Successful Cyclopolymerization of 1,6-Heptadiynes Using First

Apr 11, 2017 - Cyclopolymerization (CP) of 1,6-heptadiynes using olefin metathesis catalysts is a useful method for producing various conjugated polye...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/Macromolecules

Successful Cyclopolymerization of 1,6-Heptadiynes Using FirstGeneration Grubbs Catalyst Twenty Years after Its Invention: Revealing a Comprehensive Picture of Cyclopolymerization Using Grubbs Catalysts Cheol Kang, Eun-Hye Kang, and Tae-Lim Choi* Department of Chemistry, Seoul National University, Seoul 08826, Korea S Supporting Information *

ABSTRACT: Cyclopolymerization (CP) of 1,6-heptadiynes using olefin metathesis catalysts is a useful method for producing various conjugated polyenes. Unfortunately, commercially available user-friendly Grubbs catalysts have long been known to be inactive toward CP. However, recent mechanistic studies revealed that the problem did not lie with the intrinsic activities of Grubbs catalysts but the stability of the propagating carbenes, as decomposed carbene species catalyzed a [2 + 2 + 2] cycloaddition side reaction instead of CP. Fortunately, by adding weakly coordinating ligands such as pyridines as stabilizers, a highly active and fast-initiating third-generation Grubbs catalyst could successfully promote living CP. However, there was no report of CP using the much cheaper but less active first-generation Grubbs catalyst (G1), which has been widely used for more than 20 years. Believing that G1 should also be able to catalyze CP, we came up with three strategies to enhance the activity of G1 toward CP. By categorizing numerous additives into three distinct classes and conducting extensive reaction screening, we discovered two new excellent additives to G1: benzoic acid and sodium benzoate, both of which successfully produced various trans-selective conjugated polyenes with molecular weights of up to 23 kDa. Interestingly, additives optimal for CP catalyzed by G1 were not suitable for CP catalyzed by Grubbs catalysts containing N-heterocyclic carbene ligands and vice versa, thus implying that their activation mechanisms were distinctly different. Therefore, we conducted kinetic studies and mechanistic investigation by 1H and 31P NMR spectroscopy to reveal that these two additives, despite having very similar chemical structures, enhance the CP efficiency via very different mechanisms; benzoic acid accelerated phosphine dissociation and stabilized the propagating carbene, while sodium benzoate mediated the exchange of an anionic ligand to afford a more active catalyst. Additionally, these additives also suppressed or retarded the [2 + 2 + 2] cycloaddition side reaction to further enhance the efficiency and selectivity of CP. In brief, this study revealed new insights into the use of Grubbs catalysts to facilitate CP two decades after their invention, and this should greatly broaden the utility of CP because G1 is the cheapest and the most readily available olefin metathesis catalyst.



well-defined Schrock catalyst along with the concept of α- and β-addition to explain the mechanism and regioselectivity of the resulting polymer structures.10,11 Based on that study, both Schrock and Buchmeiser groups went on to ingeniously solve the regiochemistry problem by modifying the ligands on the

INTRODUCTION

Cyclopolymerization (CP) of terminal diynes via an olefin metathesis reaction is one of the most powerful tools for the synthesis of conjugated polymers by a chain-growth mechanism (Scheme 1A).1 In the early days, chemists used ill-defined catalysts such as Ziegler−Natta catalyst,2,3 MoCl5, Mo(CO)6, and WCl54−9 without sufficiently understanding the mechanistic details of CP. Subsequently, the Schrock group reported groundbreaking results on the first living CP process using a © XXXX American Chemical Society

Received: March 7, 2017 Revised: April 2, 2017

A

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

to a successful CP using Grubbs catalysts was not to enhance the activity of the catalysts, but all about the stabilization of the propagating carbenes in order to inhibit their decomposition and the competing [2 + 2 + 2] cycloaddition reaction. In this regard, we achieved a well-controlled living CP by using a fastinitiating third generation Grubbs catalyst (Figure 1, G3) in a weakly coordinating tetrahydrofuran (THF) solvent or with weakly coordinating pyridine additives as stabilizing ligands in noncoordinating solvents.28,29 These advances in CP greatly enhanced the monomer scope and the complexity of the resulting polymer structures to produce various brush,30 starshaped,31 bridged,32,33 and ionic18,34,35 polyacetylenes or those containing six-membered rings36−41 and heterocycles as well.18,34,42,43 Even though there are many reports and mechanistic studies detailing successful CP using Grubbs catalysts, only highly active and expensive Grubbs catalysts containing N-heterocyclic carbene (NHC) ligands are known to effectively promote CP thus far. Nevertheless, a less active but much cheaper firstgeneration Grubbs catalyst (Figure 1, G1) has been widely used in the past two decades in many organic and polymerization reactions such as cross-metathesis (CM), ring-closing metathesis (RCM), enyne metathesis, acyclic diene metathesis (ADMET) polymerization, and ring-opening metathesis polymerization (ROMP).44,45 In particular, G1 was a much more efficient catalyst than G2, HG2, and G3 in tandem ringopening/ring-closing polymerization,46 thus showing that G1 is highly useful in synthetic chemistry. Although G1 has been known for more than 20 years to be an inactive catalyst for CP, itnonethelessshowed good activities for both intra- and intermolecular enyne metathesis reactions.47,48 As such, we questioned the reason behind the failure of G1 to catalyze CP. On the basis of our previous reports on the importance of catalyst stability during CP,27−29 as well as other studies that highlighted the enhancement of G1’s activity by controlling the equilibrium between the 16-electron precatalyst and the 14electron active catalyst,49−57 we believed that CP using G1 should be possible as well. Herein, we report the first successful CP of 1,6-heptadiyne derivatives and the detailed mechanistic studies of the reaction using G1 and simple additives. Three different strategies of enhancing CP were proposed, and various additives were grouped into three distinct categories. Thereafter, the additives were screened extensively to identify the best reaction condition, thereby maximizing the polymerization efficiency to achieve polymerization selectivity of up to 100% and to afford conjugated polymers with Mn values of up to 23 kDa. Additionally, various kinetic studies on CP were carried out, and especially the propagating carbenes were closely monitored using 1H and 31P NMR spectroscopy to reveal how the additives affected CP and its competing side reaction, [2 + 2 + 2] cycloaddition. This strategy has enabled us to gain a deeper understanding of the important factors that influence CP. In short, this report provides a clear, definitive, and comprehensive evaluation of CP using Grubbs catalysts.

Scheme 1. Competition between Cyclopolymerization and [2 + 2 + 2] Cycloaddition of 1,6-Heptadiynes Using Grubbs Catalysts

Mo−alkylidene catalysts and prepared conjugated polymers containing six-12,13 or five-membered rings14,15 via selective βand α-addition, respectively. Meanwhile, Ru-based Grubbs catalysts, which are air-stable and tolerant to various functional groups, have previously failed to promote CP16 and thus led to the misbelief that they were not sufficiently active for CP (Figure 1, G1, G2, and HG2). However, the Buchmeiser group

Figure 1. Chemical structures of common Ru catalysts.

reported a breakthrough on the first successful α-selective CP to produce soluble polyacetylenes that contained fivemembered rings exclusively using Ru−alkylidenes by modifying a second-generation Hoveyda−Grubbs catalyst with electronwithdrawing groups such as trifluoroacetate or isocyanate (Figure 1, Buch-I and Buch-II).16−21 Since Buchmeiser group’s pioneering development of CP using these new Ru−alkylidene catalysts, several reports suggested that certain aromatic side products such as the dimer and trimer of diynes were generated during the reaction instead of the desired conjugated polymers (Scheme 1B).19,21 In those reports, the Buchmeiser group suggested that a mechanism based on competing intermolecular cyclization by a series of backbiting olefin metathesis reaction pathway was responsible for these side products during CP. This mechanism seemed certainly plausible as it was also proposed by several organic chemists (e.g., the Blechert and Witulski groups) who had reported the same cyclization reaction of either triynes or a combination of diynes and terminal alkynes.22,23 In contrast, another study by the Pérez-Castells group suggested that the cyclization reaction (leading to the dimer and trimer) was promoted by a completely different [2 + 2 + 2] cycloaddition mechanism, catalyzed via ruthenacyclopentatriene intermediates by some unknown Ru complexes after the Grubbs catalysts underwent decomposition, certainly not via the olefin metathesis mechanism.24−26 To put an end to this argument, our group recently conducted several control experiments to conclude that it was indeed the decomposed Ru complex that produced those aromatic side products via a [2 + 2 + 2] cycloaddition mechanism during CP.27 In other words, the key



RESULTS AND DISCUSSION Our first strategy to enhance CP began with using phosphinefree catalysts58,59 or adding various phosphine trapping agents to facilitate the dissociation of phosphine,49−57 thereby increasing the concentration of the active 14-electron catalyst. First, we tested the reactivity of G1 toward CP of the M1 monomer (M/I = 25) in dichloromethane (DCM) as a control B

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 1. Screening Various Additives To Obtain the Maximum CP Efficiency and Selectivity

entry 1 2 3

a

cat.

4 5 6

G1 HG1 G1Py2 G1 G1 G1

7

G1

8

G1

9

G1

10

G1

11

G1

12 13 14 15

G1 G1 G1 G1

16 17 18

G1 G1 G1

conva (%)

P1a (%)

D1 (T1)a (%)

P1/(P1 + D1 + T1)

entry

>99 82 34

8 18 23

52 (38) 44 (17) 6 (3)

0.08 0.23 0.72

19 20 21

G1 G1 G1

AlCl3 (5 equiv) CuCl trifluoroacetic acid (0.5) 4-nitrobenzoic acid (3.85) PhCOOH (4.20)

62 42 98

16 2 0

38 (7) 26 (8) 41 (8)

0.26 0.06 0

22 23 24

HG1 HG1 G1

67

39

28

0.58

25

G1

88

86

2

0.98

26

G1

hexanoic acid (4.88) pivalic acid (5.03)

56

46

7

0.87

27

G1

54

40

6

0.87

28

G1

2,4-dinitrophenol (4.11) THF (solvent) pyridine 3-chloropyridine 3,5dichloropyridine CH3COONa PhCOONa CF3COONa

>99

0

59 (36)

0

29

G1

94 25 31 36

3 23 26 16

59 (29) 0 3 16

0.03 1.0 0.90 0.50

30b 31b 32b 33c

99 >99 67

93 97 48

0 0 13

1.0 1.0 0.79

34c 35c

additive (pKa) none none none

conva (%)

P1a (%)

D1 (T1)a (%)

P1/(P1 + D1 + T1)

CH3COOLi CH3COOK CH3COONH4

92 92 35

19 84 23

48 (19) 0 0

0.22 1.0 1.0

44 77 98

8 42 29

15 18 37 (31)

0.35 0.70 0.30

81

30

26 (5)

0.49

81

11

55 (12)

0.14

36

18

12 (2)

0.56

99

62

27 (5)

0.70

97

84

G2 G2 G2 HG2

PhCOOH PhCOONa CH3COOAg (1 equiv) CH3COOAg (2 equiv) PhCOOAg (1 equiv) PhCOOAg (2 equiv) CF3COOAg (2 equiv) CF3COOAg (5 equiv) none PhCOOH PhCOONa none

80 68 58 94

17 66 41 8

49 (11) 1 16 54 (22)

0.22 0.99 0.72 0.10

HG2 HG2

PhCOOH PhCOONa

84 84

68 11

8 51 (12)

0.89 0.15

cat.

additive

6

0.93

Calculated from 1H NMR spectroscopy (Figure S17). b20 h reaction. c1 h reaction.

pathway (entry 6). However, to our surprise, carboxylic acids with pKa of ca. 4−5 significantly improved the efficiency of CP and greatly suppressed dimer formations (entries 7−10). Among them, benzoic acid was the best additive, affording 86% conversion to P1 with 98% CP selectivity (entry 8). Other acids, either slightly stronger or weaker, also favored CP over the competing [2 + 2 + 2] cycloaddition with selectivity between 58% and 87% (entries 7, 9, and 10); these were much better results than those obtained using Lewis acids or without any additive at all. Therefore, using an acid additive with a proper pKa value was important as the catalyst might become unstable with stronger acids, while weaker acids might not trap the phosphine effectively to activate the catalyst sufficiently. On the contrary, the pKa of the additive was not the only determining factor because 2,4-dinitrophenol, which has a similar pKa to benzoic acid, completely suppressed CP but promoted [2 + 2 + 2] cycloaddition exclusively (entry 11). In short, achieving a proper balance between the dissociation of phosphine by protonation and a weak coordination by benzoic acid greatly enhanced CP, while removing phosphine (either by completely trapping or eliminating it) was unhelpful because ensuring the stability of the catalyst by reversible coordination of the phosphine was very important as well. As the second strategy, we then decided to add various weakly coordinating ligands to stabilize the propagating

experiment and found that the competing [2 + 2 + 2] cycloaddition reaction did predominate over the desired CP reaction (52% + 38% vs 8%: 8% selectivity for CP) (Table 1, entry 1). This was similar to previous CP results using G2 and HG2 catalysts.27 We also tested the same CP with phosphinefree G1 catalysts such as HG158 and G1-Py2.59 HG1 produced a slightly higher amount of the conjugated polymer, P1 (18%), compared to G1 (8%); however, 44% of the dimer (D1) and 17% of the trimer (T1) were formed meaning that the selectivity for CP was only 23% (entry 2). With G1-Py2 catalyst containing a more labile pyridine ligand, the yield of P1 increased only slightly to 23% even though the formation of the side products, D1 and T1, was highly suppressed (entry 3). This was a big disappointment because the analogous G3 (containing the NHC ligand) is the best catalyst for living CP, thus implying that the stabilization mechanism of G1 might differ from those NHC ligand-containing catalysts. Second, a Lewis acid additive such as AlCl3, known to increase the reactivity of G1 toward RCM,49,50 was used and this slightly improved the selectivity for CP (26%). However, still large amounts of D1 and T1 were formed over P1 (38% + 7% vs 16%, entry 4). Disappointingly, another Lewis acid, CuCl,49,50 seemed to deactivate both pathways (entry 5). In response, various Brønsted acids were screened, and we observed that a strong acid, trifluoroacetic acid, completely shut down the CP C

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 2. Optimization of CP with Benzoic Acid and Sodium Benzoate Additives

a

entry

additive

M/I/Add

conc (M)

time (h)

conva (%)

P1a (%)

D1 (%)a

yield (%)

Mnb (kDa)

Đb

1 2 3 4 5 6 7

PhCOOH PhCOOH PhCOOH PhCOOH PhCOONa PhCOONa PhCOONa

25/1/10 25/1/10 25/1/10 50/1/15 25/1/5 25/1/8 50/1/8

0.5 1.0 2.0 2.0 0.5 1.0 1.0

23 20 24 48 24 3 24

88 90 98 85 92 99 87

86 90 98 83 83 96 82

2 0 0 2 9 0 0

82 87 93 72 62 90 74

12.5 14.7 14.6 20.2 11.7 12.3 23.3

1.70 1.64 1.60 1.68 1.69 1.67 1.53

Calculated from 1H NMR. bDetermined by THF size-exclusion chromatography (SEC) calibrated using polystyrene (PS) standards.

increased the CP efficiency to 84% with 93% selectivity (entry 29). Overall, new Ru complexes containing acetate or benzoate ligands showed slightly higher activities toward CP than G1 alone, but these efficiencies were still much lower than that afforded by benzoic acid or sodium benzoate, presumably because of the lower stability of the new complexes.64 In order to get a better insight, we monitored the changes in carbene signals by 1H NMR after adding silver benzoate to G1 and observed multiple complicated carbene signals, implying the formation of various complex carbenes (Figure S2). However, adding silver trifluoroacetate to G1 produced only a couple of new carbene signals, hence suggesting that these more welldefined Ru complexes substituted by electron-withdrawing trifluoroacetate ligands were much more efficient for CP (Figure S3). Since we discovered that the benzoic acid and sodium benzoate additives greatly improved CP of G1, we became curious how these new additives would affect CP involving the NHC-containing Grubbs catalysts, G2 and HG2. Similar to the previous report,27 the CP of M1 by G2 and HG2 (M/I = 25), without using any additive, produced D1 and T1 predominantly (entries 30 and 33). Interestingly, both benzoic acid and sodium benzoate somewhat enhanced the CP efficiency of G2 respectively producing 66% and 41% of P1; nonetheless, these conversions were much lower than those afforded by G1 (entries 31 and 32 vs entries 8 and 17). Furthermore, the addition of benzoic acid to another NHC-containing HG2 moderately increased the conversion toward P1 to 68% (entry 34), but replacing it with sodium benzoate resulted in an unsuccessful CP with only 15% selectivity (entry 35). As a summary, it is interesting that the additives that were optimal for G1 were not the best additives for G2 and HG2, whereas the best additive or condition for G2 and HG2 did not work at all for the CP of G1. These results strongly implied that an appropriate choice of additive, based on the specific type of catalysts, was essential for a successful CP because the activation mechanisms of phosphine-containing G1 and NHC-containing G2 and HG2 might be different. Based on the additive screening results, the reaction conditions for CP were further optimized using the two best additivesbenzoic acid and sodium benzoate (Table 2). As the concentration of the monomer increased from 0.5 to 2.0 M, the CP efficiency increased up to 98% with a perfect reaction selectivity to give P1 with Mn of 15 kDa using the benzoic acid additive (M/I = 25, entries 1−3). Furthermore, using a higher amount of sodium benzoate (from 5 to 8 equiv) and a higher

carbene, thereby suppressing the catalyst decomposition during CP. First, THF (a weakly coordinating solvent), which worked well for HG2 and G3,27−30 was tested for the CP of G1; disappointingly, the experiment was a complete failure as D1 and T1 were formed almost exclusively (entry 12). Moreover, various pyridine ligands, which also functioned as excellent additives for HG2 and G3,27,29 produced only 16%−26% of P1 with almost no or small amounts of D1 (entries 13−15). Clearly, these results showed that the activation of CP using G1 was very different from those of G2, HG2, and G3 because a simple stabilization of G1 via weak coordination could not enhance CP sufficiently. Inspired by the results of using carboxylic acids as additives, we screened various carboxylate salts and observed that sodium acetate and sodium benzoate were excellent additives, while sodium trifluoroacetate showed only a moderate effect (entries 16−18). We also screened other acetate salts containing different counter cations such as Li+, K+, and NH4+, but they were less efficient compared to sodium acetate (entries 19−21). Among the carboxylate salts tested, sodium benzoate gave the best result, showing full consumption of the monomer with 100% selectivity toward P1 (entry 17). Additionally, the phosphine-free catalyst HG1 was retested with the two best performing additivesbenzoic acid and sodium benzoate (entries 22 and 23). However, both additives could not sufficiently enhance the CP efficiency, suggesting that the extra phosphine ligand in G1 was essential for the successful CP. The last strategy to enhance CP involved exchanging the anionic X-type ligand, chloride, in G1 as inspired by the reports from the Buchmeiser group, who demonstrated the first successful CP using Buch-I via this strategy.16−21 The Buchmeiser group treated various silver salts to HG2 and isolated the corresponding modified Ru catalysts, which contained NHC and electron-withdrawing ligands. Likewise, we compared the efficiencies of CP by adding various silver carboxylates to G1 in order to produce the corresponding carboxylated Ru complexes in situ.60−68 Initially, 1 or 2 equiv of silver acetate was added to G1 (entries 24 and 25), but only 30% of P1 was formed, which was only scarcely better than just using G1 on its own (entry 1). It was more disappointing to find that 1 or 2 equiv of silver benzoate (which we had hoped to deliver better results) led to even poorer results (11−18% P1, entries 26 and 27) than the case of silver acetate. However, using 2 equiv of silver trifluoroacetate significantly increased the CP efficiency up to 62%, with 32% of side products (entry 28). Furthermore, using a higher loading of the additive (5 equiv) D

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 3. Cyclopolymerization of Various Monomers by G1

PhCOOH additivea c

entry monomer 1 2 3 4 5 6 7 8 9 10 11

M2 M3 M4 M4 M5 M6 M7 M7 M8 M8 M9

M/I

conv (%)

polymer (%)

25 25 25 50 25 25 25 50 25 50 25

73 84 98 90 74 88 97 88 95 65 98

65 75 88 76 46 80 89 81 84 55 75

c

dimer (%) 1 6 10 14 11 5 8 7 6 6 4

c

PhCOONa additiveb Mnd

c

yield (%)

(kDa)

Đd

conv (%)

polymer (%)

62 70 64 64 46 76 78 73 77 43 59

10.2 11.7 9.5 9.9 6.4 11.1 12.0 15.7 9.7 10.7 7.7

1.46 1.67 1.45 1.62 1.39 1.70 1.53 1.71 1.76 1.88 1.56

91 61 97 58 89 93 >99 63 99 93 87

87 49 95 44 68 71 87 57 93 82 58

c

dimerc (%)

yield (%)

Mnd (kDa)

Đd

0 4 0 6 0 0 0 97%, Figure S5) while CP using G3 produced a 5.4:1 mixture of E:Z-olefins.70 To broaden the monomer scope, various 4-bis-substituted 1,6-heptadiyne monomers (M2−M4) and 4-monosubstituted monomers (M5−M9) were polymerized using G1 with the addition of benzoic acid and sodium benzoate, respectively E

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules well as a high Mn of 12 kDa (entries 9 and 10). Lastly, the triisopropylsilyl (TIPS) ether-containing monomer M9 gave 75% and 58% conversions to P9 (Mn of 8 kDa) using benzoic acid and sodium benzoate, respectively (entry 11). Based on the results obtained from the monomer screening, the CP of G1 with the two additives showed comparable or complementary results, but dimer formation was more effectively suppressed using sodium benzoate. These results implied that these two additives might work differently during the course of CP. In order to investigate how the additives enhanced CP, a series of in situ NMR experiments were carried out to monitor the reaction kinetics of CP and changes in the propagating carbenes (Figure 2). Initially, the carbene signal of the original G1 appeared at 20.02 ppm, and adding the monomer M9 alone immediately generated a new carbene peak (i.e., sharp singlet) at 20.44 ppm (Figure 2A, left). This signal’s intensity increased up to 14% during the early stage of the CP, and its decay was rather slow throughout the reaction. Subsequently, other new broad signals appeared at 20.28−20.34 ppm, whose intensity gradually increased to 20% while that of the aforementioned singlet (at 20.44 ppm) slowly decreased (Figures 2A and 2B, left). We assigned the singlet at 20.44 ppm to the carbene with a degree of polymerization (DP) of 1 generated by the first monomer addition to the catalyst (Figure 2, 9a). The broad peak at 20.28−20.34 ppm was assigned to the actual propagating carbene with DP > 1 (Figure 2, 9b), formed by further monomer additions.71 To further support this argument, we conducted a control experiment using a just 1:1 mixture of G1 and the monomer M9 (M/I = 1) and observed that the 20.44 ppm singlet formed exclusively (Figure S7). Additionally, we monitored the changes in the carbenes of G3 with pyridine additives and observed a similar phenomenon, whereby two different signals corresponding to the DP = 1 carbene and the DP > 1 propagating carbene were detected separately (Figure S8). Fortunately, G1 contains two phosphines, and this allowed us to monitor the formation of new carbenes by 31P NMR, which also showed two new peaks (Figure S11). Similar to the 1H NMR result, the two new neighboring peaks at 36.8 and 37.2 ppm of the 31P NMR spectrum were assigned to the DP = 1 carbene (9a) and the actual propagating carbene (9b), respectively. Gratifyingly, changes in the population of these two 31P NMR peaks were consistent with those of 1H NMR (Figures 2B and 2C, left), thereby providing further supporting information about the propagating carbenes. On the basis of these results, we realized that the amount of the actual propagating carbene without any additive was very low (below 20%), and more importantly, conversion of the propagating carbene from 9a to 9b was very slow or inefficient. This observation agreed with the previous studies, which reported that initiation of ruthenium vinylidene was much slower than that of ruthenium benzylidene during ROMP.72 This implied that further CP propagation by G1 was retarded because of the low efficiency of phosphine dissociation from the conjugated carbene 9a. Consequently, the ratio of the actual propagating carbene (DP > 1) to the consumed G1 was lower than 20% during CP (Figure 2E), thus leading to an inefficient polymerization and eventual carbene decomposition occurred to produce the dimer as a major side product by [2 + 2 + 2] cycloaddition (Figure 2D, left). In order to understand how the additives enhanced the CP efficiency, the same investigation described above was carried out using the benzoic acid additive. Compared to the previous

Figure 2. (A) 1H NMR spectra monitoring the initial and propagating carbenes with M/I = 10. (B) Plots of changes for various carbene signals obtained from 1H NMR vs time. (C) Plots of changes for various 31P NMR signals vs time. (D) Plots of total conversion, conversions of polymerization, and dimerization vs time. (E) Ratios of the actual propagating carbene for DP > 1 to the conversion of the initial G1 over time.

case, the initial consumption of G1 with the additive was twice as fast (0.61 vs 1.2 mM/min, Table S1), with a full initiation in 6 h. The DP = 1 carbene at 20.44 ppm (9a) increased and decreased quickly at the early stage. More importantly, the DP F

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

species resulting from a mixture of M9, G1, and benzoic acid additive), we added just 1 equiv of silver benzoate and 5 equiv of tricyclohexylphosphine to promote X-type ligand exchanges (Figure 3) and observed the same new peak appearing at 20.52

> 1 propagating carbene (9b) formed rapidly (up to 34%, Figures 2A and 2B, middle), indicating that both the initiation of G1 and the conversion from 9a to 9b was accelerated by benzoic acid; this occurred presumably by trapping one of the phosphines to facilitate the formation of a 14-electron species of the conjugated propagating carbene. Furthermore, benzoic acid seemed to act as a weakly coordinating ligand to stabilize the propagating carbene, thereby preventing its decomposition. As a result, the ratio of the propagating carbene to the consumed G1 was above 50% (Figure 2E), and the initial polymerization rate was 6 times higher than that without using any additive (0.043 vs 0.25 mM/min, Table S1 and Figure 2D, middle). In general, the same changes in the formations of 9a and 9b were observed by both 31P and 1H NMR analyses (Figures 2B and 2C, middle). More importantly, 31P NMR analysis also showed the formation of 3−5% of HPCy3+ at 54.8 ppm (Figure S12), confirming the role of benzoic acid. This leaves a majority of the tricyclohexylphosphine available for active coordination to G1 throughout CP, indicating that analogous to atom transfer radical polymerization (ATRP)73−76achieving an appropriate equilibrium between the active (14-electron Ru) and the dormant species (16electron Ru) of the catalyst ensured a successful CP. This was why the CP of phosphine-free HG1 and G1-Py2, even with the use of additive, were much less efficient than that achieved using G1. Additionally, we observed that the amount of the DP > 1 propagating carbene increased by 11% (from 31% to 42% using M1, Figure S14) as the concentration increased from 0.2 to 0.6 M, suggesting that a high concentration (2.0 M) enhanced the CP efficiency (Table 2, entries 1−3). Lastly, further investigations with the sodium benzoate additive revealed that the initial consumption of G1 was ironically 0.36 times lower than that even without any additive (0.61 vs 0.22 mM/min, Table S1). This might result from the competitive coordination of the benzoate anion against the monomer. However, the conversion of 9a to 9b was rather fast and on top of the usual propagating carbene 9b at 20.28−20.34 ppm, a completely new propagating carbene, 9c at 20.52 ppm, appeared as monitored by 1H NMR spectroscopy. Interestingly, the intensity of these two peaks (9b and 9c) grew up to 24% each to give a combined value of close to 50% (Figures 2A and 2B, right). The changes in the population of 9a (36.8 ppm) and 9b (37.2 ppm), as well as the new peak corresponding to 9c at 23.6 ppm, based on 31P NMR analysis matched those of the 1H NMR analysis (Figures 2B and 2C, right, Figure S13). Interestingly, there was a significant upfield shift for this new 9c peak compared to the carbenes of 9a−9b (23.6 ppm vs 36.8 and 37.2 ppm), and this implied for some changes in the ligand sphere of the Ru complex. Even though the initiation rate was the slowest with the sodium benzoate additive, the propagation was the fastest as the initial polymerization rate was 23 times faster than using only G1 or 4 times faster than utilizing the benzoic acid additive (0.043 vs 0.25 vs 0.98 mM/min, Table S1 and Figure 2D). We suspected that this high reactivity was due to the formation of the more active carbene species 9c corresponding to the peak at 20.52 ppm in 1H NMR or 23.6 ppm in 31P NMR analyses. Thus, we carried out control experiments to identify the origin of this new peak. Initially, we added sodium benzoate to G1, but this did not produce any new carbene species, which meant that the benzylidene G1 on its own did not react with sodium benzoate at all (Figure S15). However, after the in situ preparation of the DP > 1 propagating carbene (9b) (the major

Figure 3. An exchange reaction of chloride ligand with silver benzoate on the propagating carbene.

ppm. Therefore, we identified the new peak as a Ru carbene species containing one chloride ligand and one benzoate ligand each (9c). We believe that the observed complex 9c (as well as 9a and 9b) contains two phosphines instead of one because the relative integration of the corresponding carbenes observed by 1 H NMR was similar to that of observed by 31P NMR. Also, the chemical shift of 9c is similar to that of the initial G1 catalyst. On the basis of these investigations, we concluded that sodium benzoate enhanced the reactivity of G1 by partially exchanging one chloride ligand with a benzoate ligand (9c), while other benzoate anions stabilized the original 9b species by weak coordination to suppress carbene decomposition, similar to the role of benzoic acid. Therefore, these two combined factors led to a high population of the propagating carbene ratio (up to 50%, Figure 2E) and, hence, a successful CP. Ironically, silver benzoatea much stronger exchanging reagent compared to sodium benzoategave complex mixtures of unstable carbenes and thus was a poor additive for the CP of G1 (Table 1, entries 26 and 27). However, it is clear that Buchmeiser’s original strategy for activating Hoveyda−Grubbs catalysts by exchanging the chloride ligand with electron-withdrawing acetates7 certainly worked for the CP of G1 as well. Lastly, we investigated how various additives affected the undesired side reaction by testing the same [2 + 2 + 2] cycloaddition reaction using another catalyst, RuCp*(cod)Cl, to produce aromatic dimers and trimers independently (Figure G

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules 4).77,78 Without using any additive, 1 mol % of the catalyst rapidly converted M1 to D1 and T1, at an initial reaction rate

benzoate also slowed down the reaction by 28%, with an initial rate of 4.3 mM/min (Table S2). From these control experiments, one could suggest that these additives not only activated and stabilized the propagating carbenes of G1 but also suppressed or retarded the [2 + 2 + 2] cycloaddition side reaction by poisoning or coordinating to the decomposed G1. By combining all the data, we could finally reveal a comprehensive picture of how the reactions between G1 and 1,6-heptadiyne monomers proceeded with and without additive, as summarized in Scheme 2. G1 without additive formed DP = 1 conjugated carbene, but it was difficult for this carbene species to undergo further propagation due to a slow or inefficient phosphine dissociation. Without any external stabilizing ligands, it inevitably decomposed, and this decomposed Ru complex predominantly catalyzed the [2 + 2 + 2] cycloaddition reaction to produce dimers (Scheme 2, pathway A). However, adding benzoic acid to the polymerization accelerated both phosphine dissociation and the initiation of G1 by protonating the phosphine. More importantly, this additive also accelerated the formation of the actual propagating carbene from the DP = 1 carbene species, whose phosphine dissociation was very sluggish. Moreover, benzoic acid also stabilized the propagating carbene species by weak coordination, thereby suppressing carbene decomposition. It also poisoned the decomposed catalyst to hamper the dimerization side reaction (pathway B). Interestingly, another excellent additivesodium benzoatesuccessfully promoted CP via a different mechanism from that of benzoic acid despite their similar structures. The benzoate anion partially substituted one chloride ligand on G1, and this in situ formation of the new catalyst afforded a more active catalytic species to facilitate CP. Furthermore, sodium benzoate also seemed to stabilize the catalysts and this enhanceed CP in a similar manner to benzoic acid (pathway C).

Figure 4. Kinetic profiles of [2 + 2 + 2] cycloaddition by RuCp*(cod) Cl with various additives.

of 6.0 mM/min (Table S2). Interestingly, adding pyridine to the reaction almost stopped the conversion, presumably because its coordination to RuCp*(cod)Cl poisoned the catalyst. This agreed with our previous observation that pyridine also suppressed the side reaction, [2 + 2 + 2] cycloaddition of monomers, during CP catalyzed by HG2.11 Furthermore, benzoic acid also significantly retarded the [2 + 2 + 2] cycloaddition catalyzed by RuCp*(cod)Cl (1.7 mM/min, Table S2), presumably via a similar mechanism. Adding sodium



CONCLUSION In conclusion, 20 years after the invention of first-generation Grubbs catalyst, we finally report the first successful cyclo-

Scheme 2. Proposed Scheme Showing How the Additives Affect CP Catalyzed by G1

H

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Macromolecules



polymerization of 1,6-heptadiyne derivatives by using simple additives to activate G1, which has been known to be inactive for this particular reaction. To achieve a successful CP, we attempted three strategies: (1) using phosphine-free catalysts or phosphine trapping agents to activate the catalyst, (2) stabilizing the propagating carbene by adding coordinating ligands, and (3) promoting anionic ligand exchange of G1. Through an extensive screening of three different classes of additives, both benzoic acid and sodium benzoate proved to be the best additives for promoting an efficient CP, being compatible with a broad monomer scope and capable of producing various conjugated polyenes with Mn of up to 23 kDa. Furthermore, detailed mechanistic investigation by kinetic studies and monitoring the changes in the carbene complexes by 1H and 31P NMR analyses revealed how these two additives enhanced the CP efficiency via two different mechanisms. Benzoic acid increased both the initiation and propagation rates by partially trapping phosphine and stabilizing the catalyst via a weak coordination, respectively. In contrast, sodium benzoate retarded initiation owing to its competitive coordination to the catalyst, but more importantly, it produced more active propagating carbene species by undergoing a partial exchange of chloride ligand with benzoate ligand, which, then, greatly accelerated CP. Finally, we conducted independent control experiments of [2 + 2 + 2] cycloaddition (catalyzed by RuCp*(cod)Cl with and without these additives) and concluded that these additives either retarded or suppressed the dimerization side reaction. This also explained why these additives provided a high CP selectivity over the side reaction. In short, these results have provided detailed insights to improve our understanding of cyclopolymerization using Grubbs catalysts, and this should greatly broaden the utility of CP including access to new 2D-nanomaterials79 because G1 is the cheapest and the most readily available olefin metathesis catalyst.



ASSOCIATED CONTENT

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.7b00488. Experimental procedures, characterizations, NMR spectra for new compounds and polymers, SEC traces, and other supporting experiments (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected] (T.-L.C.). ORCID

Tae-Lim Choi: 0000-0001-9521-6450 Notes

The authors declare no competing financial interest.



REFERENCES

(1) Choi, S.-K.; Gal, Y.-S.; Jin, S.-H.; Kim, H.-K. Poly(1,6heptadiyne)-Based Materials by Metathesis Polymerization. Chem. Rev. 2000, 100, 1645−1681. (2) Stille, J. K.; Frey, D. A. Polymerization of Non-conjugated Diynes by Complex Metal Catalysts. J. Am. Chem. Soc. 1961, 83, 1697−1701. (3) Gibson, H. W.; Bailey, F. C.; Epstein, A. J.; Rommelmann, H.; Kaplan, S.; Harbour, J.; Yang, X.-Q.; Tanner, D. B.; Pochan, J. M. Poly(1,6-heptadiyne), a free-standing polymer film dopable to high electrical conductivity. J. Am. Chem. Soc. 1983, 105, 4417−4431. (4) Kim, Y.-H.; Gal, Y.-S.; Kim, U.-Y.; Choi, S.-K. Cyclopolymerization of dipropargylsilanes by transition-metal catalysts. Macromolecules 1988, 21, 1991−1995. (5) Ryoo, M.-S.; Lee, W.-C.; Choi, S.-K. Cyclopolymerization of diethyl dipropargylmalonate by transition metal catalysts. Macromolecules 1990, 23, 3029−3031. (6) Jang, M.-S.; Kwon, S.-K.; Choi, S.-K. Cyclopolymerization of diphenyldipropargylmethane by transition metal catalysts. Macromolecules 1990, 23, 4135−4140. (7) Kang, K.-L.; Kim, S.-H.; Cho, H.-N.; Choi, K.-Y.; Choi, S.-K. A new class of conjugated ionic polyacetylene. Cyclopolymerization of dihexyldipropargylammonium salts by metathesis catalysts. Macromolecules 1993, 26, 4539−4543. (8) Jeon, S. J.; Shim, S. C.; Cho, C. S.; Kim, T. J.; Gal, Y. S. Catalytic cyclopolymerization and copolymerization of 1,6-heptadiynes by Mo(CO)6. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2663−2670. (9) Gal, Y. S.; Jin, S. H.; Choi, S. K. Poly(1,6-heptadiyne)-based functional materials by metathesis polymerization. J. Mol. Catal. A: Chem. 2004, 213, 115−121. (10) Fox, H. H.; Schrock, R. R. Living cyclopolymerization of diethyl dipropargylmalonate by Mo(CH-t-Bu) (NAr)[OCMe(CF3)2]2 in dimethoxyethane. Organometallics 1992, 11, 2763−2765. (11) Fox, H. H.; Wolf, M. O.; O’Dell, R.; Lin, B. L.; Schrock, R. R.; Wrighton, M. S. Living Cyclopolymerization of 1,6-Heptadiyne Derivatives Using Well-Defined Alkylidene Complexes: Polymerization Mechanism, Polymer Structure, and Polymer Properties. J. Am. Chem. Soc. 1994, 116, 2827−2843. (12) Schattenmann, F. J.; Schrock, R. R.; Davis, W. M. Preparation of Biscarboxylato Imido Alkylidene Complexes of Molybdenum and Cyclopolymerization of Diethyldipropargylmalonate To Give a Polyene Containing only Six-Membered Rings. J. Am. Chem. Soc. 1996, 118, 3295−3296. (13) Schattenmann, F. J.; Schrock, R. R. Soluble, Highly Conjugated Polyenes via the Molybdenum-Catalyzed Copolymerization of Acetylene and Diethyl Dipropargylmalonate. Macromolecules 1996, 29, 8990−8991. (14) Anders, U.; Nuyken, O.; Buchmeiser, M. R. Stereoselective Cyclopolymerization of 1,6-Heptadiynes: Access to Alternating cistrans-1,2-(Cyclopent-1-enylene)vinylenes by Fine-Tuning of Molybdenum Imidoalkylidenes. Angew. Chem., Int. Ed. 2002, 41, 4044−4047. (15) Anders, U.; Nuyken, O.; Buchmeiser, M. R.; Wurst, K. FineTuning of Molybdenum Imido Alkylidene Complexes for the Cyclopolymerization of 1,6-Heptadiynes To Give Polyenes Containing Exclusively Five-Membered Rings. Macromolecules 2002, 35, 9029− 9038. (16) Krause, J. O.; Zarka, M. T.; Anders, U.; Weberskirch, R.; Nuyken, O.; Buchmeiser, M. R. Simple Synthesis of Poly(acetylene) Latex Particles in Aqueous Media. Angew. Chem., Int. Ed. 2003, 42, 5965−5969. (17) Krause, J. O.; Nuyken, O.; Buchmeiser, M. R. Factors Relevant for the Ruthenium−Benzylidene-Catalyzed Cyclopolymerization of 1,6-Heptadyines. Chem. - Eur. J. 2004, 10, 2029−2035. (18) Halbach, T. S.; Krause, J. O.; Nuyken, O.; Buchmeiser, M. R. Stereoselective Cyclopolymerization of Polar 1,6-Heptadiynes by Novel, Tailor-Made Ruthenium-Based Metathesis Catalysts. Macromol. Rapid Commun. 2005, 26, 784−790. (19) Kumar, P. S.; Wurst, K.; Buchmeiser, M. R. Factors Relevant for the Regioselective Cyclopolymerization of 1,6-Heptadiynes, N,NDipropargylamines, N,N-Dipropargylammonium Salts, and Diproparg-

S Supporting Information *



Article

ACKNOWLEDGMENTS

The financial support from Mid-Career Basic Science Research Program and the Nano-Material Technology Development Program through NRF is acknowledged. K.C. is supported by NRF-Global PhD Fellowship program (Grant NRF2015H1A2A1030158). I

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules yl Ethers by RuIV−Alkylidene-Based Metathesis Initiators. J. Am. Chem. Soc. 2009, 131, 387−395. (20) Buchmeiser, M. R.; Schmidt, C.; Wang, D. Cyclopolymerization of N,N-Dipropargyl-3,4-dialkoxyanilines Using Schrock and GrubbsHoveyda Initiators: Influence of Initiator Structure on the Mode of Insertion. Macromol. Chem. Phys. 2011, 212, 1999−2008. (21) Sudheendran, M.; Horecha, M.; Kiriy, A.; Gevorgyan, S. A.; Krebs, F. C.; Buchmeiser, M. R. Cyclopolymerization-derived blockcopolymers of 4,4-bis(octyloxymethyl)-1,6-heptadiyne with 4,4dipropargyl malonodinitrile for use in photovoltaics. Polym. Chem. 2013, 4, 1590−1599. (22) Peters, J.-U.; Blechert, S. Ruthenium-catalysed conversion of triynes to benzene derivatives − a novel metathesis cascade. Chem. Commun. 1997, 1983−1984. (23) Witulski, B.; Stengel, T.; Fernández-Hernández, J. M. Chemo and regioselective crossed alkyne cyclotrimerisation of 1,6-diynes with terminal monoalkynes mediated by Grubbs’ catalyst or Wilkinson’s catalyst. Chem. Commun. 2000, 1965−1966. (24) Medina, S.; Domínguez, G.; Pérez-Castells. Efficient Generation of Pyridines by Ruthenium Carbene Mediated [2 + 2 + 2] Cyclotrimerization. Org. Lett. 2012, 14, 4982−4985. (25) Alvarez, S.; Medina, S.; Domínguez, G.; Pérez-Castells, J. [2 + 2+2] Cyclotrimerization of Alkynes and Isocyanates/Isothiocyanates Catalyzed by Ruthenium−Alkylidene Complexes. J. Org. Chem. 2013, 78, 9995−10001. (26) Alvarez, S.; Medina, S.; Domínguez, G.; Pérez-Castells, J. Ruthenium Alkylidene-Catalyzed Reaction of 1,6-Heptadiynes with Alkenes. J. Org. Chem. 2015, 80, 2436−2442. (27) Kang, E. H.; Kang, C.; Yang, S.; Oks, E.; Choi, T. L. Mechanistic Investigations on the Competition between the Cyclopolymerization and [2 + 2 + 2] Cycloaddition of 1,6-Heptadiyne Derivatives Using Second-Generation Grubbs Catalysts. Macromolecules 2016, 49, 6240− 6250. (28) Kang, E.-H.; Lee, I. S.; Choi, T.-L. Ultrafast Cyclopolymerization for Polyene Synthesis: Living Polymerization to Dendronized Polymers. J. Am. Chem. Soc. 2011, 133, 11904−11907. (29) Kang, E.-H.; Yu, S. Y.; Lee, I. S.; Park, S. E.; Choi, T.-L. Strategies to Enhance Cyclopolymerization using Third-Generation Grubbs Catalyst. J. Am. Chem. Soc. 2014, 136, 10508−10514. (30) Kang, E.-H.; Lee, I.-H.; Choi, T.-L. Brush Polymers Containing Semiconducting Polyene Backbones: Graft-Through Synthesis via Cyclopolymerization and Conformational Analysis on the Coil-to-Rod Transition. ACS Macro Lett. 2012, 1, 1098−1102. (31) Mayershofer, M. G.; Nuyken, O.; Buchmeiser, M. R. Bi- and Trinuclear Ruthenium Alkylidene Triggered Cyclopolymerization of 1,6-Heptadiynes: Access to An−X−An Block and (An)3X Tristar Copolymers. Macromolecules 2006, 39, 3484−3493. (32) Song, W.; Han, H.; Wu, J.; Xie, M. Ladder-like polyacetylene with excellent optoelectronic properties and regular architecture. Chem. Commun. 2014, 50, 12899−12902. (33) Song, W.; Han, H.; Wu, J.; Xie, M. A bridge-like polymer synthesized by tandem metathesis cyclopolymerization and acyclic diene metathesis polymerization. Polym. Chem. 2015, 6, 1118−1126. (34) Vygodskii, Y. S.; Shaplov, A. S.; Lozinskaya, E. I.; Vlasov, P. S.; Malyshkina, I. A.; Gavrilova, N. D.; Kumar, P. S.; Buchmeiser, M. R. Cyclopolymerization of N,N-Dipropargylamines and N,N-Dipropargyl Ammonium Salts. Macromolecules 2008, 41, 1919−1928. (35) Song, W.; Han, H.; Liao, X.; Sun, R.; Wu, J.; Xie, M. Metathesis Cyclopolymerization of Imidazolium-Functionalized 1,6-Heptadiyne toward Polyacetylene Ionomer. Macromolecules 2014, 47, 6181−6188. (36) Naumann, S.; Unold, J.; Frey, W.; Buchmeiser, M. R. Regioselective Cyclopolymerization of 1,7-Octadiynes. Macromolecules 2011, 44, 8380−8387. (37) Lee, I. S.; Kang, E.-H.; Park, H.; Choi, T.-L. Controlled cyclopolymerisation of 1,7-octadiyne derivatives using Grubbs catalyst. Chem. Sci. 2012, 3, 761−765. (38) Unold, J.; Wang, D.; Frey, W.; Buchmeiser, M. R. Stereo- and regioselective cyclopolymerization of chiral 1,7-octadiynes. Polym. Chem. 2013, 4, 4219−4233.

(39) Park, H.; Lee, H.-K.; Choi, T.-L. Faster cyclopolymerisation of 4,4-disubstituted 1,7-octadiynes through an enhanced Thorpe−Ingold effect. Polym. Chem. 2013, 4, 4676−4681. (40) Park, H.; Lee, H.-K.; Kang, E.-H.; Choi, T.-L. Controlled cyclopolymerization of 4,5-disubstituted 1,7-octadiynes and its application to the synthesis of a dendronized polymer using Grubbs catalyst. J. Polym. Sci., Part A: Polym. Chem. 2015, 53, 274−279. (41) Herz, K.; Unold, J.; Hänle, J.; Schowner, R.; Sen, S.; Frey, W.; Buchmeiser, M. R. Mechanism of the Regio- and Stereoselective Cyclopolymerization of 1,6-Hepta- and 1,7-Octadiynes by High Oxidation State Molybdenum−Imidoalkylidene N-Heterocyclic Carbene Initiators. Macromolecules 2015, 48, 4768−4778. (42) Schmidt, C.; Buchmeiser, M. R. Peculiarities of the RuIVAlkylidene Triggered Cyclopolymerization of N-(bis(alkyloxy)aryl)Containing 1,6-Heptadiynes. Macromol. Symp. 2007, 254, 370−374. (43) Song, J.-A.; Park, S.; Kim, T.-S.; Choi, T.-L. N-Containing 1,7Octadiyne Derivatives for Living Cyclopolymerization Using Grubbs Catalysts. ACS Macro Lett. 2014, 3, 795−798. (44) Grubbs, R. H. Handbook of Metathesis, 2nd ed.; Wiley-VCH: Weinheim, 2015; Vols. 2, 3. (45) Grela, K. Olefin Metathesis: Theory and Practice; Wiley-VCH: Weinheim, 2014. (46) Lee, H. K.; Bang, K. T.; Hess, A.; Grubbs, R. H.; Choi, T. L. Multiple Olefin Metathesis Polymerization That Combines All Three Olefin Metathesis Transformations: Ring-Opening, Ring-Closing, and Cross Metathesis. J. Am. Chem. Soc. 2015, 137, 9262−9265. (47) Poulsen, C. S.; Madsen, R. Enyne Metathesis Catalyzed by Ruthenium Carbene Complexes. Synthesis 2003, 1, 1−18. (48) Diver, S. T.; Giessert, A. J. Enyne Metathesis (Enyne Bond Reorganization). Chem. Rev. 2004, 104, 1317. (49) Dias, E. L.; Nguyen, S. T.; Grubbs, R. H. Well-Defined Ruthenium Olefin Metathesis Catalysts: Mechanism and Activity. J. Am. Chem. Soc. 1997, 119, 3887−3897. (50) Sanford, M. S.; Henling, L. M.; Grubbs, R. H. Synthesis and Reactivity of Neutral and Cationic Ruthenium(II) Tris(pyrazolyl)borate Alkylidenes. Organometallics 1998, 17, 5384−5389. (51) Lynn, D. M.; Mohr, B.; Grubbs, R. H. Living Ring-Opening Metathesis Polymerization in Water. J. Am. Chem. Soc. 1998, 120, 1627−1628. (52) Lynn, D. M.; Mohr, B.; Grubbs, R. H.; Henling, L. M.; Day, M. W. Water-Soluble Ruthenium Alkylidenes: Synthesis, Characterization, and Application to Olefin Metathesis in Protic Solvents. J. Am. Chem. Soc. 2000, 122, 6601−6609. (53) Gallivan, J. P.; Jordan, J. P.; Grubbs, R. H. A neutral, watersoluble olefin metathesis catalyst based on an N-heterocyclic carbene ligand. Tetrahedron Lett. 2005, 46, 2577−2580. (54) Forman, G. S.; McConnell, A. E.; Tooze, R. P.; Janse van Rensburg, W.; Meyer, W. H.; Kirk, M. M.; Dwyer, C. L.; Serfontein, D. W. A Convenient System for Improving the Efficiency of FirstGeneration Ruthenium Olefin Metathesis Catalysts. Organometallics 2005, 24, 4528−4542. (55) Ledoux, N.; Allaert, B.; Schaubroeck, D.; Monsaert, S.; Drozdzak, R.; Van Der Voort, P.; Verpoort, F. In situ generation of highly active olefin metathesis initiators. J. Organomet. Chem. 2006, 691, 5482−5486. (56) Meyer, W. H.; McConnell, A. E.; Forman, G. S.; Dwyer, C. L.; Kirk, M. M.; Ngidi, E. L.; Blignaut, A.; Saku, D.; Slawin, A. M. Tin and iron halogenides as additives in ruthenium-catalyzed olefin metathesis. Inorg. Chim. Acta 2006, 359, 2910−2917. (57) Simocko, C.; Wagener, K. B. Effects of Boron-Containing Lewis Acids on Olefin Metathesis. Organometallics 2013, 32, 2513−2516. (58) Kingsbury, J. S.; Harrity, J. P. A.; Bonitatebus, P. J.; Hoveyda, A. H. A Recyclable Ru-Based Metathesis Catalyst. J. Am. Chem. Soc. 1999, 121, 791. (59) Trnka, T. M.; Dias, E. L.; Day, M. W.; Grubbs, R. H. Ruthenium alkylidene complexes coordinated with tricyclohexylphosphine and heterocyclic N-donor ligands. Arkivoc 2002, 13, 28−41. (60) Wu, Z.; Nguyen, S. T.; Grubbs, R. H.; Ziller, J. W. Reactions of Ruthenium Carbenes of the Type (PPh3)2(X)2RuCH-CHCPh2 J

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules (X = Cl and CF3COO) with Strained Acyclic Olefins and Functionalized Olefins. J. Am. Chem. Soc. 1995, 117, 5503−5511. (61) Chang, S.; Jones, L.; Wang, C.; Henling, L. M.; Grubbs, R. H. Synthesis and Characterization of New Ruthenium-Based Olefin Metathesis Catalysts Coordinated with Bidentate Schiff-Base Ligands. Organometallics 1998, 17, 3460−3465. (62) Buchowicz, W.; Mol, J. C.; Lutz, M.; Spek, A. L. A novel ruthenium carbene dimer that is active in the metathesis of internal alkenes; the crystal structure of Ru2(CHPh)2(CF3CO2)2(mCF3CO2)2(PCy3)2(m-H2O). J. Organomet. Chem. 1999, 588, 205−210. (63) Sanford, M. S.; Henling, L. M.; Day, M. W.; Grubbs, R. H. Ruthenium-Based Four-Coordinate Olefin Metathesis Catalysts. Angew. Chem., Int. Ed. 2000, 39, 3451−3453. (64) Buchowicz, W.; Ingold, F.; Mol, J. C.; Lutz, M.; Spek, A. L. Novel Ruthenium(II)2 Carboxylates as Catalysts for Alkene Metathesis. Chem. - Eur. J. 2001, 7, 2842−2847. (65) Nieczypor, P.; Buchowicz, W.; Meester, W. J.; Rutjes, F. P.; Mol, J. C. Synthesis and application of a new polystyrene-supported ruthenium carbene catalyst for alkene metathesis. Tetrahedron Lett. 2001, 42, 7103−7105. (66) Conrad, J. C.; Amoroso, D.; Czechura, P.; Yap, G. P.; Fogg, D. E. The First Highly Active, Halide-Free Ruthenium Catalyst for Olefin Metathesis. Organometallics 2003, 22, 3634−3636. (67) Vehlow, K.; Maechling, S.; Köhler, K.; Blechert, S. Versatile Rubased metathesis catalysts designed for both homogeneous and heterogeneous processes. J. Organomet. Chem. 2006, 691, 5267−5277. (68) Buchowicz, W.; Makal, A.; Woźn iak, K. The bis(trifluoroacetate) analogue of the first-generation Grubbs catalyst: Synthesis, X-ray structure, and metathesis activity of [Ru(CF3CO2)(η2CF3CO2)(CHPh) (PCy3)2]. J. Organomet. Chem. 2009, 694, 3179− 3183. (69) Jung, K.; Kang, E.-H.; Sohn, J.-H.; Choi, T.-L. Highly β-Selective Cyclopolymerization of 1,6-Heptadiynes and Ring-Closing Enyne Metathesis Reaction Using Grubbs Z-Selective Catalyst: Unprecedented Regioselectivity for Ru-Based Catalysts. J. Am. Chem. Soc. 2016, 138, 11227−11233. (70) Kang, E.-H.; Choi, T.-L. Coil-to-Rod Transition of Conjugated Polymers Prepared by Cyclopolymerization of 1,6-Heptadiynes. ACS Macro Lett. 2013, 2, 780−784. (71) It seems that even DP = 2 and DP > 2 carbene peaks are observed separately since this broad peak contains two distinct peaks. As shown in Figure 2A, middle, the peak at 20.32 ppm disappears and 20.29 ppm increases as polymerization proceeds, implying that those two peaks would be DP = 2 and DP > 2 carbenes, respectively. (72) Schwab, P.; France, M. B.; Ziller, J. W.; Grubbs, R. H. A Series of Well-Defined Metathesis Catalysts-Synthesis of [RuCl2(= CHR′) (PR3) 2] and Its Reactions. Angew. Chem., Int. Ed. Engl. 1995, 34, 2039−2041. (73) Matyjaszewski, K.; Xia, J. Atom Transfer Radical Polymerization. Chem. Rev. 2001, 101, 2921−2990. (74) Ouchi, M.; Terashima, T.; Sawamoto, M. Precision Control of Radical Polymerization via Transition Metal Catalysis: From Dormant Species to Designed Catalysts for Precision Functional Polymers. Acc. Chem. Res. 2008, 41, 1120−1132. (75) Ouchi, M.; Terashima, T.; Sawamoto, M. Transition MetalCatalyzed Living Radical Polymerization: Toward Perfection in Catalysis and Precision Polymer Synthesis. Chem. Rev. 2009, 109, 4963−5050. (76) Matyjaszewski, K. Atom Transfer Radical Polymerization (ATRP): Current Status and Future Perspectives. Macromolecules 2012, 45, 4015−4039. (77) Kirchner, K.; Calhorda, M. J.; Schmid, R.; Veiros, L. F. Mechanism for the Cyclotrimerization of Alkynes and Related Reactions Catalyzed by CpRuCl. J. Am. Chem. Soc. 2003, 125, 11721−11729. (78) Yamamoto, Y.; Arakawa, T.; Ogawa, R.; Itoh, K. Ruthenium(II)Catalyzed Selective Intramolecular [2 + 2 + 2] Alkyne Cyclotrimerizations. J. Am. Chem. Soc. 2003, 125, 12143−12160.

(79) Yang, S.; Shin, S.; Choi, I.; Lee, J.; Choi, T.-L. Direct Formation of Large-Area 2D Nanosheets from Fluorescent Semiconducting Homopolymer with Orthorhombic Crystalline Orientation. J. Am. Chem. Soc. 2017, 139, 3082−3088.

K

DOI: 10.1021/acs.macromol.7b00488 Macromolecules XXXX, XXX, XXX−XXX