Filtration experiments

Since MIM adsorbers are designed to achieve selective separation in flow through mode, we have investigated the kinetics of selective binding during filtration, the breakthrough curves of MIP and NIP membranes and the MIP/NIP selectivity and substance

83 7. In situ synthesis of molecularly imprinted nanoparticles in porous support membranes using

nonsolvating polymerization solvents selectivity in flow through SPE experiments. These tests were carried out on a PH3 T FAP polymer modified membrane.

As a first step, we have investigated the kinetics of template adsorption on the MIP membrane, a factor often overlooked in MIM SPE developments. If the sample is driven too fast through the membrane the analyte cannot equilibrate with the MIP binding sites, therefore the adsorbed amount will depend also on the flow rate.

1 mL 10^-6 M terbutylazine in MeCN was percolated through the composite membrane using different sample throughput times (i.e. different flow rates). The bound amount expressed in percentage can be seen in Figure 7.5.

Figure 7.5 Variation in the bound fraction with sample application time. 1 mL 10^-6 M terbutylazine in MeCN is applied to the PH3 T FAP MIM.

The sorbed amount on the membrane did not change any more if the sample throughput time was 6 minutes or longer. This shows that 6 minutes is sufficient for the equilibrium saturation of the MIP adsorber. It should be noted that using sample application time of one minute, only one third of the saturation value was achieved. The 6 minute sample throughput time corresponds to approximately 170 L min^-1 flow rate i.e. almost 10 bed volumes of sample can be percolated in one minute. In further filtration experiments the sample application time was set to assure equilibrium binding on the membrane.

In a next experiment where the same 1 mL 10^-6 M terbutylazine sample was applied to the membrane in small fractions (200 L aliquots) the breakthrough curve was obtained (see Section 3.3.2.3.). In each percolated fraction the concentration of the template was quantified and plotted as percentage of the initial concentration against the cumulative volume passed (Figure 7.6).

84 7. In situ synthesis of molecularly imprinted nanoparticles in porous support membranes using

nonsolvating polymerization solvents

Figure 7.6 Breakthrough curve of a PH3 T FAP MIM applying 200 L aliquots of 10^-6 M terbutylazine in MeCN

From the figure one can see that the imprinted composite membrane completely adsorbs terbutylazine from the first 200 L percolated volume. The breakthrough volume obtained either from the breakthrough curve by the equal area method or by measuring the retained amount of terbutylazine (see Figure 7.6) was approximately 430 L (c.a. 25 bed volumes). This means that in equilibrium saturation the MIM binds 4.3 10^-10 mole terbutylazine when 10^-6 M solution is applied.

In a similar experiment where 10^-6 M terbutylazine in MeCN was percolated through a nonimprinted PH3 T FAP membrane no solute retention was observed at all. This is again a proof of the strong selective recognition capability of the imprinted membrane.

In SPE filtration experiments the PH3 T FAP MIM was tested against its nonimprinted counterpart and its selectivity was assessed towards other structurally related triazine herbicides (ametryn, prometryn, atrazine) and a nonrelated compound, phenytoin. Aqueous samples were prepared containing a mixture of the triazines in 10^-6 M concentration and phenytoin in 10^-5 M concentration. 1 mL sample was loaded on both the imprinted and nonimprinted polymer membranes. This was followed by two washing steps with 250-250 L MeCN:water=30:70 and an elution step with 2 mL MeOH. This simple protocol was not an optimized procedure, however, earlier studies have shown that the terbutylazine MIP exhibits high selectivity in highly aqueous MeCN:water mixtures¹¹ and terbutylazine can be eluted with MeOH.² During sample application all the analytes were fully retained on the MIP membranes. This is due to the rather hydrophobic nature of MAA-EGDMA polymers. While in the two washing steps only 12% terbutylazine has been eluted from the membrane, already 82% of the nonrelated compound phenytoin has been removed, although it was applied in a tenfold excess. Prometryn was retained in the washing steps almost as well as terbutylazine but 44% ametryn and 52% atrazine has been removed. This shows that the imprinted particles exhibit high selectivity against dissimilar compounds, but recognize structural analogs of the template. The selectivity order is similar to earlier reported ones obtained in elution chromatography on TOL/PO MIP particles.¹¹

The MIP/NIP selectivity of the PH3 T FAP polymer membrane has also been assessed in the SPE experiments. Figure 7.7 shows the recovered fraction of terbutylazine on the imprinted and nonimprinted membrane in each step.

85 7. In situ synthesis of molecularly imprinted nanoparticles in porous support membranes using

nonsolvating polymerization solvents

Figure 7.7 Recoveries in the different steps of the SPE procedure on PH3 T FAP MIP and NIP membranes after the application of 10^-6 M aqueous terbutylazine. (wash 1 and wash 2:

250 L MeCN:water=30:70 mixture, elution: 2 mL MeOH)

Both polymers bound the template completely from the aqueous sample due to hydrophobic interactions. However, in the two aqueous MeCN washing steps 44% of the analyte has been removed from the NIP and only 12% from the MIP. This confirms that the imprinted membrane selectively binds its template even though the SPE conditions have not been optimized.

The above experiments also suggest that the novel microparticle- or nanoparticle-modified MIMs can be used in analytical sample preparation to selectively extract the analyte from the sample - either in organic or in aqueous media. Examples for the first option can be found in food or environmental analysis where solid samples like soil, grain, and baby food are first extracted with an organic solvent and further purification is needed, whereas aqueous samples are often encountered in environmental and bioanalysis.

86 7. In situ synthesis of molecularly imprinted nanoparticles in porous support membranes using

nonsolvating polymerization solvents