Letter pubs.acs.org/acscatalysis
Silane as Chain Transfer Agent for the Polymerization of Ethylene Catalyzed by a Palladium(II) Diimine Catalyst Michael G. Hyatt† and Damien Guironnet*,‡ †
Department of Chemistry, University of Illinois Urbana−Champaign, Urbana, Illinois 61801, United States Department of Chemical and Biomolecular Engineering, University of Illinois Urbana−Champaign, Urbana, Illinois 61801, United States
‡
S Supporting Information *
ABSTRACT: A strategy for the synthesis of semitelechelic polyethylene through the palladium diimine-catalyzed chain-transfer polymerization of ethylene using various silanes as chain-transfer agents is reported. Polymer molecular weight and end-group chemical structure can be tuned by varying the chain-transfer agent as well as its concentration. NMR spectroscopy confirms that the silicon of the chain-transfer agent is incorporated into the polymer. The stability of the catalyst toward polar monomer enables the chain-transfer polymerization of semitelechelic poly(ethylene-methyl acrylate) copolymers.
KEYWORDS: chain-transfer polymerization, silane, palladium, olefin, acrylate
A
a viable solution to this issue.15 Indeed, in a CTP, the use of a chain-transfer agent (CTA) results in the formation of multiple polymer chains per catalyst, with the CTA being incorporated as one of the end groups. For example, lanthanide based catalysts have been reported to catalyze the chain-transfer polymerizations of ethylene with electron-rich CTA such as amines and phosphines to yield the corresponding amine and phosphine-terminated polyethylenes.16,17 Electron-poor CTAs such as boranes, silanes, and alanes have also been successfully incorporated into the terminus of polyethylenes.15 Most of the examples of CTP involving olefins rely on oxophilic catalysts (early transition-metal and rare-earth-metal complexes). The oxophilicity of these catalysts severely limits the scope of functional groups that one can introduce at the end of a polyolefin. In fact, thus far, linear semitelechelic polyolefins produced by CTP have been reported to include at most one heteroatom per chain.15 We report here the first example of a palladium-catalyzed chain-transfer polymerization of ethylene with tertiary silanes as CTA using the palladium(II) diimine catalyst 1 (Figure 1). The tolerance of palladium(II) diimine catalysts toward functional groups enables the synthesis of hyperbranched polyethylene with a wide array of functionality.18−20 CTP with other late transition-metal-based catalysts (Ni, Fe, and Co) have been reported using zinc and aluminum alkyls as chaintransfer agents.21−23 While the metal employed for these
s more than 50% of the world’s polymer production consists of polyolefins, increasing the molecular precision of polyolefin synthesis has the potential to directly improve the physical and mechanical properties of most of the objects surrounding us.1,2 Telechelic and semitelechelic polymers, polymers with one and two functional end-groups, can be used as building blocks for the synthesis of such advanced polymers.3,4 Telechelic polymers are most commonly synthesized by living radical or living anionic polymerizations. These polymerization methods, however, are not compatible with simple olefins, and therefore, telechelic polyolefins require alternative strategies to access directly.5−11 Different approaches have been reported for the synthesis of telechelic polyolefins. The selective quenching of the anionic polymerization of butadiene with a capping agent12 (e.g., oxygen, carbon dioxide, or epoxide) followed by the hydrogenation of the polymer13 to yield telechelic polyethylenes (PEs) remains the most common technique. The ring-opening metathesis polymerization of cyclic olefins in the presence of a chain-transfer agent also results in telechelic polyolefins.14 While very elegant, these approaches do not employ the widely available ethylene and propylene monomers. Mecking et al. recently reported the direct synthesis of telechelic polyethylene by copolymerization of 2-vinylfuran with ethylene.6 The unique reactivity of the catalyst toward the comonomer enables the exclusive formation of polyethylene with difuran end-groups. The living catalytic polymerization of olefin has been reported to yield semitelechelic polyolefins.5 The main drawback of this approach, however, is that each molecule of catalyst yields only one polymer chain. Chain-transfer polymerization (CTP) offers © XXXX American Chemical Society
Received: April 22, 2017 Revised: July 21, 2017
5717
DOI: 10.1021/acscatal.7b01299 ACS Catal. 2017, 7, 5717−5720
Letter
ACS Catalysis
that occur prior to chain-transfer (Scheme 1). Therefore, performing the reaction at a lower silane concentration was expected to result in the synthesis of a silane-terminated polyethylene instead of the hydrosilylation product. Indeed, a series of polymerizations performed in the presence of lower amounts of silane yielded hyperbranched polyethylenes. Molecular weight analysis by gel permeation chromatography of the polymers showed a systematic decrease in the molecular weight with increasing silane concentration and a broadening of the molecular weight distribution (Table 1). It is worth noting
Figure 1. Catalyst 1.
Table 1. Effect of HSiEt3 Concentration on Mn with Catalyst 1a
catalysts is less electronegative, the oxophilicity of the CTAs limits the diversity of functional groups that can be introduced as polymer end-groups. Mechanistically, in a metal-mediated chain-transfer polymerization, the chain-transfer agent cleaves the polymer chain from the metal center and yields a new complex capable of reinitiating the polymerization.15 Triethylsilane has previously been used to cleave PE chains from living palladium(II) diimine catalyzedethylene polymerizations, resulting in PE with saturated end groups24 and presumably in the formation of the corresponding palladium silyl complex. This reaction represents the first step of a chain-transfer polymerization. A phenanthroline-based palladium(II) catalyst was reported to be active toward the hydrosilylation of olefins.25,26 In the hydrosilylation reaction mechanism, a palladium silyl catalyst formed in situ inserts one olefin. Recently Brookhart et al. has also demonstrated that a diimine palladium silyl complex can insert olefin.27 These insertions correspond to the second step of a chain-transfer polymerization mechanism. These reports prompted us to test if tertiary silanes could be used as CTA for palladium(II) diimine-catalyzed olefin polymerizations. In a preliminary experiment, catalyst 1 and 311 mM (835 equiv) HSiEt3 in dichloromethane were combined and pressurized to 40 psi of ethylene at room temperature. After 16 h, GC analysis of the crude mixture showed the catalytic formation of SiEt4 (47 TON), the hydrosilylation product of ethylene. We also observed a small amount of SiEt3butyl (0.3 TON) which results from the insertion of two ethylene molecules before silane addition (Scheme 1). This successful hydrosilylation experiment
entry
HSiEt3 (mM)
yield (g)
TOFb h−1
Mnc g/mol
Đ
1 2 3 4 5 6 7 8
0.00 3.35 6.68 8.89 13.24 19.89 26.16 32.73
2.77 2.61 2.47 2.73 2.65 2.23 2.42 2.45
354 331 315 349 338 285 309 311
158 700d 71 900 40 800 41 300 23 500 18 300 14 100 11 500
1.14 1.49 1.61 1.64 1.50 1.56 1.60 1.85
Conditions: 17.5 μmol of catalyst 1; 48 mL of chlorobenzene; 5 °C; 200 psi ethylene; reacted for 16 h. bTOF is based on gravimetric yield of polymer and moles of catalyst used. cMn was determined by GPC. d Mn was determined based upon polymer yield and moles of catalyst added (see Supporting Information). a
that in absence of silane, entry 1, the polymerization is living as illustrated by the narrow molecular weight distribution of the polymer.28 This means that in the absence of chain-transfer agent, only one polymer chain is formed per metal center. The large decrease in polymer molecular weight in the presence of silane and only a small decrease in catalyst activity imply that multiple polymer chains are produced per metal center, showing that silane does indeed act as a CTA. Based on the kinetics of a chain-transfer polymerization, the degree of polymerization (DP) should follow the Mayo equation (eq 1) under steady-state conditions and in the absence of any other chain termination mechanisms.29 Here kp and ktr represent the rate constants for chain propagation and chain-transfer, respectively. The order of ethylene dependence for the CTP is not known for this system and therefore was left to an unknown power of x. Further mechanistic studies are currently being performed, including the ethylene dependency, and will be published later. Nonetheless, by performing all the polymerizations at constant temperature and pressure the [ethylene]x term is a constant throughout the study. Equation 1. Theoretical Dependence of DP on [Silane]
Scheme 1. Proposed Mechanism of Hydrosilylation/ Polymerization
k tr[silane] 1 1 = + DP DPO k p[ethylene]x
(1)
Applying eq 1 to the polymerization data presented in Table 1 shows a linear relationship between the inverse of degree of polymerization and the silane concentration as seen in Figure 2. This confirms that HSiEt3 is a well-behaved CTA under these reaction conditions. Kinetic information about the system can also be determined as the inverse of the slope of the line corresponds to the ratio ktr/kp[ethylene]x ∼ 0.069, which suggests that the rate constants of chain transfer and chain propagation are within an order of magnitude. This is significantly different from lanthanide and group-IV-mediated CTP with silane, where propagation is significantly faster than chain transfer.30
confirmed that the in situ-formed palladium(II) diimine silyl complex can indeed insert ethylene, and thus, we postulated that tertiary silanes could be used as CTAs for palladium(II) diiminecatalyzed olefin polymerizations. The main difference between a hydrosilylation reaction and a chain-transfer polymerization is the number of olefin insertions 5718
DOI: 10.1021/acscatal.7b01299 ACS Catal. 2017, 7, 5717−5720
Letter
ACS Catalysis
by integrating this signal versus the protons of the backbone as well as accounting for polymers formed from the initial amount of 1 to give a Mn = 38 400 g/mol, which is in very good agreement with the value determined by GPC = 38 900 g/mol, hence confirming that the polymer is indeed silicon end-functionalized. The tunability and robustness of this chain-transfer polymerization technique was demonstrated by varying the substituent groups on the silane. At equal concentration and under identical reaction conditions, various silanes were investigated as shown in Table 2. The decrease of the silane steric bulk, substituting an Table 2. Comparing the Effect of Various Silanes on Mna
Figure 2. Relationship between DP and triethylsilane concentration.
This is of importance as slow chain transfer requires the use of a larger amount of CTA from which only a small fraction will be consumed. Having established that chain transfer occurs in the presence of silane, we proceeded to prove that the polymer was indeed silicon end modified. We attempted to isolate the palladium silyl precursor to showcase that it can perform the living polymerization of ethylene similarly to 1. While the formed palladium silyl complex appears to be stable under polymerization conditions, multiple attempts to isolate and characterize the in situ made palladium silyl complex remain unsuccessful. At room temperature, the addition of 2 equiv of HSiEt3 to a dichloromethane-d2 solution of 1 resulted in the detection of methane and partial consumption of 1 as well as in the near quantitative consumption of the silane. The palladium silyl presumably formed could not be detected by 1H NMR as it decomposed too rapidly to palladium black. This palladium black is responsible for the catalytic conversion of the HSiEt3 into H2 and Et3Si-SiEt3. (It is important to note that H2 formation is a safety concern as it could result in an increase of the pressure of the reaction vessel). The possibility of H2 formation during the polymerization raised the question of whether silane was the true CTA and if the polymer had the desired silicon end group. Indeed, H2 is a very good chain-transfer agent for olefin polymerization and thus in situ formed H2 during the polymerization could be the culprit for the decrease of molecular weight as a result of the silane addition.31 The only difference between the product of CTP by H2 and silane is the identity of the end-group. If silane is indeed a CTA then the polyethylene formed should contain a silicon end group. Therefore, analyzing the polymer end-groups proved to be critical. The low molecular weight branched polymer formed could not be easily precipitated and thus separated from Et3Si-SiEt3, a byproduct formed after ethylene pressure is released. The 1H NMR methylene signals of Et3Si-SiEt3 overlap with the expected 1H signals of the PE-CH2Si(CH2-CH3)3 making assignments problematic. Therefore, polymers with a more linear topology were made at higher ethylene pressure as they could be precipitated and thus separated from any undesired silane containing small molecule.32 A polyethylene made at 400 psi with HSiEt3 was characterized by {1H, 29Si} gHMBC NMR spectroscopy (cf. Supporting Information). Two cross-peaks corresponding to a tetraalkyl silane molecule were detected. While the methyl protons overlapped with the PE signals, the methylene protons of the PE-CH2-Si(CH2-CH3)3 could be detected in the 1H NMR at δ = 0.52 ppm. The molecular weight of the polymer was determined
entry
silane
yield (g)
TOFb (h−1)
Mnc (g/mol)
Đ
1d 2 3 4
HSiEt3 HSiEt2Me HSi(ipr)3 HSi(OEt)3
2.65 2.38 2.67 2.07
338 302 340 264
23 500 13 100 179 000 51 700
1.50 1.70 1.13 1.50
a Conditions: 17.5 μmol of catalyst 1; 48 mL of chlorobenzene; 13.24 mM silane; 5 °C; 200 psi ethylene; reacted for 16 h. bTOF is based on gravimetric yield of polymer and moles of catalyst used. cMn was determined by GPC. dIncluded for comparison.
ethyl group for a methyl group (Table 2, entries 1 vs 2) led to a faster rate of chain transfer while increasing the steric bulk of the alkyl substituents, substituting ethyl groups for isopropyl groups, led to the complete inhibition of the chain-transfer reaction (Table 2, entry 3). The ability to tune the rate of chain transfer through steric variations and concentration of the silane enables a fine control of the degree of polymerization. In Table 2, approximately 15.0% of the HSiEt3 is consumed based on the yield and Mn, while 25.9% of the faster HSiEt2Me is consumed. This attribute could be of interest in the case of implementing an expensive CTA by minimizing the amount of unreacted CTA. Finally, for the first time, alkoxysilane (Table 2, entry 4) could be used as a chain-transfer agent, resulting in a PE made via CTP that contains multiple heteroatoms per chain. As palladium diimine catalysts are known to copolymerize ethylene and methyl acrylate, we investigated whether our chaintransfer polymerization system was compatible with the insertion of polar comonomer to further showcase the functional group tolerance of the new chain-transfer polymerization method.33−37 In a control experiment, ethylene and methyl acrylate (MA) copolymerization yielded a polymer with a Mn of 31 900 g/mol and a dispersity of 1.40, while in the presence of HSiEt3 the Mn decreases to 9000 g/mol and the molecular weight distribution increased to 1.56 As the TOF decreases somewhat from 141 h−1 for the control, to 104 h−1 in the presence of HSiEt3, this shows the system is still able to perform chain transfer in the presence of polar monomer. (cf. Supporting Information). 1H DOSY NMR of the copolymer showed that the silane and acrylate moieties had the same diffusivity coefficient, confirming that the polymer contains both silicon and methyl ester groups (cf. Supporting Information). Determination of Mn by NMR gives a value of 8550 g/mol which is again in good agreement with that determined by GPC. The ability of this system to perform CTP in the presence of polar monomers and heteroatoms is a clear advantage of using a LTM over oxophillic catalysts. Interestingly, in this system the decrease in Mn was actually followed by a decrease in the incorporation of MA from 0.7 to 0.45 mol% upon addition of silane. We hypothesize that this difference is caused by the competition between the coordination of the silane and the methyl acrylate that results in an overall lower acrylate insertion. 5719
DOI: 10.1021/acscatal.7b01299 ACS Catal. 2017, 7, 5717−5720
Letter
ACS Catalysis
(10) Patil, V. B.; Saliu, K. O.; Jenkins, R. M.; Carnahan, E. M.; Kramer, E. J.; Fredrickson, G. H.; Bazan, G. C. Macromol. Chem. Phys. 2014, 215, 1140−1145. (11) Franssen, N. M. G.; Reek, J. N. H.; de Bruin, B. Chem. Soc. Rev. 2013, 42, 5809−5832. (12) Goldberg, E. J. Hydroxyl-Terminated Polymers. US3055952A, Sept 25, 1962. (13) Shiono, T.; Naga, N.; Soga, K. Makromol. Chem., Rapid Commun. 1991, 12, 387−392. (14) Pitet, L. M.; Hillmyer, M. A. Macromolecules 2011, 44, 2378− 2381. (15) Amin, S. B.; Marks, T. J. Angew. Chem., Int. Ed. 2008, 47, 2006− 2025. (16) Amin, S. B.; Marks, T. J. J. Am. Chem. Soc. 2007, 129, 10102− 10103. (17) Kawaoka, A. M.; Marks, T. J. J. Am. Chem. Soc. 2004, 126, 12764− 12765. (18) Chen, Y.; Wang, L.; Yu, H.; Zhao, Y.; Sun, R.; Jing, G.; Huang, J.; Khalid, H.; Abbasi, N. M.; Akram, M. Prog. Polym. Sci. 2015, 45, 23−43. (19) Guo, L.; Chen, C. Sci. China: Chem. 2015, 58, 1663−1673. (20) Guo, L.; Dai, S.; Sui, X.; Chen, C. ACS Catal. 2016, 6, 428−441. (21) Valente, A.; Mortreux, A.; Visseaux, M.; Zinck, P. Chem. Rev. 2013, 113, 3836−3857. (22) Hue, R. J.; Cibuzar, M. P.; Tonks, I. A. ACS Catal. 2014, 4, 4223− 4231. (23) Van Meurs, M.; Britovsek, G. J. P.; Gibson, V. C.; Cohen, S. A. J. Am. Chem. Soc. 2005, 127, 9913−9923. (24) Gottfried, A. C.; Brookhart, M. Macromolecules 2003, 36, 3085− 3100. (25) Lapointe, A. M.; Rix, F. C.; Brookhart, M. J. Am. Chem. Soc. 1997, 119, 906−917. (26) Widenhoefer, R. A. Acc. Chem. Res. 2002, 35, 905−913. (27) Chen, Z.; Liu, W.; Daugulis, O.; Brookhart, M. J. Am. Chem. Soc. 2016, 138, 16120−16129. (28) Gottfried, A. C.; Brookhart, M. Macromolecules 2001, 34, 1140− 1142. (29) Mayo, F. R. J. Am. Chem. Soc. 1943, 65, 2324−2329. (30) Fu, P.-F.; Marks, T. J. J. Am. Chem. Soc. 1995, 117, 10747−10748. (31) Stürzel, M.; Mihan, S.; Mülhaupt, R. Chem. Rev. 2016, 116, 1398− 1433. (32) Guan, Z.; Cotts, P. M.; McCord, E. F.; McLain, S. J. Science 1999, 283, 2059−2062. (33) Johnson, L. K.; Mecking, S.; Brookhart, M. J. Am. Chem. Soc. 1996, 118, 267−268. (34) Dai, S.; Chen, C. Angew. Chem., Int. Ed. 2016, 55, 13281−13285. (35) Dai, S.; Sui, X.; Chen, C. Angew. Chem., Int. Ed. 2015, 54, 9948− 9953. (36) Dai, S.; Zhou, S.; Zhang, W.; Chen, C. Macromolecules 2016, 49, 8855−8862. (37) Zhai, F.; Solomon, J. B.; Jordan, R. F. Organometallics 2017, 36, 1873−1879. (38) Mecking, S.; Johnson, L. K.; Wang, L.; Brookhart, M. J. Am. Chem. Soc. 1998, 120, 888−899.
Higher methyl acrylate incorporation was achieved by lowering the pressure of ethylene. Performing the reaction in the presence of silane at 100 psi instead of 200 psi of ethylene gives a copolymer with 0.89 mol% MA with a further decrease in Mn to 4200 g/ mol; however, TOF drops significantly to 19 h−1. This drop in activity is consistent with the previously reported copolymerization results with the same catalyst.38 In conclusion, we have demonstrated the first example of silane as a chain-transfer agent for the polymerization of olefin by a cationic palladium(II) diimine catalyst. The rate of chain transfer can be precisely controlled by varying the concentration and substituents of the silane. The merit of using late transition metal catalysts for the polymerization of olefin is their compatibility with heteroatoms. This strength is highlighted by the use of alkoxysilanes as CTA as well as the compatibility of the system to perform ethylene polar monomer copolymerizations. An in-depth kinetic analysis of the rates involved as well as expanding this technique to other late-transition-metal-based olefin polymerization catalysts is under way.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acscatal.7b01299. All experimental methods, materials, and characterization (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Damien Guironnet: 0000-0002-0356-6697 Notes
The authors declare the following competing financial interest(s): An invention disclosure related to this work has been filed: Guironnet, D; Hyatt, G. M.; U.S. Patent Application filed November 2016 (serial number 62/420,819).
■
ACKNOWLEDGMENTS The authors thank the department of Chemical and Biomolecular Engineering at the University of Illinois, Urbana−Champaign for a startup fund.
■
REFERENCES
(1) Mazzolini, J.; Espinosa, E.; D’Agosto, F.; Boisson, C. Polym. Chem. 2010, 1, 793−800. (2) Eagan, J. M.; Xu, J.; Di Girolamo, R.; Thurber, C. M.; Macosko, C. W.; Lapointe, A. M.; Bates, F. S.; Coates, G. W. Science 2017, 355, 814− 816. (3) Tasdelen, M. A.; Kahveci, M. U.; Yagci, Y. Prog. Polym. Sci. 2011, 36, 455−567. (4) Lo Verso, F.; Likos, C. N. Polymer 2008, 49, 1425−1434. (5) Zhao, Y.; Wang, L.; Xiao, A.; Yu, H. Prog. Polym. Sci. 2010, 35, 1195−1216. (6) Jian, Z.; Falivene, L.; Boffa, G.; Sánchez, S. O.; Caporaso, L.; Grassi, A.; Mecking, S. Angew. Chem., Int. Ed. 2016, 55, 14378−14383. (7) German, I.; Kelhifi, W.; Norsic, S.; Boisson, C.; D’Agosto, F. Angew. Chem., Int. Ed. 2013, 52, 3438−3441. (8) Pitet, L. M.; Amendt, M. A.; Hillmyer, M. A. J. Am. Chem. Soc. 2010, 132, 8230−8231. (9) Shea, K. J.; Walker, J. W.; Zhu, H.; Paz, M.; Greaves, J. J. Am. Chem. Soc. 1997, 119, 9049−9050. 5720
DOI: 10.1021/acscatal.7b01299 ACS Catal. 2017, 7, 5717−5720
