Catalytic asymmetric trifluoromethylthiolation via enantioselective [2,3]-sigmatropic rearrangement of sulfonium ylides

Article abstract of DOI:10.1038/nchem.2789

The trifluoromethylthio (SCF 3) functional group has been of increasing importance in drug design and development as a consequence of its unique electronic properties and high stability coupled with its high lipophilicity. As a result, methods to introduce this highly electronegative functional group have attracted considerable attention in recent years. Although significant progress has been made in the introduction of SCF 3 functionality into a variety of molecules, there remain significant challenges regarding the enantioselective synthesis of SCF 3 -containing compounds. Here, an asymmetric trifluoromethylthiolation that proceeds through the enantioselective [2,3]-sigmatropic rearrangement of a sulfonium ylide generated from a metal carbene and sulfide (Doyle-Kirmse reaction) has been developed using chiral Rh(II) and Cu(I) catalysts. This transformation features mild reaction conditions and excellent enantioselectivities (up to 98% yield and 98% e.e.), thus providing a unique, highly efficient and enantioselective method for the construction of C(sp 3)-SCF 3 bonds bearing chiral centres.

Full text of DOI:10.1038/nchem.2789

ARTICLES
PUBLISHED ONLINE: 5 JUNE 2017 | DOI: 10.1038/NCHEM.2789
Catalytic asymmetric trifluoromethylthiolation via
enantioselective [2,3]-sigmatropic rearrangement
of sulfonium ylides
Zhikun Zhang, Zhe Sheng, Weizhi Yu, Guojiao Wu, Rui Zhang, Wen-Dao Chu, Yan Zhang
and Jianbo Wang
*
The trifluoromethylthio (SCF3) functional group has been of increasing importance in drug design and development as a
consequence of its unique electronic properties and high stability coupled with its high lipophilicity. As a result, methods
to introduce this highly electronegative functional group have attracted considerable attention in recent years. Although
significant progress has been made in the introduction of SCF3 functionality into a variety of molecules, there remain
significant challenges regarding the enantioselective synthesis of SCF3-containing compounds. Here, an asymmetric
trifluoromethylthiolation that proceeds through the enantioselective [2,3]-sigmatropic rearrangement of a sulfonium ylide
generated from a metal carbene and sulfide (Doyle–Kirmse reaction) has been developed using chiral Rh(II) and Cu( )
I
catalysts. This transformation features mild reaction conditions and excellent enantioselectivities (up to 98% yield and 98%
e.e.), thus providing a unique, highly efficient and enantioselective method for the construction of C(sp3)–SCF3 bonds bearing
chiral centres.
he introduction of SCF3 into small organic molecules has
In our efforts to develop new methods of this sort, we recognized
recently attracted considerable attention in the fields of phar- that the [2,3]-sigmatropic rearrangement of a sulfonium ylide is a
T
maceuticals, agrochemicals and materials science1–5. This is unique method to construct C(sp3)–S bond20–27, and thus we
attributed to the fact that the incorporation of an SCF3 group into sought to exploit this type of transformation to introduce SCF3
organic molecules can significantly affect their physicochemical groups into organic compounds. Moreover, based on previous
and pharmacokinetic properties due to its electron-withdrawing studies28–37, we hypothesized that such a reaction should proceed
nature, high lipophilicity and metabolic stability. As a result, the stereoselectively and possibly permit the enantioselective construc-
development of efficient approaches to construct C–SCF3 bonds tion of these molecules. In the catalytic reaction, the sulfonium
has been studied intensively, and new methods based on both ylides for the [2,3]-sigmatropic rearrangement are formed through
nucleophilic and electrophilic trifluoromethylthiolation reagents the reaction of allyl or propargyl sulfides and metal carbene
have been reported6–13. In contrast, the catalytic asymmetric tri- species generated by Rh(II)- or Cu(I)-catalysed reaction with diazo
fluoromethylthiolation has remained a challenging task, and there substrates (Doyle–Kirmse reaction). The catalytic enantioselective
are few examples in the literature (Fig. 1a)14–19. In 2013, a highly variant of this reaction is attractive; however, until now the enantios-
enantioselective quinidine-catalysed trifluoromethylthiolation of electivities that have been achieved remain mediocre23–40. The chal-
β-ketoesters derived from indanones using N-trifluoro- lenge lies in discriminating between the heterotopic lone pairs of
methylthiophthalimide as the electrophilic “SCF3” source was sulfur by a chiral metal carbene (Fig. 1b). In 2002, we achieved mod-
reported14. A similar reaction system has also been applied to erate enantioselectivity (78% e.e.) with a chiral copper catalyst,
enantioselective trifluoromethylthiolation of oxindoles15. In the which still represents the highest e.e. value for a [2,3]-sigmatropic
same year, the quinine-catalysed enantioselective trifluoro- rearrangement of a sulfonium ylide under catalytic conditions34.
methylthiolation of β-ketoesters by employing electrophilic With our double asymmetric induction approach by exploring
trifluoromethylthiolation reagent was disclosed16. Later, the same diazo substrate bearing chiral auxiliary, high stereoselectivities
reagent was used in the trifluoromethylthiolation of β-ketoesters have been achieved later37. Herein we report the highly enantioselec-
with copper–boxmi complex catalysts17. Cinchona alkaloid- tive Rh2L*4- and Cu(I)/L*-catalysed [2,3]-sigmatropic rearrange-
catalysed enantioselective trifluoromethylthiolation of oxindoles has ment of sulfonium ylides, which are generated from allyl or
been developed by also employing in situ generated electrophilic propargyl trifluoromethyl sulfides (Fig. 1b). The reactions represent
trifluoromethylthiolation reagent18. More recently, an enantioselec- a unique approach for the enantioselective construction of C(sp3)–
tive trifluoromethylthiolating lactonization using an indane-based SCF3 bonds, alongside the highest enantioselectivities ever achieved
chiral sulfide as the catalyst was reported19. Regardless of this for the Doyle-Kirmse reaction.
remarkable progress during the past few years, the field is still in
its infancy, as the reported reactions are limited in scope, mainly
focused on β-ketoester type substrates. Thus, further development of
novel approaches to achieve enantioselective trifluoromethylthiolation
is still highly desirable.
Results and discussion
Our previous studies have demonstrated that both chiral Cu(I) and
Rh(II) catalysts are effective in enantioselective Doyle–Kirmse reac-
tion34–39, so we first focused on employing commercially available
Beijing National Laboratory of Molecular Sciences (BNLMS), Key Laboratory of Bioorganic Chemistry and Molecular Engineering of Ministry of Education,
College of Chemistry, Peking University, Beijing 100871, China. e-mail: wangjb@pku.edu.cn
*
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
1
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2789
ARTICLES
(R = Me) as the substrates, and the commonly used non-chiral
Rh2(OAc)4 as the catalyst (Table 1, entry 1). Gratifyingly, the
expected product 3 was formed as detected by GC–MS and 19F
NMR, and further confirmed by 1H NMR and 13C NMR.
Subsequently, we proceeded to screen a series of previously devel-
oped41,42 chiral Rh(II) catalysts in order to realize the enantio-
selective version of the reaction (Table 1, entries 2–6). It was
apparent that Rh2(S-TBSP)4 and Rh2(S-DOSP)4 afforded moderate
yields and enantioselectivities (Table 1, entries 2–3). With
Rh2(S-DOSP)4 as the catalyst, we further studied the effect of reac-
tion temperature (Table 1, entries 7–12). Significantly improved
results in terms of both yield and e.e. could be obtained by carrying
out the reaction at low temperature. The reaction afforded optimal
results at –30 °C, whereas the reaction became sluggish and the
yields were diminished when carried out below –40 °C.
Subsequently, we examined the effects of the ester moiety of the dia-
zoacetate. We concluded that the enantioselectivities were consider-
ably diminished as the steric bulk of the ester moiety increased
(Table 1, entries 13–16).
a
O
O
Cat: cinchona alkaloid
or CuL*
SCF3
CO2R
*
CO2R
( )n
( )n
+
"
SCF3" source
O
+
"
SCF3" source:
I
OSCF3
N
SCF3
O
b
F3C
+
F3CS
R'
N2
MLn*
S
–
R
R
R'
R
*
R'
S
F3C
Non-racemic free ylide
Figure 1 | Catalytic asymmetric trifluoromethylthiolation. a, Asymmetric
trifluoromethylthiolation with electrophilic trifluoromethylthiolating reagents.
b, Asymmetric trifluoromethylthiolation through enantioselective
Doyle–Kirmse reaction.
Finally, the effect of solvent was investigated (Table 1, entries 17–
21). Notably, previous studies have indicated that Rh2(S-DOSP)4-
catalysed asymmetric reactions of donor–acceptor diazo compounds
chiral Rh(II) catalysts, which have been successfully applied in a are favoured in nonpolar solvents43. We found that in the current
series of catalytic asymmetric carbene transfer reactions41,42. The reaction the nonpolar solvents cyclopentane and n-pentane also
[2,3]-sigmatropic rearrangement of sulfonium ylide generated afforded improved enantioselectivities. With n-pentane as the
from a metal carbene and allyl trifluoromethyl sulfide has not solvent, the reaction at −30 °C with 1 mol% Rh2(S-DOSP)4 gave
been previously documented, so to verify if such a reaction works 95% isolated yield and 91% e.e. (Table 1, entry 21). As anticipated,
with CF3-bearing allyl sulfide an initial experiment was carried a similar yield and selectivity, but reversed chirality, could be
out with allyl trifluoromethyl sulfide 1 and phenyldiazoacetate 2 obtained by using Rh2(R-DOSP)4 (Table 1, entry 22).
Table 1 | Optimization of the reaction conditions.
X
H
N
ArO2S
CO2Me
H
O
O
N
O
N2
Rh2L*4
SCF3
CO2R
Rh Rh
SCF3
+
Rh Rh
Ph
CO2R
Solvent, T
2–24 h
Ph
*
3
Rh2(5S-MEPY)4 (X = CH2)
Rh2(4S-MEOX)4 (X = O)
Rh2(4S-MPPIM)4 (X = PhCH2CH2N)
Rh2(S-TBSP)4 (Ar = p-tBuC6H4)
1
2
Rh2(S-DOSP)4 (Ar = p-C12H25C6H4)
Entry*
2, R
Rh2L4
T (°C)
t (h)
Solvent
Yield (%)†
e.e. (%)‡
1
2
3
4
5
6
7
8
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
Rh2(OAc)4
40
40
40
40
40
40
20
0
−10
−30
−40
−60
−30
−30
−30
−30
20
20
20
20
−30
−30
5
2
2
3
3
5
2
3
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
PhCH3
72
38
44
44
<5
15
85
93
89
90
78
<5
72
53
61
0
Rh2(S-TBSP)4
Rh2(S-DOSP)4
Rh2(5S-MEPY)4
Rh2(4S-MEOX)4
Rh2(4S-MPPIM)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(R-DOSP)4
48
56
2
13
20
58
65
68
72
72
70
74
56
11
9
4
9
10
11
12
13
14
15
16
17
18
19
20
21
22
12
24
12
12
12
12
12
12
12
12
24
24
iPr
tBu
Ph
Et
Et
Et
Et
Et
Et
81
7
63
58
69
86
95§
85§
74
79
79
52
91
n-Pentane
Cyclopentane
CHCl3
n-Pentane
n-Pentane
−89
*Entry 1 was carried with 1 (0.2 mmol), 2 (0.24 mmol) and Rh2(OAc)4 (1.1 mol%); entries 2–20 were carried with 1 (0.2 mmol), 2 (0.2 mmol) and Rh2L*4 (0.13 mol%); entries 21–22 were carried with 1 (0.22
mmol), 2 (0.2 mmol) and Rh2L*4 (1.0 mol%). †Unless otherwise noted, the yields refer to 19F NMR (400 MHz) yield with trifluorotoluene as the internal standard. ‡Determined by HPLC on a chiral stationary
phase (for details, see Supplementary Section 9). §Isolated yield with preparative thin-layer chromatography. Rh2(S-TBSP)4: dirhodium(II) tetrakis[(S)-N-(p-butylphenylsulfonyl)prolinate]; Rh2(5S-MEPY)4:
dirhodium(II) tetrakis[methyl 2-pyrrolidone-5(S)-carboxylate]; Rh2(4S-MEOX)4: dirhodium(II) tetrakis[methyl 2-oxazolidone-4(S)-carboxylate]; Rh2(4S-MPPIM): dirhodium(II) tetrakis[methyl 1-(3-
phenylpropanoyl)-2-oxoimidazolidine-4(S)-carboxylate]; Rh2(S-DOSP)4: dirhodium(II) tetrakis[(S)-N-(p-dodecylphenylsulfonyl)prolinate].
2
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2789
ARTICLES
Table 2 | Substrate scope for the [2,3]-sigmatropic rearrangement.
Rh2(S-DOSP)4
(0.5–1.0 mol%)
SCF3
SCF3
N2
SCF3
SCF3
+
or
or
R CO2R´
*
R CO2R´
*
3a–m, 5a–n
n-Pentane, –30 °C
R
CO2R´
6
1
2–24 h
2a–m; 4a–n
7a–m
SCF3
3a, X = H, 95%, 91% e.e.
3b, X = Me, 88%, 89% e.e.
3c, X = F, 98%, 91% e.e.
3d, X = Br, 76%, 86% e.e.
3e, X = OMe, 73%, 79% e.e.
3f, X= Ph, 64%, 88% e.e.
SCF3
MeO2C
CO2Et
*
3k, 44%, 86% e.e.
3l, 88%, 84% e.e.
3m, 84%, 92% e.e.
CO2Et
*
X
SCF3
CO2Et
*
SCF3
3g, X = F, 99%, 75% e.e.
3h, X = Cl, 98%, 91% e.e.
3i, X = I, 98%, 85% e.e.
3j, X = Me, 99%, 90% e.e.
CO2Et
SCF3
*
X
CO2Et
*
S
SCF3
CO2Me
5a, X = H, 99%, 98% e.e.
5b, X = Me, 99%, 95% e.e.
5c, X = OMe, 74%, 76% e.e.
5d, X = F, 98%, 98% e.e.
5e, X = Cl, 99%, 98% e.e.
5f, X= Br, 98%, 98% e.e.
5g, X= I, 84%, 96% e.e.
*
5l, 76%, 93% e.e.
SCF3
CO2Me
SCF3
*
CO2Me
X
*
5m, 89%, 94% e.e.
5n, 82%, 93% e.e.
5h, X= CF3, 94%, 96% e.e.
5i, X= CO2Me, 87%, 93% e.e.
SCF3
SCF3
5j, X = F, 94%, 93% e.e.
5k, X = Me, 76%, 86% e.e.
CO2Me
*
Me
CO2Me
*
X
SCF3
7g, X = H, 74%, 90% e.e.
7h, X = F, 68%, 86% e.e.
7i, X= Br, 54%, 85% e.e.
7j, X = CF3, 85%, 91% e.e.
7k, X = CO2Me, 60%, 85% e.e.
SCF3
7a, X = H, 65%, 90% e.e.
7b, X = Me, 55%, 87% e.e.
7c, X = F, 74%, 92% e.e.
7d, X = OMe, 61%, 67% e.e.
CO2Me
CO2Et
*
*
X
X
SCF3
SCF3
CO2Et
7e, 33%, 81% e.e.
7f, 65%, 84% e.e.
*
CO2Me
*
7l, 50%, 70% e.e.
Cl
F
SCF3
SCF3
CO2Et
*
7m, 51%, 90% e.e.
CO2Me
S
*
Me
Reaction conditions: detailed procedures can be found in the Supplementary Section 3. Generally, with 0.2 mmol scale, the solution of 2 or 4, 1 or 6 and Rh2(S-DOSP)4 (1.0 or 0.5 mol%) in pentane was stirred at
−30 °C for 2–24 h. All of the yields refer to the isolated products with preparative thin-layer chromatography. The e.e. value was determined by HPLC on a chiral stationary phase (for details, see Supplementary
Section 9). For 3j, the reaction was carried out at −5 °C.
F3CS
F3CS
NO2
(1) LiAlH4
CO2Et
R
*
O
O
(2)
Ph
Ph
p-O2NC6H4CCl
O
3f, 88% e.e.
8, 88% e.e.
Figure 2 | Assignment of the absolute configuration of the product 3f. Absolute configuration confirmation through two-step derivatization. Crystals of 8
were grown by diffusion of hexane into a diethyl ether solution of 8. Supplementary Section 7 gives the detailed crystal data.
With the optimized reaction conditions in hand, the substrate aryldiazoester bearing an electron-donating methoxy group on the
scope of this reaction was then investigated with a series of aryl dia- aromatic ring gave diminished yield and enantioselectivity (3e,
zoesters. As shown in Table 2, various aryl diazoesters 2a–m reacted 73%, 79% e.e.). The low yield is due to side reactions, mainly dimer-
smoothly with allyl trifluoromethyl sulfide 1 to afford the corre- ization of the diazoester. This is perhaps due to the relatively weak
sponding chiral trifluoromethyl sulfides 3a–m in good-to-excellent interaction between less electron-deficient carbenic carbon and the
yields and good enantioselectivities. The reaction with sulfur of 1. The low yield in the case of 3k is mainly attributed to the
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
3

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2789
ARTICLES
Table 3 | Substrate scope for the Cu(I)-catalysed [2,3]-sigmatropic rearrangement.
R R
L*
O
O
N2
Cu(CH3CN)4PF6 (5 mol%)/L*
SCF3
OtBu
+
SCF3
N
N
Ar
t
p-Xylene, 30 °C, 6 h
Ar
CO2 Bu
*
O
(1R, 2S)
R = o-MeC6H4
9a–k
1
10a–k
10a, X = H, 86%, 90% e.e.
10b, X = F, 91%, 90% e.e.
10c, X = Cl, 93%, 89% e.e.
10d, X = Br, 98%, 88% e.e.
10e, X = Ph, 89%, 90% e.e.
10f, X = Me, 81%, 90% e.e.
SCF3
10h, X = Me, 91%, 91% e.e.
10i, X = OMe, 93%, 83% e.e.
SCF3
X
t
CO2 Bu
*
t
CO2 Bu
*
X
10g, X = OMe, 52%, 80% e.e.
SCF3
SCF3
tBuO2C
t
S
CO2 Bu
*
10k, 50%, 90% e.e.
S
10j, 92%, 90% e.e.
Reaction conditions: 1 (0.2 mmol), 9a–k (0.4 mmol), Cu(CH3CN)4PF6 (5 mol%), L* (6 mol%), p-xylene (2 ml), 30 °C, 6 h. All of the yields refer to isolated yield by preparative thin-layer chromatography. The e.
e. values were determined by HPLC on a chiral stationary phase (for details, see Supplementary Section 9). For 10f, the reaction time was 10 h.
poor solubility of the diazo substrates in n-pentane. The steric bulk obtained for aryl diazoesters in general. The reaction with aryldia-
also affected the reaction as shown by the case of 3j, for which the zoesters bearing electron-poor or electron-neutral aromatic rings
reaction temperature needed to be raised to −5 °C. For comparison, afforded good yields and enantioselectivities. In the case of 10g,
it is noteworthy that the steric bulk of the ester moiety significantly in which the diazo substrates bear an electron-rich aromatic ring,
affects the reaction outcome as shown in Table 1 (entries 14–16)
the reaction did not give good yield of the product and the enantios-
Next, we investigated the reaction scope with respect to vinyldia- electivity was only moderately high. This is attributed to the fact that
zoacetates 4a–n. For the reaction with vinyldiazoacetates the catalyst such diazo compounds are less stable, thus resulting in significant
loading could be reduced to 0.5 mol%. As summarized in Table 2, amounts of side products. The diazoester bearing a heteroaromatic
excellent yields and enantioselectivities could be obtained with ring also gave a good e.e., although only a moderate yield (10k). The
most substrates. We found that the steric of the ester moiety had absolute configuration of the product 10j was determined as S by X-
a noticeable effect on enantioselectivity, similar to previous obser- ray crystallography (Supplementary Section 7.2). Unfortunately, the
vations: the reaction with ethyl ester gave only 87% e.e., whereas reaction with vinyldiazoacetates under the same conditions gave
methyl ester afforded 98% e.e. Electronic effects were also observed: poor enantioselectivities, and the reaction with propargyl sulfide 6
the reaction with vinyldiazoacetate bearing an electron-rich was sluggish.
methoxy substituent gave diminished yield and e.e. (5c).
Gratifyingly, the reaction with styryldiazoester and propenyldiazoe- of sulfur ylide generated from metal carbene under catalytic reaction
ster also worked smoothly (5m and 5n). has been investigated previously. An outstanding question is
The reaction mechanism of the [2,3]-sigmatropic rearrangement
We further expanded the [2,3]-sigmatropic rearrangement to whether the [2,3]-sigmatropic rearrangement proceeds through a
propargyl sulfide 6 in which the allenyl sulfides were the corre- free ylide or a metal-bound ylide28–33,44. In the case of the [2,3]-
sponding products35. As shown in Table 2, the reaction of trifluor- sigmatropic rearrangement involving an oxonium ylide, the reaction
omethyl propargyl sulfide
vinyldiazoacetates 4 afforded the corresponding sulfide products backside displacement of the metal catalyst45–47. However, the asym-
in moderate to good yields and enantioselectivities. metric induction model established for oxonium ylide failed to cor-
6 with aryldiazoacetates 2 or is proposed to proceed with a metal-bound ylide through a formal
To assign the absolute configuration of the newly generated rectly predict the absolute configuration of the rearrangement
stereogenic carbon centre of the product, the optically active products in the current study. Moreover, our previous study on
product 3f was converted into ester 8 through reduction and acyla- Cu(I)-catalysed reactions suggested that the [2,3]-sigmatropic
tion (Fig. 2). The X-ray crystallographic analysis of single crystal of 8 rearrangement most likely proceeds through a free ylide34,37. For
determined the absolute configuration of the stereogenic carbon the Rh(II)-catalysed reaction, we propose that the enantioselection
centre to be R (Supplementary Section 7.1).
is due to the formation of a chiral free sulfur ylide through the differ-
Encouraged by the high enantioselectivities achieved by chiral Rh entiation of the heterotopic lone pairs of the sulfur atom by a chiral
(II) catalysts in Doyle–Kirmse reaction, we investigated the same metal carbene. The high enantioselectivity observed in this reaction
reaction with Cu(I) catalysts bearing chiral ligands. Cu(I) catalysts system may be attributed to the following factors: (1) the Rh2(S-
are low cost and tunable with chiral ligands, and our previous DOSP)4 has great face selectivity with aryldiazoacetates and vinyl-
studies have suggested they are effective catalysts in achieving high diazoacetates to generate the sulfur ylide in high enantioselectivity;
stereoselectivity in Doyle–Kirmse reaction34–37. As shown in Table 3, (2) the free sulfur ylide has sufficient configurational stability; (3)
after extensive screening of the ligands, we found that ligand L* the subsequent concerted [2,3]-sigmatropic rearrangement is facile
afforded the optimal results in terms of both yield and enantioselectiv- with complete transfer of the chirality from sulfur to carbon48,49
ity (for the details of the reaction optimization, see Supplementary Control experiments were carried out to gain insights into the
.
Section 5.2). With the optimized reaction conditions in hand, we mechanism of the trifluoromethylthiolation reaction. First, the
investigated the scope of the substrates with a series of aryldiazoesters. dependence of diastereoselectivity and the Rh(II) catalysts was inves-
It was observed that good yields and enantioselectivities could be tigated. If the diastereoselectivity varied with the Rh(II) catalysts
4
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2789
ARTICLES
a
Rh2L4
12 (%)
d.r.
Ph
*
Rh2(OAc)4
40:60
43:57
43:57
42:58
44:56
79
98
55
99
99
N2
F3CS
Ph
*
Rh2L4 (0.5 mol%)
CH2Cl2, –10 °C
Rh2(S-DOSP)4
Rh2(5S-MEPY)4
Rh2(S-BTPCP)4
Rh2(S-PTAD)4
+
SCF3
Ph
Ph
CO2Et
CO2Et
11
2a
12
b
N2
Rh2L*4
Diazo substrate
e.e. (%)
Yield (%)
Rh2L*4
R
R´
2a
2e
2j
4a
4i
2a
2a
2a
2a
2a
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
Rh2(S-DOSP)4
14a, 89
14b, 66
14c, 57
14d, 38
14e, 39
0
1
2
3
3
2
3
3
2
4
2a,e,j; 4a,i
S
+
*
R
14a-e
R´
S
13
+
S
Rh2(S-DOSP)4
.
Rh2(S-PTTL)4 2EtOAc 14a, 36
R1
X
R2
Rh2(S-BTPCP)4
Rh2(S-PTAD)4
Rh2(5S-MEPY)4
Rh2(4S-MEAZ)4
14a, 39
14a, 16
14a, 10
14a, 22
–
–
Achiral free ylide
X =
X = Rh2L*4
Ylide associated with chiral catalyst
c
N2
+
CF3
+
S
CF3
Ph
S
Ph
CO2Et
Ph
+ Rh2L4
Rh2(S-DOSP)4
2a
+
100% consumption
CO2Et
Rh2L4
Ph
–
Ph
CO2Et
n-Pentane, –30 °C
of 2a
–
Ph
SCF3
24 h
I
II
15
d
CF3
Ph
Ph
More steric
interaction
Less steric
interaction
–
–
+
S
(R)
+
(S)
S
CF3
EtO2C
EtO2C
i
ii
CF3
S
ii
i
S
EtO2C
CO2Et
–
Ph
CF3
O
O
–
O
O
i
ii
i
S
ii
S
+
(S)
(R)
Ph
+
CF3
CF3
Sterically
blocked
Sterically
blocked
SCF3
CO Et
F3CS
Ph
Model A (less favoured)
Model B (favoured)
Ph
CO Et
(S)
2
(R)
2
Figure 3 | Mechanistic experiments. a, Study on the dependence of diastereoselectivities on the Rh(II) catalysts. b,c, control experiments with sulfides 13 (b)
and 15 (c) for proving the free ylide process of [2,3]-sigmatropic rearrangement. d, Proposed model of asymmetric induction. Rh2(S-BTPCP)4: dirhodium(II)
tetrakis[(S)-(+)-[(1S)-1-(4-bromophenyl)-2,2-diphenylcyclopropanecarboxylate]; Rh2(S-PTAD)4: dirhodium(II) tetrakis[(S)-(+)-(1-adamantyl)-(N-phthalimido)
acetate]; Rh2(S-PTTL)4: dirhodium(II) tetrakis[N-phthaloyl-(S)-t-leucinate]; Rh2(4S-MEAZ)4: dirhodium(II) tetrakis[methyl-2-oxaazetidine-4(S)-carboxylate].
The labels i and ii refer to the enantiotopic lone pairs of the trifluoromethylallylthioether that furnish the R and S products, respectively.
used in the reaction, the [2,3]-sigmatropic rearrangement is likely to catalysts and the e.e. values of the products in all the reactions are
proceed through Rh(II)-associated ylide and vice versa. This is due close to zero (0–4% e.e.) (Fig. 3b), further supporting a free ylide
to the fact that the diastereoselectivity-determining step (or the mechanism. In another control experiment, the sulfur ylide 15 gen-
C–C bond forming step) is the [2,3]-sigmatropic rearrangement of erated an ylide which cannot undergo the [2,3]-sigmatropic
the ylide. Consequently, if the rearrangement proceeds through rearrangement because of the lack of a double bond. It was observed
the ylide without the Rh(II)-catalyst attached, the outcome of the that under the same reaction conditions the diazo substrate 2a was
diastereoselectivity of this C–C bond forming step should be irrele- consumed completely to give unidentified products, indicating that
vant to the Rh(II) catalyst. As shown in Fig. 3a, the reaction of sulfide the Rh(II) catalyst is easily dissociated from the ylide (from I to II),
11 and diazoacetate 2a with different Rh(II) catalysts afforded the which further catalyses the dinitrogen extrusion of the diazo sub-
product 12 with essentially identical diastereoselectivities, thus sup- strates (Fig. 3c).
porting a free ylide mechanism. Furthermore, we tested the reaction
According to the guiding principles proposed by Davies on
of diazoester with symmetric diallyl sulfide 13 in the presence of a donor/acceptor Rh(II) carbene reactions50, an asymmetric induction
chiral Rh(II) catalyst (Fig. 3b). If the [2,3]-sigmatropic rearrange- model is proposed (Fig. 3d). Owing to the steric effect of the chiral
ment occurs with the chiral Rh(II) catalyst associated with the ligand, the lone pair of the electrons of sulfur (i) and (ii) attacks the
ylide, the reaction is expected to give the product 14 with some carbenic carbon from right. When the lone pair electrons (ii) attack
enantioselectivity; conversely, if the reaction proceeds through free the carbenic carbon (model A) the steric hindrance and electronic
sulfur ylide, there should be no enantioselectivity since the ylide repulsion between the trifluoromethyl group and the ligand makes
intermediate is achiral. The experiments were carried out with a this process less favourable than mode B, in which the lone pair elec-
combination of various diazoesters and a series of chiral Rh(II) trons (i) attack the carbenic carbon and thus the relatively flexible
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
5

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2789
ARTICLES
2. Ojima, I. Fluorine in Medicinal Chemistry and Chemical Biology
(Blackwell, 2009).
3. Cametti, M., Crousse, B., Metrangolo, P., Milani, R. & Resnati, G. The fluorous
effect in biomolecular applications. Chem. Soc. Rev. 41, 31–42 (2012).
4. Nakajima, T. Fluorine compounds as energy conversion materials. J. Fluorine
Chem. 149, 104–111 (2013).
5. Wang, J. et al. Fluorine in pharmaceutical industry: fluorine-containing drugs
introduced to the market in the last decade (2001–2011). Chem. Rev. 114,
2432–2506 (2014).
allyl substituent points to the ligand. As a result, the sulfonium ylide
with chiral sulfur centre of (R) configuration is formed, followed by
a concerted [2,3]-sigmatropic rearrangement to transfer the chirality
from sulfur to carbon48.
For the Cu(I)-catalysed reaction, similar mechanistic exper-
iments were carried out (for details, see Supplementary Section
5.4). The reaction of sulfide 11 and diazo substrate 9a catalysed by
Cu(I) catalysts with a series of chiral ligands afforded the products
with essentially identical diastereoselectivities. Moreover, the reac-
tion with symmetric sulfide 13 gave no enantioselectivity. These
results suggest that for the Cu(I)-catalysed Doyle–Kirmse reaction,
the [2,3] sigmatropic rearrangement also proceeds through free
6. Chu, L. & Qing, F.-L. Oxidative trifluoromethylation and
trifluoromethylthiolation reactions using (trifluoromethyl) trimethylsilane as a
nucleophilic CF3 source. Acc. Chem. Res. 47, 1513–1522 (2014).
7. Toulgoat, F., Alazet, S. & Billard, T. Direct trifluoromethylthiolation reactions:
the “renaissance” of an old concept. Eur. J. Org. Chem. 2415–2428 (2014).
8. Shao, X., Xu, C., Lu, L. & Shen, Q. Shelf-stable electrophilic reagents for
trifluoromethylthiolation. Acc. Chem. Res. 48, 1227–1236 (2015).
9. Shen, C. et al. Recent advances in C–S bond formation via C–H bond
functionalization and decarboxylation. Chem. Soc. Rev. 44, 291–314 (2015).
10. Xu, X.-H., Matsuzaki, K. & Shibata, N. Synthetic methods for compounds having
CF3–S units on carbon by trifluoromethylation, trifluoromethylthiolation,
triflylation, and related reactions. Chem. Rev. 115, 731–764 (2015).
11. Yang, X., Wu, T., Phipps, R. J. & Toste, F. D. Advances in catalytic
enantioselective fluorination, mono-, di-, and trifluoromethylation, and
trifluoromethylthiolation reactions. Chem. Rev. 115, 826–870 (2015).
12. Zhang, K., Xu, X. & Qing, F. Recent advances of direct trifluoromethylthiolation.
Chin. J. Org. Chem. 35, 556–569 (2015).
13. Chachignon, H. & Cahard, D. State-of-the-art in electrophilic
trifluoromethylthiolation reagents. Chin. J. Chem. 34, 445–454 (2016).
14. Bootwicha, T., Liu, X., Pluta, R., Atodiresei, I. & Rueping, M.
N-Trifluoromethylthiophthalimide: a stable electrophilic SCF3-reagent and its
application in the catalytic asymmetric trifluoromethylsulfenylation. Angew.
Chem. Int. Ed. 52, 12856–12859 (2013).
15. Rueping, M., Liu, X., Bootwicha, T., Pluta, R. & Merkens, C. Catalytic
enantioselective trifluoromethylthiolation of oxindoles using shelf-stable
N-(trifluoromethylthio) phthalimide and a cinchona alkaloid catalyst. Chem.
Commun. 50, 2508–2511 (2014).
16. Wang, X., Yang, T., Cheng, X. & Shen, Q. Enantioselective electrophilic
trifluoromethylthiolation of β-ketoesters: a case of reactivity and selectivity bias
for organocatalysis. Angew. Chem. Int. Ed. 52, 12860–12864 (2013).
17. Deng, Q.-H., Rettenmeier, C., Wadepohl, H. & Gade, L. H. Copper–boxmi
complexes as highly enantioselective catalysts for electrophilic
trifluoromethylthiolations. Chem. Eur. J. 20, 93–97 (2014).
18. Zhu, X.-L. et al. In situ generation of electrophilic trifluoromethylthio reagents
for enantioselective trifluoromethylthiolation of oxindoles. Org. Lett. 16,
2192–2195 (2014).
19. Liu, X., An, R., Zhang, X., Luo, J. & Zhao, X. Enantioselective
trifluoromethylthiolating lactonization catalyzed by an indane-based chiral
sulfide. Angew. Chem. Int. Ed. 55, 5846–5850 (2016).
20. Kirmse, W. & Kapps, M. Reaktionen des diazomethans mit diallylsulfid und
allyläthern unter kupfersalz-katalyse. Chem. Ber. 101, 994–1003 (1968).
21. Doyle, M. P., Griffin, J. H., Chinn, M. S. & van Leusen, D. Highly effective
catalytic methods for ylide generation from diazo compounds. Mechanism of
the rhodium-and copper-catalyzed reactions with allylic compounds. J. Org.
Chem. 46, 5094–5102 (1981).
22. Li, A.-H., Dai, L.-X. & Aggarwal, V. K. Asymmetric ylide reactions: epoxidation,
cyclopropanation, aziridination, olefination, and rearrangement. Chem. Rev. 97,
2341–2372 (1997).
23. Braverman, S. & Cherkinsky, M. [2,3]-Sigmatropic rearrangements of
propargylic and allenic systems. Top. Curr. Chem. 275, 67 (2007).
24. Reggelin, M. [2,3]-Sigmatropic rearrangements of allylic sulfur compounds. Top.
Curr. Chem. 275, 1–65 (2007).
25. Wang, J. in Comprehensive Organometallic Chemistry III Vol. 11 (eds Mingos,
D. M. P. & Crabtree, R. H.) 151–178. (Applications II: Transition Metal
Compounds In Organic Synthesis 2, Elsevier, 2007).
26. Zhang, Y. & Wang, J. Catalytic [2,3]-sigmatropic rearrangement of sulfur ylide
derived from metal carbene. Coord. Chem. Rev. 254, 941–953 (2010).
27. West, T. H., Spoehrle, S. S. M., Kasten, K., Taylor, J. E. & Smith, A. D. Catalytic
stereoselective [2,3]-rearrangement reactions. ACS Catal. 5, 7446–7479 (2015).
28. Nishibayashi, Y., Ohe, K. & Uemura, S. The first example of enantioselective
carbenoid addition to organochalcogen atoms: application to [2,3]-sigmatropic
rearrangement of allylic chalcogen ylides. Chem. Commun. 1245–1246 (1995).
29. Fukuda, T. & Katsuki, T. Co(III)-salen catalyzed carbenoid reaction:
stereoselective [2,3]-sigmatropic rearrangement of S-ylides derived from allyl
aryl sulfides. Tetrahedron Lett. 38, 3435–3438 (1997).
ylide, which is consistent with our previous studies34–37
.
Conclusion
In summary, we have developed a new approach toward the enan-
tioselective construction of the chiral C(sp3)–SCF3 bond through
Rh(II)- and Cu(I)-catalysed [2,3]-sigmatropic rearrangements of sul-
fonium ylides generated from metal carbenes and sulfides. These
reactions occur under mild conditions and exhibit wide substrate
compatibility. It is worth mentioning that this is the only successful
catalytic asymmetric Doyle–Kirmse reaction, with high yields and
enantioselectivities demonstrated over a wide range of aryldiazoace-
tates and vinyldiazoacetates. Exploring the mechanism of this
classic reaction in detail supports a pathway that involves the trans-
fer of chirality from sulfur of the chiral free ylide intermediate to the
carbon of the product in both Rh(II)- and Cu(I)-catalysed systems,
which will have further implications in the development of new
reactions that exploit this process. From the viewpoint of organic
synthesis, the reaction is a practical method toward accessing a
series of chiral building blocks containing quaternary stereocentres
bearing an SCF3 group, which are potentially useful drug
development candidates.
Methods
General procedure for Rh2(S-DOSP)4-catalysed [2,3]-sigmatropic
rearrangement. Aryldiazoacetates. Under a nitrogen atmosphere, pentane (2–4 ml),
allyl trifluoromethyl sulfide 1 (31 mg, 0.22 mmol, 1.1 equiv.) or propargyl
trifluoromethyl sulfide 6 (34 mg, 0.24 mmol, 1.2 equiv.) were successively added to a
dry 10 ml Schlenk reaction tube. To the solution was then added the aryldiazoacetate
2 (0.2 mmol, 1.0 equiv.), and the reaction tube was immersed in a −30 °C bath. After
5 min, a solution of Rh2(S-DOSP)4 (0.5 mol%) in 0.25 ml pentane was added
dropwise to the reaction tube. The reaction solution was stirred for 12 h. If the
typical colour of diazo compounds did not disappear, an additional solution of
Rh2(S-DOSP)4 (0.5 mol%) in 0.25 ml pentane was added. The reaction was
terminated when the colour of diazo compounds completely disappeared. The
solvent was removed with rotary evaporation under reduced pressure to leave a crude
mixture, which was purified by preparative thin layer chromatography to afford
pure product.
Vinyldiazoacetates. Under a nitrogen atmosphere, pentane (2–6 ml), allyl
trifluoromethyl sulfide 1 (28 mg, 0.20 mmol, 1.0 equiv.) or propargyl
trifluoromethyl sulfide 6 (28 mg, 0.20 mmol, 1.0 equiv.) were successively added to a
dry 10 ml Schlenk reaction tube. The vinyldiazoacetate 4 (0.24 mmol, 1.2 equiv.) for
1 and vinyldiazoacetate 4 (0.26 mmol, 1.3 equiv.) for 6 was added to the solution.
Then the reaction tube was immersed in a−30 °C bath. After 5 min, a solution of
Rh2(S-DOSP)4 (0.5 mol%) in 0.25 ml pentane was added dropwise to the reaction
tube. The reaction was stopped when the colour of diazo compound disappeared.
The solvent was removed with rotary evaporation under reduced pressure to leave a
crude mixture, which was purified by preparative thin layer chromatography to
afford pure product.
Data availability. Crystallographic data have been deposited at the Cambridge
Crystallographic Data Centre (CCDC) as CCDC 1502511 (8) and 1528864 (10j) and
can be obtained free of charge from the CCDC via http://www.ccdc.cam.ac.uk/
getstructures.
Received 5 November 2016; accepted 27 April 2017;
published online 5 June 2017
30. Itoh, K., Fukuda, T., Kitajima, H. & Katsuki, T. New aspect of carbenoid reaction:
exploitation of new asymmetric synthesis using chiral carbenoid species. Yuki
Gosei Kagaku Kyokaishi 55, 764–773 (1997).
References
1. Bégué, J.-P. & Bonnet-Delpon, D. Bioorganic and Medicinal Chemistry of
Fluorine (Wiley, 2008).
6
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.2789
ARTICLES
31. Fukuda, T., Irie, R. & Katsuki, T. Catalytic and asymmetric [2,3]-sigmatropic
rearrangement: Co(III)-salen catalyzed S-ylide formation from allyl aryl sulfides
and their rearrangement. Tetrahedron 55, 649–664 (1999).
32. McMillen, D. W., Varga, N., Reed, B. A. & King, C. Asymmetric copper-
catalyzed [2,3]-sigmatropic rearrangements of alkyl- and aryl-substituted allyl
sulfides. J. Org. Chem. 65, 2532–2536 (2000).
33. Kitagaki, S., Yanamoto, Y., Okubo, H., Nakajima, M. & Hashimoto, S.
Enantiocontrol in tandem allylic sulfonium ylide generation and [2,3]-
sigmatropic rearrangement catalyzed by chiral dirhodium(II) complexes.
Heterocycles 54, 623–628 (2001).
34. Zhang, X. et al. Catalytic asymmetric [2,3]-sigmatropic rearrangement of sulfur
ylides generated from copper(I) carbenoids and allyl sulfides. J. Org. Chem. 67,
5621–5625 (2002).
45. Li, Z., Parr, B. T. & Davies, H. M. L. Highly stereoselective C−C bond formation
by rhodium-catalyzed tandem ylide formation/[2,3]-sigmatropic rearrangement
between donor/acceptor carbenoids and chiral allylic alcohols. J. Am. Chem. Soc.
134, 10942–10946 (2012).
46. Li, Z. et al. Scope and mechanistic analysis of the enantioselective synthesis of
allenes by rhodium-catalyzed tandem ylide formation/[2,3]-sigmatropic
rearrangement between donor/acceptor carbenoids and propargylic alcohols.
J. Am. Chem. Soc. 134, 15497–15504 (2012).
47. Parr, B. T. & Davies, H. M. L. Highly stereoselective synthesis of cyclopentanes
bearing four stereocentres by a rhodium carbene-initiated domino sequence.
Nat. Commun. 5, 4455 (2014).
48. Trost, B. M. & Hammen, R. F. New synthetic methods. Transfer of chirality from
sulfur to carbon. J. Am. Chem. Soc. 95, 962–964 (1973).
35. Zhang, X., Ma, M. & Wang, J. Catalytic asymmetric [2,3]- sigmatropic
rearrangement of sulfur ylides generated from carbenoids and propargyl sulfides.
Tetrahedron: Asymm. 14, 891–895 (2003).
36. Zhang, X., Ma, M. & Wang, J. Catalytic asymmetric [2,3]-sigmatropic
rearrangement of sulfur ylides generated from carbenoids and allenic 2-
methylphenyl sulfide. Chin. J. Chem. 21, 878–882 (2003).
49. Trost, B. M. & Biddlecom, W. G. Asymmetric induction in a [2,3]-sigmatropic
rearrangement. J. Org. Chem. 38, 3438–3439 (1973).
50. Davies, H. M. L. & Morton, D. Guiding principles for site selective and
stereoselective intermolecular C–H functionalization by donor/acceptor
rhodium carbenes. Chem. Soc. Rev. 40, 1857–1869 (2011).
37. Ma, M., Peng, L., Li, C., Zhang, X. & Wang, J. Highly stereoselective [2,3]-
sigmatropic rearrangement of sulfur ylide generated through Cu(I) carbene and
sulfides. J. Am. Chem. Soc. 127, 15016–15017 (2005).
38. Liao, M. & Wang, J. Highly efficient [2,3]-sigmatropic rearrangement of
sulfur ylide derived from Rh(II) carbene and sulfides in water. Green Chem. 9,
184–188 (2007).
39. Zhang, H., Wang, B., Yi, H., Zhang, Y. & Wang, J. Rh (II)-catalyzed [2, 3]-
sigmatropic rearrangement of sulfur ylides derived from cyclopropenes and
sulfides. Org. Lett. 17, 3322–3325 (2015).
40. Tyagi, V., Sreenilayam, G., Bajaj, P., Tinoco, A. & Fasan, R. Biocatalytic synthesis
of allylic and allenyl sulfides through a myoglobin-catalyzed Doyle–Kirmse
reaction. Angew. Chem. Int. Ed. 55, 13562–13566 (2016).
Acknowledgements
The authors acknowledge financial support from the National Basic Research Program of
China (973 programme no. 2015CB856600) and Natural Science Foundation of China
(grant no. 21332002). We thank Y. Wang and Z. Yu (Peking University) for the discussion
on the reaction mechanism. We greatly appreciate W.-X. Zhang and N. Wang (Peking
University) for the assistance in obtaining X-ray crystallographic structures.
Author contributions
Z.Z., Z.S., W.Y., G.W. and R.Z. performed the experiments. W.-D.C. and Y.Z. participated
in the discussion and helped the measurement of optical purities. J.W. conceived and
supervised the project. Z.Z. and J.W. wrote the manuscript.
41. Doyle, M. P. & Forbes, D. C. Recent advances in asymmetric catalytic metal
carbene transformations. Chem. Rev. 98, 911–935 (1998).
42. Davies, H. M. L. & Beckwith, R. E. J. Catalytic enantioselective C-H activation
by means of metal-carbenoid-induced C-H insertion. Chem. Rev. 103,
2861–2903 (2003).
43. Davies, H. M. L. & Nikolai, J. Catalytic and enantioselective allylic C–H
activation with donor–acceptor-substituted carbenoids. Org. Biomol. Chem. 3,
4176–4187 (2005).
Additional information
Supplementary information and chemical compound information are available in the
online version of the paper. Reprints and permissions information is available online at
www.nature.com/reprints. Publisher’s note: Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations. Correspondence and
requests for materials should be addressed to J.W.
44. Aggarwal, V. K., Ferrara, M. & Hainz, R. [2,3]-Sigmatropic rearrangement of
allylic sulfur ylides derived from trimethylsilyldiazomethane (TMSD).
Tetrahedron Lett. 40, 8923–8927 (1999).
Competing financial interests
The authors declare no competing financial interests.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
7
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.