Enantioselective Copper-Catalyzed sp2/sp3 Diborylation of 1-Chloro-1-Trifluoromethylalkenes.

Fluorine-containing organoboron compounds have emerged as novel building blocks in chemical synthesis; among them, fluorinated sp2/sp3 diborylated compounds are particularly appealing, since they might undergo chemoselective and diversified transformations of different C-B bonds with fluorinated functionality, thus bringing versatility and complexity to the eventual products. However, expedient, synthetic strategies for the construction of such fluorinated diborylative compounds are very sparse. Herein, we disclose enantioselective Cu-catalyzed sp2/sp3 diborylations of 1-chloro-1-trifluoromethylalkenes, leading to diborylated compounds bearing a gem-difluoroalkenyl moiety; most intriguingly, the new formed C-B bonds include one stereoselective and optically pure Csp3-B bond. Further transformations on the eventual products demonstrated the values of our presented strategy.

Introduction

Organoboron compounds, as one of the most important building blocks, have been widely used in organic synthesis,[1−7] and numerous methods have been developed to construct them.[8−13] Among different types of organoboron compounds, mono-organoborons are the most well-studied[14−23] (Figure A, left). Compared with mono-organoboron compounds, diborons are especially fascinating synthons since the two boron moieties could bring more complex and diversified structural elaborations.[24−27] There are several types of diborons that have appeared in the literature;[28,29] however, the two C–B bonds of the same type show up on the same structure.[30−36] In sharp contrast, molecules that contains two different types of C–B bonds, especially with one enantioenriched Csp3–B bond, are less-developed (Figure A, right).[37−42]

Figure 1

Challenges in constructing diborons with two different types of C–B bonds and our strategy.

Fluorine-containing organoboron compounds have emerged as novel synthons in chemical synthesis due to the existence of two important functionalities—fluorine and boron on one molecule scaffold, which brings versatility and complexity to the eventual products.[43−53] Among them, the fluorinated ones bearing multiple types of C–B bonds are particularly appealing, since they might undergo chemoselective and diversified transformations on different C–B bonds, leading to well-controlled and multifunctionalized complex targeted molecules.[54] However, successful examples of constructions of multiborylated compounds with distinct C–B bonds are very rare, particularly for enantioselective synthesis, and most of them are focused on the transformations of allenes. For instance, Morken and co-workers[38−40] reported a Pd-catalyzed protocol for asymmetric diborations from allenes. Recently, Tang and Ding[37] et al. developed a similar strategy to realize asymmetric diboration once again from allenes in which a chiral tertiary boronic ester was constructed (Figure B, left). Of note, the two aforementioned synthetic methods could not lead to fluorine-containing multiborylated molecules. In 2020,[54] our group disclosed novel Cu-catalyzed regio- and stereodivergent chemoselective diborylations of CF3-containing 1,3-enynes, rendering three different sp2/sp3 diborylated products bearing CF3 functionality (Figure B, right). As we all know, fluorinated compounds have unique and fascinating physicochemical and biological properties, and the fluorine atom is a bioisostere replacement of the hydrogen atom;[55−58] specifically, the gem-difluoroalkenyl moiety has been considered as isosteric and isopolar synthons to carbonyl groups, which play a critical role in pharmaceuticals and drug designs, and is thus one of the most attractive fluorine-containing structural motifs.[59−63] Given the importance of fluorinated sp2/sp3 diborylated compounds and the paucity of their efficient construction, as well as our long-term interests in both organofluorine[64−71] and organoboron[72−79] chemistry, herein, we report the first expedient Cu-catalyzed enantioselective sp2/sp3 diborylations of 1-chloro-1-trifluoromethylalkenes, in which one Csp2–B bond and one Csp3–B bond are constructed simultaneously in a one-vessel strategy along with the formation of a gem-difluoroalkenyl functionality (Figure C). The reaction features readily accessible starting materials, high enantioselectivity, and broad substrate scope with excellent functionality tolerance. Most intriguingly, two different types of C–B bonds are chemo- and stereoselectively constructed together with the formation of a gem-difluoroalkenyl moiety, thus leading to one new type of boron and fluorine-substituted chiral allylic boronates, which could be readily converted into various valuable products as demonstrated by further synthetic applications in this Article.

Results

Optimization Studies of the Racemic Process

To commence the investigation of racemic diborylation, 1-(2-chloro-3,3,3-trifluoroprop-1-en-1-yl)-4-methoxybenzene (1) was chosen as a model compound, together with B2pin2 as the borylation reagent, BuONa as the base, in the presence of copper salt, and with PCy3 as the ligand (in Table ). Pleasingly, the combination afforded the desired product 2a in a 37% yield when catalytic CuCl2 was evaluated (Table , entry 1). Different copper salts, such as CuF2, CuBr2, CuSO4, CuOTf, CuTc, CuCl, and Cu(OTf)2, were explored (Table , entries 2–8); among these salts, Cu(OTf)2 provided the best result with 87% yield. Subsequently, various reaction temperatures were investigated, and it turned out that 60 °C was the temperature that led to the optimal outcome (Table , entries 9–12). Different amounts of base caused insignificant reductions in the yields of this reaction (Table , entries 13–15).

Table 1

Optimization Studies of the Racemic Processa

entry	catalyst	base	T (°C)	yield (%)
1	CuCl2	tBuONa	60	37
2	CuF2	CuCl2	60	52
3	CuBr2	tBuONa	60	46
4	CuSO4	tBuONa	60	45
5	CuOTf	tBuONa	60	63
6	CuTc	tBuONa	60	26
7	CuCl	tBuONa	60	54
8	Cu(OTf)2	tBuONa	60	87 (81b)
9	Cu(OTf)2	tBuONa	40	49
10	Cu(OTf)2	tBuONa	50	68
11	Cu(OTf)2	tBuONa	70	59
12	Cu(OTf)2	tBuONa	80	63
13c	Cu(OTf)2	tBuONa	60	75
14d	Cu(OTf)2	tBuONa	60	82
15e	Cu(OTf)2	tBuONa	60	78

Conditions: the reaction was carried out with 1 (0.5 mmol), B2pin2 (3.5 equiv), [Cu] (5 mol %), PCy3 (6 mol %), and BuONa (1.5 equiv) in THF (2.5 mL) at 60 °C for 24 h.

Isolated yield.

BuONa 1.0 equiv.

BuONa 1.8 equiv.

BuONa 2.0 equiv. The yield was determined by GC using n-dodecane as an internal standard.

Substrate Scope for the Racemic Process

After identification of the optimal reaction conditions (Table , entry 8), the substrate generality of this Cu-catalyzed diborylation of 1-chloro-1-trifluoromethylalkenes was investigated (Figure ). Various unactivated aliphatic 1-chloro-1-trifluoromethylalkenes were examined and gave satisfactory to excellent reaction outcomes. A variety of substituents such as Me, MeO, CF3, F, Cl, COOMe, as well as Br on the benzene ring were very compatible with the transformation to afford the desired products 4–12 in good yields. Substitution position did not affect the outcomes at all, since bromo groups at the meta or ortho positions of the benzene ring also smoothly afforded the target products 10 and 12 in 78% and 70% yields, respectively; no loss of efficiency was ever observed compared with the para position (11, 69%). Aromatic rings containing aliphatic substituted alkenes with different chain lengths and different substitutions were all good candidates and transformed into the corresponding target products under our standard conditions in good to excellent yields (13–22). Heteroatomic cyclic rings, such as furan, thiophene, tetrahydropyran, and various N-protected piperidines, were all tolerated in this transformation, and the corresponding desired products were obtained in decent yields (23–28). Long, linear aliphatic chains with various functionalities (ether, ester, sily ether, and Br) also showed good capabilities to lead to the target molecules 29–32 in good to excellent yields. Cyclic rings, such as cyclohexyl (33 and 37), cyclopentyl (34 and 36), cyclopropyl (35), and adamantyl (38), also demonstrated good compatibility. In terms of the phenylcyclopropyl ring, initially, no desired product 35 was obtained due to the steric hindrance that comes from the cyclopropane ring, but after increasing the reaction temperature and prolonging the reaction time, the target product 35 was afforded in decent yield. A series of substrates that were derived from bioactive or drug molecules like naproxen, ibuprofen, flurbiprofen, and indomethacin were all suited under our standard conditions to render the corresponding products in moderate to good yields (39–42).

Figure 2

Scope of the racemic reaction. (a) Reaction conditions: Cu(OTf)2 (5 mol %), PCy3 (6 mol %), BuONa (1.3 equiv), and B2pin2 (4.0 equiv) were mixed in 2.5 mL of THF, and 1 (0.5 mmol) was added subsequently at room temperature; then, the resulting mixture was stirred at 60 °C for 24 h. (b) 80 °C. (c) 36 h. (d) 16 h.

To our delight, this chemistry was not unique to aliphatic substituted 1-chloro-1-trifluoromethylalkenes; the (hetero)aromatic substituted substrates were also compatible with this transformation to generate the corresponding desired diborylation products. Various substituents such as alkyl groups (Me, Et, iPr, and Bu), ethers and thioethers (CH3O, CF3O, BnO, and CH3S), as well as a phenyl group on the benzene ring were all tolerable under our standard conditions, and the corresponding products (43–52) were obtained in a moderate to good yields. Heteroaromatic substrates such as furan, thiophene, and 2,3-dihydrobenzofuran demonstrated good reactivities as well, and the corresponding target compounds (53–55) were isolated in good yields. A disubstituted aromatic substrate was also a good candidate to lead to the appealing product 56 in 70% yield.

Inspired by the aforementioned outcomes, we envisage that substrates bearing other types of (pseudo)halogen instead of the Cl group should be compatible with our transformation as well to render the same products. Therefore, a similar substrate (Z)-1,1,1-trifluoro-4-phenylbut-2-en-2-yl trifluoromethanesulfonate was first prepared (Figure , bottom); gratifyingly, when the new substrate was subjected to our standard conditions, the reaction indeed occurred, and the corresponding desired product 13 was obtained in 80% yield with a shorter reaction time. Just like its chloro congener, the new substrates also demonstrated good functional group tolerance; both electron-donating and electron-withdrawing groups on the benzene rings were compatible with this standard condition, and the target products 57–59 were obtained in excellent yields.

Encouraged by the successful racemic synthesis of diborylative products, we were prompted to try an asymmetric version of this transformation. A set of structurally relevant chiral phosphine ligands (Table ) were extensively evaluated. Ferrocene-type ligands (L1–L3) were first chosen and showed relatively good results, delivering 42–76% isolated yields of diborylation products 60 with fair to good er values (Table , entries 1–3). Following a systematic examination of commercial chiral bidentate ligands (Table , entries 4–7), we were delighted to find that ligand L5 could provide enantioenriched 60 in 90:10 er (Table , entry 5); further examination indicated that ligand L8 could enantioselectively provide 60 in 95:5 er while with 67% yield (Table , entry 8). Subsequently, we replaced THF with MTBE as the solvent, remarkably, the yield of the desired product 60 was dramatically increased to 81% yield along with 96.5:3.5 er (Table , entry 9). Of note, after everything was added to the reaction mixture, instead of controlling the reaction temperature at 15 °C, we allowed the mixture to warm to ambient temperature; optically active 60 was obtained in 83% isolated yield with 93:7 er (Table , entry 10).

Table 2

Discovery and Evaluation of Cu-Catalyzed Asymmetric Diborylationa

Cu(OTf)2 (5 mol %), ligand (7 mol %), BuONa (1.3 equiv), and B2pin2 (4.0 equiv) were first mixed in 2.5 mL of THF, and 1 (0.5 mmol) was added subsequently at 0 °C; then, the temperature was slowly raised to 15 °C, and the mixture was stirred for 60 h.

Isolated yield.

The enantioselectivity was determined by chiral HPLC.

MTBE (methyl tert-butyl ether) (3 mL) instead of THF (3 mL).

0 °C and then slowly raised to 25 °C and stirred for 60 h.

Having arrived at the optimized conditions for the enantiocontrolled reaction, we next explored the scope of 1-chloro-1-trifluoromethylalkenes (Figure ). As with the racemic reactions, a wide range of aliphatic substituted alkenes could be successfully converted to enantioenriched forms of target products. A variety of substituents such as Me, MeO, F, Cl, Br, and ester on the para position of the benzene ring were all very compatible under the standard conditions, delivering the respective products in good to excellent yields with high enantioselectivities (61–67). When the bromo group was on the meta or ortho positions of the benzene ring, the corresponding products could be unperturbed and obtained in good yields albeit with a slight reduction in enantioselectivity (68 and 69). Aryl-substituted alkyl chains with different lengths were all suitable substrates and delivered the desired products (70–75) in excellent yields with high enantiocontrol. The absolute configuration of the major enantiomer was determined as S by the single crystal structure of 75. Heteroatomic cyclic rings, including furan, thiophene, tetrahydropyran, and N-protected piperidines, were also tolerated very well under our chiral standard conditions, and the target molecules (76–81) were rendered in excellent yields with high enantioselectivity. In addition, ether and ester were also compatible in our transformations to lead to the corresponding enantioenriched products (82 and 83). Cyclic rings, such as cyclohexyl (84 and 87), cyclopentyl (85 and 86), and adamantane (88), were all smoothly transformed into the desired products in excellent yields and beautiful er values. Most remarkably, in addition to aliphatic substituted 1-chloro-1-trifluoromethylalkenes, we found that their styrene counterparts also demonstrated good reactivity under the chiral standard conditions with good functional group tolerance to afford the corresponding enantioenriched products with moderate yields and good er values (89–92). Furthermore, when the chloro groups on the starting alkenes were replaced with a triflate one, the desired products 93 and 94 were afforded in high yields and good er values. To further demonstrate the good functional group tolerance and broad substrate scope of our transformation, bioactive and complex molecule-derived substrates were evaluated in our chiral standard conditions as well; to our delight, those substrates were very compatible, and the desired enantioenriched diborylative target products were consistently generated in good yields with high stereoselectivities (95–97).

Figure 3

Scope of the asymmetric reaction. (a) Reaction conditions: Cu(OTf)2 (5 mol %), ligand (7 mol %), BuONa (2 equiv), and B2pin2 (4.0 equiv) were mixed in MTBE (3 mL), and 1 (0.3 mmol) was added subsequently at 0 °C; the resulting mixture was allowed to slowly rise to 15 °C and was stirred for 60 h. (b) 0 °C.

Synthetic Applications

To exemplify the synthetic applications of our chiral diborylative products in asymmetric chemical synthesis, various stereospecific transformations on the optically active diborylative products were performed (Figure ). First of all, chiral allylic alcohols 98–101[43] were readily accessible through three steps in a single-vessel reaction from the initial 1-chloro-1-trifluoromethylalkenes with moderate to good yields and high enantioselectivities (Figure A), which provides an alternative route to enable the very important allylic alcohols in a simple yet powerful way. Furthermore, the enantioenriched diborylative products are highly versatile chiral synthetic intermediates since two different types of C–B bonds exist on such molecules; therefore, they can be readily elaborated into a variety of valuable optically active compounds, by selective derivatization either on Csp2–B bonds or on Csp3–B bonds. For instance, selective functionalizations of Csp2–B bonds on chiral diboron 60, such as bromination, methylation, thiophenolation, vinylation, benzylation, as well as arylation, could lead to the corresponding products 102–107[80−83] in good yields with excellent enantioretention. Starting from compound 107, versatile transformations on Csp3–B bonds were successfully performed; for example, potassium trifluoroborate 108(84) could be easily obtained with KHF2 from 107, and 109(43) could be obtained by one-carbon homologation of allylic boronates with ClCH2Li. The chiral boron 107 could undergo a complete hydrogenation with Pd/C catalyst to render a chiral boronate 110(43) bearing a difluoromethyl group with a fair dr value but good er value; compound 111(85) was formed by cross-coupling from 107 and furan with n-BuLi and NBS; in the presence of vinylMgBr, I2, and NaOMe, a sequential reaction occurred to lead to the vinylated product 112(86) (Figure B).

Figure 4

Synthetic applications and transformations of the diborylative products.

Mechanism Studies

In order to thoroughly understand the reaction mechanism, several control experiments were carried out. First, when the reaction was performed in the absence of copper salt or base, the target product 3 was not detected, which inferred that the Cu catalyst and base played key roles for the success of this transformation (Figure A). When the reaction was performed in the absence of B2pin2 (Figure B), compound 114 was not observed; the result of this experiment inferred that alkyne 114 was not the intermediate in this transformation. When we lowered the reaction temperature to 0 °C, compound 113 with only one boron moiety was obtained in 41% isolated yield, along with 25% yield of target product 3 (Figure C). When monoboron 113 was subjected to the reaction with only 1.5 equiv of B2pin2 and 0.5 equiv of base, the desired target product 3 was obtained in 85% isolated yield (Figure D). Intriguingly, when monoboron 113 was exposed to asymmetric conditions, the corresponding desired chiral product 60 was obtained in 91% yield with a 97:3 er value (Figure E). Based on the above experiments, we have sufficient reasons to believe that compound 113 is the key intermediate of the reaction. To further validate our hypothesis, we monitored the yield changes of intermediate 113 and product 3 over time, and we found that the yield of 3 gradually increased as time went on during the reaction; meanwhile, the yields of compound 113 initially increased in the first 2 h and then decreased gradually after that (Figure F). These results further confirmed the intermediacy of compound 113 in our transformation.

Figure 5

Mechanism studies.

On the basis of the above results and previous reports,[43,45,49,87] we proposed a tentative reaction mechanism for this copper-catalyzed diborylation of 1-chloro-1-trifluoromethylalkenes, as shown in Figure . First, copper(I) alkoxide intermediate A is formed between Cu(OTf)2, ligand, and BuONa, which further reacts with diboron (B2pin2) to afford the Cu—B complex B. Subsequently, the C=C double bond of the starting material 1 inserts into the complex B to form copper(I) intermediate C; further β-chloro elimination of intermediate C gives the new olefin D (path a), or the Cu(I)–B complex goes through an oxidative addition to vinyl-Cl bond to afford intermediate C′, which further converts into olefin D after reductive elimination (path b). Then, the C=C double bond of olefin D regioselectively inserts into Cu—B complex B again to form copper intermediate E; subsequent β-fluoro elimination of E finally delivers the desired diborylative product along with LCuF (F). The formed copper(I) fluoride F reacts further with diboron (B2pin2)[88,89] to regenerate Cu—B complex B to complete the catalytic cycle. Of note, after the second C=C bond insertion to the Cu—B complex, β—F elimination occurs to lead to a new alkene (gem-difluoroalkene; thus, the final chirality center has nothing to do with the Z isomer or E isomer of the starting 1-chloro-1-trifluoromethylalkenes, and both of them could give the same single enantiomer.

Figure 6

Proposed mechanisms.

In conclusion, we developed a facile enantioselective copper-catalyzed diborylation reaction from readily accessible 1-chloro-1-trifluoromethylalkenes, commercially available diboron reagent, and inexpensive Cu catalyst to produce a diverse array of enantioenriched gem-difluoroallyl diboronates. In addition to the universality of substrates, this method is also suitable for bioactive and complex molecules and could be a very flexible conversion into versatile enantioenriched gem-difluoroallyl skeletons. We anticipate that this strategy based on the diversity of boron chemistry will simplify the synthesis and enhance structural elaborations of gem-difluoroalkene targets for chemistry, biology, and pharmaceutical and medicinal chemistry.