Copper-mediated difluoromethylation of (hetero)aryl iodides and β-styryl halides with tributyl(difluoromethyl)stannane

Article abstract of DOI:10.1002/anie.201205850

Owing to their unique properties, molecules containing the difluoromethyl group (CF₂H) are of great interest. Tributyl(difluoromethyl)stannane has now been used for the selective and efficient direct ipso difluoromethylation of aryl iodides, he

Full text of DOI:10.1002/anie.201205850

.
Angewandte
Communications
DOI: 10.1002/anie.201205850
Organofluorine Compounds
Copper-Mediated Difluoromethylation of (Hetero)aryl Iodides and
b-Styryl Halides with Tributyl(difluoromethyl)stannane**
G. K. Surya Prakash,* Somesh K. Ganesh, John-Paul Jones, Aditya Kulkarni, Kamil Masood,
Joseph K. Swabeck, and George A. Olah
Organofluorine compounds have found application in a wide
variety of fields. They occur, for example, in pharmaceuticals,
agrochemicals, materials, surfactants, and catalysts.[1] The
selective introduction of the difluoromethyl group (CF2H)
into organic molecules is of interest owing to the special
biological properties of this group, such as its enhancement of
membrane permeability, binding affinity, and bioavailabil-
ity.[2] The CF2H functionality is isosteric and isopolar with the
hydroxy (OH) group and is reasonably lipophilic.[3] At the
same time, the CF2H group is weakly acidic and capable of
participating in weak hydrogen-bonding interactions. Because
of these properties, the CF2H group is found in various
biologically active compounds, such as enzyme inhibitors,
sugars,[4] and agrochemicals.[5]
Several methods have been developed for the preparation
of CF2H-containing compounds, including the deoxofluori-
nation of aldehydes with SF4, N,N-diethylaminosulfur tri-
fluoride (DAST), and derivatives of DAST.[6] The magne-
sium-metal-mediated reductive difluoromethylation of
chlorosilanes with difluoromethyl sulfides, sulfoxides, and
sulfones has been reported,[7] as has the nucleophilic intro-
duction of a CF2H group into carbonyl compounds with
difluoromethyl phenyl sulfone,[8] (difluoromethyl)dimethyl-
phenylsilane, and (chlorodifluoromethyl)trimethylsilane.[3,9]
The direct transfer of a difluoromethyl group to a heteroarene
with zinc difluoromethanesulfinate (DFMS), which is
believed to proceed by a radical pathway, was reported by
the Baran research group.[10] However, direct access to
regiospecifically difluoromethylated arenes has been a chal-
lenge until recently.
Amii and co-workers reported a CuI-catalyzed three-step
approach for the synthesis of difluoromethyl aromatic and
Scheme 1. Methods for difluoromethylation. EWG=electron-withdraw-
ing group, TMS=trimethylsilyl.
À
heteroaromatic compounds by a C C coupling reaction
between an aryl iodide and an a-silyldifluoroacetate, followed
by hydrolysis and decarboxylation (Scheme 1).[11] Fier and
Hartwig reported a one-step copper-mediated (CuI) nucleo-
philic difluoromethylation of iodoarenes with TMSCF2H.[12]
Although the products were obtained in excellent yields, the
latter method requires a significant excess of TMSCF2H, is
limited to electron-rich and electron-neutral iodoarenes, and
is not compatible with aldehydes and ketones owing to
competitive nucleophilic addition at the carbonyl center.[13]
Herein we report the synthesis of tributyl(difluorome-
thyl)stannane and its application in the copper-mediated
difluoromethylation of iodoarenes and b-styryl halides. This
methodology has been extended to the difluoromethylation
of both activated and deactivated iodoarenes (iodohetero-
arenes), including those with carbonyl substituents, in mod-
erate to good yields, and the difluoromethylation of b-styryl
halides in good to excellent yields.
[*] Prof. Dr. G. K. S. Prakash, Dr. S. K. Ganesh, J.-P. Jones,
Dr. A. Kulkarni, K. Masood, J. K. Swabeck, Prof. Dr. G. A. Olah
Loker Hydrocarbon Research Institute and
Department of Chemistry, University of Southern California
University Park, Los Angeles, CA 90089-1661 (USA)
E-mail: gprakash@usc.edu
[**] Financial support by the Loker Hydrocarbon Research Institute is
gratefully acknowledged.
Our research group recently demonstrated that singlet
CF2 carbene can be generated from TMSCF3 in the presence
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201205850.
12090
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12090 –12094

Angewandte
Chemie
of tetra-n-butylammonium difluorotriphenylsilicate (TBAT)
and NaI.[14] This carbene adds readily across an alkene or
alkyne by a [2+1] cycloaddition reaction. We wanted to use
this methodology to insert a CF2 carbene into a metal–
hydrogen bond. While screening various compounds contain-
ing metal–hydrogen bonds, we observed that the CF2 carbene
stable. The insertion reaction can be carried out on a multi-
gram scale by this simple method to yield pure 2a, 2c, and 2d
in an inert environment. The isolation of trimethyl(difluoro-
methyl)stannane (2b) proved to be a challenge and was not
pursued further. Compound 2a is readily purified and can be
stored in air for long periods of time (weeks to months at
least) without decomposition.
À
À
could be readily inserted into the Sn H bond of nBu3Sn H to
afford nBu3SnCF2H under mild conditions in over 80% yield.
Cullen et al. had previously reported the insertion of CF2
After extensive screening studies to determine conditions
suitable for the copper-mediated difluoromethylation of
iodoarenes, we observed the desired product when CuI was
used in DMA with KF as an initiator. The reaction was found
to be sensitive to the amounts of CuI, KF, and 2a used, and
the ratio of KF to CuI. We adopted two sets of optimized
conditions: method A, in which the starting material, a b-
styryl halide, iodonaphthalene, or iodo-substituted aldehyde/
ketone, is treated with 1.3 equivalents of CuI, 3 equivalents of
KF, and 2 equivalents of 2a for 24 h at 1008C, and method B,
which is effective for Br-, Cl-, Ph-, and CF3-substituted
iodoarenes and calls for 1.3 equivalents of CuI, 3 equivalents
of KF, and 3 equivalents of 2a, a reaction temperature of
1208C, and a reaction time of 24 h. When CsF was used as the
initiator, the desired product was formed in lower yield. Only
traces of the desired product were observed when other CuI
halides, such as CuBr and CuCl, were used. Less than 5%
yield of the product was observed by 19F NMR spectroscopy
in the control reactions when no KF was added.
À
carbene generated from Me3SnCF3 into the Sn H bond of
trimethyltin hydride under harsh conditions (1508C, 24 h);
the insertion product was obtained in 63% yield.[15] We found
calcium iodide to be an ideal initiator. In the presence of CaI2,
CF2 carbene could be generated from TMSCF3 at 458C to
give nBu3SnCF2H (2a) as the major product in less than an
hour. The reaction proceeds in a variety of polar aprotic
solvents, such as N,N-dimethylformamide (DMF), N,N-dime-
thylacetamide (DMA), and N-methylpyrrolidone (NMP;
Table 1).
Table 1: Optimization of the synthesis of nBu3SnCF2H with the initiator
CaI2.[a]
Me3SiCF3 [equiv]
Solvent
Yield of 2a [%][b]
A series of iodoarenes (iodoheteroarenes) were converted
into the respective CF2H-substituted arenes (heteroarenes) in
the solvent DMA (Table 3). The products were formed in
slightly higher yields in DMA than in analogous solvents, such
as DMF, 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone
(DMPU), and NMP. Although yields vary, the method is
extremely selective and amenable to a variety of function-
alities, including aldehydes, ketones, esters, and other halides.
Iodo-substituted heteroarenes were also converted into the
desired products (3j, 3k, and 3l). Reactions with bromo-
arenes as the starting material proceeded only in low yields,
rarely above 5%. In products 3k, 3l, and 3o, the bromo
substitutent remained intact. This selectivity could be
exploited by the use of bromides such as 3k, 3l, and 3o as
starting materials in coupling reactions to generate more
complex difluoromethylated molecules. Compounds 3l, 3m,
and 3p–r were not isolated because of their volatility or
relatively low yield.
Owing to the success of this copper-catalyzed ipso-
substitution reaction, we assumed that b-styryl substituents
should undergo a similar reaction. Our findings support this
premise: the corresponding products were formed in yields
20–30% higher than those observed for similar iodoarenes
(Table 4). b-Styryl bromides also gave the desired products,
albeit in slightly lower yields than their iodo counterparts. No
isomerization of the double bond was observed during the
reaction with the b-styryl halides, which indicates that the
reaction proceeds with retention of configuration. Substituted
styryl iodides (cis, trans, electron-rich, and electron-poor)
showed similar reactivity; the desired products were formed
by method A in good to excellent yields (60–85%; Table 4).
Both computational and experimental methods were used
to study the mechanism of the reaction. We based our
2.25
2.25
2.25
2.25
2.25
2.25
1.1
DMA
DMF
NMP
THF
THF/HMPA (1:1)
DMSO
NMP
NMP
NMP
85
72
86
0
0
0
63
79
82
1.4
1.7
[a] Reaction conditions: nBu3SnH: 8.68 mmol, CaI2: 0.34 mmol, solvent:
4 mL. [b] The yield was determined by 19F NMR spectroscopy.
DMSO=dimethyl sulfoxide, HMPA=hexamethylphosphoramide.
This procedure is quite general for the transformation of
compounds of type R3SnH into R3SnCF2H (Table 2). How-
ever, we focused on the use of 2a owing to its ease of
synthesis, nonvolatile nature, and stability in air. Further-
more, unlike most tin hydrides, such as Ph3SnH, Cy3SnH, and
Me3SnH, the starting tributyltin hydride is inexpensive, has
relatively low toxicity (because of its low volatility), and is
Table 2: Conversion of R3SnH into R3SnCF2H.[a]
[a] Reaction conditions: R3SnH (5 g), TMSCF3 (2.2 equiv), CaI2
(0.07 equiv), DMA (8 mL), 45–508C, 1 h. [b] The yield of the isolated
product is given where applicable. The yield determined by 19F NMR
spectroscopy is given in parentheses. Cy=cyclohexyl.
Angew. Chem. Int. Ed. 2012, 51, 12090 –12094
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12091

.
Angewandte
Communications
Table 3: Difluoromethylation of iodoarenes with 2a.[a]
Table 4: Difluoromethylation of b-styryl halides with 2a.[a]
[a] Reaction conditions: b-styryl halide (0.5 mmol), 2a (1 mmol), CuI
(0.65 mmol), KF (1.5 mmol), DMA (4 mL), N2 atmosphere, 20 mL
microwave vial, 1008C, 24 h. [b] The yield of the isolated product is given
where applicable. Yields in parentheses were determined by 19F NMR
spectroscopy.
a single DMF molecule in all calculations with only one
formal ligand of a copper(I) species.
Since an unstable CF2HÀ anion is not likely to exist
independently in solution (especially as we did not observe
nucleophilic addition to carbonyl groups), we then tried to
identify a reasonable mechanism for the transfer of a CF2H
À
group from the tin reagent 2a to a Cu CF2H species similar to
that reported in the literature.[17] Interestingly, no transition
state could be found by simply combining the CF2H-
substituted tin reagent (structure A in Scheme 2) with CuI
(or CuCF2H) in a plausible way. Only when we first formed
the pentacoordinate anion B were we able to successfully
optimize a transition state involving CuI (structure D).
Similarly, we were unable to find a suitable transition state
involving CuI when we started with the octahedral dianion C.
If no productive transition states exist when A or C are used
as starting points, then the necessary amounts of CuI and KF
should be closely tied, as we have confirmed experimen-
tally.[18]
[a] Reaction conditions: iodoarene (0.5 mmol), 2a (1–1.5 mmol), CuI
(0.65 mmol), KF (1.5 mmol), DMA (4 mL), N2 atmosphere, 20 mL
microwave vial, 100–1208C, 24 h. [b] The yield of the isolated product is
given where applicable. The yield determined by 19F NMR spectroscopy
is given in parentheses.
computational studies on the reaction in DMF, because this
solvent is easier to model than DMA. We also simplified the
substituents on our tin reagent from n-butyl to methyl. All
calculations were conducted with the Gaussian09[16] compu-
tational package at the B3LYP/cc-pVTZ level of theory for
neutral reactions and at the B3LYP/aug-cc-pVDZ level of
theory for reactions involving anions. Further computational
details, Cartesian coordinates, and the energies for all
structures can be found in the Supporting Information.
Rather than attempt to predict energies exactly, we
followed our reasoning that trends determined computation-
ally could help us understand the reaction better, and we
attempted to back up our computationally derived mecha-
nism with experimental data. The addition of a single DMF
molecule into the calculation (with the oxygen atom interact-
ing with the copper center) proved to stabilize the CuCF2H
species by 14.1 kcalmolÀ1. Additional solvent molecules did
not yield stable structures, so from this point forward we used
The 19F NMR spectrum of 2a at À308C[19] in [D7]DMF
with a substoichometric amount of tetramethylammonium
fluoride (TMAF)[20] showed two distinct peaks at d = À136.7
À
and À148.1 ppm, which we assign to Sn F species owing to
the large 1J coupling constant (1780 and 2024 Hz, respec-
tively) of the 119Sn satellites observed. Additional overlapping
peaks observed in the region between À125.5 and
À126.2 ppm, where we generally detect 2a, indicated that
we had potentially formed an activated CF2H species
(Scheme 2, structure B). Unfortunately, this region of the
spectrum was too crowded for the various species present to
be identified and quantified. Under these conditions, we did
not detect CF2H2 (which was observed after heating to room
12092 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12090 –12094

Angewandte
Chemie
Scheme 2. DFT calculations for copper-mediated difluoromethylation.
temperature); thus, we could conclude that 2a had not yet
decomposed to tri-n-butyltin fluoride.
In summary, we have developed an efficient procedure for
the preparation of difluoromethylated tin compounds from
TMSCF3. We have demonstrated a single-step copper-medi-
ated transformation of iodoarenes and iodoheteroarenes into
their CF2H-substituted counterparts with tributyl(difluoro-
methyl)stannane. The protocol gives even higher yields with
b-styryl halides, the reactivity of which is such that bromides
can be used to give the desired products in reasonable yields.
We have also proposed a mechanism for the formation and
utilization of CF2H via a copper intermediate. The choice of
solvent is critical in this reaction because the solvent
The proposed mechanism involves two steps. The trans-
metalation step involves the transfer of CF2H from tin to
copper and is likely to be rate-determining on the basis of our
computational studies, whereas the ipso substitution involves
the reaction between the proposed CuCF2H species and the
starting material. The direct observation of CuCF2H was
unsuccessful, probably as a result of the short half-life of the
species at higher temperatures (À308C), as reported pre-
viously.[17b] The 19F NMR peak assigned to Cu(CF2H)2 by Fier
and Hartwig[12] for difluoromethylation with TMSCF2H was
not observed when 2a was used. Therefore, we propose that
difluoromethylation with 2a is mechanistically different from
difluoromethylation with TMSCF2H. This mechanistic differ-
ence is highlighted by the preference of TMSCF2H for
iodobenzenes with electron-donating groups as opposed to
the preference of 2a for iodobenzenes with electron-with-
drawing groups. The highest-energy transition state for ipso
substitution could be either the coordination of CuCF2H to
the starting material or the oxidative addition. There are two
competing pathways for the reaction of CuCF2H, namely,
decomposition[17b] and coordination to the starting material.
This competition is the likely cause of the varying yields of the
difluoromethylated products. The product of oxidative addi-
tion is formally a copper(III) species, which seems unusual at
first, but has been postulated before as a reactive intermedi-
ate in copper-mediated coupling reactions.[21]
À
molecules must directly stabilize the Cu CF2H species.
Received: July 24, 2012
Revised: September 24, 2012
Published online: October 25, 2012
À
Keywords: C C coupling · copper · density functional
.
calculations · difluoromethylation · stannanes
[1] a) K. Uneyama, Organofluorine Chemistry, Blackwell, Oxford,
2006; b) K. Mꢀller, C. Faeh, F. Diederich, Science 2007, 317,
1881; c) Organofluorine Chemistry: Principles and Commercial
Applications (Eds.: R. E. Banks, B. E. Smart, J. C. Tatlow),
Plenum, New York, 2000; d) R. Filler, Y. Kobayashi, L. M.
Yagupolskii, Organofluorine Compounds in Medicinal Chemis-
try and Biomedical Applications, Elsevier, New York, 1993; e) T.
Hiyama, K. Kanie, T. Kusumoto, Y. Morizawa, M. Shimizu,
Organofluorine Compounds: Chemistry and Applications,
Springer, Berlin, 2000.
[2] a) J. A. Erickson, J. I. McLoughlin, J. Org. Chem. 1995, 60, 1626;
b) K. L. Kirk, Org. Process Res. Dev. 2008, 12, 305; c) G. K. S.
Prakash, S. Chacko, Curr. Opin. Drug Discovery Dev. 2008, 11,
793.
[3] A. K. Yudin, G. K. S. Prakash, D. Deffieux, M. Bradley, R. Bau,
G. A. Olah, J. Am. Chem. Soc. 1997, 119, 1572.
Although CuI should be catalytic (as we start and end with
it in our mechanism), we did not find good experimental
evidence to suggest that the reaction is truly catalytic in CuI.
At best, one could describe the reaction as substoichiometric
in CuI, depending on the reaction conditions. Thus, there may
be other pathways for decomposition of the copper species to
insoluble (aggregates) or otherwise unreactive products.
Angew. Chem. Int. Ed. 2012, 51, 12090 –12094
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 12093

.
Angewandte
Communications
[4] a) J. S. Houlton, W. B. Motherwell, B. C. Ross, M. J. Tozer, D. J.
Williams, A. M. Z. Slawin, Tetrahedron 1993, 49, 8087; b) S.
Kaneko, T. Yamazaki, T. Kitazume, J. Org. Chem. 1993, 58, 2302.
[5] W. F. Goure, K. L. Leschinsky, S. J. Wratten, J. P. Chupp, J. Agric.
Food Chem. 1991, 39, 981.
[6] a) R. P. Singh, J. M. Shreeve, Synthesis 2002, 2561; b) W. R.
Dolbier, P. Xie, R. Zhang, W. Xu, Y. Chang, K. A. Abboud, J.
Org. Chem. 2008, 73, 2469.
[7] G. K. S. Prakash, J. Hu, G. A. Olah, J. Org. Chem. 2003, 68, 4457.
[8] G. K. S. Prakash, J. Hu, Y. Wang, G. A. Olah, Eur. J. Org. Chem.
2005, 2218.
[9] G. K. S. Prakash, J. Hu, Acc. Chem. Res. 2007, 40, 921.
[10] Y. Fujiwara, J. A. Dixon, R. A. Rodriguez, R. D. Baxter, D. D.
Dixon, M. R. Collins, D. G. Blackmond, P. S. Baran, J. Am.
Chem. Soc. 2012, 134, 1494.
[11] a) K. Fujikawa, Y. Fujioka, A. Kobayashi, H. Amii, Org. Lett.
2011, 13, 5560; b) K. Fujikawa, A. Kobayashi, H. Amii, Synthesis
2012, 3015.
[12] P. S. Fier, J. F. Hartwig, J. Am. Chem. Soc. 2012, 134, 5524.
[13] a) Y. Zhao, W. Huang, J. Zheng, J. Hu, Org. Lett. 2011, 13, 5342;
b) T. Hagiwara, T. Fuchikami, Synlett 1995, 717.
[14] F. Wang, T. Luo, J. Hu, Y. Wang, H. S. Krishnan, P. V. Jog, S. K.
Ganesh, G. K. S. Prakash, G. A. Olah, Angew. Chem. 2011, 123,
7291; Angew. Chem. Int. Ed. 2011, 50, 7153.
Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg,
M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M.
Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi,
J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S.
Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.
Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G.
Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S.
Dapprich, A. D. Daniels, ꢁ. Farkas, J. B. Foresman, J. V. Ortiz,
J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[17] a) R. Eujen, B. Hoge, D. J. Brauer, J. Organomet. Chem. 1996,
519, 7; b) D. J. Burton, G. A. Hartgraves, J. Fluorine Chem. 2007,
128, 1198.
[18] Extensive experimental screening showed a high degree of
correlation between the amounts of KF and CuI required in the
reaction. High concentrations of KF relative to CuI left the
reagent unreacted, whereas high concentrations of CuI relative
to KF led mostly to by-products, such as CF2H2. We take these
results as evidence that the reaction is highly dependent on the
amount of an “active” reagent, which, on the basis of computa-
tional modeling, we believe to be [nBu3SnCF2HF]À.
[15] W. R. Cullen, J. R. Suns, M. C. Waldman, Inorg. Chem. 1970, 9,
1682.
[19] We resorted to low-temperature studies as a result of the fast
exchange of FÀ ions at high temperatures.
[16] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci,
G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.
[20] TMAF was used instead of KFas a result of its solubility in DMF
at low temperatures.
[21] H. Hu, J. P. Snyder, J. Am. Chem. Soc. 2007, 129, 7210.
12094 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 12090 –12094