Catalytic Asymmetric Nucleophilic Fluorination Using BF3·Et2o As Fluorine Source and Activating Reagent

ARTICLE

https://doi.org/10.1038/s41467-021-24278-3 OPEN Catalytic asymmetric nucleophilic ﬂuorination using BF3·Et2Oasﬂuorine source and activating reagent

Weiwei Zhu1, Xiang Zhen2, Jingyuan Wu1, Yaping Cheng1, Junkai An2, Xingyu Ma1, Jikun Liu2, Yuji Qin1, ✉ Hao Zhu2, Jijun Xue2 & Xianxing Jiang 1

1234567890():,; Fluorination using chiral catalytic methods could result in a direct access to asymmetric ﬂuorine chemistry. However, challenges in catalytic asymmetric ﬂuorinations, especially the

longstanding stereochemical challenges existed in BF3·Et2O-based ﬂuorinations, have not yet been addressed. Here we report the catalytic asymmetric nucleophilic ﬂuorination using

BF3·Et2O as the ﬂuorine reagent in the presence of chiral iodine catalyst. Various chiral ﬂuorinated oxazine products were obtained with good to excellent enantioselectivities (up to >99% ee) and diastereoselectivities (up to >20:1 dr). Control experiments (the desired

ﬂuoro-oxazines could not be obtained when Py·HF or Et3N·3HF were employed as the ﬂuorine

source) indicated that BF3·Et2O acted not only as a ﬂuorine reagent but also as the activating reagent for activation of iodosylbenzene.

1 Guangdong Provincial Key Laboratory of Chiral Molecule and Drug Discovery, School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou, China. 2 State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou, ✉ Gansu, China. email: [email protected]

NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3

eing called as “a small atom with a big ego”, fluorine acts as utilization in the area: the high toxicity and biohazard for HF- Ba significant and increasingly important role in the fields of bases, and metal fluorides poor solubility in organic solvents organic chemistry, pharmaceuticals, agrochemicals and coupled with limited strategies to control reactivity. material chemistry1–4. The fluorinated molecules often display Ideally, one low-toxic, stable and commercially cheap available higher thermal and metabolic stabilities, lower polarity, and nucleophilic fluorine reagent would drastically promote enan- weaker intermolecular interactions due to the strong C−F bond tioselective fluorine synthetic innovation and industrial develop- 5 and unique properties of F atom . Therefore, these unique ment. As a versatile Lewis acid, commercially available BF3·Et2O properties of fluorine-contained compounds make the develop- is easy to prepare and is widely being used in various organic ment of efficient strategies, especially of catalytic asymmetric transformations. As early as 1960, it was discovered that as a reaction for fluorination of molecules as one of the hottest areas nucleophilic fluorine reagent could be applied in the fluorination 2,3 33 in organic synthesis . Nevertheless, the asymmetric fluorine of ring opening of mesoepoxides . The development of BF3·Et2O organic chemistry still represents a considerable challenge to mediated reactions in half a century reveals that the BF3·Et2O can date6. In the wake of the emergence of the first electrophilic also be applied to fluorinations of alkenes34–36, alkynes37, enantioselective fluorination of enolates using chiral N-fluoro arenes38,39 and other fluorine organic chemistry40–43. Although camphorsultam reagent reported by the group of Lang7, sig- these efficient achiral methodologies have been well-established, nificant progress for enantioselective fluorination studies to date, the longstanding stereochemical challenges of the 4,8,9 have been presented because of contributing to the BF3·Et2O-based fluorination have not yet been addressed, prob- development of catalytic asymmetric methodologies for electro- ably impeded by several hurdles: intense competition for the role fl + fl fl philic uorine reagents (F reagents), such as N- uor- of BF3 between a nucleophilic uorine reagent and a Lewis acid, obenzenesulfonimide (NFSI)10–12, N-fluoropyridinium salts3, and the difficulty in achieving stereocontrol of fluorine atom, the Selectfluor (Fig. 1a)13–18. These reagents exhibited efficient competition from the uncatalyzed background reaction and other transfer of fluorine atom under the asymmetric fluorination, side reactions. Undoubtedly, the advent of enantioselective however, their industrial applications were significantly restricted approach is long overdue that would be welcome. by the poor atom economy in fluorination, expensive synthesis In terms of the operational and environmental advantages and other inherent characteristics of electrophilic reagent. Alter- associated with organocatalysis, we speculated that a metal-free natively, nucleophilic fluorine reagents (F− reagents) have been mild reaction system with a chiral iodine catalyst (CIC) might attracting considerable interest recently since the relative stability meet the aforementioned challenges21,44,45. Its activated oxidants and low-cost. Considerable advances have recently been achieved forms could form the iodine (III) catalyst with a typical structure in this field involving catalytic asymmetric fluorination of keto type of trigonal bipyramidal geometry, thus this type of robust esters19,20 and alkenes21–25 employing pyridine·HF as a fluorine organocatalyst has been commonly used for asymmetric reagent, catalytic asymmetric fluorination of allylic tri- nucleophilic addition reactions44,45. The unique stereoscopic fi fi chloroacetimidates using a combination of Et3N·HF with Iridium con guration of iodine (III) and well-de ned steric hindrance of complex26, asymmetric ring-opening fluorinations of meso- the iodine (III) catalyst bearing chiral ligand can be readily epoxides (aziranes) using PhCOF, HF-reagents or AgF as the applied, leading to complete stereo-control in fluorination of fl 27–29 fi uorine source in metal-catalyzed system , and other asym- ole ns using BF3·Et2O as a nucleophilic reagent. Rapid cyclization fl fl metric transformations in the presence of metal uorides (KF, and its simultaneous BF3·Et2O nucleophilic uorination are viable CsF or AgF)30–32 (Fig. 1b). Despite these elegant works, several with a CIC to suppress the pure intramolecular cycloaddition and practical disadvantages still discouraged their further large-scale other side reactions.

O Y O X 1 1 XH F a + F b R R O [M] R1 [F ] 1 G 1 O [M] [F ] H 3 R 3 R 3 OO 1 2 R R + G NXG R O O R1 R2 X [M] R R 2 Cat. 2 *2 S S F 2 2 R R F R Ph N Ph R R F F •(HF)n 1 1 N 1 O O R 1 R R1 1 R + - R Y R [F+] Cl R * F F F Y Y O O N N HBPTC E-LG X + P X N N N N + F 2 3 2 3 N X X F Reagents R R R R R PTC R R O O R2 R3 M F H H H H F * M E R R F R 1 KF F R O R3 Cl CsF X + O O N 1 F F [M], L L O A [F ] - (HF) R Ar F 2 O 2 N 2BF4 • n F H F 2 F H F FG R A M R 1 A 3 ArI, [O] I 2 1 R 1 * R * R F N 1 R R 3 R * 2 3 L 2 3 2 R R R R O F R R R [F ] R3 R1 3 I R Ar

c F F O 2 B - Ar F Ar F O F Me Me +I OBF F F 1 N 2 2 Me F Ar H Ar BF3·Et2O N I 3 BF3·Et2O Ar 3 Ar1 Ar Ar3 Ar3 Ar O CIC1 (15% mol) up to 99% ee NH CIC (15% mol) m NH NH NH N ArI -CPBA and > 20:1 dr m-CPBA • Et2O solvent O 4 m-CPBA 4 solvent 4 O O O Ar Ar Ar Ar4 Ar4 -H+ F F up to 85% ee B Ar O F Ar2 I O F O O O O - -BF2OBF3 -ArI R O O R R =benzyl,CIC1 N N 2 NH Ar = O O 1 F H 2 1 H Ar Ar F Ar Ar I Bn Bn R = 1,2,3,4-tetrahydronaphthalen-1-yl , CIC2 B Ar F F

Fig. 1 Examples of common fluorine reagents applied in asymmetric fluorinations. a Asymmetric fluorinations using electrophilic fluorine (F+) reagents. − − b Asymmetric fluorinations using nucleophilic fluorine (F ) reagents. c The first example of using BF3·Et2OasF Reagent for asymmetric fluorinations (this work). [F+], electrophilic fluorine reagents; [F−], nucleophilic fluorine reagents; PTC, phase-transfer catalyst; [M], metal catalyst; L, ligand; HBPTC, hydrogen bonding phase-transfer catalysis; [O], oxidant; m-CPBA, Meta-Chloroperoxybenzoic acid.

2 NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3 ARTICLE

“ ” 2 – O CIC (15 mol%) F complex substituents on Ar ring (Fig. 3b, 35b 42b), again BF ·Et O (10 equiv) O N 3 2 leading to good yields (45–72%) in high to excellent diastereos- H m-CPBA (1.2 equiv) N electivities (10:1–>20:1dr) and enantioselectivities (80-99% ee). It 1a o 1b DCE, -25 C, 48 h is worth noting that the catalytic asymmetric aminofluorination CIC1, R1 =R2 =benzyl, of complex natural product structures could also be achieved 70% yield, 86% ee, and > 20:1 dr fi 1 2 ef ciently in high to excellent diastereoselectivities and O I O CIC2,R =benzyl,R = 1,2,3,4-tetrahydronaphthalen-1-yl, R2 O O R2 70% yield, 74% ee, and > 20:1 dr enantioselectivities. O O CIC3,R1 =methyl,R2 =ethyl, 1 1 Additionally, the gram-scale experiment was conducted to R R 68% yield, 73% ee, and > 20:1 dr fl CIC4,R1 =methyl,R2 = t-butyl, evaluate the applicability of our asymmetric uorination method 65% yield, 74% ee, and > 20:1 dr by using 4a (Fig. 3c), the desired product was obtained with CIC5,R1 =methyl,R2 =benzyl, excellent diastereoselectivity (>20:1 dr) and enantioselectivity 70% yield, 76% ee, and > 20:1 dr CIC6,R1 =methyl,R2 =1-iPr-4-Methylcyclohexane-2-yl , (92% ee). The result suggested that our protocol was promising in 3 I I 3 R R 66% yield, 74% ee, and > 20:1 dr future industrial applications. It was interesting that compared to CIC9, R3 =H, CIC7,R1 =methyl,R2 = adamantyl, the PTC catalyzed process46, different fluorinated products could 68% yield, 60% ee, and 9:1 dr 58% yield, 72% ee, and 15:1 dr CIC10,R3 =ethyl, CIC8,R1 = i-propyl, R2 =benzyl, be obtained from the same substrates using current process 66% yield, 62% ee, and 10:1 dr 60% yield, 35% ee, and 10:1 dr (indicating a different catalytic progress). The relative and absolute configurations of the products were determined by X- Fig. 2 Optimizing CIC structure in the model reaction. Notations for CICs: ray crystal structure analysis of 4b (see the Supplementary the reaction of 1a (0.1 mmol), BF ·Et O (1.0 mmol), m-CPBA (0.12 mmol) 3 2 Information). and catalyst (15% mol) was carried out in 4.0 ml of 1,2-dichloroethane (DCE) at −25 °C for 48 h; isolated yields are reported. The ee values were determined by chiral HPLC analysis and the dr values were determined by Asymmetric aminofluorination of N-(2-(prop-1-en-2-yl)phe- 19F NMR. nyl)benzamides. To further understanding of the scope of this catalytic system, substrates 43a–54a were employed to undergo Herein, we introduce a catalytic asymmetric fluorination the fluorination process. To our delight, various substituted N-(2- fl strategy that involves BF3·Et2O as nucleophilic uorine source (prop-1-en-2-yl)phenyl)benzamides including either electron- with chiral iodine catalysts. Various F-contained products were donating substituents or steric hindrance substituents at differ- obtained with good to excellent diastereoselectivities (>20:1 dr) ent positions on the Ar4 ring, as well as 3,5-ditrifluoromethyl and enantioselectivities (up to >99% ee) (Fig. 1c). substituents could be tolerated, affording the corresponding fluorinated products (43b–54b) with high enantioselectivities Results and discussion (80–85% ee) and isolated yields (80–88% yield) (Fig. 4). These Catalyst optimization. After several initial trials (Supplementary substrate scope expanding experiments showed the hypothesis to Table 1), we set out to optimize the model catalytic asymmetric introduce the combination of BF3·Et2O and CIC into direct, fl aminofluorination of N-cinnamylbenzamide (1a) in the presence catalytic asymmetric uorinations can be achieved. This method fl of 15 mol% of ligand loading using BF ·Et O as the fluorine expanded the structures of the uorinated products and provided 3 2 fl reagent in DCE at −25 °C with m-CPBA as an oxidate (Fig. 2). a benign access to asymmetric nucleophilic uorinations. The significant difference in stereoselectivity was observed under these reaction conditions, whereas good yields were afforded with Mechanism studies. On the basis of the experimental results the addition of structure diverse chiral iodine catalyst CICs. These described above, we have proposed a possible mechanism to results suggested that the substituents of the catalysts have a explain the stereochemistry of the catalytic asymmetric nucleophilic strong influence on the stereochemistry of the reaction. In gen- fluorinations (Fig. 5). To gain a better understanding of the process eral, compared to spiro-CICs, the linear CICs could give the of this catalytic fluorination system, we also conducted control higher stereoselectivity. The CIC1 was proved to be the best experiments (Fig. 5a) and density functional theory (DFT) calcu- fl catalyst, providing the desired chiral uorinated oxazine 1b with lations (Fig. 6). It is worth noting that we could not obtain the fl excellent diastereoselectivity (>20:1 dr) and enantioselectivity desired uorinated products when we used or Py·HF or Et3N·3HF 47 (86% ee) in 70% yield. instead of BF3·Et2O (Fig. 5a, equation 1). When PhIF2 was applied as the hypervalent iodine reagent and fluorine source, we Asymmetric aminofluorination of N-cinnamylbenzamides didn’t obtain the 1b as well (Fig. 5a, equation 1). Thus, these results fl using BF3·Et2O. Results of experiments under the optimized indicate that BF3·Et2O acted not only as the nucleophilic uorine conditions that probe the scope of the reaction are summarized in source, but also as the activating reagents for activation of iodo- Fig. 3. Substrate scope of the reaction was investigated with a sylbenzene (Int1). which is distinct from previously catalytic variety of substituted N-cinnamylbenzamides under the optimal nucleophilic fluorination process reported by Jacobsen’s group21. reaction conditions. As shown in Fig. 3a, variation of the elec- Based on previous work29,35,48,49, control experiments and DFT tronic properties of substituents at either Ar1 or Ar2 of N-cin- calculations, the plausible catalytic cycle was shown in Fig. 5b, and namylbenzamides with different steric parameters were tolerated, associated free energy profile was shown in Fig. 6. affording the desired products with good to excellent diaster- At first, the aryl iodine catalyst is oxidized by m-CPBA to form eoselectivities (2:1–>20:1 dr) and enantioselectivities (80–>99% Ar−I=O(Int1), and Int1 is found to be 2.8 kcal/mol lower in ee) in good yields (45–75%). Gratifyingly, the fluorinated oxazine energy than ArI. (Fig. 6). Then Int2 is formed through the 2 − = 44 products bearing Ar with high steric hindrance were still activation of Ar I ObyBF3·Et2O , the energy of this obtained in good yields and excellent stereoselectivities (85–>99% intermediate is calculated to be 23.8 kcal/mol lower than Int1 ee, 15b–24b). The naphthyl-substituted N-cinnamylbenzamides (Fig. 6). The electrophilic addition of iodine (III) to the double could also be tolerated, and gave the corresponding fluorinated bond of la forms Int334, during which the energy barrier is found products (26b, 27b and 31b) in excellent diastereoselectivities (up to be 3.6 kcal/mol. Then with the assistance of BF3, Int3 releases – - + − to >20:1dr) and enantioselectivities (86 87% ee) with good yields. anionic [BF4] to form Int3 . The nucleophilic attack of [BF4] Furthermore, as expected, the catalytic system also proved to be (generated in previous step) on Int3+ at C1 position afford the efficient for the N-cinnamylbenzamides with heterocycles or TS143, the energy barrier of this step is 13.1 kcal/mol. There were

NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3

BF ·Et O (10.0 equiv) O 3 2 Ar2 O I O CIC1 (15 mol%) F O O O O O 1 N 2 Ar H Ar m-CPBA (1.2 equiv) N Bn Bn 1 DCE, -25oC, 48 h Ar 1a-42a 1b-42b CIC1

Ar1 and Ar2 = Substituted Phenyl Groups a OMe F O F O F O F O N N N N

1b 2b 3b 4b 70% yield (> 20:1 dr) 71% yield (> 20:1 dr) 72% yield (> 20:1 dr) 75% yield (> 20:1 dr) 86% ee 99% ee 91% ee > 99% ee OEt F Cl Br F O F O F O F O F O F O Br N N N N N Br N

5b 6b 7b 8b 9b 10b 67% yield (> 20:1 dr) 71% yield (11:1 dr) 74% yield (> 20:1 dr) 70% yield (> 20:1 dr) 55% yield (11:1 dr) 67% yield (10:1 dr) 94% ee 92% ee 91% ee 91% ee 99% ee 88% ee O CF3 CN NO2 CF3 F O F O F O F O F O F O CF3 N N N N N N

11b 12b 13b 14b 15b 16b 65% yield (11:1 dr) 72% yield (6:1 dr) 60% yield (4:1 dr) 65% yield (> 20:1 dr) 70% yield (2:1 dr) 75% yield (> 20:1 dr) 91% ee 86% ee 91% ee (major) 91% ee 99% ee (major) 98% ee t-Bu OMe t-Bu OMe F O F O F O F O F O F O t-Bu OMe OMe N N N N N N

17b 18b 19b 20b 21b 22b 60% yield (5:1 dr) 70% yield (> 20:1 dr) 67% yield (10:1 dr) 65% yield (14:1 dr) 68% yield (> 20:1 dr) 64% yield (> 20:1 dr) 85% ee 99% ee 88% ee > 99% ee 98% ee 97% ee O O OCF3 Ph F O F O O O F O F O F O F O N N N N N N

23b 24b 25b 26b 27b 28b 68% yield (> 20:1 dr) 61% yield (15:1 dr) 65% yield (13:1 dr) 77% yield (> 20:1 dr) 74% yield (> 20:1 dr) 69% yield (11:1 dr) 93% ee 94% ee 91% ee 87% ee 86% ee 81% ee OMe OMe OMe F Cl Br F O F O F O F O F O F O N N N N N N

Me 29b F3C 30b* 31b* Me 32b Me 33b Me 34b 71% yield (2:1 dr) 75% yield (> 20:1 dr) 56% yield (4:1 dr) 68% yield (> 20:1 dr) 66% yield (> 20:1 dr) 60% yield (10:1 dr) 94% ee 80% ee 90% ee 81% ee

Ar2 = Heterocyclic Groups or "Complex Groups" b O O S S O O F O F O F O 4 F O F O N N N N N 35b 36b 37b 38b 39b 72% yield (10:1 dr) 70% yield (> 20:1 dr) 68% yield (> 20:1 dr) 61% yield (10:1 dr) 60% yield (6:1 dr) 88% ee 92% ee 92% ee 98% ee 89% ee Me O O OAc O Me H O H O O O H F O O F S H F O S H H O O O O O N † 40b N 41b 42b 58% yield (8:1 dr) 63% yield (13:1 dr) N 45% yield (> 20:1 dr) 81% ee 80% ee 92% ee

Gram-Scale Experiment

c O F BF3·Et2O (10.0 equiv) O N OMe H CIC1 (15 mol%) N OMe 4a m-CPBA (1.2 equiv) 4b 55% yield (> 20:1 dr) 1.0g, 3.74 mmol DCE, -25oC, 48 h 92% ee

Fig. 3 Substrate scope of the nucleophilic asymmetric fluorinations. Reactions were conducted on a 0.2 mmol scale with 10.0 equivalents of BF3·Et2Oin DCE (8.0 mL) at −25 °C for 48 h. The absolute configuration of 4b was assigned by X-ray crystallography (structure shown), and the configuration of all other products was assigned by analogy. a Substrate scope of N-cinnamylbenzamides bearing various substituents on Ar1 and Ar2 ring. b Substrate scope of N-cinnamylbenzamides with “complex substituents” on Ar2 ring and Ar2 = heterocyclic groups. c Gram-scale reaction with 4a, CIC1, m-CPBA and

BF3·Et2O. Isolated yields are reported. *The ee values of 30b and 31b could not be detected by HPLC. †CIC2 was employed as the catalyst for 42b. The ee values were determined by chiral HPLC analysis and the dr values were determined by 19F NMR. t-Bu, tert-butyl; Ph, phenyl; Me, methyl; Ac, acetyl.

4 NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3 ARTICLE

BF ·Et O (10.0 equiv) F Ar3 3 2 O I O CIC1 (15 mol%) 3 O O NH Ar O O O m N O 4 -CPBA (1.2 equiv) Bn Bn Ar o 4 C6H5F, -42 C, 20 h Ar 43a-54a 43b-54b CIC1

F F F F F F

O O O O O O Me N N N N N N Me t-Bu Me Me Me Me t-Bu t-Bu 43b 44b 45b 46b 47b 48b 88% yield, 85% ee 85% yield, 80% ee 82% yield, 85% ee 86% yield, 81% ee 83% yield, 83% ee 81% yield, 85% ee F F F F F F O O O O O O O N N N N N O N t CF3 OMe -Bu t-Bu OMe OEt t F3C MeO -Bu t-Bu 49b 50b 51b 52b 53b* 54b* 85% yield, 80% ee 87% yield, 85% ee 85% yield, 83% ee 83% yield, 85% ee 82% yield, 81% ee 80% yield, 80% ee

Fig. 4 Substrate scope expanding of the nucleophilic asymmetric ﬂuorinations using BF3·Et2O as the ﬂuorine source. Reactions were conducted on a 0.2 mmol scale with 10.0 equivalents of BF3·Et2OinC6H5F (8.0 mL) at −42 °C for 20 h; isolated yields are reported. *CIC2 was applied as the catalyst for the formation of 53b and 54b. The ee values were determined by chiral HPLC analysis. Me, methyl; t-Bu, tert-butyl; Et, ethyl.

B•nHF a IB, m-CPBA b BF •Et O O o F O 3 2 DCE, 0 C Ar-I=O (1) Ph N Ph N Int1 H or PhIF2 1a DCE, 0oC 1b ArI m-CPBA BF3 Et O I O + 2 O BF •Et O - [H] 1 H 3 2 Ar (2) N m N Ph IB, -CPBA Int2 2 H o O DCE, 0 C No desired F BF 1a 1a1 fluoro- O 3 product O BF •Et O N 3 2 + O F F B (3) Ph N IB, m-CPBA TS3 F H o N F F F 1a2 DCE, 0 C Ph H O B Ph F O N O O O O H BF3•Et2O Ph I H I N Ph O O O Ph (4) Ph O IB, m-CPBA O O O O Ph O 1a3 DCE, 0oC Ph Ph Ph n + - [BF2OBF3]- B• HF = Py•HF, Et3N•3HF Int4 TS1

Fig. 5 Plausible mechanism of the catalytic asymmetric nucleophilic fluorinations. a Control experiments. b Plausible catalytic cycle. intramolecular n−σ* interactions between the electron-deficient (Fig. 5b)49, which was calculated to be +11.1 kcal/mol in energy iodine (III) center and the carbonyl groups50,51. For the impacts relative to Int4+. And then Int5 is formed with 1.7 kcal/mol of of steric hindrance, the nucleophilic attack of F− on the Si face energy barrier relative to TS2. The hypervalent iodine (III) Int5 was favored (Fig. 5b), and this is in consistence with the underwent reductive elimination to afford Int6 with 6.8 kcal/mol experimental results and DFT calculations (Fig. 6b). As seen from of energy barrier. TS3 was formed by the intramolecularly the energy profile, the enantioselectivity is determined by the nucleophilic attack of the amide oxygen on C2 position with 4.5 - - fl + addition of F with [BF4] as uorine source to the cationic Int3 kcal/mol of energy barrier relative to Int6. Here the nucleophilic (Fig. 6a). The energy of TS1 versus ent-TS1, the transition state to attack of the amide oxygen takes place regioselectively at the the minor product, is compared. Surprisingly the two TSs exhibit higher substituted carbon atom of the cyclopropane unit49. Ring huge energy difference of 14.9 kcal/mol (Fig. 6b). A closer opening of the spirocyclopropyl ring TS3 takes place intramo- observation on the geometry shows that ent-TS1 is much later, lecularly via a cyclization with simultaneous ring expansion to the and bears a significantly longer I-O distance. In both TSs, the six-membered cationic Int7 (Fig. 6)49. The calculated energy hypervalent I(III) atom is stabilized by interaction with the amide barrier for this step is −38.9 kcal/mol relative to TS5. Control oxygen atom in the alkene. It could be proposed that the relative experiment (Fig. 5a, equation 2) demonstrated the necessity of direction of alkene and catalyst in ent-TS1 disabled the feasible I- Ar1 ring. Moreover, the aromatic ring (Ar2) is essential to O interaction that provides essential stabilization to the iodane, stabilize cationic Int7 (Fig. 5a, equation 3). With the assistance of - fi leading to both a later transition state and much higher energy. [BF2OBF3] anion, Int7 can be deprotonated to generate nal The formation of Int4 is achieved through the interaction product 1b (Fig. 6a). Besides, lengthening of carbon chain could between TS1 and BF , and Int4 is found to be 23.7 kcal/mol lower not result in a desired fluoro-product according to the control 3 − in energy than TS1. Then Int4 releases anionic [BF2OBF3] to experiment (Fig. 5a, equation 4). In a word, the formation of form Int4+, and Int4+ is found to be 5.5 kcal/mol lower in fluorinated oxazines follows a fluorination/1, 2-aryl migration/ energy than Int4. Dearomatization of Ar1 ring of Int4+ by cyclization cascade49. intramolecularly nucleophilic attack of the Ar1 on Si face at C2 We have disclosed an efficient asymmetric fluorinations process position (Fig. 5b) furnish the cyclopropyl compound TS2 that has enabled the development of the first highly enantioselective

NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3

Ar O I 26.6 m BzHN -CPBA Int1 BzHN F BF3 Ph Ar I 23.8 I H F Ar O Ar I Ph

in kcal/mol OBF BF3 2 BF3•Et2O HN O

G TS1 1a Int3 - BzHN Ph 0.0 3.6 BF4 4.0 F BzHN I F Int2 BF H TS3 3 -9.1 BzH F Ar F O Ph Ar O Int6 -4.5 - + BF3 I BF3 BF4 Int3 TS2 -9.0 Ar I Ph -15.8 N BF2 -19.7 Ph Ph -14.1 Int5 O -26.0 Int4 Int4+ I BzHN F 1b Ar -25.2 NHBz Ar I Ph BzHN BF2OHBF3 O BF - 2 H F -43.4 BF2OBF3 F B I 3 Ar Int7 - -BF2OBF3 F O Ph b NH Ph

TS1 (4.0) ent-TS1 (18.9)

Fig. 6 Calculated energy profile and DFT calculations for enantioselectivity of the catalytic asymmetric fluorination. a DFT-computed reaction profile for the catalytic asymmetric fluorination of 1a and BF3·Et2O in the presence of CIC1 and m-CPBA. b DFT-optimized stereo-determining transition state (TS1) structures and their relative energies for Int2.

fl uorination reaction (up to 99% ee and >20:1 dr) using BF3·Et2Oas DFT calculations. All calculations were carried out with the Gaussian the fluorine source and dual-activating reagent. Moreover, the 09 software52. The B3LYP functional53 was adopted for all calculations in combi- 54 substrate expanding experiments further demonstrated the wide nation with the D3BJ dispersion correction . For geometry optimization and fre- quency calculations, the SDD ECP and basis set55 was used for I and 6-31 G(d) for applicability of current method. This process provides not only a others56,57. The singlet point energy calculations were performed with a larger basis direct access to fluoro‐oxazine/benzoxazepine skeletons, but also a set combination, in which the def2-TZVP basis set58 was used for I, and 6–311 + G foundation for further development of new types of asymmetric (d,p)59,60 for others. The SMD implicit solvation model61 was used to account for nucleophilic fluorinations in future applications. The studies of the the solvation effect of DCE when performing single point energy calculations. applicability of this asymmetric fluorination methodology using other substrates are going on in our group. Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Methods Data availability General procedure for synthesis of 1b–42b. The substrate (0.2 mmol) and cat- All relevant data are available from the corresponding author upon reasonable request. alyst (15 mol%) were mixed into the reaction tube, and then DCE (8.0 ml) was All the data supporting the findings of this study are available within this article, and − added. The mixture was cooled to 25 °C, after stirring for 10 min at this tem- supplementary information files. The authors declare that all data generated or analyzed perature, m-CPBA (1.2 equiv.) was added in one portion, followed by addition of during this study are included in this Article (and its Supplementary Information). The − BF3·Et2O (10.0 equiv.) dropwise. The reaction was run at 25 °C for 48 h. The X-ray crystallographic coordinates for the structure of 4b are available free of charge reaction mixture was poured into saturated NaHCO3 (aq), the organic layer was from the Cambridge Crystallographic Data Centre under deposition number CCDC collected and washed with brine, dried over Na2SO4, and concentrated under 1960281. reduced pressure in the presence of basic Al2O3, Column chromatography (basic – = Al2O3, 200 300 mesh, EtOAc-hexane (0.5% Et3N) elution: hexane/EtOAc (V/V) 50:1 ~ 5:1) gave the corresponding fluorinated products. Received: 30 December 2020; Accepted: 9 June 2021;

General procedure for synthesis of 43b–54b. The substrate (0.2 mmol) and catalyst (20 mol%) were mixed into the reaction tube, and then C6H5F (8.0 ml) was added. The mixture was cooled to −42 °C, after stirring for 5 min at this tem- perature, m-CPBA (1.2 equiv) was added in one portion, followed by addition of − References BF3·Et2O (10.0 equiv) dropwise. The reaction was run at 42 °C for 20 h. The 1. Amii, H. & Uneyama, K. C−F bond activation in organic synthesis. Chem. reaction mixture was poured into saturated NaHCO3 (aq) sulotion, the organic Rev. 109, 2119–2183 (2009). layer was collected and washed with brine, dried over Na2SO4, and concentrated 2. Gouverneur, V. & Seppelt, K. Introduction: ﬂuorine chemistry. Chem. Rev. under reduced pressure in the presence of basic Al2O3, Column chromatography – 115, 563–565 (2015). (basic Al2O3, 200 300 mesh, EtOAc-Hexanes (0.5% Et3N) elution: hexanes/EtOAc (V/V) = 100:1 ~ 25:1) gave the corresponding ﬂuorinated products.

6 NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3 ARTICLE

3. Champagne, P. A. et al. Monofluorination of organic compounds: 10 years of 32. Katcher, M. H., Sha, A. & Doyle, A. G. Palladium-catalyzed regio- and innovation. Chem. Rev. 115, 9073–9174 (2015). enantioselective fluorination of acyclic allylic halides. J. Am. Chem. Soc. 133, 4. Zhu, Y. et al. Modern approaches for asymmetric construction of carbon−fluorine 15902–15905 (2011). quaternary stereogenic centers: synthetic challenges and pharmaceutical needs. 33. Edwards, J. A., Ringold, H. J. & Djerassi, C. Steroids. CXXXVI.1 Synthesis of a Chem. Rev. 118, 3887–3964 (2018). new class of potent cortical hormones. 6α-fluoro- and 6α,9α-difluoro-16α- 5. Petrone, D. A., Ye, J. & Lautens, M. Modern transition-metal-catalyzed methylprednisolone and related steroids2. J. Am. Chem. Soc. 82, 2318–2322 carbon–halogen bond formation. Chem. Rev. 116, 8003–8104 (2016). (1960). 6. Ma, J. A. & Cahard, D. Asymmetric fluorination, trifluoromethylation, and 34. Wei, Z. Y., Wang, D., Li, J. S. & Chan, T. H. Lewis acid-promoted perfluoroalkylation reactions. Chem. Rev. 104, 6119–6146 (2004). condensation of allylalkoxysilanes with carbonyl compounds. Synthesis of 7. Differding, E. & Lang, R. W. New fluorinating reagents - I. The first tetrahydropyrans. J. Org. Chem. 54, 5768–5774 (1989). enantioselective fluorination reaction. Tetrahedron Lett. 29, 6087–6090 (1988). 35. Cui, J. et al. Boron trifluoride etherate functioning as a fluorine source in an 8. Yang, X., Wu, T., Phipps, R. J. & Toste, F. D. Advances in catalytic iodosobenzene-mediated intramolecular aminofluorination of homoallylic enantioselective fluorination, mono-, di-, and trifluoromethylation, and amines. Org. Lett. 16, 1442–1445 (2014). trifluoromethylthiolation reactions. Chem. Rev. 115, 826–870 (2015). 36. Launay, G. G., Slawin, A. M. Z. & O’Hagan, D. Prins fluorination cyclisations: 9. Kohlhepp, S. V. & Gulder, T. Hypervalent iodine(III) fluorinations of alkenes preparation of 4-fluoro-pyran and -piperidine heterocycles. Beilstein J. Org. and diazo compounds: new opportunities in fluorination chemistry. Chem. Chem. 6, 41 (2010). Soc. Rev. 45, 6270–6288 (2016). 37. Yeh, M. C. P. et al. Transition-metal-free carbofluorination of TBS-protected 10. Ishimaru, T. et al. Cinchona alkaloid catalyzed enantioselective fluorination of nitrogen-containing cyclic enynols: synthesis of fluorinated azabicycles. J. Org. allyl silanes, silyl enol ethers, and oxindoles. Angew. Chem. Int. Ed. 47, Chem. 78, 5521–5529 (2013). 4157–4161 (2008). 38. Demeio, G. V. & Pinhey, J. T. Aryl fluorides from the reaction of boron 11. Miró, J., del Pozo, C., Toste, F. D. & Fustero, S. Enantioselective palladium‐ trifluoride with aryl-lead (IV) triacetates, which may be generated in situ from catalyzed oxidative β,β‐fluoroarylation of α,β‐unsaturated carbonyl aryltrimethylsilanes, triarylboroxines, and arenes. J. Chem. Soc. Chem. derivatives. Angew. Chem. Int. Ed. 55, 9045–9049 (2016). Commun. 15, 1065–1066 (1990). 12. Buckingham, F. et al. Organomediated enantioselective 18F fluorination for 39. Meio, G. D., Morgan, J. & Pinhey, J. T. Aryl fluoride syntheses involving PET applications. Angew. Chem. Int. Ed. 54, 13366–13369 (2015). reaction of aryllead triacetates with boron trifluoride-diethyl ether complex. 13. Rauniyar, V., Lackner, A. D., Hamilton, G. L. & Toste, F. D. Asymmetric Tetrahedron 49, 8129–8138 (1993). electrophilic fluorination using an anionic chiral phase-transfer catalyst. 40. Xu, Z. F., Dai, H. C., Shan, L. H. & Li, C. Y. Metal-free synthesis of (E)- fl Science 334, 1681 (2011). mono uoroenamine from 1-sulfonyl-1,2,3-triazole and Et2O·BF3 via 14. Wang, L. M. et al. Enantio‐ and diastereoselective hydrofluorination of enals stereospecific fluorination of α-diazoimine. Org. Lett. 20, 1054–1057 (2018). by N‐heterocyclic carbene catalysis. Angew. Chem. Int. Ed. 58, 7410–7414 41. Cresswell, A. J. et al. Pinacolatoboron fluoride (pinBF) is an efficient fluoride (2019). transfer agent for diastereoselective synthesis of benzylic fluorides. 15. Egami, H. et al. Asymmetric fluorolactonization with a bifunctional hydroxyl Tetrahedron Lett. 56, 3373–3377 (2015). carboxylate catalyst. J. Am. Chem. Soc. 137, 10132–10135 (2015). 42. Ding, C. H., Dai, L. X. & Hou, X. L. An efficient and highly regioselective fl fl 16. He, Y., Yang, Z. Y., Thornbury, R. T. & Toste, F. D. Palladium-catalyzed uorination of zziridines using BF3·OEt2 as uorine source. Synlett 2004, enantioselective 1,1-fluoroarylation of aminoalkenes. J. Am. Chem. Soc. 137, 2218–2220 (2004). 12207–12210 (2015). 43. Cresswell, A. J., Davies, S. D., Roberts, P. M. & Thomson, J. E. Beyond the Balz 17. Sandford, C., Rasappan, R. & Aggarwal, V. K. Synthesis of enantioenriched −Schiemann reaction: the utility of tetrafluoroborates and boron trifluoride as alkylfluorides by the fluorination of boronate complexes. J. Am. Chem. Soc. nucleophilic fluoride sources. Chem. Rev. 115, 566–611 (2015). 137, 10100–10103 (2015). 44. Yoshimura, A. & Zhdankin, V. V. Advances in synthetic applications of 18. Hiramatsu, K., Honjo, T., Rauniyar, V. & Toste, F. D. Enantioselective hypervalent iodine compounds. Chem. Rev. 116, 3328–3435 (2016). synthesis of fluoro-dihydroquinazolones and -benzooxazinones by 45. Wirth, T. Hypervalent iodine chemistry in synthesis: scope and new fluorination-initiated asymmetric cyclization reactions. ACS Catal. 6, 151–154 directions. Angew. Chem. Int. Ed. 44, 3656–3665 (2005). (2016). 46. Egami, H. et al. Dianionic phase-transfer catalyst for asymmetric aluoro- 19. Suzuki, S. et al. Iodoarene-catalyzed fluorination and aminofluorination by an cyclization. J. Am. Chem. Soc. 140, 2785–2788 (2018). Ar-I/HF·pyridine/mCPBA system. Chem. Sci. 5, 2754–2760 (2014). 47. Masanori, S., Shinichi, A. & Shoji, H. A practical synthetic method for 20. Pluta, R. et al. Metal-free catalytic asymmetric fluorination of keto esters using iodoarene difluorides without fluorine gas and mercury salts. Synthesis 2002, a combination of hydrogen fluoride (HF) and oxidant: experiment and 1802–1803 (2002). computation. ACS Catal. 8, 2582–2588 (2018). 48. Scheidt, F. et al. Enantioselective, catalytic vicinal difluorination of alkenes. 21. Banik, S. M., Medley, J. W. & Jacobsen, E. N. Catalytic, asymmetric Angew. Chem. Int. Ed. 130, 16669–16673 (2018). difluorination of alkenes to generate difluoromethylated stereocenters. Science 49. Ulmer, A. et al. A fluorination/aryl migration/cyclization cascade for the 353,51–54 (2016). metal‐free synthesis of fluoro‐benzoxazepines. Chem. Eur. J. 22, 3660–3664 22. Woerly, E. M., Banik, S. M. & Jacobsen, E. N. Enantioselective, catalytic (2016). fluorolactonization reactions with a nucleophilic fluoride source. J. Am. Chem. 50. Uyanik, M., Yasui, T. & Ishihara, K. Hypervalent compounds. Angew. Chem. Soc. 138, 13858–13861 (2016). Int. Ed. 49, 2175–2177 (2010). 23. Mennie, K. M., Banik, S. M., Reichert, E. C. & Jacobsen, E. N. Catalytic 51. Haubenreisser, S. et al. Structurally defined molecular hypervalent iodine diastereo- and enantioselective fluoroamination of alkenes. J. Am. Chem. Soc. catalysts for intermolecular enantioselective reactions. Angew. Chem. Int. Ed. 140, 4797–4802 (2018). 55, 413–417 (2016). 24. Scheidt, F. et al. Enantioselective, catalytic vicinal difluorination of alkenes. 52. Gaussian 09, rev. D.01 (Gaussian Inc., 2013). Angew. Chem. Int. Ed. 57, 16431–16435 (2018). 53. Stephens, P. J., Devlin, F. J., Chabalowski, C. F. & Frisch, M. J. Ab initio 25. Levin, M. D. et al. Catalytic enantioselective synthesis of difluorinated alkyl calculation of vibrational absorption and circular dichroism spectra using bromides. J. Am. Chem. Soc. 142, 14831–14837 (2020). density functional force fields. J. Phys. Chem. 98, 11623–11627 (1994). 26. Zhang, Q. et al. Iridium-catalyzed enantioselective fluorination of racemic, 54. Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. A consistent and accurate ab secondary allylic trichloroacetimidates. J. Am. Chem. Soc. 137, 11912–11915 initio parametrization of density functional dispersion correction (DFT-D) for (2015). the 94 elements H-Pu. J. Chem. Phys. 132, 154104 (2010). 27. Bruns, S. & Haufe, G. Enantioselective introduction of fluoride into organic 55. Bergner, A. et al. Ab initio energy-adjusted pseudopotentials for elements of compounds: First asymmetric ring opening of epoxides by hydrofluorinating groups 13-17. Mol. Phys. 80, 1431–1441 (1993). reagents. J. Fluor. Chem. 104, 247–254 (2000). 56. Hariharan, P. C. & Pople, J. A. The influence of polarization functions on 28. Haufe, G. & Bruns, S. Salen chromium complex mediated asymmetric ring molecular orbital hydrogenation energies. Theor. Chim. Acta 28, 213–222 opening of meso‐ and racemic epoxides with different fluoride sources. Adv. (1973). Synth. Catal. 344, 165–171 (2002). 57. Gnan, N. et al. Dynamical properties of different models of elastic polymer 29. Kalow, J. A. & Doyle, A. G. Enantioselective ring opening of epoxides by rings: Confirming the link between deformation and fragility. J. Chem. Phys. fluoride anion promoted by a cooperative dual-catalyst system. J. Am. Chem. 56, 2257–2261 (1972). Soc. 132, 3268–3269 (2010). 58. Weigend, F. & Ahlrichs, R. Balanced basis sets of split valence, triple zeta 30. Zhu, J. T., Tsui, G. C. & Lautens, M. Rhodium‐catalyzed enantioselective valence and quadruple zeta valence quality for H to Rn: Design and nucleophilic fluorination: ring opening of oxabicyclic alkenes. Angew. Chem. assessment of accuracy. Phys. Chem. Chem. Phys. 7, 3297 (2005). Int. Ed. 51, 12353–12356 (2012). 59. Clark, T., Chandrasekhar, J., Spitznagel, G. W. & Schleyer, P. V. Efficient 31. Pupo, G. et al. Asymmetric nucleophilic fluorination under hydrogen bonding diffuse function-augmented basis sets for anion calculations. III. The 3-21+G phase-transfer catalysis. Science 360, 638–642 (2018). basis set for first-row elements, Li-F. J. Comput. Chem. 4, 294–301 (1983).

NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-24278-3

60. Krishnan, R., Binkley, J. S., Seeger, R. & Pople, J. A. Self-consistent molecular Additional information orbital methods. XX. A basis set for correlated wave functions. J. Chem. Phys. Supplementary information The online version contains supplementary material 72, 650–654 (1980). available at https://doi.org/10.1038/s41467-021-24278-3. 61. Marenich, A. V., Cramer, C. J. & Truhlar, D. G. Universal solvation model based on solute electron density and on a continuum model of the solvent Correspondence and requests for materials should be addressed to X.J. deﬁned by the bulk dielectric constant and atomic surface tensions. J. Phys. Chem. B 113, 6378–6396 (2009). Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Acknowledgements Reprints and permission information is available at http://www.nature.com/reprints We appreciate the financial support from the National Natural Science Foundation of Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in China (No. 91853106), the Program for Guangdong Introducing Innovative and fi Enterpre-neurial Teams (No. 2016ZT06Y337), Guangdong Provincial Key Laboratory of published maps and institutional af liations. Construction Foundation (No. 2019B030301005), Shenzhen Science and Technology Program (JSGG20200225153121723), and the Fundamental Research Funds for the Central Universities (No. 19ykzd25). We also thank Dr. Chaoxian Yan for his help in Open Access This article is licensed under a Creative Commons DFT calculations. Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Author contributions Commons license, and indicate if changes were made. The images or other third party X.X.J. and W.W.Z. conceived of and directed the project; W.W.Z. developed the catalytic material in this article are included in the article’s Creative Commons license, unless asymmetric fluorinations; W.W.Z., X.Z. and J.K.A. conducted most of the experiments; indicated otherwise in a credit line to the material. If material is not included in the J.Y.W., Y.P.C. and W.W.Z. conducted the DFT calculations and provided mechanism article’s Creative Commons license and your intended use is not permitted by statutory analysis; X.Y.M., J.K.L, Y.J.Q, H.Z. and J.J.X. helped the conduction of the project; and regulation or exceeds the permitted use, you will need to obtain permission directly from X.X.J. and W.W.Z. co-wrote the manuscript. the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Competing interests The authors declare no competing interests. © The Author(s) 2021

8 NATURE COMMUNICATIONS | (2021) 12:3957 | https://doi.org/10.1038/s41467-021-24278-3 | www.nature.com/naturecommunications