Solvent Effects on Kinetics of 2-Hydroxyethyl Methacrylate Semibatch

Article pubs.acs.org/IECR

Solvent Effects on Kinetics of 2‑Hydroxyethyl Methacrylate Semibatch Radical Copolymerization Kun Liang, Thomas R. Rooney, and Robin A. Hutchinson* Department of Chemical Engineering, Queen’s University, Dupuis Hall, Kingston, Ontario K7L 3N6, Canada ABSTRACT: Radical copolymerizations of 2-hydroxyethyl methacrylate (HEMA) with n-butyl methacrylate (BMA) and n-butyl acrylate (BA) were carried out in xylene, DMF, and n-butanol solutions. Solvent effects on copolymerization propagation kinetics were investigated using pulsed laser polymerization (PLP) combined with size exclusion chromatography (SEC) as well as proton NMR, while starved-feed higher temperature semibatch reactions were carried out in different solutions to simulate industrial production. Solvent choice, through its influence on hydrogen bonding of HEMA monomer, has a significant impact on the copolymer composition and propagation rate coefficient and thus influences the semibatch polymerization behavior of BMA/HEMA, as previously found for the styrene/HEMA system. The presence of HEMA leads to increased polymer molecular weight, a result attributed to branching reactions involving dimethacrylate impurity. Although H-bonding (and solvent choice) influences BA/HEMA kinetics, its relative effect is negligible on semibatch operation under the conditions studied.

■

INTRODUCTION Acrylic resins used as polymeric binders in solvent-borne automotive coatings are complex copolymers containing reactive functional groups (often hydroxyl). A starved-feed semibatch process is usually employed to produce these copolymers at high temperature (>120 °C) in solution via free radical copolymerization. Good reactor control is required to ensure that the copolymer composition and distribution of functional groups is uniform among the copolymer chains produced over the course of the batch. This uniformity is required, as the functional groups on the low molecular weight (MW) chains react on the surface of the vehicle, cross-linking to form the final high MW coating.1−4 Thus, a good understanding of copolymer chain-growth kinetics under these higher-temperature conditions is required.5 A comprehensive model has been developed to describe the kinetic complexities of acrylates, methacrylates, and styrene copolymerizations at high temperatures.6,7 The modeling strategy is general, as all acrylic monomers follow family type behavior in terms of propagation kinetics,8 copolymerization reactivity ratios,9,10 and even secondary reactions: all methacrylate monomers depropagate,11,12 and all acrylate monomers backbite to form midchain radicals of lower reactivity.13,14 However, there are differences in the reactivities of functional monomers such as 2-hydroxyethyl methacrylate (HEMA) that influence their polymerization behavior. These variations have been systematically studied utilizing the pulsedlaser polymerization (PLP) technique combined with size exclusion chromatography (SEC) measurement of the resulting polymer molecular weight distribution (MWD) and NMR determination of copolymer composition in order to simultaneously yield information about copolymerization reactivity ratios as well as overall propagation rate.10,15 It was found that HEMA produces methacrylate-enriched copolymer in styrene-rich bulk monomer mixtures relative to alkyl methacrylates such as butyl methacrylate (BMA)10 and that the reactivity of the HEMA when copolymerized with styrene (ST) is also significantly influenced by solvent choice.15 © 2013 American Chemical Society

Building on previous homopolymerization studies by Beuermann, 16−18 the effect of solvent on ST/HEMA copolymerization was attributed to intermolecular hydrogen bonding between monomer units or between monomer and solvent that affects monomer reactivity and thus both copolymer composition and copolymer-averaged propagation rate (kp,cop). In the presence of dimethyl formamide (DMF), a solvent that disrupts H-bonding between HEMA units, the relative reactivity of HEMA in copolymerization became similar to that of BMA.15 An opposite effect was observed when nbutanol (BuOH) was used as a solvent for ST/BMA copolymerization: H-bonding between the BMA carbonyl group and the alcohol increased the reactivity of BMA so that it approached that of HEMA. It was subsequently shown that these kinetic effects have a significant influence on polymerization rate and polymer molecular weights (MWs) of ST/HEMA copolymers produced at high temperature under starved-feed conditions through a comparison of semibatch free radical copolymerization of ST/ HEMA in xylene (a nonpolar solvent) and DMF (a polar solvent which disrupts HEMA hydrogen bonding15) to the copolymerization of ST/BMA under identical conditions.19 Whereas changing the solvent had no influence on free monomer levels or polymer MWs for the ST/BMA system, significant effects were observed for ST/HEMA; free monomer levels were higher in DMF solution (compared to xylene), in agreement with the kinetic studies finding a lowered value of kp,cop. Moreover, the free monomer had a higher fraction of HEMA in DMF than in xylene, due to the reduced HEMA reactivity in the polar solvent. There was also a significant increase in polymer MW of ST/HEMA compared to ST/BMA Special Issue: John Congalidis Memorial Received: Revised: Accepted: Published: 7296

August 21, 2013 October 9, 2013 October 17, 2013 October 17, 2013 dx.doi.org/10.1021/ie4027549 | Ind. Eng. Chem. Res. 2014, 53, 7296−7304

Industrial & Engineering Chemistry Research

Article

Figure 1. PLP/SEC/NMR results for BMA/HEMA copolymerization in different solvents at 90 and 100 °C. (■, 100 °C in bulk; □, 90 °C in bulk; ●, 90 °C in 50 vol % BuOH; ▲, 90 °C in 50 vol % DMF). (a) Mole fraction of HEMA in copolymer (FHEMA) as a function of HEMA mole fraction in the monomer phase ( f HEMA). The solid curve represents the terminal model prediction for bulk copolymerization, while the dashed curve is the diagonal line. (b) Copolymer-averaged propagation rate coefficients (kp,cop) data vs 2-hydroxyethyl methacrylate (HEMA) monomer mole fraction. Terminal model predictions at 100 and 90 °C are indicated by the solid and dashed lines, respectively.

attached to the laser in order to regulate the pulse output repetition rate at a value between 10 and 100 Hz. Monomer/ solvent mixtures with 5 mmol·L−1 2,2-dimethoxy-2-phenylacetophenone (DMPA) photoinitiator were added to the quartz cell and exposed to laser energy, with temperature controlled by a circulating oil bath. Monomer conversions were kept below 5% to avoid significant compositional drift, and temperature was controlled to within ±1 °C during pulsing. The resulting PLP samples were precipitated in diethyl ether and then centrifuged to separate the solids from liquid. The polymers were dried in a vacuum oven and then redissolved in tetrahydrofuran (THF) for SEC analysis. Semibatch reactions were performed in a 1 L LabMax reactor system equipped with an agitator and reflux condenser. Initiator and monomer feed rates as well as reaction temperature were automatically controlled, following the procedures used previously.19 The reactor was charged with 215 g of solvent and brought up to the reaction temperature of 138 °C. The monomer mixture and initiator solution were continuously fed at a fixed rate over 6 h with initiator fed for an extra 15 min; the total initiator charge was 1.5 mol % of the monomer charge. Samples of approximately 2 mL were drawn from the reactor at specified times into ice-cold 4-methoxyphenol (1 g·L−1) solution to terminate the reaction. The residual monomer concentration in the semibatch samples was determined using a Varian CP-3800 gas chromatograph (GC) setup.12,21 Calibration standards were prepared by mixing measured quantities of HEMA and BMA (or BA) monomers into a known mass of acetone, and the corresponding linear calibration curve was constructed by plotting peak area versus monomer concentration. Sizeexclusion chromatography (SEC) analyses of all polymer samples (both PLP and semibatch) were performed using a Waters 2960 separation module with a Waters 410 differential refractometer (RI detector) and a Wyatt Technology Light Scattering detector (LS detector). Calibration for the RI detector was established using eight narrow dispersity

arising from the presence of ppm levels of ethylene glycol dimethacrylate (EGDMA), an impurity formed by esterification between HEMA and ethylene glycol during monomer preparation.20 The ST/HEMA studies demonstrated that the insights gained from the small scale kinetic experiments could be applied to interpret the operation of the starved-feed semibatch process used to produce acrylic copolymers for automotive coatings and also showed that hydroxyl-functional monomers can have a significant influence on the reaction kinetics. To further explore and generalize these effects, PLP-SEC and NMR analyses complemented by semibatch reactor studies are applied to solution copolymerizations of BMA/HEMA and BA/HEMA.

■

EXPERIMENTAL SECTION HEMA (97% purity containing 200−220 ppm monomethyl ether hydroquinone inhibitor and 100−300 ppm EGDMA impurity), BMA (99% purity, containing 10−15 ppm monomethyl ether hydroquinone as inhibitor), and BA (99% purity, containing 10−60 ppm monomethyl ether hydroquinone as inhibitor) were purchased from Sigma Aldrich and used as received. Due to the difficulty in separating EGDMA from HEMA20 no purification was undertaken in order to mimic industrial practices. tert-Butyl peroxyacetate (TBPA), used as initiator at 138 °C, was obtained from Arkema. The solvents, a xylene isomeric mixture with boiling point range between 136 and 140 °C, n-butanol (BuOH, 99% purity), and dimethylformamide (DMF, 99.8% purity) were obtained from Sigma-Aldrich and used as received. Low conversion polymerizations were conducted in a pulsed laser setup consisting of a Spectra-Physics Quanta-Ray 100 Hz Nd:YAG laser that is capable of producing a 355 nm laser pulse of duration 7−10 ns and energy of 1−50 mJ per pulse. The laser beam is reflected twice (180°) into a Hellma QS165 0.8 mL jacketed optical sample cell used as the PLP reactor. A digital delay generator (DDG, Stanford Instruments) is 7297

dx.doi.org/10.1021/ie4027549 | Ind. Eng. Chem. Res. 2014, 53, 7296−7304


Article

Figure 2. Solvent effects on copolymer compositions from low conversion PLP experiments at 50 °C (■, in bulk; ●, in 50 vol % BuOH; ▲, in 50 vol % DMF). (a) BA/HEMA copolymerization, described by mole fraction of HEMA in copolymer (FHEMA) as a function of HEMA mole fraction in the monomer phase ( f HEMA). Lines are terminal model predictions with literature reactivity values (solid line)26 and best-fit curves to bulk (dotted line) and solution (dashed line) experimental data. (b) BA/BMA copolymerization system, described by mole fraction of BMA in copolymer (FBMA) as a function of BMA mole fraction in the monomer phase ( f BMA) compared to predictions using literature reactivity ratios.32 (See text for further details.)

mixture calculated assuming volume additivity, and ϕmon is the volume fraction of monomer in the monomer/solvent mixture. The mole fraction of HEMA in the copolymer (FHEMA) is determined by proton NMR, with both kp,cop and FHEMA measurements carried out for varying levels of HEMA mole fraction in the monomer (f HEMA) mixture. The detailed PLPSEC results with copolymer composition analysis by NMR are tabulated elsewhere.25 As indicated in Figure 1a, and as in agreement with previous literature for HEMA copolymerized with methyl methacrylate (MMA)26 and BMA,27 HEMA-rich copolymer is formed in bulk due to the increased reactivity of HEMA monomer to radical addition relative to BMA.28,29 Monomer reactivity ratios in bulk are estimated as rBMA = 0.35 ± 0.28 and rHEMA = 1.49 ± 0.67, assuming that radical reactivity depends only on the terminal unit of the growing chain such that the mole fraction of HEMA in the copolymer depends on monomer mole fractions and the monomer reactivity ratios according to

polystyrene standards ranging in molecular weight from 890 to 3.55 × 105 g·mol−1, while the MWs of the copolymers were calculated as a weighted average of the poly(HEMA), poly(BMA), and poly(BA) homopolymer values obtained by universal calibration using known Mark−Houwink parameters.7,10 The output from the LS detector provides the absolute molar mass without the need for calibration standards but with knowledge of the dn/dc values.7,10 In all cases MW averages from the two detectors were within 15% of each other; averages calculated from the RI detector are reported, due to better reproducibility. Proton NMR was used to determine the composition of copolymer produced by PLP/SEC, as described previously.10,15 In addition, the macromonomer content (MM%, reported per 100 BA repeat units in the polymer) of the final BA copolymer semibatch samples was determined by proton NMR, with 13C NMR used to measure the branching level (BL%) according to previously developed procedures.13,22

■

RESULTS AND DISCUSSION PLP-SEC Studies. The general expectation for methacrylate−methacrylate copolymerization is that copolymer composition will not vary greatly from the composition of the monomer mixture, as seen in previous studies with alkyl methacrylates.23,24 However, the relative reactivity of HEMA is increased due to intermolecular H-bonding, an influence that is disrupted by DMF or can be induced in BMA in the presence of BuOH.15 Thus, PLP studies of BMA/HEMA copolymerization were carried out in bulk as well as in 50 vol % BuOH and DMF solutions at 90 and 100 °C. The details of the PLP-SEC technique are well-documented.8,15 In brief, the copolymeraveraged propagation rate coefficient kp,cop (L·mol−1·s−1) is deduced from k p,cop =

MW0 1000ϕmonρmon t0

FHEMA =

2 + fHEMA fBMA rHEMAfHEMA 2 2 + 2fHEMA fBMA + rBMAfBMA rHEMAfHEMA

(2)

As expected from the previous study of HEMA with ST,15 the introduction of DMF as solvent decreases the HEMA reactivity to that of BMA, such that the copolymer composition lies close to the diagonal, with no preferential HEMA incorporation (rHEMA ≅ rBMA ≅ 1). In contrast, use of BuOH as a solvent increases the BMA reactivity toward that of HEMA, again forcing the copolymer composition data toward the diagonal line (rHEMA ≅ rBMA ≅ 1). The corresponding kp,cop data are presented in Figure 1b. The bulk kp,cop data, collected at both 90 and 100 °C, are wellrepresented by the curves generated according to the terminal model using the reactivity ratios reported above with literature BMA and HEMA homopropagation rate coefficients:30,31

(1)

where MW0 is the polymer molecular weight at the first inflection point of the low conversion PLP-generated MWD as determined by SEC, ρmon (g·mL−1) is the density of monomer

k p,cop =

7298

2 2 rHEMAfHEMA + 2fHEMA fBMA + rBMAfBMA ⎛ rHEMAfHEMA ⎞ ⎛ rBMAfBMA ⎞ ⎜ ⎟ + ⎜ ⎟ ⎝ k p,HEMA ⎠ ⎝ k p,BMA ⎠

(3)



Article

Since the terminal model fits both copolymer composition and kp,cop data well for the bulk BMA/HEMA system, it can be inferred that the radical reactivity in methacrylate/methacrylate copolymerization is only dependent on the type of terminal unit, with negligible influence from the identity of the methacrylate unit in the penultimate position. DMF decreases the propagation rate coefficient significantly, where kp,cop remains close to the BMA bulk value over a large range of monomer mixture compositions. This result can be explained by the reduction of kp,HEMA in DMF by the disruption of intermolecular H-bonding, such that HEMA reactivity becomes similar to that of BMA.18 Meanwhile, BuOH increases the propagation rate coefficient of BMA by roughly 25% (from 2000 to 2500 L·mol−1·s−1 at 90 °C).17 As BuOH has no effect on the HEMA propagation rate coefficient, the kp,cop values quickly converge to the values measured for the bulk system. These H-bonding effects found for the BMA/HEMA system are consistent with the results of the previous studies on ST/ HEMA and ST/BMA solution polymerization kinetics.15 Copolymerization of BA and HEMA at 50 °C was also carried out in bulk, BuOH, and DMF solutions using the PLPSEC technique for kp,cop determination coupled with NMR analysis of copolymer composition. In addition, bulk and solution copolymer composition data were obtained for BA/ BMA copolymerizations. Experimental copolymer composition results are shown in Figure 2. Terminal model kinetics are applied to the data to estimate monomer reactivity ratios (with 95% confidence intervals from nonlinear parameter estimation) for BA/HEMA in bulk as rBA = 0.18 ± 0.09, rHEMA = 5.54 ± 3.27, which are very close to the literature values of rBA = 0.168 and rHEMA = 5.414 reported by Varma and Patnaik,26 as shown in Figure 2a. The additions of BuOH and DMF solvents to the BA/ HEMA system have the same effect on HEMA incorporation as observed for its copolymerization with ST and with BMA, albeit to a lesser extent: compared to the bulk copolymerization system there is a decreased level of HEMA incorporation in solution. These results can be interpreted using the same explanations: introduction of DMF disrupts the intermolecular H-bonding of HEMA making it less reactive, while the addition of BuOH leads to H-bonding formation with BA, increasing its reactivity to radical addition. To fit the composition data obtained for solution copolymerizations, the monomer reactivity ratios are estimated as rBA = 0.29 ± 0.14 and rHEMA = 3.57 ± 2.17, where a single fit is applied to the data obtained in BuOH and in DMF as the results cannot be distinguished. Interestingly, these values are close to those reported by Aerdts et al.32 for bulk BA/BMA copolymerization, rBA = 0.395 and rBMA = 2.279. Thus, it is found that BA/HEMA copolymer compositions in BuOH or DMF approach those of the BA/ BMA bulk system. In Figure 2b, the terminal model prediction for BA/BMA copolymer composition using the Aerdts et al.32 monomer reactivity ratios is compared to the experimental BA/ BMA copolymerization data in bulk and BuOH from this study. The very good fit indicates that any possible H-bonding effects influence both BA and BMA reactivities to the same extent, such that the net effect on the reactivity ratios is negligible. The HEMA/BA copolymerization propagation rate coefficients (kp,cop) at 50 °C were also determined by PLP-SEC; due to the insolubility of HEMA-rich copolymer in THF, the highest HEMA monomer composition examined was 50 mol %. Results are plotted in Figure 3, with detailed experimental data tabulated elsewhere.25 As found in previous PLP-SEC studies of

Figure 3. BA/HEMA copolymerization propagation rate coefficients (kp,cop) data at 50 °C vs 2-hydroxyethyl methacrylate (HEMA) monomer mole fraction (■, in bulk; ●, in 50 vol % BuOH; ▲, in 50 vol % DMF). The terminal model prediction for BA/HEMA in bulk is indicated by the solid curve.

acrylate/methacrylate systems,9,33 kp,cop values drop quickly from the acrylate homopolymerization value (>10 000 L·mol−1· s−1) upon addition of methacrylate monomer. The terminal model predictions (solid line in Figure 3) using the bulk monomer reactivity ratios fail to describe the kp,cop behavior of BA/HEMA copolymerization, especially for the BA-rich mixtures. The underprediction of kp,cop by the terminal model is in agreement with the findings from previous methacrylate/ acrylate copolymerization studies9,33 that found it necessary to employ the penultimate model of copolymerization propagation34 to represent the data. Due to the limited data set, such a fit has not been attempted for the BA/HEMA bulk system. Although the kp,cop behavior over the entire composition range cannot be reviewed, the solvent effect on kp,cop is still easy to identify. From Figure 3, it is clear that BuOH increases kp,cop significantly. As discussed previously, the addition of BuOH does not affect HEMA homopropagation but significantly increases the rate coefficient of BMA homopropagation; limited data in the literature indicates that BuOH has a similar effect on BA kp.22 In contrast, the addition of DMF decreases the kp value of HEMA and has no effect on BMA kp; assuming that DMF does not influence BA kp, the decrease in kp,cop seen in Figure 3 can be attributed to decreased HEMA reactivity in the presence of H-bonding disruptor DMF. Thus, these kinetic studies provide evidence of the increased HEMA reactivity during bulk copolymerization relative to the same system with BMA due to intermolecular H-bonding and demonstrate how the influence can be reduced by adding DMF to the system. Similarly, addition of an alcohol to a BMA or BA system enhances the monomer reactivity compared to the bulk system. The results are consistent with the previous study of HEMA (and BMA) copolymerized with ST,15 indicating the generality of the findings. The next step is to examine the influence of these solvent effects under higher temperature semibatch operation. Starved-Feed BMA/HEMA Semibatch Copolymerization. Semibatch BMA/HEMA copolymerizations with 12.5 and 25 wt % HEMA were carried out in 35 wt % xylene or DMF (final solvent fraction) at 138 °C. The free monomer profiles shown in Figure 4 are typical for starved-feed semibatch operation.6,7 As found for ST/HEMA semibatch copolymeriza7299



Article

Free monomer levels measured in BMA homopolymerization in xylene are similar to those measured for BMA/HEMA copolymerization in the same solvent. However, the total monomer concentrations in DMF are always higher than those in xylene, in agreement with the results shown in Figure 1b in which the kp,cop values in DMF are significantly lower than those for bulk polymerization. (It can be assumed that the bulk kp,cop value is also representative of the comonomer mixture in xylene.) HEMA monomer fractions in DMF are higher than those in xylene, which can also be connected to the kinetic studies; as shown in Figure 1a, a higher HEMA monomer fraction in DMF is required compared to that in bulk to achieve the same level of HEMA incorporation in the copolymer. For the copolymerizations in DMF, the HEMA fraction in the reactor is very close to that in the feed (represented by the lines in Figure 5) after the initial transient period, as expected for a system with reactivity ratios close to unity. Note that the copolymer composition produced by the semibatch process always matches the monomer feed ratios, due to the starvedfeed policy. While the dependencies of free monomer levels on solvent choice are well-explained by the results from the PLP-SEC experiments, the large increase in polymer MW resulting from the addition of HEMA to the BMA system needs some consideration. As shown in Figure 5 and summarized in Table 1, the addition of 25 wt % HEMA results in an increase in final

Figure 4. Monomer concentration ([HEMA] and [BMA]) profiles for semibatch copolymerization at 138 °C with BMA/HEMA mass ratio of 75/25 (■, □) and 87.5/12.5 (▲, Δ) in the feed. Final polymer content is 65 wt % in solution, with 1.5 mol % initiator fed to the reactor relative to monomer. Filled symbols indicate reactions in DMF, and open symbols indicate reactions in xylene.

tions,19 the reactions conducted in DMF have significantly higher free monomer concentrations than observed for polymerizations in xylene. Although decreasing the HEMA fraction in the feed to 12.5 wt % reduced the free HEMA in the reactor, monomer concentrations remain significantly higher in DMF compared to xylene. In order to analyze these results further, Figure 5 compares total monomer concentration, fraction of HEMA in the free monomer, and polymer weight-average MW (Mw) values for BMA/HEMA copolymerizations to BMA homopolymerization.

Table 1. Weight-Average MW Values (Mw, in kg·mol−1) of BMA Homopolymer, ST/BMA,19 ST/HEMA,19 and BMA/ HEMA Copolymers Produced by Semibatch Copolymerization at 138 °C with 65 wt % Polymer in Solution ST/BMA

in xylene in DMF

ST/HEMA

BMA/HEMA

BMA

25% BMA

12.5% HEMA

25% HEMA

12.5% HEMA

25% HEMA

15.5 -

16.8 16.3

23.5 28.9

41.7 55.6

15.2 18.0

25.5 35.9

polymer Mw values by more than a factor of 2 for copolymerization in DMF compared to BMA homopolymerization (from 16 to 36 kg·mol−1), and an increase by 50% to 26 kg·mol−1 in xylene. However, the Mw values obtained with 12.5 wt % HEMA in the mixture are similar to those measured for BMA homopolymer produced under identical semibatch conditions. These findings were also observed for ST/HEMA copolymerization under the same operating conditions,19 for which the addition of HEMA led to an even greater increase in polymer MW compared to ST/BMA (Table 1). While the increase in MW with HEMA content has been attributed to the presence of EGDMA impurities in the HEMA monomer feed which leads to branch formation,19 it is not clear why the impact is greater in DMF compared to xylene solution. To further explore the effect of solvent on polymer MW, semibatch homopolymerizations of BMA with 100 ppm added EGDMA were carried out in xylene and DMF at 138 °C. The significant influence of the impurity is shown in Figure 6, where the addition of 100 ppm EGDMA increases polymer Mw values from 15 to over 30 kg·mol−1 in xylene and to over 40 kg·mol−1 in DMF. Thus it can be concluded that the increased copolymer MW is due to the presence of EGDMA impurity and not a direct influence of HEMA addition to the recipe. The reason behind the observed interactions between branching

Figure 5. Total monomer concentration, mole fraction HEMA in monomer, and polymer weight-average molecular weight profiles for BMA/HEMA semibatch copolymerizations at 138 °C with BMA/ HEMA mass ratio of 75/25 (■, □), 87.5/12.5 (▲, Δ), and 100/0 (○) in the feed. Final polymer content is 65 wt % in solution, with 1.5 mol % initiator fed to the reactor relative to monomer. Filled symbols indicate reactions in DMF, and open symbols indicate reactions in xylene. Lines represent HEMA fraction in monomer feed for 25% (solid) and 12.5% (dash) experiments. 7300



Article

BuOH on polymer structure when introduced as a comonomer, compared to copolymerization of BMA with BA. Figure 7 compares BA/BMA and BA/HEMA semibatch polymerizations produced with 25 wt % methacrylate under identical operating conditions. Solvent choice has no effect on BA/BMA polymer MWs or free monomer levels, a result also reported for ST/BMA copolymerization.19 The total monomer concentration for the system remains below 0.5 mol·L−1, significantly lower than found for methacrylate/methacrylate (see Figure 5) and methacrylate/styrene19 polymerizations. The lower free monomer concentration results from the higher kp,cop values for the acrylate containing system and leads to lower polymer MW values; the Mw values of 5−7 kg·mol−1 are less than half those reported in Table 1 for methacrylate/ methacrylate and methacrylate/styrene systems. This behavior is consistent with previous experimentation and is well represented by mechanistic models developed for the systems.6,7,14 For the BA/HEMA semibatch copolymerization, solvent choice does not have a large effect on total free monomer concentration or composition; as shown in Figure 8, any observed difference is smaller than the uncertainty in the experimental data. An examination of the Mayo−Lewis plots explains this result: the difference in comonomer composition required to produce copolymer with 25 wt % HEMA in xylene (bulk) and DMF is much smaller when BA (Figure 2a) is used as comonomer rather than BMA (Figure 1a). Once again, the results obtained during semibatch operationin this case the lack of solvent effect observed for HEMA copolymerized with BA compared to HEMA copolymerized with BMAare wellexplained by analysis of the kinetic studies. It is interesting to observe that HEMA has little effect on polymer MWs compared to BA homopolymerization or BA/ BMA copolymerization, as summarized in Table 2. The poly(BA-co-HEMA) Mw values are only slightly higher than those of poly(BA-co-BMA), a finding very different from the effect of substituting HEMA for BMA during copolymerization with ST or BMA (see Table 1). These differences suggest that the branching reaction caused by the dimethacrylate impurity in HEMA, and how it is influenced by solvent choice, is of much

Figure 6. BMA semibatch homopolymerization (65 wt % final polymer content and 1.5 mol % initiator relative to monomer), polymer weight-average molecular weight experimental profiles for BMA with 100 ppm added EGDMA in xylene (●), DMF (■), and without EGDMA in xylene (○) at 138 °C.

reactions (polymer MW) and solvent choice is not obvious. Matsumoto et al.35 proposed that H-bonding involving monomer units incorporated into the polymer is significant and may lead to increased polymer MW through gelation. However, the influence of solvent on incorporated EGDMA units, present in ppm levels, cannot be directly studied. Starved-Feed BA/HEMA Semibatch Copolymerization. To complete the study of HEMA-solvent interactions on starved-feed higher temperature operations, BA/HEMA semibatch experiments with 25 wt % HEMA were also carried out in 35 wt % solvent (xylene or DMF) at 138 °C. In addition to the effect of solvent on BA/HEMA kinetics demonstrated in Figures 2 and 3, it has recently been shown that H-bonding can affect acrylate backbiting kinetics; the replacement of xylene by BuOH as a solvent for semibatch polymerization of BA led to a reduction in the level of quaternary carbons (formed by backbiting) detected in the polymer using 13C NMR.22,36 Thus, the question is whether HEMA exerts a similar influence as

Figure 7. Monomer concentration and polymer weight−average molecular weight experimental profiles for (a) BA/BMA and (b) BA/HEMA semibatch copolymerizations at 138 °C where BMA or HEMA is 25 wt % of feed with 1.5 mol % initiator relative to monomer and final polymer content is 65 wt % in xylene (■) or DMF (▲). 7301



Article

Figure 8. Total monomer concentration and mole fraction HEMA in comonomer profiles for BA/HEMA semibatch copolymerizations at 138 °C where BA/HEMA mass ratio in feed is 75/25 with 1.5 mol % initatior relative to monomer and final polymer content is 65 wt % in xylene (■) or DMF (▲).

Figure 9. Polymer weight-average MW profiles for semibatch homopolymerization of BA with and without 100 ppm added EGDMA at 110 °C in 70 wt % xylene.

Table 2. Weight-Average MW Values (kg·mol−1), Macromonomer Levels (MM%), and Branching Levels (BL %) of Final Polymer Produced by Semibatch (Co)Polymerization of BA, BA/BMA, and BA/HEMA with 35 wt % Solvent at 138 °C BA −1

Mw (kg·mol ) MM%/BL%a

in xylene in DMF in xylene in DMF

BA/BMA (25 wt % BMA)

BA/HEMA (25 wt % HEMA)

5.94

6.77

8.06

4.66 0.38/9.78

6.56 0.26 (0.20)/ 4.35 (3.44) 0.27 (0.21)/ 4.42 (3.49)

7.76 0.21 (0.16)/ 3.25 (2.50) 0.23 (0.17)/ 3.38 (2.60)

0.43/9.85

backbiting.37 A recent computational study38 reports that the acrylate backbiting rate coefficient is increased when a methacrylate unit is adjacent to the site of the H-atom abstraction from an acrylate moiety; thus, further investigation and modeling on the influence on copolymer content on branching levels is required. Substitution of HEMA for BMA in the copolymer reduces branching (from 4 to 3%) and increases polymer MW (from 7 to 8 kg·mol−1) slightly, in accordance with the previously reported effect of H-bonding on acrylate backbiting.22,36 However, more experimentation is required to determine whether these changes are significant.

■

CONCLUSIONS The PLP-SEC technique combined with NMR copolymer composition analysis has been employed to investigate the influence of solvent on HEMA copolymerization kinetics with BMA and BA both in bulk and in solution. It was found that HEMA preferentially incorporates (compared to BMA) into copolymer in bulk (or nonpolar xylene) due to the effect of intermolecular H-bonding on HEMA monomer reactivity. BuOH and DMF influence H-bonding of BMA (or BA) and HEMA, respectively, such that they demonstrate similar reactivity in both solvents; kp,cop values in DMF are reduced compared to bulk while those in BuOH are increased. These kinetic investigations are used to explain the concentration and composition of free monomer levels measured during BMA/HEMA and BA/HEMA semibatch copolymerizations in xylene and DMF under starved-feed higher-temperature conditions. As found previously for ST/ HEMA,19 the total monomer concentration and f HEMA values are higher for BMA/HEMA copolymerizations conducted in DMF than in xylene, in agreement with the lower kp,cop values and lowered relative incorporation rates measured by the kinetics experiments. The MWs of the HEMA containing copolymers (with either ST or BMA) are significantly higher than those of ST/BMA and BMA polymers produced under identical semibatch conditions. Experiments have verified that the increased MW is not directly related to HEMA H-bonding effects, but instead results from the dimethacrylate impurity found in HEMA monomer.

a

Values reported per 100 BA repeat units and (in parentheses) per 100 total repeat units in the polymer.

reduced importance for the BA/HEMA system. In order to confirm this finding, the Mw profiles of semibatch experiments performed for BA homopolymerization (110 °C in xylene) with and without 100 ppm added EGDMA are compared in Figure 9. In contrast to BMA homopolymerization (Figure 6), the addition of EGDMA has negligible effect on poly(BA) MWs. This result, while not understood on a mechanistic level, explains why BA/BMA and BA/HEMA copolymers differ little in MW and why absolute MW levels are so much lower than found for other HEMA-containing copolymers. Table 2 also summarizes the branchpoint (formed by intramolecular chain transfer to form a midchain radical followed by monomer addition) and macromonomer (unsaturated chain ends formed by scission of the midchain radical) levels measured for the final polymer samples. The values for BA homopolymer9−10 branches and 0.3−0.4 unsaturated chain ends per 100 repeat unitsare in agreement with those reported for similar reaction conditions.13,14 The addition of methacrylate (either 25% BMA or HEMA) to the system reduces the occurrence of branchpoints to 3−4 per 100 BA repeat units, as the presence of methacrylate units (which do not contain an easily abstractable H-atom, and which form less reactive radicals) is expected to reduce the probability of 7302



Article

(11) Hutchinson, R. A.; Paquet, D. A., Jr.; Beuermann, S.; McMinn, J. H. Investigation of Methacrylate Free-Radical Depropagation Kinetics by Pulsed-Laser Polymerization. Ind. Eng. Chem. Res. 1998, 37, 3567. (12) Wang, W.; Hutchinson, R. A.; Grady, M. C. Study of Butyl Methacrylate Depropagation Behavior using Batch Experiments in Combination with Modeling. Ind. Eng. Chem. Res. 2009, 48, 4810. (13) Wang, W.; Nikitin, A. N.; Hutchinson, R. A. Consideration of Macromonomer Reactions in Butyl Acrylate Free Radical Polymerization. Macromol. Rapid Commun. 2009, 30, 2022. (14) Nikitin, A. N.; Hutchinson, R. A.; Wang, W.; Kalfas, G. A.; Richards, J. R.; Bruni, C. Effect of Intramolecular Transfer to Polymer on Stationary Free Radical Polymerization of Alkyl Acrylates, 5Consideration of Solution Polymerization up to High Temperatures. Macromol. React. Eng. 2010, 4, 691. (15) Liang, K.; Hutchinson, R. A. Solvent Effects on Free-Radical Copolymerization Propagation Kinetics of Styrene and Methacrylates. Macromolecules 2010, 43, 6311. (16) Beuermann, S.; Nelke, D. The Influence of Hydrogen Bonding on the Propagation Rate Coefficient in Free-Radical Polymerizations of Hydroxypropyl Methacrylate. Macromol. Chem. Phys. 2003, 204, 460. (17) Beuermann, S. Impact of Hydrogen Bonding on Propagation Kinetics in Butyl Methacrylate Radical Polymerizations. Macromolecules 2004, 37, 1037. (18) Beuermann, S. Solvent Influence on Propagation Kinetics in Radical Polymerizations Studied by Pulsed Laser Initiated Polymerizations. Macromol. Rapid Commun. 2009, 30, 1066. (19) Liang, K.; Hutchinson, R. A. Solvent Effects in Semibatch Free Radical Copolymerization of 2-Hydroxyethyl Methacrylate and Styrene at High Temperatures. Macromol. Symp. 2013, 325, 203. (20) Montheard, J.-P.; Chatzopoulos, M.; Chappard, D. 2Hydroxyethyl Methacrylate (HEMA): Chemical Properties and Applications in Biomedical Fields. J. Macromol. Sci., Part C: Polym. Rev. 1992, 32, 1. (21) Li, D.; Grady, M. C.; Hutchinson, R. A. High Temperature Semibatch Free-Radical Copolymerization of Butyl Methacrylate and Butyl Acrylate. Ind. Eng. Chem. Res. 2005, 44, 2506. (22) Liang, K.; Hutchinson, R. A. The Effect of Hydrogen Bonding on Intramolecular Chain Transfer in Polymerization of Acrylates. Macromol. Rapid Commun. 2011, 32, 1090. (23) Stahl, G. A. Copolymerization of Methyl Methacrylate and Dodecyl Methacrylate. J. Polym. Sci., Polym. Chem. Ed. 1979, 17, 1883. (24) Grassie, N.; Torrance, B. J. D.; Fortune, J. D.; Gemmell, J. D. Reactivity Ratios for the Copolymerization of Acrylates and Methacrylates by Nuclear Magnetic Resonance Spectroscopy. Polymer 1965, 6, 653. (25) Liang, K. Free Radical Copolymerization of Hydroxy-Functional Monomers: Kinetic and Semibatch Studies. Ph.D. Thesis, Queen’s University, Kingston, ON, 2013. (26) Varma, I. K.; Patnaik, S. Copolymerization of 2-Hydroxyethyl Methacrylate with Alkyl Acrylates. Eur. Polym. J. 1976, 12, 259. (27) Fernández-García, M.; Torrado, M. F.; Martínez, G.; SánchezChaves, M.; Madruga, E. L. Free Radical Copolymerization of 2Hydroxyethyl Methacrylate with Butyl Methacrylate: Determination of Monomer Reactivity Ratios and Glass Transition Temperatures. Polymer 2000, 41, 8001. (28) Ito, K.; Uchida, K.; Kitano, T.; Yamada, E.; Matsumoto, T. Solvent Effects in Radical Copolymerization between Hydrophilic and Hydrophobic Monomers: 2-Hydroxyethyl Methacrylate and Lauryl Methacrylate. Polym. J. 1985, 17, 761. (29) Fernández-Monreal, C.; Sanchez-Chaves, M.; Martinez, G.; Madruga, E. L. Stereochemical Configuration of 2-Hydroxyethyl Methacrylate/Styrene Copolymers Obtained in N,N′-DimethyIformamide Solution over A Whole Range of Conversion. Acta Polym. 1999, 50, 408. (30) Beuermann, S.; Buback, M.; Davis, T. P.; Gilbert, R. G.; Hutchinson, R. A.; Kajiwara, A.; Klumperman, B. L.; Russell, G. T. Critically Evaluated Rate Coefficients for Free-Radical Polymerization:

While H-bonding (and solvent choice) also exerts an influence on BA/HEMA kinetics, the relative effect is smaller than that found for the styrene/methacrylate and methacrylate/ methacrylate systems. In addition, free monomer and f HEMA values are lower in acrylate copolymerizations compared to the other systems. Thus, neither the solvent choice nor the substitution of HEMA for BMA has a significant impact on semibatch operation (free monomer levels, polymer MWs, branching levels) to produce acrylate copolymer with 25 wt % methacrylate content. Surprisingly, the presence of EGDMA impurity in the HEMA does not have a significant influence on acrylate (co)polymer MW values, a finding worthy of additional study.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: (R.A.H) [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding from the Natural Science and Engineering Council (NSERC) of Canada and from Axalta Coating Systems is gratefully acknowledged.

■ ■

DEDICATION Dedicated to the memory of John Congalidis, a longtime colleague, collaborator. and friend. REFERENCES

(1) Zimmt, W. S. Coatings from Acrylic Polymers. CHEMTECH 1981, 11, 681. (2) Bauer, D. R.; Dickie, R. A. Application of Network Structure Models to Optimization of Bake Conditions for Thermoset Coatings. J. Coating Technol. 1986, 58, 41. (3) Tilak, G. Y. Thermosetting Acrylic Resins - A Literature Review. Prog. Org. Coat. 1985, 13, 333. (4) Adamsons, K.; Blackman, G.; Gregorovich, B.; Lin, L.; Matheson, R. Oligomers in the Evolution of Automotive Clearcoats: Mechanical Performance Testing as A Function of Exposure. Prog. Org. Coat. 1997, 34, 64. (5) Grady, M. C.; Simonsick, W. J.; Hutchinson, R. A. Studies of Higher Temperature Polymerization of n-Butyl Methacrylate and nButyl Acrylate. Macromol. Symp. 2002, 182, 149. (6) Wang, W.; Hutchinson, R. A. Recent Advances in the Study of High Temperature Free Radical Acrylic Solution Copolymerization. Macromol. React. Eng. 2008, 2, 199. (7) Wang, W.; Hutchinson, R. A. A Comprehensive Kinetic Model for High Temperature Free-Radical Production of Styrene/Methacrylate/Acrylate Resins. AIChE J. 2011, 57, 227. (8) Beuermann, S.; Buback, M. Rate Coefficients of Free-Radical Polymerization Deduced from Pulsed Laser Experiments. Prog. Polym. Sci. 2002, 27, 191. (9) Buback, M.; Feldermann, A.; Barner-Kowollik, C.; Lacík, I. Propagation Rate Coefficients of Acrylate−Methacrylate Free-Radical Bulk Copolymerizations. Macromolecules 2001, 34, 5439. (10) Liang, K.; Dossi, M.; Moscatelli, D.; Hutchinson, R. A. An Investigation of Free-Radical Copolymerization Propagation Kinetics of Styrene and 2-Hydroxyethyl Methacrylate. Macromolecules 2009, 42, 7736. 7303



Article

3. Propagation Rate Coefficient for Alkyl Methacrylates. Macromol. Chem. Phys. 2000, 201, 1355. (31) Buback, M.; Kurz, C. Free-Radical Propagation Rate Coefficients for Cyclohexyl Methacrylate, Glycidyl Methacrylate and 2-Hydroxyethyl Methacrylate Homopolymerizations. Macromol. Chem. Phys. 1998, 199, 2301. (32) Aerdts, A. M.; German, A. L.; van der Velden, G. P. M. Determination of the Reactivity Ratios, Sequence Distribution and Stereoregularity of Butyl Acrylate-Methyl Methacrylate Copolymers by Means of Proton and Carbon-13 NMR. Magn. Reson. Chem. 1994, 32, S80. (33) Hutchinson, R. A.; McMinn, J. H.; Paquet, D. A., Jr.; Beuermann, S.; Jackson, C. A Pulsed-Laser Study of Penultimate Copolymerization Propagation Kinetics for Methyl Methacrylate/ nButyl Acrylate. Ind. Eng. Chem. Res. 1997, 36, 1103. (34) Fukuda, T.; Kubo, K.; Ma, Y.-D. Kinetics of Free Radical Copolymerization. Prog. Polym. Sci. 1992, 17, 875. (35) Matsumoto, A.; Ueda, A.; Aota, H.; Ikeda, J. Effect of Hydrogen Bonds on Intermolecular Crosslinking Reaction by Introduction of Carboxyl Groups in Free-Radical Crosslinking Monomethacrylate/ Dimethacrylate Copolymerizations. Eur. Polym. J. 2002, 38, 1777. (36) Liang, K.; Hutchinson, R. A.; Barth, J.; Samrock, S.; Buback, M. Reduced Branching in Poly(Butyl Acrylate) via Solution Radical Polymerization in n-Butanol. Macromolecules 2011, 44, 5843. (37) González, I.; Asua, J. M.; Leiza, J. R. The Role of Methyl Methacrylate on Branching and Gel Formation in the Emulsion Copolymerization of BA/MMA. Polymer 2007, 48, 2542. (38) Cuccato, D.; Mavroudaksi, E.; Moscatelli, D. Quantum Chemistry Investigation of Secondary Reaction Kinetics in AcrylateBased Copolymers. J. Phys. Chem. A 2013, 117, 4358.

7304


Solvent Effects on Kinetics of 2-Hydroxyethyl Methacrylate Semibatch

Recommend Documents