High-Throughput Synthesis and Screening of a Library of Random and Gradient Copoly(2-oxazoline)s

Abstract

ReceiVed July 8, 2005 Introduction. The living cationic ring opening polymerization (CROP; see Scheme 1) of 2-oxazolines was discovered in 1966.1-3 After this discovery, the resulting biocompatible poly(2-oxazoline)s have been used for a wide variety of applications.4-6 Moreover, amphiphilic structures are easily accessible by the sequential polymerization of a hydrophilic monomer (e.g., 2-methylor 2-ethyl-2-oxazoline) and a hydrophobic monomer (e.g., 2-nonylor 2-phenyl-2oxazoline). Unlike block copoly(2-oxazoline)s, random copoly(2-oxazoline)s have not been extensively investigated so far. The possibility of copolymerizing different 2-oxazolines has been demonstrated by several groups;7-9 however, reactivity ratios have only been determined for copolymerizations involving 2-phenyl-2-oxazoline (PhOx).10 Due to the low reactivity of PhOx, the copolymerizations resulted in blocky structures instead of random or gradient copolymers. Moreover, the properties of copoly(2-oxazoline)s have not been studied in detail. Inspired by this lack of knowledge in the literature, we decided to investigate the synthesis and properties of a library of random copolymers from 2-methyl-2-oxazoline (MeOx), 2-ethyl-2-oxazoline (EtOx), and 2-nonyl-2-oxazoline (NonOx). Such systematical copolymerization studies and corresponding structure-property investigations are reported for the first time. In polymer science, this report is the first example of a high-throughput approach in which kinetic details, library synthesis, and property screening were investigated simultaneously, providing both fundamental and application-directed insights. Results and Discussion. The currently described work was preceded by detailed (automated parallel) kinetic investigation of the homopolymerizations of four monomers (MeOx, EtOx, NonOx, and PhOx) with four initiators at four molecular weights and two temperatures.11 The optimal conditions from this kinetic screening (polymerization with benzyl bromide as initiator at 100 °C in N,N-dimethylacetamide) were applied to synthesize copolymers from combinations of MeOx, EtOx, and NonOx utilizing a Chemspeed ASW2000 synthesis robot.12 For each combination, nine copolymers were synthesized with 0-100 mol % (steps of 12.5 mol %) of the second monomer, resulting in 27 parallel polymerizations. For each of the copolymerizations, the monomer conversion in time was investigated by automated sampling from the polymerization mixtures for GC analysis.13 The measured monomer-to-solvent ratios of the polymerization mixtures before heating (Figure 1) clearly demonstrate the gradual change in monomer composition for the different copolymerization series: With increasing mole fraction of the second monomer, the ratio of the second monomer to solvent increases linearly, and the ratio of the first monomer to solvent decreases linearly. Figure 2 (top row) depicts the resulting kinetic plots in time for the 50 mol % copolymerizations. Similar kinetic plots were obtained for all other copolymerizations (only one of the polymerizations failed during the library synthesis cq. kinetic screening). The linearity of the first-order kinetic plots (ln{[M]0/[M]t}) confirmed a constant concentration of living polymer chains as expected for a living polymerization. Moreover, monomodal GPC traces with narrow molecular weight distributions (PDI < 1.3) were obtained for the endsamples, which is indicative of a living polymerization mechanism, as well. The kinetic plots revealed a slightly faster consumption (higher reactivity) of MeOx, as compared to EtOx and NonOx. To further elucidate the copolymer compositions, the reactivity ratios (r1 ) k11/k12 and r2 ) k22/k21) were determined from the relation between the fraction of monomer A in the monomer feed (f1) and the incorporated fraction of monomer A in the polymer (F1) at both ∼20% and ∼60% monomer conversion (Figure 3.15, bottom row). For living polymerizations, the reactivity ratios should be calculated at higher monomer conversions (20% or higher) since different reactivities during the initiation process are commonly observed.15,16 Also for the polymerization of 2-oxazolines, it is well-known that the polymerization rate during initiation can be different from the final polymerization rate.17 In the first instance, the reactivity ratios were determined using the classical Mayo-Lewis terminal model (MLTM) utilizing nonlinear least-squares fitting of the data in Figure 2, bottom.18 However, the applied MLMT method is strictly only valid for 0% monomer conversion, and therefore, reliable reactivity ratios can only be determined from extrapolation of the data obtained at low monomer conversions up to ∼10%. To accurately determine the * Phone: +31 40 247 5303. Fax: +31 40 247 4186. E-mail: u.s.schubert@tue.nl. Scheme 1. Schematic Representation of the CROP of 2-Oxazolines Initiated with Benzyl Bromide 145 J. Comb. Chem. 2006, 8, 145-148