Xing
Wang
b,
Yanping
Zhang
b and
Miao
Hong
*ab
aSchool of Chemistry and Material Sciences, Hangzhou Institute for Advanced Study, University of Chinese Academy of Sciences, 1 Sub-lane Xiangshan, Hangzhou 310024, China
bState Key Laboratory of Organometallic Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032, China. E-mail: miaohong@sioc.ac.cn
First published on 14th June 2023
This contribution presents the development of a general Lewis pair (LP) catalyst for efficient and/or controlled polymerizations of inert biomass-derived acrylic monomers, including methyl crotonate (MC), (E,E)-methyl sorbate (MS), and β-angelica (β-AL). Through a comprehensive study on the interaction between bis(aryloxide)alkylaluminum Lewis acids (LAs) and N-heterocyclic carbene (NHC) Lewis bases (LBs), a new frustrated Lewis pair (FLP) has been constructed which comprises MeAl(BHT)2 (BHT: 2,6-di-tert-butyl-4-methylphenoxide) and 1,3-di-tert-butyl-4,5-dimethylimidazol-2-ylidene (Me-ItBu). Such a FLP can mediate efficient polymerizations of MS, MC, and β-AL regardless of the addition sequence, on account of its stability without the formation of abnormal Lewis adducts and the noninteracting FLP feature that enables sufficiently “free” LAs and LBs with suitable steric hindrance for catalysis. Moreover, a high degree of control over the polymerization of MS has also been achieved using the MeAl(BHT)2/Me-ItBu FLP, affording PMSs with high Mn up to 600.3 kg mol−1. An exclusive initiation via a basic mechanism in the polymerization of MS has been revealed, leading to the formation of linear PMSs with a unique conjugated diene chain end that brought about significantly enhanced thermal stability of the resultant PMSs.
Scheme 1 Structures of inert biomass-derived acrylic monomers, NHC LBs, and bis(aryloxide)alkylaluminum LAs investigated in this work, as well as the different interaction modes between LAs and LBs. |
Lewis pair polymerization (LPP), catalyzed by non-interacting frustrated Lewis pairs (FLPs) or classical Lewis adducts (CLAs) that can dissociate into FLPs in the presence of a suitable solvent or monomer, has been demonstrated to be powerful in mediating the polymerizations of polar monomers due to synergistic/cooperative monomer activation by both Lewis base (LB) and Lewis acid (LA) sites of the Lewis pair (LP) catalyst.21–24 Since Chen and co-workers uncovered the first LPP in 2010,25 significant advances have been made in the efficient and/or controlled/living LPPs of various petroleum-based polar vinyl monomers (e.g. methacrylates,26–33 acrylates,34–37 acrylamides,38,39 divinyl monomers,39–43 dialkyl vinylphosphonates,44,45 vinyl pyridines,46,47 and 2-isopropenyl-2-oxazoline47).
Recently, the scope of polar vinyl monomers has also been extended to biomass-derived monomers,14–20,48–51 especially the challenging ones that are generally inaccessible or efficiently realized by traditional polymerization techniques. For example, Chen et al. reported that MC can be polymerized into a high-MW polymer (Mn = 161.0 kg mol−1, Đ = 1.62, TOF = 41.7 h−1) by a LP catalyst consisting of an N-heterocyclic carbene (NHC) LB 1,3,4-triphenyl-4,5-dihydro-1H-1,2,4-triazol-5-ylidene (TPT) and the sterically encumbered methylaluminum bis(2,6-di-tert-butyl-4-methylphenoxide) [MeAl(BHT)2, Scheme 1] LA.14 Takasu et al. showed that the cooperativity of the MeAl(BHT)2 LA with the 1,3-di-tert-butylimidazolin-2-ylidene (ItBu, Scheme 1) NHC LB was capable of mediating the exclusive 1,4-addition polymerization of MS and also controlling the topology to produce a cyclic polymer (Mn = 28.7 kg mol−1, Đ = 1.37, TOF = 4.2 h−1, −20 °C).15,16 In sharp contrast, the LB alone only led to dimerization (MC) or oligomerization (MS). Subsequently, the two-step H-transfer cyclization mechanism in LPP of MS was established by Chen et al., thereby achieving the precision synthesis of cyclic polymers (e.g. c-PMS and c-PMMA) with predictable Mn and low Đ values.17 Recently, Zhang, Chen, and coworkers reported that the combination of an N-heterocyclic olefin (NHO) LB with PhAl(BHT)2 or iBuAl(BHT)2 (Scheme 1) enabled the living/controlled polymerization of (E,E)-alkyl sorbates at room temperature (RT), leading to the formation of linear PMSs with Mn up to 287.0 kg mol−1 (Đ = 1.16, TOF = 66.7 h−1) which can subsequently convert into intriguing alternating polymers via simple post-functionalization.18 Shortly after, a B(C6F5)3-based organic LP catalyst was also developed by the same group for topology-controlled 1,4-addition polymerization of (E,E)-alkyl sorbates.19 In the case of β-AL, we reported the first polymerization of β-AL by constructing an efficient LP catalyst based on the MeAl(BHT)2 LA and the 1,3-diisopropyl-4,5-dimethylimidazol-2-ylidene (IiPr, Scheme 1) NHC LB, providing a novel acrylic bioplastic with high heat and solvent resistance.20
As revealed by the above overview, LPP has been demonstrated to be an effective methodology for polymerizing MC, MS, and β-AL. However, the development of a LP catalyst, which is general to these inert biomass-derived acrylic monomers, has remained an unmet challenge. In this work, the comprehensive investigation on the interaction between RAl(BHT)2 (R = CH3, Et, iBu) LAs and NHC LBs has led to the disclosure of a new and stable FLP which behaves as a general catalyst for efficient LPPs of MS, MC, and β-AL. The precise control over the molecular weight in LPP of MS and the exclusive basic initiation mechanism have also been revealed.
Run | LA | LB | M. | [M.]0:[LA]0:[LB]0 | Time | Conv.b (%) | M n, GPCc (kg mol−1) | Đ | I* (%) |
---|---|---|---|---|---|---|---|---|---|
a Conditions: the polymerizations were carried out by premixing the LA and the monomer first followed by adding the LB unless otherwise noted: RT, LB = 12 μmol, [β-AL]0 = 0.6 M in CH2Cl2, [MC]0 = 3 M in toluene, [MS]0 = 1 M (runs 15–17), 3 M (run 18), 6 M (run 19), 8 M (runs 20 and 21) in toluene. b Monomer conversion measured by 1H NMR spectroscopy using mesitylene as the internal standard. c Number-average molecular weight (Mn) and molecular weight distribution (Đ = Mw/Mn) determined by GPC at 40 °C coupled with a multi(18)-angle light scattering detector and a refractive-index detector [eluent: β-AL (DMF), MC (DCM), MS (DCM)]. d Initiation efficiency (I*)% = Mn(calcd)/Mn(exptl) × 100, where Mn(calcd) = [MW(monomer)] × ([monomer]0/[LB]0) × conv.% + MW(end groups). e The polymerization was carried out by premixing the LA and LB first for 30 min followed by addition of the monomer. f About 10% of monomers were converted into dimers, rather than polymers. | |||||||||
1 | MeAl(BHT)2 | ItBu | β-AL | 200:2:1 | 20 min | >99 | 12.9 | 1.49 | 152.0 |
2e | MeAl(BHT)2 | ItBu | β-AL | 200:2:1 | 24 h | 10f | — | — | — |
3 | MeAl(BHT)2 | ItBuiPr | β-AL | 200:2:1 | 210 min | >99 | 3.7 | 1.25 | 530.0 |
4 | MeAl(BHT)2 | ItBuMe | β-AL | 200:2:1 | 40 min | >99 | 5.2 | 1.28 | 377.0 |
5 | MeAl(BHT)2 | Me-ItBu | β-AL | 200:2:1 | 20 min | >99 | 11.8 | 1.60 | 164.7 |
6e | MeAl(BHT)2 | Me-ItBu | β-AL | 200:2:1 | 20 min | >99 | 12.0 | 1.50 | 163.3 |
7 | iBuAl(BHT)2 | Me-ItBu | β-AL | 200:2:1 | 10 min | >99 | 13.0 | 1.31 | 150.8 |
8 | iBuAl(BHT)2 | Me-ItBu | β-AL | 300:4:1 | 5 min | >99 | 21.0 | 1.34 | 93.3 |
9 | MeAl(BHT)2 | ItBuiPr | MC | 200:2:1 | 16 h | >99 | 8.4 | 1.95 | 220.0 |
10 | MeAl(BHT)2 | ItBuMe | MC | 200:2:1 | 14 h | >99 | 9.1 | 2.13 | 238.4 |
11 | MeAl(BHT)2 | Me-ItBu | MC | 200:2:1 | 7 h | >99 | 16.7 | 1.77 | 120.0 |
12e | MeAl(BHT)2 | Me-ItBu | MC | 200:2:1 | 7 h | >99 | 18.0 | 1.74 | 111.1 |
13 | iBuAl(BHT)2 | Me-ItBu | MC | 200:2:1 | 24 h | 49.5 | 5.8 | 1.71 | 170.2 |
14 | MeAl(BHT)2 | Me-ItBu | MC | 200:4:1 | 3 h | >99 | 25.5 | 1.66 | 78.5 |
15 | MeAl(BHT)2 | Me-ItBu | MS | 100:2:1 | 0.5 h | >99 | 26.1 | 1.09 | 48.3 |
16e | MeAl(BHT)2 | Me-ItBu | MS | 100:2:1 | 0.5 h | >99 | 25.2 | 1.09 | 50.3 |
17 | MeAl(BHT)2 | Me-ItBu | MS | 400:3:1 | 3.5 h | >99 | 74.7 | 1.20 | 67.6 |
18 | MeAl(BHT)2 | Me-ItBu | MS | 800:8:1 | 1 h | >99 | 104.3 | 1.23 | 96.8 |
19 | MeAl(BHT)2 | Me-ItBu | MS | 1200:24:1 | 1 h | >99 | 156.6 | 1.25 | 96.7 |
20 | MeAl(BHT)2 | Me-ItBu | MS | 2000:40:1 | 1 h | >99 | 308.2 | 1.41 | 82.0 |
21 | MeAl(BHT)2 | Me-ItBu | MS | 4000:80:1 | 3 h | >99 | 600.3 | 1.65 | 84.1 |
To gain mechanistic insights into this abnormal phenomenon, the stoichiometric reaction of MeAl(BHT)2 and ItBu in benzene-d6 at RT was monitored by 1H NMR. As shown in Fig. S1,† at the early stage of the reaction (within 30 min), MeAl(BHT)2 and ItBu behaved as a noninteracting FLP because the chemical resonances were hardly changed in 1H NMR spectra before and after the reaction. However, besides an FLP, a new species could also be detected. In fact, the MeAl(BHT)2/ItBu FLP gradually converted into this new species with a prolonged reaction time, and the complete consumption of the FLP was observed after 5 h (Fig. S1c vs. d vs. f†). Slow evaporation of the solvent of the resultant mixture afforded single crystals suitable for X-ray diffraction (SC-XRD) analysis. The molecular structure reveals that this new species is attributed to the “abnormal” Lewis adduct (ALA-1, Fig. 1), as a result of the attack of MeAl(BHT)2 at the 4-position of the ItBu backbone and subsequent migration of the corresponding hydrogen atom to the former carbene carbon atom. Overall, the bond lengths around the Al center in the “abnormal” adduct are nearly identical to those in the MeAl(BHT)2/IiPr normal adduct,20 except for a noticeably short Al–C distance between MeAl(BHT)2 and NHC moieties in the “abnormal” adduct (2.0480 Å) compared to that in the MeAl(BHT)2/IiPr adduct (2.1223 Å),20 indicative of a stronger Al–C α-bond in the “abnormal” adduct. Moreover, due to the steric repulsion between tert-butyl groups of the ItBu moiety and the tert-butyl group of the BHT ligand at O(1), the rotation of the Al–C(1) bond is severely hindered. Accordingly, as shown in 1H and 13C NMR spectra (Fig. S2 and 3†), two tert-butyl groups in the ItBu moiety are chemically inequivalent at RT on the NMR time scale (1H NMR: 1.36 vs. 0.66 ppm, 13C NMR: 59.33 vs. 56.85 ppm). The upfield shift of the tert-butyl group at N(2) is presumably due to the shielding effect imposed by the fixed BHT ligand at O(2).
It is worth mentioning that the formation of the “abnormal” adduct is irreversible at RT or high temperature (80 °C). When the isolated ALA-1 was attempted for β-AL polymerization, no polymerization was observed, indicative of the complete loss of reactivity once formed. The stoichiometric reactions of ItBu with sterically more demanding EtAl(BHT)2 and iBuAl(BHT)2 also led to the formation of ALAs (Scheme 1), which were fully characterized by single-crystal X-ray diffraction analysis (Fig. 1) and NMR spectroscopy (Fig. S4–7†). The noticeable difference in these molecular structures is that Al(1)–C(1) bond lengths in ALA-2 (2.0627 Å) and ALA-3 (2.0533 Å) are slightly longer than that in ALA-1 (2.0480 Å). The characteristic signal attributed to –C+H– in these ALAs was observed at similar resonances [7.14–7.16 ppm (1H NMR) and 126.6–126.7 ppm (13C NMR)]. The formation of an ALA has also been reported previously in B(C6F5)3/ItBu which led to a loss of the reactivity towards H2.52 However, this interaction mode has not been observed in Al-based LPs which either form CLAs [Al(C6F5)3/ItBu,25 Al(C6F5)3/NHO,26 MeAl(BHT)2/IiPr,20 AlPh3/PMe344] or FLPs [e.g. MeAl(BHT)2/NHO27]. The reason for RAl(BHT)2/ItBu to form an ALA is presumably twofold: reactive C–H bonds in ItBu and high reactivity of sterically encumbered noninteracting RAl(BHT)2 and ItBu.
To shut down the undesired “abnormal” interaction, two different strategies have been explored in this work. The first one is to replace ItBu with a less sterically hindered NHC LB. In this context, dissymmetric ItBuiPr and ItBuMe, where one of the tert-butyl groups of ItBu was substituted with a smaller iso-propyl or Me group (Scheme 1), were utilized as LBs. The stoichiometric reaction revealed that the MeAl(BHT)2 LA selectively interacts with the carbine center of LBs, quantitatively yielding clean CLAs as confirmed by a significant change of chemical resonances after the reaction (Fig. S8–11†). For example, the signal of Al-CH3 in the MeAl(BHT)2 moiety shifts downfield from −0.26 ppm to 0.38 ppm (CLA-1) and 0.33 ppm (CLA-2), while the signals of –HCCH– in the NHC moiety shift upfield (CLA-1: 6.35, 6.13 ppm; ItBuiPr: 6.75, 6.55 ppm; CLA-2: 6.20, 5.65 ppm; ItBuMe: 6.67, 6.34 ppm) after the reaction. The formation of a CLA between MeAl(BHT)2 and ItBuiPr was further confirmed by SC-XRD analysis (Fig. 2). The bond length of Al–C1 in CLA-1 (2.148 Å) is identical to that in the MeAl(BHT)2/IiPr CLA (2.1223 Å),20 but noticeably longer than those in ALAs (2.0480–2.0627 Å, vide supra), indicative of a dative Al–C bonding distance. It is noteworthy that these CLAs are stable in the solution or solid state at RT, where their conversion into an ALA is undetectable.
Fig. 2 X-ray crystal structure of CLA-1 obtained by the stoichiometric reaction of ItBuiPr and MeAl(BHT)2. Hydrogen atoms have been omitted for clarity, and ellipsoids are drawn at 30% probability. |
The second strategy is to develop an NHC that does not provide reactive C–H bonds. Accordingly, Me-ItBu (Scheme 1), with 4,5-positions blocked by Me groups, was then selected to examine the interaction with RAl(BHT)2 (R = Me, Et, iBu). The monitoring of stoichiometric reactions by 1H NMR indicated that RAl(BHT)2 and Me-ItBu behaved as noninteracting FLPs because the chemical resonances hardly changed before and after the reaction (Fig. S12–14†). Different from the MeAl(BHT)2/ItBu FLP that gradually converts into an ALA, Me-ItBu/RAl(BHT)2 FLPs are stable and essentially retain the orthogonal reactivity of both LA and LB sites. Therefore, these new FLPs are potentially general catalysts for efficient LPPs of inert biomass-derived acrylic monomers (MS, MC, and β-AL).
In the case of the MC monomer, on fixing MeAl(BHT)2 as the LA, the influence of LB structures on the polymerization behavior was identical to that in LPPs of β-AL, and MeAl(BHT)2/Me-ItBu (FLP-1) stood out as the best LP catalyst (Table 1, runs 9–11). Moreover, in FLP-1-mediated LPP, there were also essentially no effects of the addition sequence on the polymerization outcome (Table 1, run 11 vs. 12). Different from the LPP of β-AL, the attempt to enhance the activity and Mn through the utilization of FLP-3 with the sterically more demanding iBuAl(BHT)2 LA was unsuccessful (Table 1, run 13), presumably due to the impeded activation and conjugate-addition of the MC monomer by the crowded LA center. Increasing the MeAl(BHT)2:Me-ItBu ratio from 2:1 to 4:1 allowed a noticeable increase of activity and Mn (Table 1, run 14), achieving quantitative monomer conversion in 3 h and affording PMC with a relatively high Mn of 25.5 kg mol−1.
Moving to the MS monomer, MeAl(BHT)2/Me-ItBu (FLP-1) exhibited high activity regardless of the addition sequence (Table 1, runs 15 and 16). More importantly, the resulting PMSs possessed very narrow dispersities (Đ = 1.09), revealing the controlled characteristics of this LPP. To verify it, matrix-assisted laser desorption/ionization time-of-flight mass spectroscopy (MALDI-TOF MS) analysis of the low-MW PMS produced by FLP-1 was performed. As shown in Fig. 3, only one set of molecular ion peaks was observed, attributed to the linear PMS with the CH2CHCHCH2CH(CO2Me)– initiation chain end and the H– termination chain end. This initiation chain end could be further confirmed from 1H and 1H–1H COSY NMR spectra of the low-MW PMS with the observation of characteristic chemical signals at 5.12, 6.26, 6.06, 5.58, and 3.02 ppm (Fig. S15 and 16†). These results indicate that the polymerization is initiated via a basic pathway through the deprotonation of MeAl(BHT)2-activated MS by Me-ItBu (Scheme 2), rather than a nucleophilic pathway, presumably due to the strong basicity of Me-ItBu. The formation of a clean [Me-ItBu-H]+/enolaluminate ion pair active species with E and Z isomers in a ratio of ca. 1:2 in the stoichiometric reaction between MeAl(BHT)2, Me-ItBu, and MS, as evidenced by the disappearance of the MS monomer signal and the appearance of the Me-ItBu-H+ signal (Fig. S17–19†), is another evidence for the basic deprotonation initiation mechanism. In short, the exclusive initiation mechanism and the absence of back-biting termination side reaction clearly confirm the controlled characteristics of this LPP.
Fig. 3 MALDI-TOF MS spectrum of the low-MW PMS and plots of m/z value (y) vs. the number of MS repeat units (x) for molecular ion peaks. |
Accordingly, with an increase in the [MS]0:[Me-ItBu]0 ratio from 100:1 to 1200:1 (Table 1, runs 15 and 17–19), efficient and controlled polymerizations were observed, which consumed all monomers within 1–3.5 h and afforded PMSs with a linear increase of Mn values and low Đ values (1.09–1.25, Fig. 4A) at moderate to near-quantitative initiation efficiency (I*: 48.3–96.8%). In addition, keeping [MS]0:[MeAl(BHT)2]0:[Me-ItBu]0 at a fixed ratio of 400:3:1, the Mn of the PMS increased linearly with an increase of the monomer conversion (Fig. 4B), while the corresponding dispersity remained at a very low value during polymerization (Đ: 1.09–1.20). In these cases, the GPC curves of all PMS samples showed narrow and unimodal distributions, and gradually shifted to the higher MW region with the increase of monomer conversion and the monomer-to-initiator ratio (Fig. 4C and D). It is noteworthy here that MeAl(BHT)2/Me-ItBu is especially effective for the polymerization of MS, in which 2000 equivalents of MS can be rapidly polymerized within 1 h (Table 1, run 20). Even though 4000 equivalents of MS were employed, quantitative MS conversion could also be accomplished within 3 h (Table 1, run 21). As a result, a high-Mn PMS up to 600.3 kg mol−1 with an isolated yield of 5.9 g could be readily prepared, despite a relatively broad Đ of 1.65 which should be caused by unfavorable monomer diffusion at the late stage of polymerization due to the increased viscosity of the high-MW polymer solution. Overall, the utilization of an excess amount of the LA relative to the LB is necessary to fully activate the monomer for high activity and suppress the chain transfer side reaction for controlled polymerization.
The thermal properties of the resultant PMSs were examined by differential scanning calorimetry (DSC) and thermal gravimetric analysis (TGA) analyses (Fig. 5). It was found that the glass transition temperatures (Tg) of PMSs were ca. 30 °C regardless of Mn, about 10 °C higher than those in previous reports.15,18 The onset degradation temperatures (Td, defined by the temperatures of 5% weight loss) increased from 270.3 to 283.5 °C when the Mn of the PMS increased from 25.2 to 308.2 kg mol−1. It is worth noting here that these Td values are about 70 °C higher than those in a previous report,18 indicative of significantly higher thermal stability of PMSs in this work.
Fig. 5 The second heating scans of DSC curves (left) and TGA curves (right) of PMSs with different Mn values. |
The above intriguing result prompted us to explore the influence of the polymer microstructure on the thermal stability. The 1H NMR spectrum of the representative PMS sample is shown in Fig. S20.† The characteristic resonance attributed to the methyl group (>CHCH3) in the 1,4-addition unit can be observed at 0.95 ppm, while the resonance attributed to the methyl group (CHCH3) in the 1,2-addition unit (1.23 ppm) is negligible, thus demonstrating that LPP of MS proceeds exclusively in a 1,4-addition manner. The appearance of splitting signals of –CHCOOCH3 and >CHCH3 in the 13C NMR spectrum (Fig. S21†) at 54.81–56.48 ppm and 38.99–39.77 ppm (ref. 55) indicated that the resultant PMS possesses a trans double bond. The trans structure was further confirmed by the FT-IR analysis (Fig. S22†) where the out-of-plane bending of –CHCH– at 969 cm−1, together with in-plane deformation and stretching vibration at 1075 and 1693 cm−1, was observed, respectively.55 To provide more information about the stereoregularity, the hydrogenation reaction of the PMS was performed. The 13C NMR spectrum of the resultant saturated PMS-H revealed the threo-rich structure of the PMS (threo:erythro = 82.7:27.3) with a diisotacticity of 52.4%.55 Overall, the PMS produced by the MeAl(BHT)2/Me-ItBu FLP has a trans, threo-rich, 1,4-addition structure, similar to that prepared by iBuAl(BHT)2 or PhAl(BHT)2/NHO.18 Therefore, the influence of the main-chain structure on the thermal stability can be ruled out. Considering that the PMS produced in this work has a conjugated diene chain-end structure generated by a basic initiation mechanism, different from the NHO+– chain end formed by the nucleophilic initiation pathway in a previous report,18 the unique chain-end structure is thereby speculated to be the key factor for enhanced thermal stability.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2239989–2239991 and 2239993. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3py00546a |
This journal is © The Royal Society of Chemistry 2023 |