Molecular imprinting using precipitation polymerization at high monomer concentrations

MAA/EGDMA based polymer particles prepared in toluene/PO mixtures at high monomer concentrations have been previously imprinted with terbutylazine in our group.¹¹ We aimed to extend the method’s applicability toward basic functional monomers targeting acidic templates. Thus, naproxen and diclofenac imprinted polymers were synthesized using 4-VPy as functional monomer. Methacrylate crosslinkers, EGDMA and TRIM were used because DVB did not afford monodisperse microspheres as discussed above. Two different solvent systems have been utilized; chloroform/PO and toluene/PO, both in a 1:1 ratio.

Low temperature N2 adsorption/desorption isotherm measurements were performed to obtain porosity data about the polymers. For comparison, bulk polymers from 4-VPy and TRIM were also prepared in neat chloroform and toluene with the same monomer composition and the same 3:1 solvent/monomer ratio. Data derived from the isotherms are shown in Table 5.4. An early study of MIPs highlighted that toluene is a good pore forming agent in MAA/EGDMA based polymers while chloroform is not.¹⁴⁴ This is in concordance with our findings that the bulk polymer prepared in toluene had a very high specific surface area (435 m² g^-1) as opposed to the one synthesized in chloroform (17 m² g^-1). All the nonimprinted polymer particles, however, exhibited low BET surfaces similar to that of the bulk polymer synthesized in chloroform, even when toluene was used as co-solvent. This is due to the fact that the microgel particles formed during phase separation preserved their individuality during polymerization. They did not become interconnected or aggregated into a space filling macrogel as in bulk polymerization. The latter would have led to the formation of micro-, meso- or macropores and to a higher surface area.

66 5. MIP microspheres prepared by precipitation polymerization at high monomer loading

If we compare the polymers prepared with different crosslinkers (TRIM or EGDMA), we can observe that TRIM crosslinked MIPs exhibit increased specific surface area in both solvent systems compared to their nonimprinted counterpart whereas polymers prepared with EGDMA do not show this difference. In case of EGDMA crosslinker and chloroform/PO solvent even a somewhat reduced specific surface area was found for the MIP. Also the total pore volume and the volume of micropores are increased in TRIM crosslinked MIPs compared to the NIP but their micropore diameter is reduced in the presence of the template.

Interestingly, this implies that, depending on the type of crosslinker the presence of the template might or might not have an effect on the morphology of the polymer.

Table 5.4 Porosity data of studied polymer compositions Polymer

SBET, m²g^-1

Vtot, cm³g^-1

Vmicro, cm³g^-1

dmicro, Functional nm

monomer Crosslinker Solvent MIP/NIP (template)

4-VPy TRIM CHCl3:PO

MIP

(diclofenac) 105 0.098 0.046 2.41

NIP 38 0.064 0.014 2.63

4-VPy TRIM toluene:PO

MIP

(diclofenac) 68 0.068 0.027 2.11

NIP 17 0.040 0.005 2.70

4-VPy EGDMA CHCl3:PO

MIP

(diclofenac) 17 0.029 0.004 2.75

NIP 28 0.030 0.005 2.60

4-VPy EGDMA toluene:PO

MIP

(diclofenac) 25 0.038 0.009 2.21

NIP 24 0.038 0.009 2.09

4-VPy EGDMA toluene:PO

MIP

(naproxen) 24 0.040 0.009 2.19

NIP 24 0.038 0.009 2.09

4-VPy TRIM toluene

(bulk) NIP 435 0.760 0.166 2.08

4-VPy TRIM CHCl₃

(bulk) NIP 17 0.039 0.002 3.37 Template binding properties of the different polymers have been assessed in equilibrium batch rebinding measurements in 100 µM toluene solution of diclofenac or naproxen and characterized by calculating their distribution coefficients (see Section 3.2.4).

Figure 5.11 presents the obtained results for the MIPs in comparison with their nonimprinted counterpart. Distribution coefficient ratios for MIP and NIP were calculated and shown in the figure to characterize imprinting efficiency. Except for the diclofenac templated polymer synthesized in toluene/PO considerable differences between MIP and NIP could be observed in each case. For naproxen the toluene/PO, while for diclofenac the chloroform/PO solvent medium provided better imprinting efficiency. EGDMA crosslinker was proven to be superior over TRIM in both cases. We must note here though, that the molar ratio of the functional monomer and the crosslinker has not been optimized and was set to the same 1:5 value in both cases. However, TRIM being a trifunctional crosslinker, can be efficient in much lower concentrations, therefore it is reasonable to expect that in a lower molar ratio it could have led to improved imprinting efficiency.

67 5. MIP microspheres prepared by precipitation polymerization at high monomer loading

Figure 5.11 Distribution coefficients on naproxen (A) and diclofenac (B) imprinted polymers in 100 µM toluene solution of the template using a phase ratio of 60. The ratio of the distribution coefficients on MIP and its respective NIP is indicated above the column pairs.

The selectivity of the diclofenac MIP was further studied in comparison with other structurally similar acidic non-steroidal anti-inflammatory drugs, ketoprofen, naproxen and ibuprofen and a non-related basic β-blocker drug, propranolol in equilibrium batch rebinding experiments. These were performed both in toluene and in MeCN. The distribution coefficients are plotted in Figure 5.12 together with the structure and logP values of the studied compounds. In the apolar toluene (Figure 5.12A) more hydrophilic compounds with lower logP values like ketoprofen and naproxen exhibit high distribution coefficients both on the MIP and the NIP, that is, the polymer behaves like a normal stationary phase. Ibuprofen being more hydrophobic shows much lower binding in toluene. Acid-base interactions also prevail between the 4-VPy-based polymers and the acidic drugs since propranolol, a basic compound, with similar hydrophilicity as ketoprofen and naproxen exhibits only very weak binding on the sorbent. Diclofenac, being the most hydrophobic of all, has almost as high distribution coefficient on the MIP as ketoprofen but low binding on the NIP, indicating successful imprinting. In the polar MeCN (Figure 5.12B) only the template, diclofenac shows high binding on the MIP. This interaction is predominantly due to the imprinted selective sites and not only to nonspecific hydrophobic binding since the NIP binds much less diclofenac.

As the specific surface area of the MIP is even somewhat smaller than that of the NIP (see Table 5.4), we can be assured that the enhanced binding on the MIP cannot be attributed to an increased specific surface area.

Figure 5.12 Selectivity of diclofenac imprinted 4-VPy/EGDMA polymer prepared in chloroform/PO. Batch rebinding measurements were carried out in 100M toluene (A) and acetonitrile (B) solution of the respective compound with a phase ratio of 60. The ratio of the

distribution coefficients on MIP and its respective NIP is indicated above the column pairs.

68 5. MIP microspheres prepared by precipitation polymerization at high monomer loading