Open Access Article
Chang-Sheng
Zhang
a,
Ya-Ping
Shao
a,
Fu-Min
Zhang
*a,
Xue
Han
a,
Xiao-Ming
Zhang
a,
Kun
Zhang
b and
Yong-Qiang
Tu
*ac
aState Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000, P. R. China. E-mail: tuyq@lzu.edu.cn; zhangfm@lzu.edu.cn
bSchool of Biotechnology and Health Sciences, Wuyi University, Jiangmen, 529020, Guangdong, P. R. China
cSchool of Chemistry and Chemical Engineering, Shanghai Jiao Tong University, Shanghai 200240, P. R. China
First published on 23rd June 2022
A novel classical kinetic resolution of 2-aryl-substituted or 2,3-disubstituted cyclobutanones of Baeyer–Villiger oxidation catalyzed by a Cu(II)/SPDO complex is reported for the first time, producing normal lactones in excellent enantioselectivities (up to 96% ee) and regioselectivities (up to >20/1), along with unreacted ketones in excellent enantioselectivities (up to 99% ee). The current transformation features a wide substrate scope. Moreover, catalytic asymmetric total syntheses of natural eupomatilones 5 and 6 are achieved in nine steps from commercially available 3-methylcyclobutan-1-one.
Over the past few years, our group has developed structurally unique spiro-pyrrolidine oxazoline (SPDO)-derived catalysts,8,9 which exhibit excellent properties in some catalytic asymmetric transformations, especially aerobic oxidative coupling reactions.9a,c Therefore, we wondered whether SPDO catalysts could be used as suitable ligands to catalyse the challenging classical kinetic resolution of 2-substituted cyclobutanone under B–V oxidation conditions.5w If successful, this transformation could not only solve the above-mentioned synthetic challenges but also further expand this rarely reported Cu-catalyzed asymmetric B–V oxidation, which would simultaneously provide two synthetically useful chiral products. Herein, we wish to report our research results as a communication (Scheme 1c).
:
1). Subsequently, other chiral SPDO ligands L2 and L3 were used to replace L1, but no better result was obtained (entries 5 and 6, Table 1).11 To our delight, the improved regioselectivity of lactones b and c was obtained in different halogenated solvents (entries 7–9, Table 1). For comprehensive consideration of the enantioselectivity of the two desired products 1a and 1b and regioselectivity of 1b and 1c, different ratios of mixed solvent of tetrahydrofuran (THF) and haloalkane were further screened. Better results were observed when the mixed solvent (THF
:
CHBr3 = 1
:
1) was used (entries 10–12, Table 1).11 When the model reaction was performed at −40 °C, improvements in the ee value of lactone 1b to 92% and regioselectivity ratio of 1b/1c to 15/1 were obtained, although the ee value of unreacted ketone 1a was only 72% (entry 13, Table 1). Inspired by Feng's excellent work in the B–V oxidations of a variety of substrates,5v–y various additives were introduced into the reaction mixture to further improve the enantioselectivity of ketone 1a, and the combination of Al(OiPr)3 and 4 Å molecular sieves (MS) showed the best reaction results.11 In this case, the enantioselectivity of unreacted ketone 1a increased to 91% ee, and the enantioselectivity of lactone 1b and regioselectivity of 1b/1c were also less influenced (entry 15, Table 1). On further decreasing the reaction temperature to −50 °C, no better result was obtained (entry 16, Table 1). Notably, the ligand ent-L1 exhibited similar reaction results, providing the enantiomers ent-1a and ent-1b, respectively (entry 17, Table 1). Therefore, the reaction parameters listed in entry 15 (Table 1) were selected as the optimal reaction conditions for subsequent investigations.
| Entry | Ligand | Lewis acid | Solvent (mL) | T (°C) | Time (h) | 1a: yieldb (%) | 1b + 1c: yieldb (%) | rsd |
|---|---|---|---|---|---|---|---|---|
| eec (%) | 1b: eec (%) | 1b/1c | ||||||
| a Reaction conditions: unless otherwise noted, the reactions were performed with 1a (0.2 mmol), Cu(NTf2)2 (10 mol%), ligand (12 mol%) and m-CPBA (1.0 equiv.) in THF (2.0 mL) at 0 °C. b Isolated yield. c Determined by UPC2 analysis. d The regioselectivity (rs) of b/c was determined by 1H NMR of crude products. e 2 mL THF and 2 mL haloalkane were used. f 4 Å MS (60 mg) was added. g Al(OiPr)3 (50 mol%) was used. | ||||||||
| 1 | L1 | Cu(OTf)2 | THF | 0 | 18 | 43/56 | 55/87 | 1.1/1 |
| 2 | L1 | Cu(NTf2)2 | THF | 0 | 10 | 45/70 | 52/74 | 4.7/1 |
| 3 | L1 | Cu(BF4)2·6H2O | THF | 0 | 10 | 50/34 | 45/86 | 1.7/1 |
| 4 | L1 | Cu(ClO4)2·6H2O | THF | 0 | 10 | 52/30 | 40/84 | 1.2/1 |
| 5 | L2 | Cu(NTf2)2 | THF | 0 | 18 | 48/35 | 50/45 | 3.8/1 |
| 6 | L3 | Cu(NTf2)2 | THF | 0 | 18 | 43/41 | 54/40 | 3.9/1 |
| 7 | L1 | Cu(NTf2)2 | DCM | 0 | 6 | 46/40 | 48/47 | 15.0/1 |
| 8 | L1 | Cu(NTf2)2 | DCE | 0 | 5 | 49/30 | 48/46 | 14.1/1 |
| 9 | L1 | Cu(NTf2)2 | CHCl3 | 0 | 5 | 46/57 | 50/70 | 11.2/1 |
| 10e | L1 | Cu(NTf2)2 | THF/DCM | 0 | 10 | 43/80 | 52/69 | 9.0/1 |
| 11e | L1 | Cu(NTf2)2 | THF/CHCl3 | 0 | 10 | 46/77 | 51/80 | 8.5/1 |
| 12e | L1 | Cu(NTf2)2 | THF/CHBr3 | 0 | 18 | 48/75 | 45/88 | 8.0/1 |
| 13e | L1 | Cu(NTf2)2 | THF/CHBr3 | −40 | 36 | 54/72 | 45/92 | 15.0/1 |
| 14e,f | L1 | Cu(NTf2)2 | THF/CHBr3 | −40 | 36 | 51/66 | 46/94 | 14.8/1 |
| 15e,f,g | L1 | Cu(NTf2)2 | THF/CHBr3 | −40 | 36 | 43/91 | 52/92 | 12.5/1 |
| 16e,f,g | L1 | Cu(NTf2)2 | THF/CHBr3 | −50 | 60 | 48/85 | 48/90 | 13.0/1 |
| 17e,f,g | ent-L1 | Cu(NTf2)2 | THF/CHBr3 | −40 | 36 | 44/−91 | 52/−92 | 12.5/1 |
With the optimal reaction conditions in hand, the substrate scope of 2-aryl cyclobutanone was investigated (Table 2). In most cases, high ee of γ-lactone b (82–94%) and unreacted ketone a (78–99%), decent rs of γ-lactones b/c, and good selectivity factor (s-factor) were obtained.5v,12 For the substrates bearing mono-substituents on the aryl ring, some experimental phenomena were observed from the reaction results, including the following: (1) The substituents on the aromatic ring at the para-position affected the reaction results, that is, the steric hindrance of substituents on the aromatic ring was greater, reaction time was longer, and ratio of γ-lactones b/c was lower (entries 1–5, Table 2). (2) The substituents at the ortho-position dramatically affected the desired B–V oxidation, and poor results were obtained even at room temperature (entry 3, Table 2). (3) An electron-donating group (EDG) on the arenes accelerated the reaction and improved the ratio of b/c (entries 7–10, Table 2), while electron-withdrawing groups (EWGs) exhibited the opposite effect on the reaction results (entries 11–15, Table 2), especially regarding the rs of γ-lactones (entries 7–10 & 11–14, Table 2). For bis-substituted aryl substrates, ketones 16a–19a or lactones 16b–19b were concisely isolated with excellent selectivity (entries 16–19, Table 2), and substituents at the 3,4-positions had less effect on the results. More importantly, the 3,4,5-tri-OMe-substituted substrate reacted well in terms of the resulting enantioselectivity and rs (entries 17 & 20, Table 2). Although 1-naphthyl cyclobutanone showed moderate rs with an excellent ee (entry 22, Table 2), the greater steric hindrance of the 2-naphthyl and 1-pyrenyl analogues improved the rs, enantioselectivities, and reactivities, and just 0.6 equiv. of m-CPBA was needed (21a, 23a). Notably, the absolute configuration of lactones 6b, 16b, and 21b was further determined by X-ray analysis. Additionally, substrates with heteroaromatic benzothiophenyl (24a), thienyl (25a, 26a), and benzofuranyl (27a) substituents were well tolerated, resulting in ketones 24a–27a and lactones 24b–27b with slightly decreased enantioselectivities; unsatisfactorily, substrate 24a showed moderate regioselectivity (4.0/1). Overall, whether for aryl rings with a mono-substituent or multi-substituents or for heteroaryl rings, the desired B–V oxidation reactions proceeded well in most cases, providing two valuable synthetic building blocks (2-aryl-substituted cycloketones and γ-lactones) with excellent enantioselectivities.
| Entry | Substituents | Time | a: yield/ee | b + c: yield/b: ee | b/c rs | s-factor |
|---|---|---|---|---|---|---|
| a Reaction conditions: unless otherwise noted, the reactions were performed with a (0.2 mmol), Cu(NTf2)2 (10 mol%), L1 (12 mol%), 60 mg 4 Å MS and m-CPBA (1.0 equiv.) in THF/CHBr3 (2.0/2.0 mL) for the indicated time at −40 °C. Isolated yield. ee was determined by UPC2 analysis. The rs of b/c was determined by crude 1H NMR. b The reaction was performed at 20 °C, and the ee value of 3c was 44.8%. c THF/Et2O/CHBr3 (1.0/1.0/2.0 mL) was used. d 0.6 equiv. of m-CPBA was used. e The ee value of 22c was 95%. f The ee value of 24c was 85%. g The absolute configuration of the products was confirmed by X-ray analysis. | ||||||
| 1 | 1a: R = Ph | 36 h | 43%/91% | 52%/93% | 12.5/1 | 88 |
| 2 | 2a: R = 4-Me-Ph | 48 h | 43%/98% | 57%/90% | 10.0/1 | 87 |
| 3 | 3a: R = 2-Me-Ph | 72 h | 52%/7% | 44%/61% | 1.5/1b,c | 4.4 |
| 4 | 4a: R = 4-iPr-Ph | 50 h | 44%/90% | 52%/91% | 9.2/1c | 65 |
| 5 | 5a: R = 4-tBu-Ph | 50 h | 53%/78% | 43%/92% | 10.0/1c | 57 |
| 6 | 6a: R = 4-Ph-Ph | 42 h | 45%/90% | 54%/91% | 12.5/1g | 65 |
| 7 | 7a: R = 4-MeO-Ph | 28 h | 44%/95% | 54%/91% | 16.6/1 | 79 |
| 8 | 8a: R = 3-MeO-Ph | 38 h | 47%/95% | 51%/91% | 12.4/1 | 79 |
| 9 | 9a: R = 4-PhO-Ph | 36 h | 45%/99% | 48%/93% | 12.5/1 | 145 |
| 10 | 10a: R = 4-BnO-Ph | 39 h | 43%/99% | 52%/88% | 13.0/1 | 82 |
| 11 | 11a: R = 4-F-Ph | 39 h | 49%/82% | 50%/92% | 6.3/1c | 61 |
| 12 | 12a: R = 3-F-Ph | 39 h | 48%/88% | 50%/91% | 5.0/1c | 62 |
| 13 | 13a: R = 4-Cl-Ph | 36 h | 43%/87% | 56%/92% | 5.9/1c | 68 |
| 14 | 14a: R = 4-Br-Ph | 38 h | 45%/85% | 50%/92% | 6.3/1c | 65 |
| 15 | 15a:R = 3-MeCO2Ph | 38 h | 45%/92% | 53%/91% | 6.1/1 | 70 |
| 16 | 16a: R = 2,3-dihydrobenzofuranyl | 38 h | 40%/95% | 56%/86% | 16.7/1g | 49 |
| 17 | 17a: R = 3,4-diMeO-Ph | 39 h | 40%/99% | 58%/87% | 14.3/1 | 75 |
| 18 | 18a: R = 3-F-4-MeO-Ph | 38 h | 47%/94% | 49%/89% | 12.5/1 | 61 |
| 19 | 19a: R = 3-Cl-4-MeO-Ph | 38 h | 46%/97% | 52%/90% | 11.1/1 | 80 |
| 20 | 20a: R = 3,4,5-triMeO-Ph | 48 h | 48%/91% | 48%/85% | >20/1 | 39 |
| 21 | 21a: R = 2-naphthyl | 42 h | 48%/91% | 51%/92% | >20/1d,g | 76 |
| 22 | 22a: R = 1-naphthyl | 38 h | 41%/98% | 56%/94% | 6.0/1e | 149 |
| 23 | 23a: R = 1-pyrenyl | 42 h | 44%/96% | 55%/89% | >20/1d | 67 |
| 24 | 24a: R = 3-benzothiophenyl | 48 h | 42%/84% | 57%/90% | 4.0/1c,f | 50 |
| 25 | 25a: R = 3-thienyl | 36 h | 45%/90% | 51%/90% | 17.1/1c | 58 |
| 26 | 26a: R = 2-thienyl | 48 h | 40%/98% | 54%/86% | 12.0/1 | 60 |
| 27 | 27a: R = 2-benzofuranyl | 36 h | 42%/96% | 53%/82% | 10.0/1 | 39 |
After expansion of the substrate scope of 2-substituted cyclobutanones, we turned our attention to more synthetically challenging 2,3-disubstituted cyclobutanones with an additional functional group at the ketone ring (Table 3). The reaction of 2,3-disubstituted cyclobutanones also proceeded well (28a–32a). With respect to 3-aryl-substituted substrates, excellent enantioselectivities (94–98% ee) and good rs (8.3/1–12.5/1 rs) were obtained with either EDG (OMe) or EWG (CF3) substituents at the para-position of aryl substrates. Notably, this reaction was performed on the gram scale, and products 28a and 28b were isolated without a decrease in the yield and selectivity (entry 2, Table 3). For the 3-methyl-2-aryl-substituted substrate, the reaction also proceeded well under the optimal reaction conditions; both an excellent ee of ketone 32a (91% ee) and excellent rs of 32b and 32c (>20/1) were observed, although with a moderate ee of γ-lactone 32b. Further improvement in the ee value of lactone ent-32b will be discussed in the asymmetric total synthesis of natural products.
| Entry | Substituents | Time | a: yield/ee | b + c: yield/b: ee | b/c rs | s-factor |
|---|---|---|---|---|---|---|
| a Reaction conditions: unless otherwise noted, the reactions were performed with a (0.2 mmol), Cu(NTf2)2 (10 mol%), L1 (12 mol%), 60 mg 4 Å MS. and m-CPBA (1.0 equiv.) in THF/CHBr3 (2.0/2.0 mL) for the time indicated at −40 °C. Isolated yield. ee was determined by UPC2 analysis. The rs of b/c was determined by crude 1H NMR. b 1.0 g 28a was used. c ent-L1 (12% mol) was applied. d Cu(OTf)2 (10 mol%) was applied to replace Cu(NTf2)2. e The absolute configuration of 28b was confirmed by X-ray analysis. | ||||||
| 1 | 28a: R1 = Ph, R2 = Ph | 36 h | 46%/96% | 52%/94% | 12.5/1e | 127 |
| 2 | 28a: R1 = Ph, R2 = Ph | 38 h | 48%/95% | 52%/94% | 12.5/1b | 121 |
| 3 | 29a: R1 = Ph, R2 = 4-F-Ph | 37 h | 49%/96% | 50%/94% | 8.3/1 | 127 |
| 4 | 30a: R1 = Ph, R2 = 4-CF3-Ph | 37 h | 48%/95% | 50%/93% | 9.1/1 | 103 |
| 5 | 31a: R1 = Ph, R2 = 4-MeO-Ph | 37 h | 48%/98% | 51%/96% | 12.5/1 | 226 |
| 6 | 32a: R1 = 3,4,5-triMeO-Ph, R2 = Me | 26 h | 40%/91% | 50%/86% | >20/1c | 42 |
| 7 | 32a: R1 = 3,4,5-triMeO-Ph, R2 = Me | 32 h | 48%/91% | 48%/92% | 19.0/1c,d | 121 |
After completing the study of generality of this classical kinetic resolution of 2-aryl-substituted and 2,3-disubstituted cyclobutanones, we focused our efforts on exploring the synthetic application of this unexplored methodology and targeted eupomatilones 5 and 6 isolated from the Australian shrub Eupomatia (Scheme 2).13,14 Starting from commercially available 3-methylcyclobutan-1-one (33), Pd-catalyzed α-arylation of cyclobutanone with 1-bromo-3,4,5-trimethoxybenzene (34) was carried out to prepare racemic precursor 32a for the B–V reaction. However, poor yield and/or diastereoisomeric ratio (dr) was observed under the tested reaction conditions.11 To our delight, after extensive investigation, the desired racemate 32a was obtained in 5.6
:
1 dr using a more sterically hindered ligand (X-Phos), and the dr of (±)-32a was further increased to 8.3
:
1 by treating with p-toluenesulfonic acid in refluxing chloroform. To improve the dissatisfactory results of the desired lactone 32b achieved under the optimal reaction conditions (entry 6, Table 3), slightly adjusted reaction conditions (Cu(OTf)2 and ent-L1 were applied)11 were used to furnish the expected lactone ent-32b in 48% yield with an excellent ee of 92% (entry 7, Table 3). With chiral lactone ent-32b in hand, mono-bromination of the aryl ring with N-bromosuccinimide (NBS) and the subsequent Pd-catalyzed Suzuki–Miyaura reaction with 1,3-benzodioxole-5-boronic acid (35) gave biaryl compound 3613b,15in 90% yield and 94% ee (two steps). To reverse the configuration of the methyl group at the β-position of γ-lactone, compound 36 was treated with phenylselenyl bromide (PhSeBr) and lithium bis(trimethylsilyl)amide (LiHMDS) at −78 °C and then oxidized with 30% H2O2 to yield an unsaturated lactone. The subsequent catalytic hydrogenation of the resulting lactone with Rh/Al2O3 in ethyl acetate (EA) at 40 °C afforded γ-butyrolactone 37 with high diastereoselectivity.16 Finally, the total synthesis of eupomatilones 5 and 6 was concisely completed in just a one-step transformation. Treatment of 37 with Eschenmoser's salt in THF at −78 °C and subsequent elimination produced eupomatilone-5 in 67% yield and 94% ee, while eupomatilone-6 was obtained through stereoselective methylation using LiHMDS and MeI at −78 °C (70% yield, 95% ee).13e The spectral data of two synthetic natural products were consistent with the reported literature.13b,d–g
Footnote |
| † Electronic supplementary information (ESI) available. CCDC 2164218–2164220, 2169290, and 2169293. For ESI and crystallographic data in CIF or other electronic format see https://doi.org/10.1039/d2sc02079c |
| This journal is © The Royal Society of Chemistry 2022 |