High Molar Mass Polycarbonate via Dynamic Solution Transcarbonation Using Bis(methyl salicyl) Carbonate, an Activated Carbonate

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website. • The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page numbers.


Introduction
Aromatic polycarbonates are an important class of engineering plastics that find uses in a wide range of applications due to the combination of a high toughness, optical clarity, and heat distortion resistance. [1] The conventional, phosgenefree, route for bisphenol-A (BPA) polycarbonate is a high temperature melt polymerization using diphenyl carbonate (DPC) as the carbonate donor. The required high temperatures, however, make it difficult to incorporate thermally labile or volatile monomers via the conventional melt process. Using an activated carbonate such as bis(methyl salicyl) carbonate (BMSC) instead of DPC has proven to ameliorate this situation in melt polymerization allowing for significantly lower temperatures. [2][3][4] Recently, we reported the efficient synthesis of BPA-based copolycarbonates in solution at relatively low temperatures using BMSC as the carbonate donor. [5] We demonstrated the incorporation of thermally labile monomers and showed the feasibility of producing multiblock copolycarbonates via this route. In that study, however, we did not focus on any kinetic aspects and, except for one polymerization at 60°C, all the reported polymerizations were carried out at 120°C. In the current work, the emphasis lies on the kinetic and thermodynamic aspects of the solution polymerization of BPA and BMSC and we show that this polymerization is an almost ideal equilibrium polymerization and that very high molar masses are obtained at low temperatures without the need for removing the methyl salicylate condensate (Scheme 1).

Results and Discussion
First, kinetic experiments were carried out at 120, 90, and 60°C in ortho-dichlorobenzene (o-DCB) in a closed system (details in Table 2A-C); no condensate was removed during the reaction. During these polymerization reactions, samples were taken at regular time intervals and molar mass distributions were determined via size exclusion chromatography (SEC). The absolute weightaverage molar masses, M w , estimated via the Mark-Houwink parameters (see the Experimental Section) are plotted in Figure 1.   Table 2, experiments A-C.
It is clear from the data in Figure 1 that the polymerization rate increases with increasing temperature, but that the maximum achievable M w decreases.
The significantly faster rate at 120°C as compared to 60°C is expected and explains the difference in molar masses obtained at these two temperatures after 1 h of reaction as reported in our previous publication. [5] Assuming a negligible contribution from the reverse reaction on the initial rates and any chain length dependence of the kinetics, the bimolecular rate coefficients, k, for the forward reaction in Scheme 1 and as defined in Equation (1), were estimated from the data shown in Figure 1 − where the subscript "0" refers to the initial conditions, the following equation can be derived (see Supporting Information) that was used to evaluate the rate coefficients The conversions, p, in Equation (4), are obtained from the number-average degree of polymerization, x n , data via Equation (5), which is obtained from a rearrangement of the general Carothers equation Analyses of the data using Equation (4) (see Supporting Information for details) yields the following estimates for k: k ≈ 3, 30, and 80 L mol −1 s −1 at 60, 90, and 120°C, respectively. Hence, the reaction rate at 120°C is about 20-30 times faster than at 60°C . A note should be made here that an exact determination for x n is difficult because of the inherent experimental uncertainties in determining the low M side of the molar mass distribution by SEC (see Supporting Information for details [6] ). Therefore, we estimated x n as M n /127 and as M w /254 (assuming a theoretical dispersity, Ð, of 2 and ignoring the masses of the end-groups); the differences in the obtained values for k using these two approaches were small.
The second important observation in Figure 1 is that a temperature-dependent maximum M w is obtained. This observation is not surprising when considering that methyl salicylate, the condensate, is not removed from the reaction and the (net) reaction is expected to stop when equilibrium is reached. Again, assuming chain length-independent kinetics, the obtained data now enable us to estimate the equilibrium constants, K eq , for this polymerization via Equation (6) and the results are summarized in Table 1 K The obtained values of the equilibrium constants are much larger than what is commonly reported for polyesters (K eq ≈ 1) and polyamides (K eq ≈ 10 2 -10 3 ). [7] These high equilibrium constants are explained by the fact that the methyl salicylate condensation product is stabilized by a strong intramolecular hydrogen bond. [3] Furthermore, the equilibrium constant decreases with increasing temperature, which is in line with reports in the literature for other systems. [8][9][10][11] The temperature dependence of K eq is most clearly seen in a so-called Van't Hoff plot, which is shown in Figure 2 and allows for the estimation of the standard enthalpy www.advancedsciencenews.com www.mcp-journal.de  (6); f) Data taken from experiment E in Table 2 (and Figure   S6b, Supporting Information), all other data from the kinetic experiments shown in Figure 1 (A-C, Table 2).
The linear fits of the two data sets (with K eq based on x n = M w /254 and x n = M n /127) lead to slightly different values for ΔH 0 and ΔS 0 . The use of x n = M w /254 leads to ΔH 0 = −19 kJ mol −1 and ΔS 0 = 13 J mol −1 K −1 , whereas the use of x n = M n /127 results in ΔH 0 = −11 kJ mol −1 and ΔS 0 = 28 J mol −1 K −1 . We believe the true values to lie within these ranges.
In order to confirm that indeed the polymerization equilibrium was obtained and to establish whether the system is truly reversible, a series of experiments was carried out in which the temperature was varied in a cyclic fashion (experiments D-F in Table 2). When equilibrium was achieved, a sample was taken and quenched in cold tetrahydrofuran (THF) before determination of the molar mass distribution by SEC. At 120°C, the system was left to react for 2 h to make sure that the equilibrium weightaverage molar mass (M w,eq ) was reached after which the reaction mixture was slowly cooled to 60°C and left overnight. This was resulted in an increase in M w,eq from 11 × 10 3 g mol −1 at 120°C to 16 × 10 3 g mol −1 at 60°C. This experiment was performed in three cycles in a closed system and without removal of methyl salicylate. Results of the experiments are summarized in Figure 3A, clearly showing the reversibility of the equilibrium polymerization process. The molar mass distributions for the experiment in which the temperature was cyclically varied between 120 and 60°C are shown in Figure 3B. The molar mass distributions are the same for every cycle at 120°C, whereas the distributions of the polymers at 60°C slightly vary. These small differences are attributed to the lower solubility of the high molar mass polycarbonate in o-DCB at 60°C. The same experiment was also performed between 120°C and room temperature ( Figure 3A, exp F) and also here a variation at low temperatures was observed because of the low solubility of the polymer. When the temperature was increased to 160°C, a decrease in molar mass was observed, but now (because of the good solubility) the same molar mass was reproduced in every cycle ( Figure 3A, exp E). For details on the latter two experiments, the reader is referred to the Supporting Information.
Although the equilibrium molar masses obtained in these experiments are already close to typical molar masses of commercial polycarbonates (M w ≈ 17 × 10 3 g mol −1 for optical quality Lexan resin [1a] ), we tried to push the molar mass to higher values by removing the methyl salicylate before carrying out a second polymerization step in solution. Starting from a polycarbonate synthesized at 60°C (M w ≈ 16 × 10 3 g mol −1 ), we removed the methyl salicylate by precipitation in hexane (1:15 v/v) and continued the polymerization at 120°C (because of the lower polymer solubility at 60°C). In this step, M w increased to 46 × 10 3 g mol −1 and after repeating this step once more, the obtained M w of 67 × 10 3 g mol −1 ( Figure 3C) was very close to the theoretical maximum of 70 × 10 3 g mol −1 (calculated from the Carothers equation for p = 1 and r = 0.9929). This clearly illustrates the favorable equilibrium conditions for this system even at 120°C. It is interesting to note here that the obtained M w is about twice as high as the M w of typical commercial sheet-grade polycarbonates. [1a] From the results above, it is clear that equilibrium polymerization at lower temperatures would be desirable for obtaining higher molar masses, but is limited by the poorer polymer  Table 2). C) Molar mass distributions of polycarbonate synthesized in an equilibrium polymerization at 60°C before removal of methyl salicylate (M w ≈ 16 × 10 3 g mol −1 , dotted line) and after removal of methyl salicylate and subsequent equilibrium polymerizations at 120°C: once (M w ≈ 46 × 10 3 g mol −1 , dashed line) and twice (M w ≈ 67 × 10 3 g mol −1 , solid line). solubility. Therefore, polymerizations were conducted at 120°C using lower starting concentrations of BMSC (i.e., [BMSC] 0 = 0.76, 0.30, and 0.07 m) while the ratios of monomers and catalysts were kept the same. The obtained molar mass distributions are shown in Figure 4. As can be seen from this figure, lowering the concentration results in a significant change of the molar mass distribution at lower M. Since at lower concentrations the chance of a reactive end group finding another chain or monomer decreases, the probability for cyclics formation increases and this seems clearly to be the case here.
The experiments related to Figure 4 (i.e., exps. I, D, and G in Table 2, respectively) were conducted in a cyclic fashion between 120 and 60°C, similar to what was done in experiment D (Figure 3a); the results shown in Figure 4 correspond to the first heating at 120°C. However, in contrast to the case of experiment D, the reactions at lower concentrations are less reversible, as the maximum attainable M w at 60°C decreases in every cycle (see Figure S6, Supporting Information). This is also consistent with the presence of cyclics, which are not able to participate further in chain extensions.

Conclusions
In this work, it was shown that the solution polymerization of BMSC and BPA is characterized by a very high equilibrium constant, allowing for the synthesis of high molar mass polycarbonate without the need for removal of the condensate. Kinetic and equilibrium studies show that the system can be adequately described by the general Carothers' equation in combination with the determined equilibrium constants. Without removal of the condensate, the system was shown to be completely reversible when changing the temperature. Only at very low starting monomer concentrations, the formation of cyclics led