Jiaqiang Lia,
Qing Liua,
Hang Shena,
Ruofeng Huanga,
Xiaohui Zhanga,
Yan Xiong*ab and
Changguo Chen*a
aSchool of Chemistry and Chemical Engineering, Chongqing University, Chongqing 400044, China. E-mail: xiong@cqu.edu.cn; cgchen@cqu.edu.cn
bChongqing Institute of Green and Intelligent Technology, Chinese Academy of Science, Chongqing 400714, China
First published on 30th September 2015
A novel ether/BF3 reductive system has been described, in which diphenylmethanols and their ether and ester derivatives are used as starting materials. Reductions are performed in ether under reflux and an argon atmosphere, and the addition of extra water is beneficial to this reduction. A series of alkanes are able to be prepared with good to excellent yields. A deuterated experiment exhibits that the reductive hydrogen is generated from ether. The mechanism is discussed in detail to explain the observed reactivity.
BF3·OEt2, generated from the donor–acceptor action between BF3 and Et2O,14 has shown high activity in many synthetic procedures like alkylation,15a–d cyclization,15e rearrangement,15f coupling,15g cyanation15h and etherification reactions.15i It was also utilized in the reduction of alcohols,16a,b carbonyl compounds16c and epoxides16d via hydride transfer from hydrosilanes. Our studies have demonstrated that the superacid BF3–H2O promoted benzylation reaction via a carbonium intermediate.17 We found that when iodobenzene was involved as both arene and solvent, the benzylation product was not produced and the undesired diphenylmethane was obtained in 47% yield by a reductive mechanism. The source of hydrogen was thought to come from the diphenylmethanol or diethyl ether by disproportionation, accompanied with the generation of reductive hydrogen. However, no diphenylketone was detected, which indicated that hydrogen was released very probably from the selfoxidation of ether ligand of BF3·OEt2. To test this theory, control experiments were performed in which diphenylmethanol and BF3·OEt2 in dichloromethane led to a 22% yield (see ESI†) and the later deuterated experiment resulted in the D-labelled methene.
The studies on triphenylmethyl cation, in the stable form of trityl salt, can date back to one century ago.18 Due to its importance in polymer of ether, it was regarded as an excellent initiator of polymerization and kinetically studied extensively covering the hydride ion transfer between triphenylcarbonium and ether. However, for this H transfer, there is no report on relatively unstable diphenylcarbonium ion, due to the unavailability of the salt. Herein, we would like to report our work in BF3·OEt2-promoted reductive reaction of diphenylcarbonium, using diphenylmethanols as well as their ether and ester derivatives as substrates, to corresponding alkanes, employing various ethers as hydrogen sources (Scheme 1).
Considering the hydrogen generated from the ether, we chose initially tetrahydrofuran (THF) as a solvent for the reduction of diphenylmethanol (1a), and as a result the yield was improved by up to 53% (Table 1, entry 1). As mentioned before,17 mixing BF3·OEt2 with H2O resulted in the formation of BF3–H2O which has shown elevated reactivity in the benzylation of arenes. We assessed the influence of different amounts of H2O in BF3·OEt2 promotion. 0.48 equivalent of water gave the best result while lower or higher amount of water turned out to be detrimental to this reduction (Table 1, entries 2–4). Then, the treatment of diphenylmethanol with catalytic amount (20 mol%) of BF3·OEt2 gave rise to diphenylmethane 2a in 20% yield (Table 1, entry 5). Increasing the amount of BF3·OEt2 from 0.6 to 0.8 equivalent resulted in a large improvement of the yield up to 61% (Table 1, entries 6 and 7). 1.0 equivalent of BF3·OEt2 furnished the desired product in 63% yield with the same level of reactivity compared to 1.2 equivalents of BF3·OEt2 (Table 1, entry 8 vs. 3). However, a lower yield was observed when further increasing the amount of BF3·OEt2 (Table 1, entry 9). A solvent screening with diethyl ether, and 2-methyl tetrahydrofuran was conducted and gave moderate yields (Table 1, entries 10 and 11). When methyl tert-butyl ether (MTBE) and isopropyl ether were utilized as solvent, the reactions went complex and the attempt to separate the product failed (Table 1, entries 12 and 13). To our delight, employing butyl ether as solvent exhibited excellent reactivity (Table 1, entry 14). Variations of the temperature from 100 °C to 160 °C were tested, in which an excellent yield of 87% was obtained under 160 °C (Table 1, entries 15–17). The optimal reaction conditions were established: 1.0 mmol diphenylmethanol, 1.2 equivalents of BF3·OEt2 in 1 mL of butyl ether in argon atmosphere under reflux for 2 h (Table 1, entry 17).
Entry | BF3·OEt2 (eq.) | H2O (eq.) | Solvent | Yieldb (%) |
---|---|---|---|---|
a Conditions: 1a (1.0 mmol), BF3·OEt2 (specified), H2O (specified) in solvent (1.0 mL) under argon and reflux (oil bath of 120 °C) for 2 h.b Isolated yields.c Complex reaction.d Oil bath of 100 °C.e Oil bath of 140 °C.f Oil bath of 160 °C. | ||||
1 | 1.2 | — | THF | 53 |
2 | 1.2 | 0.36 | THF | 55 |
3 | 1.2 | 0.48 | THF | 62 |
4 | 1.2 | 0.60 | THF | 52 |
5 | 0.2 | 0.48 | THF | 20 |
6 | 0.6 | 0.48 | THF | 52 |
7 | 0.8 | 0.48 | THF | 61 |
8 | 1.0 | 0.48 | THF | 63 |
9 | 1.5 | 0.48 | THF | 52 |
10 | 1.2 | 0.48 | Et2O | 53 |
11 | 1.2 | 0.48 | 2-Methyl THF | 55 |
12 | 1.2 | 0.48 | MTBE | —c |
13 | 1.2 | 0.48 | Isopropyl ether | —c |
14 | 1.2 | 0.48 | Butyl ether | 83 |
15 | 1.2 | 0.48 | Butyl ether | 80d |
16 | 1.2 | 0.48 | Butyl ether | 84e |
17 | 1.2 | 0.48 | Butyl ether | 87f |
a Conditions: 1 (1.0 mmol), BF3·OEt2 (1.2 mmol), H2O (0.48 mmol), oil bath of 160 °C, butyl ether (1.0 mL) under argon for 2 h. Isolated yields.b THF as solvent. Oil bath of 120 °C. |
---|
With the optimal reaction conditions in hand, diverse substituted diphenylmethanols bearing mono-/di-halo, methyl, and methoxyl groups were investigated with moderate to excellent yields Table 2. Considering the potential utilization of fluorine-containing compounds in pharmaceuticals and functionalized materials,19 fluoro-substituted diphenylmethanols were chosen as substrates such as 4-fluoro- (1b), 4,4′-difluoro- (1c), 3,4-difluoro- (1d) and 2-fluoro- (1e) substituents. As a result, the corresponding alkanes were obtained in yields of 72–93%. The ortho-fluoro substituent (1e) gave a relatively lower yield of 72% than the para-fluoro substitution, probably due to a large extend to the significant steric effects. The chloro- and bromo- substituted aromatics are capable of being further functionalized by SNAr reactions, Grignard reactions and coupling reactions, etc. Structurally diverse chloro- and bromo- substituents (1f–j) were chosen as substrates, and the yields of 53–94% were obtained. It is worthwhile noting that ortho-substitution gave lower yields in the reduction process (1f vs. 1g, h and 1i vs. 1j). When compounds with different halogen groups such as 1k (F, Cl) and 1l (F, Br) were used as substrates, good yields of 68% and 59% were obtained. Diphenylmethanols bearing electron-donating methyl groups such as methyl (1m) and methoxyl (1n) gave the yields of 93% and 24%, respectively. However compared with methyl substituent 1m, dramatically decreased yields were observed by using diphenylmethanol bearing electron-withdrawing trifluoromethyl (1o) and ester (1p) groups. Naphthalene derivative (1q) worked well providing the corresponding product in 76% yield. Interestingly, triphenylmethanol (1r) and triphenylmethanthiol (1s) gave rise to triphenylmethane in 95% and 91%, which suggests that thiol possesses the similar reactivity in this case. Despites many attempts, aliphatic alcohols and aromatic primary or secondary alcohols, did not undergo any reduction process and only adamantanol (1t) generated the corresponding adamentane in an isolated yield of 29%. Heterocyclic substrates (1u and 1v) were disable to give the dehydroxylation products. It is notable that prop-1-ene-1,1-diyldibenzene (2w) was obtained with 99% yield through β-H elimination when 1,1-diphenylpropan-1-ol (1w) was used as substrate, and 1,1-diphenylethanol (1x) only gave ethene-1,1-diyldibenzene (2x) in 55% yield (Scheme 2).
A series of leaving groups has been investigated in reductive reactions as well, such as alkoxy, hydroxymethoxy, and acetyoxy groups (Table 3). The diphenylmethane was produced in yields ranging from 37 to 90% depending on the nature of the leaving group. Methoxy diphenylmethane 3a provided the diphenylmethane in 71% yield. Regardless of linear (3b, 3c), branched (3d) or cyclic (3e) alkoxy diphenylmethyl ethers employed, the reductive products were obtained in good yields. Benzyloxy (3f) and allyloxy (3g) substituted substrates gave similar levels of yields, 37% and 42% respectively, which was probably explained by the released benzylalcohol and allylalcohol which further trap the promoter BF3·OEt2 to form carbonium ion. Interestingly, when doubling the amount of BF3·OEt2 in these two cases, both yields were improved to 60% and 59%. The reaction of the hydroxyethyl substituted substrate (3h) led to 2a in 70% yield, close to the yields obtained for alkoxy substitutions. Both acetoxy (3i) and trifluoroacetoxy (3j) were better leaving groups which resulted in yields of 83% and 90%. Diphenyl ether 3k provided the diphenylmethane in 60% yield under standard conditions. Doubling the amount of BF3·OEt2 improved the yield to 86%.
A large-scale experiment was carried out to demonstrate both the practicality and effectiveness of our method. 74% yield of diphenylmethane (2a) was obtained when treating 2 g of 1a under standard reaction conditions (Scheme 3).
Further experiments were performed to gain a better understanding of the reaction mechanism. Deuterated experiments employing THF-d8 were performed and the results showed the incorporation of the deuterium on the product 2m′ and 92% of D on 2m′, which was readily identified by 1H NMR analysis after silica column chromatography (Scheme 4). It is well-known that the p-σ hyperconjugation exists between the σ electron of α-H of ether and the lone pair electron on oxygen and hence the α-H could be activated through this p-σ hyperconjugation probably resulting in the formation of reductive hydrogen (H−), accompanied with the formation of oxonium ion, which could react with another ether molecule to produce the resonance stabilized species.18c–h Meanwhile, the diphenylmethyl cation could be produced in the presence of BF3 from diphenylmethanol (1). The carbonium ion interacted with reductive hydrogen would give rise to alkane product (2) (Fig. 1).
In summary, we have developed a novel ether/BF3 reductive system for diphenylmethanols and their ether and ester derivatives, presenting a metal-free strategy and affording the corresponding alkane products in good to excellent yields for most cases. The D-labelled experiment showed the reductive hydrogen was generated from ethers. The favourable safety profile, ease to handling and environmentally benign nature make this methodology particularly attractive and practical.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra14775a |
This journal is © The Royal Society of Chemistry 2015 |