CT 4/99: Molecular recognition using surface template polymerization

CHEMTECH
April 1999
CHEMTECH 1999, 29(4), 12-18.
Copyright © 1999 by the American Chemical Society.

ENABLING SCIENCE

Molecular recognition using surface template polymerization

This technique, which uses self-assembly of functional host molecules, facilitates the use of water-soluble substances and has great potential for biological and clinical applications.

Kazuya Uezu
Masahiro Yoshida
Masahiro Goto
Shintaro Furusaki

M olecular imprinting is a technique for preparing polymeric materials that are capable of high molecular recognition. This method usually involves cross-linking the functional monomers in the presence of template organic molecules or ions by radical polymerization, then removing the target molecules. The imprinted polymers selectively bind again with the template molecules or ions in a mixture of chemical species. The concept of molecular-imprinted polymers was introduced by Wulff and Sarhan in the early 1970s (1). The Wulff and Sarhan method relied on a polymeric receptor that used reversible covalent bonding for molecular-imprinted polymerization.

Molecular imprinting has been successfully applied to create recognition sites that use noncovalent interactions (hydrogen bonding and/or electrostatic interactions) in synthetic polymers for a variety of molecules, including nucleotide bases (2), steroid hormones (3), dipeptide derivatives (4), amino acids (5), and artificial antibodies (6).

We have developed a novel molecular-imprinting technique called surface template polymerization (Figure 1). This technique allows us to overcome the fundamental drawbacks of conventional polymer imprinting methods (7-15). These drawbacks include slow rebinding kinetics and difficulty in handling water-soluble templates such as metal ions and biological components such as amino acids and proteins. In contrast, surface-imprinted polymers are prepared by emulsion polymerization using a functional host molecule, an emulsion stabilizer, a polymer matrix-forming monomer, and a print molecule. This method is characterized by allowing these molecules to self-assemble at the water-oil interface to form recognition sites. A functional host molecule, which is amphiphilic in nature, forms a complex with a print molecule during emulsion; thus, the complex formed remains at the reaction surface. After the matrix is polymerized, the coordination structure is eventually "imprinted" on the polymer surface, not in the polymer matrix.

Figure 1.

Preparation strategy
To prepare highly selective surface-imprinted polymers, it is necessary to fix the binding sites for a target molecule rigidly on the polymer surface. The assessment of a suitable functional host molecule, which forms binding sites and recognizes the target molecule, is a matter of great importance. In the beginning, we considered the requirement of functional host molecules with strong binding ability for the target molecule and high interfacial activity. We prepared a Zn(II)-imprinted polymer by surface template polymerization using dioleyl phosphoric acid (DOLPA) as a functional host molecule and divinylbenzene (DVB) as a polymer matrix-forming monomer (9).

Both the binding ability for Zn(II) and the interfacial activity of DOLPA are very high.

The polymer exhibited an imprinting effect as measured by the amount of metal ions adsorbed; however, the Zn(II)-imprinted polymer showed poor selectivity for Zn(II) when both Zn(II) and Cu(II) were present in aqueous solution. The poor selectivity was probably due to insufficient rigidity of the polymer matrix causing increased swelling of the imprinted polymers. In comparison, DVB polymers with no functional host molecules have more rigid matrixes and exhibit insignificant swelling. Thus, the structure of the functional host molecule has a dominant effect on the rigidity of the polymer matrix. It has been established that host molecules that have long alkyl chains as hydrophobic units affect matrix rigidity.

To make the polymer matrix more rigid and create stronger interactions between the functional host molecules and imprint molecules, we used two approaches to prepare Zn(II)-imprinted polymers: postirradiation with -rays and a novel functional host molecule design.

Postirradiation with -rays
A Zn(II)-imprinted polymer was prepared with DOLPA as a functional host molecule and DVB as a polymer matrix-forming monomer. After drying in vacuo, the imprinted polymer was irradiated with (10). The matrix was made rigid by the irradiation.

The pH dependence of adsorption of Zn(II) and Cu(II) by the Zn(II)-imprinted polymers with DOLPA is shown in Figure 2. No difference is noted for Zn(II) sorption by the imprinted polymers with and without postirradiation with -rays. An irradiated blank polymer that did not include DOLPA adsorbed no metal ions (data not given). These results clearly show that -ray irradiation does not destroy the Zn(II)-DOLPA complex and produces few functional groups such as carbonyl groups. In contrast, the Cu(II)-binding ability of the imprinted, -irradiated polymer decreased markedly. Thus, the imprinted, -irradiated polymers can distinguish Zn(II) coordination from that of Cu(II).

Figure 2.

The improved selectivity of the imprinted,-irradiated polymers is attributed to the cross-linking induced in the polymer matrix by irradiation, which renders the polymer matrix rigid and thereby enhances the stability of the binding sites toward Zn(II) recognition. Furthermore, the double bond in the oleyl chains of DOLPA can form single bonds with free radicals on each end to combine DOLPA rigidly with the polymer matrix.

These results show that rigid and dimensionally stable metal ion-imprinted polymers that recognize metal coordination can be prepared by anchoring the functional host molecule (e.g., DOLPA) to the polymer surface. It is well-known that properties such as matrix rigidity bring about poor mass transfer in conventionally imprinted polymers. The ability to enhance the matrix rigidity without decreasing mass transfer is therefore an important advantage in surface template polymerization. Furthermore, the combination of the surface template polymerization with -ray irradiation offers a potential technique for constructing highly selective, molecule-recognizing polymers that are applicable to the sorption of various water-soluble substances.

Host molecule design
To fix the recognition sites more rigidly, we considered the design requirements for the functional host molecules:

strong binding ability for the target metal ions,

high interfacial activity, and

no detrimental effect on the polymer matrix.

To fulfill these requirements, we designed the functional host DDDPA (diphenyldodecyldiphosphonic acid) (11). DDDPA

The two phosphonic acid groups were prepared to produce strong binding with Zn(II), the target metal ion. The two benzene rings were included in the hydrophilic portion of DDDPA to preserve the polymer matrix, which is formed by DVB. Alkyl chains, which link the two phenylphosphonic acid units, controlled the interfacial activity of the functional host molecules and their solubility in organic solvents.

Zn(II)-imprinted polymers derived from this multifunctional host are expected to be highly selective toward Zn(II) over Cu(II) because the polymers combine both the rigid polymer matrixes and the strong binding ability due to the specificity of the multifunctional host molecule.

The interfacial activity of DDDPA is 8 times greater than that of DOLPA. Therefore, DDDPA satisfied the second requirement in the design of functional hosts. The pH dependence for sorption of Zn(II) and Cu(II) on a Zn(II)-imprinted polymer prepared with DDDPA is shown in Figure 3. The percent sorption was enhanced with increased pH for both ions. However, the imprinted polymer adsorbed Zn(II) much more effectively than it adsorbed Cu(II) over the entire pH range. The ability of the imprinted polymers to adsorb Zn(II) is significantly higher than for Cu(II) because of the high interfacial activity of DDDPA and the strong interaction between Zn(II) and DDDPA. It should be noted that Zn(II) is completely separated from Cu(II) in aqueous solutions with pH ~3. This high selectivity is produced by the Zn-imprinted cavity on the surface of polymers.

Figure 3.

On the basis of our design guideline, we also synthesized several other functional host molecules.

We then determined the desirable ranges for the swelling ratio (rigidity of the recognition sites) of Zn(II)-imprinted polymers and the interfacial activity of functional host molecules (Figure 4) (12). The functional host molecules that have aromatic rings and a high interfacial activity, such as DDDPA and n-DDPA (n-dodecylphosphonic acid), were able to fix the recognition sites rigidly on the polymer surface and, consequently, produced a high selectivity for zinc ions. If only the functional hosts exhibited appropriate interfacial activity and strong binding characteristics, then the polymer would have demonstrated an excellent template effect by being irradiated with

-rays, as in the case of DOLPA.

Figure 4.

Separation of lanthanoid elements
Because of their similar chemical and physical properties, lanthanoids behave almost identically. Consequently, they are in great demand for the production of novel advanced materials used in various electronic, optical, and magnetic devices. An efficient process for separating lanthanoids is still under study.

Currently, the simplest, most efficient process for separating lanthanoids is column separation using an appropriate stationary phase. The advantage of this process is that it produces a high concentration of lanthanoids. The disadvantage is that an expensive chelating reagent is used as a selective eluent, because a highly selective cation exchanger has not been developed as a stationary phase. Therefore, we have tried to prepare materials that are highly selective toward lanthanoid elements by using surface template polymerization.

Most of the organophosphate-lanthanoid(III) complexes have a nonacoordinate structure, similar to the tricapped trigonal prism coordination in a lanthanoid(III) series (16). Thus, size recognition is required to prepare the surface-templated polymers for the effective separation of lanthanoid elements.

We prepared Dy(III)-imprinted polymers by incorporating the functional host DOLPA by surface template polymerization using water-in-oil emulsions (13). The pH dependence of Dy(III), Ce(III), and La(III) sorption on the Dy(III)-imprinted or unimprinted polymer is shown in Figure 5. The percentage of adsorption was enhanced with increased pH for all lanthanoid ions. However, the imprinted polymer adsorbed Dy(III) much more effectively than Ce(III) and La(III) did over the entire pH range.

Figure 5.

In the adsorption by the unimprinted polymer, much lower selectivity for Dy(III) was observed than with the imprinted polymer. The ion radii of Dy(III), Ce(III), and La(III) in the nonacoordinate structure were 1.083, 1.196, and 1.216 Å, respectively. The selectivity for Dy(III) is controlled by the size of the cavity that the functional host molecules create on the polymer surfaces. However, no imprint effect was observed in the sorption test in which the La(III)-imprinted polymer was used (Figure 6, below). This observation indicates that Ce(III) and Dy(III) can invade the La(III)-fitted cavity and adhere to the recognition sites. This is because their ion radii are smaller than that of La(III) and their affinity with the functional host molecules is naturally higher than that of La(III).

Figure 6.

The imprinted polymer for Ce(III), whose ion radius is midway between those of La(III) and Dy(III), showed intermediate selectivity between that of the other two metal ions: Dy(III) was significantly adsorbed, whereas La(III) adsorption decreased on the Ce(III)-imprinted polymers. This result also supports the effect of ionic size exclusion. When the smallest ion of the three, Dy(III), was imprinted on the polymer surface, La(III) and Ce(III) could not invade the cavity. On the basis of these results, we hypothesize that the improved selectivity for lanthanoid(III) by the surface-imprinted polymers originates from the synergistic effect of a natural affinity for a functional host molecule and size exclusion by the cavity formed on the polymer surface.

Preparation of enantioselective polymers
In addition to highly selective metal-imprinted polymers, we have prepared an enantioselective polymer by incorporating the functional host n-DDPA by surface template polymerization with water-in-oil emulsions (14). The pH dependence on adsorption of D- and L-TrpOMe on the L-TrpOMe-imprinted polymer is shown in Figure 7.

Figure 7.

The percentage of adsorption on the polymers was enhanced with increased pH, indicating that the phosphonic acid groups on the polymers play a predominant role in binding with the amino acid derivative in their ionized phosphate forms. The L-TrpOMe-imprinted polymer prepared in the presence of L-TrpOMe had a higher affinity for L-TrpOMe over D-TrpOMe in the entire pH range. However, the unimprinted polymer prepared similarly but in the absence of L-TrpOMe afforded no evidence of enantiomeric separation in this instance, because the phosphonic groups are randomly distributed on the polymer surface. In addition, the D-TrpOMe-imprinted polymer effectively adsorbed D-TrpOMe over L-TrpOMe. We were surprised to find that the surface-imprinted polymer could recognize three independent domains of chiral molecules. We identified two interactions between n-DDPA and L-TrpOMe as electrostatic interaction (-NH³⁺-^-O=P ) and hydrogen bonding (-NH-O=P) by using ¹H NMR and Fourier transform infrared (FTIR) spectroscopic measurements. The third interaction is assumed to be the hydrophobic interaction between the methoxyl group in the substrate and the polymer matrix, DVB.

New directions
Ordinary imprinting techniques have been limited to template structures that are soluble in organic solvents. However, this novel approach facilitates the use of water-soluble substances and has great potential for the use of imprinted polymers in a range of biological and clinical applications. Furthermore, surface-templated polymers provide high sorption rates for target molecules, because the recognition sites are formed on polymer surfaces. In this technique, the interfacial activity of the functional host is a vital factor in producing high selectivity for the metal ion on the recognition site. It also is an important factor for the firm attachment of the functional host molecule onto the polymer matrix. We prepared highly selective imprinted polymers by using two approaches: the design and synthesis of novel functional host molecules and postirradiation with -rays to make the polymer matrix more rigid.

Recently, the molecular-imprinting technique has been expanded to applications in the field of biomimetics. This novel surface-imprinting technique creates artificial biocatalysts that mimic a variety of enzymes. Using the newly synthesized functional host molecule, oleylimidazole, an enzyme-mimic polymer has been prepared by imprinting a substrate analogue (N--tert-Boc-L-histidine) through the complex formation between a cobalt ion and the imidazole moiety (Figure 8) (15). An oleyl chain was introduced to the imidazole derivative as the functional host molecule to enhance the interfacial activity. The catalytic properties of artificial biocatalysts were investigated by comparing the hydrolysis reaction of an amino acid ester (N-tert-Boc-L-alanine p-nitrophenyl ester) with several control experiments. The imprinted polymer exhibits much higher catalytic activity than the control polymer. These results suggest that complementary specific recognition sites were constructed by the imprinting guest molecule and by the functional host molecules that are specially positioned on the polymer surface. We hope that our molecular surface-imprinting technique for preparing artificial biocatalysts will find useful applications in the future.

Figure 8.

References