Analytica Chimica Acta 597 (2007) 5866

Study of fibrinogen adsorption on hydroxyapatite and TiO2 surfaces by

electrochemical piezoelectric quartz crystal impedance
and FTIRATR spectroscopy
Qin Yang, Youyu Zhang , Meiling Liu, Min Ye, YuQin Zhang, Shouzhuo Yao
Key Laboratory of Chemical Biology and Traditional Chinese Medicine Research (Ministry of Education),
College of Chemistry and Chemical Engineering, Hunan Normal University, Changsha 410081, PR China
Received 20 August 2006; received in revised form 13 February 2007; accepted 12 June 2007
Available online 19 June 2007

Abstract
The electrochemical piezoelectric quartz crystal impedance (EQCI), a combined technique of piezoelectric quartz crystal impedance (PQCI),
electrochemical impedance (EI), and Fourier transform infrared spectroscopyattenuated total internal reflectance spectroscopy (FTIRATR) were
used to in situ study the adsorption process of fibrinogen onto the surface of biomaterialsTiO2 and hydroxyapatite (Ca5 (PO4 )3 OH, HAP). The
equivalent circuit parameters, the resonance frequencies and the half peak width of the conductance spectrum of the two biomaterial-modified
piezoelectric quartz crystal (PQC) resonances as well as the FTIRATR spectra of fibrinogen during fibrinogen adsorption on TiO2 and HAP
particles modified electrode surface were obtained. The adsorption kinetics and mechanism of fibrinogen were investigated and discussed as well.
The results suggested that two consecutive steps occurred during the adsorption of fibrinogen onto TiO2 and hydroxyapatite (HAP) surface. The
fibrinogen molecules were firstly adsorbed onto the surface, and then the rearrangement of adsorbed fibrinogen or multi-layered adsorption occurred.
The FTIRATR spectroscopy investigations showed that the secondary structure of fibrinogen molecules was altered during the adsorption and
the adsorption kinetics of fibrinogen related with the variety of biomaterials. These experimental results suggest a way for enriching biological
analytical science and developing new applications of analytical techniques, such as PQCI, EI, and FTIRATR.
2007 Published by Elsevier B.V.
Keywords: Electrochemical quartz crystal impedance; Fourier transform infrared spectroscopyattenuated total internal reflectance spectroscopy; Adsorption kinetics;
TiO2 ; Hydroxyapatite; Fibrinogen

1. Introduction
Fibrinogen is a major plasma protein and one of the most
relevant proteins adsorbed on the surfaces of blood contact
biomaterials. It plays a crucial role in the regulation of both
haemostasis and thrombosis [1,2]. Fibrinogen is a 340 kDa protein containing two sets of three polypeptide chains (, , and ).
The simplest fibrinogen model comprises three spherical regions
(one E domain and two D domains) connected by two narrow
rods [3]. Previous studies [13] have shown that the N-termini
of all six chains folds into the central E domain and the Ctermini of the two sets of and chains folds into the two

Corresponding author. Tel.: +86 731 8865515; fax: +86 731 8865515.
E-mail addresses: zhangyy@hunnu.edu.cn,
yyshmr@yahoo.com.cn (Y. Zhang).
0003-2670/$ see front matter 2007 Published by Elsevier B.V.
doi:10.1016/j.aca.2007.06.025

outer D domains. As the two chains are longer than or

chain, their C-termini depart from the D domains and fold back
to form two small globular domains located close to the central
E domain. The total length of the molecule is about 47.5 nm.
Under physiological condition, fibrinogen is negatively charged
(Ip : 5.5) with most of the negatively charged residues residing in the E and D domains. The C (C-termini of chain)
domains, which are rich in Arg and Lys residues, are actually positively charged at neutral pH condition. Moreover, the
E and D domains are more hydrophobic than the C domains
[3].
It is well known that HAP and TiO2 are excellent biomaterials due to their biocompatibility. They have been widely used
as dental and orthopedic implant materials. A basic study of
protein adsorption onto TiO2 and HAP nano-particles will contribute to the elucidation of the mechanisms involved in their
biocompatibility.

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

Many kinds of methods, including circular dichroism and fluorescence [4], surface plasmon resonance [5], high-performance
liquid chromatography [6] and UVvis spectrophotometer [7]
have been employed to investigate the binding of fibrinogen to
biomaterials and polymers. However, some of these methods are
sensitively limited, or cannot provide real time information on
protein adsorption behavior.
Electrochemical quartz crystal microbalance (EQCM) has
been widely used to monitor the electrochemical process, due to
the ability to in situ respond to the changes in mass loading on
the electrode surface down to the monolayer or sub-monolayer
level [810]. However, the conventional EQCM technique only
provides piezoelectric information on the oscillating frequency
of the piezoelectric quartz crystal (PQC), and thus gives no direct
insight into foreign film rigidity [11].
In situ measurement of electrochemical piezoelectric quartz
crystal impedance (EQCI), a combination of piezoelectric quartz
crystal impedance (PQCI) and electrochemical impedance
(EI), can characterize the piezoelectric quartz crystal (PQC)
resonance more precisely and provide multidimensional piezoelectric information simultaneously during electrochemical
perturbations. PQCI analysis is based on the Butterworthvan
Dyke (BVD) equivalent electrical circuit [1113] that contains a motional arm (the motional resistant R1 , the motional
capacitance C1 , and the motional inductance L1 in series), in
parallel with a static arm (the static capacitance C0 ). R1 , L1 , C1
and C0 can be obtained by simultaneously fitting experimental conductancefrequency (Gf) and susceptancefrequency
(Bf) curves to corresponding theoretical ones obtained from
the BVD model. In addition, the PQC resonance is frequently

analyzed using the resonant frequency (f0 = 1/2 L1 C1 )

and the half-peak width of a Gf curve (fG1/2 ) [11]. Electrochemical impedance (EI) allows the measurements of the
interfacial capacitance of the electrode/solution interface as
well, and it can therefore be used to study the film characters
of biomaterial-modified electrodes [14,15]. The electrochemical
impedance (Z) can be represented as a sum of the real (Zre) and
imaginary (Zim ) components (Z = Zre + jZim , where j = 1)
that originate generally from the resistance and capacitance
of an electrolytic cell, respectively. The surfaceadsorption
kinetics as well as mass-transport parameters can be obtained
by measuring the electrochemical impedance spectrum at
a fixed measurement frequency with appropriate equivalent
circuits.
The EQCI has been used as a powerful method for studying
interaction of protein with drug [16,17]. In order to enrich the
applications of analytical techniques such as EQCI, FTIRATR,
etc., on biological analytical science, we made an attempt to
elucidate some basic aspects of the biocompatibility of TiO2
and HAP nano-particles by EQCI and FTIRATR investigations on the protein adsorption. In this work, we focused
our attentions on the adsorption behaviors of fibrinogen on
TiO2 and HAP nano-particle surface and detail investigations
about the kinetics and mechanism of the adsorption. TiO2 and
HAP nano-particle modified gold electrodes were prepared,
EQCI technique was adopted in this work. In addition, in
order to explore the secondary structure alteration of protein

59

molecules during the adsorption, Fourier transform infrared

spectroscopyattenuated total internal reflectance (FTIRATR)
technique, a sensitive diagnostic tool for determining secondary structure of proteins and monitoring the nature of
changes in the conformation of protein [18,19], was used in
this work. On the basis of the multi-dimensional information
provided by EQCI analysis, the process of fibrinogen adsorption onto TiO2 and HAP nano-particle surface was studied.
The adsorption kinetics and the secondary structure changes
of fibrinogen were studied by FTIRATR spectroscopy method
as well.
2. Experimental
2.1. EQCI measurements
The EQCI measurement device used in this work was
described by Xie et al. [11]. It is composed of a HP4395A
network/spectrum/impedance analyzer and a CHI660A electrochemical workstation (CH Instruments Co., USA). Synchronous
conductance (G) and susceptance (B) measurements were conducted on the HP4395A controlled by a user-written Visual
Basic (VB) 5.0 program. BVD equivalent circuit parameters
were obtained during experiments using the same VB program
by fitting each group of G and B to the BVD model based on a
GaussNewton non-linear least-squares fitting algorithm and a
selection of R1 , C0 , f0 , and 1/C1 as estimation parameters. The
simultaneous electrochemical experiments were conducted on
a CHI660A electrochemical workstation. The electrochemical
impedance data were analyzed with the EQUIVCRT software.
The HP 4395A network/spectrum/impedance analyzer was used
to measure the response of PQC. AT-cut quartz crystals (9 MHz,
12.5 mm in diameter) with gold electrodes (6 mm in diameter) on both sides were used and one of the electrodes was
contacted with the solution and served as the work electrode
and the other side placed in air. Electrochemical experiments
were performed on the CHI660A electrochemical workstation
(CH Instruments Co., USA). A carbon rod (0.5 mm in diameter, 1 cm in length) and a saturated calomel electrode (SCE)
were used as auxiliary electrode and reference electrode, respectively. All the potentials measured in this article are relative to
the SCE.
To remove possible surface contamination, the PQC electrode was first treated with nitric acid, and then potential cycling
between 0 and 1.5 V in 0.2 M HClO4 solution until reproducible
cyclic voltammograms were obtained.
A certain amount of nanocrystallite titanium dioxide (TiO2 )
powder, which was purchased from Degussa (P25, ca. 80%
anatase and 20% rutile) was dispersed into doubly distilled
water, and then ultrasonic oscillated for 30 min to form TiO2
suspension (TiO2 , 104 M). Nanocrystallite HAP powder was
pre-prepared and characterized according to a published paper
[20], and HAP suspension (HAP, 104 M) was prepared in the
same way as the preparation of TiO2 suspending solution. Then,
the gold surface of the PQCs were coated with TiO2 or HAP
by dispersing TiO2 or HAP suspension on it and vaporizing
water, and then were baked under an infrared lamp for 5 h in

60

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

order to consolidate the coupling of the biomaterials and gold

surface.

3. Results and discussion

3.1. Characterization of modied electrodes

2.2. Surface characterization and spectrum determination

All infrared spectra were recorded at room temperature
with a Nicolet Nexus 670 FTIR spectrometer (Nicolet Instrument Co., Madison, WI) equipped with a ZnSe attenuated
total reflectance (ATR) accessory, a DTGS detector and a
KBr beam splitter. TiO2 and HAP thin films were prepared
by overnight room-temperature evaporation of 200 L of suspending liquid on the horizontal surface of a 45 single
reflection ZnSe ATR prism. Adsorption experiments were carried out by placing 20 g/L fibrinogen (in PBS) stock solution
in contact with the biomaterial films for 1 h. The IR spectra
were collected with 32 scans and 4 cm1 resolution and the
buffer spectral contribution was subtracted according to the
method previously published [21,22]. The water vapor spectral
contribution was also removed by standard spectral subtraction procedures. All spectral manipulations were performed
using GRAMS AI (Galactic Industries Corp., Salem, NH,
USA).
A JEOL JSM-6700 field emission scanning electron
microscopy (SEM, Japan) and a Y-2000 X-ray diffraction
(XRD, China) were employed to characterize the size of TiO2
and HAP particles. Specific surface area of TiO2 and HAP
were measured by using surface area analyzer (Micromeritics Tristar-3000, USA) with BET nitrogen adsorption analysis
method.

3.1.1. Characterization by IR, X-ray, BET nitrogen

adsorption and SEM of TiO2 and HAP
TiO2 with 80% anatase and 20% rutile was purchased from
Degussa and HAP particle was prepared according to Ref.
[20]. The HAP particles were characterized by FTIR and Xray diffraction analysis (the results of FTIR and X-ray spectra
were not shown here). In IR spectrum, the characteristic peaks
of phosphate groups (550600, 964, 10201120 cm1 ), OH
(632 cm1 ), PO (563.5, 602.5 cm1 ), PO(H) (874.0 cm1 )
were found, which are accordant with the report by Raynaud
[23]. The X-ray diffraction spectrum was almost completely
overlapped with the standard spectrum of Ca5 (PO4 )3 OH. The
BET surface areas of TiO2 and HAP were measured by using
Micromeritics Tristar-3000 surface, which were 87.33 m2 /g and
18.98 m2 /g for TiO2 and HAP, respectively. The result suggested
that the specific surface area of TiO2 was much higher than that
of HAP. The surface morphologies of the TiO2 and HAP modified gold electrodes before and after protein adsorption were
also investigated with SEM method. Fig. 1A and B show the
SEM images of TiO2 and HAP modified surfaces. It was found
that the particle size of TiO2 and HAP was ca. 50 and 400 nm,
respectively, and the HAP particles are formed by loose agglomerated fine needle-like crystals. Fig. 1 B and D shows the SEM
images after fibrinogen adsorbed on HAP and TiO2 . The particle
size increased and the surface became smooth, indicating that a
protein film was formed on the surface of TiO2 and HAP.

Fig. 1. SEM images of TiO2 (A), HAP (B) modified electrode surfaces and fibrinogen immobilized TiO2 (C), HAP (D) surfaces.

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

3.1.2. Characterization by cyclic voltammograms and

electrochemical impedance spectroscopy
In order to clarify the difference among the modification
films of the PQC electrodes, cyclic voltammograms (CV) and
electrochemical impedance (EI) measurements were carried out
before and after TiO2 and HAP modification. The experiments
were conducted in a 2 mM K3 Fe(CN)6 /K4 Fe(CN)6 PBS solution
(50 mM, pH 7.4) at a scanning rate of 50 mV/s. Fig. 2A shows
that the redox couple of K3 Fe(CN)6 /K4 Fe(CN)6 exhibits a more
irreversible behavior after TiO2 or HAP covered on the gold electrodes. Fig. 2B and C exhibits the electrochemical impedance
spectroscopy of the bare gold electrode, TiO2 and HAP modified electrode before (B) and after (C) fibrinogen adsorption.
The Nyquist diameter of the TiO2 electrode is much larger than
that of the bare gold electrode, but it is slightly smaller than that
of HAP electrode. Which also indicates that the redox couple
of K3 Fe(CN)6 /K4 Fe(CN)6 exhibits a more irreversible behavior
after TiO2 or HAP covered on the gold electrodes. The above
results suggested that the active sites on the gold electrodes surface were occupied by TiO2 or HAP particles and the electron
exchange ability of the electrode surfaces was further weakened
after the adsorption of fibrinogen molecules.

3.2. Piezoelectric electrochemical impedance studies of the

brinogen adsorption
Figs. 3 A1and B1 show f0 and R1 responses during the
adsorption of fibrinogen on TiO2 and HAP modified electrodes.
Fixed volume of stock fibrinogen solution was added continuously into the test solution at fixed time intervals until the
sustained f0 change was obtained. The experimental results show
that the adsorption of fibrinogen on the two surfaces was saturated when the final concentration of fibrinogen reached ca.
1.768 g/L in our system. Thus, 2.0 g/L fibrinogen concentration
was selected in following experiments. It should be noted that
addition of the same volume of background electrolyte solution caused no response, indicating that the responses of the
PQC were resulted from the adsorption of fibrinogen onto the
electrode.
Fig. 3 A2 and B2 show typical responses of f0 , R1 , C0 ,
fG1/2 and Cs obtained simultaneously during the adsorption
of fibrinogen on TiO2 and HAP electrodes. Additions of fibrinogen in all cases led to abrupt increases in R1 , C0 and fG1/2 but
decreases in f0 and Cs . As indicated by the arrows in Fig. 3
A1 and B1, increases of fibrinogen concentration in the solution
induced large R1 and fG1/2 responses in both cases, but there
were some differences between the adsorption of fibrinogen on
TiO2 and HAP. Obvious increases in R1 and fG1/2 could be
observed in the case of HAP modified electrode after the fibrinogen concentration increased to 0.990 g/L, while the increases of
protein concentration caused relatively small changes in R1
and fG1/2 after the saturated adsorption of fibrinogen in the
case of TiO2 modified electrode. Change in R1 reflects some
changes in the surrounding environment of the PQC resonance.
The increases in R1 can, in principle, be understood from the
increases in solution density and viscosity as well as the viscosity of the protein layers adsorbed onto the electrode surface.
Therefore, the sudden increase of R1 may result predominantly
from the formation of the surface fibrinogen layer after fibrinogen adsorption, or it may result from the viscosity effect of the
fibrinogen surface layer.
Sauerbrey equation [24] described a frequencymass relationship for the load or removal of a rigid and thin film, which
can be used for the solution system when the influence of the
visco-elasticity is negligible [25,26]. A net liquid-loading effect
for a PQC with one side contacting solution can be characterized
by the following equation [27]:
2L1q fG1/2L R1L
= 2fL1L = 4L1q f0L
4L1q f0L

Fig. 2. Cyclic voltammograms (A) of bare gold electrode (solid line), TiO2
electrode (dashed line) and HAP electrode (dotted line), scan rate: 50 mV/s. And
electrochemical impedance spectroscopes for bare gold (), TiO2 modified ()
and HAP modified () electrodes before (B) and after (C) fibrinogen adsorption,
10 mHz 100 kHz, 5 mV r.m.s., 0.17 V vs. SCE. All experiments were carried
out in 2 mM K3 Fe(CN)6 /K4 Fe(CN)6 PBS solution (50 mM, pH 7.4).

61

fq /

f0g c 66
(1)

where fG1/2L , f0L , R1L and L1L are changes in fG1/2

(the half peak width of the conductance curve), f0L (the resonant frequency, and f0 = 1/[2(L1 C1 )1/2 ]), R1L and L1L due to
the variations of the solution density and viscosity, respectively.
f0g is the resonant frequency in air, L1q is the motional inductance for the PQC in air, and c 66 (2.9571010 N m2 ) is the lossy
piezoelectrically stiffened quartz elastic constant. According to

62

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

Fig. 3. Time course of simultaneous responses of f0 and R1 during adsorption of fibrinogen onto TiO2 (A1) and HAP (B1) modified gold electrodes. Arrows
indicate additions of fibrinogen solutions (20 L, 100 g/L) to a final fibrinogen concentration of 0.199, 0.398, 0.596, 0.794, 0.990, 1.186, 1.381, 1.575 and 1.768 g/L,
respectively. And time course of simultaneous responses of f0 , R1 , C0 , fG1/2 and Cs during adsorption of fibrinogen onto TiO2 (A2) and HAP (B2) modified
gold electrodes. All experiments were carried in 50 mM PBS, pH 7.4, 10 Hz, 5 mV r.m.s., 0.1 V vs. SCE.

this equation, the slope of f0 /R1 for a net density/viscosity

effect on the 9 MHz PQC resonance is about 10 Hz/. Obviously, for an investigated system, the larger the absolute value
of f0 /R1 is, the weaker the viscous effect and the stronger the
mass effect will be. In this experiment, the values of f0 /R1 for
fibrinogen adsorption on TiO2 and HAP surface were 75.3 and
30.9 Hz/, respectively. Therefore, for fibrinogen adsorption
on TiO2 modified electrode, the effect of the density/viscosity
could be neglected, but for the HAP modified electrode, the
effect cannot be neglected.
It was found that the protein adsorption process could be
described as two consecutive steps occurring at the interface, and
the time-dependent responses of f0 and Cs during the protein
adsorption could be well simulated by the following experiential
equation [28]:
Y(t) = a0 + a1 e(t/t1 ) + a2 e(t/t2 )

(2)

where Y(t) can be either frequency response or interface capacity

response, a0 , a1 , a2 , 1 and 2 are estimated parameters, and t
is the adsorption time. The parameter a0 is the total frequency
or interfacial capacitance shift at t . These two exponential
functions suggest that two different kinetic steps were incorporated into the process of protein adsorption on the electrode
surface. The two steps are the adsorption of protein molecules

and the rearrangement of the adsorbed protein molecules on

the electrode surface, respectively. We fitted the f0 and Cs
responses to Eq. (2) using a non-linear fitting program embedded in Sigmaplot Scientific Graphing Software Version 2.0 and
the relative sum of the residual squares is defined as
N
(rfit rexp )2
q
qr = N
= 1 N
(3)
2
2
1 rexp
1 rexp
where rfit and rexp denote the fitted and experimentally obtained
response values of frequency or interfacial capacitance, respectively, and N is the number of the response signal points. The
fitted results are listed in Table 1 and plotted as circles in Fig. 3
which indicate that the adsorption of fibrinogen molecules onto
TiO2 and HAP surface fit the two consecutive reactions kinetic
model well. In other words, the adsorption of fibrinogen could be
divided into two consecutive reaction steps as described above.
As mentioned above, a0 is the total frequency or interfacial
capacitance response value at t . Therefore, it could be
inferred from Table 1 that the amounts of proteins adsorbed
onto these surfaces were different and the time periods reach
to adsorption equilibriums were also different. Compared to
the adsorption of fibrinogen onto TiO2 , much larger values of
f0,max (a0 ) were obtained in the case of fibrinogen adsorption onto the HAP modified surfaces. Fibrinogen adsorption

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

63

Table 1
Parameters obtained by fitting responses of f0 and amide II (1545 cm1 ) FTIRATR absorbance given in Figs. 3 and 5 to Eq. (2)
Adsorbent surface

Methods

Responses

a0

a1

a2

1 (s)

2 (s)

qr (104 )

On TiO2

ATR
EQCI

On HAP

ATR
EQCI

Absorbance
f0 (Hz)
Cs (F)
Absorbance
f0 (Hz)
Cs (F)

0.1242
1136.86
0.4791
0.1497
5663.0
0.022

0.0436
561.17
0.2628
0.0900
2455.10
0.0156

0.0806
500.20
0.2475
0.0599
2871.10
0.0060

36.566
103.62
223.61
32.378
67.79
22.01

630.95
1660.06
2121.46
1451.70
2243.37
1731.49

0.102
2.43
0.54
0.105
5.60
0.654

onto TiO2 and HAP surfaces resulted in about 15 and 183 

changes in motional resistance, respectively. Subtracted the f0
changes caused by viscosity effect, 150 and 1830 Hz, from
the total f0 changes (shown in Table 1), then the f0 changes
caused by mass effect were 987 and 3833 Hz, corresponding to the surface fibrinogen mass coverage of 1.67 106 and
6.49 106 g/cm2 , respectively. Usually, great surface area will
lead more adsorption. From the measurement of surface area, it
is know that the surface area of TiO2 and HAP were 87.33 and
18.98 m2 /g, respectively. Considering the surface area of TiO2
(87.33 m2 /g) is much great than that of HAP (18.98 m2 /g), the
above fact suggests that the surface area of the particles should
not be the main factor which affects fibrinogen adsorption onto
TiO2 and HAP surface and some unknown factors predominate
the adsorption of fibrinogen on TiO2 and HAP surface. We will
discuss them in Section 3.3.2.
C0 is related to the interfacial double layer capacitance of
the charged surface [16]. However, the present results prove
that C0 is not a direct measure of the double layer capacitance and may be determined by other factors that are unclear
at present. In our experiment, the C0 increased in all cases, and
that Cs decreased in the case of fibrinogen adsorption onto TiO2
surface, but increased in the case of HAP. The Cs data represents the interfacial capacitance according to a simplified Rs Cs
equivalent circuit of EIS in the absence of electroactive species
[25]. The changes of Cs may be interpreted as the effect of the
hydrophilic properties of the TiO2 and HAP films. It is generally accepted that important interactions between proteins and
biomaterials are hydrophobic/hydrophilic interactions and electrostatic interactions [29]. In pH 7.4 solutions, the C domains
of fibrinogen were actually positively charged. After adsorption, the positively charged C domains of fibrinogen interacted
with the negatively charged surface (TiO2 , Ip 5), the hydrophobic end (D and E domains) of fibrinogen will cover the surface
of the electrode, resulting in the decrease in hydrophilicity and
subsequent reduction in ion transfer. Therefore, the interfacial
capacitance decreased after fibrinogen adsorption onto the TiO2
film. For the case of HAP, increase in Cs was observed during fibrinogen adsorption. We suppose that another interaction
between protein and HAP was the main contributing factor of the
adsorption, which resulted in increase in the hydrophilicity of
the electrode surface. We are going to discuss it in the following
paragraph.
As mentioned above, fibrinogen bound to TiO2 mainly come
from the electrostatic interactions between positive charges C
domains of fibrinogen and TiO2 negative charges [30,31]. In

the case of HAP, the protein binding includes both nonspecific

attraction between protein positive charges and HAP, and specific complexing of protein carboxyls with the calcium ions
of HAP [32]. The protein carboxyls of hydrophobic D and E
domains also interact with calcium ions of HAP, thus hydrophilic
region was exposed on HAP surface. An increase in Cs was
observed.
3.3. FTIRATR spectroscopy studies
3.3.1. Studying of the process of brinogen adsorption
Fig. 4 shows the FTIR spectrum of fibrinogen in solution
(20 g/L in PBS, pH = 7.4) in the spectral region between 1200

Fig. 4. IR spectra of fibrinogen adsorbed on the bare (A) and modified (TiO2
coated, B; HAP coated, C) ZnSe prisms for 60 min from PBS solution of 20 g/L
fibrinogen, pH 7.4, containing 50 mM PBS. The background spectrum is from
a solution of 50 mM PBS.

64

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

drawn from FTIRATR tests. Furthermore, the values of 1 and

2 corresponding to TiO2 were different with that corresponding
to HAP. The results indicate that the values of 1 and 2 were
related to the nature of the electrode surfaces as well.

Fig. 5. Band intensity changes during fibrinogen adsorption on bare ZnSe prism
(A), on TiO2 coated ZnSe prism (B) and on HAP coated ZnSe prism (C).

and 4000 cm1 . Proteins exhibit characteristic bands in this

spectral range, which emerged from vibrations in the peptide
linkages. There were two dominant bands, namely amide I band
around 1650 cm1 and amide II band around 1550 cm1 . The
other bands at 1460 and 1400 cm1 were assigned to the CH2
deformation and to the vibrations of the amino acid side chains,
respectively.
Curve of absorbance band intensity against adsorption time
is shown in Fig. 5. The plot shows that the absorbance increased
slightly at the first 1500s, and then the absorbance reached to
the equilibrium state when the TiO2 was exposed to the protein solution. While fibrinogen interacted with the HAP surface,
the absorbance increased sharply during the initial 250s, then the
increasing trend slowed down in the following period from 1250
to 1500s, the absorption curve reached a plateau. The change in
IR absorbance intensity corresponding to amide II (1545 cm1 )
shown in Fig. 5 indicates that the FTIRATR experiment results
are consistent with PQCI experiments results. The experimental
data were fitted using nonlinear fitting program with Eq. (2) and
the results are also listed in Table 1. Comparing with EQCI
experiments, the values of 1 and 2 were obviously small.
Because the FTIRATR tests were carried out in the systems
without stirring while EQCI tests were carried out in the systems
with stirring, we deduced that the differences in time constants
may be due to the shearing effects of stirring. The immobilization of protein molecules onto solid surface became difficult
when the solution was being intensively stirred. Therefore, the
values of 1 and 2 drawn from EQCI tests were great than that

3.3.2. Studying of secondary structure of adsorbed

brinogen
A curve fitting procedure can be applied to estimate quantitatively the area of each component representing a type of
secondary. Curve or peak fitting is based on the original algorithm of non-linear peak fitting. The fitting procedure was carried
out always in the same way in order to allow such a comparison for protein before and after adsorption on the solid
surface. Before starting fitting, initial estimates of band widths
and band heights were given. The amide I band was deconvoluted with a Gaussian non-line function and the full width
at half height (FWHH) of the widest resolvable peak. Band
positions used as input parameters were determined from second derivative FTIR spectra of fibrinogen in PBS solution and
adsorbed on TiO2 and HAP surface. The percentage of each
secondary structure was calculated from the integrating areas of
the component bands in amide I. The assignments of component
bands in amide I of proteins [19,33] were 16101640 cm1 to
-sheet, 16401650 cm1 to random coil, 16501658 cm1 to
-helix and 16601700 cm1 to -turn structure, respectively.
In addition, bands around 1615 and 1618 cm1 are assigned to
side chains and could be evaluated in solution and on modified
surfaces [34], respectively. The percentages of the secondary
structure calculated from the band areas, the relative error and
correlation coefficients are shown in Table 2, and the representative results for curve fitting are given in Fig. 6. Compared with
the protein in solution, the conformation of fibrinogen molecules
changed when it was adsorbed on both TiO2 and HAP surfaces.
The -helix content of adsorbed fibrinogen obviously decreased
and was mainly transformed to -sheet, while the -turn and random coil contents were less changed. Furthermore, the results
listed in Table 2 indicate that fibrinogen adsorption on HAP surface led more significant change in the secondary structures of
the protein molecule than that on TiO2 surface. These results
can be explained as follows. Generally, the TiO2 particles are
non-charged at around pH5 [2,31]. Under the adopted experimental condition, the predominant TiO2 surface groups are
Ti2 O , TiOH, and in the second order Ti2 OH, while
the main protein functional groups are RCOO and RNH3 +
[35,36]. These groups on the electrode surface could interact
with the ionized residues in the protein molecules through electrostatic interaction, TiOH2 + :NH2 R, Ti2 O+ NH3 R, and
hydrogen bonding interaction, TiOH COOR. Thus, we
think that the electrostatic interactions and hydrogen bonding

Table 2
Secondary structures (%) of fibrinogen in bulk solution and adsorbed on the TiO2 and HAP coated surfaces after 60 min
Fibrinogen

-Helix

-Turn

-Sheet

Random

Side chain

Correlation coefficient

In bulk solution
On TiO2
On HAP

38.7 0.14
36.9 0.72
35.4 0.37

25.4 0.11
24.8 0.81
22.0 0.56

23.8 0.43
25.6 0.73
26.5 0.68

9.4 0.49
10.7 0.46
11.3 0.25

2.52 0.12
1.74 0.23
3.19 0.15

0.9997
0.9994
0.9998

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

65

Fig. 6. Amide I (16001700 cm1 ) bands and fitted curves of free fibrinogen (A), fibrinogen on TiO2 (B) and fibrinogen on HAP (C) in the 50 mM PBS when
fibrinogen was adsorbed 60 min, pH 7.4. The background spectrum is from a solution of 50 mM PBS.

interaction leaded to the changes in secondary structure of the

protein molecules. For the case of HAP, it has been found that the
HAP particles are non-charged when the pH value lies between
4.3 and 7.5, depending on the experimental techniques employed
[37,38]. In pH 7.4 PBS buffer, the surfaces of HAP particles
are net negatively charged, but there are still positive (calcium)
and negative (phosphate) charged sites on the particle surface.
Therefore, there are two main factors to affect the change in
the secondary structure of protein molecules, which are the
electrostatic interactions between positive charges of the protein surface and biomaterials, and the complexing interaction
between protein carboxyls and the calcium ions of HAP, i.e.,
HAPPO4 2+ NH3 R and HAPCa2+ OOCR. It might be reasonable that more active sites which responsible to fibrinogen
adsorption exist on HAP surface and the interaction between
HAP with fibrinogen molecule is stronger than that in the case
of TiO2 under the test condition. [39,40]. Therefore, we could
conclude that more active sites on HAP surface and stronger
interactions between fibrinogen and HAP resulted in more conformational changes in adsorbed fibrinogen molecules than that
in the case of TiO2 . Following the above explanations, the experimental fact that more fibrinogen was adsorbed (larger mass
coverage of fibrinogen) on HAP than on TiO2 were reasonable,
although the specific surface area of HAP was quite small than
that of TiO2.
4. Conclusions
In summary, through simultaneous measurement of quartz
crystal impedance and electrochemical impedance related equivalent circuit analysis, we have been able to simultaneously
obtain the equivalent circuit parameters, resonant frequencies
and the interfacial capacitance, which reflected the physical and
chemical properties between the interfaces of the solid and the
liquid. According to obtained parameters, the adsorption kinetics and mechanism were analyzed. FTIRATR spectroscopy
as a useful tool has been used to study the secondary structure conformations of protein molecules. Experiments proven
that HAP surface bears more bioactivity as comparing with the
TiO2 surface and the piezoelectric impedance and electrochemical impedance spectroscopy are sensitive, simple and reliable
methods to study the interactions of protein and biomaterials. While the EQCI techniques are combined with FTIRATR
spectroscopy, we can obtain more information of the structure

changes. The method will play an important role in the biochemical field in the future.
Acknowledgments
This work was supported by the National Natural Science
Foundation of China (20675030, 20675029, 20335020), the
Basic Research Special Program of the Ministry of Science and
Technology of China (2003CCC00700), and Scientific Research
Fund of Hunan Provincial Education Department (06A035).
References
[1] J.L. Brash, T.A. Horbett (Eds.), ACS Symposium Series No. 343, American
Chemical Society, Washington, DC, 1986.
[2] T.A. Horbett, J.L. Brash (Eds.), ACS Symposium Series No. 602, American
Chemical Society, Washington, DC, 1995.
[3] Y. Lin, J. Wang, L.J. Wan, X.H. Fang, Ultramicroscopy 105 (2005)
129136.
[4] Y.L. Chen, X.F. Zhang, Y.D. Gong, N.M. Zhao, T.Y. Zeng, X.Q. Song, J.
Colloid Interface Sci. 214 (1999) 3845.
[5] T. Endo, S. Yamamura, N. Nagatani, Y. Morita, Y. Takamura, E. Tamiya,
Sci. Technol. Adv.; T. Endo, S. Yamamura, N. Nagatani, Y. Morita, Y.
Takamura, E. Tamiya, Sci. Technol. Adv. Mater. 6 (2005) 491500.
[6] M. Ombelli, R.J. Composto, Q.C. Meng, D.M. Eckmann, J. Chromatogr.
B 826 (2005) 198205.
[7] X.F. Zhang, Y. Ma, H. Liu, F.G. de, Sa. Paula, P.R. Brown, J.A. Dain, Anal.
Biochem. 325 (2004) 255259.
[8] X.L. Mao, L.J. Yang, X.L. Su, Y.B. Li, Biosens. Bioelectron. 7 (2006)
11781185.
[9] X.L. Su, Y.B. Li, Biosens. Bioelectron. 21 (2005) 840848.
[10] X.D. Su, Y.J. Wu, W. Knoll, Biosens. Bioelectron. 21 (2005) 719726.
[11] Q.J. Xie, Y.Y. Zhang, Y. Yuan, Y. Guo, X. Wang, S.Z. Yao, J. Electroanal.
Chem. 484 (2000) 4154.
[12] M. Thompson, A.L. Kipling, W.C. Duncan-Hewitt, L.V. Rajakovic, B.A.
Cabic-Vlasak, Analyst 116 (1991) 881890.
[13] H.L. Bandey, M. Gonsalves, A.R. Hillman, A. Glidle, S. Bruckenstein, J.
Electroanal. Chem. 410 (1996) 219227.
[14] R.J. Pei, Z.L. Cheng, E.K. Wang, X.R. Yang, Biosens. Bioelectron. 16
(2001) 55361.
[15] Y.X. Hou, S. Helali, A.D. Zhang, N. Jaffrezic-Renault, C. Martelet, J.
Minic, T. Gorojankina, M.A. Persuy, E. Pajot-Augy, R. Salesse, Biosens.
Bioelectron. 21 (2006) 13931402.
[16] Y.Y. Zhang, Y.S. Fun, H. Sun, D.R. Zhu, S.Z. Yao, Sens. Actuators B 10
(8) (2005) 933942.
[17] Y.M. Long, L.H. Nie, J.H. Chen, S.Z. Yao, J. Colloid Interface Sci. 263
(2003) 106112.
[18] S.E. Moulton, J.N. Barisci, A.J. McQuillan, G.G. Wallace, Colloids Surf.
A 220 (2003) 159167.

66

Q. Yang et al. / Analytica Chimica Acta 597 (2007) 5866

[19] Y. Liu, M.X. Xie, M. Jiang, Y.D. Wang, Spectrochim. Acta Part A 61 (2005)
22452251.
[20] P. Parhi, A. Ramanan, A.R. Ray, Mater. Lett. 58 (2004) 36103612.
[21] H.T. Zeng, K.K. Chittur, W.R. Lacefield, Biomaterials 20 (1999) 377
384.
[22] F. Dousseau, M. Therrien, M. Pezolet, Appl. Spectrosc. 43 (1989) 538
542.
[23] S. Raynaud, E. Champion, D. Bernache-Assollant, P. Thomas, Biomaterials
23 (2002) 10651072.
[24] G.Z. Sauerbrey, Z. Phys. 155 (1959) 206222.
[25] D.A. Buttry, A.J. Bard (Eds.), Electroanalytical Chemistry, 17, Marcel
Dekker, New York, 1991, pp. 185.
[26] T. Sato, S.T. erizawa, Y.B. Okahata, Biochim. Biophys. Acta 1285 (1996)
1420.
[27] H. Muramatsu, E. Tamiya, I. Karube, Anal. Chem. 60 (1988) 21422146.
[28] P. Bernabeu, L. Tamisier, A. De Cesare, A. Caparani, Electrochim. Acta
33 (1988) 11291136.
[29] W. Norde, Adv. Colloid Interface Sci. 25 (1986) 267340.

[30] A.D. Roddick-Lanzilotta, P.A. Conner, A.J. McQuillan, Langmuir 14

(1998) 64796484.
[31] K.D. Kobson, P.A. Conner, A.J. McQuillan, Langmuir 13 (1997)
26142616.
[32] M.J. Gorbunoff, S.N. Timasheff, Anal. Biochem. 136 (1984) 440445.
[33] S. Tunc, M.F. Maitz, G. Steiner, L. Vazquez, M.T. Pham, R. Salzer, Colloids Surf. B 42 (2005) 219225.
[34] Y.B. Yan, Q. Wang, H.W. He, H.M. Zhou, Biophys. J. 86 (2004) 16821690.
[35] F.Y. Oliva, L.B. Avalle, O.R. Camara, C.P.D. Pauli, Part I: J. Colloid Interface Sci. 261 (2003) 299311.
[36] S. Servagent-Noinville, M. Revault, J. Colloid Interface Sci. 221 (2000)
273283.
[37] I.S. Harding, N. Rashid, K.A. Hing, Biomaterials 26 (2005) 68186826.
[38] D.P. Jiang, G.S. Premachandra, C.J. Johnston, S.L. Hem, Phosphate adjuvant, Vaccine 23 (2004) 693698.
[39] G. Bernardi, M.-G. Giro, C. Gaillard, Biochim. Biophys. Acta 278 (1972)
409420.
[40] G. Marguerie, Biophys. Acta 494 (1977) 172181.