on perfect "%
column packing since the presence of only a few void areas in a column will "&
significantly affect the theoretical plate height in the column and decrease "'
resolution. A comparison of a poorly packed column and an adequately packed "(
column is shown in Animation 3.5. ")
"*
#+
#"
##
#$
#%
#&
#'
"*
"
#
Animation 3.5 The Effect of Adequate Column Packing on Peak Broadening in $
HPLC Columns. %
&
The particle size of the stationary phase also significantly affects the '
operation of an HPLC by determining the shape of the van Deemter curve. (
Figure 3.6 shows the van Deemter curves for a variety of particle-sized stationary )
phases. Recall from our discussions in Chapter 1, better separations occur at *
the minimum point (linear velocity) in the van Deemter curve and a more stable "+
system operation will occur with a large range of linear velocities. In HPLC, ""
smaller sized stationary phase particles will yield a van Deemter curve that is "#
#+
essentially flat (refer to Figure 3.6). Thus, a range of flow rates will produce "
optimum analyte separation. #
$
%
Figure 3.6 The Effect of Stationary Phase Particle Size on the Optimum Flow &
Rate in HPLC. From R.E. Majors, Effect of Particle Size on Column Efficiency in '
Liquid-Solid Chromatography, 1973, 11, 88. Reproduced from the J ournal of (
Chromatographic Science by permission of Preston Publications, A Division of )
Preston Industries, Inc. *
"+
3.3.8 Detectors: There are a variety of detectors available for use with ""
HPLC. These range from near universal detectors that will respond to any "#
organic chemical structure to specialty detectors that only respond to a few "$
analytes. Both have their advantages. Universal detectors provide economic "%
use for a large diversity of chemicals but specific analyte detection can be "&
hampered if two or more analytes elute from the analytical column at the same "'
retention time. Specialty detectors provide near-interference free detection of "(
#"
only a few analytes. Common detectors, along with their use and approximately "
detection limits are given in Table 3.1. #
Table 3.1. Common Commercially Available HPLC detectors, Applications, and $
Detection Limits. %
Detector Application(s) Detection Limits
UV-Visible Absorbance For compounds that
absorb in the UV or
visible range
pg quantities
Fluorescence For compounds capable
of fluorescence
(especially polyaromatic
hydrocarbons)
fg quantities
Refractive Index (RI) For alcohol, sugar,
saccharide, fatty acid,
and polymer analysis
with refractive indices
different from the solvent
ng quantities
Electrochemical For analyzing a wide
range of compounds
high pg quantities
Conductivity for IC Mainly for inorganic ions ~ng quantities
Evaporative Light
Scattering (ELS)
For a wide variety of
compounds that lack
UV/Vis chromophores
including triglycerides,
sugars and natural
products
g quantities
##
Fourier Transfer Infrared For compounds with
vibrational functional
groups
g quantities
Mass Spectrometry A universal detector g to pg quantities
depending on the type of
mass spectrometer
"
The most common detector is the UV-Vis detector that comes standard on #
basic HPLC systems. This is a near-universal detector since most organic $
compounds have a chromophore capable of adsorbing UV or visible %
wavelengths. The focus of this textbook, mass spectrometer detectors, will be &
discussed in the next chapter. '
3.4 Advanced and Specialty LC Systems (
3.4.1 Ultra-High Pressure Liquid Chromatography, U-HPLC: HPLC )
technology made a significant advance with the development of ultra high *
pressures systems. These systems, currently operate at 15 000 to 20 000 psi, "+
offer significant time and cost saving per analysis. The increased pressure ""
allows samples to pass through the column faster and results in significant "#
increases in column efficiency (and decreases in theoretical plate height) by "$
minimizing longitudinal diffusion processes. These systems offer higher column "%
efficiencies (resolution per column length) and allow the systems to operate at a "&
much wider range of linear velocities, flow rates, and backpressures. Nanobore- "'
sized columns offer up to 95 percent decreases in solvent use and interface well "(
with mass spectrometers. Higher pressure systems are under development that ")
will possibly operate at up to 100 000 psi and yield even greater column "*
efficiencies. #+
3.4.2 Ion Chromatography, IC: Ion chromatography is a form of HPLC #"
where the system is modified to analyze for mainly inorganic ions such as ##
#$
nutrient ions (i.e. nitrate, sulfate, phosphate, etc.) as well as many organic anions "
and metal cations. The systems require a special ion exchange analytical #
column that is specific to the analytes of interest, a relatively simple conductivity $
detector, and a unique ion suppressor column for removing ions other than the %
analytes that would generate electrical conductivity in the detector. Columns and &
tubing are usually composed of PEEK to avoid reactions on metal surfaces. '
Analytical columns are mainly divided into columns for cations and anions and (
are similar in diameter to standard HPLC columns but slightly longer in length. )
Cationic exchange columns have active strong acid sites such as sulfonic acid (- *
SO
3
-
H
+
) or weak acid sites such as carboxylic acid groups (-COO
-
H
+
) that "+
preferentially exchange with analyte cations. Common anionic exchange sites ""
include the strongly basic tertiary amines (-N(CH
3
)
3
+
OH
-
) and the weakly basic "#
primary amine group (-NH
3
+
OH
-
). Both exchange groups are usually placed on "$
porous microparticles of silica. As the injected sample passes through the "%
analytical column, shown below for the sulfonic acid exchange surface, the "&
following reaction occurs "'
-R-SO
3
-
H
+
solid
+ Metal Cation
+
solution
<--- > "(
(-R-SO
3
-
)
n
Metal Cation
+
solid
+ nH
+
solution
")
"*
Similarly for anion analysis, the reaction would be #+
-NH
3
+
OH
-
solid
+ Anion
-
solution
<--- > #"
(-NH
3
+
)
n
Anion
-
solid
+ nOH
-
solution
##
At appropriate flow rates, an equilibrium migration front is set up for each analyte #$
passing through the column. Since each metal cation will have a different affinity #%
for the stationary ion exchange resin, each analyte will elute from the column at a #&
different time. As in all types of chromatograph (in the absence of mass #'
spectrometry detection), identification is largely based on retention time. #(
#%
The detector in IC is a simple conductivity detector where a potential is "
placed across two electrodes. In the absence of ions in the solution passing #
through the detector, minimal current is transmitted and no signal is generated by $
the electronics. As each packet of analytes, cations or anions, pass through the %
detector a current is generated that is proportional to the concentration of &
analytes. In order to detect low concentrations of analytes, the background '
signal (current) of the mobile phase must be maintained as low as possible. (
Since all samples have counterions (an equal concentration of cations or anions) )
half of the ionic components must be removed prior to entering the detector. For *
example, detecting the presence of cations require the removal of anions before "+
the column effluent reaches the detector. The removal of these ions is ""
accomplished with an ion suppressor device shown in Figure 3.7. This device is "#
positioned within the ion chromatography system as shown in Figure 3.8. "$
"%
Figure 3.7 Photograph of Two Ion Suppressor Devices. "&
#&
"
Figure 3.8 Overview of the Components used to Separate Analytes in an Ion #
Chromatography System. Source: Dionex Corporation Product Manual ASRS $
300 CSRS 300, Figure 7, page 20. Reprinted with permission, courtesy of %
Dionex Corporation, Sunnyvale, California. &
As presented in Figure 3.8, the sample first passes through the analytical '
column where the cations or anions are separated relative to their affinity for the (
solid-phase exchange resin. The effluent from the analytical column then enters )
the ion suppressor device, along with acid or base to keep the ion suppressor *
column active. For anionic analysis, strong acid (H
+
) is used in the suppressor "+
device; for cationic analysis, strong base (OH
-
or COO
-
) is used in the device. In ""
both cases the analytes of interest are unchanged and unretained. As a "#
consequence, the following suppression reactions occur depending on the mode "$
of operation (cation or anion analysis). "%
In anion analysis, cations in the sample are exchanged for H
+
in the suppressor "&
column and neutralized in the center flow through portion of the suppressor "'
column by the following reaction "(
#'
H
+
solution
+ Cl
-
solution
+ resin
+
OH
-
solid
--- > "
resin
+
Cl
-
solid
+ H
2
O #
Thus, only anions, such as Cl
-
, contribute to the conductivity in the detector. $
In cation analysis, anions in the sample are exchanged with HCO
3
-
(or OH
-
) in %
the suppressor column and neutralized in the center flow through portion of the &
suppressor column by the following reaction '
cation
+
solution
+ HCO
3
-
solution
+ resin
-
H
+
solid
--- > (
resin
-
cation
+
solid
+ H
2
CO
3solution
)
Thus, only cations, such as Ca
2+
, contribute to the conductivity in the detector. *
Again, note that these reactions convert all ionic species, except for the analytes, "+
to non-conductive chemicals. These reactions in relation to the flow through the ""
ion suppressor are illustrated in Figure 3.9. In both cases, non-ionic species are "#
produced and the interfering ions are removed (suppressed) from the solution. "$
This significantly decreases the background ions in solution and allows for low "%
detection limits (parts per trillion to parts per billion in the injected sample). Ion "&
chromatography systems also operate well with MS detectors. "'
#(
"
Figure 3.9 Cation suppression in an ion suppression device. Source: Dionex #
Corporation Product Manual ASRS 300 CSRS 300, Figure 4, page 9. Reprinted $
with permission, courtesy of Dionex Corporation, Sunnyvale, California %
&
3.4.3 Super Critical Fluid Chromatography, SCF: A liquid turns into a '
super critical fluid (SCF) when its temperature rises above the critical (
temperaturethe temperature where it can no longer exist as a liquid no matter )
how much pressure is applied. Super critical fluids exist in a state between a *
liquid and a gas and have the penetration abilities of a gas and the dissolving "+
power of a liquid. As a result, SCF is a mixture of GC and HPLC and in some ""
cases is superior to GC or HPLC. Super critical CO
2
, the same matrix used to "#
selectively extract caffeine from coffee beads and nicotine from tobacco "$
products, can be used as a mobile phase in chromatography. Super critical fluid "%
#)
chromatography is a special form of HPLC where a near-identical system is used "
but the mobile phase, as noted above, is super critical CO
2,
. The system is #
therefore pressured and temperature controlled to maintain the super critical $
fluid. SCF is a form of normal phase chromatography that is used for the %
analysis of thermally labile molecules. The same types of packed and capillary &
columns that are used in HPLC are utilized in SCF. Due to the nature of the '
super critical fluid, packed columns can actually contain more theoretical plates (
than capillary column. Also, the shape of the van Deemter curve is different from )
those observed in GC and HPLC in that a minimum plate height exists over a *
very broad range of linear velocities. SCF can be used for a variety of "+
separations but it is most commonly used in the separation of chiral compounds ""
in the pharmaceutical industry. In place of a solvent gradient the "#
chromatographer uses pressure programming and the affinity of the stationary "$
phases to separate complex mixtures of analytes. Pressure programming in "%
SCF is analogous to temperature programming in GC and gradient programming "&
in HPLC. Methods of detection include UV/Vis, mass spectrometry, FID (as in "'
GC but unlike in HPLC) and evaporative light scattering. "(
3.5 Applications ")
HPLC extends the capabilities of chromatographic separations past the "*
abilities of capillary GC by allowing the analysis of thermally unstable analytes #+
but at the expense of poorer resolving power due to the packed column nature of #"
the system. These limitations are being slowly overcome with the use of capillary ##
columns and ultra high pressure systems. This section highlights a few HPLC #$
applications. As with GC, major column manufacturers and distributors provide #%
very useful websites that contain example chromatograms for common analytes. #&
These sources can be extremely helpful in selecting the appropriate column for #'
your needs. In addition, technical help is available from professional #(
chromatographers at these companies with more complex separations. The #)
chromatograms below were selected from numerous ones that are available from #*
Agilent technologies at http://www.chem.agilent.com/en- $+
#*
us/Search/Library/Pages/ChromatogramSearch.aspx and Dionex Corporation at "
http://www.dionex.com/en-us/columns-accessories/ion-chromatography. #
The first example we will illustrate is for polyaromatic hydrocarbons that $
can be analyzed very efficiently by GC. While HPLC also offers adequate %
separation (as shown in Figure 3.10 below) as compared to GC, HPLC allows &
improved detection limits with the use of fluorescence detection. '
(
$+
Figure 3.10 Separation of PAHs by HPLC. Source: Copyright 2006 Agilent "
Technologies, Inc. Reproduced with Permission #
Applications that cannot be performed by GC include the isolation of $
Vitamin A (Figure 3.11), the separation of proteins (Figure 3.12) and the %
separation of food coloring agents (Figure 3.13). &
'
$"
Figure 3.11 Isolation of Vitamin A by HPLC. Source: Copyright 2006 Agilent "
Technologies, Inc. Reproduced with Permission #
$
%
&
Figure 3.12 The separation of proteins by HPLC. Source: Copyright 2006 '
Agilent Technologies, Inc. Reproduced with permission. (
$#
"
#
$
Figure 3.13 The separation of food coloring agents by HPLC. Source: %
Copyright 2006 Agilent Technologies, Inc. Reproduced with permission. &
'
$$
A separation of anions in drinking water by IC is shown in Figure 3.14. "
#
$
Figure 3.14 Separation of anions by IC. Source: Dionex Corporation, Inc. %
Reprinted with permission, courtesy of Dionex Corporation, Sunnyvale, &
California. '
(
3.6 Summary )
This chapter described the basic applications and operation of an HPLC *
system. Modern HPLC systems significantly expand the capabilities of "+
chromatography to include the relatively delicate biomolecules that are so ""
abundant in nature. Current research in ultra high pressure systems may result "#
in HPLC performances that match those of capillary column GC in the near "$
future. Basic non-confirmatory detectors systems were described in this chapter "%
$%
and confirmatory identification by mass spectrometer techniques that are "
applicable to all forms of chromatography will be covered in Chapter 4. #
$
3.7 Questions %
1. Contrast the advantages and disadvantages of thin layer chromatography &
(TLC) versus modern HPLC. '
(
2. What does HPLC stand for? )
*
3. What are the advantages of dual reciprocating pumps have over syringe "+
pumps? ""
"#
4. How much does a basic HPLC system cost? "$
"%
5. What are the sub-categories of liquid chromatography? "&
"'
6. What is the difference between normal phase HPLC and reverse phase "(
HPLC? Which is most commonly used today? ")
"*
7. What chemical factors determine if a chemical will be analyzed in a GC or #+
LC? #"
##
8. Can moderately volatile, thermally stable chemical be analyzed on an LC? #$
#%
9. Why do we filter analyte solutions before injection into an HPLC? #&
#'
10. Draw a basic HPLC system and label all of the components. #(
#)
$&
11. Why are pressurized gases used in HPLC? "
#
12. What two preparatory steps must be taken before a solvent can be used as $
an HPLC mobile phase? %
In general, what is the maximum pressure limit of standard HPLC systems? &
'
13. What is the purpose of the proportioning valve? How does this reduce the (
cost of an HPLC? )
*
14. What is the difference in isocratic and gradient programming? Why is "+
gradient programming sometimes necessary? ""
"#
15. Why are dual piston pumps preferred over single piston pumps? "$
"%
16. What is the purpose of a pulse damper? "&
"'
17. Why are six-port valves used for injecting samples in HPLC? "(
")
18. Draw and explain how a six-port valve works. "*
#+
19. Why are in-line filters used in HPLC systems? #"
##
20. What is the composition of the stationary phase and purpose of the guard #$
column? #%
#&
21. What are common stationary phases used in reverse phase HPLC? #'
#(
22. Why do chromatographers purchase their analytical columns instead of self- #)
packing their own? #*
$'
"
23. How will a poorly packed column affect performance? #
$
24. What is the relationship between performance (resolution) and stationary %
phase particle size? &
'
25. Compile a list of HPLC detectors and provide a list of chemicals each can be (
used to analyze. )
*
26. Name three advanced types of LC. "+
""
27. Why is U-HPLC superior to standard HPLC? "#
"$
28. How does IC differ from standard HPLC? "%
"&
29. What is the purpose of the suppressor column in IC? "'
"(
30. Draw a suppressor column for cation analysis in IC. Explain how it works. ")
Write all requisite chemical reactions. "*
#+
31. Although not shown in this textbook, attempt to draw a suppressor column #"
for anion analysis in IC. Explain how it works. Write all requisite chemical ##
reactions. #$
#%
32. What is a super critical fluid? #&
#'
33. What types of compounds are usually separated by SCF chromatography? #(
#)
34. What type of gradient is used in SCF chromatography? #*
$(
"
35. Review each of the chromatograms in Section 3.5 and determine what #
intermolecular force is involved in these attractions. $
%
&
3.8 References '
Agilent Technologies Internet site at http://www.home.agilent.com (
Dionex Corporations Internet site at http://www.dionex.com )
Supelco-Aldrich Internet site at http://www.sigmaaldrich.com *
Waters Corporation Internet site at http://www.waters.com "+
""
!"#$%%"&' )%*+,&-#.-&*/$/

By Nicole }ames

4.1: Intiouuction

The sepaiation of compounus baseu on theii movement when exposeu to an electiic
fielu was fiist obseiveu in 18u7 by Feiuinanu Fiieuiich Reu, who noticeu the
movement of clay paiticles in watei when a constant electiic fielu was applieu. The
theoiy of electiophoiesis was iefineu in the eaily 19uus by Naiian Smoluschowski,
anu fuithei iefineu in 19S7 by Aine Tiselius, who won the 1948 Nobel Piize in
Chemistiy foi his woik.

Electiophoiesis was initially conuucteu in polyaciylamiue oi agaiose gels on a
slaba technique calleu !"#$ &'" '"'()*+,-+*'!.!wheie chaigeu molecules aie
applieu to one enu of the slab anu an electiic fielu is applieu ovei the length of the
slab. The molecules migiate uown the slab at uiffeient iates accoiuing to the chaige-
to-size (mz) iatio. Slab gel electiophoiesis is still wiuely useu in the fielus of
biology anu biochemistiy on laige molecules such as nucleic acius anu pioteins to
show ielative concentiations of molecules, the puiity of a sample, oi, when useu in
conjunction with a stanuaiu, to iuentify compounus. Slab gel electiophoiesis
geneially has long analysis times (often 2u-4u minutes pei sample), low efficiency,
uifficulties in analysis, anu is unable to uefinitively iuentify compounus in a sample
as migiation time is not necessaiily unique to each compounu. Auuitionally, slab-gel
electiophoiesis is uifficult to automate, making it veiy time-intensive to iun
multiple samples.

Capillaiy electiophoiesis (CE) is wiuely useu as an alteinative to slab gel
electiophoiesis. uel meuia aie not necessaiy in capillaiy electiophoiesis as capillaiy
tubes aie themselves anti-convective. Stellan Bjiten piefoimeu the fiist woik in
capillaiy electiophoiesis in 1967, using millimetei-uiametei capillaiy tubes. By the
eaily 198us, the uiametei of capillaiy tubes hau been ieuuceu to 7Sm. Capillaiy
electiophoiesis geneially iuns fastei than slab electiophoiesis, pioviues bettei
piecision anu accuiacy, uses fewei ieagents anu is moie easily automateu. Capillaiy
electiophoiesis can also analyze smallei molecules than slab electiophoiesis, thus
expanuing the iange of possible analytes. By paiiing capillaiy electiophoiesis with
mass spectioscopy, it is possible to obtain confiimatoiy iesults. This makes
capillaiy electiophoiesis an intensely useful anu poweiful instiument foi many
scientific uisciplines. An oveiview of a CE system is shown in Figuie 4.1.

Figuie 4.1 0veiview of a mouein CE system.

4.2: Electiophoiesis anu Capillaiy Electiophoiesis

The sepaiation of compounus in capillaiy electiophoiesis uepenus on the velocity of
inuiviuual compounus. The velocity is given by the electiophoietic mobility (e) of
the compounu anu the applieu electiic fielu, (E)

!
velocity =
e
E Eqn 4.1

The electiophoietic mobility is a measuie of the paiticle's tenuency to move
thiough the meuium given the applieu electiical fielu anu thus changes with the
meuium anu the paiticle; tabulated values of electrophoretic mobility often differ from
the experimentally observed electrophoretic mobility. Consequently, experimentally
determined mobilities are called effective mobilities, and can change radically with
different solvents and with different solution pH. The electiophoietic mobility
uepenus on the fiictional uiag (Ff) the meuium exeits on the paiticle anu the
electiical foice (Ef) exeiteu to move the paiticle such that

!

e
=
E
f
F
f
Eqn 4.2

wheie the electiical foice is uepenuent on the chaige of the ion (q) anu the stiength
of the electiical fielu (E)

!
E
f
= qE Eqn 4.S

The fiictional foice can be uesciibeu as

!
F
f
= "6#$rv Eqn 4.4

wheie i is the ion iauius, v is the ion velocity, anu ! is the solution viscosity.

In slab electiophoiesis, the sample is placeu on the gel, anu then the electiical fielu
is applieu. The foice of the electiical fielu causes the sample to piogiess uown the
slab, acceleiating until the uiag--oi fiiction--expeiienceu by inteimoleculai
inteiactions with the meuium equals the foice causeu by the electiical fielu, anu
then the sample pioceeus at a constant iate. 0nce the sample has ieacheu a steauy
velocity, it can be shown that

!

e
=
q
6"#r
Eqn 4.S

This equation shows that highly chaigeu, small paiticles move quicklyespecially
with a low viscosity fluiuwheieas less-chaigeu, laigei paiticles move moie slowly,
especially thiough a moie viscous meuium.


4.2.1: Electro-Osmotic Flow (EOF)

Also called electroendosmotic flow, electro-osmotic flow is the movement of liquid in a
porous material (such as a capillary tube) caused by a difference in potential across the
material.

When a charged particle is put in contact with a liquid in a capillary tube, a double-layer-
-or electrical double layer (EDL)--forms at the wall of the capillary (see Figure 4.2); this
occurs at the interface of the glass capillary wall and the bulk solution. The first layer is
surface charge, and can be positive or negative depending on the material. As capillaries
are generally borosilicate glass, the numerous silanol (SiOH) groups cause the charge of
the first layer to be negative. This layer is also sometimes called the Stern layer or
Helmholtz layer. The second layer is made up of ionic particles in solution that are
electronically attracted to the charge of the capillary surface. As the particles in this layer
are not fixed, but move as a result of electrical and thermal energy, it is called the diffuse
layer.


Figure 4.2: Capillary double-layer

The diffuse layer has a net charge, whereas the bulk, uncoordinated solution generally
does not. When an electrical potential is placed across the capillary tube, the diffuse layer
is pulled to one side. As the diffuse layer progresses to one side of the capillary tube, it
drags the bulk solution along with it, creating a flow (specifically, the electro-osmotic
flow) of the solution through the cathode. Because the species in the solution are charged,
they separate in solution as discussed above; for most applications
*
, positively-charged
analytes move toward the anode given normal (negatively charged capillary) conditions.

The electro-osmotic flow (EOF) can be described in terms of velocity or mobility. The
mobility can be calculated from Equation 4.6, where: " (unitless) is the dielectric
constant, a measure of a materials ability to respond to a dialectrics ability to store
charge; # (V) is the zeta potential, discussed later; and ! (
!
Pa s =
kg
s m
) is the solution
viscosity.

!

EOF
=
"#
$
Eqn 4.6

Since the mobility is the ability to respond to electric potential, the velocity is the
mobility multiplied by the electric potential:

!
V
EOF
= E
"#
$
Eqn 4.7

The zeta potential (#) is the potential due to the double layer on the capillary wall, and is
directly correlated with the charge on the capillary. Because the charge on the capillary
varies as a function of pH, the zeta potential also varies with pH, meaning the mobility
and velocity of the EOF are highly pH dependent. At high pH, the silanol groups are
deprotenated, meaning the net charge of the capillary wall is greater. Thus, the EOF is
significantly greater at high pH than at low pH; in fact, the EOF can sometimes vary
more than an order of magnitude between a pH of 2 and a pH of 12.


Figure 4.3: Pump induced flow vs. electro-osmotic induced flow


*
Foi small monoatomic ions, the solute mobility can sometimes be gieatei the E0F; foi these
conuitions, anions anu cations will move in opposite uiiections.
A marked difference between electro-osmotic flow and flow induced by a pump is the
cross section of the flow. Because a pump applies a force across the entire cross sectional
area of the liquid in a capillary or tube, the friction between the wall and the liquid will
cause the liquid at the center of the cross section to move faster than the liquid closer to
the wall. Thus, the liquid velocity gradient flows parabolically down through the
capillary. However, if a liquid is moved by electro-osmotic flow, the bulk flow is caused
primarily by the acceleration of the cations near the capillary wall, and the force on the
liquid is uniformly distributed. This results in a flat flow profile. As seen in (Figure 4.3),
the flat flow profile drops off sharply at the capillary wall; this is caused by friction
against the direction of flow, but to a much lower extent than observed in laminar flow
caused by pumps. An even (or flat) flow results in compact peaks whereas a parabolic
laminar flow results in a broad peak because the analyte is spread over a larger area.
Tighter peaks are easier to integrate and facilitate the isolation of peaks that elute at
similar times.

Table 4.1 summarizes the variables that control electro-osmotic flow.

Variable Effect on EOF Possible Problems
pH directly proportional can alter structure and charge of analyte
Ionic Strength inversely proportional sample heating and adsorption
Temperature inversely proportional changes in viscosity
Electric Field directly proportional can cause heating
Organic Modifiers usually inversely proportional can alter selectivity of instrument
Hydrophilic Polymer usually inversely proportional dependent on polymer
Covalent Coating dependent on coating coating often unstable
Table 4.1: EOF variables

4.2.2: Peak Resolution and Solute Zones

Peak resolution in capillary electrophoresis is dependent on many variables, including
longitudinal diffusion, injection length, sample adsorption to capillary walls,
electrodispersion, Joule heating, and detector cell size. The loss of peak resolution is
often referred to as zone broadening in CE. As in other forms of chromatography, this
results in reduced separation of compounds and increased difficulty in accurate peak
integration.

4.2.2.1: Longitudinal Diffusion

Under normal conditions, diffusion along the capillary tube ( longitudinal diffusion) is
one of the more substantial contributors to zone broadening. A solutes propensity to
diffuse throughout the solvent depends on the solute and the solvent. A longitudinal
diffusion coefficient can be provided for a given system, with a lower coefficient
meaning the solute will form narrower zones.

The equation for the number of theoretical plates in a capillary tube is given in Equation
4.8, where N is the number of theoretical plates,
a
is the solute mobility, E is the electric
field, I is the effective capillary length, and D is the diffusion coefficient of the solute.

!
N =

a
EI
2D
Eqn 4.8

This equation only accounts for longitudinal diffusion, although there are often other
variables that affect zone broadening.


4.2.2.2: Injection Length

When introducing a sample to a CE system, the volume of liquid that enters the capillary
is called the sample plug (See Figure 4.4). If the sample plug is very spread out, analytes
will not separate into tight peaks and tailing or broadening can occur. A practical
injection should be between 1% and 2% of capillary length.


Figure 4.4: Diagram of sample introduction and plug length

4.2.2.3: Sample Adsorption

Interactions of the analyte with the capillary wall negatively affect zone broadening. In
some cases, complete adsorption of the analyte to the capillary wall can occur. As most
capillary tubing is made of borosilicate glass, this is most substantial with positively
charged analytes that can interact strongly with the hydroxide groups on the surface of
the glass. Prolonged sample-surface interactions can result in peak tailing and zone
broadening.

4.2.2.4: Electrodispersion

The capillary in a CE system contains an ambient level of ions, called the background
electrolyte (BGE). For ideal separation, the BGE should be uniform in and between all
solute zones. Therefore, the background solvent is generally always a buffer.


4.2.2.5: Joule Heating

Joule heat is generated by passing electric current through a conductor and is directly
proportional to the amount of power used. Joule heating can result in non-uniform
temperature gradients throughout the capillary tube. Temperature increases are often
around 10C, although increases of up to 70C or higher can occur. While uniform
increases in temperature generally do not affect zone broadening, uneven increases in
temperature leading to temperature gradients can cause substantial peak broadening
and/or tailing. Heat can relatively easily dissipate through capillary walls, which can
result in a large difference in temperature between the center of the capillary and the
internal edges of the capillary. These gradients can result in differences in solvent
viscosity, and thus air or liquid cooling of the capillaries is often necessary. Using thicker
capillary tubes with a large outer radius and a smaller inner radius can also help combat
this problem. Smaller sample volumes and higher surface-to-volume ratios limit the Joule
heating, even when hundreds of volts per centimeter are applied.

4.2.2.6: Detector Cell Size

Detectors on CE systems average a signal over a finite volume. Naturally, if the volume
averaged is too large, it can combine signals from different analytes. In this way, zone
broadening due to the size of the detector cell is similar to zone broadening due to a long
injection plug. Generally, any variance due to detector cell size can be ignored if the
length of the detection cell is roughly one third the sample zone length.


4.3: Samples

4.3.1: Sample Introduction

For loading a sample, a stable DC power supply of about 30kV (200-300mA) is required.
Instruments are made with the ability to switch polarity to facilitate a variety of methods
of operation. Usually, current is monitored and maintained to compensate for some
temperature-induced changes in buffer viscosity.

Sample volumes are generally incredibly small. Volumes as low as 5L can be sufficient
for several injections. This can be disadvantageous if the sample is very dilute, because
being limited on the loading volume of a dilute sample can reduce sensitivity. Loading
too much of a sample can result in bad peak shapes due to field inhomogeneities if the
conductivity of the sample is different from that of the buffer. Additionally, if the
injection length is greater than the diffusion-controlled zone, peak broadening will occur.

4.3.1.1: Types of Sample Injection

There are several methods for introducing samples into CE systems. It is important to
note that some sample is always injected when inserting a capillary into the sample
reservoir, due to capillary action. This volume is called the Zero-Injection, and the effect
of this initial injection on sample migration and analysis is called the Zero-Injection
Effect. In most cases, this phenomenon is insignificant, but it can be important in systems
being employed for quantitative analysis.

4.3.1.1.1: Hydrodynamic Injection

Hydrodynamic injection introduces a sample by applying a pressure on the sample
reservoir. This physically forces solution through the capillary. The volume of sample
loaded is independent of the sample matrix; instead, the volume depends on the details of
the system, such as capillary dimensions, the viscosity of the buffer, the pressure applied,
and the amount of time that pressure was applied. Pressures and times generally range
from 25-100mbar and 0.5-5 seconds.



Figure 4.5: Hydrodynamic injection setup

Similarly, hydrodynamic injection can be performed by siphoning, or raising the input
reservoir 5-10cm above the output reservoir while keeping the levels of each equal. This
technique is generally employed when the equipment necessary for pressure injection is
unavailable.

4.3.1.1.2: Electrokinetic Injection

In electrokinetic injection, a voltage 3-5 times lower than that required for separation is
applied between the input and output reservoirs for approximately 10-30 seconds. Sample
is loaded according to the difference in voltage across the capillary and the amount of
each individual analyte that is loaded is dependent on the analytes electrophoretic
mobility. As such, electrokinetic injection cannot be used quantitatively. This technique
is useful in situations where the media in capillary is very viscous making hydrodynamic
injection difficult. Additionally, it requires no extra instrumentation.


Figure 4.6: Electrokinetic injection setup

4.3.1.1.3: Field Amplified Sample Stacking (FASS) and Field Amplified
Sample Injection (FASI)

Samples zones are sharpened when the conductivity of the sample matrix is substantially
lower than that of the buffer. Essentially, this is the reverse of electro-dispersion,
described previously. Thus, samples with low salt content can be actively sharpened
and/or large quantities of them can be loaded onto a capillary before zone-broadening
effects begin to occur. When this phenomenon is utilized to allow larger sample lengths
to be loaded hydrodynamically, the concentration is called field amplified sample
stacking (FASS). When the injection is performed electrokinetically, it is called field
amplified sample injection (FASI).

In the case of FASI, if the conductivity of the sample varies or is unstable, results can be
unreliable. In such a case, FASS can be used instead and, while peak resolution may be
affected by variability in sample conductivity, the peak areas are still reliable.

Often 100 times more sample can be loaded by FASS before separation is greatly
affected. The actual amount of sample that can be loaded by FASS is regulated by the
samples conductivity and factors such as Joule heating. As the conductivity of the
sample must necessarily be much lower than that for the running buffer, the applied
electric field can induce a large degree of heating within the sample plug, which can
change the viscosity of the sample and result in degassing and/or boiling.

Additionally, due to the difference in ionic strength between the sample plug and the
running buffer, the electro-osmotic mobility will vary between the sample plug and the
buffer. This can result in a parabolic flow profile between the plug and the running buffer
that may cause zone broadening. Similarly, a substantially different pH between the
sample plug and the buffer can also result in differences in analyte mobility, so a buffer
with a similar pH to the sample solution should be chosen.

4.3.2: Sample Purification

CE is attractive as a preparatory purification source because of its separatory strength.
However, CE can only be used to purify small amounts of analyte, as the sample loading
limit is low. However, small-volume fractions can be collected for sequencing DNA
fragments, establishing biological activity of substances, and ascertaining the purity of a
collected peak by reinjection.

Autosamplers can be programmed to run a sequence of samples so that, in the event that
multiple runs must be made and fractions collected and combined, the instrument does
not need to be attended for the entire period.

4.4: Methods of Operation

4.4.1: Capillaiy Zone Electiophoiesis (CZE)

Capillaiy Zone Electiophoiesis is the stanuaiu methou of CE. Sepaiations aie
peifoimeu by a uiffeience in migiation time as pieviously uesciibeu.


Figuie 4.7 Biagiam of a CZE system.

4.4.2: Nicellei Electiokinetic Chiomatogiaphy (NEKC)

Nicellei electiokinetic chiomatogiaphy is a combination of electiophoiesis anu
chiomatogiaphy. If one uissolves suifactants (foi example, 8-9mN SBS) in the
iunning buffei, micelles--often anionic--foim anu tiavel against the E0F. The E0F,
howevei, is still fastei anu the net motion of the micelles is in the same uiiection as
the E0F. Solutes paitition between the micelles anu the buffei thiough hyuiophobic
anu electiostatic inteiactions. In this way, the micelles act as a stationaiy phase
which is not necessaiily stationaiy, but slowei with iespect to the E0F. The moie
hyuiophobic compounus migiate moie slowly thiough the capillaiy uue to theii
inteiactions with the micelles. All neutial compounus will elute between the E0F,
which elutes fiist, anu the micelles, which elute last. (See Figuie 4.8).

The suifactants can be anionic, cationic, non-ionic, zwitteiionic, mixtuies of these,
salts, oi micioemulsions. Beciuing which suifactant to use is baseu on the specific
inteiactions necessaiy to sepaiate taiget compounus.


4.8: Example micellei electiokinetic chiomatogiam

Suifactants may affect the E0F anu solute-wall inteiactions. ueneially, using a high
pB buffei will maintain the E0F anu uiminish changes in solute-wall inteiactions.
Auuitionally, a consistent tempeiatuie is especially impoitant as changes in
tempeiatuie will affect the concentiations of suifactant iequiieu to foim micelles as
well as the paitition coefficient(s) of the system.


4.4.S: Capillaiy Electio-Chiomatogiaphy (CEC)

Capillaiy electio-chiomatogiaphy is miniatuiizeu liquiu chiomatogiaphy. An
applieu electiic fielu pumps liquiu thiough a packeu capillaiy column. The silica
capillaiies aie packeu with silica-baseu ieveiseu-phase paiticles of 1-Sm uiametei.
The migiation times of uiffeient solutes vaiy as a function of theii paititioning
between the mobile anu stationaiy phases. If the packeu phase is chaigeu, theie is
also an electiophoietic component to the paititioning. 0n aveiage, theie aie
1uu,uuu theoietical plates pei peak in a CEC system.

As the only uiffeience in a CEC setup is the capillaiy tubing useu, CZE systems can be
easily set up foi CEC analysis.




4.4.4: Capillaiy uel Electiophoiesis (CuE)

Capillaiy gel electiophoiesis is essentially gel electiophoiesis conuucteu in a
capillaiy tube. The "gel" is not necessaiily stiff anu soliu as it is in slab gel
electiophoiesis. To avoiu confusion, it is usually iefeiieu to as a polymei matiix.
Nany polymei matiices aie available anu offei uiffeient sepaiating piopeities.

The polymei is uissolveu in buffei anu loaueu into the capillaiy tube. ueneially, the
polymei concentiation necessaiy foi auequate sepaiation is inveisely piopoitional
to the size of the analyte. As low viscosity polymei solutions uo not expeiience the
same capillaiy action as most iegulai buffei solutions, the capillaiy gel is often
pumpeu with piessuie anu the capillaiy wall is coateu to eliminate E0F.

Theie aie geneially 1u million theoietical plates pei metei of capillaiy tubing. CuE
is most commonly useu to sepaiate single-stianueu oligonucleotiues, pioteins, anu
BNA fiagments.

As in CZE, it is possible to auu agents such as chiial selectois anu ion-paiiing agents
by covalently bonuing them to the gel oi the iunning buffei. This can gieatly
inciease the sepaiation possibilities of a single system.


4.4.S: Capillaiy Isoelectiic Focusing (CIEF)

Capillaiy isoelectiic focusing sepaiates peptiues anu pioteins baseu on theii
isoelectiic point (pI)--the point at which they caiiy no net chaige. A pB giauient is
foimeu within the capillaiy using ampholyteszwitteiionic molecules with both an
aciuic anu basic moiety.

The capillaiy is filleu with a mixtuie of solutes anu ampholytes, anu a pB giauient is
foimeu with a basic solution at the cathoue anu an aciuic solution at the anoue.
When an electiic fielu is applieu, ampholytes anu solutes migiate thiough the pB
giauient until they ieach the pB that coiiesponus to theii isoelectiic point, wheie
they become unchaigeu. The solute zones will maintain a naiiow bieauth as any
solute which migiates into the zone of a uiffeient pB becomes chaigeu anu migiates
back into the pB of its isoelectiic point. Aftei this equilibiium occuis, a steauy state
is ieacheu anu hyuiaulic piessuie is useu to pass the zones thiough to the uetectoi.

EOF must be minimized to prevent flushing ampholytes through the capillary.

4.4.6: Capillary Isotachophoresis (CITP)

Capillary isotachophoresis is a moving boundary technique using two buffer systems.
The buffer systems trap the analyte zones and all zones lay in between leading and
terminating electrolytes, where the leading electrolytes are those in the buffer with a
higher EOF and the terminating electrolytes are those in the buffer with the lower EOF.

This bracketing affects the electric field experienced by each zone, creating different
electric fields for each analyte. The electric field varies such that each analyte moves at
the same velocity as that established by the leading electrolyte. Because velocity is
defined as the product of the mobility and the electric field (velocity=mobilityfield), this
varying field acts to compliment the mobility such that all analytes move at the same
velocity. Zones are formed and focused much in the same way as they are in CIEF; as
soon as an analyte drifts away from its zone, it experiences a new electric field such that
it either accelerates or decelerates until it migrates back to its zone and the steady state is
reestablished.

According to the Kohlrausch Regulating Function, the concentration of analyte in a zone
is determined by the concentration of the leading electrolyte. Analytes with a higher
concentration than the leading electrolyte experience broadened until the concentrations
match, whereas analytes with a lower concentration than the leading electrolyte sharpen.
In this way, CITP can be used as a preconcentration step prior to other methods of
analysis. Generally, up to 30-50% of the capillary can be filled with sample while still
maintaining good separation and concentration.

Also due to the Kohlrausch Regulating function, increasing the concentration of the
leading and/or trailing electrolytes can result in better separation and can improve
detection of analytes that can be difficult to separate. However, when applying CITP, it
can be difficult to identify a buffer system with leading and trailing electrolytes that have
the appropriate mobilities while still maintaining the appropriate pH for the sample
analysis. Additionally, only cations or anions can be analyzed by a given buffer system.

4.5: Detectors

Detection in CE can be a significant challenge, as the small injection volumes require the
samples to be relatively concentrated for reasonable detection. Many of these techniques
are similar to those employed in liquid chromatography.

4.5.1: UV-vis Spectrophotometry

UV-visible absorption is the most common detection technique for CE. Detection is
nearly universal, and with fused-silica as the sample holder, wavelengths from 200nm
through the visible spectrum can be used without interference. Generally, the capillary
itself is used as the sample holder and passes through a UV-vis detector.

Because the capillary tube is so thin compared to conventional path lengths, care needs to
be taken in the design of the detector. The beam of light must be of small radius and
focused directly on the capillary for accurate results. A disadvantage of this is that,
correspondingly, few photons reach the detector and a generally high baseline is
obtained. The optical beam is generally passed through a slit to restrict its width and
center it on the capillary tube.

One approach to decrease the detection limit is by increasing the path length of the flow
cell. This can be done by increasing the diameter of the capillary, but by doing this there
is also an increase in the current and resulting Joule heating. Two ways of increasing path
length without increasing capillary diameter are depicted in Figure 4.10. The method
shown in 4.10A, called the bubble cell does not require an increase in current, as the
detection occurs after separation. When a zone enters the bubble, the velocity decreases
as it expands to fill the capillary, it contracts along the capillary and overall concentration
remains the same, even with increasing path length. There is a third method, which is a
combination of 4.10a and 4.10b; the capillary bends to increase the path length, but the
capillary is also bubbled along the path.


4.10: a, left) Capillary bubble cell for UV-Vis detection; b, right) capillary cell.


When trying to detect a given analyte, care must be taken to ensure the running buffer is
not optically active in the range of the analyte. Most peptides and carbohydrates absorb
around 200nm, and consequently many biological buffers, such as HEPES, CAPs, and
tris, should not be used in systems analyzing peptides and carbohydrates.

A normal S/N for a 75mm id capillary is 62.5, and up to 650 on an increased path length
cell.

4.S.2: Inuiiect Photometiy

Inuiiect photometiy utilizes a monitoiing ion that is 0v-active to tiace the piesence
of a non 0v-active compounu. An ion of the iunning buffei is chosen, geneially to
have a high 0v absoibance intensity, anu its absoibance is monitoieu. When the
sample zone passes thiough the uetectoi, it will be accompanieu by a coiiesponuing
uiop in the 0v absoiption uue to the monitoiing ion.

A veiy simplifieu explanation foi this phenomenon is that the monitoiing ion is
uisplaceu by an analyte with the same chaige baseu on electioneutiality. Bowevei,
the uisplacement iatio can be gieatei than one anu analytes with chaiges opposite
to the monitoiing ion can be uetecteu. This can be bettei explaineu by a closei
analysis of the Kohliausch Regulation Function.


4.S.S: Lasei-Inuuceu Fluoiescence (LIF)

Lasei-inuuceu fluoiescence auapts fluoiescence spectiometiy to CE in the same
mannei as 0v-vis spectiometiy. A lasei is useu to excite the analyte, anu
fluoiescence is measuieu via a PNT. 0ften the lasei is aligneu non-collineaily, anu a
ball lense is useu to focus the light on the centei of the capillaiy. ueneially, a
ieflective ellipsoiu is glueu to the capillaiy to uiiect fluoiescence towaiu the PNT.

A uiawback of LIF is that most compounus of inteiest in CE aie not fluoiescent.
Bowevei, a numbei of ueiivatizing agents can be useu to act as maikeis foi the
compounu. These aie geneially oiganic aiomatics anu iequiie excitation between
2Sunm anu Suunm.


4.S.4: Contactless Conuuctivity

Contactless conuuctivity uetectois (CCBs) place an actuatoi electioue anu ieceivei
electioue along the capillaiy tube, as uepicteu in Figuie 4.11. An oscillating
fiequency is applieu at the actuatoi electioue anu uetecteu at the ieceivei electioue
with the liquiu insiue acting as a iesistoi. As a sample zone passes, the conuuctivity
changes. Since analytes aie chaigeu, the conuuctivity geneially incieases--except foi
when samples have a lowei ionic stiength than the iunning buffei.


Figuie 4.11 CCB setup.

Contactless conuuctivity uetectois aie univeisal anu woik well foi inoiganic
analytes that aie uifficult to measuie by othei methous. CCBs function well with low
sample volume, which is auvantageous foi use in CE. Auuitionally, it geneially has a
uetection limit in the ppm-ppb iange foi small oiganic ions.

4.S.S: Nass Spectiometiy

Nass spectiometiy is a paiticulaily attiactive uetecting stiategy as it is
confiimatoiy anu can give valuable stiuctuial infoimation about analytes. Bowevei,
it is uifficult to tiansition fiom a liquiu-phase sepaiatoiy technique to a gaseous,
vacuum-baseu uetection technique. To an extent, this pioblem is alleviateu by the
small amount of solvent anu the low flow iate associateu with CE.

4.S.S.1: Electio-Spiay Ionization (ESI)

Electio-spiay ionization is an iueal technique foi inteifacing CE anu NS, as it accepts
small volumes at a low flow iate anu iesults in gaseous, ionizeu paiticles (foi a
uetaileu uesciiption go to section S.S.1). Auuitionally, ESI is well-suiteu to laigei,
biological molecules, which aie fiequently analyzeu by CE. Bowevei, CE anu ESI aie
both uepenuent on electiic fielus. In oiuei to couple them in the simplest of cases,
the enu electioue of the CE instiumentation is shaieu with the ESI electioue.
Bowevei, in some situations the uiffeience in the magnetic fielu can be seveial
oiueis of magnituue anu the electioues may be of the opposite sign.

Some situations can be iemeuieu by giounuing the enu electiolyte of the CE system,
anu pioviuing voltage to the electio-spiay system fiom the NS. If the potential
applieu to the ESI is negative, then positive ion will entei the NS anu this is calleu
positive ion moue. If the CE voltage is much gieatei than the ESI voltage, a giounu
path foi the cuiient is pioviueu by a iesistoi sink. In geneial, when the ESI voltage is
much uiffeient fiom the CE voltage, the vaiiability in the electiic fielu can affect the
sample sepaiation in the CE system.

4.S.S.1.1: Byuiaulic Inteifacing

A hyuiaulic flow iate is necessaiy to foim the electio-spiay. The amount of
hyuiaulic flow necessaiy is uepenuent on many vaiiables, incluuing the E0F anu
capillaiy uimensions. In geneial, theie aie two methous foi intiouucing a hyuiaulic
flow: sheath flow anu sheathless flow.

4.S.S.1.1.1: Sheath Flow Inteiface

Smith, et al. inventeu the tii-axial capillaiy spiayei, similai to the one shown below.




Figuie 4.12: Sheath flow inteiface

Solvent flows between the capillaiy anu the spiayei tip, anu combines with the
capillaiy solution immeuiately befoie the nebulizing gas. In this way, a hyuiaulic
flow inuepenuent of the E0F is pioviueu by the sheath solvent.

The majoi uisauvantage of the sheath flow inteiface is the fact that the sample is
necessaiily uiluteu by the sheath solvent. Bespite this, the sheath flow inteiface is
one of the most ieliable methous of inteifacing CE anu NS, although seveial
alteinative methous exist but aie not "as iobust," accoiuing to Agilent liteiatuie. (It
shoulu be noteu that Agilent possesses the uistiibution iights on the tii-axial
sheath.)

4.S.S.1.1.2: Sheathless Flow Inteiface

Seveial methous of uniting the CE capillaiy with the ESI system exist Bowevei, in
all of these systems, electiochemical piocesses such as oxygen anu hyuiogen
foimation can occui. The foimation of gas bubbles that aie not fiequently venteu
can cause incieaseu iesistance. In the tii-axial system, evolveu gases aie fiequently
flusheu uue to sheath flow anuoi the nebulizei gas, but in sheathless systems
incieaseu iesistance can foim a substantial pioblem.

In 2uu7, Noini publisheu a technique that uoes not iequiie sheath flow anu uoes not
cieate electiochemical piocesses at the electioues. A thin, poious segment of glass
acts as a liquiu anu electiical contact outsiue the capillaiy. The tube is eithei
giounueu oi placeu at the ESI voltage. This system has iecently become available
commeicially thiough Beckman-Coultei.


4.S.S.2: Atmospheiic Piessuie Chemical Ionization (APCI) anu
Atmospheiic Piessuie Photo Ionization (APPI)

ESI is highly efficient foi many systems. Bowevei, foi systems wheie the iunning
buffei is not especially volatile anuoi the molecules aie not stiongly chaigeu,
alteinative ionization techniques may be moie efficient. Atmospheiic piessuie
chemical ionization (APCI) anu atmospheiic piessuie photo ionization (APPI) can
both be useu foi systems like these.

APCI ielies on the evapoiation of an analyte-containing aeiosol, wheie a chaigeu
ieagent gas colliues with analyte molecules anu confeis the chaige. In APPI, an
analyte aeiosol is again evapoiateu to a vapoi, anu then photons ionize the analyte.
In situations wheie the analyte is not easily photoionizeu, a buffei gas may be
photoionizeu anu then tiansfei the chaige to the analyte.

APCI anu APPI both iequiie substantial heating of the sample anu theiefoie aie not
appiopiiate foi analytes that easily uegiaue at highei tempeiatuies. Typically, APCI
anu APPI systems aie heateu to between 2uu-SuuC. Auuitionally, low moleculai
weight (<1kB) analytes aie bettei suiteu to APPI anu APCI as they aie moie volatile.
4.6: Citeu anu Consulteu Refeiences

Dolnik, V. DNA sequencing by capillary electrophoresis (review). J. Biochem. Biophys.
1999, 41, 103-119.

Lauer, H.H.; Rozing, C.P. High Performance Capillary Electrophoresis; Agilent
Technologies: Germany, 2010.

Petersen, J; Okorodudu, A.; Mohammad, A.; Payne, D. Capillary Electrophoresis and its
application in the clinical laboratory. Clinica Chimica Acta. 2003, 330, 1-30.

Tagliaro, F.; Manetto, G.; Crivellente, F.; Smith, F. A brief introduction to capillary
electrophoresis. Forensic Science International. 1998, 92, 75-88.

Tagliaro, F.; Turrina, S.; Smith, F. Capillary electrophoresis: principles and applications
in illicit drug analysis. Forensic Science International. 1996, 77, 211-229.
1
Chapter 5 1
Basic Mass Spectrometry 2
3
5.1 Introduction and History 4
5
The earliest forms of mass spectrometry go back to the observation of 6
canal rays by Goldstein in 1886 and again by Wien in 1899. Thompsons later 7
discovery of the electron also used one of the simplest mass spectrometers to 8
bend the path of the cathode rays (electrons) and determine their charge to mass 9
ratio. Later, in 1928, the first isotopic measurements were made by Aston. 10
These basic experiments and instruments were presented to most readers in 11
first-year general chemistry. More modern aspects of mass spectrometry are 12
attributed to Arthur J effrey Dempster and F.W. Aston in 1918 and 1919. Since 13
this time there has been a flurry of activity [not only concerning minor advances 14
in components of mass spectrometers such as different types of instrument 15
interfaces (direct injection, GC, and HPLC)] to different ionization sources 16
(electron and chemical ionization) but also new types of ion separators. For 17
example, double focusing magnetic sector mass filters were developed by 18
Mattauch and Herzog in 1934 (and recently revised into a new type of mass 19
filter), time of flight MS by Stephens in 1946, ion cyclotron resonance MS by 20
Hipple and Thomas in 1949, quadrupole MS by Steinwedel in 1953, and ion trap 21
MS by Paul and Dehmelt in the 1960s. 22
23
Mass spectrometry was first coupled with GC as a means of sample 24
introduction in 1956 by Golhke et al. and with HPLC via electro-spray ionization 25
in the mid 1980s (Blakely and Vestal, 1983; Yamashita and Fenn, 1984). New 26
methods of mass spectrometry are constantly under development and even as 27
recent as 1985, Hillenkamp and Michael Karas developed the MALDI technique 28
(a laser-based sample introduction device) that radically advanced the analysis 29
of protein structures and more types of mass analyzers will certainly be 30
developed. This chapter will deal only with basic mass spectrometer instruments 31
2
used in the analysis of organic chemicals exiting GC and HPLC systems, and is 1
also applicable to effluents from ion chromatographic systems. One of the most 2
comprehensive Internet summaries of the history of mass spectrometry can be 3
found at http://masspec.scripps.edu/mshistory/timeline/timeline.php. 4
5
5.2 Sample Introduction from GC and Analyte Ionization 6
7
The purpose of coupling GC with MS is to provide confirmatory 8
identification with minimal effort. Prior to the common availability of mass 9
spectrometers, confirmatory identification was possible but required twice the 10
effort. GC analysis alone can provide confirmatory analysis, but it is usually 11
necessary to analyze a sample using two different columns. With capillary 12
systems, it is possible to perform two independent analyses by installing two 13
different capillary columns into one injector system and monitoring each column 14
effluent with a separate detector. If the same retention time and concentration 15
are obtained, the identity of a compound is determined and the results are 16
considered confirmatory. 17
18
Capillary column systems are more easily interfaced with a mass 19
spectrometer than packed columns. The high flow rate of packed columns (30 to 20
60 mL/min) created problems in maintaining the necessary low pressure of a 21
mass spectrometer. On the other hand, capillary columns typically have a flow 22
rate between 1 and 5 mL/min which has a minimal effect on the low pressure MS 23
requirements. The GC and MS are interfaced by inserting the effluent end of the 24
capillary column into the MS with a standard nut and ferrule system near the 25
ionization source (Section 5.1.2a). Since GC analytes are volatile, the interface 26
and MS must be maintained at temperatures and pressures that keep the analyte 27
(or ionized form) in a volatile form. 28
29
As implied in the previous paragraph, mass spectrometer systems require 30
a low operating pressure, typically 10
-5
to 10
-6
Torr through out the system 31
3
(ionization source, mass analyzer, and detector). This is necessary to avoid 1
collisions between ionized molecules. If collisions are prominent, the mass 2
resolving capabilities will be effected which decreases the detection limit and the 3
resolution. Collisions also affect the interpretative value of the mass spectrum 4
preventing identification. 5
6
The MS works by (1) ionizing each analyte as it exits the GC column, (2) 7
accelerating and focusing the ionized compound and its fragments into the mass 8
analyzer, (3) separating the fragments in the mass analyzer based on mass to 9
charge (m/z) ratios, and (4) detecting the fragments as they exit the mass 10
analyzer. There are a variety of ionization systems and mass analyzers that 11
achieve these results. The following sections are dedicated to a simple 12
description of most common ones. 13
14
5.2.1 Analyte ionization 15
16
Analytes can be introduced into the ionization zone of a MS in two states, 17
a solid or a vapor. Solids can be introduced by depositing milligram quantities of 18
pure analyte onto a metal probe or in a matrix that is inserted into the ionization 19
chamber. These more direct forms of ionization do not require the interfacing of 20
a separatory instrument such as GC or LC since relatively pure analytes are 21
directly placed into the MS. More commonly, analytes enter the MS system in a 22
pure form (a peak) after separation by a capillary column GC. The MALDI 23
technique, an increasingly popular tool described below, does not neatly fit into 24
either of these categories but is included below due to its powerful applications 25
for biological systems. Irrespective of the samples state, analytes must be 26
ionizated into positively charged ions, and are in some cases broken into 27
fragments before they can be detected. Almost every compound has a unique 28
fragmentation pattern that can subsequently be used for conclusive identification 29
purposes. This pattern is dependent on the type of ionization source used and 30
the stability of the energized analyte molecule. Below we will divide the 31
4
ionization techniques into those for solid, non-volatile analytes and volatile 1
analytes entering the MS from a GC. 2
3
5.2.1.1 Ionization Techniques for Solid Non-Volatile Analytes 4
5
Field Desorption: Field Desorption (FD) techniques are relatively simple 6
and do not require analyte separation in a GC since only one compound is 7
introduced into the MS at a time. As noted in the heading above, compounds 8
analyzed by this technique tend to be non-volatile, have high molecular weights, 9
and degrade at higher temperatures. Analytes are introduced to the system on a 10
probe made of carbon fibers that has been lightly coated with pure analyte. A 11
high current is applied between the probe and an adjacent electrode. The 12
current moves the ionized analyte towards the end of the carbon fibers by charge 13
attraction, where the molecules are ionized into the vapor (plasma) phase. Then 14
they enter the mass analyzer and then the detector. The breaking of bonds 15
within the analyte (fragmentation) is rare in FD techniques, thus the spectrum 16
only contains the molecular ion. Many older inexpensive bench-top systems 17
used to come with a direct probe build into EI systems. However, this feature 18
has been removed due to the high number of service calls to clean out the MS 19
units when too much analyte was placed on the probe. Service technicians refer 20
to these analyte-rich probes as having peanut butter placed on them. 21
22
5.2.1.2 Ionization Techniques for Volatile Analytes Entering the MS from a GC 23
24
5.2.1.2a Electron Ionization or Electron Impact (EI): Electron ionization of 25
analytes is referred to as a hard ionization technique since it causes bonds to be 26
broken within a sample molecule (fragmentation). Neutral, radical, and positively 27
charged species are produced from fragmentation. Neutral and radical species 28
are not affected by the accelerator plates or mass analyzer and are removed by 29
the vacuum. Positive ions are accelerated towards the mass analyzer and some 30
either (1) collide with a surface in the source (typically the accelerator plate) or 31
5
(2) enter into the mass analyzer through the slit in the electronic lens. The ions 1
that collide with any surface are neutralized and removed by the vacuum. The 2
ions that enter into the mass analyzer are separated by mass to charge ratios. 3
The high degree of fragmentation can be an advantage in compound 4
identification. When more ion fragments are created, the more unique the 5
fragmentation pattern, and the more confirmatory analyte identification will be. 6
On the other hand, the detection of the molecular ion in EI can be difficult, which 7
is often a goal of MS analysis in organic chemistry. 8
9
Electron ionization works by forcing the stream of pure analytes exiting the 10
GC through a beam of high energy electrons in the MS. Electrons are created by 11
heating a metal filament, usually tungsten, to a temperature high enough to expel 12
electrons. Electrons are drawn toward an anode, passing through the stream of 13
analyte molecules. It is important to note that electrons do not actually impact 14
analyte molecules as implied by the name electron impact. The high energy of 15
the electron (70 eV) is actually transferred to an analyte when the electronic 16
transition of the analyte matches the frequency of the electron. The exact 17
electron energy was selected through experimentation. It was found that a 70 eV 18
electron energy source resulted in the most reproducible spectra and in a high 19
degree of fragmentation. This 70 eV condition is now the standard and all 20
computer libraries of fragmentation are based on this energy level. 21
22
The animation below shows a beam of electrons that is generated by a 23
heated filament at the bottom of the figure that is accelerated toward the anode 24
at the top of the figure. When different analytes (in this case butane) exit the GC 25
column (the brown column on the left) and cross through the electron beam, an 26
electron from the sample molecules is removed. Once the molecular ion is 27
formed, they are forced to the right by repulsion from a positively charge 28
accelerator plate on the left (not shown) and drawn toward the negatively 29
charged accelerator plate to the right. Some butane molecules also fragment 30
into smaller ions. The prevalence of this process is underestimated by the 31
6
animation due to space restraints. The molecular ion and fragments would next 1
enter the mass analyzer (shown later). 2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Animation 5.1. Illustration of an Electron Impact Chamber. 21
22
After the energy transfer between the electron beam and the analyte, the 23
energy causes the molecule to become unstable and frequently cleave bonds. 24
The fragmentation patterns are predictable because the types of bond cleavages 25
a molecule undergoes is related to its structure (Chapter 6). The ionization rate 26
is predicted to be between one in a thousand to one in a million of the molecules 27
entering the ionization chamber. This level of successful ionization should be 28
noted since MS detection limits are approximately one part per million and below 29
(injected analyte concentration). In early systems, the instrument only ionized 30
and detected approximately one millionth of the number of molecules that were 31
7
injected; today this has been improved to about one in a thousand or more. Two 1
examples of EI spectra are shown in Figures 5.1 and 5.2; note the extensive 2
fragmentation of each analyte. 3
4
5
Figure 5.1. Fragmentation of Cyclohexanol by EI. 6
7
8
1
Figure 5.2. Fragmentation of Decanoic Acid Methyl Ester by EI. 2
3
4
5.2.1.2b Chemical Ionization (CI): Today, most mass spectrometers can 5
perform both electron ionization and chemical ionization, with different 6
interchangeable ionization units. The CI unit is less open to diffusion of the 7
reagent gas in order to contain the reagent gas longer and promote chemical 8
ionization. Several reagent gases are used including methane, propane, 9
isobutane, and ammonia, with the most common being methane. CI is referred 10
to as a soft ionization technique since less energy is transferred to the original 11
analyte molecule, and hence, less fragmentation occurs. In fact, one of the main 12
purposes of using CI is to observe the molecular ion, represented by M
.+
or M
.-
, 13
or a close adduct of it, such as MH
+
, MH
+2
, or M plus the chemical ion (i.e. 14
M+CH
3
with methane as the reagent gas or M+NH
3
with ammonia as the reagent 15
gas). Notice again that neutral, negative, and positive fragments are produced 16
but only the positive fragments are of use in positive CI detection, while negative 17
ion fragments are detected in negative CI mode. 18
9
1
This section will limit its discussion to CI and methane, the most common 2
reagent gas. Methane enters the ionization chamber at about 1000 times the 3
concentration of the analyte molecules. While the electron beam in EI is usually 4
set at 70eV, in CI lower energy levels are used near the range of 20 to 40 eV. 5
This energy level produces electrons that react with methane to form CH
4
!+
, 6
CH
3
+
, and CH
2
!+
. These ions rapidly react with unionized methane in the 7
following manner: 8
9

!
CH
4
+
+ CH
4
" CH
5
+
+ CH
3
Rxn 5.1
CH
3
+
+ CH
4
" C
2
H
5
+
+ H
2
10
11
The CH
5
+
and C
2
H
5
+
ions collide with the analytes (represented by M) and 12
form MH
+
and (M-1)
+
by proton and hydride transfer 13
14

!
CH
5
+
+ M " MH
+
+ CH
4
proton transfer
C
2
H
5
+
+ M " MH
+
+ C
2
H
4
proton transfer Rxn 5.2
C
2
H
5
+
+ M " (M- 1)
+
+ C
2
H
6
hydride transfer
15
16
Note that several types of ions can occur, (M+1)
+
or MH
+
from proton transfer, 17
(M-1)
+
from hydride transfer, and M+CH
3
+
and even M+C
2
H
5
+
from additions. By 18
inspecting the mass spectrum for this pattern, the molecular mass of the analyte 19
can be deduced. Similarly, if other reagent gases are used, such as propane, 20
isobutene, and ammonia, similar proton and hydride transfer and adduct 21
formations can occur. The usual goal of CI is to obtain a molecule weight for the 22
molecular ion that would usually not be present in an EI spectra. 23
24
A relatively simple illustration of a CI chamber and its reactions is shown 25
in the animation below. This animation is similar to the EI animation, but the 26
continuous addition of a reagent gas, methane, causes the gas to be ionized by 27
the beam of electrons. Subsequently, the ionized methane reacts with analytes 28
10
exiting the GC column. Methane is preferentially ionized by the beam of 1
electrons due to its significantly higher concentration as compared to analytes 2
from the GC. Positively charged fragments are drawn into the focusing lens and 3
mass analyzer by a positively charged repeller plate (not shown) and the 4
negatively charged accelerator plate. 5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Animation 5.2. Illustration of a CI Chamber and Reagent Gas-Analyte Reactions. 23
24
Chemical ionization is most commonly used to create positive ions, but 25
some analytes, such as those containing acidic groups or electronegative 26
elements (i.e. chlorinated hydrocarbons) will also produce negative ions that can 27
be detected by reversing the polarity on the accelerator and detector systems. 28
Some of these analytes produce superior detection limits with CI as opposed to 29
EI, while others only give increased sensitivity (slope of the response to 30
concentration line). Negative ions are produced by the capture of thermal 31
11
electrons (relatively slower electrons with less energy than those common in the 1
electron beam) by the analyte molecule. Thermal electrons are present from the 2
low energy end of the distribution of electrons produced by the lower-energy CI 3
source (~20 eV as opposed to 70 eV in EI). These low energy electrons arise 4
mostly from the chemical ionization process but also from analyte/electron 5
collisions. Analyte molecules react with thermal electrons in the following 6
manner, where R-R is the unreacted analyte molecule and R represents an 7
organic group. 8
9

!
R- R + e
-
" R- R
-
(by associative resonance capture) Rxn 5.3
R- R + e
-
" R + R'
-
(by dissociative resonance capture)
R- R' + e
-
" R
+
+ R'
-
+ e
-
(by ion pair production)
10
11
The identification of negative ion fragmentation patterns of analytes can 12
be used in the same manner as in EI or positive ion CI. But note that extensive 13
fragmentation libraries exist only for 70eV electron ionization (EI). Many analysts 14
create their own reference libraries with the analysis of reference materials that 15
will later be used for the identification of unknown analytes extracted from 16
samples. 17
18
Figures 5.3 and 5.4 contain CI spectra for the same compounds analyzed 19
by EI in Figure 5.1 and 5.2, respectively. Note the obvious lack of fragmentation 20
with the CI source and the presence of molecular ions in the CI spectra. 21
22
12
1
Figure 5.3. Fragmentation of Cyclohexanol by CI. 2
3
4
Figure 5.4. Fragmentation of Decanoic Acid Methyl Ester by CI. 5
6
13
To summarize, for GC-MS systems, individual analytes exit the GC 1
column, are ionized, and fragmented using electron or chemical impact 2
(ionization). Since the detector in a MS is universal (responds to any positively 3
charged ion) it is necessary to separate the molecular ion and its fragments by 4
their mass or mass to charge ratio. This process is completed in a mass 5
analyzer, which is explained in the section below. But first, some mass analyzers 6
require the beam of ion fragments to be focused and all require the ion fragments 7
to be accelerated in a linear direction. 8
9
5.2.1.3 Repulsion and Accelerator Plates, Slits, and Electronic Focusing Lens: 10
11
Ions, regardless of the way they are generated, need to be accelerated 12
into the mass filter/analyzer in order to separate ions of different masses. Since 13
the majority of the ionization sources produce positively charge species, the most 14
common way of accelerating ions is to place a positively charged plate on the 15
upstream side of the system. This plate repels the cations toward the mass 16
filter/analyzer. Most systems require ions to have a minimum velocity, so 17
negatively charged plates are placed on the downstream side of the instrument, 18
just prior to the mass filter, to accelerate the ion in that direction (shown earlier in 19
the EI and CI animations). The accelerator plates also act as slits since a 20
relatively small hole is present in the middle of the plates that allow some of the 21
ions to pass through the plate/slit and into the mass filter. 22
23
Accelerator plates/slits can also act as gates to the mass filter. This is 24
accomplished by placing a positive charge on the slit that will repel the entry of 25
an ion fragment or packet of ions to the system. Gates are used to hold up the 26
entry of new ions to the mass filter until all of the ions have passed through to the 27
detector. After this, the polarity on the gate is returned to negative and a new set 28
of ion fragments is allowed to enter the mass filter. This type of gating system is 29
important in the time-of-flight mass filters discussed in Section 5.5.4. 30
31
14
Some systems, especially the quadrupole mass filter require the stream of 1
ions to be focused into a narrow point in order to allow successful mass to 2
charge separation. One such electrical lens is the Einzel lens that is analogous 3
to a focusing lens in an optical spectrophotometer. Figure 5.5 illustrates how an 4
Einzel lens works. Six plates are in parallel, three on each side, and are exposed 5
to the potentials shown below. These potentials set up a set of electrical field 6
lines that act to bend the ions near the outside of the plates toward the center. 7
Ions are focused to a small point for entry into the mass filter. The series of 8
lenses stretch the length of a given beam of ions since ions on the outside (near 9
the plates) have to travel a longer distance to reach the focal point. 10
11
12
Figure 5.5. An Einzel Lens (Electronically Focusing Lens). 13
14
The Einzel lens above is shown and explained as six horizontal plates. In 15
practice, Einzel lens are vertical plates with a hole in each plate. Thus, the 16
applied electrical potential creates three-dimensional field lines that focuses the 17
ion beam to a point where the entrance slit/hole to the next component is located. 18
19
Electrostatic, magnetic, and time-of-flight instruments have only repulsion 20
and accelerator plates. In addition to these plates, quadrupole instruments have 21
a focusing lens to help introduce the ions towards the center of the mass 22
filter/analyzer. 23
24
5.3 The Introduction of Samples from HPLC 25
26
15
At this point it is noteworthy to recall the differences between GC and LC. 1
Chapter 2 defined GC as a technique applicable to relatively volatile, thermally 2
stable compounds. These restrictions greatly limited the types and number of 3
compounds that could be analyzed by GC, and GC-MS. LC, discussed in 4
Chapter 3, uses a mobile phase in the analysis of many of the compounds 5
analyzed by GC, and also can be used to analyze the plethora of biomolecules 6
that are non-volatile and thermally unstable at even slightly elevated 7
temperatures. While the conditions used in LC greatly extends the applications 8
of chromatography, it has historically suffered difficulties with mass spectrometry 9
interfaces. Most of the various forms of LC, especially HPLC types discussed in 10
Chapter 3, can be interfaced with MS today. 11
12
The largest difficulties in interfacing LC with MS is the removal of the 13
mobile phase solvent prior to introduction to the MS mass analyzer and the 14
transfer and ionization of nonvolatile analyte molecules into the gas phase. The 15
first attempt at an LC-MS interface was to place the effluent droplets from the LC 16
onto a supposed chemical resistant conveyer belt that transported the liquid into 17
the MS ionization chamber. The conveyer belt was then cleaned and returned to 18
the HPLC effluent for more sample. However, these early attempts resulted in 19
inefficient removal of the analytes from the conveyer belt and analyte residue 20
being left on and released from the belt during subsequent MS runs. This 21
problem was significantly compounded with 4.5 mm diameter HPLC columns 22
with flow rates in the range of 1 mL/min. The later use of 300 to 75 mm long 23
capillary columns improved flow rate problems. The invention of Electro Spray 24
Ionization (ESI) solved all of the major problems associated with sample 25
introduction to MS. ESI was first conceived in the 1960s by Malcolm Dole at 26
Northwestern University, but it was not put into practice until the early 1980s by 27
J ohn B. Fenn of Yale University (and resulted in his Noble prize in 2002). Its 28
common use today has been one of the most important advances in HPLC and 29
today allows routine identification of biological macromolecules. 30
31
16
5.3.1 Electro-Spray Ionization (ESI) Sample Introduction 1
2
Today, the most common form of LS-MS interface is the ESI sample 3
introduction system. An overview of this system is shown in Figure 5.6. 4
Samples can be introduced via a syringe or an HPLC system (convention or 5
capillary column type). A restriction in the syringe needle or HPLC column 6
causes the solvent containing the analytes to form droplets. An electrical 7
potential, discussed in the next paragraph, is placed between the sample inlet 8
and the first cone. This cone separates the sample introduction from the vacuum 9
chamber in the MS. For high flow HPLC applications N
2
gas is used to 10
evaporate the solvent or mobile phase and de-solvate the analyte molecules. 11
This is usually unnecessary for capillary columns or nano- applications. After 12
desolvation and charge formation occur, as discussed below, the charged 13
molecules enter a slightly heated transfer capillary tube and pass through two 14
more cones that are used to control the vacuum. Finally, the positively charged 15
ions enter a mass analyzer such as the quadrupole shown in Figure 5.6. 16
17
18
19
20
Figure 5.6 Overview of an Electro Spray Ionization (LC-MS) Interface. 21
22
The heart of ESI is the desolvation and charge formation shown in Figure 23
5.7. Ionization in ESI is referred to as a soft ionization and is really not 24
17
ionization but charge formation since no real ionization source is present. 1
Charge formation occurs by evaporating the solvent by passing a dry gas counter 2
current to the movement of droplets. While at the same time the droplets are 3
passed along a charged field (from 2.5 to 4 kV) between the tip of the sample 4
introduction point and the first cone. Charge formation occurs by one of two 5
proposed mechanisms, (1) Ion Evaporation Model where the droplet reaches a 6
certain radius such that the field strength at the surface of the droplet becomes 7
large enough to assist the field desorption of solvated ions and (2) Charged 8
Residue Model where electrospray droplets undergo evaporation and fission 9
cycles, resulting in gas-phase ions that form after the remaining solvent 10
molecules evaporate. 11
12
The Charged Residue Model is the most accepted theory and is explained 13
in the following. As the droplets pass from left to right, desolvation occurs in the 14
present of the dry N
2
gas. At the same time, the charged field results in the 15
collection of a positive charge on the droplet. As this process continues, from left 16
to right, the droplet shrinks until it reaches a point where the surface tension can 17
no longer sustain the charge accumulation, this point is referred to as the 18
Rayleigh limit. Above the Rayleigh limit, Rayleigh fission (also known as 19
Coulombic explosion) occurs and the droplet is ripped apart forming smaller 20
charged droplets containing the analyte molecules. This process continues until 21
desolvation is complete and the charge is transferred to the ionized and now 22
gaseous analyte molecule. The resulting charged molecules can be singly or 23
multiply charged (refer to Figure 5.7). The positively charged ions enter the 24
mass analyzer. Simple molecules result in a single mass to charge ion while 25
complex molecules result in a Gaussian distribution of mass to charge ions 26
yielding a single molecule molecular mass for identification purposes. As noted 27
above, the ionization process is considered to be a soft ionization, thus, if 28
structural identification is required the parent ion is usually analyzed by tandem 29
MS where it is fragmented into smaller fragments for identification. Nano-spray 30
versions of this process have recently become available. 31
18
1
2
Figure 5.7 Charge Formation in ESI. 3
4
5.3.2 (NEW) Atmospheric Pressure Chemical Ionization 5
6
5.3.2 Matrix Assisted Laser Desorption/Ionization, MALDI: The MALDI 7
technique has revolutionized the analysis of large molecular weight non-volatile 8
compounds, especially synthetic polymers and biopolymers with molecular 9
weights up to 300 000 Daltons. Unlike the Field Desorption technique that 10
desorb and ionize pure analyte from a probe, MALDI volatilizes a mixture of a 11
matrix and analyte in order to transport the non-volatile analyte into a vapor 12
phase. 13
14
The MALDI technique is completed in two steps. First, a solution of 15
solvent, analyte, and matrix compound are thoroughly mixed and placed on a 16
disk to dry. As the solvent evaporates crystals of matrix containing evenly 17
dispersed analyte molecules are formed. For the second step, the coated disk is 18
placed in the vacuum chamber of the MS. Then the disk is repeatedly pulsed 19
with a laser in the UV or visible spectrum depending on the matrix (Table 1.2). 20
During each laser pulse, the matrix molecules are rapidly volatilized 21
19
(sublimated/abulated) and carry the individual analyte molecules into a low 1
pressure plasma. The wavelength of the laser is selected to heat and volatize 2
the matrix and to avoid significant heat or degrade the analyte molecules. 3
Analyte molecules are mostly ionized in the vapor phase by photoionization, 4
excited-state proton transfer, ion-molecule reactions, desorption of preformed 5
ions and most commonly by gas-phase proton transfer in the expanding plume 6
by photoionized matrix molecules. 7
8
After the analyte molecules are ionized (to cations) they are drawn toward 9
the negative accelerator plate and into the mass filter. A time-of-flight mass filter 10
is always used because of its rapid scanning abilities and large mass range. The 11
introduction of ions into the flight tube is controlled so that all ions reach the 12
detector before the next group enters into the TOF tube. This requires carefully 13
spacing the laser pulses and electric gates (discussed in Section 5.5.4). The 14
spectrum of the analysis is considerably clean since only pure analyte is 15
introduced into the MS and essentially no fragmentation occurs (matrix 16
molecules/ions can be ignored by the mass filter due to their relatively low mass). 17
Ionized analytes can acquire +1, +2, and +3 charges and multiple molecules can 18
form dimer and trimer peaks (combined fragments of two or three molecular 19
ions), so the confirmational molecular weights can easily be determined. A very 20
simple illustration of a MALDI-Time-of-Flight MS (the most common combination) 21
is shown in Animation 5.3. 22
23
Table 5.1. Frequently Used Matrix Compounds 24
25
Matrix Compound Active Wavelength (nm)
Nicotinic acid 220-290
Benzoic acid derivatives such as
Vanillic acid
266
Pyrazine-carboxylic acid 266
3-Aminopyrazine-2-carboxlic acid 337
20
Cinnamic acid derivatives such as
Caffeic acid
266-355
3-Nitrobenzylalcohol 266
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Animation 5.3. Illustration of a MALDI-TOF MS System. 22
23
5.3.3 Proton Transfer Reaction Ionization (PTR): PTR is a relatively 24
recent addition to mass spectrometry (1995) that was originally developed for GC 25
and LC, there is not reason that it can not be used for CE. It was developed at 26
the Institut fr Ionenphysik at the Leopold-Franzens University in Innsbruck, 27
Austria by Hansel et al. (1995). As shown in Figure 5.8, the PTR consists of a 28
21
reaction chamber where water vapor is ionized to gas phase ions by hollow 1
cathode discharge via the following reactions 2
3

!
e
-
+ H
2
O " H
2
O
+
+ 2 e
-
Rxn 5.4
e
-
+ H
2
O " H
2
+
+ O + 2 e
-
e
-
+ H
2
O " H
+
+ OH + 2 e
-
e
-
+ H
2
O " O+ + H
2
+ 2 e
-
4
5
These products undergo ion-water vapor reactions in a short drift tube to form 6
7

!
H
2
+
+ H
2
O " H
2
O
+
+ H
2
Rxn 5.5
H+ + H
2
O " H
2
O
+
+ H
O+ + H
2
O " H
2
O
+
+ O
H
2
O
+
+ H
2
O " H
3
O
+
+ OH
8
9
The hydronium ion (H
3
O
+
) is end product and the primary reacting ion that 10
ionizes organic analytes in the reaction drift tube via the reaction 11
12

!
H
3
O
+
+ R " RH
+
+ H
2
O Rxn 5.6 13
14
Unlike in TOF or ion mobility MS, reaction ions are not subjected to a electrical 15
potential in the drift tube but are moved through the system by placing a low 16
pressure vacuum pump at the interface of the PRT drift tube and the inlet to the 17
mass filter (refer to Figure 5.8). Analyte cations created in the drift tube enter a 18
mass filter where they are separated by the operating parameters of each mass 19
filter and are detected with an electron mulitiplier. 20
21
22
1
2
Figure 5.8 Illustration of a Proton Transfer Reaction MS System. Reprinted 3
with permission from Ionicon Analytik Gesellschaft, Innsbruck, Austria. 4
5
A PTR-MS is illustrated via the link in Animation 5.4. 6
7
http://www.uibk.ac.at/ionen-angewandte-physik/umwelt/research/pics/animation.gif 8
9
Animation 5.4. Illustration of a Proton Transfer ReactionMass Spectrometer. 10
11
Advantages of the PTR-MS include (1) low fragmentation with allows 12
improved detection limits due to the formation of more molecular ions, (2) direct 13
sampling of atmospheric gases (no sample preparation), (3) real time 14
measurements, (4) high mobility due to the lack of gas cylinders, relative ease of 15
operation only requiring electrical power and distilled water, and part per billion 16
detection limits. 17
23
5.3.4 Fast Ion Bombardment (FAB): Another technique for ionizing large 1
bio-molecules (up to and greater then 10,000 Daltons) is to bombard them with 2
ions of argon or xenon; this is also referred to as a liquid secondary ion source. 3
First, analytes are embedded in a matrix such as glycerol, thioglycerol, m- 4
nitrobenzyl alcohol, crown ethers, sulfolane, 20-nitrophenyloctyl ether, 5
diethanolamine or triethanolamine. An electronic impact (EI) source similar to 6
that described in the GC ionization section is used to ionize Ar or Xe gas at a 7
pressure of 10
-5
torr. Ar and Xe ions are accelerated towards the matrix 8
containing the analytes and their impact sputters off positive and negative 9
analytes ions (mostly molecular ions) that enter a mass spectrometer for mass 10
determination. 11
12
13
5.4 The Introduction of Samples from a Capillary Electrophoresis System 14
15
Years ago, if you wanted to own a CE-MS system you had to purchase the CE 16
and MS separately and hire the MS manufacturer or vender to interface the two 17
instruments. Recently (~2008) you are now able to purchase off-the-shelf 18
interfaced instruments from chromatography vendors. CE-MS interfaces are 19
designed and operate in much the same way as the HPLC-MS interface, with two 20
exceptions. While HPLC columns can be composed of metal that readily 21
conduct the electrical potential to ionize the analytes, the CE columns are only 22
composed of fused silica. As a result the effluent of the CE column must be 23
coated with a conducting metal sheath. Also, ss you will recall from Chapter 4 on 24
CE, minimal solvent flow results in CE, only from the dragging of solvent by the 25
electrophoretic mobility of the buffer ions. Thus, CE is almost ideal for MS 26
interfaces and is far superior to HPLC interfacing since very little solvent must be 27
removed prior to entry into the MS vacuum system. Other than these two 28
differences, CE-MS operates like HPLC-MS. Solvent droplets, containing 29
analytes, are created at the end of the fused silica column, and are charged by 30
the electrical potential placed between the metal sheath and the metal cone at 31
24
the entry to the MS system (Figure 5.9). Solvent is evaporated with a drying gas 1
that flows counter current to the movement of the solvent droplets. Charge 2
transfer occurs through Coulombic explosion and the de-solvated and ionized 3
anionic or cations (depending on the potential) are accelerated through the MS 4
interface cone. CE-MS has finally reached a level of maturity and dependability 5
that promises significant advances in many areas of analytical separation and 6
quantification, especially protein studies. 7
8
Figure 5.9. CE-MS Interface. 9
10
5.5 Common Mass Filters (Mass Analyzers) 11
12
Mass analyzers separate the molecular ion and its fragments by ion 13
velocity, mass, or mass to charge ratio. A number of mass filters/analyzers are 14
available for GC, LC and CE interfaces, but not all are commercially available. 15
These can be used individually or coupled in a series of mass analyzers to 16
improve mass resolution and provide more conclusive analyte identification. This 17
text will only discuss the most common ones. 18
19
The measure of power of a mass analyzer is resolution, the ratio of the 20
average mass (m) of the two adjacent peaks being separated to the mass 21
difference (!m) of the adjacent peaks, represented by 22
23
Rs =m/!m 24
25
1
Resolution (R
S
) is achieved when the midpoint between two adjacent peaks is 2
within 10 percent of the baseline just before and after the peaks of interest (the 3
valley between the two peaks). Resolution requirements can range from high 4
resolution instruments that may require discrimination of a few ten thousands 5
(1/10 000) of a gram molecular weight (0.0001) to low resolution instruments that 6
only require unit resolution (28 versus 29 Daltons). Resolution values for 7
commonly available instruments can range from 500 to 500 000. 8
9
Before introducing the various types of mass analyzers, remember our 10
current location of the mass analyzer in the overall MS system. The analyte has 11
been ionized, underwent fragmentation, been accelerated, and in some cases 12
focused to a focal point with a velocity towards the mass analyzer. Now the 13
packet of ion fragments needs to be separated based on their momentum, kinetic 14
energy, or mass-to-charge ratio (m/z). Often the terms mass filter and mass 15
analyzer are used interchangeable, as is done in this text. But, first a 16
controversy in the literature needed to be addressed with respect to how a mass 17
filter actually separates ion fragments. 18
19
Some resources state that all mass analyzers separate ions with respect 20
to their mass to charge ratio while others are more specific and contend that only 21
quadrupoles separate ions by mass to charge ratios. The disagreement in 22
textbooks lies in what components of the MS are being discussed. If one is 23
discussing the affect of the accelerator plates and the mass filter, then all mass 24
filters separate based on mass to charge ratios. This occurs because the charge 25
of an ion will be a factor that determines the velocity a particle of a given mass 26
has after interacting with the accelerator plate in the electronic, magnetic sector, 27
and time of flight mass analyzers. But after the ion has been accelerated, a 28
magnetic section mass filter actually separates different ions based momentums 29
and kinetic energies while the time of flight instrument separates different ions 30
based on ion velocities (arrival times at the detector after traveling a fixed length). 31
26
In the other case, no matter what the momentum or velocity of an ion, the 1
quadrupole mass analyzer separates different ions based solely on mass to 2
charge ratios (or the ability of the ion to establish a stable oscillation in an 3
oscillating electrical field). These differences may seem semantic but some MS 4
users insist on their clarification. For the discussions below, in most cases, mass 5
to charge will be used for all mass analyzers. 6
7
5.5.1 Magnetic sector mass filter: It has been known for some time that 8
the trajectory of a point charge, in our case a positively charged molecular ion or 9
fragment, can be altered by an electrical or magnetic field. Thus, the first MS 10
systems employed permanent magnets or electromagnets to bend the packets of 11
ions in a semi-circular path and separated ions based on their momentum and 12
kinetic energy. Common angles of deflection are 60, 90, and 180 degrees. The 13
change in trajectory of the ions is caused by the external force of the magnetic 14
field. The magnitude of the centripetal force, which is directly related to the ions 15
velocity, resists the magnetic fields force. Since each mass to charge ratio has a 16
distinct kinetic energy, a given magnetic field strength will separate individual 17
mass to charge ratios through space. A slit is placed in front of the detector to 18
aid in the selection of a single mass to charge ratio at a time. 19
20
A relatively simple mathematical description will allow for a better 21
understanding of the magnetic field and the ions centripetal force. First, it is 22
necessary to compute the kinetic energy (KE) of an ion with mass m possessing 23
a charge z as it moves through the accelerator plates. This relationship can be 24
described by 25
26

!
KE = 1/2 mv
2
= zeV Eqn 5.1 27
28
where e is the charge of an electron (1.60 x 10
-19
C), v is the ion velocity, and V 29
is the voltage between the two accelerator plates (shown in the Animation 1.5 30
below). Fortunately in EI and CI, most ions have a charge of +1. As a result, an 31
27
ions kinetic energy will be inversely proportional to its mass. The two forces that 1
determine the ions path, the magnetic force (F
M
) and the centripetal force (F
C
), 2
are described by 3
4

!
F
M
= B z ev Eqn 5.2 5
6
and 7
8

!
F
C
= (mv
2
)/r Eqn 5.3 9
10
where B is the magnetic field strength and r is the radius of curvature of the 11
magnetic path. In order for an ion of particular mass and charge to make it to the 12
detector, the forces F
M
and F
C
must be equal. This obtains 13
14

!
B z e v = (mv
2
)/r Eqn 5.4 15
16
which upon rearrangement yields 17
18

!
v = (V z e r) / m Eqn 5.5 19
20
Substituting this last equation into our first KE equation yields 21
22

!
m/z = (B
2
r
2
e) / 2V Eqn 5.6 23
24
Since e (the charge of an electron) is constant and r (the radius of curvature) is 25
not altered during the run, altering the magnetic field (B) or the voltage between 26
the accelerator plates (V) will vary the mass to charge ratio that can pass through 27
the slit and reach the detector. By holding one constant and varying the other 28
throughout the range of m/z values, the various mass to charge ratios can be 29
separated. One option is to vary the magnetic field strength while keeping the 30
voltage on the accelerator plates constant. 31
28
1
However, it is difficult to quickly vary the magnetic field strength. The 2
resulting slow scan rate is especially problematic with capillary column GCs since 3
the peak width is narrow. Using a magnetic sector instrument could complicate 4
identification of a compound if two or more peaks emerge from the GC during a 5
single scan, especially in the relatively fast elution of peaks from a capillary 6
column GC. Generally, several complete mass to charge scans are desired for 7
accurate analyte identification. This can be overcome in modern magnetic sector 8
instruments by rapidly sweeping the voltage between the accelerator plates, in 9
order to impart different momentums on the ion fragments, as opposed to 10
sweeping the field strength. Due to the operational advantages of this technique, 11
most electromagnets hold the magnetic field strength (B) and vary the voltage (V) 12
on the accelerator plates. 13
14
The magnetic sector mass filter is illustrated in Animation 5.5 below. 15
Although B and r are normally held constant, this modern design is difficult to 16
illustrate, so we will illustrate a magnetic sector MS where B, the magnetic field, 17
is varied to select for different ions. As a particular peak (compound) enters the 18
MS from a GC, it is ionized/fragmented by an EI in the animation. The ions are 19
then uniformly accelerated by the constant voltage between the two accelerator 20
plates/slits on the left side of the figure. As the different ions travel through the 21
electromagnet, the magnetic field is varied to select for different m/z ratios. Ions 22
with the same momentum or kinetic energy (and therefore mass) pass through 23
the exit slit together and are measured by the detector, followed by the next ion, 24
and so on. 25
26
27
28
29
30
31
29
1
2
3
4
5
Animation 5.5. Illustration of a Magnetic Sector MS. 6
7
While magnetic sector mass filters were once the only tool to create a 8
mass spectrum, they are becoming less common today. This is due to the size 9
of the instrument and its weight. As a result, many magnetic sector instruments 10
have been replaced by quadrupole systems that are much smaller, lighter, and 11
able to perform extremely fast scans. Magnetic sector instruments are still used 12
in cases where extremely high resolution is required such as double-focusing 13
instruments (Section 5.5.6). 14
15
5.5.2 Quadrupole mass filter: Quadrupole mass filters have become the 16
most common type of MS used today due to their relatively small size, light 17
weight, low cost, and rapid scan times (less than 100 ms). This type of mass 18
filter is most commonly used in GC applications and to some extent in LC 19
systems because they are able to operate at a relatively high pressure (5 x 10
-5
20
torr). The quadrupole has also gained widespread use in tandem MS 21
applications (a series of MS analyzers). 22
23
Despite the fact that quadrupoles produce the majority of mass spectra 24
today as mentioned earlier, they are not true mass spectrometers. Actual mass 25
spectrometers produce a distribution of ions either through time (time of flight 26
mass spectrometer) or space (magnetic sector mass spectrometer). The 27
quadrupoles mass resolving properties are instead a result of the ions 28
stability/trajectory within the oscillating electrical field. 29
30
30
A quadrupole system consists of four rods that are arranged an equal 1
distance from each other in a parallel manner. Paul and Steinwegen theorized in 2
1953 that hyperbolic cross-sections were necessary. In practice, it has been 3
found that circular cross sections are both effective and easier to manufacture. 4
Each rod is less than a cm in diameter and usually less then 15 cm long. Ions 5
are accelerated by a negative voltage plate before they enter the quadrupole and 6
travel down the center of the rods (in the z direction). The ions trajectory in the z 7
direction is not altered by the quadrupoles electric field. 8
9
The various ions are separated by applying a time independent dc 10
potential as well as a time dependent ac potential. The four rods are divided up 11
into pairs where the diagonal rods have an identical potential. The positive dc 12
potential is applied to the rods in the X-Z plane and the negative dc potential is 13
applied to the rods in the Y-Z plane. The subsequent ac potential is applied to 14
both pairs of rods but the potential on one pair is the opposite sign of the other, 15
and is commonly referred to as being 180 out of phase (Figure 5.9). 16
17
Mathematically the potentials that ions are subjected to are described by 18
the following equations: 19
20

!
"
X- Z
= +(U + V cos #t)
and Eqn 5.7
"
Y- Z
= - (U + V cos #t)
21
22
where ! is the potential applied to the X-Z and Y-Z rods respectively, " is the 23
angular frequency (in rad/s) and is equal to 2#v where v is the radio frequency of 24
the field, U is the dc potential and V is the zero-to-peak amplitude of the radio 25
frequency voltage (ac potential). The positive and negative signs in the two 26
equations reflect the change in polarity of the opposing rods (electrodes). The 27
values of U range from 500 to 2000 volts and V in the above equation ranges 28
from 0 to 3000 volts. 29
31
1
2
32
1
2
3
Figure 5.10 AC and DC Potentials in the Quadrupole MS. 4
5
33
To understand the function of each pair, consider the rods in the X-Z plane 1
in isolation. For now, imagine that only an ac potential is applied to the rods. 2
Half the time when the potential was positive, ions (cations) would be repelled by 3
the rods charge and would consequently move towards the center of the rods. 4
Likewise, when the potential was negative, ions would accelerate towards the 5
rods in response to an attractive force. If during the negative ac potential, an ion 6
comes into contact with the rod, it is neutralized and is removed by the vacuum. 7
The factors that influence whether or not a particle strikes the rod during the 8
negative cycle include the magnitude of the potential (its amplitude), the duration 9
of time the ions are accelerated towards the rod (the frequency of the ac 10
potential), the mass of the particular ion, the charge of the ion, and its position 11
within the quadrupole. 12
13
Now imagine that a positive dc potential (at a fraction of the magnitude of 14
the ac potential) is applied to the rod in the X-Z plane. This positive dc potential 15
alone would focus all of the ions towards the center of the pair of rods. When the 16
ac and dc potentials are applied at the same time to the pair of rods in the X-Z 17
plane, ions of different masses respond differently to the resulting potential. 18
Heavy ions are largely unaffected by the alternating current and as a result 19
respond to the average potential of the rods. This results in heavy ions being 20
focused towards the center of the rods. Light ions, on the other hand, will 21
respond more readily to the alternating ac current. Ions that are sufficiently light 22
will have an unstable trajectory in the X-Z plane and will not reach the detector. 23
Only ions heaver than a selected mass will not be filtered out by the X-Z 24
electrodes. As a result, the X-Z plane electrodes only filter light ions and form a 25
high pass mass filter (Figure 5.11). 26
27
Now look at the other pair of rods and the converse of the ac/dc potentials. 28
The rods in the Y-Z plane have a negative dc voltage and the ac potential is the 29
same magnitude but the oppose sign as the potential applied to the X-Z plane. 30
Heavy ions are still mostly affected by the dc potential, but since it is negative, 31
34
they strike the electrode and are unable to reach the detector. The lighter ions 1
respond to the ac potential and are focused towards the center of the 2
quadrupole. The ac potential can be thought of as correcting the trajectories of 3
the lighter ions, preventing them from striking the electrodes in the Y-Z plane. 4
These electrical parameters result in the construction of a low pass mass filter. 5
6
When both the electrodes are combined into the same system, they are 7
able to selectively allow a single mass to charge ratio to have a stable trajectory 8
through the quadrupole. Altering the magnitude of the ac and dc potential 9
changes the mass to charge ratio that has a stable trajectory resulting in the 10
construction of mass spectra. Different ions possess a stable trajectory at 11
different magnitudes and reach the detector at different times during a sweep of 12
the ac/dc magnitude range. The graph of the combined effect, shown in Figure 13
5.10c, is actually a simplification of the actual stability diagram. 14
15
35
1
2
Figure 5.11 A Conceptual Stability Diagram 3
4
One way to generate an actual stability diagram is to perform a series of 5
experiments where a single mass ion is introduced into the quadrupole. The dc 6
36
and ac voltages are allowed to vary and the stability of the ion is mapped. After 1
performing a great number of experiments the resulting plot would look like 2
Figure 5.12. The shaded area under the curve represents values of ac and dc 3
voltages where the ion has a stable trajectory through the potential and would 4
reach the detector. The white space outside the stability diagram indicates ac 5
and dc voltages where the ion would not reach the detector. 6
7
While any ac and dc voltages that fall inside the stability diagram could be 8
utilized, in practice, quadrupoles keep the ratio of the dc to ac potential constant, 9
while the scan is performed by changing the magnitude of the ac and dc 10
potential. The result of this is illustrated as the mass scan line intersecting the 11
stability diagram in Figure 5.12. The graphs below the stability diagram 12
correspond to specific points along the scan and help to illustrate the ions 13
trajectories in the X-Z and Y-Z plane (Figure 5.12). While the mass to charge 14
ratio of the ion remains constant in each pair of horizontal figures, the magnitude 15
of the applied voltages are changing while their ratio stays constant. As a result, 16
examining points along the mass scan line in Figure 5.12 is equivalent to shifting 17
the position of the high and low pass mass filters with respect to the x axis 18
illustrated in Figure 5.11. Even though the mass is not changing for the stability 19
diagram discussed here, the mass that has a stable trajectory is altered. 20
21
37
1
Figure 5.12 Stability Diagram for a Single Ion Mass. Used with permission from 2
the J ournal of Chemical Education, Vol. 75, No. 8, 1998, p. 1051; copyright 3
1998, Division of Chemical Education, Inc. 4
5
In the above figure, the graph corresponding to point A indicates that the 6
ion is too light to pass through the X-Z plane because of the high magnitude of 7
the ac and dc potentials. As a result, its oscillation is unstable, and it eventually 8
impacts the electrode/rod. The motion of the particle in the Y axis is stable 9
because the combination of the ac potential as well as the negative dc potential 10
38
yields a stable trajectory. This is the graphical representation of the ac potential 1
correcting the trajectory of the light ions in the Y-Z plane. At point B the 2
magnitude of voltages has been altered so the trajectories of the ion in both the 3
X-Z and Y-Z plane are stable and the ion successfully reaches the detector. At 4
point C, the ion has been eliminated by the low mass pass filter. In this case, the 5
ac potential is too low to allow the ion to pass through the detector and it strikes 6
the rod. This is caused by the ions increased response to the negative dc 7
potential in the Y-Z plane. The trajectory in the X-Z axis is stable since the dc 8
potential focusing the ion towards the center of the poles overwhelms the ac 9
potential. 10
11
Until this point, the stability diagram shown above is only applicable to a 12
single mass. If a similar experiment were to be performed using a different 13
mass, the positions of the ac and dc potential on the x and y axes would be 14
altered but the overall shape of the curve would remain the same. Fortunately, 15
there is a less time consuming way to generate the general stability diagram for a 16
quadrupole mass filter using a force balance approach. This derivation requires 17
a complex understanding of differential equations and is beyond the scope of an 18
introductory text, but it can be explained graphically (Figure 5.13). The 19
parameters in the axes are explained below the figure. 20
21
39
1
Figure 5.13 The General Stability Diagram 2
3
While this derivation is particularly complex, the physical interpretation of 4
the result helps explain how a quadrupole is able to perform a scan. The final 5
solution is dependent on six variables, but the simplified two-variable problem is 6
shown above. Utilizing the reduced parameters, a and q, the problem becomes 7
a more manageable two-dimensional problem. While the complete derivation 8
allows researchers to perform scans in multiple ways, we will focus only on the 9
basic mode that makes up the majority of mass spectrometers. For the majority 10
of commercially available mass spectrometers, the magnitude of the ac potential 11
(V) and the dc potential (U) are the only parameters that are altered during run 12
time and allows a sweep of mass to charge ranges. The rest of the parameters 13
that describe K
1
and K
2
are held constant. The values for K
1
and K
2
in the 14
general stability diagram can be attributed to the following equations: 15

!
K
1
=
2e
r
2
"
2
Eqn 5.8 16

!
K
2
=
4e
r
2
"
2
Eqn 5.9 17
The parameters that make up K
1
and K
2
are exactly what we predicted 18
when listing the variables earlier that would affect the point charge. Both K terms 19
depend upon the charge of the ion e, its position within the quadrupole r, and the 20
40
frequency of the ac oscillation !. These parameters can be altered, but for the 1
majority of applications remain constant. The charge of the ion (e) can be 2
assumed to be equivalent to positive one for almost all cases. The distance from 3
the center of the quadrupole (r) is carefully controlled by the manufacturing 4
process and an Einzel lens that focuses the ions into the center of the 5
quadrupole and is also a constant. Also the angular frequency (") of the applied 6
ac waveform can be assumed to be a constant for the purposes of most 7
spectrometers and for this discussion. 8
9
The first important note for the general stability diagram is the relationship 10
between potential and mass. The general stability diagram (Figure 5.13) is 11
illustrated where there is an inverse relationship between the two. Figure 5.12 12
shows the lighter ions (m-1) are higher on the mass scan line and the heavy ions 13
(m+1) are lower on the line. This is why in Figure 5.12 at point A, the molecule 14
was too light for the selected frequencies, and it was too heavy at point C. 15
16
From the general stability diagram, it is also possible to explain how an 17
instruments resolution can be altered. The resolution is improved when the 18
mass scan line intersects the smallest area at the top of the stability diagram 19
(Figure 5.14). The resolution can be improved when the slope of the mass line is 20
increased. The resolution will subsequently increase until the line no longer 21
intersects the stability diagram. While it would be best for the line to intersect at 22
the apex of the stability diagram, this is impractical due to fluctuations in the ac 23
(V) and dc (U) voltage. As a result, the line intersects a little below this point 24
allowing the quadrupole to obtain unit resolution. 25
26
Once the resolution has been determined, the ratio of the ac to dc 27
potential is left unchanged throughout the scan process. Again, to perform a 28
scan, the magnitude of the ac and dc voltages is altered while their ratio stays 29
constant. This places a different mass to charge inside the stability diagram. For 30
example, if the ac and dc voltages are doubled, the mass to charge ratio of the 31
41
selected ion would also be doubled as illustrated in the second part of Figure 1
5.14. By scanning throughout a voltage range, the quadrupole is able to create 2
the majority of mass spectra produced in todays chemical laboratories. But it 3
should be noted that quadrupole mass filters have a upper limit of approximately 4
650 amus. 5
6
42
1
2
3
Figure 5.14 Quadrupole Mass Scan. Used with permission from the J ournal of 4
Chemical Education, Vol. 63, No. 7, 1998, p. 621; copyright 1986, Division of 5
Chemical Education, Inc. 6
7
Now that we have given a detailed description of the factors influencing the 8
movement of a charged particle through the quadrupole, it is advantageous to 9
summarize the entire process as a physicist would do in the form of a force 10
43
balance. This is the origin of the governing equation where the French scientist 1
E. Mathieu balanced the equations for the motion of ionized particles to the 2
potential forces (electrical potentials) encountered in a quadrupole mass 3
analyzer. This six-parameter differential equation, known as the Mathieu 4
equation, is represented by 5
6

!
d
2
u
d"
2
+ [a
u
+ 2quCos2" ]u = 0 Eqn 5.10
where
a =
4eU
#
2
r
0
2
m
and q =
2eV
#
2
r
0
2
m
7
8
and where u is either the x or y directional coordinate, " is the redefining of time 9
(t/2), e is the charge of the ion, U is the magnitude of the dc potential, ! is the 10
angular frequency (2pf) of the applied ac waveform, r
o
is the distance from the 11
center axis (the z axis) to the surface of any electrode (rod), m is the mass of the 12
ion, and V is the magnitude of the applied ac or radio frequency waveform. By 13
using the reduced terms, a and q, the six-parameter equation (e, w, r
o
, m, U, and 14
V) can be simplified to a two-parameter equation (involving a and q). Thus, when 15
the two opposing forces are balanced, the movement of a charged particle in an 16
electrical field, the particle will pass through the quadrupole and strike the 17
detector. 18
19
20
View Animation 5.6 for an illustration of how the trajectory of ions of different 21
masses are changed during a mass scan. 22
23
24
25
26
44
1
3
12
13
Animation 5.6. Illustration of a Quadrupole Mass Filter. 14
15
5.5.3 Quadrupole ion trap mass filter: While the operation of the ion trap 16
was characterized shortly after the linear quadrupole in 1960 by Paul and 17
Steinwedel, its application in the chemical laboratory was severely limited. This 18
was due to difficulties associated with manufacturing a circular electrode and 19
performance problems. These performance problems were overcome when a 20
group at Finnigan MAT lead by Stafford discovered two breakthroughs that lead 21
to the production of a commercially available ion trap mass filter. The first ion 22
trap developed used a mode of operation where a single mass could be stored in 23
the trap when previously all of the ions had to be stored. Their next important 24
discovery was the ability for 1 mtorr of helium gas to improve the instruments 25
resolution. The helium molecules collisions with the ions reduced their kinetic 26
energy and subsequently focused them towards the center of the trap. 27
28
After these initial hurdles were cleared, many new techniques were 29
developed for a diverse set of applications especially in biochemistry. This is a 30
45
result of its comparative advantage over the quadrupole when analyzing high 1
molecular mass compounds (up to several thousand m/z units) to unit resolution 2
in commonly encountered instruments. The ion trap is also an extremely 3
sensitive instrument which allows a molecular weight to be determined with a 4
small number of molecules. The ion trap is also the only mass filter that can 5
contain ions that need to be analyzed for any significant duration of time. This 6
allows the instrument to be particularly useful in monitoring the kinetics of a given 7
reaction. The most powerful application of the ion trap is its ability to be used in 8
tandem mass spectrometry (section 5.5.7). 9
10
The ion trap is made up of a single ring electrode that is placed in the X-Y 11
plane between two end cap electrodes (Figure 5.15). Both an ac and dc voltage 12
can be applied to the ring electrode while only an ac voltage can be applied to 13
the end cap electrodes. The two end cap electrodes and the ring electrode 14
ideally have a hyperbolic shape to establish an ideal field however in practice, 15
non-ideal fields can operate effectively. While the ion trap is applying force to the 16
charged ions in three directions, the problem can be simplified into a two- 17
dimensional problem. Since the ring is symmetrical, the force in any direction is 18
always the same. As a result of this symmetry, movement of the molecules can 19
be expressed in terms of r and z where
2 2
y x r + = where x and y are 20
coordinates. For commercially available instruments, r
0
(the distance from the 21
center of the trap to the ring electrode is either 1.00 or 0.707 cm. 22
23
46
1
2
Figure 5.15 A Cross Section of the Ion Trap 3
4
After the sample molecules have been ionized by the source, they enter 5
into the ion trap through an electric gate located on a single end cap electrode. 6
This gate functions in the same fashion as the one that is utilized in time of flight 7
mass spectrometry (Section 5.5.4). The gates purpose is to prevent a large 8
number of molecules from entering into the trap. If too many sample molecules 9
enter into the trap, the interaction with other molecules becomes significant 10
resulting in space-charge effects, a distortion of the electrical field that minimizes 11
the ion traps performance. Once the ions enter the trap, their collisions with the 12
helium gas focus the ions towards the center of the trap. An ac frequency is also 13
applied to the ring electrode to assist in focusing the ions towards the center of 14
the trap. 15
16
In the ion trap, the ring electrode oscillates with a very high frequency 17
(typically 1.1 MHz) while both the end cap electrodes are kept at a ground 18
47
potential (U equals 0 Volts). This frequency causes the ion to oscillate in both 1
the r and z direction (Figure 5.16). The oscillation in the r direction is an 2
expected response to the force generated by the ring electrode. The oscillation 3
in the z direction, on the other hand, may seem counter intuitive. This is a 4
response to both the grounded end cap electrodes and the shape of the ring 5
electrode. When the ac potential increases, the trajectory of the ion becomes 6
unstable in the z direction. The theoretical basis for this motion will be discussed 7
later. While it would be convenient to describe the ion traps function as a point 8
charge responding to an electrical field, the complexity of the generated field 9
makes this impractical. 10
11
12
48
Figure 5.16 The Trajectories of a Single Mass Within the Electrical Field. Figure 1
6 from Wong and Cooks, 1997. Reprinted with permission of Bioanalytical 2
Systems, Inc., West Lafayette, IN. 3
4
The simplest way to understand how the ion trap creates mass spectra is 5
to study how ions respond to the electrical field. It is necessary to begin by 6
constructing a stability diagram for a single ion. Imagine a single mass to charge 7
ratio being introduced into the ion trap. Then, the ac and dc voltages of the ring 8
electrode are altered and the ions stability in both the z and r directions are 9
determined simultaneously. If this experiment was performed multiple times, the 10
stability diagram for that single mass would look similar to Figure 5.17. 11
12
13
14
Figure 5.17 A Single Mass Stability Diagram for an Ion Trap. Adapted from 15
Figure 5 from Wong and Cooks, 1997. Reprinted with permission of Bioanalytical 16
Systems, Inc., West Lafayette, IN. 17
18
49
The yellow area indicates the values of the ac and dc voltages where the 1
given mass has a stable trajectory in the z direction but the ions trajectory in the 2
r direction is unstable. As a result, the ion strikes the ring electrode, is 3
neutralized, and removed by the vacuum. The blue area is voltages where the 4
ion has a stable trajectory in the r direction, but not in the z direction. At these 5
voltages, the ion exits the trap through the slits in the end cap electrode towards 6
a detector. The detector is on if the analyst is attempting to generate a mass 7
spectrum, and can be left off if the goal is to isolate a particular mass to charge 8
ratio of interest. The gray-purple area is where the stability in both the r and z 9
direction overlap. For these voltages, the ion has a stable trajectory and remains 10
inside the trap. 11
12
Similar to the quadrupole mass filter, differential equations are able to 13
expand the single mass stability diagram to a general stability diagram. The 14
derivation of this result requires an in depth understanding of differential 15
equations, so only the graphical result will be presented here (Figure 5.18). As 16
with the linear quadrupole mass filter, the solution here is simplified from a six- 17
variable problem to a simpler two-variable problem. 18
19
50
1
2
Figure 5.18 A General Stability Diagram. Adapted from Figure 5 from Wong and 3
Cooks, 1997. Reprinted with permission of Bioanalytical Systems, Inc., West 4
Lafayette, IN. 5
6
From the general stability diagram it becomes visible how scans can be 7
performed by just altering the ac voltage on the ring electrode. But before we 8
discuss the ion traps operation it is necessary to understand the parameters that 9
affect ions stability within the field. The terms K
1
and K
2
are characterized by the 10
following equations: 11
12

!
K
1
=
4e
r
0
2
"
2
Eqn 5.11
K
2
=
#8e
r
0
2
"
2
Eqn 5.12
13
14
51
As expected, these parameters are very similar to the ones that resulted 1
from the general stability diagram for the quadrupole mass filter. These 2
parameters, like in the quadrupole, are also kept constant during a scan. The 3
charge of the particle (e), the distance from the center of the trap to the ring 4
electrode (r
0
), and the radial frequency of the ac voltage (") are all kept constant 5
during the run. While it would be possible to alter both the ac and dc voltages, in 6
practice it is only necessary to alter the ac voltage (V) of the ring electrode. The 7
dc voltage (U) on the other hand, is kept at zero. If the dc voltage is kept at a 8
ground potential, increasing the ac voltage will eventually result in an unstable 9
trajectory in the z direction. When ac voltage creates a q
z
value that is greater 10
than 0.908, the particle will be ejected from the trap towards a detector through 11
the end cap electrode. As illustrated below, the q
z
value is dependent on the 12
mass to charge ratio of the particle, each different mass has a unique ac voltage 13
that causes them to exit the trap. 14
15
For example, lets place four different ion masses into the ion trap where 16
each has a single positive charge. The general stability diagram in Figure 5.19 is 17
identical to Figure 5.18 except that it is focused around a dc voltage (U) of zero 18
and the scale is enlarged; thus, a
x
is equal to zero through a scan. A mass scan 19
is performed by starting the ring electrode out at a low ac voltage. Each distinct 20
mass has a unique q
z
value, which is visually illustrated by placing these particles 21
on the stability diagram. As the ac frequency begins to increase, the q
z
values 22
for these masses also increases. Once the q
z
value becomes greater than 23
0.908, the ions still have a stable trajectory in the r direction but now have an 24
unstable trajectory in the z direction. As a result, they are ejected out of the trap 25
through the end cap electrode towards the detector. 26
27
28
29
52
1
Figure 5.19 A Stability Diagram During a Sample Scan 2
3
The stability diagram above at A, B, and C was the result of taking a 4
snapshot of the ac voltage during the scan and placing each mass at its 5
corresponding q
z
values for that particular voltage. In this mode of operation, the 6
lightest masses (m
1
) are always ejected from the trap (Figure 15.19 B) before the 7
heaver masses (m
2
). The heaviest masses (m
3
and m
4
) still remain in the trap at 8
point C. To eject these ions, a very large ac voltage is necessary. This voltage 9
is so high that it becomes extremely difficulty to eject ions over a m/z value of 10
650. Since it is impractical to apply such high voltages to the electrode and its 11
53
circuits, a new method of operation needed to be discovered so the ion trap 1
could analyze more massive molecules. 2
3
As a result, resonance ejection was developed to extend the mass range 4
of the ion trap to a m/z value of several thousand. Under normal scanning 5
conditions, ions oscillate at a given frequency depending on their q
z
value which 6
is a function of its mass, charge, and the amplitude of the ac voltage. This 7
frequency is referred to as the ions secular frequency. It was discovered that an 8
ac voltage applied to the end cap electrodes would only affect one ions secular 9
frequency. The effected ions oscillation in the z direction would increase linearly 10
until it was ejected from the trap. Resonance ejection can be conceptualized as 11
a hole inside the stability diagram at any chosen q
z
value. Then the ac voltage 12
of the ring electrode can be altered so any mass can have the same q
z
value as 13
the hole and exit the trap in the z direction (Figure 5.20). This mode of 14
operation not only extended the mass analyzers mass range, but it also made it 15
possible to eject ions from the trap in any order. Before this mode of operation 16
existed, it was only possible to eject the ions in order from lightest to heaviest. 17
Figure 5.20 illustrates how it is possible to eject the heaviest ion (m
4
) before the 18
lighter ion (m
3
). 19
20
54
1
2
Figure 5.20 A Sample Resonance Ejection Scan 3
4
The resonance ejection mode of operation is one reason why the ion trap 5
is such a valuable tool. It not only greatly extends the mass range of the mass 6
analyzer, but it also increased its applications in tandem spectroscopy (section 7
5.5.8). The ability to isolate any given mass under several thousand amu is an 8
extremely powerful tool. Through the use of both modes of operation, the ion 9
trap has become a valuable tool in performing many specialized mass 10
separations. 11
12
View Animation 5.7 at this time for an illustration of how an ion trap mass 13
filter contains and ejects ions of given mass to charge ratios. 14
15
16
17
18
55
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Animation 5.7. Illustration of an Ion Trap Mass Filter. 19
20
5.5.4 Time-of-Flight (TOF) mass filter: While time-of-flight mass filters 21
were one of the first MS systems to be developed, they had limited use due to 22
their need for very fast electronics to process the data. Developments in fast 23
electronics and the need for mass filters capable of resolving high mass ranges 24
(such as in MALDI systems) has renewed interest in time of flight systems. TOF 25
is used exclusively with MALDI systems and also has other applications, as in 26
HPLC where high molecular weight compounds are encountered. 27
28
Entry into the TOF mass filter is considerably different than with other 29
mass filters. The entry has to be pulsed or intermittent in order to allow for all of 30
the ions entering the TOF to reach the detector before more ions are created. 31
56
With sources that operate in a pulsing fashon such as MALDI or field desorption, 1
the TOF functions easily as a mass analyzer. In sources that continually produce 2
ions such as a GC system or an EI source, the use of a TOF is more difficult. In 3
order to use a TOF system with these continuous sources, an electronic gate 4
must be used to create the necessary pulse of ions. The gate changes the 5
potential on an accelerator plate to only allow ions to enter the TOF mass filter in 6
pulses. When the slit has a positive charge, ions will not approach the entryway 7
to the mass analyzer and are retained in the ionization chamber. When all of the 8
previously admitted ions have reached the detector, the polarity on the 9
accelerator(s) is again changed to negative and ions are accelerated toward the 10
slit(s) and into the TOF mass analyzer. This process is repeated until several 11
scans of each chromatographic peak have been measured. (This type of 12
ionization and slit pulsing will be shown in the animation below). The other way 13
to interface EI with TOFs is to operate the EI source in a pulsing mode. This is 14
achieved by maintaining a constant negative polarity on the accelerator plate/slit, 15
and pulsing the EI source. This method can also periodically introduce packets 16
of ions into the TOF mass filter. 17
18
Whichever type of ionization and entry into the TOF mass filter is used the 19
remainder of the process is the same. Prior to developing the mathematics 20
behind TOF separations a simple summary is useful. Mass to charge ratios in 21
the TOF instrument are determined by measuring the time it takes for ions to 22
pass through the field-free drift tube to the detector. The term field-free is 23
used since there is no electronic or magnetic field affecting the ions. The only 24
force applied to the ions occurs at the repulsion plate and the acceleration 25
plate(s) where ions obtain a similar kinetic energy (KE). All of the ions of the 26
same mass to charge ratio entering the TOF mass analyzers have the same 27
kinetic energy and velocity since they have been exposed to the same voltage on 28
the plates. Ions with different mass to charge ratios will have velocities that will 29
vary inversely to their masses. Lighter ions will have higher velocities and will 30
57
arrive at the detector earlier than heavier ones. This is due to the relationship 1
between mass and kinetic energy. 2
3

!
KE = mv
2
/2 Eqn 5.13 4
5
The kinetic energy of an ion with a mass m and a total charge of q =ze is 6
described by: 7
8

!
mv
2
/2 = q V
S
= z e V
S
Eqn 5.14 9
10
where V
S
is potential difference between the accelerator plates, z is the charge 11
on the ion, and e is the charge of an electron (1.60 x 10
-19
C). The length (d) of 12
the drift tube is known and fixed, thus the time (t) required to travel this distance 13
is 14
15

!
t = d/v Eqn 5.15 16
17
By solving the previous equation for v and substituting it into the above equation 18
we obtain 19
20

!
t
2
=
m
z

d
2
2V
S
e
"
#
$
%
&
'
Eqn 5.16 21
22
In a TOF mass analyzer, the terms in parentheses are constant, so the mass to 23
charge of an ion is directly related to the time of travel. Typical times to traverse 24
the field-free drift tube are 1 to 30 ms. 25
26
Advantages of a TOF mass filter include their simplicity and ruggedness 27
and a virtually unlimited mass range. Additionally, virtually all ions produced in 28
the ionization chamber enter the TOF mass filter and traverse the drift tube. 29
However, TOF mass filters suffer from limited resolution, related to the relatively 30
58
large distribution in flight times among identical ions (resulting from the physical 1
width of the plug of ions entering the mass analyzer). Animation 5.8 illustrates 2
how a pulsed accelerator plate/slit acts as a gate for a TOF mass filter system. 3
4
5
6
7
8
9
10
11
12
13
Animation 5.8. Illustration of a Traditional TOF Mass Filter. 14
15
Animation 5.9 illustrations how a pulsed accelerator plate/slit acts as a gate for a 16
reflective TOF mass filter system. (The system shown is actually for the analysis 17
of metal isotopes with an Inductively Coupled Plasma (ICP), but the reflective 18
TOF works the same for organic analytes. 19
20
21
22
23
24
25
26
27
28
29
30
31
59
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
36
Animation 5.9. Illustration of a Reflective TOF Mass Filter 37
38
Ion Mobility Mass Spectrometry: 39
40
If you have been in an airport recently you have seen or your luggage has 41
been analyzed by an Ion Mobility Spectrometer (IMS). Although originally 42
developed by Earl W. McDaniel of Georgia Institute of Technology in the 1950s, 43
IMS systems have gained popularity recently due to their versatilitythey can 44
designed for specific classes of compounds, they have excellent detection limits, 45
and they can be manufactured to be light-weight and mobile. 46
47
60
The basic design is similar to the TOF mass filter. Important differences 1
are that they use an easier ionization source, they can be operated at 2
atmospheric pressure and therefore do not necessarily require pressurized gases 3
or high vacuum pumps, and as a result of their atmospheric pressure sample 4
introduction they have superior detection limits. Samples are introduced at 5
atmospheric pressure and ionized by corona discharge, atmospheric pressure 6
photoionization (APPI), electrospray ionization (ESI), or a radioactive source 7
such as a small piece of
63
Ni or
241
Am, similar to the thoses used in ionization 8
smoke detectors or GC electron capture detectors. The ionized analytes are 9
then introduced to the drift tube by a gate valve similar to the one described 10
earlier in this section for TOF mass filters. However, the IMS drift tube is 11
different in that it can be operated at atmospheric pressure and is a counter 12
current environment. The analytes travel from left to right in the one-meter drift 13
tube due to a 10-30 kV potential difference between the inlet and exit. As the 14
analytes are mobile due to the potential they travel through a buffer gas that is 15
passed from right to left in the drift tube (and atmospheric gases are commonly 16
used). Separation of different analtyes is achieved due to each ion having a 17
different mass, charge, size and shape (the ion mobility). As the ions are 18
electrically drawn toward the detector, the ions cross sectional area strikes buffer 19
gases and its velocity is reduced based on its size and shape. Larger ions will 20
collide with more buffer gas and be impeded, travel slower, and arrive at the 21
detector after longer times in the drift tube. Detectors for IMS are usually 22
relatively simple Faraday cups but better detection limits can be obtained with an 23
EM. 24
25
The most common use of IMS is for volatile organic molecules. IMS has 26
been expanded for use in gas, liquid, and super critical fluid chromatography. 27
28
5.5.5 Double Focusing Systems: The magnetic sector MS described 29
earlier is referred to as a single-focusing instrument since it only uses the 30
magnetic component to separate ion mass to charge ratios. This can be 31
61
improved by adding a second electrostatic-field based mass filter, and is referred 1
to as double focusing. A magnetic field instrument focuses on the distribution of 2
translational energies imparted on the ions leaving the ionization source as a 3
means of separation. But in doing so, the magnetic sector instruments broaden 4
the range of kinetic energies of the ions, resulting in a loss of resolution. If we 5
combine both separation techniques by passing the ions separately through an 6
electrostatic field (to focus the kinetic energy of the ion packet) and magnetic 7
field (to focus the translational energies of the ion packet), we will greatly improve 8
our resolution. In fact, by doing this we can measure ion masses to within a few 9
parts per million (precision) which results in a resolution of ~2500. Compare this 10
to the unit resolutions (28 versus 29 Daltons) discussed at the beginning of this 11
section (under resolution). On the downside, these instruments can be costly. 12
13
5.5.6 Fourier Transform Ion Cyclotron Mass Spectrometry: (by Nicole 14
J ames) 15
16
Developed by Alan G. Marshall and Melvin B. Comisarow at the University 17
of British Columbia, the use of FT-ICR MS first began in 1974 with approximately 18
235 instruments in use by 1998. FT-ICR MS has higher mass resolution and 19
accuracy than any other MS system and can detect multiple mass-to-charge ratio 20
ions simultaneously. However, FT-ICR MS can be prohibitively expensive at $1- 21
2 million for a standard instrument. 22
The general steps of an FT-ICR MS experiment are: (1) ion formation 23
outside of the detector; (2) ion focusing and accumulation; (3) transportation of 24
ions into a Penning trap; (4) selection of ions based on mass-to-charge ratio and 25
ejection of these ions from the Penning trap; (5) excitation; (6) detection; (7) fast 26
Fourier transform of the digital time-domain signal; (8) conversion of frequency to 27
mass-to-charge ratio. 28
29
Ion-Cyclotron Motion 30
62
If a moving ion is exposed to a uniform magnetic field, it is subject to a 1
force dependent on the mass, charge and velocity of the ion. If an ion does not 2
collide with another particle and hit off its natural course, the magnetic field will 3
bend the ions path into a circular orbit. 4
5
6
7
Figure 5.21a Illustration of a particle in a magnetic field. 8
9
The motion of the ion can be described by the equation below, where w is the 10
unperturbed ion cyclotron frequency, B
0
is the magnetic field in Tesla, q is the 11
charge in Coulombs, and m is the mass in micrograms. 12

!
w =
qB
o
m
13
This equation can be rearranged into the following equation where v is the 14
velocity, and z is the charge of the ion in units of elemental charge (e.g., +1, +2, 15
etc). 16

!
v =
w
2"
=
1.5356 x 10
7
B
o
m/z
17
18
It is important to note that the above equation is dependent on only the mass-to- 19
charge ratio of the ion and not its velocity. This makes ion-cyclotron resonance 20
especially useful in mass spectrometry, as one does not need to focus 21
translational energywhich requires longer experiment times, larger apparatus 22
and more powerful electronicsin order to obtain high-accuracy results. 23
The radius of the circle an ion makes when exposed to the magnetic field 24
can be found by the equation below, where r is radius in meters, and T is the 25
temperature in Kelvin: 26
63

!
r =
1.3365 x 10
-6
zB
0
mT 1
2
One can see from the above equation that an ion with a mass of 100 amu and a 3
charge of +1 in a magnetic field of 1 Tesla at room temperature (298 K) would 4
have a radius of 0.2 mm; the same ion with triple the magnetic field (3 T) at room 5
temperature would have a radius of 0.077 mm. Thus, ions can be easily confined 6
to a relatively small orbit by a reasonable magnetic field; this is called ion 7
trapping and is vital to ICR-MS because the longer (approximately 1s) 8
experiment times require one to be able to retain the ions in a designated space. 9
Additionally, a 3T magnetic field is easily attainable for commercially available 10
electronics. The largest FT ICR MS built as of 2010 can attain a magnetic field of 11
15T, allowing one to confine an ion with an m/z value of 60,000. 12
13
Ion Cyclotron Excitation and Detection 14
A number of ions at a specific mass-to-charge ratio spinning in an ion- 15
cyclotron orbit does not, itself, generate an observable electric signal, because 16
(a) the ions were randomly placed (i.e, incoherent; ions are spread throughout 17
the radio of orbit) as they began orbiting, meaning that an ion at a specific 18
position will have its charge cancelled out by an ion half an orbit away from it, 19
leaving no net electrical current, and (b) the radius of the orbits are generally too 20
small to be detectable, even if all ions were in the same phase. Thus, ions must 21
be excited in order to be detected. 22
Particles in an ion-cyclotron orbit can be excited by applying an oscillating 23
or rotating uniform electric field at or near the frequency of ions of a given mass- 24
to-charge ratio. This excitation can be used for three purposes: (1) accelerating 25
the ions into a larger orbital radius for detection, (2) accelerating the ions to a 26
larger orbital that is ejected from the ion trap, and (3) increasing the kinetic 27
energy of an ion to the point that it further ionizes or reacts with another 28
molecule. For the purposes of this text, excitation in order to accelerate the ions 29
for detection is most significant. 30
64
Applying an oscillating or rotating radio-frequency electric field in 1
resonance with (at the same frequency as) a specific m/z value or range applies 2
a force on the ion(s) that continuously enlarges the circular orbit of the ion(s) at 3
one pointin other words, the orbiting ions begin to spiral outward. Ions of 4
different types will spiral outward at different rates. The post-excitation orbit for 5
an ion excited for a period of time, t, is shown in the following equation, where E
0
6
is the applied electric field and B
0
is the magnetic field: 7

!
r =
E
0
t
2B
o
8
9
The fact that the above equation is independent of the mass-to-charge ratio of 10
the ion means that all ions can be excited by a radio-frequency electric field to 11
enlarged ion-cyclotron orbits for detection. This simultaneous detection vastly 12
decreases both the time an experiment will take and the amount of analyte 13
required. 14
When a group of ions with the same mass-to-charge ratio are excited, 15
they are pushed off-axis due to their spiraling nature. By pushing the ions off 16
axis, not all ions have a partner ion half a cycle awaythe ions are considered 17
to be cohered. A cohered packet of orbiting ions causes a difference in current 18
between opposing detection plates within the ion trap; this differential current can 19
be modeled as an image current opposing the current on the detection plates; 20
this image current is proportional to the number of coherent orbiting ions. This is 21
the ICR signal; the ICR signal increases linearly with increasing ion-cyclotron 22
radius after excitation and with increasing ion charge. Throughout most of the 23
frequency range possible on the instrument, the signal-to-noise ratio (S/N) is 24
proportional to the differential current observed. The number of ions required for 25
a S/N ratio of 3:1 on a standard instrument using standard parameters is 26
approximately 190 ions. Other detection processes have been designed to such 27
high accuracy and detection that they are able to detect a single ion and have 28
been used to corroborate the theory that protons and anti-protons do, in fact, 29
have the same mass (Gabrielse et al, 1990). 30
65
1
2
Figure 5.21b Excitation and Detection of an Ion. 3
4
The Penning Trap 5
The most common ion trap used in FT-ICR MS is the Penning trap, 6
designed in the 1950s by Hans Georg Dehmelt, who named it after Frances 7
Michel Penning for his work on the Penning gauge. The ion-cyclotron motion 8
induced by a radial magnetic field contains ions radially, but it is necessary to 9
add an axial electric field in order to trap the ions axially. Thus, the motion of an 10
ion inside a Penning trap is essentially the combination of three distinct motions: 11
the cyclotron, magnetron (a component of ion-cyclotron motion), and the axial 12
motion. The axial containment is accomplished by introducing two end-cap 13
electrodes. The end-cap electrodes are coupled by capacitance, which allows 14
for a nearly perfect rf electric field to be used for the ion-cyclotron excitation 15
without any negative effects on other electronics. Opposing plates with an 16
electric field applied across them within the Penning trap are used as detector 17
plates. 18
19
66
1
2
Figure 5.21c Diagram of a Penning Trap. 3
4
Analysis of Results 5
The signal detected by an experiment is in units of current per time. To 6
extract mass-to-charge data, one must apply a Fourier transform. In general, a 7
Fourier transform (FT) takes a time-based signal and converts it into a frequency- 8
based plot. Since the initial function is a function of time, it is typically called the 9
time domain; the frequency plot is called the frequency domain, or the frequency 10
domain representation of the initial function. More specifically, the Fourier 11
transform uses the fact that almost any function can be degraded into a sum of 12
sine and cosine waves; each component sine and/or cosine wave represents a 13
periodic component of the data. By finding each component sine or cosine wave, 14
one can make a frequency plot by representing a specific sine or cosine wave as 15
a peak on a plot of amplitude versus the frequency of the wave. The sharper the 16
peak, the more exact the periodicity is; in most real-life applications, the peak 17
will be somewhat broadnot just a vertical line. 18
67
1
2
3
Figure 5.21d Graphic Representation of the Fourier-Transform Process where a 4
time domain signal is transformed to a frequency output. 5
6
A Fourier-transform of the (time-domain) ICR response results in a 7
frequency plot that can be mass-corrected to result in a mass spectrum. 8
Obtaining this mass spectrum with most other types of MS would have required 9
sweeping slowly across the entire range of mass-to-charge ratios; being able to 10
quickly and simultaneously detect all mass-to-charge ratios decreases the time, 11
effort and supplies that must be used to test a sample. In addition, the greatly 12
increased resolution means that FTICR-MS will continue being an extremely 13
powerful instrument. 14
15
68
1
2
Figure 5.21e Overall Schematic of an Ion Cyclotron Mass Spectrometer. 3
4
5.5.7 Orbitrap Analyzers (by Nicole J ames) 5
6
Designed in 2005 by Alexander Makarov, the Orbitrap mass spectrometer 7
features a mass resolution of up to 150,000, high mass accuracy (2-5ppm, 8
compared with approximately 20ppm for quadrapole systems), a mass-to-charge 9
ratio range of 6,000 and a dynamic range larger than 1,000. 10
The Orbitrap works similarly to an FT ICR-MS: all ions are identified 11
simultaneously by reading an image current of oscillations that are unique to a 12
given mass-to-charge ratio and a Fourier transform is applied to the data to 13
isolate individual signals. However, the Orbitrap requires no magnet, no RF field, 14
and no excitation sequence. Despite this, Orbitrap systems generally cost at 15
least $600,000. 16
Ions are first ionized by a given source; given the large m/z range, 17
Orbitrap systems are often used to study biological molecules such as proteins, 18
peptides, oligsaccharidesconsequently, one of the most common ionization 19
methods is ESI. The ions are then transported to a storage cell, generally a 20
69
storage quadrapole, which is kept at a vacuum near 10
-3
mbar. A series of 1
transfer lenses gradually increases the electric field experienced by the ions until 2
they are at the level of the Orbitrap. 3
After ions have been transferred into the Orbitrap, the system uses only 4
electrostatic (DC) fields. The Orbitrap itself is composed of an outer barrel 5
electrode, an inner spindle electrode, and two endcap electrodes. Upon 6
introduction into the Orbitrap, stable ion trajectories will result in orbiting around 7
the center electrode while also oscillating in the z-direction. The motion in the z- 8
direction can be described as an harmonic oscillator, which is described in 9
equation 5.5.6.1, where # is oscillation frequency, z is the ion charge, m is the 10
ion mass and k is the field curvature. 11
12

!
" =
z
m
k 13
14
15
Figure 5.22 The Orbitrap (reprinted from WikiPedia via the GNU Free 16
Documentation License) 17
18
While the frequency of orbiting the central electrode is also dependent on the 19
ions mass-to-charge ratio, this frequency is also dependent on the ions energy 20
and when it was introduced into the Orbitrap, whereas oscillations in the z- 21
70
direction are independent of energy and any initial parameters. The oscillations in 1
the z-direction are read by the image current produced on the end-cap 2
electrodes. While all ions of a given mass to charge ratio oscillate in phase for 3
hundreds of thousands of oscillations, small imperfections in the Orbitrap or 4
orbital shape, along with occasionally collisions with background gas molecules 5
(despite the 10
-10
mbar vacuum) can result in the loss or displacement of some 6
ions, ultimately resulting in a slow decrease in the intensity of the signal until it is 7
completely lost in instrument noise. This results in a free induction decay (FID), 8
similar to that which is acquired in NMR analysis. A Fourier Transform of the FID 9
results in a mass spectrum. 10
11
12
5.5.8 Tandem Mass Spectroscopy: Mass spectroscopy is commonly 13
referred to as a confirmatory technique since there is little doubt (error) in the 14
identity of an analyte. To be even more certain of an analytes identify, two or 15
even three, mass spectrometers can be used in series (the output of one MS is 16
the input of another MS). Most often a soft ionization source, such as chemical 17
ionization, is used in the first MS and allows for selection of the molecular ion in 18
the first MS, while a harder ionization is used in the second MS to create 19
fragments. A subsequent MS will select for a specific ion fragment from the 20
second MS and further fragment it for identification. This technique allows a 21
molecular ion (or ion fragment) to be isolated in the first MS, subsequently 22
fragmented in the second and third MS, and identified based on its final fragment 23
pattern. You should be able to see the confirmatory nature of this technique. 24
25
Mass filters of choice for use in tandem include magnetic sector, 26
electrostatic, quadrupole, and ion trap systems. In the absence of HPLC or GC 27
introduction, tandem MS offers many of the same advantages of a single GC-MS 28
or HPLC-MS system but it is much faster since the analyst does not have to wait 29
on the chromatography portion of the analysis. For example, chromatograph 30
separations take from minutes to hours prior to entry into a MS, while tandem MS 31
71
systems (without GC) require only milliseconds. But of course, this saving in 1
time is considerably more expensive than simple chromatographic-based MS 2
systems. 3
4
5.6 Ion Detectors 5
6
Once the analytes have been ionized, accelerated, and separated in the 7
mass filter, they must be detected. This is most commonly completed with an 8
electron multiplier (EM), much like the ones used in optical spectroscopy. In MS 9
systems, the electron multiplier is insensitive to ion charge, ion mass, or chemical 10
nature of the ion (as a photomultiplier is relatively insensitive to the wavelength of 11
a photon). EMs for MS systems can be a series of discrete dynodes as in the 12
photomultiplier or they can be continuous in design. Most commonly, continuous 13
EMs are used. Continuous EMs are horn shaped and are typically made of glass 14
that is heavily doped with lead oxide. When a potential is placed along the length 15
of the horn, electrons are ejected as ions strike the surface. Ions usually strike at 16
the entrance of the horn and the resulting electrons are directed inward (by the 17
shape of the horn), colliding sequentially with the walls and generating more and 18
more electrons with each collision. Electrical potentials across the horn can 19
range from high hundreds of volts to 3000 V. Signal amplifications are in the 10 20
000 fold range with nanosecond response times. Animation 1.10 illustrates the 21
response of a continuous electron multiplier as ions, separated in a mass filter, 22
strike its surface. 23
72
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
43
Animation 5.10. Illustration of a Continuous-Dynode Electron Multiplier. 44
45
46
47
48
49
50
73
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
41
Animation 5.11. Illustration of a Discrete-Dynode Electron Multiplier. 42
43
Another form of MS detector is the Faraday Cup that counts each ion 44
entering the detector zone. These detectors are less expensive but provide no 45
amplification of the signal and are not used in typical instruments due to their 46
poor detection limits. 47
48
One of the latest detectors to reach the market is a microchannel plate, a 49
form of an array transducer also called an electrooptical ion detector (EOID). 50
The EOID is a circular disk that contains numerous continuous electron 51
74
multipliers (channels). Each channel has a potential applied across it and each 1
cation reaching the detector will generate typically up to 1000 electrons. The 2
electrons produce light as they impinge on a phosphorescent screen behind the 3
disk containing the channels. An optical array detector using fiber optic 4
technology records the flashes of light and produces a two dimensional 5
resolution of the ions. The advantage of an EOID is their ability to greatly 6
increase the speed of mass determinations by detecting a limited range of 7
masses simultaneously, thus reducing the number of discrete magnetic field 8
adjustments required over a large range of masses. EOIDs have not been readily 9
incorporated into instruments as initially anticipated. 10
11
5.7 Three-Dimensional Aspects of GC-MS 12
13
Typical chromatographic peaks were illustrated in earlier chapters. But as 14
each chromatographic peak enters the MS it is fragmented and separated into a 15
series of ion fragments. When graphed together on an x, y, and z plot, the x-axis 16
represents time and traces the arrival of each compound at the chromatographic 17
detector and the z-axis represents the total detector response that is related to 18
analyte concentration. The mass-to-charge spectrum of each chromatographic 19
peak is represented by a series of lines that are parallel to the y-axis and show 20
the arrival of molecular fragments at the MS detector. Again, detector response 21
and concentration are represented by the height of each peak. This is illustrated 22
for one chromatographic peak in Figure 5.22. 23
24
75
1
2
Figure 5.23. The Three-Dimensional Nature of a GC-MS Analysis. 3
4
5.8 Summary 5
6
In this chapter we illustrated the utility of combining chromatography and 7
MS systems. A variety of possible components provide for interesting 8
instruments that can be used to analyze a broad range of analytes. Hard and 9
soft ionizations techniques provide for the determination of the molecular weight 10
of the analyte, as well as unique fragmentation patterns for confirmational 11
identification of an unknown chemical structure. More inexpensive instruments, 12
such as quadrupole and time of flight mass spectrometers, allow only unit 13
resolution of ions while double focusing instruments yield the determination of 14
differences with resolution of four decimal points in masses. Mass spectrometry, 15
like NMR, is one of the most powerful techniques available to chemists and it is 16
becoming more and more important. While most of the instruments presented in 17
the chapter have detection limits in the sub parts per million range, extremely 18
lower detection limits (10
-15
moles) have been obtained in research-grade 19
instruments. 20
21
76
A summary of mass filters and their characteristics is given below in Table 1
5.2. 2
3
Table 4.2 Summary of Mass Filter Features. Source: Company Literature and 4
Personal Communiqu David Koppenaal, Thermal Scientific & EMSL, Pacific 5
National Laboratory. 6
7
Type of Mass Filter Resolution Detection
Limit
Approximate Instrument
Price
Routine Mass Filters Coupled with ICP
Single Quadrupole 250-500 low ppb
high ppt
$80 000 - $100 000
Ion Trap 1 000
10 000
ppb $250 000 - $300 000
Time of Flight 3000
10 000
high ppt $300 000 - $400 000
Double Focusing 10 000
20 000
mid to high
ppt
$750 000 - $1 000 000
Fourier Transform Ion
Cyclotron
200 000
1 000 000
ppb $1 000 000 +
New Mass Filters
Magnetic Sector /
Multi-collector with
the Mattauch-Herzog
Geometry
~500 high ppb $350 000 - $400 000
Proton Transfer
Reaction Ionization
Chamber
Depends on
type of
mass filter
ppt $120 000
Orbital Trap
(Electrostatic Ion
Trap)
150 000
200 000
ppb $600 000 (currently only
available with HPLC)
77
1
5.9 Questions 2
3
1. Why are most mass filters maintained at a low pressure? 4
5
2. List the common ways samples are introduced into a MS system. 6
7
3. How can solid samples be introduced into a MS? 8
9
4. Draw and explain how the interface between a GC and a MS works. 10
11
5. Why do capillary columns, versus packed columns, work best for MS 12
interfaces? 13
14
6. Explain the difference between hard and soft ionization in GC-MS. 15
16
7. Why does soft ionization reduce the fragmentation of analytes in GC-MS? 17
18
8. Write the chemical reactions occurring when methane is used in soft 19
ionization. 20
21
9. Draw and explain how the interface between a LC and a MS works. 22
23
10. What is the major problem with interfacing LC (ESI) to MS? 24
25
11. Explain how MALDI works. What types of samples is it commonly used for. 26
What type of MS is it commonly coupled with? 27
28
12. Draw and explain how the interface between a CE (ESI) and a MS works. 29
30
13. Explain resolution with respect to mass filters. Give relevant resolution 31
numbers. 32
33
14. Draw and explain how a magnetic sector mass filter works. 34
35
15. Draw and explain (in detail) how a quadrupole mass filter works. 36
37
16. The governing equation of the quadrupole mass filter consists of a six- 38
parameter differential equation. Which two parameters are used to control the 39
mass filter? 40
41
17. What is the purpose of the dc voltage in the quadrupole MS? 42
43
18. What is the purpose of the ac cycle in the quadrupole MS? 44
45
78
19. How does the low mass and high mass filters work to create a stable cation 1
region in the quadrupole MS? 2
3
20. Explain the mass scan line in the quadrupole MS figures. 4
5
21. What is the purpose of sweeping the dc-ac voltages? 6
7
22. Extend the concepts of a linear quadrupole mass filter, explained above, to 8
explain how the quadrupole ion trap mass filter works. 9
10
23. How is the mass range of a quadrupole ion trap mass filter extended? 11
12
24. Explain the concept of resonance ejection in ion trap mass filters. 13
14
25. Draw and explain how a time-of-flight mass filter works. 15
16
26. Contrast traditional TOF and ion mobility MS. 17
18
27. Draw and explain how a PTR-MS works. 19
20
28. Give a brief explanation of how an Ion Cyclotron works. 21
22
29. Draw and explain how a double focusing mass filter works. What are its 23
advantages? 24
25
30. What is tandem mass spectrometry? 26
27
31. What types of detectors are used in mass spectrometry? 28
29
32. Use the date in Table 5.2 to contrast the various types of mass filters. Which 30
is the most economical? Which has the best mass resolution? 31
32
33
5.10 References 34
35
Blakely C.R., Vestal M.L. Thermospray interface for liquid chromatography/mass 36
spectrometry. Anal. Chem., 1983, 55(4), p.750 37
38
G.Gabrielse, X. Fei, L.A. Orozco, R.L. Tjoelker, J . Haas, H. Kalinowsky, T.A. 39
Trainor, W. Kells, Phys. Rev. Lett. 65 (1990) 1317-1320 40
41
79
Golhke R.S., McLafferty F., Wiley B., Harrington D. First demonstration of 1
GC/MS, Bendix Corporation, 1956 2
3
Hansel, A., A. J ordan, R. Holzinger, P.Prazeller, W. Vogel, and W. Lindinger. 4
Proton transfer reaction mass spectrometry: on-line trace gas analysis at 5
the ppb level. International J ournal of Mass Spectrometry and Ion 6
Processes, 1995, 149/150, pp. 609-619 7
8
de Hoffmann, E.; Stroobant, V. Mass Spectrometry Principles and 9
Applications, 2nd ed.; J ohn Wiley & Sons: New York, 2002. 10
11
J onscher, K.; Yates, J . The Quadrupole Ion Trap Mass Spectrometer A 12
Small Solution to a Big Challenge. Analytical Biochemistry 1997, 244. 1- 13
15. 14
15
Kanu AB, Dwivedi P, Tam M, Matz L, Hill HH (J anuary 2008). "Ion mobility-mass 16
spectrometry". J Mass Spectrom 43 (1): 122. 17
18
March, R. An Introduction to Quadrupole Ion Trap Mass Spectrometry. 19
Journal of Mass Spectrometry 1997, 32 351-369. 20
21
A.G. Marshall, C.L. Hendrickson, Int. J . Mass. Spec. 215 (2002) 59-75 22
23
G. Marshall, C.L. Hendrickson, G.S. J ackson, Mass. Spec. Rev., 17 (1998) 1-35 24
25
Miller, P.; Denton, M. The Quadrupole Mass Filter: Basic Operating 26
Concepts. Journal of Chemical Education 1986, 63 (7), 617-622. 27
28
Skoog, D.; Holler, F.; Nieman, T. Principles of Instrumental Analysis, 5th 29
ed.; Saunders College Publishing: Philadelphia, 1992. 30
31
80
Steel, C.; Henchman, M. Understanding the Quadrupole Mass Filter through 1
Computer Simulation. Journal of Chemical Education 1998, 75 (8), 1049- 2
1054. 3
4
N.I. Tarantin, Phys. Res. B 126 (1997) 392-395 5
6
M. Ubieto-Diaz, D. Rodriguez, S. Lukic, Sz. Nagy, S. Stahl, K. Blaum, Int. J 7
Mass. Spec. 288 (2009) 1-5 8
9
Wong, P.; Cooks, R. Ion Trap Mass Spectrometry. Current Seperations.com 10
and Drug Development 1997, 16 (3). 11
12
Yamashita M., Fenn, J .B. Electrospray Ion Source. Another Variation on the 13
free-jet theme. J ournal of Physical Chemistry, 1984, 88(20), p.4451 14
Zolotov, Yu. A. (2006). Ion Mobility Spectrometry. Journal of Analytical 15
Chemistry, 61 (6): 519. 16
1
CHAPTER 6 1
2
GC EI and CI Fragmentation and Interpretation of Spectra 3
4
6.1 Introduction 5
6
Before discussing fragmentation and interpretation, it is important to 7
understand the many ways mass spectra are utilized. For the analytical chemist, 8
a mass spectrum is useful for two applications. The first is the relatively simple 9
case when the analyst is looking for a particular compound in a sample and has 10
a reference material to compare spectra. The second occurs when an analyst 11
observes the presence of an unknown and wishes to identify it. The mass 12
spectrum allows an experienced analyst to identify the compound or at a 13
minimum narrow the possibilities down to a few compounds from the millions of 14
potential chemicals. Then, a reference standard can be more easily selected 15
from this knowledge to confirm the identity of this unknown. A similar situation 16
exists for the synthetic chemist except their analytical tool box is much larger. 17
Sometimes synthetic chemists are attempting to synthesize a final product that is 18
known (for example in an industrial process line). Here, the mass spectrum of 19
the synthesized product is compared to a reference standard. Other times, a 20
synthetic chemist is attempting is attempting to make a new compound where a 21
reference standard is impossible to find. 22
23
All four problems center on the same difficult task, identifying the structure 24
of a compound under various conditions. There are three main instruments that 25
perform this task for organic compounds, infrared spectroscopy, mass 26
spectroscopy and nuclear magnetic resonance (NMR). It is very important that 27
both synthetic and analytical chemists are able to choose the best tool for their 28
particular problem. The mass spectrometer has a few advantages over the other 29
analytical methods. Mass spectroscopy, when coupled with either gas or liquid 30
chromatography, can analyze a complex mixture that an NMR or IR could not. A 31
2
MS is also the only way to determine the molecular mass of a compound. The 1
largest advantage for analytical chemists is that mass spectroscopy can 2
elucidate structural information from a very small amount of a compound (part 3
per million quantities). 4
5
The MS has a distinctive advantage over IR spectroscopy in that there is 6
more structural information can be determined, though the information contained 7
in a mass spectrum is more difficult to interpret. The MS has an advantage over 8
NMR in that it can be performed more quickly. However, both IR spectroscopy 9
and mass spectroscopy have a distinct disadvantage when analyzing 10
compounds with multiple functional groups. For these types of compounds and 11
when the analyst has mg quantities of a relatively pure compound, NMR is 12
usually the best analytical tool. 13
14
As a result of these advantages and disadvantages, mass spectroscopy is 15
normally utilized to perform three tasks. The first is in analytical chemistry when 16
there is a small concentration of analyte. The second is identifying compounds 17
that contain few functional groups; a common procedure in industrial synthesis. 18
The third is confirming steps in a complex synthesis of a new product to 19
determine the molecular mass and possible some structural information. The 20
products in the third example, however, are usually always checked by NMR. 21
22
After choosing to use mass spectroscopy, the selection of gas 23
chromatography or liquid chromatography is equally important. Gas 24
chromatography is utilized for volatile and thermally stable compounds (up to 300 25
C). Liquid chromatography is usually utilized for all other compounds since it 26
has poorer resolution than GC. As a result, this chapter will focus upon 27
interpreting structural information from the types of compounds commonly 28
analyzed with GC-MS. Once a chemist is able to determine the identity of a 29
compound from a mass spectrum, their problem has been solved. 30
31
3
6.2 Creation of the Spectra 1
2
As sample molecules exit the GC column and enter into the mass 3
spectrophotometer, they encounter an energy source. For the purposes of this 4
chapter, the source is an electron impact tungsten filament at 70 eV (Section 5
5.2.1.2a). Energy emitted from the source removes a single electron from a 6
sample molecule. This is the most basic reaction and is illustrated by methanol 7
below (Figure 6.1). 8
9

!
e
-
+ CH
3
OH " [CH
3
OH]
+
+ 2e
-
Rxn 6.1 10
11
After these products move through the mass spectrometer, the detector is 12
only sensitive to the positively charged molecules and is not sensitive to any 13
neutral or radical molecules. The detector transforms the number of molecules 14
into an electrical signal, and a computer or integrator translates this individual 15
signal peaks into a bar graph. The abundance is plotted as a function of a 16
molecules mass divided by its charge (m/z) on a bar graph (Figure 6.1). Since 17
almost all of the fragments detected by the GC-MS have only a single positive 18
charge, m/z is also a measurement of the molecules mass. 19
20
4
1
Figure 6.1 Mass Spectrum of Methanol 2
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 3
permission from NIST. 4
5
There are more bars on the graph than just the mass of the sample 6
molecule. These other peaks are attributed to the cleavage of bonds in the 7
original sample molecule. These fragments allow for the original structure of the 8
sample molecule to be determined by looking at its various components (Section 9
6.8). Since the energy of the source exceeds the ionization energy of the sample 10
molecule, the excess energy that is not utilized in the removal of a single electron 11
is distributed over various electronic, vibrational, and rotational degrees of 12
freedom (Section 6.6). Fractionation occurs when this energy exceeds the 13
activation energy of any bond cleavage (Section 6.6). This feature allows the 14
instrument to distinguish between compounds with the same molecular mass and 15
constitutional isomers. The major fragments for methanol (Figure 6.1) can be 16
attributed to the following reactions. 17
18
5

!
[CH
3
OH]
+
" CH
3
O
+
+ H
[CH
3
OH]
+
" CH
2
O
+
+ H
2
Rxns 6.2
[CH
3
OH]
+
" CH
3
+
+ OH
1
2
The two most important peaks in any mass spectrum are the base peak 3
and the molecular ion peak. The base beak (also referred to as the parent peak) 4
is the largest peak in the spectrum. In the case of methane, the base peak is the 5
peak at m/z 31 corresponding to the CH
3
O
+
fragment. Since the absolute height 6
of any peak is dependent on the concentration of the sample, the other peaks in 7
the spectrum are referenced as a percentage of the base peak and referred to as 8
relative abundance. This normalization of peak heights greatly aids in 9
identification of fragmentation pattern and therefore analyte identification. 10
11
The molecular ion peak corresponds to an analyte molecule that has not 12
undergone fragmentation. In Figure 6.1, the molecular ion peak is caused by the 13
[CH
3
OH]
+
ion and corresponds to m/z 32. The molecular ion peak is often 14
referred to as the M
+
ion. The molecular ion is used as a reference point in 15
identifying the other fragments. For example, the peak corresponding to m/z 15 16
is referred to as both M OH and M 17. 17
18
6.3 Identifying the Molecular Ion Peak 19
20
The molecular ion peak is both an important reference point and is integral 21
in identifying an unknown compound. While it may seem that the molecular ion 22
peak should be the most abundant peak in the spectrum, this is not the case for 23
the majority of compounds. Compounds like alcohols, nitrogen containing 24
organics, carboxylic acids, esters, and highly branched compounds may 25
completely lack a visible molecular ion. In these cases, it is critical that fragment 26
peaks are not mistakenly identified as the molecular ion peak in order to avoid 27
misidentification of an analyte. Obtaining a chemical ionization spectrum can 28
assist in correctly identifying the molecular ion (Section 5.2.1.2b). 29
6
1
Even without a CI spectrum of the compound, other rules can assist in 2
ruling out potential masses as the molecular ion. The nitrogen rule is one 3
valuable tool for identifying the molecular ion. This rule indicates that if a 4
molecular ion has an odd mass it must have an odd number of nitrogen and that 5
a molecular ion with an even mass must lack nitrogen atoms or contain an even 6
number of them. Since the majority of organic compounds analyzed with the 7
GC-MS contain either zero or one nitrogen atom, the rule practically states an 8
odd molecular ion is attributed to a single nitrogen and an even molecular ion 9
indicates the sample lacks nitrogen (Figure 6.2). This rule only applies to 10
compounds that contain carbon, hydrogen, nitrogen, oxygen, sulfur, halogens, 11
and a few other less common elements. Since the majority of organic 12
compounds that are analyzed using the GC-MS are made up of these elements, 13
this stipulation is practically ignored. 14
15
16
Figure 6.2 The Nitrogen Rule - The mass spectrum of N,N-dimethyl-ethanamine 17
illustrates the presence of an odd molecular ion and even fragments. 18
7
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 1
permission from NIST. 2
3
This rule is a result of nitrogens unique property. Nitrogen has an even 4
atomic mass but bonds with three other atoms in its most stable form. Other 5
atoms that have even molecular weights like carbon, oxygen, and sulfur bond 6
with an even number of other atoms. Atoms that bond with an odd number of 7
other atoms like hydrogen, chlorine, bromine, and iodine have odd molecular 8
weights. This rule is invaluable when a chemist knows that a compound lacks 9
nitrogen. This can occur if a sample is prepared from a synthesis whose 10
products and solvents lack nitrogen atoms. In this case, any odd peak cannot be 11
attributed to the molecular ion of the analyzed compound. 12
13
Most fractionation excluding rearrangements (Section 6.6) occurs when a 14
single bond is broken. The nitrogen rule indicates that when a molecule with an 15
even mass produces a fragment by breaking a single bond, the fragment will 16
have an odd mass. When the samples mass is odd, fragmentation via a similar 17
pathway will give an even fragment as long as the nitrogen is still contained in 18
the observed fragment. Since this is generally the observed trend (See 19
Stevensons Rule Section 2.6), analyzing the major fragments can help 20
determine if the molecular ion should be even or odd. Practically, if the major 21
fragments are mostly odd, the molecular ion is likely even and contains no 22
nitrogen. If the major fragments are even, the molecular ion is likely odd and 23
contains one nitrogen atom as shown in Figure 6.3. 24
25
8
1
Figure 6.3. The Use of the Nitrogen Rule in Determining the Molecular Ion - 2
Should the faint peak at m/z 60 be attributed to the presence of C
13
or is it the 3
molecular ion? The presence of the base peak at 45 in combination with our 4
knowledge about the nitrogen rule suggests that the peak at m/z 60 is likely the 5
molecular ion because even molecular ion usually produce odd molecular 6
fragments by breaking single bonds. Given this spectrum is of Isopropyl alcohol, 7
our deduction is correct although chemical ionization techniques could verify the 8
molecular mass of the sample. 9
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 10
permission from NIST. 11
12
Since molecular ions fragment in predictable ways, the presence of certain 13
fragmentation peaks can suggest that a particular peak is the molecular ion. The 14
observed fragments must be able to be attributed to logical losses. The 15
existence of a M - 15 peak from the loss of CH
3
, a M - 18 peak from the loss of 16
H
2
0, or a M - 31 from the loss of OCH
3
are a few examples of these logical 17
fragments. 18
19
9
The opposite is true for fragments that are not logical. These peaks 1
suggest that a particular peak is not the molecular ion. Some illogical 2
fragmentation peaks include peaks that is 3 to 14 mass units away from the peak 3
suggest that the identified peak is likely not the molecular ion peak. The loss of 4
fragments of mass units 1-3 can result from the loss of up to three hydrogen 5
atoms. From 14 to 18, multiple peaks can be explained from the loss of CH
3
, 6
oxygen, a hydroxide ion, or water. The loss of fragments from the 19-25 range is 7
also unlikely except in the case of fluorinated compounds which produce M - 19 8
(loss of F) and M - 20 (loss of HF). 9
10
The molecular ion is difficult to identify with chemical ionization because 11
there is no definitive test. While these patterns can greatly assist in identifying 12
the molecular ion, they should not be trusted as confirmatory. Complex 13
rearrangements can potentially result in the misidentification of the molecular ion. 14
As a result, it is good practice to double check with a soft ionization technique 15
such as chemical ionization when in doubt of the identity of the molecular ion. 16
17
2.4 Use of the Molecular Ion 18
19
Once the identity of the molecular ion has been determined much can be 20
learned about the compound. One extremely valuable piece of information that 21
can be determined from a high resolution mass spectrometer is the molecular 22
formula of an unknown analyte. If a molecular ion was identified to be at m/z 80 23
on an instrument with unit resolution little could be determined about the 24
molecular formula. For example, some of the many possible molecular formulas 25
include C
4
H
4
N
2
(80.0375), C
5
H
4
O(80.0262), and C
6
H
8
(80.0626). A high 26
resolution instrument measurement of this peak at 80.0372 .0005 would 27
indicate that the empirical formula is C
4
H
4
N
2
. Extensive tables and computer 28
programs are used to perform this technique of a routine basis. 29
30
10
Once the molecular formula is known, it becomes possible to determine 1
the degree of unsaturation. This allows the analyst to know the number of pi 2
bonds and rings that are in their structure. The elements of unsaturation can be 3
computed by using the following equation. 4
5
6
7
where H is the number of H atoms, X is the number of halogen atoms (F, Cl, Br, 8
and I), and N is the number of nitrogen atoms in the chemical formula. As the 9
equation indicates, the number of oxygen atoms does not affect the degree of 10
unsaturation. By using this equation the molecular formula C
4
H
4
N
2
has four 11
degrees of unsaturation. The combination of the molecular formula with the 12
degrees of unsaturation is important tools in identifying a particular compound. 13
14
The molecular ion along with other information from IR and NMR spectra 15
can allow the identity of an unknown to be determined. If all three techniques 16
can be utilized on a sample, the strengths of each allow for the easiest 17
identification. Since IR identifies the unknowns functional groups, the mass of 18
these groups are first subtracted from the mass of the molecular ion. This mass 19
frequently represents the mass of the carbon and hydrogen contained in a 20
sample. Taking this number and dividing by twelve will the number of carbon 21
atoms and a fraction representing the number of hydrogen atoms. It is important 22
to not blindly trust this method. If the molecular ion minus the mass of the 23
functional groups gives 85, dividing by 12 would give a molecular formula of C
7
H. 24
Attempting to create this molecule will quickly indicate that a more logical 25
molecular formula would be C
6
H
13
. 26
27
For the example unknown analyte illustrated in Figure 2.4, the IR 28
spectrum indicates the compounds functional groups. The sharp peak observed 29
at around 1710 cm
-1
indicates the presence of a carbonyl group. The large round 30
11
peak centered around 3000 cm
-1
suggests that the compound is a carboxylic 1
acid. The large peak slightly above 1600 cm
-1
could indicate that the unknown 2
contains an alkene or even possibly an imine. 3
4
5
6
12
1
Figure 2.4 IR, Mass Spectrum, and NMR of an Unknown Analyte 2
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 3
permission from NIST. 4
5
After talking a mass spectrum of the compound, it is necessary to identify 6
the molecular ion. We can identify the peak at 86 to be the molecular ion using 7
the nitrogen rule (discussed above). Because its major fragments are both odd, 8
69 and 41, it is reasonable that its molecular ion should be even. The peaks at 9
85 (loss of H) and 71 (loss of CH
3
) can be explained in a logical fashion further 10
confirming the m/z 86 peak as the molecular ion. 11
12
From above, the compounds known mass is 86, thus, we can confirm that 13
the compound is not an imine because it has an even molecular weight indicating 14
that it does not contain an odd number of nitrogen atoms. Now it becomes 15
possible to identify something about the carbon backbone of the atom. By 16
subtracting the mass of the carboxylic acid functional group (COOH) a mass of 17
41 is obtained. The IR spectrum indicates that the remainder of the molecule is 18
likely only made up of carbon and hydrogen. This allows the analyst to deduce 19
that the rest of the molecule is made up of 3 more carbon atoms and 5 hydrogen 20
atoms. From taking these two easy measurements, one is able to determine that 21
this compounds molecular formula is C
3
H
5
COOH. From this molecular formula 22
13
we can determine that the degree of unsaturation is two. One degree of 1
unsaturation is attributed to the acid functional group while the other is a double 2
bond since a three carbon ring is extremely unlikely. 3
4
While both the IR and MS are able to determine a great deal about a 5
compounds identity, NMR is necessary to identify this compound. The peak 6
below 12 ppm is a result of the hydrogen on the carboxylic acid group. The 7
doublet peaks 6.3 and 5.7 ppm are split by approximately 1.4 Hz which indicates 8
that they are geminal protons. The fact that the peaks are doublets indicates that 9
there are only two hydrogen atoms connected to the vinyl group. The presence 10
of a methyl group is indicated by the peak at 2 ppm. As a result the unknown 11
compound is methacrylic acid whose structure is shown below. 12
13
14
15
While utilizing IR and NMR, in combination with the mass spectra, made 16
the identification of this compound relatively simple, these tools are not always 17
available. If the analyte of interest is in a complex mixture or there is only a small 18
concentration or quanity (parts per million), both IR and NMR are not effective 19
tools. As a result, it is necessary to be able to identify as much information is 20
possible from the mass spectrum alone. The rest of this chapter will be devoted 21
to such a task by observing common fractionation trends in various types of 22
compounds. 23
24
2.5 Identification of Analytes using Isotopic Ratios 25
26
Since the majority of elements have two or more isotopes, the ratio of 27
these isotopes can be a powerful tool in deriving the composition of unknown 28
samples. Prominent peaks will have a smaller peak one mass unit higher than 29
14
the prominent peak due to the presence of one
13
C in some of the sample 1
molecules. Background noise and a lack of resolution in the majority of mass 2
spectrometers prevent the ratio of various isotopes from being an identification 3
technique for all compounds. 4
5
However, some isotopes are so prominent that they can easily be 6
observed with a quadrupole mass spectrophotometer with unit resolution. 7
Chlorine, bromine, and sulfur can all be identified by their isotopic rations. Their 8
exact isotopic ratios are summarized in Table 2.1. Compounds containing 9
chlorine have a M+2 peak that is 25% of the molecular ion (Figure 2.5a). 10
Compounds containing bromine have a M+2 peak that is approximately the same 11
height as a M
+
peak (Figure 2.5b). Compounds containing sulfur have an 12
unusually large M+2 (Figure 2.5c). 13
14
Table 2.1 Isotopic Abundances of Common Elements 15
Element M
+
M + 1 M + 2
hydrogen
1
H 100.0%
carbon
12
C 98.9%
13
C 1.1%
nitrogen
14
N 99.6%
15
N 0.4%
oxygen
16
O 99.8%
18
O 0.2%
sulfur
32
S 95.0%
33
S 0.8%
34
S 4.2%
chlorine
35
Cl 75.5%
37
Cl 24.5 %
bromine
79
Br 50.5%
81
Br 49.5%
iodine
127
I 100.0%





16
15
1
Figure 2.5A of Ethyl Chloride illustrates the presence of a M++2 peak that is 2
about 25% of the M+peak. 3
4
5
Figure 2.5B of bromoethane has a characteristic M+2 peak that has a similar 6
intensity as the M+peak. 7
16
1
Figure 2.5C of 2-Propanethiol contains a larger than usual M+2 peak, a pattern 2
observable in sulfur containing compounds. 3
4
Figure 2.5 Isotopic Identification 5
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 6
permission from NIST. 7
8
Compounds can also contain any combination of multiple chlorine and 9
bromine atoms. These samples will produce distinct peaks due to the various 10
combinations of the isotopes. A compound containing two bromines will have a 11
M+2 peak twice the size of the M+peak and a M+4 peak the same size as the 12
M+peak (Figure 2.6a). A compound containing two chlorines will have a M+2 13
peak that is two thirds the size of the M
+
peak and a M+4 peak that is ten percent 14
of the molecular ion (Figure 2.6b). 15
16
17
1
Figure 2.6A shows the mass spectrum the dibrominated compound 1,3- 2
dibromopropane. 3
4
Figure 2.6B shows the mass spectrum of the dichlorinated compound 1,2- 5
dichloroethane. 6
7
18
Figure 2.6 Polybrominated and Polychlorinated Compounds 1
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 2
permission from NIST. 3
4
Iodine is more difficult to detect because it is one of the few compounds 5
that is monoisotopic. Despite this fact, the large atomic mass of iodine allows for 6
its identification. The combination of a peak at m/z 127 (I
+
) and a large gap of 7
127 mass units between fragments containing iodine and fragments lacking in 8
iodine allows for the compounds identification (Figure 2.7). 9
10
11
12
Figure 2.7 Mass Spectrum Containing Iodine 13
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 14
permission from NIST. 15
16
2.6 Fragmentation 17
18
19
While the molecular ion is one of the most important peaks in the spectra, 1
it is also important to gain information from the peaks that are a result of 2
fragmentation. The goal of interpreting mass spectra is identifying the structure 3
of the molecular ion by examining pieces (fragments) of the original molecule. 4
The frequency and size of the fragments is dependent on the structure and bond 5
energy of the sample molecule. This property has resulted in the creation of 6
unique and reproducible spectrum for a wide variety of compounds. 7
8
Before fragmentation can be discussed, it is necessary to develop a new 9
notation because the cation fragments that will be encountered are not present in 10
other branches of chemistry due to their high reactivity. The presence of the 11
vacuum in the instrument prevents collisions with other molecules allowing these 12
reactive cations to exist. The academic convention for notation is to either 13
represent the charge as a delocalized one (Example A below) or localized it on 14
either a ! bond (Example B) or on a heteroatom (Example C). 15
16
a) b) c) 17
18
The process that creates these observed fragments is the result with their 19
interaction with the energy released from the source. This energy both removes 20
a single electron, while the excess energy is distributed over various degrees of 21
freedom. This distribution converts electronic energy into electronic, rotational, 22
and vibration energy. The molecular ion is created when the sample molecule 23
returns to its ground state via a relaxation. Other times this energy exceeds the 24
activation energy of fragmentation and this energy is released via the breaking of 25
bonds. 26
27
e- +R R ! [R R]
+
+2e- 28
[R R]
+
! R

+R
+
29
[R R]
+
! R
+
+R

30
20
1
The fragmentation of a single bond can produce two peaks, one from R
+
2
and the other from R
+
, since the instrument can only detect the positive ion. 3
According to Stevensons rule, if two fragments are in competition to produce a 4
cation, the fragment with the lowest ionization energy will be formed more 5
frequently (Figure 2.8). 6
7
8
Figure 2.8 An Illustration of Stevensons Rule 9
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 10
permission from NIST. 11
12
The fragmentation of a bond can proceed through two pathways, either 13
homolytic or heterolytic cleavage. In the heterolytic cleavage, a pair of electrons 14
move towards the charged site as illustrated by the double headed arrow 15
producing a cation and a radical. 16
21
1
2
3
The fragmentation produced by a hemolytic cleavage results from the 4
movement of single electrons. 5
6
7
8
For simplicity, usually only one set of arrows is drawn to illustrate the 9
movement of electrons. 10
11
12
13
These fragmentation patterns are usually the result of a functional group 14
contained in the compound. As a result, the bonds that typically break are either 15
located one, or two carbon atoms away from the functional group. These carbon 16
atoms are referred to as the " and # atoms. 17
18
19
20
The bond between the functional group Y and the " carbon is called the " 21
bond and the bond between the " and # carbons is the # bond. 22
23
2.7 Rearrangements 24
25
22
Some fragments are the result of the cleavage of multiple bonds. The 1
removal of water from an alcohol is only one example. The nitrogen rule (Figure 2
2.3) is helpful in identifying peaks that are produced via a rearrangement. If a 3
molecular ion has an even molecular weight then generally peaks of even 4
molecular weight were created from a rearrangement. If a molecule has an odd 5
molecular weight, then its rearrangements peaks will also be odd. 6
7
One rearrangement is the loss of water from a primary alcohol. The 8
mechanism is illustrated with butanol. 9
10
11
12
These rearrangements are favored because the low energy transitions 13
help stabilize the products. Other rearrangements such as the McLafferty 14
rearrangement will be explored in greater detail in the following sections. 15
16
2.8 Identification of Compounds 17
18
The ability to identify unknown samples is one of the most powerful uses 19
of a mass spectrometer. This, however, requires an understanding of 20
fractionation patterns for commonly encountered compounds. The following 21
trends are only applicable to electron impact with a source at 70 eV. These 22
trends are not comprehensive but are rather a selection of common fragments 23
that are most useful in properly identifying common types of organic chemicals. 24
The actual likelihood of fragmentation is related to the activation energy of 25
the reaction, the ability for rearrangements to occur, and the stability of the 26
products. Trends that were observed in organic chemistry are helpful in 27
predicting fragmentation patterns. Thinking about the stability of the products as 28
23
more or less stable cations and radicals is not entirely theoretically accurate but 1
is usually a good, practical way to predict the spectrum of a molecule. 2
3
2.9 Fragmentation of Hydrocarbons 4
5
There are two types of hydrocarbons that are analyzed with the GC-MS. 6
One is long chain hydrocarbons and the other is the hydrocarbon portion of 7
molecules containing other functional groups. Identifying the structure of these 8
hydrocarbons can be difficult since rearrangements that are not easily explained 9
are frequently observed. It is especially important to utilize reference compounds 10
and GC retention times whenever possible to confirm the identity of the 11
compound. 12
13
2.9.1 Fragmentation of Straight Chain Alkanes 14
15
Straight chain alkanes always produce a molecular ion even in long chain 16
compounds where the molecular ion is usually faint. The base peak in the 17
spectra is usually the peak at m/z 57 corresponding to the C
4
H
9
carbocation 18
surrounded by other smaller peaks due to the rearrangement of hydrogen atoms. 19
These groups are separated by 14 mass units resulting from the loss of another 20
CH
2
group. The largest peak in each cluster is caused by the loss of (CH
2
)
n
CH
3
21
resulting in a fragment of molecular formula C
m
H
2m+1
. The subsequent fragments 22
after the C
4
peak decrease in an exponential fashion to a minimum at M-C
2
H
5
. 23
The M CH
3
peak is weak in smaller compounds and absent in long chain 24
compounds due to the relative instability of the methyl radical. The molecular ion 25
is the unique identifiable peak in straight chain alkanes longer than eight carbon 26
atoms. 27
28
24
1
2
Figure 2.9 Fragmentation of a straight Chain Alkane 3
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 4
permission from NIST. 5
6
The example illustrated in Figure 2.9 illustrates these common trends 7
discussed above. The prominent peaks at C
m
H
2m+1
combined with the decaying 8
intensity of these peaks indicates this compound is an alkane. The base peak at 9
m/z 57 caused by C
4
H
9
is further confirmation that there is a lack of other 10
functional groups causing other prominent fragments. The molecular ion at m/z 11
170 indicates that this compound is dodecane. 12
13
2.9.2 Fragmentation of Branched Alkanes 14
15
Branched alkanes have a smaller molecular ion that at times may be 16
absent in highly branched compounds. In larger compounds branched alkanes 17
contain peaks at C
m
H
2m+1
, similar to straight chain alkanes. They are 18
distinguished by the lack of the smooth exponential decay beginning at the C
3
or 19
25
C
4
carbon (Figure 2.9). This is caused by the increased frequency of 1
fractionation at the branch since it results in a secondary rather than a primary 2
carbocation and is hence favored. The loss of the largest alkyl fragment at the 3
brancing cite is favored because it helps to stabilize the radical. 4
5
This mass spectrum of a C
12
alkane (determined from the molecular ion by CI at 6
m/z 170) lacks the exponential decay seen in Figure 2.9 indicating the chain is 7
branched. The intensity of the peak at m/z 71 indicates a favored C
5
fragment 8
and the fragment at m/z 127 indicates a favored C
9
fragment suggesting that a 9
methyl group on fourth carbon. 10
26
1
Figure 2.10 Fragmentation of a Branched Alkane 2
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 3
permission from NIST. 4
5
The fragmentation at the branching point is often accompanied by 6
hydrogen rearrangement causing the C
n
H
2n
peak to be more prominent and 7
sometimes larger than C
n
H
2n+1
peak. 8
Identifying branched alkanes in organic compounds that contain another 9
functional group is also an important task. The alkane portion of these molecules 10
is usually smaller and is more goverened by the stability of the produced radical 11
and cation. Since an ethyl radical is more unstable than a methyl radical, the 12
methyl radical will occur less frequently. Similarly, tertiary carbocations are more 13
stable than secondary, which are more stable than primary. As the alkane 14
portion of any molecule becomes larger, the presence of the C
n
H
2n+1
peaks 15
become more prominent. 16
17
2.9.3 Fragmentation of Cyclic Alkanes 18
27
1
The ring structure of cyclic alkanes increases the intensity of the molecular 2
ion. Its stability also increases the likelihood that side chains will fragment at the 3
$ bond to the ring. The fragmentation of the cyclic structure is usually most often 4
caused by the loss of more than two carbon atoms. The loss of a methyl radical 5
occurs less frequently because of the instability of the methyl radical in 6
comparison to the neutral ethylene molecule at M-28 or an ethyl radical at M-29. 7
8
9
10
Figure 2.11 Fragmentation of a Cyclic Alkane 11
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 12
permission from NIST. 13
14
28
2.9.4 Fragmentation of Alkenes 1
2
The molecular ion of alkenes, is usually distinct especially in compounds 3
containing multiple double bonds. Alkene fragments, like alkane fragments are 4
situated in clusters 14 units apart. In alkenes, the C
n
H
2n-1
and C
n
H
2n
peaks are 5
more intense than the C
n
H
2n+1
peak of alkanes. 6
7
The presence of double bonds also allows for the production of 8
resonance-stabilized cations. Allylic cleavage results in an allylic cation. 9
10
11
12
Determining the position of the double bonds in the sample molecule is 13
especially difficult and usually requires reference spectra because of double 14
bonds migration. Cyclic alkenes also undergo a retro-Diels-Alder fragmentation 15
by the following mechanism. 16
17
18
19
2.9.5 Fragmentation of Aromatics 20
21
The presence of an aromatic ring in a compound results in a prominent 22
molecular ion. A common peak at M 1 results from the cleavage of a hydrogen 23
molecule from the benzene ring. Alkyl substituted benzene rings result in a 24
prominent peak at m/z 91 (Figure 2.12). In most cases, the peak at m/z 91 is the 25
result of a tropylium ion caused by the following rearrangement. 26
27
29
1
2
The peak observed in most aromatic compounds at m/z 65 results from 3
the elimination of an acetylene molecule from the tropylium ion. 4
5
6
7
Benzene rings with highly branched substituted groups produce fragments 8
larger than m/z 91 by intervals of 14 units. The largest of these peaks will result 9
in a highly substituted cations and a large radical, like a simpler branched 10
alkanes. The fragment at m/z 105 in Figure 2.12 is relatively small since it 11
produces a primary carbocation and an unstable methyl radical. Substituted 12
benzene rings also first undergo $ cleavage followed by hydrogen rearrangement 13
producing a grouping of peaks at m/z 77 from C
6
H
5
+
, m/z 78 from C
6
H
6
+
, and m/z 14
79 from C
6
H
7
+
. 15
16
30
1
Figure 2.12 Fragmentation of an Aromatic 2
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 3
permission from NIST. 4
5
Side chains with a more than two carbon atoms create a peak at m/z 92 6
(Figure 2.12). Unbranched groups result in a more prevalent peak than do 7
branched groups. 8
9
10
11
2.10 Fragmentation of Alcohols 12
13
The molecular ion of alcohols is usually small and sometimes 14
undetectable especially in tertiary alcohols. The identification of the molecular 15
31
ion is complicated by the prevalence of a M-1 peak caused by the loss of a single 1
hydrogen from the " carbon in primary and secondary alcohols. 2
3
Alcohols also frequently cleave to give resonance stabilized cations due to 4
the breaking of the # bond. As a result of this clevage, primary alcohols show a 5
prominent peak at m/z 31 (Figure 2.13). 6
7
8
9
The presence of a m/z 31 peak is not confirmation of a primary alcohol. It 10
is necessary for the peak to be relatively large in comparison to other peaks in 11
the spectrum. This is because secondary alcohols and sometimes even tertiary 12
alcohols can undergo a rearrangement resulting in a peak at m/z 31. 13
14
15
16
32
Alcohols also frequently undergo the rearrangement described in Section 1
2.7 resulting in a M-18 peak from the loss of water. This peak is most easily 2
visible in primary alcohols but can be found in secondary and tertiary alcohols as 3
well. Primary alcohols also can lose both water and an alkene. 4
5
6
7
Primary alcohols also produce a M-2 peak caused by R-CH=O+and M-3 8
attributed to R-C%O
+
. Alcohols with carbon chains containing methyl groups 9
frequently loose both the methyl group and water at M-33. 10
11
12
Figure 2.13A can be identified as an alcohol because of the characteristic peak 13
at M-H
2
O, M - 33, and m/z 31. The peak at m/z 31 can attributed to a primary 14
alcohol because it is one of the larger peaks in the spectrum. 15
33
1
Figure 2.13B has a small peak at m/z 31 but the base peak in the spectrum 2
indicates that this alcohol is not a primary alcohol. The presence of a M - Et and 3
M - CH3 peak indicates that this four carbon alcohol (determined from its 4
molecular mass) is the secondary alcohol 2-butanol. 5
6
34
Figure 2.13C illustrates trends common to tertiary alcohols. The spectrum is 1
easily discernable since the single prevalent peak is characterized by M - CH
3
. 2
The lack of a molecular ion helps to confirm the spectrum of a 3
tertiary alcohol. 4
5
Figure 2.13 Fragmentation of Three Alcohols 6
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 7
permission from NIST. 8
9
Cyclic alcohols fragment similar to straight chain alcohols in that they give 10
a M-1 peak from the loss of hydrogen and an M-18 peak from the loss of water. 11
They also create a peak at m/z 57 via a complex ring cleavage. 12
13
Aromatic alcohols, unlike other alcohols, have a prominent molecular ion 14
peak due to the stability of the aromatic group. Phenols usually give a weaker 15
peak at m/z 77 attributed to a rearrangement and can be identified by two peaks 16
at M CO and M - COH. 17
18
19
20
21
2.11 Fragmentation of Ketones and Aldehydes 22
23
35
Both ketones and aldehydes give prominent molecular ion peaks though 1
the M+peak is more prominent in ketones. The majority of compounds in these 2
categories undergo an important rearrangement, the McLafferty rearrangement. 3
4
5
6
This rearrangement is mediated by the p systems of the carbonyle group 7
but can occur in other p systems such as in nitriles (Section 2.17). The only 8
ketones and aldehydes that do not undergo this rearrangement lack a three- 9
carbon side chain allowing for the necessary hydrogen donation. 10
11
2.11.1 Ketones 12
13
One major fragment of ketones is the creation of the resonance stabilized 14
acylium ion resulting from the cleavage of the $ bond. The base peak in the 15
spectrum is usually caused by the removal of the larger alkyl group since it forms 16
a more stable radical illustrated by 4-Octanone (Figure 2.14). 17
18
36
1
2
While ketones undergo a single McLafferty rearrangement 3
described above, they also undergo a subsequent McLafferty rearrangement. 4
5
37
1
2
3
The second rearrangement is mediated by the p system of the alkene 4
group. The ketone functional group is often easily discernable due to the 5
prevalent fragments and rearrangements described above. The configuration of 6
the carbon structure can be difficult to discern. Reduction of the carbonyl group 7
to a methylene group is commonly performed to determine the complete 8
structure of the molecule. 9
10
38
1
Figure 2.14 Fragmentation of a Ketone 2
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 3
permission from NIST. 4
5
The base peak in Figure 2.14 is the result of a McLafferty rearrangement 6
and an " cleavage. 7
8
9
10
2.11.2 Fragmentation of Cyclic Ketones 11
12
Cyclic ketones major cleavage is also at the $ bond. Due to the ring 13
structure, this cleavage will be detected as the molecular ion unless another 14
39
bond is broken. Saturated cyclic ketones produce a fragment at m/z 55 1
illustrated by cyclohexanone. 2
3
4
5
In cyclohexanone, this peak is the base peak. In absence of a reference 6
standard, cyclic ketones are difficult to identify given the difficulty explained 7
earlier in determining the composition of the alkyl portion of the ketone. 8
9
2.11.3 Fragmentation of Aromatic Ketones 10
11
Aromatic ketones create fragments via almost identical pathways as 12
aliphatic ketones. One prominent peak, and usually the base peak, is the result 13
of the cleavage of the less stable alkyl fragment resulting in the ArC%O fragment 14
located at m/z 105. The alpha cleavage resulting in a benzyl radical is 15
infrequent given the stability of the competing reaction (Figure 2.15). 16
17
40
1
2
3
The cleavage of the bond " to the aromatic group results in a peak at m/z 4
77. 5
6
7
8
Further fragmentation results in a peak at m/z 55 after the loss of HC%CH 9
Some aromatic ketones undergo the typical McLafferty rearrangement if the 10
other alkyl component contains an abstractable hydrogen atom. 11
12
41
1
2
3
4
Figure 2.15 Fragmentation of a Cyclic Ketone 5
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 6
permission from NIST. 7
8
2.11.4 Fragmentation of Aldehydes 9
10
The major peaks observed in spectrums of aldehyde are the result of the 11
same $ cleavage as in ketones. This fragmentation results in an M-1 peak and a 12
peak at M-R from the COH
+
ion. The presence of an M-1 peak helps to identify 13
the aldehyde but the hydrocarbon rearrangement at C
2
H
5
prevents the M-R (m/z 14
29) peak from being truly useful. Another prominent peak is the McLafferty 15
42
rearrangement located at m/z 44. The only aldehydes that do not contain this 1
peak are ones that lack an the necessary hydrogen atom for this rearrangement. 2
3
Straight chain aldehydes have unique features that help in identification. 4
These compounds will have a M - 18 fragment from the loose of water, M 28 5
from the loss of ethylene, M 43 loss of CH2=CHO and M 44 from the loss of 6
CH2=CHOH (Figure 2.16). 7
8
9
Figure 2.16 Fragmentation of an Aldehyde 10
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 11
permission from NIST. 12
13
The patterns resulting from aromatic ketones are almost identical to those 14
governing aromatic aldehydes. The characteristic molecular ion is accompanied 15
by a M-1 peak from the loss of hydrogen. The ArC%O fragment looses CO to 16
form the phenyl ion at m/z 77 that further degrades to give a peak at m/z 51. 17
18
2.12 Fragmentation of Carboxylic Acids 19
43
1
The molecular ion of straight chain carboxylic acids is weak but usually 2
present. The prominent and often times the base peak results from the 3
McLafferty rearrangement. 4
5
6
7
Short chain carboxylic acids give prevalent peaks at M OH and M 8
CO
2
H. In larger carboxylic acids these peaks are less prevalent. Long chain 9
carboxylic acids are better identified by the fragments at C
n
H
2n-1
O
2
(Figure2.17). 10
There is also the presence of the hydrocarbon fragment at C
m
H
2m+1
illustrated in 11
Section 2.9. 12
13
14
Figure 2.17 Fragmentation of a Carboxylic Acid 15
44
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 1
permission from NIST. 2
3
Aromatic acids have a more prominent molecular ion peak but undergo 4
similar fractionation to short chain hydrocarbons. They produce large peaks at M 5
OH and M CO
2
H. Aromatic acids can also loose water if an ortho group 6
contains an abstractable hydrogen atom. 7
8
9
10
2.13 Fragmentation of Ethers 11
12
The molecular ion peak is usually weak in ethers. The oxygem atom 13
mediates the major fragment and creates a # cleavage that results in a 14
resonance stabilized cation. This peaks is prominent and sometimes are the 15
base peak. 16
17
18
19
The fragment can also undergo a subsequent rearrangement which 20
typically creates the base peak when the $ carbon is substituted. 21
22
45
1
2
Ethers also produce prominent alkyl fragments when the CO bond (" 3
bond) is broken and the fragment containing oxygen is a radical. 4
5
6
7
8
Figure 2.18 Fragmentation of an Ether 9
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 10
permission from NIST. 11
12
46
The base peak in Figure 2.18 is the result of both a # cleavage and the 1
above rearrangement. 2
3
4
5
Aromatic ethers have a slightly different pattern of fragmentation. They 6
produce prominent molecular ions due to the stability of the benzene ring. The 7
major fractionation occurs at the # bond to the aromatic ring. This fragment can 8
decompose further with the loss of CO. 9
10
11
12
Aromatic ethers also cleave at the bond $ to the ring to create a peak at 13
m/z 78 and 77 due to hydrogen migration. 14
15
16
17
47
When the alkyl portion of the sample is larger than two carbons, the # is 1
accompanied by hydrogen migration caused by the presence of the aromatic 2
group. This cleavage results in a peak located at m/z 94. 3
4
5
6
2.14 Fragmentation of Esters 7
8
The molecular ion peak of straight chain esters is sometimes discernable. 9
A prevalent peak and often the base peak results from the familiar McLafferty 10
rearrangement. The size of the alcohol that formed the ester and the presence 11
of $ substituents can normally be discerned by the mass of these two peaks. 12
13
14
15
The cleavage of the above bonds results in other fragments, however 16
these peaks are too small to be of great significance. For example, hexanoic 17
acid methyl ester produces the following fragments. 18
19
20
48
1
2
The resonance stabilized ion gives a discernable peak for 3
almost all esters. The R
+
ion is prominent in short chain esters but is barely 4
visible in esters with more than six carbon atoms. For hexanoic acid methyl 5
ester, the R
+
ion is only 9.5% of the base peak. 6
7
Until this point, this chapter has covered individual functional groups in 8
isolation. Since esters have both an alcohol and an acid component, 9
fractionation patterns can be observed from both of these types of compounds. 10
The prevalence of the fragments described earlier is dependent on the size of 11
each part of the ester. The increased size of each portion results in a unique 12
rearrangement. 13
14
When the acid portion is the major component like in hexanoic acid methyl 15
ester, the fractionation pattern is partially characterized by typical acid peaks. 16
For these straight chain esters, cleavage of successive carbon atoms gives an 17
alkyl fragment and a fragment containing oxygen. This pattern results in the 18
familiar grouping of fragments spaced 14 units apart with the largest fragment in 19
the cluster resulting from the C
n
H
2n-1
O
2
ion (Figure 2.19). 20
21
49
1
Figure 2.19 Fragmentation of an Ester 2
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 3
permission from NIST. 4
5
The base peak in Figure 2.19 is a result of the McLafferty rearrangement. 6
7
8
9
When the alcohol portion of the ester is the prominent portion of the ester, 10
fragments similar to that of an alcohol is observed. These esters will loose a 11
molecule of acid like alcohols loose a molecule of water. 12
13
50
1
2
Like alcohols, the prevalence of this rearrangement is so frequent that the 3
molecular ion is normally absent from the spectra. These long chain alcohols will 4
also loose the alkyl fragment from the alcohol accompanied by two hydrogen 5
migration. 6
7
8
9
10
In aromatic esters, the molecular ion is prominent due to the aromatic 11
group. There are two distinctive types of aromatic esters that have their own 12
unique fractionation patterns. Esters synthesized from aromatic acids mostly 13
undergo $ cleavages. 14
15
51
1
2
The loss of OR results in the base peak because of the multiple 3
resonance forms stabilizing the cation. When the alkyl portion of the alcohol 4
becomes longer, the McLafferty rearrangement and the loss of R with two 5
hydrogen migrations explained above is more favorable. Increasing the size of 6
the R chain will cause the alkyl portion to retain the charge. 7
8
9
10
The presence of an aromatic group in the alcohol that formed the ester 11
results in the creation of the CH
3
C%O
+
ion. These esters also undergo a 12
rearrangement that results in the loss of a ketene molecule. 13
14
15
16
2.15 Fragmentation of Amines 17
18
The presence of nitrogen in an amine can be detected by its odd 19
molecular weight and the even fragments that it produces (Section 2.3). Often 20
52
times the presence of the molecular ion in longer straight chain amines is not 1
detectable. In these cases, chemical ionization techniques are often used in 2
determining the molecular mass in order to determine the presence of nitrogen. 3
4
The base peak in most amines results from the cleavage of the # bond. 5
The loss of the largest branch (R) is preferred because the larger alkly fragment 6
stabilizes the produced radical. 7
8
9
10
Like alcohols, if the $ carbon is bonded to a hydrogen atom, a M H peak 11
is usually visible. In primary amines with an unbranched $ carbon, cleavage of 12
the # bond produces a peak at m/z 30. This peak is not conclusive proof of a 13
primary amine because secondary and tertiary amines undergo a rearrangement 14
similar to that of alcohols. 15
16
17
18
Amines also produce even fragments caused the cleavage of C C bonds 19
farther away from the functional group. The fragment containing the nitrogen 20
group usually retains the charge resulting at peaks characterized by C
n
H
2n+2
N 21
spaced at 14 units. There is also the less prevalent hydrocarbon pattern of 22
C
n
H
2n+1
, C
n
H
2n
, and C
n
H
2n-1
. 23
53
1
2
54
1
2
Figure 2.20 Fragmentation of Three Amines - The mass spectrum is easily 3
recognized as an amine due to its odd molecular ion and the presence of even 4
fragments. The base peak in each spectrum, due to b cleavage distinguishes 5
between the primary, secondary, and tertiary amine. 6
Spectra from the NIST/EPA/NIH Mass Spectral Library. Reprinted with 7
permission from NIST. 8
9
Cyclic amines produce a discernable molecular ion peak unless the $ 10
carbon is substituted. The loss of hydrogen from the $ carbon is also a 11
prominent peak. The ring is cleaved when the # bond is broken and subsequent 12
alkene molecules fragment from the remaining ring structure. 13
14
The molecular ion of an aromatic amine is expectedly intense. The loss of 15
the hydrogen atom bonded to the nitrogen gives a prominent peak at M 1. 16
Similar to ethers, the loss of HCN from the aniline ion produces peaks at C
5
H
6
17
and C
5
H
5
. Unlike ethers however, the hetero atom not the aromatic group 18
controls the major pathways of fractionation resulting in # cleavage. 19
55
1
2
3
2.16 Fractionation of Amides 4
5
The molecular ion of most straight chain amides is usually discernable 6
which allows the nitrogen atom to be identified. The fractionation pattern is 7
dependent on the length of the alkyl chain and the degree of substitution of the 8
nitrogen group. Primary amides give a prevalent peak from the McLafferty 9
rearrangement. 10
11
12
13
In primary amides that lack an abstractable hydrogen atom for the 14
McLafferty rearrangement, this accounts for the base peak. In the other amides 15
including secondary and tertiary, the base peak is created by the McLafferty 16
rearrangement at m/z 59. Primary amines also produce a peak at m/z 86 as a 17
result of the following cleavage. 18
19
56
1
2
When the alkyl groups bonded to the nitrogen are longer than two 3
carbons, another rearrangement is discernable. 4
5
6
7
Aromatic amides have a more prominent molecular ion peak. The 8
common fragments are characterized by the loss of NR
2
to form a resonance 9
stabilized cation followed by the subsequent loss of CO. 10
11
12
13
2.17 Fragmentation of Nitriles 14
15
The presence of the nitrogen can usually be identified by the odd 16
molecular weight according to the nitrogen rule (Section 2.3). This identification 17
technique is usually unable to identify a nitrile because these compounds lack a 18
molecular ion. The presence of a M-1 peak complicates the identification of the 19
57
molecular ion. This peak is formed by a loss of an " hydrogen to form a 1
resonance stabilized cation. 2
3
4
5
A prominent and frequent base peak is the result of the McLafferty 6
rearrangement at m/z 41 in compounds whose " carbon is not branched. This 7
peak, however, is unable to confirm that a compound is a nitrile because 8
hydrocarbon chains frequently form a peak at C
3
H
5
. 9
10
11
12
A unique peak at m/z 97 is characteristic of nitriles that contain a straight 13
chain of seven carbons or more. 14
15
16
17
2.18 Reviewing General Principals 18
19
58
After surveying common compounds encountered in organic chemistry 1
and their corresponding spectra we are able to make some generalizations about 2
fractionation patterns. When dealing with a compounds molecular ion; 3
4
1. The molecular ion of ethers, carboxylic acids, aldehydes, and nitrogen 5
containing molecules such as amides and nitriles can be very faint or 6
potentially absent. The molecular ion of alcohols and branched 7
compounds is almost always undetected. 8
2. Increasing a compounds size or the branching of the alkly portion will 9
decrease the intensity of the molecular ion. 10
3. Cyclic structures, elements of unsaturation, and aromatic groups increase 11
the intensity of the molecular ion. 12
13
Unknown compounds, in the absence of a reference standard, can often be 14
more difficult but not impossible to discern given the prevalence of multiple 15
fragments. The majority of compounds, however, will abide by the following 16
rules. 17
18
1. Resonance stabilized cations are favored because they help delocalize 19
the positive charge throughout the molecule. 20
2. The cleavage of bonds is favored at substituted carbon atoms that 21
produce the most stable cation. As a result, tertiary cations are favored 22
over secondary which are favored over primary which are more stable 23
than CH
3
+
. 24
3. The longest chain is eliminated most frequently because the greater 25
number of carbon atoms allows for the delocalization of the radical. 26
4. The combination of rule two and three can be good predictors of the 27
prevalence of fragments. Achieving a balance between the stability of the 28
radical and the cation produces the most prevalent peaks. The following 29
example illustrates this point. 30
59
1
2
5. The # bond to the heteroatom is frequently broken since the heteroatoms 3
non bonding electrons allow for resonance forms that stabilize the cation. 4
6. Rearrangements account for prominent peaks in the spectrum such as the 5
loss of water from an alcohol or the McLafferty rearrangement. 6
7
Besides having a general set of guidelines that govern general fractionation, it 8
is also important to be able to identify patterns that are indicative of particular 9
functional groups. As a result, a condensed table of the commonly observed 10
fragmentation patterns is listed in the table below. 11
12
Table 2.2 A Review of Common Fragmentation Patterns 13
Functional Group Observed Fragments M/Z Value
C
n
H
2n+1
43, 57, 71,
M CH
3
M 15
M CH
2
CH
3
M 29
Straight Chain
Alkanes
M CH
2
CH
2
CH
3
M 43
Branched Alkanes C
n
H
2n
Various
Cyclic Alkanes M H
2
C=CH
2
M 28
C
n
H
2n-1
Various
Alkenes
C
n
H
2n
Various
60

91

77
Aromatics

56
M H
2
O M 18
M (H
2
O & H
2
C=CH
2
) M 46 Alcohols
M (CH
3
& H
2
O) M 33
Primary Alcohols CH
2
OH 31

43 +R
Ketones

Various

44
COH 29
M H
2
O M 18
M H
2
C=CH
2
M 28
Aldehydes
M H
2
C=CHOH M 44
61

60
M OH M 17
M CO
2
H M 45
Carboxylic Acids
C
n
H
2n-1
O
2
73, 87,
# cleavage various
Ethers

various

74, 88,

various
R
+
various
Esters

various

various
Amines
C
n
H
2n+2
N 58, 72,
Primary Amines CH
2
NH
2
30
CH
2
NH
2
30 Amides

44
62

59


86
M H M - 1
Nitriles

41
1
2.19 Searchable Databases 2
3
The proliferation of databases and the number of compounds that they 4
contained has made the interpretation of spectra less important. These 5
databases cover over 200,000 compounds, the two most commonly used 6
databases are the one produced by the National Institute of Standards and 7
Technology (NIST) along with the Wiley Registry of Mass Spectral Data. Using 8
these databases can be of great assistance when performing routine analysis 9
and sometimes are the only way to positively identify a particular compound. 10
11
These databases cannot be the exclusive tool that chemists rely on to 12
interpret data. Like other tools, it is necessary to know when and how to use it. 13
Databases are a perfect tool for performing routine analysis when the analyte 14
and the reference standard have a high percent match. When the quality of the 15
match becomes low, it is necessary to access the validity of the database match. 16
It is also necessary to understand these fractionation patterns when performing 17
research especially when synthesizing new compounds that are not contained in 18
63
the published databases, and for which there is obviously no reference 1
compound. Only the combination of manual interpretation along with the usage 2
of a library can the composition of an unknown sample truly be discerned. 3
4
6.20 References 5
6
de Hoffmann, E.; Stroobant, V. Mass Spectrometry Principles and Applications, 7
2nd ed.; J ohn Wiley & Sons: New York, 2002. 8
9
Linstrom, P.; Mallard, W., Eds., NIST Chemistry WebBook, NIST. Standard 10
Reference database Number 69, J une 2005, National Institute of Standards and 11
Technology, Gaithersburg MD, 20899 (http://webbook.nist.gov). 12
13
McLafferty F.; Tureek, F. Interpretation of Mass Spectra, 4th ed.; University 14
Science Books: California, 1993. 15
16
Silverstein R.; Bassler, G.; Morrill, T. Mass Spectrometry. Spectrometric 17
Identification of Organic Compounds, 5th ed.; J ohn Wiley & Sons: New York, 18
1991; 3-89. 19
20
Wade, L. Infrared Spectroscopy and Mass Spectrometry. Organic Chemistry, 6th 21
ed.; Pearson Prentice Hall: New J ersey, 2006; 508-558. 22
23
24
6.21 Questions 25
26
Determine the identity of the products of the given reaction by using the mass spectra. 27
Then propose a likely mechanism of fragmentation for the numbered peaks. 28
29
30
31
1)
Cl
NH
3
+
32
a) 33
64
67
54
1
b) 2
43
56
3
4
2)
+ NBS
h!
5
6
65
1
3)
+ H
2
SO
4
H
2
O
2
3
4
5
6
4)
1. BH
3
THF
2. H
2
O
2
, OH
7
66
1

31
70
42
55
2
3
5)
MgBr
O
+
2. H
3
0
+
4
5

59
73
55
31
6
7
67
6)
+ HI
1
2
3
7)
1. O
3
2. (CH
3
)
2
S
4
5
68
a) 1
43
71
58
2
b) 3
4
5
8)
+
HCl
6
7
69

77
63
57
56
1
2
3
70
9)
OH
Cl
O
+
1
2
43
56
87
101
3
10)
OH
1. NaH
2. EtBr
4
5
71
1
11)
1. KMnO
4
, NaOH
2. H
+
2
3
a) 4
5
6
b) 7
72
1
2
73
12)
Br
+ NaCN
1
2

41
29
3
13)
C
O
O
+
NH
3
4
5
6
7
74
14)
OH
Et
O
+ CH
3
OH
1
2
3
4
6.22 Problem Solutions 5
6
7
1) 8
9

Cl
NH
3
+ +
NH
2
10
11
a) The M
+
peak at 82 suggests that the spectrum is of cyclohexene. 12
13
14
H
2
C
CH
2
CH
CH
CH
2
H
2
C
CH
2
CH
2
CH
CH
CH
2
H
2
C
CH
2
CH
2
CH
CH
CH
2
H
2
C
15
75
CH
3
CH
2
CH
CH
CH
H
2
C
CH
2
CH
CH
CH
H
2
C
m/z 67
CH
2
CH
2
CH
CH
CH
H
2
C
H
1
m/z 67 2
m/z 54
+
3
4
b) The odd M+peak at 99 suggests the nitrogen containing compound 5
cyclohexanamine. 6
7
NH
2
NH
2
NH
2
NH
2
m/z 43
8
NH
2
NH
2
m/z 56
9
2) 10
76

+ NBS
h!
Br
1
2
The pair of peaks at 121 and 123 as well as at 149 and 151 indicates that the 3
compound contains bromine. These, in combination with other prominent 4
fragments, suggest that the compound is 2-bromo-2-methlybutane. 5
6
7
Br Br
m/z 121
8
Br
m/z 71
9
10
11
3) 12
+ H
2
SO
4
H
2
O
OH
13
14
The peak at m/z 87 is deceiving because it is not the molecular ion in this case. 15
By utilizing other prominent peaks (m/z 73 and 45) and our knowledge of organic 16
chemistry, we can deduce that the mass spectrum is of 2-pentanol. 17
18
O
H
O
m/z 87
19
OH
OH OH
m/z 73
20
77
1
OH H
2
C
H
OH CH
2
H
CH
CH
2
CH
m/z 55
2

OH
OH
OH
m/z 45
3
CH
CH
2
CH
OH
H
H
2
C OH
m/z 31
4
5
4) 6
1. BH
3
THF
2. H
2
O
2
, OH
OH
7
8
The molecular ion is missing from the spectrum which is to be expected from a straight 9
chain alcohol. While it would be possible to determine the molecular weight from a soft 10
ionization technique, the existence of other prominent peaks indicates that the spectrum is 11
of 1-pentanol. 12
78
O
H
H
2
C
H
O
H
CH
2
H
CH
2
CH
2
m/z 70
1
H
CH
CH
2
CH
2
CH
2
O
H
CH
CH
2
CH
2
CH
2
CH
CH
2
CH
2
CH
2
HC
m/z 55
2
H
CH
CH
2
CH
2
CH
2
O
H
CH
CH
2
m/z 42
3
4
5
79
5) 1
MgBr
O
+
2. H
3
0
+
OH Et
2
3
The molecular ion is missing from the spectrum which is to be expected. The existence 4
of other prominent peaks indicates that the spectrum is of 2-methyl-2-butanol. 5
6
CH
3
OH Et
OH Et OH
Et
m/z 73
7
OH Et OH OH
m/z 59
8
H
2
C
CH
3
OH Et
H
CH
2
m/z 55
9
H
2
C OH
m/z31
10
11
12
6) 13
+ HI I
14
15
The molecular ion at 184 as well as the large gap in the spectrum in combination with the 16
peak at 127 indicates that the spectrum contains iodine and can be attributed to tert-butyl 17
iodide. 18
19
80
CH
3
I
I I
m/z169
1
I
+
2
m/z 127 3
I
m/z 57
4
5
6
7) 7

1. O
3
2. (CH
3
)
2
S
O
H
O
8
9
10
a) The molecular ion at 86 doesnt help distinguish compound a from b. The more 11
intense molecular ion suggests that a) is the ketone, 2-pentanone, rather than the 12
aldehyde. 13
14
15
C
CH
3
O
C
O
C
O
m/z 71
16
C
O H
2
C
H
C
O
H
m/z 58
17
81
C
O
C
O
C
O
m/z43
1
2
3
b) The presence of the M-1 peak helps to indicate that compound b) is the aldehyde, 4
pentanal. 5
H
O
C
O
C
O
m/z 85
6
CH
2
CH
2
C
H
O
H
2
C
H
H
3
C C
H
O
m/z58
7
C
H
O CH
H
C
H
O
H
m/z 44
8
C
H
O
C
H
O
C
H
O
m/z 29
9
10
11
8) 12

+
HCl
Cl
13
82
1
While the molecular ion is absent from the spectrum, the peaks at m/z 77 and 63 indicate 2
the presence of a chlorine atom in the sample indicating that the compound is 2-chloro- 3
butane. 4
5
Cl
Cl Cl
m/z 77
6
Cl
Cl Cl
m/z63
7
Cl
m/z57
8
Cl H
m/z 56
9
10
11
9) 12
OH
Cl
O
O
O
+
13
14
While the molecular ion is abscent, the other prominent peaks indicate both the functional 15
group and the hydrocarbon structure of the compound sec-butyl acetate. 16
17
83
O
O
O
O
O
O
m/z 101
1
O
O
O
O
O
O
m/z 87
2
3
CH
O
O
H
C
H
CH
CH
m/z 56
4
O
O
O O
m/z 43
5
O
O
CH
3
m/z 15
6
7
8
10) 9
10

OH
1. NaH
2. EtBr
O
11
12
84
The molecular ion of this ether is stabilized by the aromatic group. At m/z 122 it 1
indicates that the spectrum is of ethoxybenzene. 2
3
4
O
CH
2
CH
2
H
H
H
H
O
H
O
H
m/z 94
5
O
m/z 77
6
7
8
9
85
11) 1
C
O
OH
HO
O
1. KMnO
4
, NaOH
2. H
+
2
3
4
a) The molecular ion at 102 fails to distinguish between the two expected carboxylic 5
acids. The prevelant peaks at m/z 87 and 57 as indicates that the compound is pivalic 6
acid. 7
8
H
3
C C
O
OH C
O
OH
C
O
OH
m/z 87
9
C
O
OH
m/z 57
10
11
b) The missing molecular ion can be expected from a carboxylic acid. The prevelant 12
peaks at m/z 73 and 60 indicates that the compound is pentanoic acid. 13
14
HO
O
HO
O
m/z73
15
HO
O
H
C
H
HO
O
H
m/z 60
16
17
18
86
12) 1
Br
+ NaCN
N C
2
3
The lack of a molecular ion is to be expected from a nitrile. The base peak at m/z 41 4
indicates that the compound is butanenitrile. 5
6
N
C
CH
2
H
N
C
H
m/z41
7
N C CH
2
CH
3
m/z 29
8
9
10
13) 11
C
O
O
+
NH
3
C
O
NH
2
12
13
14
H
2
C
CH
2
CH
2
C
NH
2
O
H
3
C
H
2
C
H
2
C
CH
2
C
NH
2
O
m/z 86
15
16
17
18
19
87
14) 1
OH
Et
O
+ CH
3
OH
O
Et
O
2
3
4
O
Et
O H
2
C
H
O
Et
O
H
m/z 102
5
O
Et
O
Et
O
Et
O
m/z 99
6
7
O
Et
O
O
O
O
O
m/z 59
8
9
10
1
CHAPTER 7

Proper Laboratory Protocol and Sample Laboratory Experiments

7.1 Preliminary Information

7.1.1 Instrument and Instrumental Settings

This section will provide a brief introduction to the basic settings and
operation of GC (with any detector other then MS) and GC-MS
instrumentation. Examples of these settings will be provided in the
experiments that follow in this section. Some topics are redundant to
previous discussions but are included here for the purpose of clarity.
While section 7.1 concentrates on GC and GC-MS applications, most
of the experimental applications in sections 7.4 and 7.5 can be
extended to LC.

7.1.1.a Temperature Settings: There are four main
temperature controlled regions on a GC-MS. The first region is the
injector, which is set at least 20 degrees higher than the final oven
temperature. The column oven is set to run either an isothermal
mode or in as a temperature program, with the latter being the most
common. The oven temperature is usually initially set between 10
and 15 degrees below the boiling point of the solvent, held at this
point during the split-less mode of the injection, followed by one or
more temperature ramps, and typically held at a high final
temperature to remove late eluting analytes that may or may not be of
interest. Modern GC-MS systems automatically return to the starting
temperature after a given time. The oven is initially held at a
relatively low temperature (compared to the boiling point of the
solvent) to concentrate the analytes at the head of the column. If a
higher temperature is used, the solvent will rapidly volatize and
spread the analytes over a broad area and decrease peak resolution.
The detector is always set at a constant temperature 15 to 20
degrees above the maximum temperature of the oven. The injector
and detector are held at higher temperatures to prevent
recondensation of analytes onto this surfaces which would interfere
with quantification due to peak tailing and potential cross
contamination between samples. The final temperature region is the
MS vacuum chamber. It can be set above or at a lower temperature
2
than the GC components (~150 to 250 degrees C in the quadrupole
mass analyzer) due to the more volatile nature of analytes at low
pressure.

7.1.1.b Gas Flow: As noted in the GC chapter, all gases in
GC-MS, and in most GC applications, must be 5-nine quality (99.999
percent pure). These typically include a He or H
2
as a carrier gas,
Ar/CH
4
for makeup gas for electron capture detectors, and H
2
and
compressed air (lower grade) for flame ionization detectors in GC.
For GC-MS, only He is required for the carrier gas, with CH
4
being
commonly used in chemical ionization mode. Even at this purity, the
gas must be purified further by passing it through a resin trap that has
a high affinity for specific contaminants, including water, atmospheric
oxygen in some cases, and hydrocarbons. Although the presence of
these contaminants would result in a high detector background, the
main reason such high purity gases are needed is due to the use of
temperature programming. Two contrasting examples will
demonstrate the need for high purity gas purifiers. For the first case,
imagine running the GC-MS with a high-temperature isothermal oven
setting. At this temperature all contaminants in the gases will pass
freely through the system unretained in the separation column and a
high, but steady, background detector signal would result. For the
second case, imagine a temperature-programmed analysis where
initially the column oven is at a temperature lower than the boiling
point of any contaminants in the carrier gas. As carrier gas passes
through the analytical column, contaminants would be adsorbed to
the stationary phase at the beginning of the column. As the
temperature program progresses, these contaminants would volatilize
and appear as peaks in the chromatogram. The height of the
contaminant peaks (concentration) would be inconsistent since it
would depend on the time and the amount of carrier gas passing
through the column between runs. The contaminants would result in
additional problems if they co-eluted with an analyte of interest.

The gas pressure in the supply tank is usually between 2000 and
2500 psi (up to 17000 kPa). Instruments require that this pressure be
reduced with step-down or secondary regulators that drop the
pressure to 100 psi (700 kPa) or less, depending on the instrument
and gas. Integrated regulators or mass-flow controllers further
reduce the pressure to 5 to 20 psi at the head of the capillary column,
3
resulting in a flow of 1 to 5 mL/minute depending on the internal
diameter of the capillary column.

7.1.1.c. Vacuum Chamber. As helium enters the MS unit, it
must be evacuated to minimize secondary collisions with the ionized
analytes. Two vacuum pumps are used to accomplish this. First, a
rotary vacuum pump evacuates the gases to approximately 10
-2
torr.
Then a molecular turbo pump reduces the pressure to 10
-4
to 10
-6

torr.

7.1.2 Maintenance

7.1.2.a. Gas Filters: As noted in the previous section, ultra
high purity gases are purified even further with resin filters (traps).
These filters must be replaced periodically, usually after 5 to 10 tanks
of gas depending on the size of the filter.

7.1.2.b. Septa. The interface where samples are introduced
into the instrument is a silicone gum septum with a Teflon backing on
the injector side of the septum. This allows the sharp needle of the
syringe to be easily inserted into the injector chamber and the sample
to be introduced. As more and more injections are made, the septa
develops a slight perforation in it that will eventually leak carrier gas
and allow the loss of sample during an injection. Therefore the
septum must be replaced periodically, typically daily or just prior to a
new run of standards and samples. Septa are relatively inexpensive
so this is not a major cost issue.

7.1.2.c. Injection Syringes and Needles. Syringes can wear
with time depending on the type of samples injected. Dirty samples
will quickly clog the syringe by leaving residue in the barrel that
interferes with the movement of the plunger. This can usually be
avoided by numerous rinses between samples. However, it is
sometimes necessary to disassemble the syringe and rinse with acid,
polar organic solvent, and a nonpolar organic solvent. Injection
needles can also plug with a piece of the septum. Most syringes
come with a thin wire to remove this plug but this technique is rarely
successful and the syringe is usually replaced. Syringes for manual
injection are as inexpensive as $15, but autosampler syringes can
easily cost $100.
4

7.1.2.d. Column Fitting. Columns are attached to the injector
and detector ports with threaded nuts and ferrules, a soft hollow
conical-shaped device that fits snugly around the column and fits into
a receptor secured with a nut. As the nut is tightened, the ferrule is
compressed around the column, preventing gas leaks. As the
temperature is repeatedly raised and lowered, leaks can result from
the cycling expansion and contraction, so the ferrule nuts need to be
tightened periodically (weekly to monthly). Ferrules can be made of
Teflon, ceramic, graphite, and composites of ceramic and graphite.

7.1.2.e. Glass Wool Plugs in the Injector Liner. Most injector
liners have glass wool inserted into them to aid in the uniform mixing
of the volatized solvent and analytes with the carrier gas. Over time
(weeks to months) these liners accumulate pieces of the septum
(referred to as septum worms) and nonvolatile components of the
injected sample. Therefore the liners are routinely replaced when
discoloration or evidence of cross contamination occurs. The
frequency of replacement is directly related to the presence of
nonvolatile components in the samples, and can range from weekly
to yearly replacement cycles.

7.1.2.f. MS Tuning. The mass spectrometer, specifically the
mass analyzer, must be calibrated with respect to mass, typically on
a weekly basis. Some applications require daily tuning. Modern MS
systems have an automated tuning sequence. For electron ionization
systems, perfluorotribuylamine (PFTBA) is used. A small mass of
volatilized PFTBA is introduced into the ionization chamber and the
system automatically adjusts to correspond to its mass. Parameters
such as repeller and accelerator voltages and gain on the EM are
adjusted to achieve a given detector response. After this, the system
can be tuned for any mass unit.

7.1.2.g. Ion Lens. The repulsion and acceleration lenses may
accumulate nonvolatile residues when dirty samples are analyzed.
Depending on the quality of a sample and the frequency of use,
lenses will need to be taken out and rinsed with solvents, dried, and
reassembled. The typical sign of a dirty lens is the need to apply
higher than normal voltages to these lenses during the MS tuning
procedure.
5

7.1.2.f. Gain on the Electron Multiplier. For a given mass of
tuning compound (PFTBA), a specific counts per second of ions
hitting the EM is expected. This is adjusted by increasing or
decreasing the gain (potential) across the EM. As the EM ages, it
may require an excess gain to be applied and the EM will need to be
replaced.

7.1.2.h. Rotary Pump Oil. The rotary pump is lubricated with
special grade oil. The gauge level should be checked monthly, and
oil added if necessary. The oil should be replaced at least annually,
sooner in high use situations and when dirty samples are analyzed.
Many or most of the unionized analytes, contaminates, and solvents
eventually reside in the rotary pump oil. Rotary pumps usually
require semi-annual replacement due to oil leaks.

7.1.2.i. Analyte calibration. While not a normal part of
maintenance, instruments are normally calibrated at least daily with
analytes of interest.

7.1.3 Trouble Shooting

A variety of problems will be experienced when using a GC-MS for
prolonged time. A few of the most obvious are discussed below.
Instrument manuals normally come with a trouble-shooting guide.

7.1.3.a. Leak Detection. Atmospheric leaks will occur from
time to time. The most likely sources of these leaks are the column
fittings and the door to the MS vacuum chamber. Leaks may be
present if mass numbers corresponding to N
2
, O
2
, H
2
O, CO
2
, and Ar
appear in the spectra. System leaks are easily checked by setting
the instrument to constant monitoring mode and then spraying
canned Freon at each fitting and watching for a detector response. A
readily available leak detection agent is DustOff that contains
difluoroethane (CAS #75-37-6).

7.1.3.b. Contamination of the GC-MS system. Unfortunately,
all systems become contaminated with time. The key to minimizing
time locating the source of contamination is to systematically isolate
6
each system and therefore the source of contamination. A good
practice is the following.

-First, look for the obvious. What was the last thing changed prior to
the presence of contamination? Was a septum, liner, column, gas
filter, or gas tank recently changed?

-Check each potential source for problems, especially the filters and
liners. There have also been cases of contaminated 5-nine gas being
delivered from suppliers and contaminated injector liners direct from
the factory.

- Check the solvent for contamination by eliminating sample
introduction and only running solvent"

-An easy way to isolate the injector and check for contamination,
without taking it apart, is to cool the injector and conduct a
temperature run without sample injection. If the contamination is not
present when the injector is cooled, a contaminated injector is likely.

-Identify your contaminate with the spectra library. If your
contamination is the analyte, then the contamination is likely to be on
the front-end of the GC-MS system (syringe or injector liner).
Hydrocarbon contamination from oils is possible and will be indicative
when ions are present at 43, 57, 71, and 87 mass units. Siloxanes
are indicative at mass units of 73 and 207 mass units. Phenyl
degradation from column degradation will be present at 281 mass
units. Phthalates are ubiquitous in the environment and will give an
ion peak at 149 mass units.

7.1.3.c. Plugged Needle. As noted in section 7.1.2, needles
frequently become plugged with pieces of septum. This is indicated
when a sample is thought to be injected but no ions or peaks appear,
including the solvent.

7.1.3.d. Broken Columns. Another explanation for a lack of
detector response is a broken column. This is easily observed by
cooling the oven and inspecting the column. Never allow the column
to rub against a surface as it will wear off the protective coating of the
column and promote a break in the column.
7

7.1.3.e. Low Sensitivity/High Gain on the EM. This is indicative
of a worn out detector.


7.2 Preliminary Experiments: Getting to Know Your Instrument

7.2.1 Autotuning the MS

MS instruments must be tuned frequently to ensure correct
identification of ion mass to charge ratios; modern instruments have
an automated sequence or menu to do this. Most instruments use
perfluorotribuylamine (PFTBA) that is stored in a vial in the MS.
During the tuning procedure, a valve is opened to allow a small,
consistent mass of PFTBA to enter the ionization chamber. Typical
concentrations of vapor range from 1 to 10 ppm PFTBA. As PFTBA
passes through the MS, the instrument optimizes several settings to
obtain the maximum detector response (counts per second) for
selected ion fragments of the tuning compound. Results from one of
the most common brands on the market (Agilent 5975C) are shown in
Figure 7.1.
8

Figure 7-1. A Typical Electron Ionization Tune File from an Agilent
5875 EI-Quadrupole Mass Spectrometer.

Interpretation of the tune file. printout: The center plot in Figure 7.1 is
a chromatogram of PFTBA showing the abundance of each ion as a
9
function of temperature. Below the plot are the observed counts per
second for three m/z ratios (69 amu, 219 amu, and 502 amu) and
their corresponding C-13 isotope-containing ions (the small peak to
the immediate right of each tune peak) after the instrument has been
successfully tuned. The counts per second for each ion are given in
the Abund (abundance) column below the chromatogram. The 69,
219, and 502 ions are used to calibrate the m/z values over the entire
range of the spectrum.

Now look at the top left-hand side of the figure. This contains
expanded scale enlargements of the three m/z peaks. Recall that the
quadrupole mass analyzer only yields unit mass resolution. Each of
the peaks shown in Figure 7-1 is the result of 10 data point
measurements evenly spread across the single amu measurement.

There are several objectives of the tune function. One objective is to
calibrate the mass analyzer with respect to mass, so the instrument
assigns the large peaks at 69 and 219 to these masses, while the
isolated ion at 502 is calibrated to the 502 m/z value. A second
objective is to obtain unit resolution as shown in the enlarged plots
where the presence of C-12 and C-13 in each of the ions is resolved.
A third objective of the tune is to calibrate the instrument where peak
height can be used in the counts per second measurements instead
of peak area since peak height calculations are faster to calculate
and thus allow faster analysis. This last objective is accomplished by
normalizing the width at half peak maximum for each of the three ion
peaks to similar or near identical values. Each of these objectives is
accomplished by sequentially adjusting the parameters listed on the
top right-hand side of the figure. These include the voltages of the
Repeller, Ion Focus (IonFcus), entrance lens (EntLens), entrance
offset (EntOffs), AmuGain and Amu offset (AmuOffs). Recall that the
repeller is located on the upstream side of the ionization source and
is positively charged to push the ionized molecules (cations) toward
the mass analyzer. Most of the inertia/velocity imposed on the ion is
from charge placed on the repeller. The other lens focuses the ions
into the center of the trajectory towards the mass analyzer. The
mass width of the peak is primarily adjusted by the AmuGain and
AmuOffs parameters. All of the other parameters shown in the top
right corner of the figure are normally held constant.

10
Each of the parameters are adjusted sequentially until the maximum
counts per second, resolution, and similar half peak width are
achieved; as one parameter is changed, the instrument readjusts the
previously adjusted parameters for optimum performance. Finally the
EMVolts (potential applied across the electron multiplier) is adjusted
so that the 69 m/z ion has a counts per second of approximately 500
000.

Leaks can be detected in the tune process by reviewing the Air/Water
Check line of data located immediately below the center total ion
chromatogram (TIC). The presence of H
2
O, N
2
, O
2
, CO
2
and N
2
/H
2
O
are shown here and should be present at no more than 10 percent of
the total 69 m/z ion counts. If values higher than this are
encountered, a leak is present, and is usually located at the vacuum
door or column inlet fitting.

7.2.2 Optimizing Analyte Separations with a Temperature
Program

The goal of chromatography is to separate a complex mixture
of compounds. Some separations are relatively simple while others
require experimentation to optimize the instrumental settings.
Analyte separations are controlled by the temperature settings of the
column and oven. Usually the initial temperature of the oven is set at
approximately 10 to 15 degrees below the boiling point of the solvent.
After injection, the oven temperature may or may not be held at this
value for a few minutes. Next, the oven and column temperature is
increased as slow as needed to allow separation of the compounds
but as fast as possible to minimize the instrument run time. Finally,
after all of the analytes have reached the detector, the instrument is
usually held at a high temperature to allow any high boiling
compounds to exit the column. The key to an adequate separation is
to determine each of these temperatures, noting the need to achieve
adequate separation in a minimum amount of time, especially in an
industrial setting where cost (and time) efficiency is mandatory.

In this experiment, we will show the optimization of the
temperature program for a set of hydrocarbons normally found in
gasoline, the subject of the next lab.

11
Experimental Procedures:

Chemicals and Supplies:
A Pasteur pipet for each analyte
One 10-mL volumetric flask
Neat (pure) samples of benzene, decane, ethyl benzene, n-
heptane, isooctane, toluene, m-xylene, and o-xylene.

Instrumental Settings:
GC-FID Settings (Flame Ionization Detector)
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.53 mm; 1.5 m phase coating
Injection Volume: 1.00 L
Splitless Injection for: 1.00 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.2 mL/min.
Linear Velocity: 14 cm/s
Injector Temperature: 230
o
C
Detector Temperature: 250
o
C
Oven Program: varied as described below.

Sample Preparation:

The dilution solvent will be pentane because it has a very low
boiling point and most other dilution solvents would co-elute with one
or more analytes. Prepare a qualitative standard, as described
below, for injection into the GC.

-Add two drops of each analyte to approximately 10 mL of pentane,
cap in an air-tight vial, and mix the solution.
-Inject this solution into the GC using a variety of temperature
programs. Start with a relatively low temperature isothermal program
(50 C) for an extended time (20-30 minutes). Next, use a relatively
high temperature isothermal program (150 C for 15 minutes). You
will not obtain complete separation for either of these programs.
Finally, use a temperature program starting from a temperature just
below the boiling point of your analyte with the lowest boiling point
and program an increase of 5 C per minute to a final temperature
12
approximately 10 C above the boiling point of your analyte with the
highest boiling point. Refer to the next experiment in section 7.2.3 for
optimum temperature programming instructions. When all peaks
have been separated, the elution order will be: benzene, n-heptane,
isooctane, toluene, ethyl benzene, m-xylene, o-xylene, and decane.


7.2.3 Obtaining a Linear Calibration Line

After the temperature program has been optimized, the next
task is to calibrate the instrument. As discussed in Chapter 1,
instruments easily generate numbers but the analyst must always
question the validity of numbers until they are sufficiently scrutinized.
Chromatographic analysis has a special feature over most other
analyses since the very nature of chromatography allows the analysis
of several compounds at one time. If quantitative work is being
performed, the instrument must be calibrated with respect to each
analyte. This experiment will illustrate proper calibration of a GC-MS.
We will use several components of gasoline as our analytes and
service station samples of gasoline as our sample.

In this experiment, the analyst will (1) obtain reference standards of
several components of gasoline, (2) make dilutions of the reference
standards (in pentane) ranging from 1.00 ppm (parts per million) to
100 ppm, (3) inject these standards into the instrument, (4) analyze
the samples (at an appropriate dilution), (5) use the software to
calibrate the instrument, and (6) analyze the results (perform a linear
least squares on the calibration line and calculate the concentration
of each component in the gasoline sample).

Experimental Procedures

Chemicals and Supplies:
25-L, 50-L, and 100-L glass microsyringes
1.00-mL and 2.00-mL Class A pipets
Eight 10-mL volumetric flasks
Two 25-mL volumetric flasks
One 250-mL volumetric flask
Neat (pure) samples of benzene, decane, ethyl benzene, n-
heptane, isooctane, toluene, m-xylene, and o-xylene.
13

Instrumental Settings:
GC-FID Settings (Flame Ionization Detector)
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.53 mm; 1.5 m phase coating
Injection Volume: 1.00 L
Splitless Injection for: 1.00 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.2 mL/min.
Linear Velocity: 14 cm/s
Injector Temperature: 230
o
C
Detector Temperature: 250
o
C
Oven Program: 40
o
C for five minutes, 4
o
C to 200
o
C, hold for 10
minutes

GC-MS Settings:
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.25 mm; 0.25 m phase coating
Injection Volume: 1.00 L
Splitless Injection for: 0.50 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.2 mL/min.
Linear Velocity: 40 cm/s
Injector Temperature: 230
o
C
Detector Temperature: 250
o
C
Oven Program: 40
o
C for five minutes, 4
o
C to 200
o
C, hold for 10
minutes

Calibration and Sample Preparation:

Calibration standards containing the major components of
unleaded gasoline are required. An external calibration procedure
will be used, with an internal standard to correct for injector errors
and detector drift. The dilution solvent will be pentane because it has
a very low boiling point and most other dilution solvents would co-
14
elute with one or more analytes. Prepare a stock calibration
standard, as described below, and use this standard to perform serial
dilutions (using pentane containing decane as an internal standard) to
obtain a range of calibration standards.

NOTES: (1) To minimize the volume (and expense) of GC grade
solvents used, dilutions will be made with micro-syringes. This
method is less accurate then when using Class A pipets, but will be
sufficient for our demonstrations here. (2) All compounds used in
this lab are very volatile and flammable. Work in a fume hood away
form hot plates, flames, and combustion sources. To minimize
volatilization of analytes during solution preparation, place
approximately 10 to 15 mL of pentane in the volumetric flask. Since
pentane has the lowest boiling point, it will be the first to volatilize,
leaving the other analytes in solution.

Procedures:

(1) To add each analyte to the flask, fill a microsyringe to the desired
volume (in Table 3.1 below), place the syringe needle on the inside
neck of the flask (not in the solution), empty the syringe, withdraw it,
and immediately rinse the walls of the flask with 1-3 mL of pentane.
Rinse the syringe thoroughly with clean pentane and repeat the
process. After all of the analytes have been added to the flask, fill it
to the mark with pentane. This solution is the stock solution of each
analyte.

Table 7.1 Preparation Guide for the Stock Calibration Solution.
Analyte
(>99%
neat)
Boiling
Point
o
C
Density
of liquid
L of pure
analyte to be
added to a 25
mL volumetric
flask
Resulting ppm
concentration in
flask
Benzene 80 0.874 29.0 1010
Ethyl
Benzene
136 0.867 29.0 1010
n-
Heptane
98 0.684 37.0 1010
Isooctane 99 0.692 36.0 996
15
Toluene 111 0.865 29.0 1000
m-Xylene 138 0.868 29.0 1010
o-Xylene 144 0.870 29.0 1010
Decane
(Internal
Standard)
174 0.73 14.0 101

(2) All solutions injected into the GC must contain internal standard
(decane). Make 250 mL of pentane-internal standard solution for
dilutions by adding 35 mL of pure decane with a microsyringe to a
250-mL volumetric flask and then filling the flask to the mark with
pentane. Cap, mix, and use to make the following solutions.

(3) Make dilutions of the ~1000 mg/L solution made in step 1,
according to the table below. Fill each flask with the internal
standard-pentane solution made in step 2.

Table 7.2 Preparation of GC-MS Calibration Standards.

1 2 3 4
Approx. Conc.
of each Analyte
(ppm)
Solution to be
Used in Dilution
mL of Solution
from Column 2
to be added to
Volumetric
Flask
Volumetric
Flask Size to
Use
100. Stock 1000
ppm
1000 (1.00 mL) 10.00
80.0 Stock 1000
ppm
2000 (2.00 mL) 25.00
40.0 Stock 1000
ppm
1000 (1.00 mL) 25.00
20.0 100. ppm 2000 (2.00 mL) 10.00
10.0 100.0 ppm 1000 (1.00 mL) 10.00
4.00 40.0 ppm 1000 (1.00 mL) 10.00
2.00 20.0 ppm 1000 (1.00 mL) 10.00
1.00 10.0 ppm 1000 (1.00 mL) 10.00
0.400 4.00 ppm 1000 (1.00 mL) 10.00
0.200 2.00 ppm 1000 (1.00 mL) 10.00

16
(4) The compounds in pure gasoline are at too high of a
concentration to be analyzed directly on the GC-MS. Most of the
major constituents in gasoline are present between 5 and 20 percent
on a mass basis. To dilute the gasoline to an acceptable level, add
40.0 mL to 100 mL of pentane-internal standard solution. Several
samples of unleaded gasoline should be analyzed. Suggestions for
selecting samples include brand, octane rating, and the presence of
methanol and MTBE. Note: if methanol or MTBE are present in your
sample, the calibration standards must also include these
compounds.

(5) Analyze the standards and diluted samples by GC-MS using the
instrumental conditions given earlier. Use the MS to identify each
peak in the spectra and then calibrate your instrument. Calculate the
% composition of each analyte. Finally, analyze the spectrum of
each compound and review the fragmentation rules from Chapter 2.

Results:

Each compound should produce a linear calibration line over
the concentration range of your external standards. Most modern
instruments will do this relatively automatically. After you calculate
the concentration of each analyte in your gasoline sample, convert
the ppm concentrations to percent by mass. Compare this to
published composition available on the Internet. NOTE: the power of
chromatography is the separation of complex mixtures which we have
accomplished in this experiment.

7.2.4 Electron (hard) versus Chemical (soft) Ionization.

As noted in Chapter 1, the most common form of ionization in
MS is electron ionization (EI) that is considered a hard source since is
creates numerous fragments and allows for a unique fragmentation
pattern. Several spectral libraries and computer search/match
routines are available to aid in analyte identification. In contrast,
chemical ionization (CI) is a milder form of ionization. Chemical
ionization is rarely used for fragmentation pattern recognition, but is
used to observe or obtain the molecular mass of the molecular ion.
This experiment shows the electron and chemical ionization of three
compounds.
17

EXPERIMENTAL PROCEDURES:

Chemicals and Supplies:
A 25 ppm solution of 2,2,66,-tetrachlorobiphenyl in isooctane
A 50 ppm solution of cyclohexanol in methanol
A 50 ppm solution of decanoic acid methyl ester in methanol

GC-MS Settings:
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.25 mm; 0.25 m phase coating
Injection Volume: 1.00 L
Splitless Injection for: 0.50 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.2 mL/min.
Linear Velocity: 40 cm/s
Injector Temperature: 250
o
C
MS Transfer Line
Detector Temperature: 230
o
C
Quadrupole Temperature: 150
o
C
Oven Program: 55
o
C and hold for zero minutes, 5
o
C to 250
o
C,
hold for ten minutes
Total Run Time: 49 min.

Procedures:
Inject the standard solutions and analyze them using the
instrument conditions given above.


RESULTS:

Spectra of the three compounds for EI and CI are shown below.

18


Figure 7.2. Fragmentation of Cyclohexanol by EI.



19

Figure 7.3. Fragmentation of Cyclohexanol by CI.

First, note the presence of the molecular ion using both ionization
techniques. As expected, extensive fragmentation of cyclohexanol
occurs for the EI analysis and follows the rules for fragmentation of
alcohols given in Chapter 6, while minor fragmentation occurs in the
CI analysis.

20


Figure 7.4. Fragmentation of Decanoic Acid Methyl Ester by EI.


21

Figure 7.5. Fragmentation of Decanoic Acid Methyl Ester by CI.

Similar results are found for decanoic acid methyl ester; extensive
fragmentation occurs during EI, but not during CI. Furthermore, for
CI the molecular ion is more pronounced and the M+C
2
H
5
ion is
present in significant concentrations.


22

Figure 7.6. Fragmentation of 2,2,6,6-tetrachlorobiphenyl by EI.

Figure 7.7. Fragmentation of 2,2,6,6-tetrachlorobiphenyl by CI.

23
The analyte, 2,2,6,6-TCB is so stable, even under the conditions in
the EI chamber that the molecular ion is still a dominant peak. Again,
some fragmentation occurs during EI, while additions are observed
for the CI technique.

When to use EI and CI: Most MS analysis uses EI because it
yields easily identified (via a fragmentation library) and unique
fragmentation patterns. However, CI is used in two main cases: (1)
when the point of the analysis is to obtain information about the
molecular weight of the molecular ion and (2) when a better (lower)
detection limit can be obtained using CI. Chemical ionization can be
used in two modes, positive and negative.

As noted in section 1.5.1.2b: Chemical ionization is most
commonly used to create positive ions, but some analytes, such as
those containing acidic groups or electronegative elements (i.e.
chlorinated hydrocarbons) will also produce negative ions that can be
detected by reversing the polarity on the accelerator and detector.
Some of these analytes produce superior detection limits with CI as
opposed to EI, while others only give increased sensitivity (slope of
the response to concentration line). Negative ions are produced by
the capture of thermal electrons (relatively slower electrons with less
energy than those common in the electron beam) by the analyte
molecule. Thermal electrons are present from the low energy end of
the distribution of electrons produced by the lower-energy CI source
(~20 eV as opposed to ~70 eV in EI). These low energy electrons
arise mostly from the chemical ionization process but also from
analyte/electron collisions.

7.3 Concept Illustrative Experiments

7.3.1 Advantages of GC-MS over GC

Capillary columns provide superior resolution over packed
columns, and while separations of complex mixtures are usually
complete, some samples can be problematic. This is also why a
single GC analysis for an analyte, even with a reference standard, is
not conclusive, but suggestive. As discussed in Chapter 1, GC
analysis (in the absence of MS detection) can be considered
conclusive when a sample is analyzed twice, once on one stationary
24
phase and once on a different stationary phase, and when the same
results from these two analyses confirm the present of an analyte
based on retention time.

In contrast, gas chromatography-mass spectrometer analysis
can give conclusive identification for many structures, with or without
a reference standard. But MS analysis requires that a pure
compound be introduced into the MS or that a GC be used to
separate a complex mixture of analytes. This is the purpose of this
experiment, to show the identification power of MS. Polychlorinated
biphenyls (PCBs) will be used for illustration purposes here, and
there are many other classes of compounds that can be used for this
purpose (i.e. alkanes, aromatics, etc.). There are 209 different PCBs,
ranging from monochlorobiphenyls to a single decachlorobiphenyl.
PCBs are usually separated/analyzed on a non-polar column such as
the polydimethyl siloxane phase (commonly referred to as HP-1, SP-
1, or DB-1) or the poly(phenylmethyldimethyl) siloxane phase
(commonly referred to as HP-5, SP-5, or DB-5). These columns
mainly separate non-polar analytes based on boiling points and given
the possibility of similar structures in PCBs (and other classes of
compounds), some compounds will have similar boiling points and
therefore similar retention times in the chromatogram (lack of
separation). However, given the range of boiling points of the 209
PCBs a very slow oven temperature ramp (~1.0
o
C per minute) is
necessary that results in a long analysis time (approximately 2
hours). In this experiment the lack of separation will be illustrated for
2,2-dichlorobiphenyl and 2,6-dichlorobiphenyl.

Experimental Procedures

Chemicals and Supplies:
A 25 ppm solution of 2,2-DCB in isooctane
A 25 ppm solution of 2,6-DCB in isooctane
An isooctane solution containing 2,2-DCB and 2,6-DCB (25
ppm each)

GC-MS Settings:
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
25
30 m x 0.25 mm; 0.25 m phase coating
Injection Volume: 1.00 L
Splitless Injection for: 0.50 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.2 mL/min.
Linear Velocity: 40 cm/s
Injector Temperature: 250
o
C
MS Transfer Line
Detector Temperature: 230
o
C
Quadrupole Temperature: 150
o
C
Oven Program: 70
o
C for two minutes, 5
o
C to 280
o
C, hold for 2
minutes
Total Run Time: 46 min.

Procedures:
Inject the standard solutions and analyze them using the
instrument conditions given above.

RESULTS:



Figure 7-8. Total Ion Chromatogram of a 25ppm solution of
2,2-dichlorobiphenyl and 2,6-dichlorobiphenyl.

Figure 7.8 shows the analysis results for a solution
containing both 2,2- and 2,6- dichlorobiphenyl. Note the lack of
separation; individual injections shows that 2,6-DCB elutes at
21.494 minutes while 2,2-DCB elutes at 21.506 minutes. An
injection of a combined solution does not resolve the two
analytes. A slower temperature ramp may allow the separation
of these compounds, or separation can be improved with a
26
longer column or with a thicker film coating. But again there are
instances where gas chromatography cannot adequately
separate some compounds. If only one of the compounds is
present in a GC peak we may still be able identify it using MS.
For example, review the two spectra below.

Figure 7-9. Mass Spectrum of 2,2-dichlorobiphenyl.
27

Figure 7-10. Mass Spectrum of 2,6-dichlorobiphenyl.

While these spectra look similar at first glance, distinct
differences (relative ion abundance heights) can be noted that
are used by the GC matching algorithm to identify the
compound. Recall, the library search routine mainly uses two
criteria to match an analysis with a known from the library
spectra: presence of a m/z peak and relative heights of the m/z
peaks.

Similar m/z peaks are present in each spectra but the
relative proportions are distinctly different, especially in the186-
190 m/z region. Thus, if only one of the compounds is present
in a GC peak, it can be easily identified.

As an aside, it should be noted that if milligram quantities
of the analytes could be obtained, NMR could be used to
identity their presence and abundance, even in a mixed
solution.


28
7.3.2 Advantages of GC over MS; cis- versus trans-

The experiment in Section 3.3.1 illustrated the power of MS in
identifying analytes when they could not be separated by GC. This
experiment will do the reverse, use GC to identify analytes that give
the same spectra with MS. This is important with cis- and trans-
isomers. Cis- and trans- isomers can have significantly different
physical parameters due to the rotation of functional groups around a
double bond. For example, cis-stilbene has a boiling point of 82-84
o
C, while rotation of one benzene ring around the double bond to
form trans stilbene yields a boiling point of 305-307
o
C. These can
easily be separated by chromatography but ~all cis- and trans-
isomers yield the same fragmentation pattern in MS.



cis-stilbene trans-stilbene

This experiment will use GC to separate and identify cis- and
trans- heptene. Look up the boiling points to estimate the relative
retention order.

Experimental Procedures

Chemicals and Supplies:
A 50 ppm solution of cis- and trans- in heptene

GC-MS Settings:
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.25 mm; 0.25 m phase coating
Injection Volume: 1.00 L
29
Splitless Injection for: 0.50 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.2 mL/min.
Linear Velocity: 40 cm/s
Injector Temperature: 250
o
C
MS Transfer Line
Detector Temperature: 230
o
C
Quadrupole Temperature: 250
o
C
Oven Program: 80
o
C for two minutes, 5
o
C/min to 210
o
C, hold
for 10 minutes

Procedures:

Analyze the standard solutions on a GC-MS using the
instrumental conditions given above.

RESULTS:



Figure 7-11. Total Ion Counts for the Analysis of cis- and trans-
Heptene.

30
Note the dependence of the results, retention times, on
boiling points. DB-1 and DB-5 columns separate exclusively
based on boiling points.

7.3.3 Advantages of GC over MS; Chiral separations

One of the most difficult classes of compounds to separate is
chiral compounds. In some cases these can be separated by normal
capillary column GC, usually if the compound has more than one
chiral center. Recall, the criteria that allows separation is if the
chemical structure results in a different set of physical characteristics
such as boiling point. In our design of several laboratory experiments
we accidentally came across several chiral compounds that
separated on a DB-5 capillary column (in the fragrance experiments
in section 7.5). We know this since two ion peaks that were
extremely close to each other in the chromatograph give identical and
essentially exclusive identification (99% probability of a library match
with limited or no additional matches). Chiral columns are available
but only for a limited selection of a compound structures. In general,
MS fragmentation will not distinguish between chiral compounds.


7.4 Analytical Experiments with an External Reference Standard
Calibration

7.4.1 Caffeine Concentrations in Human Urine by Nathan Conroy

As the use of drugs has become more commonplace, so has the
concern and apprehension toward the misuse and abuse of drugs.
For example, professional athletes and Olympians are subject to
random testing for performance enhancing drugs. Drug testing has
even become routine as part of many job applications. Often drug
analyses have to be designed to test for metabolites of the drug of
interest, rather than the drug itself. Cocaine drug analyses involve
not only the quantification of cocaine, by also benzoylecgonine (a
metabolite of cocaine formed in the liver) and ecgonine methyl ester
(both a metabolite and precursor of cocaine) [1]. Typically, urinary
drug analysis procedures require the use of an internal standard that
accounts for most losses during a liquid-liquid extraction to remove
analyte from the protein and compound aqueous bio-layer [1-2]. The
31
isolated solution of analyte and internal standard is then analyzed on
an instrument such as a gas chromatography-mass spectrometer
(GC-MS) or high performance liquid chromatography-mass
spectrometer (HPLC-MS). Laboratory analysis of many drugs
requires a costly license for possession of the drug, thus most
academic laboratories do not teach these extraction procedures. But
similar extraction using street legal compounds can be used as
surrogates. This caffeine lab procedure is designed to introduce
students to the techniques and procedures used in drug analyses
where caffeine is used as a surrogate for many drugs. Students will
determine the concentration of caffeine in their urine after having
consumed caffeinated beverages, and see how this concentration
changes as a function of time; as well as across different caffeine
consumption habits.

Multi-step sample preparation techniques do not quantitatively
transfer an analyte from starting material to the final extraction
solution; small percentage losses can occur at each step in a
procedure. Therefore, when trying to determine an unknown
concentration of an analyte, the analyst must account for the sum of
these experimental losses using an internal standard (ISTD). A
chosen internal standard should behave similarly to the analyte under
reaction conditions; therefore relative losses throughout sample
preparation will be equal. In other words if 20ppm ISTD (final
concentration in the extraction solution) is added in a sample, but
when the sample extract is analyzed and the instrument signal
corresponds to only 15ppm of ISTD, we known only 75% of actual
ISTD concentration was detected by the instrument. The use of an
ISTD correction procedure will account for these losses in the
analyte. For the case just stated, if the instrument detects a signal
corresponding to 12ppm analyte, the instrument will back calculate
the actual concentration of analyte in the solution to be 16ppm (a
correction of +25 percent).

Experimental Procedures

The best way to ensure an equal concentration of internal standard
across your samples is to deliver an equal amount of internal
standard to all solutions prior to diluting. Also, make sure that the
final concentration of internal standard in your sample extract is the
32
same concentration as internal standard used in making your
calibration standards. A given volume of internal standard solution is
most accurately delivered by a Hamilton-type micro-syringe where
the full volume (or near full volume) of the syringe is used. Choosing
an internal standard for a GC-MS analysis of caffeine is problematic
because most compounds that share structural similarities with
caffeine thermally degrade before volatilizing. 4-Acetylpyriding
emerges on the gas chromatogram as two separate peaks, 4-
Acetylpyrdine and its hydrated derivative, making a standardized
integration of the peak problematic. Decyl-alcohol works as an
internal standard, but is not ideal because it shares no structural
similarity to caffeine. Cyclizine would likely make an appropriate
internal standard for a caffeine GC-MS analysis but is more costly.

Chemicals and Supplies:

High-Resolution GC grade methanol
Caffeine
Ammonium chloride
Ammonium hydroxide
GC grade dichloromethane
Sodium chloride
High purity (99.999%) helium gas

Instrument Settings:

Front Inlet:
Mode: splitless
Inlet temperature: 250 C
Pressure: 10 psi
Purge Flow: 50 mL/min
Purge Time: 0.50 min.
Total Flow: 54.0 mL/min.

Injection Volume: 1.00 mL

Column Specifications:
HP-5MS 5% Phenyl Methyl Siloxane
Length: 30. m
Diameter: 250. mm
33
Film Thickness: 0.25 mm
Flow Rate: 1.2 mL/min.
Linear Velocity: 40 cm/sec.

Oven Settings:
Initial Temp.: 50 C
Initial Time: 2.00 min.
Ramp: 15.0 degrees per minute to 260 C, hold for 2.00 min.
Transfer Tube Temp.: 280 C

MS Parameters:
Solvent Delay: 4.00 min.
Ionization Source: EI
Temperatures: MS Source 230 C; MS Quad 150 C

TABLE 7.1. Approximate Peak Retention Times

Compound Retention Time
(min)
Caffeine 13.86
Decyl-alcohol 9.08
4-Acetylpyrdine 7.28


Breaking an Emulsion:

The liquid-liquid extraction used in this procedure has a tendency to
form emulsions, mixtures of two immiscible liquids. They generally
appear as either a cloudy combination of liquids, or as bubbly pockets
at the interface between the two liquids. Emulsions interfere with the
recovery of the extraction solvent and the analyte, and therefore are
problematic in analytical analyses. Emulsions can be resolved or
broken several different ways. A saturation of the aqueous phase
with sodium chloride is a first defense against emulsion formation, but
has not been shown to be sufficient with this procedure. Sonicating
solutions is another common solution, but did not prove completely
successful here. Centrifuging is another common solution. A far less
costly strategy for breaking emulsions is glass wool. Glass wool can
be used to break emulsions by packing the glass wool into the bottom
1cm-2cm of a transfer pipet, then filtering the emulsified layer through
34
the packed pipet. While glass wool does successfully break the
emulsion, the likelihood of experimental loss makes it non-ideal. If all
else fails time will break the emulsion; allow solutions to sit over night.

Evaporation/Concentration of the Extraction Solvent
One of the advantages of using organic solvents is that the final
extraction volume can be concentrated. This results in the
concentration of the analytes and improves the detection limit. The
final step in the procedure below calls for concentration of the
extraction solvent. Such a procedure is briefly given here.

Procedures:

Creating a calibration curve:
1. Make up solutions containing approximately 15ppm internal
standard and a caffeine concentration of approximately 0.5, 1,
3, 5, 10, 15, 20, 25, and 50ppm.
3. Create a new GC method using the GC and MS conditions
given above.
4. Inject samples from low to high concentration and create
calibration curve using mass spectra software.

Make up an Ammonium Buffer Solution:
1. Add a few scoops of ammonium chloride to 15-20mL of High
Resolution-GC grade methanol in a 25mL collection vial.
2. Place a calibrated pH electrode in the collection vial.
3. Add 28% ammonium hydroxide to the collection vial in single
drop increments, until the solution reaches a pH of 9.5. If the
pH exceeds 9.5, add more ammonium chloride to lower the pH.

Extraction of Caffeine from Urine:
The next step involves a liquid-liquid extraction of caffeine from
a bio-aqueous layer (urine) into a methanol and
dichloromethane solvent mixture. A high purity helium gas
stream is used to evaporate the dichloromethane/methanol
solvent (see Figure 7-12), and then the remaining material is
dissolved into HR-GC grade methanol.
1. Collect a sample of urine 0.5 to 1 hour after the consumption of
a drink containing caffeine.
2. Quantitatively transfer 2mL of urine to a 10mL collection vial.
35
3. Add ~100mg sodium chloride, 200!L ammonium buffer solution
4. Add a volume of your chosen internal standard to make the
concentration equal to that used in the calibration curve,
keeping in mind that the entire solution will eventually be
dissolved in 200 !L.
5. Add 5mL of dichloromethane and methanol in a 9:1 volume
ratio to the collection vial.
6. Mix vials vigorously for 2 minutes.
7. If necessary, use previously discussed emulsion techniques to
break the emulsion.
8. Place the recovered organic layer in a KD vial and the vial in a
warm water bath. (Warm water is sufficient, no hot plate is
needed)
9. Use a high purity helium gas stream to evaporate the solvent.
Once half the solvent has evaporated, use a transfer pipet to
wash the side of the KD vial with the remaining solvent. Repeat
this process again once half the remaining solvent has
evaporated. Continue to evaporate to dryness.
10. Dissolve the remaining residue in 200!L High Resolution-
GC grade methanol.
11. Using 100!L vial inserts in a standard automatic sampler
vial, inject methanol in the GC-MS.
12. Analyze the results using MS computer software for
identification and concentration using appropriate dilution
factors.


Figure 7-12. Solvent blow down.

When blowing off solvent with a gas stream, there is a delicate
balance between needle height and gas pressure. The gas should
36
maintain a small indentation on the surface of the solvent.
Submersing the needle will contaminate both the needle and your
sample and too high a pressure will blow the solvent out of the KD
vial. It is best to set a needle height, then start the gas pressure at 0
and increase the pressure slowly.



Results:

Caffeine concentrations vary depending on caffeine intake, but
usually are in the parts per million range. The procedure can be used
in a variety of class experiments. The simplest is to extract each
students urine for a range of levels. A more interesting experiment is
to have one or more students drink a cup of coffee or other relatively
high dose of caffeine and follow the clearance of caffeine from their
body with time.

Note that many other peaks are present in mass spectra of urine
extracts.

References:

Mul, S.J ., and G.A. Casella. "Confirmation and Quantitation of
Cocaine, Bezoylecgonine, Ecgonine Methyl Ester in Human Urine by
GC/MS." J ournal of Analytical Technology Vol.12 (1988): 153-55.

Thuyne, W.Van, and F.T. Delbeke. "Distribution of Caffeine Levels in
Urine in Different Sports in Relation to Doping Control Before and
After the Removal of Caffeine from the WADA Doping List." Int J
Sports Med 2006 27: 745-50.


7.4.2 Analysis of Cocaine Concentration on U.S. Currency by J ohn
Nelson and David Wallace

There has been an increase in the use of cocaine in the United
States since the late 1950s. Originally obtained in extremely small
doses through extraction from the coca plant by oral chewing,
cocaine is now harvested in large quantities and extremely high
Mzuri Handlin 6/30/11 3:20 PM
Comment: what is this figure for? should it
be referred to in the rest of the text
somewhere?
37
purity. It is often found in its crystalline form. This form of cocaine is
taken into the body by snorting through the nasal cavity, allowing the
drug to take effect quickly. The most common method for snorting
cocaine involves rolling paper money to form a straw by which the
cocaine can be sucked into the nose.

The most common currency used for snorting cocaine in the
United States is the one-dollar bill. Once a bill is used to snort
cocaine, it can then come into contact with other bills, transferring a
small portion of the residual cocaine. This proliferation of trace
amounts of cocaine has led previous studies to conclude that four out
of every five dollar bills have trace amounts (above 0.1 g) of cocaine
on them.

This study further examines the frequency of cocaine
contamination on dollar bills and employs methodologies to increase
precision in the measurement of cocaine concentration.

Experimental Procedures

This study largely employed the use of the procedure found in
Cocaine Contamination of United States Paper Currency by Oyler J .
et al. from 1996.

Ten one-dollar bills obtained from a random cash register in
Walla Walla, Washington were placed in glass vials that were filled
with 10.0 mL of HR-GC grade methanol. These vials were allowed to
stir for a period of 24 hours to ensure that all present cocaine was
dissolved from the currency. After stirring, the methanol was
decanted from the vials and 10 mL of sodium acetate buffer (10 mL,
2M, pH 4.0) was added to samples and allowed to mix. The buffered
samples were then filtered through solid-phase extraction (SPE)
columns to extract the cocaine from any other compounds that were
dissolved by the methanol. Each column was washed with deionized
water (1x, 2 mL) and 0.1 M HCl (1x, 1.5 mL) and aspirated to
dryness. The columns were then washed with methanol (2x, 1 mL)
and aspirated to dryness again. The cocaine analytes were then
eluted from the columns using a 80:20:2 ratio of methylene chloride,
isopropanol, and concentrated aqueous ammonium hydroxide. The
38
elution solvent was added to the columns (6x, 1 mL) and allowed to
drip into clean glass tubes.

The elution solvent was then evaporated to dryness using high purity
helium. This concentration of the elution solvent will provide a higher
signal to noise ratio when analyzed by gas chromatography (GC).
100 L of methanol was added to the glass tubes to allow the
cocaine analyte to redissolve. The methanol was then transferred to
glass gas chromatography (GC) vials for analysis.

An external calibration curve was made with concentrations of
10, 25, 40, 50, and 100 ppm cocaine. Five samples were analyzed by
GC-MS using a temperature program starting at 180
o
C and ending at
250
o
C with a ramp of 5
o
C per minute. The analyte concentration
was measured by comparing the analyte response to the calibration
curve responses to achieve part-per million concentration levels.

This procedure was repeated with the use of a derivatizing
agent (BSTFA with 1% TMCS) to increase the signal strength of the
analyte. The derivatizing agent was added to the final 100 L of
methanol in equal volume and allowed to heat in a 50
o
C oven for 1
hour. The derivatizing agent was also added to the external
calibration standards.

Results

Two separate calibration plots were constructed from the calibration
standards of each run. A linear regression line was fit to each of
these plots from which the concentration of cocaine in each of the
corresponding samples was determined. Five samples were run
along with the calibration standards without derivatizing agent, and
four derivatized samples were run with the derivatized standards.
Table 7-2 shows the cocaine concentrations obtained for each
sample. The derivatized samples and calibration curve yielded
substantially better results and detection limits than samples run
without derivatizing agentfor this reason, the use of a derivatizing
agent is crucial to obtaining precise and accurate results. Data for
both calibration lines are reproduced below.

Table 7.2 Instrument Calibration and Extraction Results
39
Cocaine Calibration - Underivatized
Concentration (ppm) Response Slope Y Intercept R2
25 96 316 -9182 0.99359
40 2136
50 6082
100 22810

Sample Number Response
Concentration
(ppm)
Amount On
Bill (mg)
1 2718 37.707 0.37707
2 1161 32.7734 0.327734
3 4633 43.775 0.43775
4 1740 34.6081 0.346081
5 8352 55.5593 0.555593

Cocaine Calibration - Derivatized
Concentration (ppm) Response Slope Y Intercept R2
10 1580 261 -604 0.989426
25 5786
40 11947
50 10947
100 25540

Sample Number Response
Concentration
(ppm)
Amount On
Bill (mg)
1 13589 52.0651 0.520651
2 14100 54.0229 0.540229
3 12512 47.9387 0.479387
4 3795 14.5402 0.145402


40


Figure 7-13. Total Ion Chromatogram of a Cocaine External
Calibration Standard. The retention time of cocaine is just under 20
mintutes.


Figure 7-14. Total Ion Chromatogram of a Sample Extract. Again,
the retention time of Cocaine is near 20 minutes.
41

Figure 7-15. The Mass Spectrum and Fragmentation of Cocaine.


References

J enkins, A.J . 2001. Drug Contamination of US Paper Currency,
Forensic Science International, Vol. 21, pp. 189-193.

Oyler, J ., W.D. Darwin, and E.J . Cone. 1996. Cocaine
Contamination of United States Paper Currency, J . of Analytical
Toxicology, Vol 20, J uly/August, pp. 213-216

7.5 Analytical Experiments without an External Reference Standard;
Conformational Identification without Quantification.

One of the most powerful applications of an MS system is its
ability to identify an analyte without a reference compound. In GC
(with other non-conclusive detectors), reference compounds are
needed to determine the retention time, the criteria for identification in
GC. In MS, the spectrum is the identifier since it can be compared to
thousands of reference spectra and a unique match is normally
achieved. The laboratory exercises below illustrate the power of MS
in identifying unknown analytes in a variety of samples.

7.5.1 Identification of Components in Liquors and Distilled Spirits

Distilled Spirits contain a range of flavors that can be identified
by GC-MS. Not only can an analyst tell what type of liquor is present
42
(i.e. gin versus whiskey) but they can also compare the presence and
abundance of select flavor compounds between different brands of a
given type of liquor.

NOTE/WARNING: Some liquors contain nonvolatile components that
will coat out on the glass liner in the injector port. The injector liner
may need to be cleaned or replaced after completing this laboratory
exercise to avoid damage to the GC column.

Experimental Procedures

Chemicals and Supplies:

Pure samples of a variety of liquors. One approach is to
contrast types of liquors (rum versus gin versus whiskey). Another
approach is to contrast brands (spiced liquor versus pure liquor).

GC-MS Settings:
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.25 mm; 0.25 m phase coating
Injection Volume: 1.00 L
Splitless Injection for: 0.50 min.
Split Flow Rate: 50 mL/min.
Column Flow: 1.3 mL/min.
Linear Velocity: 42 cm/s
Injector Temperature: 250
o
C
MS Transfer Line
Detector Temperature: 230
o
C
Quadrupole Temperature: 150
o
C
Oven Program: Initial Temp. at 60.0
o
C for zero minutes, 2.0
o
C
to 150
o
C, hold for zero minutes. Post Temp. at 280
o
C for 10 minutes.
Total Run Time: 45 min.

Procedures:
Analyze a variety of liquor samples on a GC-MS using the
instrumental conditions given above.

RESULTS:
43

It is relatively easy to distinguish between most types of liquors.
However, pure liquors (unspiced) such as rum and vodka produce
similar chromatograms and only contain ethanol, water, and a few
trace longer chain alcohols. Other liquors, as shown below, are
easily distinguished.

All of the compounds identified in the chromatogram below
were conclusively identified by the spectral library (typically 99
percent confidence/probability) and are known to be present in the
liquors based on a scientific literature search or from common
information found on company web sites or web searches. An
interesting project is to Goggle some of the compounds identified in
the spectra and research their origin and why they are added to a
specific liquor.

Rum: Rum is a fermented beverage made from sugarcane
byproducts such as molasses and juice. After fermentation, it is
distilled as a clear liquid. Double distillation yields the light rums,
while single distillation will yield darker rums that were originally
thought of as being of lower quality. From here the process becomes
brand specific and the initial rum can be aged in a variety of barrels,
including oak to impart strong flavors, or filtered through charcoal to
remove colors. Spices or color agents are then added.

Figures 7-16, 7-17, and 7-18 are chromatograms for Bacardi
Gold, Captain Morgan, and Citrus Rum, respectively. Note the lack
of compounds in the relatively pure Barcardi Gold rum, only 3-methyl-
1-butanol and acetic acid are present in measurable quantities.
Barcardi Gold has little presence of the oak flavor compounds such
as those found in the other two rum beverages. Captain Morgans
flavor is characterized by additional compounds, most notably oak
flavors and vanilla. Citrus Rum contains almond, orange, cocoa, fruit,
and lemon flavors, as well as extracts from the oak barrel aging
process.

44

Figure 7-16. Chromatogram of Barcardi Gold.



45
Figure 7-17. Chromatogram of Captain Morgan.


Figure 7-18. Chromatogram of Citrus Rum.

As a side note, most of the components shown in these figures are in
the parts per million range of concentrations.

Whiskey: Whiskey (originally Whisky from its origin with Irish
monks) refers to a broad range of alcoholic beverages that are
distilled from fermented grain mash and aged (matured) in oak
barrels (casks). The age of a whiskey refers to its time in the cask
(between fermentation and bottling) and the length of aging greatly
affects its chemical makeup and taste from the extraction of wood
components from the cask. These components include lacone (3-
methyl-4-octanolide) that has a coconut aroma, and numerous
phenolic compounds. Grains of choice include barley, malted barley,
rye, wheat, and corn, and it may be fermented from single or blends
of grains. Published flavoring chemicals include carbonyl
46
compounds, alcohols, carboxylic acids and their esters, nitrogen- and
sulphur-containing compounds, tannins and other polyphenolic
compounds, terpenes, and oxygen-containing heterocyclic
compounds and esters of fatty acids. The nitrogen compounds
include pyridines, picolines and pyrazines. After distillation, the
flavoring compounds that are common among different brands of
whiskey include fusel oils that are higher alcohols that are actually
mildly toxic and have a strong disagreeable smell and taste in high
concentrations. Hence, these are commonly removed by charcoal
and linen filtration. Other common flavor agents in whiskey are
acetals, such as acetaldehyde diethyl acetal (1,1-diethoxyethane),
the principal flavor agent in sherry. The presence of a buttery aroma
is due to diketone diacetyl (2,3-butanedione). Some whiskey blends
contain specific flavor agents. Use the chromatograms given below
to confirm the presence of these known whiskey components.


Figure 7-19. Chromatogram of Crown Royal.

47

Figure 7-20. Chromatogram of Southern Comfort.

Cognac: The cognac used here is Grand Marnier, a blend of
cognac. Cognacs are brandies produced from specific white grape
varieties. Most cognacs are distilled twice and aged in oak barrels.
After the review of the liquors above, the student should be able to
predict some of the compounds present in cognac.

48

Figure 7-21. Chromatogram of Grand Marnier.

Note that most of the flavor compounds come from the oak barrel or
are specifically added.

Peppermint Schnapps: Schnapps is usually a clear, colorless
beverage with a light fruit flavor since it is fermented from fruit. The
schnapps used in this experiment is infused with peppermint leaf
extract or specific flavors (chemicals) found in peppermint. Note the
dominant mint flavor compound in the chromatogram.

49

Figure 7-22. Chromatogram of Peppermint Schnapps.

7.5.2 Identification of Fragrances

Fragrances/perfumes provide a Holy Grail for GC-MS
analysis. As noted in many movies or from a trip to a European
fragrance shop (perfumery), a near infinite variety of combinations of
fragrances can be made. In this laboratory exercise, name brand
fragrances will be compared to their more inexpensive counterparts in
an effort to determine if a difference exists in their fingerprint based
on GC-MS. A fingerprint, in this context, is a characteristic
chromatogram of a complex mixture of compounds.



50
Perfumes consists of (1) primary scents at the parts per million
concentration, (2) modifiers that alter the primary scent to give the
perfume a certain desired character, (3) blenders (ingredients that
smooth out the transitions of a perfume between different bases; top,
middle, and base notes of a fragrance may have separate scents),
and (4) fixatives (natural or synthetic substance used to reduce the
evaporation rate).

Sources of primary scents include: (1) Plant sources (bark,
flowers and blossoms, fruits, leaves and twigs, resins, roots,
rhizomes and bulbs, seeds, woods, (2) Animal sources (Ambergris
which are lumps of oxidized fatty compounds, Castoreum from the
odorous sacs of the North American beaver, Civet Musk obtained
from the odorous sacs of the animals related to the Mongoose,
Honeycombs, Musk originally derived from the musk sacs from the
Asian musk deer), (3) and 0ther natural sources (extracts of lichens
and seaweed). Synthetic sources of the natural compounds
mentioned above are used today, as well as calone, linalool and
coumarin from terpenes, and salicylates (orchid scents) are also used
today.

Experimental Procedures

Chemicals and Supplies:

A variety of perfume samples can be analyzed. In this
experiment, Light Blue by Dolce and Gabbana, Shades of Blue by
Belcam, Drakkar Noir by Guy Karoche, Classic Match by Belcam,
Unforgivable by Sean J ohn, Unjustified by Belcam, and Bring It by
Parfums were used.

GC-MS Settings:
Capillary Column: DB-5
Poly(phenylmethyldimethyl) siloxane (5
% phenyl)
30 m x 0.25 mm; 0.25 m phase coating
Injection Volume: 1.00 L
Split Mode of Injection
Split Flow Rate: 131 mL/min.
Column Flow: 1.3 mL/min.
51
Linear Velocity: 42 cm/s
Injector Temperature: 250
o
C
MS Transfer Line
Detector Temperature: 230
o
C
Quadrupole Temperature: 150
o
C
Oven Program: Initial Temp. at 40.0
o
C for zero minutes, 2.0
o
C
to 280
o
C, hold for five minutes.
Total Run Time: 125 min.

Procedures:
Analyze a variety of perfume samples on a GC-MS using the
instrumental conditions given above.

Results


(1) The relatively expensive Light Blue and a generic blend
Shades of Blue:


Figure 7-23. Chromatogram of Light Blue (top figure) and Shades of
Blue (bottom figure).

52
Note the presence, absence, or reduced concentrations of the
components between the two perfumes. Names and chemical
structures for the numbered components are given in the following
table.

Table 7.5. Labeled Components in Chromatograms of Light Blue by
Dolce & Gabbana vs Shades of Blue by Belcam.

Name Structure
1)
diethylene glycol
monoethyl ether
(preservative)
O
O
OH

2) limonene (lemon scent)

3) a-cedrene (wood scent)

4) b-cedrene (wood scent)

5) thujopsene (wood scent)

6) cuparene (wood scent)

7) cedrol (wood scent)
HO

8)
diethyl phthalate
(preservative)
O
O
O
O

9)
methyl dihydrojasmonate
(jasmine)
O
COOCH
3

10)
isopropyl myristate (skin
binder) (CH
2
)
12
O
O

53
11)
1,3,4,6,7,8-hexahydro-
4,6,6,7,8,8-
hexamethylcyclopenta-g-
2-benzopyran (musk
scent)
O


(2) Drakkar Noir and the generic Classic Match:

Drakkar Noir is a blend of citrus, lavender, spices and woods. Top
notes are citrus, middle notes are woody and herbaceous and base
notes are woody warmed and spiced with aromatic coriander and
juniper berries, strengthened by sandalwood, patchouli and fir
balsam.



Figure 7-24. Chromatogram of Drakkar Noir (top figure) and Classic
Match (bottom figure).


Table 7.6. Labeled Components in Chromatograms of Drakkar Noir
and Classic Match.

Name Structure
54
1) limonene (lemon scent)

2)
dihydromyrcenol (lime
scent)
OH

3) linalool (spicy floral scent)
HO

4)
4-Allylanisole (minty sweet
scent)
O

5)
linalyl acetate (sweet
scent)
O
O

6)
2,6-ditertbutyl-4-
methylphenol (antioxidant)
OH

7)
diethyl phthalate
(preservative)
O
O
O
O

8)
patchouli alcohol (woody
scent)
HO

9) verymoss (woody scent) HO
OH
O
O

10) d-cadinene (woody scent)
H

11)
benzyl salicylate (floral
scent)
O
O OH







55
(3) Unforgivable by Sean J ohn vs Unjustified by Belcam Inc. vs Bring
It by Parfums de Coeur


Figure 7-25. Chromatogram of Unforgivable (top figure) and
Unjustified (bottom figure).


Figure 7-26. Chromatogram of Bring It.

Table 7.7. Labeled Components in Chromatograms of Drakkar Noir
and Classic Match.

Name Structure
1) propylene glycol
OH
OH

56
2) limonene (lemon scent)

3)
dihydromyrcenol (citrus
scent)
OH

4) tricyclene (citrus scent)

5)
methyl dihydrojasmonate
(jasmine)
O
COOCH
3

6)
1,1,3-trimethyl-3-
phenylindan

7)
acetyl cedrene (musty
scent)
OH
O

8)
isopropyl myristate (skin
binder) (CH
2
)
12
O
O

9)
1,3,4,6,7,8-hexahydro-
4,6,6,7,8,8-
hexamethylcyclopenta-g-
2-benzopyran (musk
scent)
O

10) Versalide (musky scent)
O

11)
6-tert-butyl-3-methyl-2,4-
dinitroanisole (musky
scent)
O O
2
N
O
2
N

12)
Octyl 4-methoxycinnamate
(keratin binder)
(CH
2
)
3
O
O
OCH
3



57
Comparisons of the real versus fake perfumes show distinct
similarities with respect to presence of peaks and their fingerprint.
However, closer inspection of each peak shows differences. These
subtle differences change our olfactory perception of their smell.
Note the identification of the components in each table and the type
of the compounds present and their purpose.

7.5.3 SPME-GC-MS Analysis of Wine Headspace by Bailey Arend

For many consumers, the aroma of a wine is nearly as
important as the flavor. The wine industry is obviously interested in
producing wine with pleasing and abundant aroma. More than 1000
compounds have been identified in the headspace of wine, including
alcohols, esters, carbonyls, acids, phenols, lactones, acetals, thiols,
terpenols and many more (Weldegergis, et al, 2007; Polaskova, et al,
2008)] Although human senses can detect surprisingly small
concentrations of certain volatile organic compounds in wine
headspace, analytical instrumentation provides a more specific and
precise way to measure the headspace character of wine.

The complex matrix of wine, as well as the low concentration of
some of the volatile compounds presents further obstacles in the
characterization of wine aroma. To analyze many of the compounds,
sample enrichment techniques must be employed (Weldegergis, et
al., 2007) liquid-liquid extractions using organic solvents (Andujar-
Ortiz, et al., 2009; Ortega-Heras, et al., 2002) and solid phase
extraction (SPE) (Andujar-Ortiz, et al., 2009; Dominguez, et al., 2002)
are both effective for wine analyses, however solid-phase micro
extraction (SPME) presents a major advancement in volatile
compound analyses.

SPME was first introduced in 1989 by Belardi and Pawliszyn for
analysis of organic pollutants in water. The original method involved
immersing fiber coated with fused-silica stationary phase directly in
the liquid analyte[6]. Analytes are sorbed/adsorbed onto the solid
phase, which can then be inserted directly into a gas chromatograph
(GC) injector, where the analytes are thermally desorbed and loaded
onto the GC column. The newer SPME has drawn much attention for
being versatile, yet simple. The technique does not require
expensive, high-purity, toxic organic solvents generally associated
58
with instrumental analysis and eliminates many possible sources of
error.

SPME was first modified for headspace analysis in 1993. The
new method exposed the coated fiber to the sample headspace only,
which was found to shorten the extraction time while maintaining
detection limits in the ppt range (Zhang and Pawliszyn, 1993). The
driving theory behind any SPME is the partition coefficient of the
analyte between the coating and the solvent or vapor. The partition
coefficient along with the large difference in volume between the
coating and headspace volume result in impressive concentration
factors. The mass of analyte adsorbed to the coating (n) is given by

n=C
o
V
1
V
2
K
1
K
2
/(K
1
K
2
V
1
+K
2
V
3
+V
2
),

where C
o
is the original concentration in the liquid phase, and the
volumes of the three phases in equilibrium are as follows: V
1
for the
coating, V
2
for the liquid phase, and V
3
for the headspace (Zhang and
Pawliszyn, 1993). Minimizing the ratio of headspace to sample
volume (V
3
<<V
2
) and maximizing the affinity of the coating for analyte
(large K
1
) can boost the amount of adsorbed analyte. This allows
direct injection of analyte to the GC without risking instrument
damage by injecting concentrated and sugary wine matrix. HS-SPME
also avoids many complications of matrix effects, even allowing
analysis of solid samples and human blood (Cardinali et al., 2000), as
long as the analyte is volatile (Zhang and Pawliszyn, 1993). Tat et al.
found that 50/30m Divinylbenzene/Carboxen/Polydimethylsiloxane
coated fiber gave the most sensitive and repeatable results for the
analysis of wine headspace (Tat, 2005).

Aside from fiber coatings and volume ratios, other parameters
that can affect the sensitivity of HS-SPME are exposure time,
temperature, and pH of the sample solution (wine). Exposure time is
logically related to the concentration of analyte sorbed to the fiber.
Sufficient time must be given for the system to reach equilibrium
before the equation above is valid. Temperature governs the fraction
of analyte present in the headspace and available for adsorption.
Many methods immerse the extraction vial in a heated water bath
(Tat, 2005), however, care must be taken that the fiber, headspace
and condensed phase are all in thermal equilibrium. The pH values
59
have been adjusted in some studies (Boutou and Chatonnet, 2007) to
allow multiple classes of molecules to be in their most analyzable
form. For example, to simultaneously adjust for pyrazines (which are
best analyzed at neutral to basic pH values, pKa ~0.50) and phenols
(pKa ~25, which are best analyzed at low pH values) Boutou and
Chatonnet adjusted all samples to a pH value of 7.

The method of Boutou and Chatonnet was also sufficient for
analysis of contaminants that cause off flavors in wine. Compounds
such as 2,4,6-trichloroanisole, 2,3,4,6-tetrachloroanisole, 2,4,6-
tribromoanisole have olfactory perception thresholds near 10 ng L
-1

and give wine a barnyard character (Boutou and Chatonnet, 2007).
Wines with these contaminants are referred to as corked and are
quite undesirable. Analysis of such contaminants can determine the
origin of contamination and improve wine production techniques
(Boutou and Chatonnet, 2007).

Materials and Methods:

Sampling conditions (adopted from Tat, 2005)

Sample wines and all equipment were stored at room
temperature to ensure thermal equilibrium and minimize thermal
differences between samples. Thirty-two (32.0) mL of sample wine
was pipetted into a 40-mL glass vial equipped with a septum. The
septum was pre-punctured with a sharp, hollow needle to avoid
contaminating or breaking the fiber by contact with the septum. The
extraction fiber was a Supelco 50/30m
Divinylbenzene/Carboxen/Polydimethylsiloxane Stableflex fiber
conditioned at 270C in the GC inlet for 1 hour. The fiber was
inserted into the vial via the septum before being exposed. The fiber
was exposed to the headspace while the wine was stirred. The
extraction was performed at 25C for 15 minutes. When finished, the
fiber was immediately inserted into the gas chromatograph injector,
where it remained for the entire duration of the temperature program.

Instrumental Parameters (adopted from Boutou and Chatonnel, 2007)

GC/MS analysis was performed by:
Instrument: Agilent 19091S-433
60
Column: HP-5MS 5% Phenyl Methyl Siloxoane 30.0m x 250 m x
0.25 m nominal capillary column.
Carrier gas: helium (ultra high purity, 99.999% passed through
hydrocarbon traps) programmed to flow at a constant linear rate of
54.1 mL/min for the entire run. The injector lining was ensured to be
long enough to allow full insertion of the fiber and operated by manual
injection. The injector was operated in splitless mode at 270C for the
entirety of the each run.

The oven program started at 50C for 2.0 min, and then
increased at 3.0C min
-1
to 190C. The temperature was than
increased at 50C min
-1
to 320C where it was held for 5min.
Detection was performed by an Agilent 5975C inert EI/CI MSD
quadrupolar mass detector with EI ionization (source temperature
230C, quadrupole temp. 150C, energy of constant ionization 70).
The entire method required a total of 56 min.

Results:

This study was conducted to demonstrate the ease and
versatility of HS-SPME for student chemists. For this reason, no
adjustments were made to the wine samples and the recommended
heating of the samples was not performed. Although the range of
detectable analytes was much smaller than other methods have
reported (Tat, 2005; Boutou and Chatonnel, 2007), the quick and
simple method returned multiple analyte peaks for each wine tested.

It was found to be important to thermally clean the extraction
fibers before each exposure. Samples of Black Box Merlot were run
without cleaning the fiber and unexpected compounds were observed
(shown in Figure 1). Many silica compounds were found in the blank
runs, which could possibly be attributed to degradation of the fiber
coating (shown below in Figure 2). If the fiber was baked out directly
before exposure, the presence of these compounds was minimized
(shown below in Figure 3).

The method was repeated 5 times on Black Box 2007 California
Merlot and the results were found to be reproducible. One advantage
of using boxed wine is that the affects of oxidation are eliminated.
Samples may be drawn weeks apart, whereas opening a bottled wine
61
introduces oxygen to the entire bottle and could affect the
composition of the wine (Simpson, 1978).

While the method was found to be repeatable, only a limited
number of identifiable compounds were recovered for each sample.
These compounds are shown in Figure 7-26 below and include 3-
methylbutyl acetate, ethyl hexanoate, 2-phenylethanol, diethyl
succinate, ethyl octanoate and ethyl decanoate. The chromatogram
for Chardonnay (Buckleys Cove, South Eastern Australia, 2009,
Figure 6) does not contain 2-phenylethanol and diethyl succinate
which were found in all red wine samples. It is possible that the lack
of these two compounds could be a signifier of white wine.


Figure 7-26. The compounds repeatedly found in all red wine
samples by the proposed method.

It was thought that a higher-quality wine would have a stronger
bouquet and yield more volatile compounds, however the
chromatogram of Red Table Blend from Walla Walla Village Winery
(82%Cabernet Sauvignon, 9% Merlot and 9% Cabernet Franc, Figure
7) did not show additional peaks. This method was also insufficient to
characterize contaminants in a sample of corked wine from Foundry
Vineyards in Walla Walla (2005 Red Wine), (shown in Figure 8). It is
likely that heating of the samples or adjustment of the other
parameters, such as pH, would increase the range of compounds
detectable by this method.

Although this simplified method does not engage the full potential of
HS-SPME techniques, it is sufficient to demonstrate the theory and
application of such techniques in a college/university laboratory.
62
Additional streamlined methods may be developed for other
consumable complex matrices such as whiskey or vinegar.


Chromatograms



Figure 7-27. The chromatogram for Black Box California Merlot,
2007. No blank was run and unusual peaks were observed.
63

Figure 7-28. The fiber was subjected to an entire run without
exposure to any sample. Multiple contaminants were observed,
demonstrating the need to run blanks and thermally clean the fiber
between samples.



64


Figure 7-29. The chromatogram for Black Box California Merlot,
2007. The fiber was exposed to the sample immediately after
subjecting it to a blank run. Fewer contaminants were observed.




65

Figure 7-30. The chromatogram for Buckleys Cove 2009 Shiraz from
South Eastern Australia. Relatively few volatile compounds were
detected.


3
-
m
e
t
h
y
l
b
u
t
y
l

a
c
e
t
a
t
e

3
-
m
e
t
h
y
l
b
u
t
y
l

a
c
e
t
a
t
e

66
Figure 7-31. The chromatogram for Buckleys Cove 2009
Chardonnay from South Eastern Australia differs slightly from
Buckleys Cove Shiraz (Figure 4). 2-Phenylethanol and diethyl
succinate that were found in the Shiraz were not observed in this
sample.



Figure 7-32. The chromatogram for Walla Walla Village Winerys Red
Table Blend (82%Cabernet Sauvignon, 9% Merlot and 9% Cabernet
Franc). Contrary to our prediction, no additional compounds were
detected in the higher-quality wine. Changing the sample parameters
could aid the detection of additional volatiles.

3
-
m
e
t
h
y
l
b
u
t
y
l

a
c
e
t
a
t
e

67

Figure 7-33. The chromatogram for Foundry Vineyards 2005 Red
Wine. Corked wines have a displeasing aroma caused by
trichloroanisole or tribromoanisole (Boutou and Chatonnet, 2007).
The streamlined method could not detect the presence of these
compounds, however changing the sample parameters such as
temperature or pH could improve detection of these contaminants.

References

Weldegergis, B.T., A.G.J . Tredoux, and A.M. Crouch, Application of a
Headspace Sorptive Extraction Method for the Analysis of Volatile
Components in South African Wines. J ournal of Agricultural and Food
Chemistry, 2007. 55(21): p. 8696-8702.

Polaskova, P., J . Herszage, and S.E. Ebeler, Wine flavor: chemistry
in a glass. Chemical Society Reviews, 2008. 37(11): p. 2478-2489.

Andujar-Ortiz, I., et al., Analytical performance of three commonly
used extraction methods for the gas chromatography-mass
spectrometry analysis of wine volatile compounds. J ournal of
Chromatography A, 2009. 1216(43): p. 7351-7357.

68
Ortega-Heras, M., M.L. Gonzalez-SanJ ose, and S. Beltran, Aroma
composition of wine studied by different extraction methods. Analytica
Chimica Acta, 2002. 458(1): p. 85-93.

Dominguez, C., D.A. Guillen, and C.G. Barroso, Determination of
volatile phenols in fino sherry wines. Analytica Chimica Acta, 2002.
458(1): p. 95-102.

Robert P. Belardi, J .B.P., Application of Chemically Modified Fused
Silica Fibres in the Extraction of Organics from Water Matrix Samples
and the Rapid Transfer to Capillary Columns. Water Pollution
Research J ournal of Canada, 1989. 24: p. 179-191.

Zhang, Z. and J . Pawliszyn, Headspace solid-phase microextraction.
Analytical Chemistry, 1993. 65(14): p. 1843-1852.

Cardinali, F.L., et al., The use of solid-phase microextraction in
conjunction with a benchtop quadrupole mass spectrometer for the
analysis of volatile organic compounds in human blood at the low
parts-per-trillion level. J ournal of Chromatographic Science, 2000.
38(2): p. 49-54.

Tat, L., et al., Optimization of wine headspace analysis by solid-
phase microextraction capillary gas chromatography with mass
spectrometric and flame ionization detection. Food Chemistry, 2005.
93(2): p. 361-369.

Boutou, S. and P. Chatonnet, Rapid headspace solid-phase
microextraction/gas chromatographic/mass spectrometric assay for
the quantitative determination of some of the main odorants causing
off-flavours in wine. J ournal of Chromatography A, 2007. 1141(1): p.
1-9.

Simpson, R.F., AROMA AND COMPOSITIONAL CHANGES IN
WINE WITH OXIDATION, STORAGE AND AGING. Vitis, 1978.
17(3): p. 274-287.


7.5.4 An Extraction Procedure for the Investigation of Pesticide
Residues on Strawberries by Kyle Byrd-Fisher
69

Agriculture is as vital to the health of the United States and the
world as ever. Over the course of the 20th century, the scale of food
production in the United States increased dramatically to cope with
the increased demand of the larger population. An increase in the
scale of food production is more cost effective for farmers, but it does
have some very significant draw-backs. Most notably, large-scale
production carries with it the cost of crop vulnerability. If a pest is
mobile or transferable, such as an insect or fungi, it has the potential
to damage a much larger quantity of food. For this reason, the ability
to contain pests or eliminate them is of vital importance to the modern
farmer. This is how pesticides have found a home in modern
agriculture, especially within the United States.

To better give an indication of the pervasive use of pesticides,
in 2006, of the states surveyed by the United States Department of
Agriculture Pesticide Data Program (USDA PDP), 98 percent of all
head lettuce producing acreage received applications of insecticide,
63 percent of head lettuce acreage received applications of herbicide,
and 87 percent of head lettuce acreage received applications of
fungicide

(United States, 2010). One of the issues with the broad use
of pesticides is that the health effects of human consumption have
often not been sufficiently studied by the time a pesticide is approved
by the Environmental Protection Agency (EPA). In the case of
azinphos methyl, which was approved as a coddling moth insecticide
in 1959, it was not until 2006 that the EPA decided to phase it out due
to its consumption and application toxicity

(United States EPA).
Pesticides are currently regulated by three separate branches of the
Federal Government: the EPA is responsible for approving pesticides
for use and setting tolerance levels on their presence in or on food,
the USDA is responsible for monitoring the pesticide residue levels in
food with the PDP, and the Food and Drug Administration (FDA) is
responsible for enforcing standards with the Total Diet Study. This
paper is concerned with the monitoring aspects of pesticide
regulation, specifically extraction and identification methods for
determining the presence of pesticide residues on the surfaces of
fruit. The focus is on strawberries, which, like head lettuce, have
significant quantities of pesticides applied on them every year.

Pesticides on Strawberries
70

In 2008, 2.5 billion pounds of strawberries were produced
domestically, the vast majority coming from California (United States,
DoA, 2008). Strawberry monitoring by the PDP in 2008 came in the
form of 741 samples analyzed with a total of 113,071 analyses
performed. Of these analyses, 3.3 percent detected pesticide
residues, with 46 different pesticide detections reported
1
. Of these
detections, 16 pesticides were detected on more than 5 percent of
the samples, with the top three pesticide detections being boscalid,
captan, and myclobutanil in that order (Anastassiades, 2003).

Extraction Standard Operating Procedure:

Strawberry samples were obtained from Safeway, which
purchased them from Boskovich Farms of Oxnard, CA. A sample
size of seven strawberries was obtained by washing the surface of
each strawberry twice with methylene chloride. No prior wash with
water was conducted. The methylene chloride wash was evaporated
overnight and reconstituted in 10 mL of methylene chloride. The
reconstituted methylene chloride organic wash was then put into a
separatory funnel and washed twice with a concentrated aqueous
sodium chloride solution. The remaining organic layer was then
filtered through a syringe filter into a GC vial, totaling 0.5 1.0 mL. A
methylene chloride blank was created and run on an Agilent 6890N
GC-MS (quadrupole) alongside the sample. The temperature
program was set to run at 90C for 2 minutes before ramping by 2C
per minute to 270C.

Results:

This extraction procedure produced a chromatogram showing
evidence of three sharp peaks. The first peak had a retention time of
52.716 minutes, a percent area of 19.58%, and an approximate
number of counts of 300,000. The library search report conclusively
identified this compound (quality of 96) as cis-1,2,3,6-
tetrahydropthalimide also known as cis-4-Cyclohexene-1,2-
dicarboximide. This is a breakdown product of captan, which was
identified as the third peak. The second peak had a retention time of
74.38 minutes, a percent area of 4.81%, and an approximate number
of counts of 80,000. The library search report conclusively identified
71
this compound (quality of 98) as cyprodinil or 4-cycloproplyl-6-methyl-
N-phenyl-pyrimidinamine. This compound is an insecticide
commonly applied to the foliage of grapes, almond trees, and stone
fruit targeting scab and brown rot blossom. The third peak had a
retention time of 81.401 minutes, a percent area of 75.61%, and an
approximate number of counts of 1,200,000. The library search
report conclusively identified this compound (quality of 99) as captan.
Captan was listed above as the second most common pesticide
found on strawberries and is also listed as a Pesticide Action Network
(PAN) bad actor and a probable carcinogen (EXTOXNET, 2011).
Captan is a broad-spectrum fungicide applied on many fruits and
vegetables, and is also applied to the surfaces of fruit after harvesting
to improve the fruits appearance (PAN Pesticide Database, 2011).



Figure 7-34. Total Ion Chromatogram of Strawberry Extract.

Conclusion
This investigation was not meant as a quantitative study as in
the PDP analyses, but rather as a proof of extraction/concept and
analyte identification without an external standard. Finding that
pesticides residues can be readily identified with an easy extraction
procedure is a strong starting point for a more extensive quantitative
project.

References

United States. Department of Agriculture. National Agricultural
Statistics Service. Agricultural Statistics 2010. Ed. Rich Holcomb.
Print.

72
United States. Environmental Protection Agency. California.
Azinphos-Methyl Risk Characterization Document. By Carolyn M.
Lewis. Print.

United States. Department of Agriculture. Agricultural Marketing
Service. Pesticide Data Program Annual Summary 2008. Pesticide
Data Program. Web. <http://ams.usda.gov/pdp>.

United States. Department of Agriculture. Agricultural Marketing
Service. USDA Pesticide Data Program Analytical Methods.
Agricultural Marketing Service. Web.
<http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRDC
5049940>.

Anastassiades, M.; Lehotay, S. J .; Stajnbaher, D.; Schenck, F. J .
Journal of AOAC International. 2003, 86, 412-31.

"Captan Pesticide Information Program." EXTOXNET - The
EXtension TOXicology NETwork. Cornell University. Web.
<http://extoxnet.orst.edu/pips/captan.htm>. Accessed 2011.

"Captan - Toxicity, Ecological Toxicity and Regulatory Information."
PAN Pesticide Database. Pesticide Action Network. Web.
http://www.pesticideinfo.org/Detail_Chemical.jsp?Rec_Id=PC34569,
Accessed 2011.