Switching Regioselectivity in the Asymmetric Bromocyclization of Difluoroalkenes Catalyzed by a Chiral Bisphosphine Oxide

Dedicated to Professor Hisashi Yamamoto on the occasion of his 80^th birthday

Abstract

We present an efficient approach for the enantioselective synthesis of difluoromethylene-containing oxazine compounds through 6-endo-selective bromocyclization of difluoroalkenes by using a chiral proton-bridged bisphosphine oxide complex as a catalyst precursor. The regioselectivity is significantly influenced by the solvent and the catalyst structure, and 6-endo cyclization products can be obtained preferentially with moderate to high enantioselectivity. This protocol offers complementary regioselectivity to our previously reported 5-exo-selective reaction, permitting the synthesis of diverse medicinally interesting compounds.

The incorporation of fluorine atom(s) into biologically active molecules has emerged as a key strategy for drug design and lead optimization.[1] In particular, there has been a surge of interest in the difluoromethylene (R¹–CF₂–R²) group owing to its ability to act as a bioisostere for carbonyl, sulfonyl, and ether groups.[2] Indeed, several difluoromethylene-containing drugs, such as lubiprostone,[3] gemcitabine,[4] and vinflunine,[5] have been developed to date (Figure [1]). It is noteworthy that most of these drugs possess a heterocyclic scaffold containing a difluoromethylene group at the stereogenic center as a common substructure. Nevertheless, the range of catalytic and asymmetric reactions available for the practical and versatile construction of such a privileged structural motif remains limited.[6]

Asymmetric electrophilic halocyclization of difluoroalkenes is one of the most straightforward methods to access heterocycles containing a difluoromethylene stereogenic center. In pioneering work, Toste and co-workers reported a regio- and enantioselective bromocyclization of difluoroalkenes by using a chiral anion phase-transfer catalyst, providing an efficient way to synthesize oxazolines or oxazines with a bromodifluoromethylated stereocenter, which could be readily transformed into various tetrasubstituted difluoromethylene-containing compounds.[7] Quite recently, we reported an operationally simpler and more stereoselective reaction that uses a chiral proton-bridged bisphosphine oxide complex (POHOP) 1 [8] as a catalyst precursor (Scheme [1]).[9] Our previously established concerted Lewis/Brønsted base catalysis with a chiral bisphosphine oxide is applicable to the bromination of even electron-deficient difluoroalkenes 2, and 5-exo cyclization occurs preferentially compared with 6-endo cyclization, delivering the corresponding oxazolines 3 with excellent enantioselectivities (up to 99% ee).

Importantly, this study revealed that the regioselectivity is greatly influenced by several factors, including the catalyst structure (e.g., the substituents on the phosphorus atom of 1), the brominating reagent, and the reaction temperature. In the course of the present work, we were surprised to observe a switchover of the regioselectivity when we changed the solvent (Scheme [2]). When toluene was used instead of dichloromethane, the 6-endo cyclization product 4a was obtained as the major product, albeit with moderate enantioselectivity. This marked solvent effect on the regioselectivity potentially offers a promising avenue for the selective synthesis of another class of medicinally interesting molecules. Here, we report the first general approach for the highly enantioselective and 6-endo-selective bromocyclization of difluoroalkenes 2, providing facile access to optically active oxazines with a tetrasubstituted difluoromethylated stereocenter.

Following the promising result shown in Scheme [2], we tested a series of POHOPs 1a–e under the same reaction conditions to evaluate their catalytic performance (Table [1], entries 1–5). In accord with our previous report,[9] the chemical yield and the regio- and enantioselectivity were influenced by the substituents on the phosphorus atom of 1 (entries 1–4). POHOPs 1a and 1d exhibited superior catalytic and regio- and enantiocontrolling abilities compared with the other catalysts. Although both precatalysts gave similar reaction outcomes, a slightly higher enantioselectivity toward 4a was observed with the use of 1d (entry 1 vs entry 4). Interestingly, the replacement of the binaphthyl backbone of 1d with a bi-1,3-benzodioxole resulted in a substantial improvement in the regioselectivity, affording 4a with the highest enantioselectivity among the tested POHOPs (entry 5). We next examined the effect of various brominating reagents in the presence of optimal precatalyst 1e (entries 5–10). The regioselectivity was found to depend strongly on the structure of the brominating reagent. In particular, the regioisomeric ratio (rr) reached a maximum (1:16 rr) when N-bromophthalimide (NBP) was used (entry 6). On the other hand, the brominating reagent had little impact on the enantioselectivity. Whereas N-bromosaccharin produced the desired product 4a with the highest enantioselectivity (83% ee), its regioselectivity was the lowest (entry 8). We therefore chose NBP as the optimal brominating reagent, based on the balance between regioselectivity and enantioselectivity. Aiming to improve the enantioselectivity for 4a, we conducted the reaction at a lower temperature (–50 °C; entry 11), but no improvement in enantioselectivity was observed and the conversion and regioselectivity decreased (entry 6 vs entry 11). We also tested the reaction in a toluene–dichloromethane cosolvent system at –40 °C (entry 12). Gratifyingly, the enantioselectivity of 4a was improved to 84% ee when a 9:1 ratio of toluene/dichloromethane was employed; however, both the conversion and regioselectivity decreased. Consequently, the reaction was conducted at –50 °C with a longer reaction time, leading to a slight improvement in both the conversion and enantioselectivity (entry 13). Next, the effect of the ratio of dichloromethane was examined (entries 13–15). Although increasing the ratio of dichloromethane improved the reaction efficiency and enantioselectivity, a substantial decrease in the regioselectivity was observed. Thus, we determined that a 9:1 ratio of toluene to dichloromethane was optimal for the 6-endo cyclization. To our delight, when the amount of NBP was slightly increased to 1.5 equivalents, 2a was fully consumed to afford the desired product 4a in 88% yield (entry 16). Finally, hoping to further improve the regio- and enantioselectivity, we tried the reaction at lower temperatures (–60 or –78 °C; entries 17 and 18). Unfortunately, however, no substantive improvement was observed in either case. The absolute stereochemistry of the major enantiomer 4a was assigned as an S-configuration by comparison of the retention time of HPLC with the data in the literature[7b] (see the Supporting Information).

Table 1 Optimization of the Reaction Conditions^a

Entry	1	[Br+] reagent	Solvent	Temp (°C)	Conversionb (%)	3a/4a b	Yieldc (%) and eed (%) of 3a	Yieldc (%) and eed (%) of 4a
1	1a	NBS	toluene	–40	>99	1:5.6	15, 64	85, 69
2	1b	NBS	toluene	–40	48	1:3.6	11, 60	33, 55
3	1c	NBS	toluene	–40	>99	1:2.4	29, 72	71, 50
4	1d	NBS	toluene	–40	>99	1:5.6	15, 61	85, 72
5	1e	NBS	toluene	–40	>99	1:11	8, 65	87, 77
6	1e	NBP	toluene	–40	>99	1:16	5, 64	83, 78
7	1e	NBA	toluene	–40	>99	1:8.6	10, 63	84, 78
8	1e	N-bromosaccharin	toluene	–40	>99	1:4.9	17, 75	83, 83
9	1e	DBI	toluene	–40	>99	1:10	9, 64	91, 77
10	1e	DBDMH	toluene	–40	>99	1:9.6	11, 66	86, 77
11	1e	NBP	toluene	–50	80	1:8.0	9, 62	72, 78
12	1e	NBP	toluene–CH2Cl2 (9:1)	–40	80	1:7.0	10, 73	70, 84
13e	1e	NBP	toluene–CH2Cl2 (9:1)	–50	90	1:7.2	11, 73	79, 86
14e	1e	NBP	toluene–CH2Cl2 (4:1)	–50	>99	1:5.4	15, 80	81, 88
15e	1e	NBP	toluene–CH2Cl2 (1:1)	–50	>99	1:2.8	26, 88	74, 92
16e	1e	NBPf	toluene–CH2Cl2 (9:1)	–50	>99	1:7.2	11, 75	88, 86
17e	1e	NBPf	toluene–CH2Cl2(9:1)	–60	>99	1:7.1	12, 75	84, 87
18e	1e	NBPf	toluene–CH2Cl2 (9:1)	–78	55	1:6.4	8, 71	48, 78

^a Reactions were conducted on a 0.05 mmol scale.

^b Determined by ¹H NMR analysis of the crude mixture.

^c Isolated yield.

^d Determined by chiral HPLC analysis.

^e Run for 48 h.

^f 1.5 equivalents.

With these optimized conditions in hand, we investigated the generality of the 6-endo-selective bromocyclization of difluoroalkenes 2 (Scheme [3]).[10] The desired 6-endo cyclization proceeded smoothly for amide substrates with a para-, meta-, and ortho-tolyl group, providing the corresponding products 4b–d with high regio- and enantioselectivity. Our catalytic system was compatible with both electron-donating and electron-withdrawing substituents on the phenyl group. Compounds with 4-methoxy, 4-fluoro, or 4-chloro substituents were efficiently converted into the cyclized products 4e–g with good enantioselectivity (75–82% ee). However, when a much more electron-deficient amide bearing a nitro group was employed, the reaction hardly proceeded under the optimized conditions (data not shown). The 2-furyl amide gave the 6-endo cyclization product 4h preferentially, but the enantioselectivity was modest. We also explored the scope of the substituent (R²) on the alkene group. The meta-tolyl-substituted substrate 2i gave the corresponding product 4i in 91% yield with 80% ee. Although the phenyl-substituted alkene 2j showed a relatively low reactivity and was not fully consumed, the desired product 4j was obtained in satisfactory yield with a good enantioselectivity. Bulky biphenyl- and 2-naphthyl-substituted substrates 2k and 2l were well tolerated, giving the corresponding products 4k and 4l with high regio- and enantioselectivity (85% and 86% ee, respectively). We were pleased to find that halogenated and oxygenated styrenes 2m–p were also suitable substrates. Styrenes with a halogen group (F or Cl) at the para-position reacted smoothly to afford the 6-endo cyclization products 4m and 4n, respectively with excellent regioselectivity and enantioselectivity (91% and 93% ee, respectively). A styrene bearing a triflate group (2o), which is useful for further chemical transformations, also underwent the desired reaction to provide the 6-endo cyclization adduct 4o with 84% ee, although an increased amount of NBP (2.0 equivalents) and a longer reaction time (72 h) were required to complete the reaction. In addition to sulfonate ester, the styrene with an acetyl group (2p) was also a good substrate, and the 6-endo-selective cyclization proceeded without difficulty to yield the oxazine compound 4p in 92% yield with 93% ee. In our protocol, the bisphosphine oxide catalyst reacts with the brominating agent to generate a chiral P=O⁺–Br species in situ.[8] Owing to its high electrophilicity, not only a nonactivated styrene (2j), but also electron-deficient styrenes (2m–p), were able to undergo the desired cyclization, even at a low temperature, highlighting the synthetic utility of the present catalysis.

In summary, we have achieved a highly 6-endo-selective and enantioselective bromocyclization of difluoroalkenes. Although the reason for the drastic switch in regioselectivity to 6-endocyclization simply upon changing the solvent remains unclear, our method produces structurally unique oxazines with a tetrasubstituted difluoromethylated stereocenter in high yields and with high enantioselectivities, and is expected to contribute to the creation of novel bioactive compounds. Investigation of the mechanistic details of this reaction is underway in our laboratory.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

R.H. is grateful for a Nagai Memorial Research Scholarship from the Pharmaceutical Society of Japan.

• References and Notes

• 1a Purser S, Moore PR, Swallow S, Gouverneur V. Chem. Soc. Rev. 2008; 37: 320
  CrossrefPubMedSearch in Google Scholar
• 1b Wang J, Sánchez-Roselló M, Aceña JL, del Pozo C, Sorochinsky AE, Fustero S, Soloshonok VA, Liu H. Chem. Rev. 2014; 114: 2432
  CrossrefPubMedSearch in Google Scholar
• 1c Gillis EP, Eastman KJ, Hill MD, Donnelly DJ, Meanwell NA. J. Med. Chem. 2015; 58: 8315
  CrossrefPubMedSearch in Google Scholar
• 1d Zhou Y, Wang J, Gu Z, Wang S, Zhu W, Aceña JL, Soloshonok VA, Izawa K, Liu H. Chem. Rev. 2016; 116: 422
  CrossrefPubMedSearch in Google Scholar
• 1e Inoue M, Sumii Y, Shibata N. ACS Omega 2020; 5: 10633
  CrossrefPubMedSearch in Google Scholar
• 2a Lapierre J, Ahmed V, Chen M.-J, Ispahany M, Guillemette JG, Taylor SD. Bioorg. Med. Chem. Lett. 2004; 14: 151
  CrossrefPubMedSearch in Google Scholar
• 2b Meanwell NA. J. Med. Chem. 2011; 54: 2529
  CrossrefPubMedSearch in Google Scholar
• 2c Zhou Q, Ruffoni A, Gianatassio R, Fujiwara Y, Sella E, Shabat D, Baran PS. Angew. Chem. Int. Ed. 2013; 52: 3949
  CrossrefPubMedSearch in Google Scholar
• 2d Ivanova MV, Bayle A, Besset T, Pannecoucke X, Poisson T. Chem. Eur. J. 2016; 22: 10284
  CrossrefPubMedSearch in Google Scholar
• 2e Meanwell NA. J. Med. Chem. 2018; 61: 5822
  CrossrefPubMedSearch in Google Scholar
• 3 Ambizas EM, Ginzburg R. Ann. Pharmacother. 2007; 41: 957
  CrossrefPubMedSearch in Google Scholar
• 4 Toschi L, Finocchiaro G, Bartolini S, Gioia V, Cappuzzo F. Future Oncol. 2005; 1: 7
  CrossrefPubMedSearch in Google Scholar
• 5 Frampton JE, Moen MD. Drugs 2010; 70: 1283
  CrossrefPubMedSearch in Google Scholar
• 6a Liu Y.-L, Yu J.-S, Zhou J. Asian J. Org. Chem. 2013; 2: 194
  CrossrefSearch in Google Scholar
• 6b Yu J.-S, Liao F.-M, Gao W.-M, Liao K, Zuo R.-L, Zhou J. Angew. Chem. Int. Ed. 2015; 54: 7381
  CrossrefPubMedSearch in Google Scholar
• 6c Rong M.-Y, Li J.-S, Zhou Y, Zhang F.-G, Ma J.-A. Org. Lett. 2020; 22: 9010
  CrossrefPubMedSearch in Google Scholar
• 6d Liu J, Yu L, Zheng C, Zhao G. Angew. Chem. Int. Ed. 2021; 60: 23641
  CrossrefPubMedSearch in Google Scholar
• 6e Uno H, Kawai K, Araki T, Shiro M, Shibata N. Angew. Chem. Int. Ed. 2022; 61: e202117635
  CrossrefPubMedSearch in Google Scholar
• 7a Miller E, Kim S, Gibson K, Derrick JS, Toste FD. J. Am. Chem. Soc. 2020; 142: 8946
  CrossrefPubMedSearch in Google Scholar
• 7b Miller E, Mai BK, Read JA, Bell WC, Derrick JS, Liu P, Toste FD. ACS Catal. 2022; 12: 12369
  CrossrefPubMedSearch in Google Scholar
• 8 Yamashita K, Hirokawa R, Ichikawa M, Hisanaga T, Nagao Y, Takita R, Watanabe K, Kawato Y, Hamashima Y. J. Am. Chem. Soc. 2022; 144: 3913
  CrossrefPubMedSearch in Google Scholar
• 9 Hirokawa R, Nakahara Y, Uchida S, Yamashita K, Hamashima Y. Chem. Asian J. 2023; e202300141
  CrossrefPubMed
• 10 (5S)-5-Bromo-6,6-difluoro-2-phenyl-5-(p-tolyl)-5,6-dihydro-4H-1,3-oxazine (4a):Typical ProcedurePrecatalyst 1e (12.9 mg, 0.010 mmol, 10 mol%) and difluoroalkene 2a (28.7 mg, 0.10 mmol, 1.0 equiv) were placed in a well-dried Schlenk test tube under an argon atmosphere and dissolved in anhyd toluene (1.8 mL) and CH₂Cl₂ (0.2 mL). The mixture was cooled to –50 °C, and NBP (33.9 mg, 0.15 mmol, 1.5 equiv) was added. The mixture was stirred for 48 h at –50 °C, and then the reaction was quenched by adding a 1.5 M solution of butyl vinyl ether in MeOH (1.0 mL, precooled to –78 °C) through a cannula. The mixture was then gradually warmed to rt and diluted with brine (5.0 mL). The aqueous phase was extracted with CH₂Cl₂ (3 × 10 mL). The combined extracts were dried (MgSO₄) and concentrated under reduced pressure, and the residue was subjected to ¹H NMR analysis to determine the regioisomeric ratio. The crude product was purified by column chromatography [silica gel, hexane–EtOAc (100:0 to 9:1)] to give a white solid; yield: 32.4 mg (88%, 3a/4a = 1:7.2, 86% ee); mp 103–104 °C; [α]_D ²⁰ = –0.8 [c 1.00, CHCl₃, 86% ee, (S)].HPLC (Method 1) [IJ-3, hexane–i-PrOH (99:1), flow rate 1.0 mL/min, column temp. 40 °C, λ = 254 nm]: t _R = 13.2 min (minor, R), 14.5 min (major, S); (Method 2) [OD-H, hexane/i-PrOH (99.5:0.5), flow rate 1.0 mL/min, column temp. 40 °C, λ = 254 nm]: t _R = 6.5 min (major, S), 6.9 min (minor, R). IR (neat): 2919, 1674, 1448, 1353, 1234, 1149, 1084, 1054 cm^–1. ¹H NMR (500 MHz, CDCl₃): δ = 7.99–7.98 (m, 2 H), 7.58 (d, J = 8.0 Hz, 2 H), 7.51 (t, J = 7.5 Hz, 1 H), 7.44–7.41 (m, 2 H), 7.20 (d, J = 8.0 Hz, 2 H), 4.61 (dt, J = 17.8, 3.3 Hz, 1 H), 4.52 (dt, J = 17.8, 3.8 Hz, 1 H), 2.36 (s, 3 H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ = 150.8, 139.7, 132.2, 131.9, 130.1, 129.3 (2 C), 128.4 (2 C), 127.8 (2 C), 127.6 (2 C), 120.9 (t, J = 261.7 Hz), 56.7 (t, J = 26.9 Hz), 56.1, 21.1. ¹⁹F NMR (470 MHz, CDCl₃): δ = –74.9 (br, 2 F).The spectral data for 4a were consistent with those reported in ref. 7a

Corresponding Author

Publication History

Received: 21 April 2023

Accepted after revision: 26 June 2023

Accepted Manuscript online:
26 June 2023

Article published online:
09 August 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany

• References and Notes

• 1a Purser S, Moore PR, Swallow S, Gouverneur V. Chem. Soc. Rev. 2008; 37: 320
  CrossrefPubMedSearch in Google Scholar
• 1b Wang J, Sánchez-Roselló M, Aceña JL, del Pozo C, Sorochinsky AE, Fustero S, Soloshonok VA, Liu H. Chem. Rev. 2014; 114: 2432
  CrossrefPubMedSearch in Google Scholar
• 1c Gillis EP, Eastman KJ, Hill MD, Donnelly DJ, Meanwell NA. J. Med. Chem. 2015; 58: 8315
  CrossrefPubMedSearch in Google Scholar
• 1d Zhou Y, Wang J, Gu Z, Wang S, Zhu W, Aceña JL, Soloshonok VA, Izawa K, Liu H. Chem. Rev. 2016; 116: 422
  CrossrefPubMedSearch in Google Scholar
• 1e Inoue M, Sumii Y, Shibata N. ACS Omega 2020; 5: 10633
  CrossrefPubMedSearch in Google Scholar
• 2a Lapierre J, Ahmed V, Chen M.-J, Ispahany M, Guillemette JG, Taylor SD. Bioorg. Med. Chem. Lett. 2004; 14: 151
  CrossrefPubMedSearch in Google Scholar
• 2b Meanwell NA. J. Med. Chem. 2011; 54: 2529
  CrossrefPubMedSearch in Google Scholar
• 2c Zhou Q, Ruffoni A, Gianatassio R, Fujiwara Y, Sella E, Shabat D, Baran PS. Angew. Chem. Int. Ed. 2013; 52: 3949
  CrossrefPubMedSearch in Google Scholar
• 2d Ivanova MV, Bayle A, Besset T, Pannecoucke X, Poisson T. Chem. Eur. J. 2016; 22: 10284
  CrossrefPubMedSearch in Google Scholar
• 2e Meanwell NA. J. Med. Chem. 2018; 61: 5822
  CrossrefPubMedSearch in Google Scholar
• 3 Ambizas EM, Ginzburg R. Ann. Pharmacother. 2007; 41: 957
  CrossrefPubMedSearch in Google Scholar
• 4 Toschi L, Finocchiaro G, Bartolini S, Gioia V, Cappuzzo F. Future Oncol. 2005; 1: 7
  CrossrefPubMedSearch in Google Scholar
• 5 Frampton JE, Moen MD. Drugs 2010; 70: 1283
  CrossrefPubMedSearch in Google Scholar
• 6a Liu Y.-L, Yu J.-S, Zhou J. Asian J. Org. Chem. 2013; 2: 194
  CrossrefSearch in Google Scholar
• 6b Yu J.-S, Liao F.-M, Gao W.-M, Liao K, Zuo R.-L, Zhou J. Angew. Chem. Int. Ed. 2015; 54: 7381
  CrossrefPubMedSearch in Google Scholar
• 6c Rong M.-Y, Li J.-S, Zhou Y, Zhang F.-G, Ma J.-A. Org. Lett. 2020; 22: 9010
  CrossrefPubMedSearch in Google Scholar
• 6d Liu J, Yu L, Zheng C, Zhao G. Angew. Chem. Int. Ed. 2021; 60: 23641
  CrossrefPubMedSearch in Google Scholar
• 6e Uno H, Kawai K, Araki T, Shiro M, Shibata N. Angew. Chem. Int. Ed. 2022; 61: e202117635
  CrossrefPubMedSearch in Google Scholar
• 7a Miller E, Kim S, Gibson K, Derrick JS, Toste FD. J. Am. Chem. Soc. 2020; 142: 8946
  CrossrefPubMedSearch in Google Scholar
• 7b Miller E, Mai BK, Read JA, Bell WC, Derrick JS, Liu P, Toste FD. ACS Catal. 2022; 12: 12369
  CrossrefPubMedSearch in Google Scholar
• 8 Yamashita K, Hirokawa R, Ichikawa M, Hisanaga T, Nagao Y, Takita R, Watanabe K, Kawato Y, Hamashima Y. J. Am. Chem. Soc. 2022; 144: 3913
  CrossrefPubMedSearch in Google Scholar
• 9 Hirokawa R, Nakahara Y, Uchida S, Yamashita K, Hamashima Y. Chem. Asian J. 2023; e202300141
  CrossrefPubMed
• 10 (5S)-5-Bromo-6,6-difluoro-2-phenyl-5-(p-tolyl)-5,6-dihydro-4H-1,3-oxazine (4a):Typical ProcedurePrecatalyst 1e (12.9 mg, 0.010 mmol, 10 mol%) and difluoroalkene 2a (28.7 mg, 0.10 mmol, 1.0 equiv) were placed in a well-dried Schlenk test tube under an argon atmosphere and dissolved in anhyd toluene (1.8 mL) and CH₂Cl₂ (0.2 mL). The mixture was cooled to –50 °C, and NBP (33.9 mg, 0.15 mmol, 1.5 equiv) was added. The mixture was stirred for 48 h at –50 °C, and then the reaction was quenched by adding a 1.5 M solution of butyl vinyl ether in MeOH (1.0 mL, precooled to –78 °C) through a cannula. The mixture was then gradually warmed to rt and diluted with brine (5.0 mL). The aqueous phase was extracted with CH₂Cl₂ (3 × 10 mL). The combined extracts were dried (MgSO₄) and concentrated under reduced pressure, and the residue was subjected to ¹H NMR analysis to determine the regioisomeric ratio. The crude product was purified by column chromatography [silica gel, hexane–EtOAc (100:0 to 9:1)] to give a white solid; yield: 32.4 mg (88%, 3a/4a = 1:7.2, 86% ee); mp 103–104 °C; [α]_D ²⁰ = –0.8 [c 1.00, CHCl₃, 86% ee, (S)].HPLC (Method 1) [IJ-3, hexane–i-PrOH (99:1), flow rate 1.0 mL/min, column temp. 40 °C, λ = 254 nm]: t _R = 13.2 min (minor, R), 14.5 min (major, S); (Method 2) [OD-H, hexane/i-PrOH (99.5:0.5), flow rate 1.0 mL/min, column temp. 40 °C, λ = 254 nm]: t _R = 6.5 min (major, S), 6.9 min (minor, R). IR (neat): 2919, 1674, 1448, 1353, 1234, 1149, 1084, 1054 cm^–1. ¹H NMR (500 MHz, CDCl₃): δ = 7.99–7.98 (m, 2 H), 7.58 (d, J = 8.0 Hz, 2 H), 7.51 (t, J = 7.5 Hz, 1 H), 7.44–7.41 (m, 2 H), 7.20 (d, J = 8.0 Hz, 2 H), 4.61 (dt, J = 17.8, 3.3 Hz, 1 H), 4.52 (dt, J = 17.8, 3.8 Hz, 1 H), 2.36 (s, 3 H). ¹³C{¹H} NMR (125 MHz, CDCl₃): δ = 150.8, 139.7, 132.2, 131.9, 130.1, 129.3 (2 C), 128.4 (2 C), 127.8 (2 C), 127.6 (2 C), 120.9 (t, J = 261.7 Hz), 56.7 (t, J = 26.9 Hz), 56.1, 21.1. ¹⁹F NMR (470 MHz, CDCl₃): δ = –74.9 (br, 2 F).The spectral data for 4a were consistent with those reported in ref. 7a

• Supplementary Material