Controllable catalytic difluorocarbene transfer enables access to diversified fluoroalkylated arenes

Article abstract of DOI:10.1038/s41557-019-0331-9

Difluorocarbene has important applications in pharmaceuticals, agrochemicals and materials, but all these applications proceed using just a few types of reaction by taking advantage of its intrinsic electrophilicity. Here, we report a palladium-catalysed strategy that confers the formed palladium difluorocarbene (Pd=CF₂) species with both nucleophilicity and electrophilicity by switching the valence state of the palladium centre (Pd(0) and Pd(ii), respectively). Controllable catalytic difluorocarbene transfer occurs between readily available arylboronic acids and the difluorocarbene precursor diethyl bromodifluoromethylphosphonate (BrCF₂PO(OEt)₂). From just this simple fluorine source, difluorocarbene transfer enables access to four types of product: difluoromethylated and tetrafluoroethylated arenes and their corresponding fluoroalkylated ketones. The transfer can also be applied to the modification of pharmaceuticals and agrochemicals as well as the one-pot diversified synthesis of fluorinated compounds. Mechanistic and computational studies consistently reveal that competition between nucleophilic and electrophilic palladium difluorocarbene ([Pd]=CF₂) is the key factor controlling the catalytic difluorocarbene transfer.

Full text of DOI:10.1038/s41557-019-0331-9

Articles
https://doi.org/10.1038/s41557-019-0331-9
Controllable catalytic difluorocarbene transfer
enables access to diversified fluoroalkylated arenes
Xia-Ping Fu1,5, Xiao-Song Xue2,3,5, Xue-Ying Zhang1, Yu-Lan Xiao1, Shu Zhang4, Yin-Long Guo1,
Xuebing Leng1, Kendall N. Houkꢀ ꢀ2* and Xingang Zhangꢀ ꢀ1*
Difluorocarbene has important applications in pharmaceuticals, agrochemicals and materials, but all these applications pro-
ceed using just a few types of reaction by taking advantage of its intrinsic electrophilicity. Here, we report a palladium-catalysed
strategy that confers the formed palladium difluorocarbene (Pd=CF2) species with both nucleophilicity and electrophilicity by
switching the valence state of the palladium centre (Pd(0) and Pd(ii), respectively). Controllable catalytic difluorocarbene
transfer occurs between readily available arylboronic acids and the difluorocarbene precursor diethyl bromodifluorometh-
ylphosphonate (BrCF2PO(OEt)2). From just this simple fluorine source, difluorocarbene transfer enables access to four types
of product: difluoromethylated and tetrafluoroethylated arenes and their corresponding fluoroalkylated ketones. The transfer
can also be applied to the modification of pharmaceuticals and agrochemicals as well as the one-pot diversified synthesis of
fluorinated compounds. Mechanistic and computational studies consistently reveal that competition between nucleophilic and
electrophilic palladium difluorocarbene ([Pd]=CF2) is the key factor controlling the catalytic difluorocarbene transfer.
ifluorocarbene (:CF2) is an electrophilic ground-state sin- reactions. We speculated that the reactivity of [M]=CF2 could be
glet carbene1 that has important applications in various modulated by altering the oxidation state of the central transition
D
fields2,3. In industry, difluorocarbene undergoes homocou- metal. Because palladium is a normal transition metal for the catal-
pling to synthesize tetrafluoroethene for the production of Teflon ysis process, and the transition between Pd(0) and Pd(ii) is com-
(DuPont), a widely used polymer (Fig. 1a)4. In drug production mon27, the reactivity of [Pd]=CF2 might be modulated between Pd0
and medicinal chemistry, difluorocarbene serves to construct dif- (with d10) and PdII (with d8). [Pd0]=CF2 (with d10) is expected to pos-
ferent fluorinated moieties (Fig. 1b–e), such as difluoromethyl sess unusual nucleophilic reactivity as a result of electron donation
ethers5,6, gem-difluorocyclopropanes7,8 and gem-difluoroalkenes9. from the more electron-rich Pd0, while [PdII]=CF2 (with d8) should
Difluorocarbene can also be used for the preparation of diverse maintain the electrophilic behaviour of difluorocarbene due to the
18F-labelled trifluoromethylation compounds as radioisotopes for relative electron deficiency of PdII (Fig. 1f). In this way, [Pd]=CF2
positron emission tomography studies (Fig. 1e)10–13. Despite the can present either electrophilicity or nucleophilicity in a control-
range of applications, difluorocarbene has limited reaction types lable manner, acting as catalytic species in a catalytic cycle.
(Fig. 1a–d)2–14, which mainly take advantage of its electrophilic
We sought to control the reactivity of difluorocarbene to be
nature1,2. To overcome this limitation, there have been attempts nucleophilic or electrophilic, and to diversify the difluorocarbene
over 40 years to tune the reactivity of difluorocarbene by coordina- transfer reactions to produce a variety of different compounds.
tion to transition metals15. Although various metal difluorocarbene The proposed mechanism comprises the following steps (Fig. 1f):
([M]=CF2) complexes based on groups 6 to 10 have been synthe- (1) a nucleophilic [Pd0]=CF2 species reacts with an electrophile (E)
sized and isolated, their inherently low reactivity compared to their to afford the key fluoroalkyl palladium intermediate [PdII]–CF2E;
non-fluorinated counterparts has hindered further transformation (2) electrophilic difluorocarbene(s) are inserted into [PdII]–CF2E
and posed challenges for catalysis cycling16–20. Only recently has to elongate the fluoroalkylated chain28. From these steps, it is pos-
catalytic difluorocarbene transfer been realized to install a difluo- sible to provide an array of products containing different units of
romethyl (CF2H) group onto aromatic rings21–23. However, these difluoromethylene (CF2) by varying the concentration of difluo-
methods do not involve controllable manipulation of the reactiv- rocarbene in the presence of the nucleophilic Pd carbenoid. These
ity of metal difluorocarbene between nucleophilicity and electro- catalytic steps enable a straightforward strategy to install CF2 units
philicity, and can only access one type of fluorinated compound. into organic molecules in a controllable manner, which has not been
Controllable catalytic difluorocarbene transfer remains an elusive achieved with other methods. The challenge of this strategy is to
goal and has not yet been reported, mainly due to the lack of mech- selectively produce the fluoroalkylated compound at each step as
anistic insight into the metal difluorocarbene chemistry.
a major product. We now report the achievement of controllable
A few examples have demonstrated that [M]=CF2 species of palladium-catalysed difluorocarbene transfer reactions between
groups 8 and 9 with formal d8 electron configurations (Ru0, Os0, arylboronic acids and a readily available difluorocarbene precursor
IrI, CoI) exhibit nucleophilic character16,17,24–26, providing the oppor- BrCF2PO(OEt)2 (ref. and Fig. 1f). Manipulation of the oxidation
29
tunity to develop a new paradigm for difluorocarbene transfer states of palladium leads to two different types of [Pd]=CF2 species
1Key Laboratory of Organofluorine Chemistry, Center for Excellence in Molecular Synthesis, Shanghai Institute of Organic Chemistry, University of Chinese
Academy of Sciences, Chinese Academy of Sciences, Shanghai, China. 2Department of Chemistry and Biochemistry, University of California, Los Angeles,
CA, USA. 3State Key Laboratory of Elemento-Organic Chemistry, College of Chemistry, Nankai University, Tianjin, China. 4School of Materials and Energy,
University of Electronic Science and Technology of China, Chengdu, Sichuan, China. 5These authors contributed equally: Xia-Ping Fu, Xiao-Song Xue.
*e-mail: houk@chem.ucla.edu; xgzhang@sioc.ac.cn
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
948

NATurE CHEMIsTry
Articles
a
b
d
e
O
Cl
O
N
S
O
F
F
F
CF2
O
CF2
Nu
NuCF2H
N
H
F
Cl
N
HF2CO
N
Nu = O, S, N, P, C nucleophiles
Roflumilast
NH
O
N
c
CF218F
NH
O
F
F
CF2
O
CF2
F
F
Me
F
N
H
F
Inducible T-Cell Kinase (ITK)
Inhibitor
Fluoxetin
f
Reaction with electrophile
Difluorocarbene insertion
Carbonylation
[Pd0]
CF2
Nu
E
F
F
[Pd0]=CF2
[PdII]–CF2E
[PdII]–(CF2)nCF2E
NuCO(CF2)nCF2E
[PdII]=CF2
H O
2
+
CO
Nu
Nu
Nu
NuCF2E
NuCF2CF2E
Nu(CF2)nCF2E
n = 0, 1, 2...
[Pd] catalyst
H2O
[Pd0]=CF2
[PdII]=CF2
ArCF2H
ArB(OH)2
+
ArCOCF2H
Nucleophilic
[Pd] catalyst
via [Pd]=CF2
via [Pd]=CF2
ArCF2CF2H
BrCF2PO(OEt)2
ArCOCF2CF2H
Electrophilic
CF2
Fig. 1 | Previous and present approaches to difluorocarbene transfer. a, Homocoupling of difluorocarbene is a typical reaction to generate tetrafluoroethene
for the production of Teflon. b, Reactions of difluorocarbene with heteroatom or carbon nucleophiles are commonly used to prepare difluoromethylated
compounds, but those with carbon nucleophiles have limited substrate scope. c, [2+1] cyclizations of unsaturated carbon–carbon bonds with
difluorocarbene. d, Wittig olefinations to prepare gem-difluoroalkenes. e, Examples of bioactive molecules prepared by using difluorocarbene transfer
approaches to introduce fluorinated moieties. f, Outline of the present strategy for controllable catalytic difluorocarbene transfer. Manipulation of the
oxidation states of palladium leads to two different types of [Pd]=CF2 species (with nucleophilicity or electrophilicity), which can make controllable
catalytic difluorocarbene transfer feasible and produce a variety of different fluorinated compounds.
with opposite reactivity (nucleophilicity or electrophilicity). In the a proton was proposed to be the electrophile. Unexpectedly, besides
study, we also discovered that [PdII]=CF2 can serve as a CO sur- the formation of difluoromethylated arene 4 and tetrafluoroethyl-
rogate in the presence of H2O (ref. 16), leading to a carbonylation ated arene 5, two fluoroacylated compounds, difluoromethylaryl
reaction that terminates the difluorocarbene insertion and provides ketone 6 and tetrafluoroethylaryl ketone 7, were also observed. The
fluoroacylated compounds (Fig. 1f). These transformations selec- CO probably originates from the difluorocarbene moiety, implying
tively afford four typical types of difluorocarbene transfer prod- that an electrophilic [PdII]=CF2 is involved in the reaction16. The
uct: difluoromethylated and tetrafluoroethylated arenes and their formation of these fluoroalkylated ketones is intriguing, because
corresponding fluoroalkylated ketones. By deliberately tuning the no transition-metal-catalysed perfluoroalkylation–carbonylation
reaction conditions, it is possible to manipulate the reaction to reaction has been reported previously due to the strong σ bond
selectively produce each of the four products.
between the transition metal and the perfluoroalkyl group in M–Rf
(Rf =perfluoroalkyl)30. Additionally, as a versatile functional group,
the carbonyl group offers great opportunities for downstream trans-
results and discussion
To investigate the advances of such controllable difluorocar- formations.
bene transfer reactions, we focused on transition-metal catalysis
Subtle adjustment of the reaction conditions allows the selective
with a suitable difluorocarbene precursor. After extensive efforts, production of one of these four types of compound, as shown in
BrCF2PO(OEt)2 1 was identified as a suitable difluorocarbene pre- Table 1. The reaction temperature and the concentration of 1 are
cursor in combination with hydroquinone 2 and a Brønsted base critical factors. Overall, a lower reaction temperature is favourable
(K2CO3); the concentration of difluorocarbene in the reaction can for the formation of carbonylation products 6 and 7, while a higher
be controlled by adjusting concentrations of 1, 2 and the base. We concentration of the difluorocarbene precursor 1 benefits CF2 elon-
investigated a model reaction, using arylboronic acid 3 as a nucleo- gation to produce 5 and 7. In this study, the selective formation
phile to react with 1 in the presence of 2, K2CO3, H2O, palladium of difluoromethyl arylketone 6 was the most difficult to achieve
catalyst PdCl2(PPh3)2 and a phosphine ligand (Xantphos), in which among the competitive reactions; another CO source, tungsten
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
949

NATurE CHEMIsTry
Articles
Table 1 | Optimization of the selective formation of fluoroalkylated arenes in a model reaction.
CF2H
CF2CF2H
BrCF2PO(OEt)2
Pd(PPh3)Cl2 (7.5 mol%)
Xantphos (y mol%)
O
Ph
Ph
1 (x equiv.)
PPh2
PPh2
OH
4
6
5
Hydroquinone (z equiv.)
Xantphos
+
+
H2O (w equiv.)
K2CO3 (4.0 equiv.)
1,4-dioxane, temp.
B(OH)2
COCF2H
COCF2CF2H
HO
Ph
Ph
Ph
Hydroquinone
7
3 (1.0 equiv.)
Conditionsa
4 (%)b
5 (%)b
6 (%)b
7 (%)b
Conditions A: xꢀ=ꢀ2.0, yꢀ=ꢀ11.5, zꢀ=ꢀ2.0,
wꢀ=ꢀ1.0, 100ꢀ°C
82 (80)
ND
2
6
Conditions B: xꢀ=ꢀ3.0, yꢀ=ꢀ20, zꢀ=ꢀ3.0, wꢀ=ꢀ0, 16
64 (64)
ND
20
4ꢀÅ MS, 120ꢀ°C
Conditions C: xꢀ=ꢀ2.0, yꢀ=ꢀ11.5, zꢀ=ꢀ1.0, wꢀ=ꢀ0, 15
ND
3
56 (54)
3
ND
80ꢀ°C, W(CO)6 (25ꢀmol%)
Conditions D: xꢀ=ꢀ4.0, yꢀ=ꢀ20, zꢀ=ꢀ3.0, wꢀ=ꢀ1.0, 9
73 (70)
80ꢀ°C
aThe reaction was carried out with 3 (0.3ꢀmmol, 1.0ꢀequiv.) and 1 in 1,4-dioxane (conditions A, 3.0ꢀml; conditions B, 4.0ꢀml; conditions C, 2.5ꢀml; conditions D, 3.0ꢀml) under suitable reaction conditions
(conditions A–D). b The yields were determined by 19F NMR using fluorobenzene as the internal standard; numbers given in parentheses are isolated yields. ND, not determined. MS, molecular sieves.
hexacarbonyl (W(CO)6) was added to the reaction to selectively
To demonstrate the utility of these catalytic controllable difluo-
furnish 6 with 56% yield, but a 15% yield of 4 was still obtained rocarbene transfer reactions, the direct fluoroalkylation of N-Boc-
(Table 1, conditions C). Control experiments showed that the palla- mexiletine, an anti-arrhythmic drug, was conducted by sequential
dium catalyst, Xantphos ligand and hydroquinone all play essential C–H bond boroylation38 and fluoroalkylation (Fig. 2a). Three mexi-
roles in promoting the reaction (Supplementary Table 5).
letine analogues (52–54) were selectively obtained in good yields,
With the optimized reaction conditions in hand, we exam- providing a straightforward and versatile route for applications in
ined the scope of these transformations (Table 2). A wide range of medicinal chemistry without parallel de novo synthesis. In addition,
difluoromethylated arenes can be selectively obtained under con- three diversified fluoroalkylated products can be accessed from one
ditions A with good yields31–34. Many functionalities (for example, simple starting material in a one-pot reaction, as demonstrated
thioether, trimethylsilyl (TMS), esters and ketones) showed good in the modification of herbicide nitrofen (Fig. 2b). This one-pot,
tolerance to the reaction (12–18). In the case of substrates bearing diversity-oriented synthetic strategy provides a facile and step-eco-
electron-withdrawing groups, the addition of 1.0 equiv. of LiCl ben- nomic route for applications in the discovery of different interest-
efitted the reaction (15–18). Good yields were also obtained under ing new bioactive molecules, which would otherwise be difficult to
conditions B for the selective formation of tetrafluoroethylated realize by conventional methods.
arenes; even the alkene-containing substrate reacts without forma-
To gain mechanistic insight into these processes, we conducted
tion of gem-difluorocyclopropane side product (30)7. Although the several stoichiometric experiments (Fig. 3). The preparation and
selective formation of difluoromethyl ketones appears relatively isolation of unstable [Pd0]=CF2 monomer complexes is difficult and
difficult, moderate yields were still obtained under conditions C. has not yet been reported. In our study, complexes (tBu-Xantphos)
In contrast, a variety of arylboronic acids reacted smoothly with Pd0=CF2 (A1-1) and [4,7-di-tBu(tBu-Xantphos)]Pd0=CF2 (A1-2)
1 under conditions D to selectively produce tetrafluoroethylaryl were prepared in good yields (70%) by defluorination from their
ketones. These reactions are readily scalable, as demonstrated corresponding trifluoromethyl palladium(ii) salts (PdII−CF3) (B1-1
by the gram-scale synthesis of fluoroalkylated arenes 14 and 42. and B1-2, respectively) with potassium graphite (KC8) (Fig. 3a)18.
The promising selective formation of these fluorinated products Here, the use of bulkier Bu-Xantphos and 4,7-di-tBu(tBu-Xant-
t
encourages evaluation of these controllable difluorocarbene trans- phos) ligands sterically stabilizes the vulnerable [Pd0]=CF2. The
fer reactions with complex molecules. Commercially available 13C NMR spectra of complexes A1-1 and A1-2 showed a charac-
drugs, such as clofibrate, ezetimibe and δ-vitamin E, were viable teristic positive chemical shift for metal difluorocarbene carbon
in the reactions from their corresponding arylboronic acid deriva- ([Pd0]=CF2, A1-1: δ 231.5ppm, triplet of triplet, JCF =519.9Hz,
1
tives (23, 24, 31, 48–50). In particular, the selective difluorometh- 2JCP =68.0Hz; A1-2: δ 231.0ppm, triplet of triplet, JCF =519.5Hz,
1
ylation and tetrafluoropropanoylation of ezetimibe (24 and 49) 2JCP =67.6Hz). A single-crystal X-ray diffraction study of complex
with the free hydroxyl group intact provides a convenient synthetic A1-2 further confirmed that the [Pd0]=CF2 complexes involved in
method for modulation of bioactive molecules without the need the reaction are monomers. A1-2 has a Pd–C(carbene) bond length
for a multistep protection–deprotection procedure. CF2H is a bio- of 1.78(2)Å, ~0.23Å shorter than the [PdII]–CF2H bond in difluo-
isostere of hydroxyl and thiol groups35 and can serve as a lipophilic romethyl complex C1 (Fig. 3b). The sum of the bond angles around
hydrogen bond donor36; the 1,1,2,2-tetrafluoroethyl (CF2CF2H) the carbene carbon atom in A1-2 is very close to 360° (≤359.8°),
group is a structural mimic of the trifluoromethyl group and can featuring the sp2 carbon in the [Pd0]=CF2 structure. Complexes A1
improve the lipophilicity of the molecules37. Consequently, these show unique nucleophilicity and are readily protonated with aryl-
transformations provide useful tools for modification of pharma- boronic acid or H2O at room temperature to produce complexes
ceuticals and other bioactive molecules.
[PdII]–CF2H (C1) with full conversion (Fig. 3b and Supplementary
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
950

NATurE CHEMIsTry
Articles
Table 2 | Substrate scope of the controllable palladium-catalysed difluorocarbene transfer.
CF2H
CF2CF2H
OH
R
R
Pd(PPh3)Cl2 (7.5 mol%)
Xantphos (y mol%)
R
R
BrCF2PO(OEt)2
(x equiv.)
B(OH)2
hydroquinone (z equiv.)
R
O
or
+
PPh2
PPh2
H2O (w equiv.)
K2CO3 (4.0 equiv.)
dioxane, temp.
OH
COCF2H
COCF2CF2H
(1.0 equiv.)
CF2 and CO
R = H, Xantphos
R = tBu, 4,7-di-tBu-Xantphos
R
R
Hydroquinone
Conditions Aa
CF2H
CF2H
CF2H
CF2H
MeO
CF2H
CF2H
O
O
Ph
tBu
PhO
EtS
OMe
4, 80%
8, 80%
9, 60%
10, 55%e
11, 70%
12, 70%e
CF2H
TMS
CF2H
CF2H
EtO2C
CF2H
CF2H
MeOC
CF2H
CF2H
TMS
EtO2C
MeOC
14, 73%, 60%k
15, 70%f
16, 70%f
17, 65%f
18, 60%f
13, 76%
19, 63%
CF2H
F
CF2H
CF2H
Me
Me
CF2H
O
O
O
N
F
CF2H
OEt
HO
O
20, 74%e
21, 65%
22, 75%
23, 50%g
24, 65%g
(from ezetimibe, antiatherosclerosics)
(from clofibrate,
against cardiovascular disease)
Conditions Bb
CF2CF2H
CF2CF2H
CF2CF2H
CF2CF2H
CF2CF2H
CF2CF2H
O
O
BzO
Ph
R
TMS
29, 65%
28, 50%h
5, 64%
25, R = tBu, 74%
26, R = EtS, 50%
27, 63%
30, 61%
Conditions Cc
CF2CF2H
COCF2H
Me
O
Me
O
COCF2H
COCF2H
COCF2H
OEt
Ph
PhO
EtS
31, 70%
(from clofibrate,
against cardiovascular disease)
32, 55%i
33, 60%i
6, 54%
34, 41%
Conditions Dd
COCF2CF2H
COCF2CF2H
COCF2CF2H
MeO
COCF2CF2H
COCF2CF2H
O
O
Ph
R
EtS
OMe
39, 65%
7, p-Ph, 70%
35, m-Ph, 73%
36, R = tBu, 60%
37, R = PhO, 67%
38, 50%
40, 62%
COCF2CF2H
COCF2CF2H
COCF2CF2H
COCF2CF2H
TMS
EtO2C
S
COCF2CF2H
47, 62%
43, p-CO2Et, 61%j
44, m-CO2Et, 61%j
41, p-TMS, 60%
42, m-TMS, 64%, 55%k
45, 56%
46, 52%
COCF2CF2H
Me
O
COCF2CF2H
F
COCF2CF2H
Me
Me
O
Me
Me
Me
O
OEt
N
O
F
Me
HO
Me
48, 50%
(from clofibrate,
against cardiovascular disease)
49, 70%
(from ezetimibe, antiatherosclerosics)
50, 50%
δ
(from -vitamin E, antioxidants)
Isolated yields are shown. aConditions A: xꢀ=ꢀ2.0, yꢀ=ꢀ11.5, zꢀ=ꢀ2.0, wꢀ=ꢀ1.0, 100ꢀ°C. bConditions B: xꢀ=ꢀ3.0, yꢀ=ꢀ20, zꢀ=ꢀ3.0, wꢀ=ꢀ0, 4ꢀÅ MS, 120ꢀ°C. cConditions C: xꢀ=ꢀ2.0, yꢀ=ꢀ11.5, zꢀ=ꢀ1.0, wꢀ=ꢀ0, 80ꢀ°C, W(CO)6
(25ꢀmol%). dConditions D: xꢀ=ꢀ4.0, yꢀ=ꢀ20, zꢀ=ꢀ3.0, wꢀ=ꢀ1.0, 80ꢀ°C. eReaction run at 80ꢀ°C. fUsing 1.0ꢀequiv. of LiCl. gUsing 2.0ꢀequiv. of K2CO3. hUsing 2.5ꢀmol% Pd2(dba)3. i4,7-di-tBu-Xantphos was used. jUsing
20ꢀmol% 4,7-di-tBu-Xantphos and 70ꢀmg of 4ꢀÅ MS without addition of H2O. kGram-scale synthesis.
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
951

NATurE CHEMIsTry
Articles
a
H
B(OH)2
CF2H
CF2CF2H
COCF2CF2H
(i) [Ir(cod)OMe]2 (0.5 mol%)
4,4'-di-tert-butylbipyridine (1 mol%)
B2pin2, THF, 80 °C
BrCF2PO(OEt)2
Conditions
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
(ii) NaIO4, NH4OAc
acetone/H2O, r.t.
O
O
O
O
O
BocHN
Me
BocHN
Me
BocHN
Me
BocHN
Me
BocHN
Me
N-Boc-mexiletine
51, 65%
Conditions A
Conditions B
Conditions D
52, 60%
53, 75%
54, 58%
b
B(OH)2
CF2H
CF2CF2H
COCF2CF2H
Pd(DPEphos)Cl2 (7.5 mol%)
Xantphos (20 mol%)
Hydroquinone (3.0 equiv.)
BrCF2PO(OEt)2
+
+
+
Cl
Cl
Cl
Cl
4 Å MS, K2CO3 (4.0 equiv.)
1,4-dioxane, 120 °C
O
O
O
O
Cl
Cl
Cl
Cl
55
1
56, 25%
57, 28%
58, 24%
(5.0 mmol, 1.0 equiv.)
from nitrofen
(3.0 equiv.)
Fig. 2 | Diversified synthesis of fluoroalkylated arenes. a, Selective C–H bond borylation and fluoroalkylation of pharmaceuticals. Mexiletine, an anti-
arrhythmic drug, can be readily modified to selectively provide three mexiletine analogues through the presented late-stage strategy. b, One-pot diversity-
oriented modification of agrochemicals. Nitrofen, a herbicide, can be readily modified from one simple starting material to provide three diversified
fluoroalkylated nitrofen analogues. Isolated yields are shown. B2pin2, 4,4,4′,4′,5,5,5,5′-octamethyl-2,2′-bi(1,3,2-dioxaborolane). DPEphos, (oxybis(2,1-
phenylene))bis(diphenylphosphane).
Fig. 1). So far, metal difluorocarbene protonated by H2O has not A′ overcomes a small barrier (6.8kcalmol−1) via TSA′-E to afford
been reported16,17,24–26. This extreme nucleophilicity seems unique to the much more stable trans-[(Xantphos)PdII(CF2CF2H)(OH)] (E),
the [Pd0]=CF2 monomer, because the previously isolated palladium which is −41.2kcalmol−1 downhill from A′, providing the driving
difluorocarbene trimer cannot react with H2O (ref. 23).
force for the CF2 elongation. This CF2 elongation shows a lower
Density functional theory (DFT) calculations shed light on the barrier than that for the transmetallation of arylboronic acid with
nucleophilicity of [Pd0]=CF2 (Fig. 4a). As shown in the calculated difluoromethyl palladium complex C (TSA′-E versus TStrans-C-cis-D1
,
charge distributions (Fig. 4b), the electron density on the carbon of Fig. 4c), thus enabling the difluorocarbene insertion to occur before
difluorocarbene increases substantially from the free difluorocar- transmetallation.
bene to complex A (0.07atomic units (a.u.) versus −0.03a.u.). This
We also investigated the mechanism of the carbonylation reac-
is possibly a result of electron donation from the occupied d orbital tion. We found that the [PdII]=CF2 (with d8) does exhibit elec-
of palladium to the vacant p orbital of difluorocarbene39, which is trophilicity and can generate CO easily when reacting with H2O
responsible for the property reversal of the difluorocarbene car- (Supplementary Fig. 5), suggesting that the CO to form tetrafluo-
bon from electrophilicity to nucleophilicity, and the carbene car- roethylaryl ketones originates from the [PdII]=CF2 species. This
bon can be attacked by a proton or other electrophiles. Indeed, the deduction was further supported by 18O-labelling experiment
protonation of [Pd0]=CF2 at the carbon site by arylboronic acid, via (Fig. 3d and Supplementary Fig. 6) and DFT studies (Supplementary
a four-membered ring transition state (TSA-C′), is ∼20kcalmol−1 Fig. 10). Control experiments (Fig. 3e and Supplementary Fig. 7)
more favourable over protonation at the palladium centre (TSA-C′′′
)
and DFT calculations further identified that the carbonylation reac-
(Fig. 4a), ruling out the pathway where a proton attacks at the metal tion occurs after the transmetallation of PdII fluoroalkyl species (E),
centre, as observed with other nucleophilic M=CF2 (refs. 16,24–26). and the CO insertion into the Pd–Ar σ bond is preferred over its
The barriers for the protonation of [Pd0]=CF2 by hydroquinone insertion into the Pd–CF2CF2H σ bond by 22.6kcalmol−1 (Fig. 4d).
or water are higher than that by phenylboronic acid. The predic-
Collectively, the experimental observations and DFT cal-
tion that the more strongly acidic arylboronic acid serves as a culations are consistent with the aforementioned hypothesis
more favourable electrophile confirms the nature of the reaction. (Fig. 1f). Arylboronic acid, H2O or hydroquinone function as
The more weakly acidic water and hydroquinone can also serve as electrophiles for the protonation of nucleophilic [Pd0]=CF2 at
proton donors under the current reaction conditions, as shown in the carbene carbon, while H2O serves as a nucleophile for the
Supplementary Fig. 2. The catalytic reactions of arylboronic acid hydrolysis of electrophilic [PdII]=CF2 to generate CO. Because
3 with BrCF2PO(OEt)2 can occur in the presence of deuterated difluorocarbene and CO insertions are more favourable than
hydroquinone or water, suggesting the highly nucleophilic activity the transmetallation of [(Xantphos)PdII(CF2H)(OH)] (C) (TSA′-E
of [Pd0]=CF2.
versus TStrans-C-cis-D1) (Fig. 4c) and reductive elimination of [(Ar)
The introduction of multiple CF2 units was confirmed by the (Xantphos)PdII(CF2CF2H)] (G) (TSG1-H1 versus TSG1-P61) (Fig. 4e),
reaction of the trans-[(Xantphos)PdII(CF2H)(Cl)] complex C2 with respectively, it makes the selective preparation of fluorinated
40
difluorocarbene precursor TMSCF2Br (ref. and Fig. 3c). On the compounds feasible. An outline of a possible mechanism for
basis of the DFT calculation (Fig. 4c), we assumed that [(HCF2) this controllable catalytic difluorocarbene transfer is provided in
(Xantphos)PdII=CF2(OH)] complex A′ is formed by the com- Fig. 5. The reaction begins with the formation of a [(Xantphos)
plexation of trans-[(Xantphos)PdII(CF2H)(OH)] (trans-C) with Pd0=CF2] species A between Pd0 and difluorocarbene, which
a difluorocarbene ligand41, which is a thermodynamically favour- is generated from the reaction of BrCF2PO(OEt)2 with hydro-
able process (ΔG=−9.5kcalmol−1 without barrier). Subsequently, quinone in the presence of a base (Supplementary Fig. 9).
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
952

NATurE CHEMIsTry
Articles
a
R
R
R
R
CF3CO2
R
R
KC8 (2.2 equiv.)
THF, 0 °C, 3 h
O
O
O
tBu2P
PtBu2
tBu2P
PtBu2
PtBu2
PtBu2
Pd
Pd
CF3
CF2
R = H, tBu-Xantphos
R = tBu, 4,7-di-tBu(tBu-Xantphos)
B1-1, R = H (1.0 equiv.)
B1-2, R = tBu (1.0 equiv.)
A1-1, R = H, 70%
A1-2, R = tBu, 70%
A1-2
b
X–
R
R
R
R
1,4-dioxane
O
O
H2O
or PhB(OH)2
+
tBu
2P
tBu2P
PtBu2
PtBu2
r.t.
Pd
Pd
(2.0 equiv.)
CF2H
CF2
A1-1 (1.0 equiv.)
A1-2 (1.0 equiv.)
C1, full conversion
R = H, tBu
C1
c
CF2
KOAc, THF
r.t., 10 min
O
O
+
Cl
Cl
Ph2P
PPh2
Pd
Ph2P
PPh2
CF2CF2H
Pd
TMSCF2Br (10 equiv.)
CF2H
C2
(1.0 equiv.)
E1, 65%
Determined by 19F NMR
E1
d
e
B(OH)2
O(18O)
CF2CF2H
O
Conditions D
CO (1 atm.)
Ph
O
CF2CF2H
H218O (x equiv.)
1,4-dioxane, r.t.
Ph2P
PPh2
Pd
+
Ph
BrCF2PO(OEt)2
7
Ph
CF2CF2H
59, quantitative yield
x = 2.0 equiv., 74%, 18O/16O = 7.62/100
x = 4.0 equiv., 32%, 18O/16O = 14.97/100
cis-G1
Fig. 3 | mechanistic studies. a, Preparation of [Pd0]=CF2 complexes a1 and single-crystal X-ray structure of [Pd0]=CF2 complex A1-2. b, Protonation
of complex a1 with H2O or phenylboronic acid and single-crystal X-ray structure of difluoromethyl–palladium complex C1. The anions are not shown.
c, Investigation of CF2 elongation and the single-crystal X-ray structure of tetrafluoroethyl palladium complex E1. d, 18O-labelling of compound 7. A twofold
loading amount of H218O almost doubled the ratio of 18O-labelled 7, indicating that the CO to form the tetrafluoroethyl aryl ketones is probably derived
from the hydrolysis of [PdII]=CF2. e, Reaction of cis-G1 with CO. This reaction proceeded smoothly at room temperature to give the carbonylation product
in quantitative yield, whereas the reaction of palladium complex E1 with CO failed to provide the fluoroacyl palladium complex (Supplementary Fig. 7).
These results suggest that the carbonylation step occurs after the transmetallation of complex E1. X denotes hydroxyl or other oxygen anions. Colour code:
grey, carbon; white, hydrogen; blue, palladium; orange, phosphorus; red, oxygen; green, fluorine.
Subsequently, protonation of A at the difluorocarbene ligand 11.1 kcal mol−1 higher than that for the CF2 elongation step from C
provides the key intermediate [(Xantphos)PdII(CF2H)(OH)] (C). (Fig. 4c), it is relatively difficult to selectively control the produc-
As a competitive process to CF2 elongation, C undergoes trans- tion of the desired difluoromethyl ketone while suppressing the
metallation and reductive elimination to afford ArCF2H, depend- difluorocarbene insertion products in the presence of excessive
ing on the tuning of the reaction conditions (Supplementary Figs. 8, difluorocarbene. As a result, an additional CO source and lower
11 and 12). Difluorocarbene insertion by the reaction of difluo- concentration of difluorocarbene is required to furnish the difluo-
rocarbene with
complex A′. Complex A′ subsequently undergoes intramolecu-
C
affords [(HCF2)(Xantphos)PdII=CF2(OH)] romethyl ketone (Supplementary Figs. 13 and 14).
In conclusion, a controllable way to synthesize difluoromethyl-,
lar difluoromethyl migration to generate the CF2 elongated com- tetrafluoroethyl-, difluoroacetyl- or tetrafluoropropanoyl arenes
plex [(Xantphos)PdII(CF2CF2H)(OH)] (E). Complex E undergoes from arylboronic acids has been developed and explored mecha-
transmetallation to deliver the key intermediate [(Ar)(Xantphos) nistically. The ready availability of these compounds is made pos-
PdII(CF2CF2H)] (G). Reductive elimination of G affords tetrafluo- sible by harnessing metal difluorocarbene chemistry, opening a new
roethylated arenes. Insertion of CO into the Pd–Ar σ bond of G, chapter in the preparation of fluorinated derivatives of potential
in which CO is derived from hydrolysis of [PdII]=CF2, followed by value in the pharmaceutical industry.
reductive elimination, provides tetrafluoroethyl ketones. The key
intermediate C may also undergo a similar pathway to produce
methods
difluoromethyl ketones (Supplementary Figs. 8 and 13). Because
the barrier for the formation of [(Ar)(Xantphos)PdII(CF2H)] (D),
a key intermediate for the production of difluoromethyl ketone, is
All calculations were performed using the Gaussian 09 suite of programs42.
Geometry optimizations and frequencies were calculated with the B3LYP density
functional43–45 and a mixed basis set of LANL2DZ46 for Pd and 6–31G(d)47 for other
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
953

NATurE CHEMIsTry
Articles
a
d
P
P
ΔG (kcal mol–1
)
Pd
H
CF2
2.65
O
CF2CF2H
2.47
2.19
P
P
Ph BOH
16.9
P
CF2
Pd
3.02
P
= Xantphos
C
H
O
1.89
O
Ph
Pd
P
TSA-C'''
P
Ph
TSA-C''
TSG1-H1
'
P
P
P
–72.9
TSG1-H1
'
Pd
Pd
CF2
P
A0
1.5
TSA-C''
COCF2CF2H
Ph
H1'
–97.6
H
P
P
O
H
CO
Pd
0.5
0.0
TSA-C
TSA-C
–2.9
TSA-C'
–95.5
PhOH
H2O
P
P
–97.1
2.00
2.34
TSG1-H1
Pd
CF2
CF2CF2H
1.29
P
P
1.89
2.01
H
CF2CF2H
3.48
2.49
O
CF2
Pd
2.55
2.97
O
P
–107.6
CF2CF2H
Ph BOH
Ph
PhB(OH)2
C
Pd
2.25
1.31
P
TSA-C'
cis-G1
P
P
–18.5
H2O
Pd
–27.4
cis-C
P
–28.0
CF2H
COPh
CF2H
TSG1-H1
P
P
cis-C'
Pd
PhB(OH)2
H1
Pd
Pd CF2
P
P
OH
OB(OH)Ph
P
A
b
e
CF2CF2H
P
Pd
CF2CF2H
P
–0.10
–0.03
0.01
P
P
–0.035
–0.035
–0.02
Pd
C
+
TSG1-P61
–72.9
O
0.07
–0.10
0.22
0.09
0.01
TSH1-P59
–92.5
–0.02
TSG1-H1
–95.5
A0
CF2CF2H
CO
A
P
P
Pd
–97.1
c
COPh
H1
CF2CF2H
CF2CF2H
P
O
P
P
C
CF2H
Pd
Pd
2.11
1.91
–107.6
P
P
Ph
Pd
A0
3.67
2.45
cis-G1
P
B(OH)2
HO
TSG1-H1
A0
2.33
CF2H
TStrans-C--cis-D1
–8.5
PhB(OH)2
–16.9
–118.4
P
P
CF2
OH
Pd
Ph-COCF2CF2H
TSA'-E
–124.5
59
TSA'-E
PhCF2CF2H
CF2
–19.6
CF2H
HF2CF2C
61
P
P
P
Pd
Pd
–26.4
CF2H
B(OH)3
P
OH
B(OH)2
HO
trans-C
2.49
TSE-G
–46.1
2.09
2.44
P
Pd
OH
A'
CF2
P
–59.6
2.52
2.65
CF2H
P
P
2.28
Pd
–67.6
2.40
2.06
Ph
cis-D1
B(OH)3
CF2CF2H
2.05
2.03
P
Pd
–97.1
P
OH
CF2CF2H
E
P
P
Pd
TSG1-P61
TSH1-59
Ph
cis-G1
Fig. 4 | Calculated energy profile of the palladium-catalysed difluorocarbene transfer reactions. a, DFT-computed reaction pathways and optimized key
transition state structures for protonation of [(Xantphos)Pd0ꢀ=ꢀCF2] a. Here, we chose cis-[(Xantphos)PdII(CF2H)(OH)] (cis-C) as the protonation intermediate
in calculations (also in c), because its transmetallation has a lower barrier than the direct transmetallation of [(HCF2)(Xantphos)PdII[PhB(OH)O]] (cis-C′)
(Supplementary Figs. 11 and 12). b, Computed charge model 5 (CM-5) charges of (Xantphos)Pd, difluorocarbene and [(Xantphos)Pd0ꢀ=ꢀCF2] a. c, Free energy
barrier differences between difluorocarbene insertion and transmetallation of trans-C. d, DFT-computed reaction pathways and optimized key transition
state structures for insertion of CO into complex G1. e, Free energy barrier differences between reductive elimination and CO insertion of complex G1. Colour
code: grey, carbon; white, hydrogen; blue, palladium; orange, phosphorus; red, oxygen; green, fluorine; pink, boron. All energies are reference to A0.
atoms, in conjunction with the SMD48 implicit solvation model to account for the
solvation effects of dioxane. Optimized geometries were verified by frequency
computations as minima (zero imaginary frequencies) or transition structures
(a single imaginary frequency) at the same level of theory. More accurate electronic
energies were obtained by single-point energy calculations at the SMD-M0649,50
/
6–311++G(d,p)+SDD(Pd)51,52 level of theory. All Gibbs energies in solution reported
throughout the text are in kcalmol−1, and the bond lengths are in ångstroms (Å).
The structures were generated by CYLview53.
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
954

NATurE CHEMIsTry
Articles
Hydroquinone
Base
BrCF2PO(OEt)2
CF2
ArCOCF2CF2H
(Ln)PdII=CF2
Pd0(Ln)
(Ln)Pd0=CF2
H2O
(Ln)Pd0=CF2
ArCOPdII(Ln)CF2CF2H
[PdII]
(A)
(H)
H
ROH
CO
Carbonylation
Protonation
CO
Pd(Ln)
XPdII(Ln)CF2H
(C)
ArPdII(Ln)CF2CF2H
ArCF2CF2H
O
CF2
R
H
(G)
Pd0(Ln)
Difluorocarbene
insertion
Base
CF2
CF2
XPdII(Ln)CF2CF2H
X
PdII CF2H
XPdII(Ln)CF2H
R = ArBOH, H, p-HO-C6H4
(C)
Ar
= ArB(OH)2
(E)
L
(A')
Difluoromethyl
migration
Fig. 5 | Outline of a possible pathway for controllable palladium-catalysed difluorocarbene transfer. The reaction is initiated by the formation of
palladium(0) difluorocarbene complex a, in which the difluorocarbene is generated by the reaction of BrCF2PO(OEt)2 with hydroquinone in the presence
of base. Subsequent protonation at the difluorocarbene carbon of a gives the key intermediate C, which is illustrated in more detail at the lower left of the
figure. As a process competitive to CF2 elongation, C undergoes transmetallation and reductive elimination to produce ArCF2H, depending on the tuning
of the reaction conditions (Supplementary Figs. 8, 11 and 12). Difluorocarbene insertion into C forms palladium difluorocarbene complex a′, and then
intermolecular difluoromethyl migration generates CF2 elongated complex e. Subsequent transmetallation of e produces complex G, which undergoes
reductive elimination to provide tetrafluoroethylated arenes. Alternatively, insertion of CO, which originates from the hydrolysis of [PdII]=CF2 (as shown
in the upper right of the figure), into the Pd–Ar σ bond of G, followed by reductive elimination, gives tetrafluoroethyl aryl ketones. In another competitive
reaction, the key intermediate C may also undergo carbonylation to afford difluoromethyl aryl ketones (Supplementary Figs. 8 and 13). For simplicity, all
palladium complexes are illustrated as neutral species, and all processes are depicted as being irreversible. Ln denotes Xantphos coordinated to palladium.
Data availability
10. MacNeil, J. G.Jr & Burton, D. J. Generation of trifuoromethylcopper from
chlorodifuoroacetate. J. Fluor. Chem. 55, 225–227 (1991).
Crystallographic data for the structures reported in this Article have been deposited
at the Cambridge Crystallographic Data Center under deposition numbers CCDC
1902891 (B1-1), 1916606 (B1-2), 1916607 (A1-2), 1902894 (C1-1a), 1902898
(C2), 1902901 (E1) and 1902900 (cis-G1). Copies of the data can be obtained free
of charge from https://www.ccdc.cam.ac.uk/strucutres/. All other data supporting
the findings of this study are available within the Article and its Supplementary
Information, or from the corresponding author upon reasonable request.
11. Duan, J.-X., Su, D.-B. & Chen, Q.-Y. Trifuoromethylation of organic halides
with methyl halodifuoroacetates—a process via difuorocarbene and
trifuoromethide intermediates. J. Fluor. Chem. 61, 279–284 (1993).
12. Huiban, M. et al. A broadly applicable [18F]trifuoromethylation of aryl and
heteroaryl iodides for PET imaging. Nat. Chem. 5, 941–944 (2013).
13. Levin, M. D. et al. A catalytic fuoride-rebound mechanism for C(sp3)–CF3
bond formation. Science 356, 1272–1276 (2017).
Received: 14 March 2019; Accepted: 13 August 2019;
Published online: 23 September 2019
14. Dilman, A. D. & Levin, V. V. Difuorocarbene as a building block for
consecutive bond-forming reactions. Acc. Chem. Res. 51, 1272–1280 (2018).
15. Reger, D. L. & Dukes, M. D. Molybdenum perfuorocarbene complexes.
J. Organomet. Chem. 153, 67–72 (1978).
16. Brothers, P. J. & Roper, W. R. Transition-metal dihalocarbene complexes.
Chem. Rev. 88, 1293–1326 (1988).
17. Harrison, D. J., Gorelsky, S. I., Lee, G. M., Korobkov, I. & Baker, R. T. Cobalt
fuorocarbene complexes. Organometallics 32, 12–15 (2013).
18. Harrison, D. J., Daniels, A. L., Korobkov, I. & Baker, R. T. d10 nickel
difuorocarbenes and their cycloaddition reactions with tetrafuoroethylene.
Organometallics 34, 5683–5686 (2015).
references
1. Kirsch, P. Modern Fluoroorganic Chemistry: Synthesis, Reactivity, Applications
2nd edn (Wiley-VCH, 2013).
2. Brahms, D. L. S. & Dailey, W. P. Fluorinated carbenes. Chem. Rev. 96,
1585–1632 (1996).
3. Ni, C. & Hu, J. Recent advances in the synthetic application of difuorocarbene.
Synthesis 46, 842–863 (2014).
4. Hudlicky, M. & Pavlath, A. E. Chemistry of Organic Fluorine Compounds II
(American Chemical Society, 1995).
19. Trnka, T. M., Day, M. W. & Grubbs, R. H. Olefn metathesis with
1,1-difuoroethylene. Angew. Chem. Int. Ed. 40, 3441–3444 (2001).
20. Takahira, Y. & Morizawa, Y. Ruthenium-catalyzed olefn cross-metathesis
with tetrafuoroethylene and analogous fuoroolefns. J. Am. Chem. Soc. 137,
7031–7034 (2015).
5. Miller, T. G. & Tanassi, J. W. Te preparation of aryl difuoromethyl ethers.
J. Org. Chem. 25, 2009–2012 (1960).
6. Hohlfeld, J. M. et al. Rofumilast attenuates pulmonary infammation upon
segmental endotoxin challenge in healthy subjects: a randomized placebo-
controlled trial. Pulm. Pharmacol. Ter. 21, 616–623 (2008).
7. Dolbier, W. R. Jr. & Battiste, M. A. Structure, synthesis and chemical
reactions of fuorinated cyclopropanes and cyclopropenes. Chem. Rev. 103,
1071–1098 (2003).
21. Feng, Z., Min, Q.-Q. & Zhang, X. Access to difuoromethylated arenes by
Pd-catalyzed reaction of arylboronic acids with bromodifuoroacetate.
Org. Lett. 18, 44–47 (2016).
22. Feng, Z., Min, Q.-Q., Fu, X.-P., An, L. & Zhang, X. Chlorodifuoromethane-
triggered formation of difuoromethylated arenes catalysed by palladium.
Nat. Chem. 9, 918–923 (2017).
8. Burch, J. D. et al. Tetrahydroindazoles as interleukin-2 inducible T-cell kinase
inhibitors. Part II. Second-generation analogues with enhanced potency,
selectivity and pharmacodynamic modulation in vivo. J. Med. Chem. 58,
3806–3816 (2015).
9. Fuqua, S. A., Duncan, W. G. & Silverstein, R. M. A one-step synthesis of
1,1-difuoroolefns from aldehydes by a modifed Wittig synthesis.
Tetrahedron Lett. 5, 1461–1463 (1964).
23. Deng, X.-Y., Lin, J.-H. & Xiao, J.-C. Pd-catalyzed transfer of difuorocarbene.
Org. Lett. 18, 4384–4387 (2016).
24. Clark, G. R., Hoskins, S. V., Jones, T. C. & Roper, W. R. Oxidation state
control of the reactivity of a transition metal–carbon double bond. Synthesis,
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
955

NATurE CHEMIsTry
Articles
X-ray crystal structure, and reactions of the zerovalent difuorocarbene
complex [Ru(=CF2)(CO)2(PPh3)2]. J. Chem. Soc. Chem. Commun.
719–721 (1983).
47. Hehre, W. J., Ditchfeld, R. & Pople, J. A. Self-consistent molecular orbital
methods. XII. Further extensions of Gaussian-type basis sets for use in
molecular orbital studies of organic molecules. J. Chem. Phys. 56,
2257–2261 (1972).
48. 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
defned by the bulk dielectric constant and atomic surface tensions. J. Phys.
Chem. B 113, 6378–6396 (2009).
49. Zhao, Y. & Truhlar, D. G. Density functionals with broad applicability in
chemistry. Acc. Chem. Res. 41, 157–167 (2008).
25. Brothers, P. J., Burrell, A. K., Clark, G. R., Rickard, C. E. F. & Roper, W. R.
Trifuoromethyl, difuorocarbene and tetrafuoroethylene complexes of
iridium and the crystal structures of IrI(CH3)(CF3)(CO)(PPh3)2, Ir(CF3)(C2F4)
(CO)(PPh3)2 and Ir(CF3)(CF2)(CO)(PPh3)2. J. Organomet. Chem. 394,
615–642 (1990).
26. Hughes, R. P. et al. A simple route to difuorocarbene and perfuoroalkylidene
complexes of iridium. J. Am. Chem. Soc. 127, 15020–15021 (2005).
27. Negishi, E. & de Meijere, A. Handbook of Organopalladium Chemistry for
Organic Synthesis 2nd edn (Wiley, 2002).
50. Zhao, Y. & Truhlar, D. G. Te M06 suite of density functionals for main
group thermochemistry, thermochemical kinetics, noncovalent interactions,
excited states and transition elements: two new functionals and systematic
testing of four M06-class functionals and 12 other functionals. Teor. Chem.
Acc. 120, 215–241 (2008).
28. Yang, Z.-Y., Wiemers, D. M. & Burton, D. J. Trifuoromethylcopper: a useful
difuoromethylene transfer reagent: a novel double insertion of
difuoromethylene into pentafuorophenylcopper. J. Am. Chem. Soc. 114,
4402–4403 (1992).
51. Andrae, D., Haussermann, U., Dolg, M., Stoll, H. & Preuss, H. Energy-
adjusted ab initio pseudo potentials for the second and third row transition
elements. Teor. Chim. Acta 77, 123–141 (1990).
29. Zafrani, Y., Sod-Moriah, G. & Segall, Y. Diethyl
bromodifuoromethylphosphonate: a highly efcient and environmentally
benign difuorocarbene precursor. Tetrahedron 65, 5278–5283 (2009).
30. Cordaro, J. G. & Bergman, R. G. Dissociation of carbanions from acyl
iridium compounds: an experimental and computational investigation. J. Am.
Chem. Soc. 126, 16912–16929 (2004).
52. Dolg, M., Wedig, U., Stoll, H. & Preuss, H. Energy-adjusted ab initio
pseudo potentials for the frst row transition elements. J. Chem. Phys. 86,
866–872 (1987).
53. Legault, C. Y. CYLview, 1.0b (Universite
http://www.cylview.org.
́
de Sherbrooke, 2009);
31. Fujiwara, Y. et al. A new reagent for direct difuoromethylation. J. Am. Chem.
Soc. 134, 1494–1497 (2012).
32. Fier, P. S. & Hartwig, J. F. Copper-mediated difuoromethylation of aryl and
vinyl iodides. J. Am. Chem. Soc. 134, 5524–5527 (2012).
acknowledgements
Financial support for this work was provided by the National Natural Science Foundation
of China (21425208, 21672238, 21790362 and 21421002), the National Basic Research
Program of China (973 Program) (No. 2015CB931900), the Strategic Priority Research
Program of the Chinese Academy of Sciences (no. XDB20000000). We thank H.-L. Qin
for MS analysis of 18O-labelled compound 7 and G.-Y. Li for 13C NMR analysis of the
palladium complexes. K.N.H. acknowledges the National Science Foundation (NSF)
for support (CHE-1764320). Computations were performed on the Hoffman2 cluster
at UCLA and the Extreme Science and Engineering Discovery Environment (XSEDE),
which is supported by the NSF (OCI-1053575).
33. Gu, Y., Leng, X. & Shen, Q. Cooperative dual palladium/silver catalyst for
direct difuoromethylation of aryl bromides and iodides. Nat. Commun. 5,
5405 (2014).
34. Xu, L. & Vicic, D. A. Direct difuoromethylation of aryl halides via base metal
catalysis at room temperature. J. Am. Chem. Soc. 138, 2536–2539 (2016).
35. Meanwell, N. A. Synopsis of some recent tactical application of bioisosteres in
drug design. J. Med. Chem. 54, 2529–2591 (2011).
36. Erickson, J. A. & McLoughlin, J. I. Hydrogen bond donor properties of the
difuoromethyl group. J. Org. Chem. 60, 1626–1631 (1995).
37. Vaclavik, J., Klimankova, I., Budinska, A. & Beier, P. Advances in the
synthesis and application of tetrafuoroethylene- and 1,1,2,2-tetrafuoroethyl-
containing compounds. Eur. J. Org. Chem. 2018, 3554–3593 (2018).
38. Ishiyama, T. et al. Mild iridium-catalyzed borylation of arenes. High turnover
numbers, room temperature reactions and isolation of a potential
intermediate. J. Am. Chem. Soc. 124, 390–391 (2002).
39. Jean, Y. Molecular Orbitals of Transition Metal Complexes
(Oxford Univ. Press, 2005).
40. Li, L., Wang, F., Ni, C. & Hu, J. Synthesis of gem-difuorocyclopropa(e)nes
and O-, S-, N- and P-difuoromethylated compounds with TMSCF2Br.
Angew. Chem. Int. Ed. 52, 12390–12394 (2013).
author contributions
X.Z. and X.-P.F. conceived and designed the experiments. X.Z. directed the project.
X.-P.F. performed the experiments and mechanism studies. X.-S.X. conducted the DFT
calculations and contributed parts of the mechanism analysis. K.N.H. directed the DFT
calculations. X.-Y.Z. conducted parts of the mechanistic studies. Y.-L.G. conducted MS
analysis of the palladium complexes. X.L. analysed the X-ray crystal structure of the
palladium difluorocarbene complex. X.-P.F., Y.-L.X. and X.Z. analysed the data. X.Z.,
X.-S.X., S.Z. and K.N.H. co-wrote the manuscript. All authors discussed the results and
commented on the manuscript.
41. Pu, M. I., Sanhueza, A., Senol, E. & Schoenebeck, F. Divergent reactivity of
stannane and silane in the trifuoromethylation of PdII: cyclic transition state
versus difuorocarbene release. Angew. Chem. Int. Ed. 57, 15081–15085 (2018).
42. Frisch, M. J. et al. Gaussian 09, revision D.01 (Gaussian, 2009).
43. Lee, C., Yang, W. & Parr, R. G. Development of the Colle–Salvetti correlation-
energy formula into a functional of the electron density. Phys. Rev. B. 37,
785–789 (1988).
Competing interests
The authors declare no competing interests.
additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41557-019-0331-9.
44. Becke, A. D. Density functional thermochemistry. III. Te role of exact
exchange. J. Chem. Phys. 98, 5648–5652 (1993).
Correspondence and requests for materials should be addressed to K.N.H. or X.Z.
Reprints and permissions information is available at www.nature.com/reprints.
45. Becke, A. D. A new mixing of Hartree–Fock and local density-functional
theories. J. Chem. Phys. 98, 1372–1377 (1993).
46. Hay, P. J. & Wadt, W. R. Ab initio efective core potentials for molecular
calculations. Potentials for K to Au including the outermost core orbitals.
J. Chem. Phys. 82, 299–310 (1985).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2019
Nature CHemiStrY | VOL 11 | OCTOBER 2019 | 948–956 | www.nature.com/naturechemistry
956