A Pyrrolidine Functionalized Poly[(ethylene glycol) Methacrylate] Resin as a Heterogeneous Catalyst for Aqueous Aldol Reactions

: The development of a performant aminated catalyst for aldol condensations requires the combined tuning of the active site, support and solvent system. For this purpose, a pyrrolidine group was immobilized on a swellable polymer resin. Favorable interactions between the support and water (in its role as solvent) resulted in a turnover frequency (TOF) amounting to 3.0 ± 1.5 × 10 − 3 s − 1 , despite potential inhibition of the active sites by formation of iminium species. The affinity of the solvent for the poly[(ethylene glycol) methacrylate] support resulted in efficient swelling of the catalytic material, which was shown to be key to the observed catalytic performance.


Introduction
In the past two decades, amines have gained attention as catalysts for the aldol reaction, an important carbon-carbon coupling reaction [1]. Inspired by naturally occurring enzymes, i.e., aldolase enzymes [2,3], the excellent catalytic performance of amines was first demonstrated with the amino acid proline [2,[4][5][6]. The excellent performance of proline, offering high activity and (stereo)selectivity in the aldol condensation at relatively mild conditions for a wide range of substrates [2,7,8], preluded the promising field of organocatalysis [9]. In recognition of his pioneering work on asymmetric organocatalysis, Benjamin List was awarded the 2021 Nobel Prize in Chemistry together with David W.C. Macmillan who performed similar work on the Diels-Alder reaction [10].
List et al. [7] established that the amine group in proline acts as a Lewis base active site, catalyzing the aldol condensation via an enamine reaction mechanism [2,7], while interacting cooperatively with the adjacent carboxyl group [7,11]. Proline and proline derivatives remain, however, homogeneous catalysts, exhibiting disadvantages with respect to catalyst recovery and reusability. Therefore, more sustainable heterogeneous alternatives were investigated, in particular aminated silicas and organosilicas [1]. Several key factors determining the overall catalytic performance have, in the meantime, been identified [1].

-
As active sites, primary and secondary amines provide high activity, especially in the presence of promoting groups such as hydroxyl groups [12][13][14][15][16][17]. -Additionally, the presence of (small amounts of) water was found to be favorable, both with respect to catalytic activity as well as stability [1,18,19]. In addition to these benefits, water is a widely available and environmentally benign (co-)solvent [20][21][22]. - In an aqueous environment, however, silica and organosilica supports exhibit poor stability as their siloxane bonds are prone to hydrolysis, causing structural degradation and leaching of active sites [1,19,23].
Several studies reported on the successful application of aminated polymers and polymer resins as aldol condensation catalysts [1,[24][25][26][27][28][29][30][31]. As support material, the polymer backbone can offer sufficient tolerance against problematic degradation reactions, as such contributing to enhanced catalytic stability and reusability. In particular, a swellable polymer resin consisting of poly[(ethylene glycol) methacrylate] (abbreviated PEGMA) [32] seems to exhibit properties rendering it suitable as a catalyst support [28]. As a lightly crosslinked polymer resin (with (ethylene glycol) dimethacrylate as crosslinker), the PEGMA support can swell in a range of solvents without dissolving. Furthermore, the support offers the required thermal and chemical stability for the anticipated application [28,32]. The polymeric backbone ensures tolerance to hydrolysis, and the terminal hydroxyl groups provide options for chemical modification or fulfill a promoting role as a hydrogen bond donor. In 2020, for the first time, a stable operation in an aqueous environment at an appreciable catalytic activity was obtained with a primary amine functionalized PEGMA resin, demonstrating the feasibility of such supports for aldol reaction catalysts [28]. A PEGMA supported primary amine catalyst was demonstrated to exhibit excellent stability for at least 8 h on stream in a continuous flow setup [28].
Via computational probing of electronic as well as steric effects, De Vylder et al. [33] identified a pyrrolidine function with a potentially higher performing functionality compared to the primary amine referred to above. This cyclic secondary amine combines a high nucleophilicity with minimal steric constraints [33], and was already positively assessed as a homogeneous catalyst in aqueous aldol condensation by Singh and Chimni [34]. Immobilization of pyrrolidine on a suitable solid support, hence, brings promising perspectives for establishing a highly active heterogeneous aldol condensation catalyst. Furthermore, key determining factors of the catalytic performance such as cooperative interactions with promoting groups, or solvent and support effects [1], were not taken into account in the computational study by De Vylder et al. [33] and deserve experimental validation. Therefore, in this work, pyrrolidine groups were covalently linked to the PEGMA polymer resin. The catalytic activity of the resulting aldol reaction catalyst was assessed experimentally in an aqueous as well as a non-polar environment, taking into account solvent-support interactions.

Catalyst Characterization
A series of 3-(aminomethyl)pyrrolidine functionalized PEGMA resin catalysts with varying amine loadings were synthesized. After each synthesis step (see Scheme 1), the obtained material was characterized by SEM, FTIR and 13 C solid-state NMR.
The cross-linked poly[(ethylene glycol) methacrylate] support (PEGMA) was prepared via suspension polymerization (Scheme 1, step 1). This polymerization step yielded spherical polymer beads with a diameter ranging between 10 and 200 µm, as is visible in the SEM image of the PEGMA resin (see Figure A1, Appendix A). This is in agreement with the particle size ranges reported by Tuncel et al. (40-200 µm) [32] and De Vylder et al.
(around 100 µm) [28]. The particle size distribution is provided in Figure A5, Appendix B. In the subsequent functionalization steps, SEM images indicate partial fragmentation of the PEGMA-based materials, probably due to mechanical stresses during the synthesis procedure ( Figures A2-A4, Appendix A). However, as fragmentation is considered to be a mechanical process, no effect on the catalyst performance was expected. The 13 C NMR spectrum of the PEGMA support presented Figure 1, reveals the characteristic 13 C resonances associated with the ethylene glycol and the methacrylate moieties. The presence of broad resonances at 17.8, 19.7 and 45.8 ppm corresponding to the methyl groups and methylene linkages of the polymerized acrylate, confirmed the successful synthesis of the support material. This is further supported by the FTIR spectrum presented in Figure 2, as the most intense peaks correspond to the characteristic peaks of the -O-H, -C-H, -C=O and -C-O groups within the support material, appearing around 3500 cm −1 , 2900 cm −1 , 1720 cm −1 and 1100 cm −1 , respectively. However, the presence of a small fraction of non-polymerized acrylic acid terminal groups, was evidenced in the 13 C NMR spectrum by the presence of narrow resonances at 146 ppm and 14.8 ppm, corresponding to terminal methylene and methyl groups, respectively.  After the polymerization step, a partial substitution of the hydroxyl groups by Cl groups resulted in the partly chlorinated PEGMA material, denoted as PEGMA-Cl (Scheme 1, step 2). In the FTIR spectrum of the PEGMA-Cl material, a strong peak appeared at 660 cm −1 and the 740 cm −1 peak intensified, see Figure 2. Both peaks are related to the -C-Cl stretching vibration, confirming the partial substitution of hydroxyl groups by Cl-groups. Effective functionalization with Cl-groups is furthermore supported by the 13 C NMR spectrum (red trace, Figures 3, A6 and A7, Appendix C), by the appearance of the resonance at 44.2 ppm corresponding to the -CH 2 Cl and the decrease in area of the 13 C peak at 62 ppm comprising the resonance of terminal -CH 2 OH groups. From the elemental composition of PEGMA-Cl (see Table 1), it follows that a major fraction of hydroxyl groups (>55%) remains unsubstituted, which serve as promoting groups in the resulting catalyst [13].
After substitution of a part of the Cl-groups by 1-Boc-3-(aminomethyl)pyrrolidine yielding the PEGMA-AMP-Boc material (Scheme 1, step 3), the 13 C NMR spectrum (green trace, Figures 3, A6 and A7, Appendix C) revealed a reduction in the area of the resonance of -CH 2 Cl at 44.2 ppm, and the appearance of resonances corresponding to the diamine and the protective Boc moiety, confirming successful functionalization. Only a fraction (4-24%, see Table 1) of the -CH 2 Cl groups underwent functionalization, as a residual -CH 2 Cl resonance was still present after functionalization. The characteristic peaks of the carbons in the pyrrolidine ring were observed at 55.5, 52.7, 50.8, 38.5 and 29.8 ppm. The intensity of the diamine resonances was, however, low due to the rather low active site loading, which according to CHN analysis, varied for three different batches between 0.054 ± 0.008 mmol·g −1 to 0.270 ± 0.014 mmol·g −1 (see Table 1). For PEGMA-AMP-3, the amine loading was validated by 13 C NMR as well, i.e., based on comparison of the integrated area per mass of the -CH2Cl moiety, a functionalization ratio of 24% was obtained. This agrees well with the elemental analysis data (see Table 1), showing a reduction in chloride content of 23% for PEGMA-AMP-3. The remaining hydroxyl groups on the support, may serve as promoting sites in the reaction [13]. With FTIR, the presence of amine groups could not be directly demonstrated due to the low intensity of characteristic peaks corresponding to C-N and N-H vibrations or overlap with support-related peaks. Nonetheless, an intensification at 2870 cm −1 was discerned (see Figure 2) corresponding to the -C-H stretching vibration within the pyrrolidine ring. Furthermore, the chlorine-related peaks at 660 cm −1 and 740 cm −1 significantly decreased, which supports the partial substitution of Cl groups by diamines.
Deprotection of the 3-(aminomethyl)pyrrolidine (AMP) groups was achieved by an acid treatment (Scheme 1, step 4), yielding the pyrrolidine functionalized resin catalyst, denoted as PEGMA-AMP. The presence of Boc groups in the PEGMA-AMP-Boc ( Figure 3, green trace) sample and its removal in the PEGMA-AMP ( Figure 3, purple trace) sample, was demonstrated by the respective appearance and disappearance of characteristic 13 C resonances of the Boc moiety around 154.7, 79.0 and 29.2 ppm (see Figures 3 and A7). CHN analysis confirmed no loss of amine groups during the Boc removal step. An upfield shift of the pyrrolidine resonances was observed upon the removal of the Boc with the electron-withdrawing carbonyl carbon group, also confirming their removal (see Figure 3). Hence, it can be concluded that the targeted 3-(aminomethyl)pyrrolidine (AMP) active sites were immobilized on the PEGMA support, even though there remains room for improving the amine substitution efficiency.
As a polymer resin, the affinity of the reaction environment, more particularly the solvent, for the PEGMA support is important for effective swelling of the resin and, hence, ensuring the accessibility to the catalytic sites inside of the resin [17]. To obtain insights into the support-solvent interactions, the swelling ratio S of the PEGMA resin, defined as the volume of the wet resin divided by that of the dry resin, was assessed in different solvents and solvent mixtures, see Table 2. The highest swelling ratio was obtained in pure water with the swollen resin roughly tripling in volume. With a volume increase of 80%, the PEGMA resin also swelled significantly in acetone. Due to the high affinity of water for the PEGMA resin, its swelling ratio in a 50/50 vol% acetone/water mixture was also very high (swelling ratio S = 2.9 ± 0.3), indicating that the resin absorbed a relatively large fraction of water under such conditions. In n-hexane practically no swelling was observed due to its low affinity for the PEG-based support [35,36], while in a 50/50 vol% acetone/n-hexane mixture the material swelled only to a limited extent (S = 1.4 ± 0.3). Table 2. Swelling ratio S of the PEGMA polymer resin determined in different solvents and solvent mixtures.

Catalyst Performance Assessment
The catalytic activity of the PEGMA-AMP catalyst was assessed in the model aldol reaction of 4-nitrobenzaldehyde with acetone at 55 • C in an aqueous environment (50/50 vol% acetone/water) in a batch reactor. Acetone was used as an excess reagent to avoid consecutive condensation reactions of a single acetone molecule with two 4-nitrobenzaldehdyde molecules, as well as the formation of an inhibiting iminium ion from 4-nitrobenzaldehyde, which is stabilized in a polar environment, e.g., in aqueous, environment [1,18,19,[37][38][39]. As such, acetone served both as a reagent as well as a solvent. In Figure 4, the conversion profile is presented as a function of batch time (i.e., the product of the molar amount active sites and the reaction time), allowing comparison of catalytic performance between different catalysts. A linearly increasing conversion profile was obtained, from which the turnover frequency (TOF) was determined. The main reaction product was the aldol species (see Scheme 2), formed with high selectivity, i.e., 97.7%. The secondary enone product, formed by the dehydration of the aldol product, was formed with only 2.3% selectivity. Although organocatalysts generally comply with stereoselective execution of the reaction [40], our present work focused on an optimal combination of a nucleophilic amine with minimal steric constraints, not considering stereoselectivity. The aldol and enone selectivity profiles can be found in Figure A8, Appendix D. The high aldol product selectivity could be anticipated as the two reaction products are typically in equilibrium, which in an aqueous environment is situated at the side of the aldol product [19]. The PEGMA-AMP catalyst exhibited high activity, corresponding with a TOF amounting to 3.0 ± 1.5 × 10 −3 s −1 . The reaction rate furthermore remained constant, as the conversion profile increased linearly up to conversions as high as 60%, before levelling off (see Figure 4). This indicates that the rate determining step was not strongly affected by the 4-nitrobenzaldehyde concentration and that the concentration of the enamine intermediate on the catalyst surface remained fairly constant due to the excess of acetone.
As the pyrrolidine active sites are immobilized by partial substitution of hydroxyl groups, the number of promoting hydroxyl groups per active site decreases with increasing loading. Typically, it can be expected that if the ratio of promoting groups per active site becomes too low, insufficient promotion is provided, which negatively affects the catalytic activity [13,41]. However, here, similar activities were obtained for both low-and high-loaded materials, which implies that within the investigated range, the amine active sites are sufficiently promoted by the hydroxyl groups and, hence, loading effects can be excluded.
When performing the reaction with the PEGMA-Cl material or HCl, only 0.1% conversion was obtained after 3 h, compared to 6.6% with PEGMA-AMP. Hence, unsubstituted Cl-groups or amine salts with chlorine that remain on the PEGMA-AMP catalyst are not expected to contribute significantly to its (high) activity. Note that De Vylder et al. [28] have already demonstrated that the amine group serving as an anchoring point for the pyrrolidine group is not catalytically active, due to excessive steric hinderance [28]. Hence, only the secondary amine within the pyrrolidine group is considered to be catalytically active.  As the investigated reaction of 4-nitrobenzaldehyde with acetone is a widely adopted model reaction in aldol condensation research, it is possible to compare the performance of PEGMA-AMP to that of other catalysts for which activity data are available, see Table 3. A comparison to proline-based catalysts is not feasible due to the lack of relevant activity data. Nonetheless, proline is generally considered to perform poorly in aqueous environment [8,42,43], although there are indications that supplementing the reaction environment with smaller amounts of water can lead to optimization [8,[43][44][45][46]. In comparison to a PEGMA-supported methyl-substituted secondary amine tested under identical conditions by De Vylder et al. [28], PEGMA-AMP exhibited a 10 times higher activity (see Table 3, entry 1-2). To a certain extent, this improvement could be due to the higher base strength of the pyrrolidine site (pK a > 11) [47] as compared to the methyl-substituted secondary amine site (pK a ≈ 10) [47]. After all, amines, in their role of nucleophilic sites, activate the carbonyl compound by engaging in a reversible bond [9]. However, steric effects should not be excluded as they are known to significantly affect the catalyst performance, even with relatively small substituents [11,14,15,33]. For the AMP site, steric hindrance is minimized due to the pyrrolidine structure, while the methyl substituent of the other secondary amine constitutes a small, though nonnegligible burden [33]. Table 3. Overview of batch screening conditions and obtained turnover frequencies (TOF) exhibited by the PEGMA-AMP catalyst, compared to PEGMA-supported methyl-substituted secondary and primary amines as reported by De Vylder et al. [28], and a silica-supported methyl-substituted secondary amine as reported by Lauwaert et al. [15]. The solvent is applied in a 50/50 vol% mixture with acetone. Strikingly, the PEGMA-AMP catalyst also provided a five-fold increase in activity compared to a PEGMA-supported primary amine, tested under identical conditions by De Vylder et al. [28], see Table 3, entry 3. In aqueous environment, the activity of secondary amines is generally inhibited by formation of an iminium ion, i.e., a charged species that is stabilized by the polar protic solvent [1,13,18]. Conversely, the activity of primary amines is not impaired in aqueous environments, as the analogous, inhibiting species on primary amines is a neutral imine that is not stabilized in polar protic solvents [1,18,37]. Moreover, swelling experiments indicated that the PEGMA support had a higher affinity for water than for acetone (see above). A high affinity of the support for specific species in the reaction mixture has already been demonstrated to lead to local enrichment of this species inside the catalyst, creating a local environment that is different from the bulk aqueous solution, which in turn may affect the catalyst performance [30,38,[48][49][50]. The relatively high swelling ratio in a 50/50 vol% acetone/water mixture (see above) suggests that in the reaction environment, PEGMAsupported catalysts will be enriched with water and the amine active sites are exposed to a more pronouncedly polar environment. Even under these conditions, the PEGMA-AMP catalyst exhibited a high activity with a TOF of 3.0 ± 1.5 × 10 −3 s −1 , which suggests that the intrinsic activity of the AMP site, i.e., in the absence of inhibition, might be even higher.

Entry
In a nonpolar environment, such inhibition phenomena are suppressed, which could potentially enhance the catalytic performance. It would therefore be advantageous to apply the PEGMA-AMP catalyst in a non-polar solvent. In a 50/50 vol% acetone/n-hexane mixture, however, PEGMA-AMP provided very low activity with a TOF amounting to only 6.0 ± 1.1 × 10 −6 s −1 , which is a factor 500 lower than in the aqueous environment. Given the low swelling ratio of the PEGMA support in n-hexane (see above), the poor catalytic activity is predominantly related to the poor swelling behavior of the catalyst in the acetone/nhexane mixture. Secondary amines, which are not subjected to extensive steric hindrance, have already been demonstrated to exhibit high activities in a non-polar environment; e.g., the best performing aminated aldol reaction catalyst reported in the literature, methylsubstituted secondary amine on silica, was tested under similar conditions (albeit at a lower temperature) [15], see Table 3, entry 5. A poor swelling behavior, however, compromises both the accessibility of the amine sites [17], as well as interactions between neighboring active and promoting sites [51,52]. The current PEGMA support had a higher affinity for polar solvents, such as water and acetone, while being virtually incompatible with nonpolar solvents, such as n-hexane [35,36]. Improvement of the catalytic performance could, hence, be obtained by further tuning of the support properties or selection of a compatible nonpolar solvent, such that effective swelling is guaranteed.
Finally, the Boc-protecting groups were removed by washing for 2 h in 80 mL of a concentrated HCl solution (4 M HCl in dioxane, Sigma Aldrich, Germany) [53,54] (Scheme 1, step 4). After filtering over a glass filter (Por 4), and washing twice in a large excess of dry ethanol, the material was dried overnight under vacuum at 50 • C, resulting in the final catalyst, denoted as PEGMA-AMP.

Catalyst Characterization
Scanning electron microscopy (SEM) was performed on a SEM JEOL JSM 5400 (JEOL Europe B.V., Zaventem, Belgium) at 20 kV and was used to verify the structure of the polymer particles and assess the particle size range. Polymer samples were spread on carbon tape and sputter coated with a layer of gold (<5 nm) in order to avoid charge buildup. The size of approximately 100 polymer beads was determined on 6 SEM micrographs for determining a particle size distribution.
Further validation of the synthesis procedure was performed by means of Fourier transform infrared (FTIR) spectroscopy and solid-state 13 C NMR. To ensure sufficient resolution, the material with the highest loading of functional groups was used. FTIR was performed on a Brüker Tensor 27 with an MCT detector (Brüker Optik GmbH, Ettlingen, Germany).
Measurements were performed in transmission mode at wavelengths between 400 cm −1 and 4000 cm −1 for PEGMA, PEGMA-Cl, PEGMA-AMP-Boc-3 and PEGMA-AMP-3. The samples were prepared as a sandwich structure in a mixture with KBr.
Solid-state 13 C MAS NMR was performed on a Bruker 800 MHz Avance Neo spectrometer (Bruker Belgium NV, Kontich, Belgium) equipped with a 1 H/ 2 H/ 13 C/ 15 N highresolution magic angle spinning probe. The samples were packed in a 4 mm ZrO 2 rotor and spun at 15 kHz MAS speed. 1 H decoupled 13 C NMR spectra (Waltz64 decoupling) were acquired with a π/2 pulse at 20 kHz RF strength and a recycle delay of 40 s. 512 transients were recorded for PEGMA, PEGMA-Cl and 4480 transients for PEGMA-AMP-Boc-3 and PEGMA-AMP-3.
The amine site loading and the C, H and N mass percentage were determined by elemental (CHN) analysis. The measurements were performed on a Thermo Scientific Flash 2000 Series elemental analyzer (Interscience, Ottignies-Louvain-la-Neuve, Belgium) using V 2 O 5 as catalyst. O and Cl mass percentages were derived from the C, H and N mass percentages assuming a closing mass balance.
For PEGMA-AMP-3, the amine loading was also validated by the 13 C-SS-NMR spectra. The solid-state NMR spectra were acquired for PEGMA-AMP-3 obeying a quantitative methodology [55,56] (the same coil filling factor, similar Q-factor and optimized relaxation delays). The quantification of -CH 2 Cl moieties remaining after AMP functionalization relative to the PEGMA-Cl was performed via spectral decomposition using DMFIT software [57].
The swelling ratio of the resin in the presence of a solvent was determined as the volume of the wet resin divided by that of the dry resin. Therefore, about 300 mg of the material was submersed in either 10 mL of a pure solvent or 10 mL of a solvent mixture, and allowed to swell for 60 min. As a solvent, either water (18.20 MΩ cm double deionized), acetone (99.6%, Acros organics, Spain) or n-hexane (99%, Chem-lab, Belgium) were used. The swollen material was filtered off over filter paper (MN 640 d). Subsequently, the wet resin was dried at 60 • C until no further mass loss was observed, i.e., for at least 24 h, and the volume of the absorbed solvent (V s ) was derived from the mass difference before and after drying. The swelling ratio S was then determined according to Equation (1). Based on repeat experiments, the error on the measurements was determined as 1.96 times the standard deviation, corresponding to the 95% confidence interval.
where V s is the volume of the swollen resin and V d the volume of the dry resin.

Catalyst Performance Assessment
The catalytic activity was evaluated in the aldol reaction of 4-nitrobenzaldehyde (99%, Acros organics, Shanghai, China) with acetone (99.6%, Acros organics, Spain). The reaction was performed in a batch-type reactor (Parr ® 4560 mini batch reactor, 300 mL, Parr Instrument Company, Moline, IL, USA) at 55 • C and stirring speed ±300 rpm, with 0.45 g 4-nitrobenzadehyde, 45 g acetone and a solvent, either 55 g water (18.20 MΩ cm double deionized) or 38 g n-hexane (99%, Chem-lab, Belgium). 0.25 g methyl-4-nitrobenzoate (≥99%, Sigma Aldrich, United Kingdom) was used as internal standard. For catalytic tests, 0.15 g of PEGMA-AMP was added to the reactor. The effect of the remaining Cl groups was assessed by adding either 0.26 g of PEGMA-Cl or 0.2 mL of a 4 M HCl/dioxane solution to the reactor. The reaction was monitored for 3 h by taking 8 samples (300 µL) of the reaction mixture which were analyzed by liquid chromatography performed on an Agilent 1100 reversed-phase high-performance liquid chromatography (HPLC, Agilent Technologies, Santa Clara, CA, USA). The catalyst activity was assessed by the turnover frequency (TOF), determined from the slope of the initial conversion, corresponding to differential conditions, as a function of the batch time. The latter was defined as the product of two main variables governing the reaction, namely the reaction time and the number of amine active sites present in the reactor. Repeat experiments were performed to determine the error of the measurements defined as the 95% confidence interval (1.96 times the standard deviation).

Conclusions & Perspectives
A poly[(ethylene glycol) methacrylate] (PEGMA) resin was successfully functionalized with 3-(aminomethyl)pyrrolidine (AMP). The resulting PEGMA-AMP catalyst was demonstrated to exhibit high activity in the model aldol reaction of 4-nitrobenzaldehyde with acetone at 55 • C in a 50/50 vol% mixture of acetone/water, attaining an average TOF of 3.0 ± 1.5·10 −3 s −1 . The adequate performance of the catalyst is attributed to the selection of an amine active site with both high alkalinity and minimal steric hindrance, coupled to a swellable polymer resin support comprising promoting groups. Applying the catalyst in a nonpolar environment, which would enhance the stability of the pyrrolidine active site, resulted in a low catalytic activity, due to a poor swelling of the PEGMA support. Hence, further tuning of the support as a function of the solvent used may enhance the catalyst's performance. In addition, applying other substrates and ketones, while also taking into account the enrichment and swelling effects, could broaden the scope of the catalyst.    Appendix C. Additional 13 C NMR Spectra Figure A6. 13 C NMR spectra of (from bottom to top) the PEGMA support, the chlorine-functionalized PEGMA-Cl, the protected amine-functionalized PEGMA-AMP-Boc-3 and the aminated PEGMA-AMP-3. The spectrum is scaled vertically to magnify the diamine resonances.

Appendix D. Selectivity Profiles
Aldol condensations typically yield two main products (see Scheme 2). In the aldol reaction, the so-called aldol product (a β-hydroxy aldehyde or β-hydroxy ketone) is formed as the result of the coupling of the two reactants [1]. Dehydration of this aldol yields the so-called enone product (α,β-unsaturated carbonyl compound) [1].
In the aldol condensation of 4-nitrobenzaldehyde with acetone over PEGMA-AMP as catalyst in aqueous environment, the aldol product is formed as the main product, with a selectivity amounting to 97.7 ± 0.6%. The enone product is the secondary product, with a selectivity amounting to 2.3 ± 0.6%. As presented in Figure A8, initially only the aldol product is formed. Soon, however, the system evolves towards the aforementioned distribution and from 15% conversion onwards, the aldol product and enone product are in equilibrium while the conversion remains increasing till > 85%. Figure A8. Selectivity profile of the aldol product (N) and enone product (N) in the aldol condensation of 4-nitrobenzaldehyde with acetone at 55 • C in a 50/50 vol% mixture of acetone/water with PEGMA-AMP-3 as catalyst. Error bars represent the 95% confidence interval.
In the aldol condensation of 4-nitrobenzaldehyde with acetone in a 50/50 vol% mixture of acetone/n-hexane, a similar trend was observed, though the selectivity for the aldol product was lower (91 ± 1.7%) and for the enone product was higher (9.2 ± 1.7%). This can be attributed to the low water concentration in the reaction environment. Nonetheless, it should be pointed out that due to the high affinity of water for the PEGMA support, the scarce water that was present (formed in the dehydration step) will preferentially remain absorbed by the catalyst instead of dissolving into the bulk liquid. The catalyst can get enriched with water, which will then affect the equilibrium between aldol and enone.
Hence, the dehydration of the aldol product forming the enone product and one water molecule should be regarded as reversible, at least from a thermodynamic point of view, not necessarily from a kinetic point of view, and the reaction environment can significantly affect the equilibrium between aldol and enone.