Vous êtes sur la page 1sur 28

MOLECULAR IMPRINTING

1. Introduction
Molecularly imprinted polymers (MIPs) are materials that have widespread use
for applications in the biomedical, analytical, chemical, and biological sciences
(13). The concept of molecular imprinting is related to the bioinspired principle
of a lock-and-key fit for enzymes with target ligand molecules. The lock-and-key
theory was put forth by Emil Fischer in 1894 to explain the specific interactions
of enzymes with a target molecule. In this analogy, the enzyme is regarded as the
lock that binds the target molecule represented as the key (Fig. 1). The process of molecular imprinting starts with a target molecule (ie, the key) and
builds an artificial receptor (ie, the lock) around the template, as shown in
Figure 2. The template molecule serves two purposes; first, the template provides
a space-filling three-dimensional object around which a complementary mold or
cavity can be formed. Second, the functional groups on the template can organize
interactive molecules (ie, polymerizable monomers) in a complementary specific
arrangement. The organization of polymerizable functional monomers by the
template is achieved either through covalent bonds and/or noncovalent forces
such as hydrogen bonding and electrostatic and hydrophobic interactions. In
practice, the polymerization reaction mixture consists of the template along
with functional monomers and a large excess of cross-linking monomer. An
equal volume of inert solvent (porogen) and free-radical initiator make up the
remainder of the polymerization solution. Thermal- or photo-initiated polymerization results in a highly cross-linked insoluble network polymer with the template still inside. The template is reversibly removed from the polymer network,
whereas the functional monomers remain covalently bound to the polymer itself.
Ideally, left in the polymer matrix are three-dimensional cavities that are complementary in shape to the template with desired functionality in a specific complementary arrangement.
Because of the level of molecular control afforded by MIPs, these materials
hold tremendous promise for applications such as:
1.
2.
3.
4.
5.
6.

Chemical and biological separations (4,5)


Immunoassays (6,7)
Catalysts (8)
Sensors (9,10)
Nanotechnology (11)
Biomedical (eg, drug delivery) (1214)

MIPs outperform nonspecific polymers such as silica gel, carbohydrates, or


other polymer materials for all applications that require molecular recognition.
On the other hand, biological molecules exhibit good specificity toward a large
range of molecules. However, MIP materials have superior properties over biological receptors such as the following:
1. Stability for organic solvents, aqueous solvents, and high temperatures (15)
2. Facile, low cost, and rapid preparation in any volume
1

Kirk-Othmer Encyclopedia of Chemical Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

MOLECULAR IMPRINTING

Fig. 1. An illustration of the lock-and-key fit first proposed by Emil Fischer.

3. Binding affinities that can be tailored toward any desired functional group
(including those not available to biological molecules)
4. Shelf-life of several years up to decades without loss of performance, unlike
biological molecules
Furthermore, biological molecules are generally expensive, often use animal
sources, and cannot be stored for long periods. Small molecule receptors can be a
rugged alternative, however, the synthesis of these compounds is often long and
arduous, whereas molecular imprinting can be carried out in one day. Thus,
MIPs are an important class of materials for current technical demands, and
improvements in MIP methodology are continuing to progress, including efforts
toward new formats (16), new applications (17), and new materials (18,19).

2. The Origins of Molecular Recognition in MIPs


The origins of the memory effect in MIPs for specific rebinding of the template
molecule have been attributed to the following factors:

Fig. 2. A schematic of the molecular imprinting method. A template molecule is used to


organize the functional monomers prior to their polymerization into a macromolecular
network polymer with cross-linking monomers.

MOLECULAR IMPRINTING

1. Preorganization of complementary functional groups in the polymer by the


template
2. Formation of a shape-selective cavity that is complementary to the template
As a result, the imprinting effect essentially can be considered a threedimensional effect (ie, it is the effective control of three-dimensional interactions
by the template with surrounding functional monomers and the cross-linked
matrix). Recent studies have shown that the contributions of preorganization
and shape selectivity to the molecular recognition observed depend on the
number and the types of interactions (20). For templates that can interact
with one or two functional monomers, enantioselectivity was determined to
result primarily from shape-selective interactions of the template with the
matrix of the binding cavity. However, greater numbers of functional group interactions tend to provide a greater influence on the selectivity properties of MIPs.

3. Evaluation of Binding and Selectivity


3.1. Batch Rebinding. Two methods generally are employed for the
selective binding of targets, batch rebinding, and chromatography. Batch rebinding will be discussed first, which in practice, analyzes a solution of guest substrate S in the presence of the solid polymer MIP host (Fig. 3). In this case, a
known amount of particulated MIP is added to a solution of substrate S, and
the amount of substrate remaining in solution after adsorption to the MIP is
measured and referred to as Cf (the concentration of free substrate). The difference of Cf from the total substrate added (Ct) gives the amount of substrate
adsorbed to the MIP (Cb). Following classical adsorption models, the equation
for the partition coefficient can be written as follows:
K P Cb =Cf

Fig. 3. An illustration of a batch rebinding experiment.

MOLECULAR IMPRINTING

Selectivity by any host molecule or polymer results from the differences of the
adsorption of one substrate vs another. Thus, the ratio of the partition coefficients determines the selectivity of the MIP host between two substrates, which
is referred to as the separation factor (a) or sometimes as the selectivity factor
(eq. 2).
a K P2 =K P1

If alpha is determined for a pair of enantiomers, then equation 2 can be


used directly. However, when comparing the imprinting effects of nonenantiomeric compounds, it is best to use a modified partition coefficient I, which is
referred to as the imprinting factor (21). The imprinting factor is obtained
from the ratio of the partition coefficient of a substrate on an imprinted polymer,
Kmip, and the partition coefficient Knon-mip, using the same substrate on a nonimprinted polymer with the same monomer formulation.
I K mip =K non-mip

The imprinting factor represents how many times better the substrate binds to
the imprinted polymer vs a nonimprinted (generic) polymer. In this manner,
binding resulting from nonspecific interactions are numerically removed, leaving
a value for the binding solely resulting from the imprinting effect, namely the
three-dimensional organization of monomers in the MIP. Taking the ratio of imprinting factors for two different substrates, I1 for substrate one and I2 for substrate two, we obtain the specific selectivity factor S (21) as follows:
S I 1 =I 2

For enantiomeric compounds, the value for Knon-mip is the same for both substrates; thus, the specific selectivity factor simply reduces to a.
3.2. Chromatography. As solids, MIPs are predisposed for use as chromatographic stationary phases. Once the polymers are obtained in the desired
particle size (eg, by sizing and sieving) columns can be made using traditional
chromatographic packing techniques (Fig. 4). Equations for the same analysis
using chromatographic parameters are proportional to the equations for
batch rebinding (22). The fundamental equation that relates batch rebinding

Fig. 4. Once MIPs are sized, they can be packed into high performance liquid chromatography columns.

MOLECULAR IMPRINTING

to chromatography and vice versa is as follows:


0

K K =f

where f is the phase volume ratio (volume of stationary phase/volume of mobile


phase) and k0 is the capacity factor that is found using the following equation:
0

k RV  DV =DV

where RV = retention volume of the sample (retention time  flow rate) and
DV = retention volume of an unretained sample (eg, acetone).
The relationship between K and k0 holds if both batch rebinding and chromatographic methods are under equilibrium conditions, allowing the equations
derived for the batch rebinding method to be converted to the following equations
for chromatographically derived data:
0

a k2 =k1
0

I kmip =knon-mip

S I 1 =I 2

It should be noted that MIP binding sites have a distribution in the quality of
binding sites. Because the overall performance of the MIP will be a combined
average of the ensemble of sites, the values found by these equations are most
reproducible and comparable if the same conditions and quantities are used for
each analysis. A more detailed treatment of the distribution of binding sites can
provide greater insight into the effects of the ensemble of sites on binding affinity
and selectivity (2328).

4. Development of MIP Materials


Early studies in molecular imprinting employed silica as both the matrix and the
functional groups interactions with the template (1). This idea evolved into a new
era of modern molecular imprinting in which the use of highly cross-linked
network polymers was introduced as the matrix. In this case, the polymer matrix
remained inert toward the template molecule and acted to immobilize additional
interactive monomers in their assembled positions. An example of this is shown
in Figure 5 in which methacrylic acid (MAA) (1) is complexed prior to polymerization to a derivative of adenine (2)one of the deoxyribonucleic acid (DNA)
bases (29). This approach of using an inert matrix was designed to lower nonspecific interactions with the matrix and to promote interactions within the
imprinted cavities. The interactive functional monomer, in this case MAA, has
a single polymerizable group (eg, the double bond shown in the structure for

MOLECULAR IMPRINTING

Fig. 5. An example of molecular imprinting 9-ethyladenine into a cross-linked network


polymer.

MAA) that does not cross-link the material. The primary role of the functional
monomer is to provide templatepolymer interactions, referred to (in solution)
as the prepolymer complex (see Fig. 5), which provides rebinding interactions
for the template molecule in the final polymer. The prepolymer complex is
locked into place by copolymerization with cross-linking monomers. Most polymeric MIPs use cross-linkers incorporating two or more polymerizable groups,
which form the requisite rigid network polymer to immobilize the template-organized functional monomers in their positions. Although this approach has been
the most widely reported for forming molecular-specific polymers, recent
advances have shown that incorporation of the interactive functional groups
into a cross-linking format further can improve the molecular recognition in
these materials (vide supra).
4.1. The Matrix. Traditional methods in molecular imprinting employ a
class of highly cross-linked network polymers known as macroporous (or macroreticular) polymers (30), and a diagram of the hierarchy of structure is shown in
Figure 6. The structure consists of interconnected microspheres (globules) that
are partly aggregated in larger clusters that form the bulk polymer. The size
of the spherical globules range from 10 nm to 30 nm, which create pores (macropores >50 nm in diameter) between the globules of a given cluster or even within
the interstitial space of the globules themselves (mesopores 250 nm in
diameter). The globules, in turn, are made up of nuclei that aggregate to form
micropores in the polymer (<2 nm in diameter). The surface area and the pore
size distribution reflect the internal organization of both the globules and their
clusters within the macroporous polymer and largely depend on the composition
of the polymerization mixture and the reaction conditions. The most effective

MOLECULAR IMPRINTING

Fig. 6. A depiction of the structure of a macroporous polymer produced in the presence of


a nonsolvating diluent. The first family of pores contribute significantly to the surface area
for these materials (interstitial voids between nucleii). The high pore volume of this polymer comes from the voids between the aggregates (third family of pores).

variables that control surface area and pore distribution are the percentage of
cross-linking monomer, the type and amount of porogen, the concentration of
free-radical initiator, and the reaction temperature. Typical values for the surface area of the imprinted polymers are in the range of 100 m2/g to 400 m2/g,
and the pore size distribution ranges from 7 nm to 20 nm in addition to micropores of 0.6 nm to 2 nm in diameter.
The molecular structures of several cross-linking monomers are shown in
Figure 7. The first cross-linking monomer to be employed for molecular imprinting in organic polymers was divinylbenzene (DVB) (31), which is still a useful
cross-linker today. In a comparison of ethyleneglycol dimethacrylate (EGDMA)
and butanediol dimethacrylate (BDMA) to DVB, it was determined that
EGDMA provided the best selectivity in MIPs (32), and thus, EGDMA has
been the cross-linker of choice for most MIP publications. The difference between
glycol cross-linkers EGDMA and BDMA are the two additional carbon groups in
BDMA, which allow for the loss of rigidity and result in a loss of structural integrity of the imprinted cavity. A relatively new cross-linker that has shown
sporadic improvement over EGDMA is the trifunctional monomer 2-bis-(hydroxymethyl) butanol trimethacrylate, which is commercially available (33). This
also was true of the similar cross-linker pentaerythritol triacrylate but not of

Fig. 7. Monomers commonly used for cross-linking in MIPs.

MOLECULAR IMPRINTING

the tetrafunctional cross-linker pentaerythritol tetraacrylate (34). Several bis(acrylamide) and bis-(methacrylamide) cross-linkers have been synthesized and
have been successful as imprinting matrices; however, these are not soluble in
nonpolar solvents that are needed to optimize the imprinting process (35,36).

4.2. Other Matrix Materials: Silica, Hydrogels, and Conjugated


Polymers. Imprinting in silica matrices has continued at a moderate pace,
and this subject has been reviewed recently (37). One advantage of using silica
(or other inorganic materials) for molecular imprinting is that these materials
exhibit little or no swelling in organic solvents, thus maintaining the imprinted
binding site geometries in various solvents. Other advantages include resistance
to heat toward both binding site deformation and long-term stability of the
matrix, stability under high pressures, and the wealth of fabrication methods
generally developed for silicate materials. Imprinted silica can be carried out
in bulk (38), in particles (39), or following various methods of surface imprinting
(40,41). Imprinting in conjugated materials offers the opportunity of a diagnostic
response to binding molecules (42,43). A milestone article in this field employed a
polypyrrole matrix, which was imprinted with metal cations (44); since then,
there has been further development of MIPs in electropolymerized polypyrrole,
polyphenol, and poly(phenylenediamines) (45,46).
Molecularly imprinted hydrogels recently have been reviewed (47,48) and
often are used for large biomolecules (eg, proteins) in which low cross-linking
is needed for mass-transfer of the templates in and out of the imprinted material.
For example, acrylamide hydrogels with low levels of methylene-bis-acrylamide
cross-linker have been imprinted with protein templates (49). Many other types
of hydrogel formulations have been employed for molecular imprinting other
macromolecules and small molecules as well (5052). Metal-imprinted hydrogels
have been reported in materials such as alginate (53) or in interpenetrating networks (54). Similar types of hydrogels have been used for biomolecular imprinting, which uses bioreceptors as the functional group for forming the templatefunctional monomer complex (55). More highly cross-linked hydrogels also
have been used for molecular imprinting (56), and drug delivery of template
molecules has been controlled by adjusting the cross-linker content.

5. Functional Monomers
In 1906, Paul Ehrlich stated Corpora non agunt nisi fixata, which translates to
Molecules do not act if they do not bind (57). This concept is the basis and the
first step of any molecular recognition event, including molecular imprinting.
MIPs first require binding in the solution phase to create binding sites via a prepolymer complex (Fig. 8) and for molecular recognition of the target species

Fig. 8. The relationship between solution phase binding and specific binding sites made
in the MIP.

MOLECULAR IMPRINTING

Fig. 9. Comparison of approaches for noncovalent vs covalent imprinting incorporating methacrylic acid as the functional monomer and a-methylbenzylamine as
the template.

during rebinding analyses. At least one strong bond is needed for the formation
of the prepolymer complex, and it is proposed that each complex in the prepolymer solution gives rise to each binding site (Fig. 8). By applying Le Chatliers
principle to the complex formed prior to polymerization, increasing the concentration of components or the binding affinity of the complex in the prepolymerization mixture will drive the functional monomer and template toward more
prepolymerization complex. This will increase the concentration of prepolymer
complex and is postulated to increase the number of final binding sites in the
imprinted polymer.
The nature of the functional monomertemplate interactions can be covalent or noncovalent (Fig. 9). Initial studies by Wulff used covalent interactions
to restrict the incorporation of interactive functional groups only to the binding
cavities, eliminating chances of nonspecific binding outside the binding cavity.
Early examples from this group used covalent interactions between vinylboronic
acid and carbohydrates such as n-mannopyranoside for imprinting and for the
recognition of carbohydrate derivatives via covalent rebinding (58). Carboxylate
esters have been used in several imprinted systems; however, these covalent
complexes have proven difficult to hydrolyze when incorporated into the highly
cross-linked imprinted materials. Covalent complexes that are cleaved more
easily include acetal/ketal and siloxane derivatives (58).
An interesting example makes use of thermoreversible carbamate bonds,
which leaves the isocyanate moiety that can be transformed to an amine group
in the presence of water (59,60). In this example, dithiol bonds between the template and polymerizable group were cleaved to leave free thiol groups. The free
thiol can be used as such or else oxidized to sulfonic acid for cation recognition. It
should be noted that covalent prepolymer complexes usually are not found to
generate stronger binding sites than noncovalently imprinted polymers. This

10

MOLECULAR IMPRINTING

may be a result of stronger affinity sites developing from higher order complexes
that are possible using noncovalent imprinting methods but are not possible for
covalent methods. However, as pointed out earlier, it has shown that covalently
imprinted polymers, however, do provide a narrower distribution of binding sites
(24).
The use of noncovalent interactions in MIPs was introduced in 1984 by
Andersson et al, who imprinted the L-phenylalanine ethyl ester using acrylic
acid or para vinylbenzoic acid monomers to form ionic complexes via the free
amine group of the template (61). Because of the ease and success of this
approach, most subsequent reports of molecular imprinting have been based
on noncovalent bonds, as shown in the example in Figure 5. The primary advantage of noncovalent imprinting is the simple formation of the prepolymer complex, which self-assembles by merely combining the functional monomer and
template in solution (Fig. 9). This process is in contrast to the formation of covalent prepolymer complexes that require synthetic transformations to obtain the
functional monomertemplate complex. Furthermore, removing and rebinding
the template in covalent MIPs also requires synthetic steps, which can be challenging once the functional group is incorporated in the polymer matrix. Noncovalent imprinting also can provide better binding sites vs covalent MIPs, which
result from greater flexibility in the positioning of the functional groups and
potentially larger numbers of functional groups that can interact with the template.
Some examples of the more commonly used noncovalent functional monomers are shown in Table 1 (28,6268). Several approaches have been employed
for deciding which functional monomers to use for polymer imprinting including
computational analysis, combinatorial libraries, and general Table 1. Widely
used noncovalent functional monomers chemical intuition. Most reports in the
literature use MAA because of its commercial availability, low cost, and general
success as a functional monomer. Although MAA is often an effective choice, the
optimum choice of functional monomer has several considerations. The first
requirement is that binding interactions between the template and th functional
monomers should be as complementary as possible. For example, hydrogenbonding interactions can be hydrogen donating (D) or hydrogen accepting (A).
A hydrogen-bonding pair, such as a template and a functional monomer, should
align donor groups with acceptor groups and vice versa, as shown in Figure 10. It
is often assumed that the imprinting process alone will configure binding groups
between the template and the polymer that are complementary to the template,
preorganizing the specific interactions needed for selectivity of the template over

Table 1.

Widely Used Noncovalent Functional Monomers

Functional monomer
methacrylic acid
4-vinylpyridine (4-vpy)
acrylamide
2-hydroxyethylmethacrylate
2-(diethylamino)ethyl methacrylate
2,6-bis(acrylamido/methacrylamido)pyridine

References
28,62
6365
66
67
64
68

MOLECULAR IMPRINTING

11

Fig. 10. Examples of noncovalent interactions for MIPs and corresponding bond energy
ranges.

any other analyte. However, further fine tuning can come from the directionality
of the binding interactions between the template and the functional monomers in
MIPs. Thus, after the molecular imprinting process provides preorganization of
functional groups in the binding cavity, the interactions themselves can increase
selectivity in the binding site by limiting the orientation of the interaction. For
example, in a survey of MIPs toward selective binding of 2-phenylbutyric acid,
chiral selectivity was found only for functional monomers such as 2-aminopyridine methacrylamide, which is capable of hydrogen bonding with the template in
a single, coplanar direction (Fig. 11) (69). This is what is meant by an interaction
having specific directionality. However, a primary amine on the MIP provided by

Fig. 11. Comparison of directional vs nondirectional binding interactions, which does not
provide directionality and does not limit the orientation of interaction with the template,
although it does provide a strong binding interaction.

12

MOLECULAR IMPRINTING

monomers such as N-(2-aminoethyl)methacrylamide (Fig. 11) presents a charge


that can be regarded as spherical in nature.
There are also reports of polymers imprinted under stoichiometric noncovalent interactions (70) in which the template and functional monomer are nearly
completely bound to each other. This occurs when the association constants are
considerably high (Ka > 900 M1), providing a full prepolymer complex even in a
1:1 molar ratio. A key factor of this type of molecular imprinting is that the functional monomers only are located inside the imprinted cavity, as in the case of
covalent imprinting. These imprinted polymers also exhibit high capacity (eg,
for preparative applications) and can rebind up to 98% of the theoretical amount
of template.

6. One-Monomer MIPs (OMNIMIPs)


The use of separate functional and cross-linking monomers can limit the success
of MIPs because a balance between both types of monomers is needed, which consequently reduces the amount of either monomer into a less optimal range. A
typical formulation for MIPs uses 20% functional monomer and 80% cross-linker
(71), and a further increase in the functional monomer leads to a reduction of the
amount of cross-linker. The reduction in cross-linker, in turn, can lead to degradation of the recognition or catalytic properties of the MIP through loss of material rigidity. Recently, studies have been directed toward combining the
functional monomer and the cross-linker into the same molecule. Using this
strategy, the amount of cross-linking can be maximized while simultaneously
allowing the maximum amount of functional monomer desired. The use of
these functional cross-linkers has been shown to improve significantly recognition properties of MIPs compared with equivalent functionality and cross-linking
provided by separate molecules (72).
During the investigation of functional cross-linkers, a facile approach to
MIP formation was developed (Fig. 12) using the single cross-linking monomer,
N,O-bismethacryloyl ethanolamine (NOBE) (7375). NOBE has two double
bonds that have been shown to provide adequate cross-linking as well as the
amide functionality, which provides hydrogen bonding interactions with templates. Molecularly imprinted materials formulated with NOBE have been
shown to provide polymers with superior selective binding properties vs traditionally formulated MIPs for most templates. The term OMNiMIPs (onemonomer molecularly imprinted polymers) has been coined to describe this
approach, which eliminates variables such as:

Fig. 12. An outline of the OMNiMIP imprinting strategy.

MOLECULAR IMPRINTING
1.
2.
3.
4.
5.
6.

13

What type of functional monomers (FM) to use


How many functional monomers to use
How much of each functional monomer to use
What type of cross-linker to use
The ratio of functional monomer/cross-linker (FM/XL)
The ratio of functional monomer/template

which normally complicate MIP design. The best results for traditionally formulated imprinted materials generally are determined empirically, which can be
time consuming involving synthesis and evaluation of several MIPs. With OMNiMIPs however, there is only one formulation, which removes the need to optimize
any imprinting systems. Because of the amide interactive group in NOBE,
OMNiMIPs made with this monomer are particularly effective for templates
that interact primarily via hydrogen-bonding. However, templates incorporating
ionic groups tend to work better employing traditional formulations using complementary ionic functional monomers.

7. Types of Templates
Many types of molecular templates are used for imprinting (Table 2), but one of
the more important classifications of templates is according to size. It has been
shown that for the highly cross-linked MIPs traditionally formed, templates
should have a molecular weight of 1100 g/mole or less. In general, complete
removal of the template is not obtained, and during subsequent analyses using
MIPs, small amounts of template continue to bleed from the material. To avoid
Table 2.

Selected Examples of Templates Used for Molecular Imprinting

Types of templates

Specific examples

carbohydrates

mannose
galactose, fucose
cellobiose and maltose
phenylalanine
tryptophan
oligopeptides
adenine
DNA oligomers
nicotine
theophylline
benzodiazepines
phosphorous-based nerve agents
TNT baculoviruses
common proteins: cytochrome C, BSA,
lysozyme, ribonuclease
tobacco mosaic virus
picornaviruses
baculoviruses
transition metals: lanthanides,
group I and II metals

amino acids and peptides


nucleotide bases
bioactive templates
national security agents
proteins
viruses
metal ions

References
77
78
79
80
81
9295
71,86
87
88
8991
92
9395
96,97
87,98,99
100102

14

MOLECULAR IMPRINTING

this in practice, an analog of the targeted molecule is often imprinted instead of


the actual molecule of interest. This way, any template that bleeds during binding analysis of the target molecule will be distinguishable and not confused with
the target being analyzed. A lower size limit for templates has been discussed
with respect to Propofol (2,6-diisopropylphenol), a small substituted phenol
with clinical significance, as the active substituent of the intravenous anaesthetic Diprivan (76). Table 2 (71,77102) gives a few examples of the hundreds
of different templates that have been imprinted; primary targets for small molecules have been biomolecules, bioactive molecules, national security agents, and
environmental toxins to mention a few. Another class of small templates that
have been imprinted is metal ions, sometimes referred to as ion imprinting(103). MIPs to metal ions have been fabricated using typical molecular
level prepolymer complexes (104), or in many cases, a preformed polymer is
further cross-linked under imprinting conditions (105).
For large templates, such as proteins, the traditional highly cross-linked
MIP materials cannot be used because the large templates become encapsulated
without the possibility of mass transfer in or out of the polymer matrix. Most
reports on large molecule imprinting have focused on proteins, and this area
has recently been reviewed (106,107). One approach to solve this problem is to
reduce drastically the amount of cross-linker used (49). Alternatively, several different strategies for surface imprinting or imprinting in thin films have been
explored, such as self-assembled monolayers (108), polymer brushes (109), and
nanoscale thin films (110,111). A third strategy is called the epitope approach
in which a substructure or a structure resembling a small piece of a larger molecule is imprinted (112,113). Subsequent molecular recognition by the MIP material is via the substructure imprinted. The first example of the epitope approach
used a substructure of oxytocin to imprint, which showed selective recognition of
the full oxytocin oligopeptide structure. Recently, excellent selectivity for
cytochrome c was obtained by a polymer imprinted with the nine amino acid
C-terminus of this protein. The formulation for these and other examples usually
uses approximately 30%50% of cross-linker, which is significantly less than
the >80% cross-linker used for traditional imprinting, to allow for better masstransfer kinetics of larger molecules. Some molecular recognition is lost as a
result of the lower cross-linking; however, it is hypothesized that the relatively
large number of noncovalent interactions accounts for the affinity and selectivity
by these materials.

8. Optimization of Polymerization Conditions


For multicomponent monomer systems, there is an optimum ratio of XL/FM.
Figure 13 shows data for MIPs imprinted with the L enantiomer of phenylalanine anilide (L-phe-an), using different ratios of XL/FM (114). The enantioselectivity for L-phe-an vs D-phe-an initially increases as the mole % of functional
monomer (MAA) increases; however, selectivity starts to diminish greater than
20 mole % MAA and then decreases greater than 30 mole % MAA. The loss of
selectivity by imprinted polymers having more than 2030 mole % MAA has
been postulated to result from the need for a minimum amount of cross-linker

MOLECULAR IMPRINTING

15

Fig. 13. Determination of the optimum ratio of XL/FM (adapted from Reference 114).

(EGDMA in this case) to form a rigid enough polymer network that will maintain
the fidelity of the binding site. This limits the amount of noncross-linking functional monomer MAA that can be used for the formation of the MIP binding sites.
Although empirical studies should be carried out for optimization of XL/FM for
each imprinted polymer, a good rule of thumb is to use a 4/1 ratio. Another
important ratio is the amount of functional monomer to template (FM/T) used
to form the prepolymer complex for noncovalent MIPs. Although there is a
limit to the amount of functional monomer that can be employed, the amount
of template can be increased until the point where it becomes 100% of the porogen. A study has been published evaluating a series of MIPs made with increasing amounts of the template nicotine while keeping the amount and ratio of FM/
XL constant (115). In this manner, the FM/T ratio could be varied without changing the FM/XL composition of the final
MIP.0 Figure 14 shows the selectivity of
0
nicotine relative to bipyridine a knicotine =kbipyridine by MIPs made with different FM/T ratios. Low FM/T ratios exhibit low selectivity because only a few specific sites are created from the small concentration of prepolymer complex. As the

Fig. 14. Selectivity as a function of the ratio of FM/T (adapted from Reference 115).

16

MOLECULAR IMPRINTING

FM/T ratio increases, selectivity increases (because of the increase in specific


sites) up to an optimal FM/T ratio that needs to be determined for each system;
in the example presented here, this ratio was 4/1. Further increase in the FM/T
ratio after an empirical optimum results in a decrease in selectivity. The
decrease in selectivity is attributed to the greater increase of nonspecific (or
less-specific) binding sites vs the imprinted specific binding sites.
The polymerization solvent, sometimes referred to as porogen, also plays a
significant role in molecular imprinting. There are two effects the polymerization
solvent; the first is in the formation of the prepolymer complex. For weaker noncovalent polarpolar interactions such as ionic or hydrogen-bonds, nonpolar solvents will promote prepolymer complexation of the template and functional
monomer. The use of polar solvents, especially those capable of hydrogen bonding, inhibits the formation of noncovalent prepolymer complexes, decreasing the
number of imprinted sites in the polymer (Fig. 8). The second effect of the polymerization solvent is on the formation of the porous structure of the polymer,
which is the reason it is referred to as the porogen. The details of how porogens
can be used to control the structure and morphology of these macroporous polymers can be found in the literature (30). To help understand the effects of porogen on selectivity by MIPs, a study published by Sellergren and Shea evaluated
MIPs templated with L-phe-an using several different types of porogens (116).
The results in Table 3 show that enantioselectivity is better for polymers that
are made with relatively nonpolar porogens and have a poor capacity for hydrogen bonding. A decrease in enantioselectivity is seen for MIPs made with polar
porogens with moderate capacity for hydrogen bonding, and enantioselectivity
also decreases with the polarprotic porogens.
An interesting observation from Table 3 is that the MIP made with acetonitrile as the porogen performed with higher enantioselectivity than the MIP made
with chloroform. A larger concentration of prepolymer complex would be
expected using chloroform vs the more polar acetonitrile as the porogen; thus,
a higher performance by the chloroform solvated MIP would be expected. However, it should be noted that in this study, the MIPs all were evaluated in a chromatographic mobile phase primarily made of acetonitrile. Early on in the
development of MIPs, Kempe and Mosbach suggested that using the same solvent in the mobile phase that was used as the porogen would mimic, in the
chromatographic mode, the interactions existing prior to and during the
Table 3.
116)

Selectivity of MIPs Made with Different Porogens (Adapted from Reference

Porogen
acetic acid
dimethylformamide
isopropanol
tetrahydrofuran
chloroform
acetonitrile
benzene
a

Hydrogen-bonding capabilitya
S
M
S
M
P
P
P

Separation factor (a)


1.9
2.0
3.5
4.1
4.5
5.8
6.8

P poor hydrogen bonding solvent, M moderate hydrogen bonding solvent, S strong hydrogen
bonding solvent.

MOLECULAR IMPRINTING

17

Table 4. Capacity Factors of Polymers Imprinted with 9-Ethyladenine Employing


Chloroform and Acetonitrile as Porogen and Mobile Phase (Adapted from Reference
116)
Polymer

Porogen

k0 for mobile phase 1:


85/15 CH3CN/CH3COOH

1
2

CH3CN
CHCl3

7.5
2.7

k0 for mobile phase 2:


85/15 CHCl3/CH3COOH
16.5
2.0

Swelling
(mL/mL)
1.36
2.11

polymerization (117). A possible explanation for this may lie in a link between
conditions during polymerization and those during rebinding analysis on the product network polymer. The origins of specificity in the imprinted polymer are
postulated to result from the positioning of complementary functional groups,
which then are locked covalently into place during polymerization. Different
swelling properties of different solvents, such as chloroform and acetonitrile,
may play a role in determining shape and distance parameters that are locked
into the forming polymer. To recreate and maintain these shape and distance
parameters, it is possible that optimum rebinding conditions require the same,
or very similar, swelling conditions used for polymerization. To test this hypothesis, a study was carried using two MIPs templated with 9-ethyladenine, one
using choroform as porogen and the other using acetonitrile. Selectivity was
evaluated using two mobile phase systems on each column85/15-acetonitrile/
acetic acid and 85/15-chloroform/acetic acid to recreate prepolymerization conditions as well as obtain suitable chromatograms. Table 4 shows the capacity factors of 9-EA for all four cases. These results indicate that the solvent does affect
the microenvironment of the binding sites created in the polymer. It also seems
that there is enhanced binding in polymers immersed in the solvent in which
they were polymerized.
A last consideration is the initiator, and in particular, the process of molecular imprinting most often uses radical polymerization of monomers that have
double bonds. There are three primary reasons for using radical polymerization;
first, radical polymerization eliminates any interference with a noncovalent prepolymer complex by polar intermediates that would develop from other types of
polymerization processes such as anionic, cationic, redox, condensation, or metalcatalyzed polymerization. Second, the kinetics of radical polymerization follows a
chain-growth mechanism vs step-growth kinetics for condensation-type polymerizations. The fast, radical chain-growth mechanism produces a high molecular
weight polymer in a short period of time that could improve the chances for locking-in the prepolymer complex. Third, radical polymerization is an inexpensive
and easy process to carry out.
The typical radical initiator used is azo-bis isobutyronitrile (AIBN), which is
useful for organic polymer solutions. There are various derivatives of azo-initiators available, including a water-soluble species for aqueous polymerizations.
Traditionally, 1 mole % of AIBN is introduced into the prepolymerization mixture. To investigate whether this is an optimum concentration of initiator, several polymers were imprinted with 2-aminopyridine employing different
concentrations of the AIBN initiator (118). The polymers were evaluated in the
chromatographic mode, and the results in Figure 15 show that 1 mole % AIBN

18

MOLECULAR IMPRINTING

Fig. 15. The improvement in binding affinity vs the amount of initiator used in the photoinitiated polymerization of MIPs (adapted from Reference 116).

provides the maximum increase in binding affinity by the MIP. It is interesting


to note that an increase in initiator actually has a positive effect on binding. This
result may find its origins in kinetic parameters (ie, the speed of polymerization)
or in the morphology of the polymer itself.

9. Molecular Imprinting Formats


9.1. Bulk Molecularly Imprinted Polymers. Historically, MIPs have
been formed in bulk and then ground into powders, which often are packed
into columns or cartridges requiring multimicron sized particles (Fig. 16). Unfortunately, the grinding process leads to the pulverization of the MIP into fine particles (up to 80% loss by weight) that are often not useful for subsequent
applications. A recent study has shown that it is not necessary to grind MIPs
into small particles to enhance surface area effects because of the porous structure of MIPs, and it has been shown that a larger particle size facilitates binding
in chromatographic mode (21). An alternative method to grinding MIPs and
packing into chromatographic columns is to polymerize directly within the column; however, the solvents needed to form the needed porous monolith limit
the types of templates that can be used (119,120). Therefore, several methods

Fig. 16. The initial polymer monolith in a test tube and the final powdered MIP product.

MOLECULAR IMPRINTING

19

have been developed to polymerize directly from MIP particles (121). Recently a
chromatographic comparison of different MIP formats (eg, bulk, particles, and
monoliths) has been carried out employing the same template-functional monomer complex for each system (122). The study showed that bulk imprinting still
exhibited the highest performance in binding studies vs the other formats; however, there seemed to be a greater amount of nonspecific binding.
9.2. Particles and Nanoparticles. To resolve problems associated with
bulk polymerization, one approach for the formation of MIP particles is suspension polymerization, either aqueous or in an organic solvent, in which polymerization occurs in droplets suspended in a medium. Prepolymer complexes based
on noncovalent interactions are often sensitive to water, limiting the use of aqueous suspensions. As a result, liquid perfluorcarbons have been used as a more
compatible suspension medium, which can form a suspension of the MIP mixture
in an organic solvent, with the aid of poly-fluorosurfactants. Another method
uses a linear polystyrene latex particle as a seed particle (123). The seed particles
can absorb the MIP solution, which can be polymerized to form cross-linked MIP
beads with a narrow size range. Other types of composite MIP beads can be
formed, often starting with a core particle that serves as a support for another
MIP coating (124). MIP nanoparticles also have been synthesized, which has
potential for chromatographic analysis and recently has been compared with
the dimensions and behavior of antibodies (125). A relatively simple way to
form imprinted nanoparticles is precipitation polymerization, which is carried
out at high dilution in an organic solvent, which maintains the prepolymer complex (126).
9.3. Films, Membranes, and Surface Imprinting. MIPs also can be
formed in film formats (127); however, the traditional formulations for MIPs
are often too brittle to make them practical for making films. Some reports of
composite polymer membranes integrate the MIP polymer within the pores or
as a coating on a stable membrane support (128130). In some instances, nontraditional MIP formulations have been used to create directly the membrane material (131). Surface-imprinted polymer phases are of interest, particularly for
large molecules, such as proteins, that cannot be imprinted using traditional
bulk polymerization formats (132). MIP binding sites formed on or near a surface
can allow access to large molecules, thereby improving the mass-transfer
kinetics and ultimate recognition of the template molecule (133). Both thinfilm formatted MIPs and surface-imprinted materials often are employed for sensor development toward large biomolecules (81).

10. MIP Catalysts and Microreactors


The formation of enzyme-like MIP catalysts has been reviewed (134), and many
successful systems have been demonstrated (135137). Catalytic MIPs often are
made using a biomimetic approach by imprinting a stable analog of a reactions
transition state, following similar methods used to elicit catalytic antibodies
(138). A typical example is the hydrolysis of an ester illustrated in Figure 17
in which a stable analog is imprinted that resembles the transition state of the
reaction, such as the phosphate compound shown in Figure 17a. The resulting

20

MOLECULAR IMPRINTING

Fig. 17. An energy diagram for catalyzed and uncatalyzed saponification of esters.

MIP is exposed to the ester-starting material and catalyzes a hydrolysis of the


ester (Fig. 17b). This strategy was used by Mosbach and Robinson (139), who
used the phosphonate compound in Figure 18 as the template to form MIPs
that hydrolyze p-nitrophenol acetate (the reaction shown in Fig. 17) 1.6-fold faster than a nonimprinted control polymer. More complicated systems developed
by Sellergren and Shea enabled enantioselective ester hydrolysis by MIPs (140).
Microreactors are similar to catalysts in that they can influence the rate,
regiochemistry, or stereochemistry of a reaction, with the exception that a stoichiometric amount is required (ie, no turn over). Imprinted polymers have been
used as microreactors for several reactions, for example, the regio- and stereochemical control of the hydride reduction of a keto steroid shown in Figure 18.
An interesting application of MIPs as microreactors is for anti-idiotypic imprinting, which first imprints a molecule in its entirety. The removal of the template
leaves a site that catalyzes the coupling reaction of reactive partners that most
resemble the two halves of the original molecule (Fig. 19) (141). An analogous
method, referred to as direct molding, also has been carried out directly

Fig. 18. Imprinted polymer microreactor control of regio- and stereochemistry for the hydride reduction of keto steroids.

Fig. 19. An outline of the antiidiotypic imprinting method.

MOLECULAR IMPRINTING

21

using an enzyme that catalyzes the coupling of reactive partners that bind the
enzyme active site best (142,143).

11. Applications of MIPs


11.1. Liquid Chromatography. One of the most reported applications
of imprinted polymers is in the area of chromatography, in many cases for characterization of the binding properties rather than separation applications (144).
Because of their solid nature, MIPs are predisposed to incorporation into column
formats and essentially operate by an affinity mechanism. When used as high
performance liquid chromatography stationary phases, the imprinted polymers
have been shown to separate a variety of molecules including pharmaceuticals,
inorganic ions, carbohydrates, amino acids, and nucleic acids. Direct separation
of enantiomers represents one of the more challenging tasks in separation
science, and MIPs have been proven capable of forming efficient chromatographic chiral stationary phases (CSPs). For example, the separation of the
enantiomers of nicotine has been achieved using different MIP formulations
(145,146). Because of the heterogeneous distribution of binding sites discussed
earlier, in addition to slow mass-transfer kinetics, chromatographic results
show broad peaks and extensive tailing for the template molecule. Nonetheless,
affinity separations routinely are achieved by the initial removal of extraneous
compounds that elute rapidly and generally show a narrow band, followed by a
long-term collection of the template during its broad and tailing elution. An
advantage of MIPs to general CSPs that are nontarget specific is the ability to
make a reliable prediction of the elution order of an enantiomeric separation.
Of potentially greater interest is the imprinting of racemic mixtures for chiral
separations, thus eliminating the need for chiral-pure templates (147). Through
the use of chiral imprinting monomers, diastereomeric prepolymer complexes are
formed, which result in a material that has a different binding potential for each
of the enantiomers. MIPs also have been formed with multiple templates simultaneously, which can resolve enantiomers of each of the templates (148,149).
One of the most successful applications of imprinting has been in the area of
molecularly imprinted solid phase extraction (MISPE) (150)the leading commercial enterprise of MIPs to date (151,152). Current state-of-the-art analytic
methods have the sensitivity for contamination detection and quantification of
chemo- or biomarkers, but most are not capable of direct analysis of the complex
matrices such as biofluids (eg, blood), environmental samples from land extracts
or bodies of water, and mixtures with targets of interest to national security.
Thus, samples often first are pretreated using methods such as solid phase
extraction for clean-up and preconcentration of samples for further sensitive
analyses. Solid phase extraction methods based on nonselective resins are fast
and inexpensive, but they cannot always target only the uptake of compounds
of interest. MISPE materials can provide the needed selectivity in an inexpensive format for rapid and selective clean-up methods and have been used in
many types of applications (153155). In fact, this technique seems to be
particularly suitable for extractive applications in which analyte selectivity in
complex solutions is the main problem of sample preparation. Capillary

22

MOLECULAR IMPRINTING

electrochromatography (CEC) is another promising application of MIPs, which


combines elements of liquid chromatography and electrophoresis. The mobilephase flow through CEC systems is accomplished by electroosmotic flow instead
of pressure pumping, giving rise to a stable plug-likerather than a parabolic
flow profile, which provides higher chromatographic efficiency vs other liquid
chromatography methods. Uses of MIPs in CEC have been reviewed, and MIP
formats include particles packed into fused-silica capillaries, Thin films of MIP
can be grafted onto silica particles containing surface-bound free radical initiators, and monolithic MIP phases were prepared in situ by photo-initiated
polymerization. Recently, MIP nanoparticles have been employed as a pseudostationary phase in CEC, which can move back and forth in an oscillating electric
field, which increases the number of theoretical plates with a significant
improvement in separation efficiency (156).
11.2. MIP Based Sensors. The binding sites of MIPs have been shown
to follow an exponential decaying distribution of binding affinities (see Evaluation of Binding and Selectivity) with small numbers of the more selective high
affinity sites. It also has been shown that the higher affinity and more selective
sites are filled first upon rebinding the template under equilibrium conditions
(157). Therefore, the best results by MIPs are seen at low concentrations of
template rebinding, potentially making these materials more suitable for
sensor applications on small samples of template vs separations requiring
higher concentrations. Furthermore, MIPs have been incorporated into array
formats, which can increase further the sensitivity and selectivity of analyte
identification.
Mass sensors have been used successfully for monitoring the binding of
analytes to MIP coatings. The most widely used mass sensors have been based
on quartz crystal microbalance (QCM) technology, and, to a lesser extent, with
acoustic wave devices. The use of a QCM requires the attachment of the MIP
to the crystal-gold surface, which has been accomplished by derivatization of
the surface and directly imprinting or by adhering MIP particles to the surface.
Most MIP-QCM examples have been applied to solution-phase analytes; however, there are some examples of measurements on gas phase samples (158).
Another sensitive method for measuring binding affinities that also generally
requires immobilization of MIPs to a gold surface is surface plasmon resonance
(SPR) spectroscopy. Detection limits as small as 10 pM have been obtained for
trinitrotoluene (TNT) MIPs (159), and enantioselective binding has been demonstrated using this technique (160). A significant advantage of SPR-based sensing
approaches over other sensors (eg, QCM) is that the former more easily can be
adapted to measurement in aqueous media. However, both QCM and SPR methods have the advantage of direct measurement of template binding without any
need to add an identifiable tag (eg, fluorescent tag).
Fluorescence detection also can afford sensitive detection of analytes but
requires either the analyte (161) or the MIP matrix to have fluorescent properties (162). Analytes with intrinsic fluorophores or those that have been modified
with a fluorescent tag can be measured directly, even opaque MIPs. Alternatively, functional monomers incorporating fluorophores have been used in
MIPs that exhibit changes in the fluorescent wavelength upon binding the template molecule. Other methods such as IR spectroscopy and solid-state NMR

MOLECULAR IMPRINTING

23

spectrometry can evaluate analyte uptake in MIP materials, but quantification is


difficult. Electrochemical sensors have the potential for amperometric, impedence, or conductometric sensing of MIP binding (163,164). These methods of
detection are relatively inexpensive and can incorporate the MIP materials in
a myriad of formats, including deposition directly onto electrodes, incorporation
of conducting polymers, and field-effect transistors (FETs). The FET sensors can
detect a change in the potential of the gate of these devices as a consequence of
analyte binding to the MIP matrix incorporated or deposited on the gate.
Recently, color-responsive MIPs have been developed using polymer-based
photonic crystals (165). The color change occurs from Bragg diffraction changes
in an MIP with controlled pore size caused by the shrinking of the polymer upon
the binding of the template. The method is sensitive, with detection limits down
to the picomolar range, and does not require any labeling of the template.
11.3. Batch Assays. Assays using MIPs can be run in batch-mode in
which the MIP material is added to a solution of template and binding analyzed
under equilibrium conditions that avoid problems caused by slow mass-transfer
kinetics of template binding. One promising application is the development of
immuno-type assays in which MIPs replace biological receptors (eg, antibodies)
as the recognition element (166). In one approach, the template remaining in
solution after incubation with the MIP can be measured, for example, by UV
spectroscopy or by scintillation counting of a radio-labeled template. Binding isotherms often are determined in this manner to determine the binding constants
and the number of binding sites (see Chromatography). Alternatively, assays can
be carried out by measuring the amount of a labeled template (eg, incorporating
a chromophore (167), fluorophore (168), or radioactive label) that is displaced by
a nonlabeled template. The radio-labeled template displacement assay is a superior method for measuring binding and selectivity in MIPs because the template
structure is unchanged, which is not the case with tag-modified templates.
Furthermore, the radio-labeled templates can be detected at very low concentrations, allowing measurement of the highest affinity and most selective MIP binding sites. An interesting method recently has been developed to determine the
enantiomeric excess of chiral mixtures (169).This method allows analysis of a
mixture of optical isomers on an enantioselective MIP based on a comparison
of uptake measurements to a predetermined calibration curve.

12. Summary and Conclusions


MIPs are becoming recognized as a viable science with potential for many applications. However, the study and practice of molecular imprinting can be
described by the old adage It takes a minute to learn, and a lifetime to master.
The concept of molecular imprinting is conveyed easily using diagrams such as
Figure 2. Moreover, the ease of fabrication of traditional form bulk MIPs has
attracted a wide scientific audience that often has reproduced the molecular
recognition phenomena of these materials. Yet, several aspects of molecular
imprinting are difficult to understand and control. For example, molecular
imprinting in bulk materials suffers from a broad distribution of sites, with
most sites exhibiting low affinity and selectivity. This drawback has led to

24

MOLECULAR IMPRINTING

difficulty in characterizing MIP materials accurately with respect to binding


properties. Although recent advances have helped to increase the understanding
of binding site heterogeneity, problems associated with this characteristic continue to plague the overall performance of MIPs. For example, biological receptors such as antibodies can be purified and cloned to give a single binding site
that is highly effective. Traditionally formed MIPs, however, are trapped with
an ensemble of the distribution of binding sites where the preponderance of
poor-performing sites mask the performance of high performing sites. Thus,
direct comparison with antibodies may not be entirely appropriate; however,
MIP materials are more stable than biological receptors with long shelf-lives
(on the order of decades) and can withstand different temperatures and solvents.
Furthermore, applications involving small amounts of an analyte, such as sensors and binding assay, are suitable for binding the best sites in MIPs. For the
future, improvements in MIP methodology are continuing to progress, including
efforts toward new formats (170), new applications (171), and new materials
(17,18). Many of these innovations may go far toward solving the shortcomings
of MIPs. Nonetheless, there are many current examples in which MIPs can be
powerfully effective in binding and catalytic applications.

BIBLIOGRAPHY
1. C. Alexander and co-workers, J. Mol. Recognit. 19, 106180 (2006).
2. K. Mosbach, Sci. Am. 295, 8791 (2006).
3. M. Yan and O. Ramstrom, eds, Molecularly Imprinted Materials: Science and Technology, Marcel Dekker, New York, 2005.
4. B. Sellergren, American Lab. 1420 (1997).
5. V. Pichon and K. Haupt, J. Liq. Chrom. Related Tech. 29, 9891023 (2006).
6. K. Haupt, A. G. Mayes, and K. Mosbach, Anal. Chem. 70, 3936393 (1998).
7. G. Vlatakis and co-workers, Nature 361, 645647 (1993).
8. G. Wulff, Chem. Rev. 102, 127 (2002).
9. S. A. Piletsky and A. P. F. Turner, Optical Biosens. 397425 (2002).
10. D. Kriz, O. Ramstrom, and K. Mosbach, Anal. Chem. 345A349A (1997).
11. S. Tokonami, H. Shiigi, and T. Nagaoka, Anal. Chim. Acta 641, 713 (2009).
12. M. C. Norell, H. S. Andersson, and I. A. Nicholls, J. Mol. Recogn. 11, 98102 (1998).
13. C. J. Allender and co-workers, Proceedings of the International Symposium on Controlled Release of Bioactive Materials, Controlled Release Society, Inc., St. Paul,
Minn., 1997.
14. D. A. Spivak, Adv. Drug Delivery Rev. 57, 17791794, (2005).
15. B. Sellergren and K. J. Shea, J. Chromatogr. 635, 3149, (1993).
16. S. C. Zimmerman and M. G. Lemcoff, Chem. Comm. 517 (2004).
17. J. D. Marty and M. Mauzac, Adv. Polym. Sci. 172, 135 (2005).
18. D. A. Spivak and K. J. Shea, J. Org. Chem. 64, 46274634 (1999).
19. K. J. Shea and co-workers, Macromolecules 23, 4497507 (1990).
20. R. Simon, M. E. Collins, and D. A. Spivak, Anal. Chim. Acta 91, 716 (2007).
21. S. H. Cheong and co-workers, Macromolecules 30, 13171322 (1997).
22. M. Ringo and C. Evans. Anal. Chem. 70, 315A321A (1998).
23. G. Wulff and co-workers, Markromol. Chem. 178, 28172825 (1977).
24. R. J. Umpleby and co-workers, Anal. Chem. 73, 45844591 (2001).
25. R. J. Umpleby II and co-workers, Anal. Chim. Acta 435, 3542 (2001).

MOLECULAR IMPRINTING
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.

25

R. J. Umpleby II, M. Bode, and K. D. Shimizu. Analyst 125, 12611265 (2000).


P. Sajonz and co-workers, J. Chromatogr. A 810, 117 (1998).
H. Kim and D. A. Spivak, J. Am. Chem. Soc. 125, 1126911275 (2003).
K. J. Shea, D. A. Spivak, and B. Sellergren, J. Am. Chem. Soc. 115, 33683369
(1993).
J. Hearn, P. L. Smelt, and M. C. Wilkinson, J. Colloid Interface Sci. 133, 284287
(1989).
G. Wulff and A. Sarhan, Angew. Chem. Int. Ed. Engl. 11, 341 (1972).
G. Wulff, J. Vietmeier, and H.-G. Poll, Makromol. Chem. 188, 731740 (1987).
M. Kempe and K. Mosbach, Tetrahedron Lett. 36, 35633566 (1995).
M. Kempe, Anal. Chem. 68, 19481953 (1996).
K. J. Shea and co-workers, Macromolecules 23, 44974507 (1990).
K. Tanabe and co-workers, J. Chem. Soc., Chem. Commun. 22, 23032304 (1995).
J. L. Defreese and A. Katz, From Molecularly Imprinted Materials, CRC Press, Boca
Raton, Fla., 2005, pp. 307327.
J. D. Bass and A. Katz, Chem. Mat. 15, 27572763 (2003).
M. M. Titirici and B. Sellergren, Anal. Bio. Chem. 378, 19131921 (2004).
W. G. Borghard and co-workers, Langmuir, 25, 1266112669 (2009).
K. Morihara, in ACS Symposium Series 703, Molecular and Ionic Recognition with
Imprinted Polymers, 1998 pp. 300313.
J. Li, C. E. Kendig, and E. Nesterov, J. Am. Chem. Soc. 129, 1591115918 (2007).
R. Thoelen and co-workers, Biosens. Bioelectron. 23, 913918 (2008).
G. Bidan and co-workers, J. Am. Chem. Soc. 114, 59865994 (1992).
T. L. Panasyuk and co-workers, Anal. Chem. 71, 46094613 (1999).
A. Namvar and K. Warriner, Biosens. Bioelectron. 22, 20182024 (2007).
M. Ali and co-workers, J. Control. Rel. 124, 154162 (2007).
M. E. Byrne and V. Salian, Int. J. Pharm. 364, 188212 (2008).
N. Ghasemzadeh, F. Nyberg, and S. Hjerten, J. Stellan J. Sep. Sci. 31, 39543958
(2008).
M. Watanabe and co-workers, J. Am. Chem. Soc. 120, 55775578 (1998).
M. E. Byrne, K. Park, and N. A. Peppas, Adv. Drug Deliv. Rev. 54, 149161 (2002).
K. Yamashita and co-workers, Polym. J. 35, 545550 (2003).
J. M. Wu, Y. Y. Wang, and C. L. Yan, Chin. J. Anal. Chem. 30, 14141417 (2002).
K. Yamashita and co-workers, Polym. J. 35, 545550 (2003).
T. Miyata and co-workers, Proc. Nat. Acad. Sci. U.S.A. 103, 11901193 (2006).
X. C. Liu and J. S. Dordick, J. Polym. Sci. A, Polym. Chem. 37, 16651671 (1999).
P. Ehrlich, Lancet 172, 16341636 (1907).
G. Wulff, in Ref. 3, pp. 5992.
M. A. Khasawneh, P. T. Vallano, and V. T. Remcho, J. Chromatogr. A 922, 8797
(2001).
C. D. Ki and co-workers, J. Am.Chem. Soc. 124, 1483814839 (2002).
L. Andersson, B. Sellergren, and K. Mosbach, Tetrahedron Lett. 25, 52115214
(1984).
R. Del Sole and co-workers, Molecules 14, 26322649 (2009).
E. Herrero-Hernandez, R. Carabias-Martinez, and E. Rodriguez-Gonzalo, Anal.
Chim. Acta 650, 195201 (2009).
R. Simon and D. A. Spivak, J. Chromatogr, B Anal. Technol. Biomed. Life Sci. 804,
203209 (2004).
Q.-Z. Feng and co-workers, Anal. Bioanal. Chem. 391, 10731079 (2008).
Z. Zhang and co-workers, App. Surf. Sci. 255, 93279332 (2009).
R. J. Ansell and D. Wang, Analyst 134, 564576 (2009).
P. Manesiotis and co-workers, Mat. Chem. 19, 61856193 (2009).

26
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.

104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.

MOLECULAR IMPRINTING
D. A. Spivak and K. J. Shea, J. Am. Chem. Soc. 119, 43884393 (1997).
G. Wulff, and K. Knorr, Bioseparation, 10, 257276 (2002).
D. A. Spivak, in Ref. 3, pp. 395417.
M. Sibrian-Vazquez and D. A. Spivak, J. Org. Chem. 68, 96049611 (2003).
M. Sibrian-Vazquez and D. A. Spivak, J. Am. Chem. Soc. 126, 78277833 (2004).
J. LeJeune and D. A. Spivak, Biosens. Bioelect. 25, 604608 (2009).
K. Yoshimatsu and co-workers, Analyst 134, 719724 (2009).
M. Petcu and co-workers, J. Molec. Recogn. 22, 1825 (2008).
G. Wulff and co-workers, Makromolekulare Chemie, 178, 27992816, (1977).
A. G. Mayes, L. I. Andersson, and K. Mosbach, Anal. Biochem. 222, 483488 (1994).
S. Striegler, Bioseparation, 10, 307314 (2002).
B. Sellergren, B. Ekberg, and K. Mosbach, J. Chromatogr. 347, 110 (1985).
H. Kim and G. Guiochon, Anal. Chem. 77, 64156425 (2005).
M. M. Titirici and B. Sellergren, Anal. Bioanal. Chem. 378, 19131921 (2004).
B. R. Hart and K. J. Shea, J. Am. Chem. Soc. 123, 20722073 (2001).
I. A. Nicholls, O. Ramstroem, and K. Mosbach, J. Chromatogr. A 691, 349353
(1995).
K. Farrington, E. Magner, and F. Regan, Anal. Chim. Acta 566, 6068 (2006).
H. S. Andersson and co-workers, J. Chromatogr. A 848, 3949 (1999).
G. Vlatakis and co-workers, Nature 361, 645647 (1993).
B. R. Hart, D. J. Rush, and K. J. Shea, J. Am. Chem. Soc. 122, 460465 (2000).
R. J. Umpleby and co-workers, Macromolecules 34, 84468452 (2001).
J. Matsui, M. Higashi, and T. Takeuchi, J. Am. Chem. Soc. 122, 52185219 (2000).
D. Spivak, M. A. Gilmore, and K. J. Shea, J. Am. Chem. Soc. 119, 43884393 (1997).
D. A. Spivak and K. J. Shea, Anal. Chim. Acta 435, 6574 (2001).
A. L. Jenkins and S. Y. Bae, Anal. Chim. Acta 542, 3237 (2005).
G. Bunte and co-workers, Prop. Exp. Pyrotech. 34, 245251 (2009).
L. D. V. Bolisay and co-workers, Res. Soc. Symp. Proc. 787, 2933 (2004).
H. Nishino, C.-S. Huang, and K. J. Shea, Angew. Chem., Int. Ed. 45, 23922396
(2006).
H. Shi and co-workers, Nature 398, 593597 (1999).
L. D. Bolisay, J. N. Culver, and P. Kofinas, Biomacromolecules 8, 38933899 (2007).
M. Jenik and co-workers, Anal. Chem. 81, 53205326 (2009).
T. Rao, K. Prasada, and S. Daniel, Anal. Chim. Acta 578, 105116 (2006).
G. M. Murray and E. Southard, in Ref. 37, p. 579.
S. Zhong and co-workers, J. Chongqing Univ. 7, 2327 (2008).
K. Tsukagoshi, M. Murata, and M. M. Maeda. in Molecularly Imprinted Polymers:
Man-Made Mimics of Antibodies and their Applications, Elsevier, Amsterdam,
The Netherlands, 2001, pp. 245269.
J. F. Krebs and A. S. Borovik, in Molecular and Ionic Recognition with Imprinted
Polymers, Vol. 7, ACS Symposium Series, Washington, D.C., 1998, pp. 159169.
K. Ohga, Y. Kurauchi, and H. Yanase, Bull. Chem. Soc. Jpn. 60, 444446 (1987).
T. Takeuchi and T. Hishiya, Org. Biomol. Chem. 6, 24592467 (2008).
N. W. Turner and co-workers, Biotechnol. Prog. 22, 14741489 (2006).
K. Tappura, I. Vikholm-Lundin, and W. M. Albers, Biosens. Bioelectron. 22, 912919
(2007).
B. Zdyrko, O. Hoy, and I. Luzinov, Biointerphases 4, FA17FA21 (2009).
H. Shi and co-workers, Nature 398, 593597 (1999).
A. Bossi and co-workers, Anal. Chem. 73, 52815286 (2001).
A. Rachkov and N. Minoura, Biochim. Biophys. Acta 1544, 255266 (2001).
D.-F. Tai and co-workers, 77, 51405143 (2005).
B. Sellergren, Makromol. Chem. 190, 27032711 (1989).

MOLECULAR IMPRINTING
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.

27

H. S. Andersson and co-workers, J. Chromatogr. A 848, 3949 (1999).


B. Sellergren and K. J. Shea. J. Chromatogr. 635, 3149 (1993).
M. Kempe and K. Mosbach, Anal. Lett. 24, 11371145 (1991).
D. Spivak, Ph.D. dissertation, University of California, Irvine, Irvine, Calif., 1995.
L. Schweitz, L. I. Andersson, and S. Nilsson, Anal. Chim. Acta 435, 4347 (2001).
B. Sellergren, J. Chromatogr. Lib. 67, 277300 (2003).
R. E. Fairhurst and co-workers, Biosens. Bioelectron. 20, 10981105 (2004).
J. Oxelbark and co-workers, J. Chromatogr. A 1160, 215226 (2007).
L. Zhang and co-workers, Polymer Eng. Sci. 43, 965974 (2003).
C. Sulitzky and co-workers, Macromolecules 35, 7991 (2002).
Y. Hoshino and co-workers, J. Am. Chem. Soc. 130, 1524215243 (2008).
K. Yoshimatsu and co-workers, Analyst 134, 719724 (2009).
M.-M. Titirici and B. Sellergren, Chem. Mat. 18, 17731779, (2006).
S. A. Piletsky and co-workers, J. Membr. Sci. 157, 263278 (1999).
F. Tasselli, L. Donato, and E. Drioli, J. Membr. Sci. 320, 167172 (2008).
K. Takeda and T. Kobayashi, J. Membr. Sci. 275, 6169 (2006).
H. Y. Wang and co-workers, J. Chromatogr. B 804, 127134 (2004).
S. M. Husson and D. Gopireddy, Sep. Sci. Tech. 38, 28512866 (2003).
C. J. Tan and Y. W. Tong, Anal. Bioanal. Chem. 389, 369376 (2007).
G. Wulff, Chem. Rev. 102, 127 (2002).
J.-Q. Liu and G. Wulff, J. Am. Chem. Soc. 130, 80448054 (2008).
J. Svenson, N. Zheng, and I. A. Nicholls, J. Am. Chem. Soc. 126, 85548560 (2004).
D. Carboni and co-workers, Chem. A Eur. J. 14, 70597065 (2008).
R. A. Lerner, S. J. Benkovic, and P. G. Schultz, Science 252, 659657 (1991).
D. K. Robinson and K. Mosbach, J. Chem. Soc. Chem. Comm. 969970 (1989).
B. Sellergren, R. N. Karmalkar, and K. J. Shea, J. Org. Chem. 65, 40094027 (2000).
K. Mosbach and co-workers, J. Am. Chem. Soc. 123, 1242012421 (2001).
Y. Yu and co-workers, Angew. Chem. Int. Ed. 41, 44604462 (2002).
W. G. Lewis and co-workers, Angew. Chem. Int. Ed. 41, 10531057 (2002).
B. Sellergren, Chiral Separation Techniques, 3rd ed., John Wiley & Sons, Inc.,
New York, 2007, pp. 399431.
M. Sibrian-Vazquez and D. A. Spivak, Macromolecules 36, 51055113 (2003).
D. A. Spivak and M. Sibrian-Vazquez, Bioseparation 10, 331336 (2002).
J. LeJeune and D. A. Spivak, Anal. Bioanal. Chem. 389, 433440 (2007).
A. C. Meng, J. LeJeune, and D. A. Spivak, J. Molec. Recogn. 22, 121128 (2009).
J. LeJeune and D. A. Spivak, Biosens. Bioelectron 25, 604608 (2009).
F. G. Tamayo, E. Turiel, and A. Martin-Esteban, J. Chrom. A 1152, 3240 (2007).
MIPTechnologies web information. Available at http://www.miptechnologies.com/.
Accessed October 31, 2009.
PolyIntell web information. Available at: http://www.polyintell.com/index_uk.htm#history. Accessed October 31, 2009.
L. I. Andersson, Bioseparation 10, 353364 (2001).
C. Baggiani, L. Anfossi, and V. Giovannoli, Anal. Chim. Acta 591, 2939 (2007).
R. A. Anderson and co-workers, Forensic Sci. Int. 174, 4046 (2008).
C. Nilsson, S. Birnbaum, and S. Nilsson, J. Chrom. A 1168, 212224 (2007).
N. M. Brunkan and M. R. Gagne, J. Am. Chem. Soc. 122, 62176225 (2000).
Y. Fu and H. O. Finklea, Anal. Chem. 75, 53875393 (2003).
M. Riskin and co-workers, J. Am. Chem. Soc. 131, 73687378 (2009).
P. Li and co-workers, Sensors 2, 3540 (2002).
S. Guillon and co-workers, Lab. Chip 9, 29872991 (2009).
T. Nguyen, T. Hien, and R. J. Ansell, Org. Biomolec. Chem. 7, 12111220 (2009).
D. Kriz, M. Kempe, and K. Mosbach, Sens. Actuat. B B33, 178181 (1996).

28
164.
165.
166.
167.
168.
169.
170.
171.

MOLECULAR IMPRINTING
D. Lakshmi and co-workers, Anal. Chem. 81, 35763584 (2009).
Z. Wu and co-workers, Chem. A Eur. J. 14, 1135811368 (2008).
G. Vlatakis and co-workers, Nature 361, 645647 (1993).
R. Levi and co-workers, Anal. Chem. 69, 20172021 (1997).
A. Rachkov and co-workers, Chim. Acta 405, 2329 (2000).
Y. Z. Chen and K. D. Shimizu, Org. Lett. 4, 29372940 (2002).
S. C. Zimmerman and M. G. Lemcoff, Chem. Commun. 517 (2004).
J. D. Marty and M. Mauzac, Adv. Polym. Sci. 172, 135 (2005).

DAVID SPIVAK
Louisiana State University

Vous aimerez peut-être aussi