Direct cupration of fluoroform

Article abstract of DOI:10.1021/ja2081026

We have found the first reaction of direct cupration of fluoroform, the most attractive CF₃ source for the introduction of the trifluoromethyl group into organic molecules. Treatment of CuX (X = Cl, Br, I) with 2 equiv of MOR (M = K, Na) in DMF or NMP produces novel alkoxycuprates that readily react with CF₃H at room temperature and atmospheric pressure to give CuCF₃ derivatives. The CuCl and t-BuOK (1:2) combination provides best results, furnishing the CuCF₃ product within seconds in nearly quantitative yield. As demonstrated, neither CF₃^- nor CF₂ mediate the Cu-CF₃ bond formation, which accounts for its remarkably high selectivity. The fluoroform-derived CuCF₃ solutions can be efficiently stabilized with TREAT HF to produce CuCF ₃ reagents that readily trifluoromethylate organic and inorganic electrophiles in the absence of additional ligands such as phenanthroline. A series of novel Cu(I) complexes have been structurally characterized, including K(DMF)[Cu(OBu-t)₂] (1), Na(DMF)₂[Cu(OBu-t)₂] (2), [K₈Cu₆(OBu-t)₁₂(DMF)₈(I)] ⁺ I^- (3), and [Cu₄(CF₃) ₂(C(OBu-t)₂)₂(μ³-OBu-t) ₂] (7).

Full text of DOI:10.1021/ja2081026

ARTICLE
pubs.acs.org/JACS
Direct Cupration of Fluoroform
Alessandro Zanardi, Maxim A. Novikov, Eddy Martin, Jordi Benet-Buchholz, and Vladimir V. Grushin*
The Institute of Chemical Research of Catalonia (ICIQ), Tarragona 43007, Spain
S Supporting Information
b
ABSTRACT: We have found the first reaction of direct cupration of fluoroform,
the most attractive CF3 source for the introduction of the trifluoromethyl group
into organic molecules. Treatment of CuX (X = Cl, Br, I) with 2 equiv of MOR
(M = K, Na) in DMF or NMP produces novel alkoxycuprates that readily react
with CF3H at room temperature and atmospheric pressure to give CuCF3
derivatives. The CuCl and t-BuOK (1:2) combination provides best results,
furnishing the CuCF3 product within seconds in nearly quantitative yield. As
demonstrated, neither CF3ꢀ nor CF2 mediate the CuꢀCF3 bond formation, which accounts for its remarkably high selectivity. The
fluoroform-derived CuCF3 solutions can be efficiently stabilized with TREAT HF to produce CuCF3 reagents that readily
trifluoromethylate organic and inorganic electrophiles in the absence of additional ligands such as phenanthroline. A series of novel
Cu(I) complexes have been structurally characterized, including K(DMF)[Cu(OBu-t)2] (1), Na(DMF)2[Cu(OBu-t)2] (2),
[K8Cu6(OBu-t)12(DMF)8(I)]+ Iꢀ (3), and [Cu4(CF3)2(C(OBu-t)2)2(μ3-OBu-t)2] (7).
’ INTRODUCTION
Numerous polyfluoromethanes that could be used to prepare
CF3 derivatives of Zn, Cd, and Cu (CF3Cl, CF3Br, CF2Cl2,
CF2ClBr, etc.) are on the Montreal Protocol list of substances
that deplete the ozone layer, and CF3I is available only in limited
quantities and also costly. Ruppert-type reagents, CF3SiR3, are
versatile and highly efficient, yet too expensive for larger scale
operations. Low-cost CF3CO2M (M = K, Na) must be used in a
large excess in order to produce CuCF3 species in good yield via
high-temperature (ca. 160 °C) decarboxylation. Although
FSO2CF2CO2Me decarboxylates at lower temperatures, it is
much more expensive and releases 1 equiv of SO2 per each equiv
of CuCF3 generated.
Organic derivatives bearing a trifluoromethyl group on an
aromatic ring are widely used as active ingredients of numerous
modern drugs and highly efficient crop protection agents.1ꢀ8
Trifluoromethylated aromatic building blocks that are needed for
these and other applications, such as in materials,9 are currently
manufactured via a Swarts-type process that involves exhaustive
chlorination of a methyl group on the ring with subsequent Cl/F
exchange.10 This process has a number of limitations, including
low functional group tolerance, and is environmentally un-
friendly because it generates large amounts of chlorine waste.
Efficient metal-catalyzed or -mediated cross-coupling of electro-
philes with a nucleophilic CF3 source (eq 1) is a promising
alternative to the Swarts reaction.8 Since the groundbreaking
discovery of the Cu-promoted perfluoroalkylation of iodoarenes
by McLoughlin and Thrower11 in the mid-1960s and further
developments by Kumadaki et al.12 and others, considerable
progress has been made in the area. Recent advancements in the
field include the first Cu-catalyzed trifluoromethylation reactions
of iodoarenes,13 the synthesis of well-defined Cu(I) trifluoro-
methylating agents,14 the discovery of previously inconceivable
PhꢀCF3 reductive elimination from Pd(II),15 and the develop-
ment of the first Pd-catalyzed aromatic trifluoromethylation
reactions.16,17
Fluoroform (trifluoromethane; CF3H; HFC-23; bp =
∼ꢀ82 °C),18 a side-product of Teflon manufacturing, is cheap,
available in large quantities, nontoxic, and not ozone-depleting. It
is unsurprising therefore, that CF3H has long been viewed as an
ideal source to prepare stable CF3 metal derivatives.8 Although
fluoroform (pKa = 27 in water)19 can be deprotonated by strong
ꢀ
bases such as KCH2S(O)Me, the CF3 generated is notorious
for exceedingly facile decomposition to difluorocarbene at room
temperature and even below. This decomposition can be
avoided by deprotonating CF3H at ꢀ20 to ꢀ40 °C in DMF
ꢀ
that instantly adds the resultant CF3 carbanion to give
[Me2NCH(O)CF3]ꢀ.20,21 Although this hemiaminaloate adduct
still quickly decomposes at room temperature, it is stable at ꢀ20 °C
for at least a few hours and can be used for CF3 transfer to Cu(I)
and subsequent aromatic trifluoromethylation. This strategy was
first reported by Folleas, Marek, Normant, and Saint-Jalmes20
in 2000 and is described in Scheme 1. In the first step, deprotona-
tion of CF3H in DMF at ꢀ25 °C with dimsyl potassium produced
potassium hemiaminolate A that was detected by 19F NMR.
In spite of these important recent accomplishments, the
problem of ArꢀCF3 coupling (eq 1) is far from being solved.
One of the key challenges in the area is the development of low-
cost CF3-transferring nucleophilic reagents for reaction 1.8
Received: August 27, 2011
Published: December 02, 2011
r
2011 American Chemical Society
20901
dx.doi.org/10.1021/ja2081026 J. Am. Chem. Soc. 2011, 133, 20901–20913
|

Journal of the American Chemical Society
ARTICLE
Scheme 1
Figure 1. 19F NMR monitoring of the reaction of CF3H with CuCl, t-BuOK, phen (1:1:1) and PhI in DMF (bottom to top: 1, 2, 6, 24, and 36 h after the
addition of CF3H). The signals at ꢀ25.6 (s), ꢀ63.5 (s), ꢀ80.3 (d, JFꢀH = 79.5 Hz), and ꢀ117.7 (m) ppm are from CuCF3, PhCF3, CF3H (used in
excess), and 4,40-difluorobiphenyl (internal standard), respectively.
The addition of CuI to the thus pregenerated solution of A
induced ligand exchange leading to Cu(I) hemiaminolate B and,
eventually, to CuCF3 (19F NMR). The solution containing
both B and CuCF3 was then stabilized by DMI and heated with
p-iodoanisole to give p-methoxybenzotrifluoride.20
ꢀ40 °C to avoid decomposition to CF2,20,21 our cupration
reaction occurs at room temperature to furnish the CF3Cu
product directly in high yield of up to >90%. A brief preliminary
description of part of this work has been presented in the Latest
Developments section of a recent review article.8
Numerous attempts have been made to metalate rather than
deprotonate CF3H in order to directly obtain stable CF3Cu
derivatives in one step instead of via the multistep route shown
in Scheme 1. It has been communicated,20 however, that basic
organocopper derivatives including n-Bu2CuLi, n-Bu2CuCNLi2,
n-Bu3CuCNLi3, and tert-Bu2CuCNLi2 fail to produce CF3Cu
under a variety of conditions. It is also noteworthy that Et2Zn,
n-Bu3ZnLi, and allylzinc bromide did not give rise to trifluoromethyl-
zincderivatives on contact with CF3H, either.20 Most recently, the
formationof zinc perfluoroalkyls from1H-perfluoroalkanes (RfH)
and zinc bis-2,2,6,6-tetramethylpiperidide was described.13d
The RfZn derivative was then used for in situ Rf transfer to Cu(I)
that was employed as a catalyst for perfluoroalkylation of iodoar-
enes. Direct cupration of fluoroform has not been reported.
Herein we describe the first example of direct cupration of
fluoroform, leading to valuable CF3Cu reagents in one step.
Unlike deprotonation of CF3H that is conducted at ꢀ25 to
’ RESULTS AND DISCUSSION
Originally we found that adding CF3H to a mixture of CuCl,
t-BuOK, and 1,10-phenanthroline (phen) in a 1:1:1 molar ratio
in DMF containing 3 equiv of PhI at room temperature, triggered
a remarkably clean transformation leading to PhCF3 (eq 2).
Monitoring the reaction by 19F NMR (Figure 1) showed that a
CuCF3 species (δ = ꢀ25.6 ppm) was quickly formed, which
reacted with PhI to give PhCF3 at full conversion after 36 h at
25 °C. Although the reaction was highly selective, exhibiting no
19F NMR-observable side products, the yield of PhCF3 was only
45ꢀ50%. Moreover, the yield of PhCF3 was always around 50%
in a number of highly reproducible repeats of this experiment.
GC-MS analysis of the reaction mixture revealed the presence of
t-BuOPh and unreacted PhI. No or little reaction (<5% CuCF3)
was observed in the absence of phen.
20902
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Figure 2. 19F NMR spectrum of a reaction mixture containing the CuCF3 product (ꢀ24.0 ppm) formed in 95% yield on addition of CF3H (ꢀ79.0 ppm,
d, JFꢀH = 79.3 Hz) in excess to CuCl ([Cu] = 0.5 M) and t-BuOK (1:2) in DMF at room temperature. The signal at ꢀ116.4 ppm (m) is from 4,40-
difluorobiphenyl (internal standard).
that integrated in ca. 1:19 (Figure 3). The signal at 6.05 ppm is
assigned to the OH of t-BuOH that is released upon displacement
of the hydrogen of CF3H with Cu and that is apparently in
exchange with the remaining 1 equiv of tert-butoxide. This assign-
ment is fully consistent with the observed integral ratio (∼1:19) vs
the theoretical value of 1:18.
It was then found that CF3H smoothly reacted with CuCl and
t-BuOK in a 1:2 molar ratio in DMF or NMP to produce a CuCF3
derivative in nearly quantitative yield in the absence of phen. This
finding prompted our detailed studies of the reaction, the cuprating
reagent involved, and the trifluoromethyl copper product.
Cupration of CF3H with 1:2 CuClꢀt-BuOK. Addition of
CF3H to a 1:2 mixture of CuCl and t-BuOK in DMF ([Cu] =
0.5 M) at room temperature resulted in a quick reaction. The
reaction was complete within seconds, giving rise to a CuCF3
derivative in up to >95% yield and no 19F NMR-detectable side-
products (Figure 2).
The 13C NMR spectrum (Figure 4) exhibited, apart from a
quartet at 117.4 ppm (JCꢀF = 277 Hz) from CF3H and DMF-d7
solvent peaks, a quartet at 152.3 ppm (JCꢀF = 350 Hz) from
CuCF3, and two sets of broadened resonances from two different
types of tert-butoxy groups. Line-shape analysis produced the
¼
rate of ∼62.5 sꢀ1 and ΔG = 15 kcal molꢀ1 for the exchange of
the two t-BuO groups at 25 °C.23 These figures are inconsistent
with exchange between t-BuOK and t-BuOH via proton transfer
1
that is much faster.24 Indeed, H and 13C NMR spectra of an
anhydrous DMF-d7 solution of t-BuOK and t-BuOH (1:1)
exhibited only one set of resonances for the t-BuO-group.
Therefore, the 13C NMR data not only confirmed the formation
of the CuCF3 species observed by 19F NMR, but also indicated
that after the cupration the tert-butoxide was coordinated to Cu
and also involved in exchange with t-BuOH.23,25
Importantly, the reaction did not give rise to observable
quantities of [CF3CH(O)NMe2]ꢀ (19F NMR: ꢀ78.8 ppm)20,21
ꢀ
that is well-known to form when CF3 is generated in DMF.
Performing the cupration in the presence of 2 equiv of styrene or
α-methylstyrene did not produce gem-difluorocyclopropanes
(19FꢀNMR),22 indicating that the reaction is not mediated by
CF3 or CF2. Therefore, deprotonation of CF3H leading to
CF3ꢀ is not involved in the cupration process.
When the t-BuOK to CuCl ratio was increased to 3:1, the
formation of CuCF3 also occurred in nearly quantitative yield. In this
experiment, however, side products were detected (19F NMR),
apparently from the generation of CF3ꢀ/CF2 upon slower depro-
tonation of CF3H with the extra equivalent of tert-butoxide.21
Repeating the cupration with a 1:1.5 ratio of t-BuOK to CuCl
gave CuCF3 in only ca. 20% yield. Therefore, the required stoichi-
ometry of CuCl and t-BuOK for the generation of an efficient
cuprating reagent is 1:2.
Cuprating Reagents. The reaction of alkali metal tert-but-
oxides with CuCl (1:1) has long been known26ꢀ31 to produce
copper(I) tert-butoxide, a tetramer in the solid state28ꢀ30 and a
strong base that metalates CꢀH acids such as terminal acetylenes
and cyclopentadiene,26 1,3-dinitrobenzene,31 coordinated phos-
phines,29 carboranes,32 some nonclassical hydrides,33 and MꢀH
bonds of certain main group element compounds.34,35 We found,
however, that [t-BuOCu]4, prepared as reported in the litera-
ture26 or generated in situ from CuCl and t-BuOK (1:1) in
DMF, was poorly reactive toward CF3H at 20ꢀ60 °C, giving
only small quantities of CF3Cu species (<10% by 19F NMR), if
any at all. In sharp contrast, the reaction of CF3H with 1:2
mixtures of CuCl and t-BuOK or t-BuONa in DMF occurred
within seconds at room temperature, yielding CF3Cu deriva-
tives. Apparently, a different, more reactive alkoxide species
was generated on treatment of CuCl with 2 equiv of t-BuOM
(M = K, Na).
The cupration reaction was then carried out in DMF-d7 for
studies by 1H and 13C NMR. The 1H NMR spectrum displayed,
in addition to residual solvent peaks and a quartet at 7.6 ppm
(JFꢀH = 79.3 Hz) from CF3H (excess), only two singlets, one at
6.05 ppm (Δν1/2 = 5 Hz) and one at 1.2 ppm (Δν1/2 = 12 Hz)
20903
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Figure 3. 1H NMR spectrum of a reaction mixture obtained upon addition of CF3H in excess to CuCl and t-BuOK (1:2) in DMF-d7 (see text).
Figure 4. 13C NMR spectrum of a reaction mixture obtained upon addition of CF3H in excess to CuCl and t-BuOK (1:2) in DMF-d7 (see text).
On addition of a DMF solution containing 2 equiv of t-BuOK
to a stirred sample of CuCl, the copper salt quickly dissolved and
a cloudy solution was produced. Separation of the solid by
filtration and analysis by powder X-ray diffraction indicated that
KCl was formed quantitatively. Cooling the colorless filtrate gave
uniformly shaped, very air- and moisture-sensitive thin white
needles, whose structure was determined by single-crystal X-ray
diffraction. The established structure K(DMF)[Cu(OBu-t)2] (1)
is shown in Figure 5. This novel dialkoxycuprate crystallizes as a
polymer with each K atom coordinated to the oxygen atom of one
DMF molecule and three oxygen atoms of three [Cu(OBu-t)2]
units.
suspension. Gentle heating of this mixture to ∼50ꢀ60 °C and
quick filtration produced a clear, solid-free filtrate that deposited
needle-shaped crystals on cooling to room temperature. Although
the crystals were also extremely air- and moisture sensitive and
very thin, a single-crystal X-ray crystallographic study was possi-
ble (Figure 6). This sodium cuprate also crystallized as a polymer,
yet appeared to have a different composition, Na(DMF)2[Cu-
(OBu-t)2] (2). Each sodium ion in this structure is bonded to two
DMF molecules through their oxygen atoms and two oxygen
atoms of two adjacent Cu(OBu-t)2 links.
Treatment of CuBr with 2 equiv of t-BuOM (M = K, Na) in
DMF under similar conditions, also resulted in precipitation of
alkali metal bromide, even though KBr and NaBr are slightly
more easily soluble in DMF than KCl and NaCl. However,
no precipitation took place when CuI was dissolved in DMF
When the experiment was repeated with t-BuONa, CuCl also
quickly dissolved. However, a large amount of white crystals
precipitated out within less than 1 min, resulting in a thick
20904
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Figure 5. Structure and ORTEP drawing of the structural unit K(DMF)[Cu(OBu-t)2] (1) with thermal ellipsoids drawn to the 50% probability level.
Figure 6. Structure and ORTEP drawing of the structural unit Na(DMF)2[Cu(OBu-t)2] (2) with thermal ellipsoids drawn to the 50% probability level.
Table 1. CuꢀO Bond Distances (Å) and OꢀCuꢀO Bond
Angles (deg) in 1ꢀ3 and Two Previously Reported Cuprates
[Cu(OC6H3-
Me2-2,6)2]ꢀ [Cu(OPh)2]ꢀ
1
2
3
(ref 38)
(ref 39)
CuꢀO
1.815(2)
1.825(2)
1.816(2)
1.819(2)
1.827(4)
1.832(4)
1.806(6)
1.798(8)
169.8(3)
1.816(4)
1.787(4)
177.8(2)
OꢀCuꢀO 173.10(10) 171.50(10) 175.0(2)
with an iodine atom in the center of the cluster on a 3 axis.
The cluster bears a charge of +1 that is balanced by the outer
sphere iodide anion (ꢀ1) disordered in three positions around
another 3 axis. The eight K atoms form a cube that is body-
centered by the iodine atom and face-centered by six copper
atoms, each bonded to two t-BuO groups. All of the t-BuO ligands
are μ3, capping one Cu and two K atoms of two different types.
The potassium atoms of both types are five-coordinate,
being bonded to the iodine in the center of the cluster,
three t-BuO groups, and one terminal DMF ligand. The two
different DMF molecules attached to K1 and K2 are disordered
in different orientations surrounding the iodide anion. This
structure is distinct and different from those reported37 for the
neutral polynuclear Cu(I) alkoxides [MnCu4(OCR3)8] (M = Li,
n = 4, R = Me; M = Na, n = 4, R = Et; M = Ba, n = 2,
R = Et).
The newcomplexes 1ꢀ3 are extremely rare examples of dialkoxy-
cuprates. We are aware of only two structurally characterized salts
of [Cu(OR)2]ꢀ, both being aromatic derivatives where R = 2,6-
Me2C6H338 and Ph.39 The CuꢀO bond lengths and OꢀCuꢀO
bond angles in 1ꢀ3 and the two literature compounds are
collected in Table 1. Although 1 and 2 crystallize in the polymeric
form and 3 is a cluster, the CuꢀO bond distances in 1-3 are either
similar or slightly longer, by 0.02ꢀ0.04 Å, than in the monomeric
diaryloxycuprates. The OꢀCuꢀO bond angles are nearly linear
in all of the complexes.40
Figure 7. ORTEP drawing of [K8Cu6(OBu-t)12(DMF)8(I)]+ Iꢀ (3)
with thermal ellipsoids drawn to the 50% probability level and all H
atoms omitted for clarity.
containing 2 equiv of t-BuOM (M = K, Na), in accord with the
fact that unlike potassium and sodium chlorides and bromides,
NaI and KI are easily soluble in DMF at room temperature.36
Although all of the combinations of CuX (X = Cl, Br I) with
t-BuOM (M = K, Na) in a 1:2 molar ratio were found to be
reactive toward CF3H (see below), we suspected that the high
solubility of KI and NaI in DMF might influence the structure
of the cuprating species. Indeed, cooling a solution obtained
by dissolving CuI and t-BuOK (1:2) in DMF at room tempera-
ture prompted crystallization of a remarkable cationic cluster
[K8Cu6(OBu-t)12(DMF)8(I)]+ Iꢀ (3; Figure 7). This high-sym-
metry cation crystallizes in the hexagonal space group R-3c (167)
20905
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Scheme 2
The data presented above account for the ca. 50% yield of
PhCF3 consistently observed in reaction 2. As shown in
Scheme 2, the reaction of CuCl with 1 equiv of t-BuOK produces
[t-BuOCu]4 that is unreactive or poorly reactive toward CF3H. In
the presence of 1 equiv of phen, however, one-half of the Cu(I) is
sequestered in the form of the bis-chelate [Cu(phen)2]+. This
changes the molar ratio of CuCl to t-BuOK in the system to 1:2,
leading to the formation of the dialkoxycuprate that is reactive
toward fluoroform. However, the amount of the CF3Cu species
generated and consequently the yield of the trifluoromethylated
product (PhCF3) of its reaction with PhI theoretically cannot
exceed 50%, as calculated on the basis of the total quantity of Cu(I)
used for the reaction.
Cuprating Reagents Based on CuCl, CuBr, CuI, and Var-
ious MOR (M = K, Na). Stability of the CF3H-Derived CuCF3
Reagents. All possible combinations of CuX (X = Cl, Br, I) with
2 equiv of MOR (M = K, Na; R = Me, Et, t-Bu) in DMF and NMP
at [Cu] = 0.5 M were studied in order to identify the most
efficient reagent for the cupration. Potassium alkoxides were
found to be superior in terms of both the yield of the CuCF3
product and its stability. For instance, the 2MeONa-CuCl system
gave only small quantities of CuCF3 on reaction with fluoroform,
whereas use of KOMe under similar conditions furnished the
desired CuCF3 product in 60ꢀ65% yield. The yield was much
lower however (ca. 10%), when KOMe was replaced with KOEt.
Although Na(DMF)2[Cu(OBu-t)2] generated in situ from CuCl
and t-BuONa in DMF (see above) smoothly reacted with CF3H,
the CuCF3 complex produced was less stable, decomposing
quickly at room temperature.
The best results were obtained with the originally discovered
CuCl-t-BuOK (1:2) system that consistently produced the
CuCF3 species in >90% yield on treatment with CF3H in DMF
or NMP. Slightly lower yields (75ꢀ90%) were achieved when
CuBr was used under identical conditions. With CuI as a Cu(I)
source, the yields were in the range of 40ꢀ85%, and the resultant
CuCF3 solutions were decomposing faster than the ones derived
from CuBr and CuCl. Even faster decomposition was observed
when the reaction of CF3H with CuI-t-BuOK (1:2) in NMP was
carried out in the presence of 2 equiv of KI.
The stability of the CF3Cu generated from CF3H and CuCl-t-
BuOK (1:2) in DMF with [Cu] = 0.5 M at room tempera-
ture was studied by measuring 19F NMR spectra of the freshly pre-
pared reaction solutions containing PhCF3 as an internal standard
(Table 2). As can be seen from this table, the yield of the CF3Cu,
as determined 5ꢀ10 min after its formation (95%), dropped to
approximately 80%, 75%, and 65% after 2, 4, and 18 h, respec-
tively. After that, no significant further decomposition was observed.
Similar results were obtained in NMP: the originally measured 96%
Table 2. Spontaneous Decomposition of CuCF3 Species
Generated from CF3H and CuCl-t-BuOK (1:2) in DMF with
[Cu] = 0.5M at 22ꢀ25 °C
experiment 1a
CuCF3 yield, %
experiment 2a
CuCF3 yield, %
time, h
time, h
0.1
0.7
95
87
84
80
77
76
74
72
64
64
61
0.1
0.5
94
89
85
81
78
76
75
72
65
65
63
1.2
1.0
1.7
1.5
2.5
2.3
3.1
2.9
4.0
3.4
4.2
4.2
17.7
23.5
42.5
17.5
23.4
42.3
a Results of two similar experiments are presented to demonstrate
reproducibility.
yield of the CF3Cu product dropped to 76% after 4 h and further to
68% after 14 h, at which point the decomposition ceased. Addition of
DMI (20% by volume) to DMF before the cupration did not provide
significant stabilization to the CF3Cu product.41
A series of studiesof the decomposition of the CF3Cu solutions
in DMF resulted in the following important observations:
1 A well-pronounced concentration effect on stability of the
CuCF3 product was observed. At [Cu] = 0.75 and 1.0 M the
measured yields were lower (ca. 70ꢀ85%) than at the standard
0.5 M concentration of Cu(I). Also, at higher [Cu], the CuCF3
product decayed more rapidly, suggesting that the decompo-
sition is unlikely to be a unimolecular process.
2 Freshly prepared CuCF3 solutions reacted with PhI to
produce both PhCF3 and t-BuOPh. In contrast, the addition
of excess PhI to the CuCF3 solutions after ca. one-third of
the organocopper complex had decomposed and the de-
composition ceased (see Table 2) led to the formation of
PhCF3 and, at most, only trace quantities of the ether.
3 The decomposition process was accompanied by precipita-
tion of KF (identified by powder X-ray diffraction), which
ceased when the decomposition stopped.
4 As a result of numerous attempts to isolate the CuCF3
species produced in the cupration reaction, four struc-
tures were determined by single-crystal X-ray diffraction:
[Cu2(C(OBu-t)2)2(μ-Cl)2] (4),[Cu2(C(OBu-t)2)2(μ-Br)2] (5),
20906
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Figure 10. ORTEP drawing of [Cu2(C(OBu-t)2)2(μ-Br)2](5) with
thermal ellipsoids drawn to the 50% probability level and all H atoms
omitted for clarity.
Figure 8. Structures of crystallographically characterized complexes
4ꢀ7.
Figure 11. ORTEP drawing of [Cu2(C(OBu-t)2)2(μ-OBu-t)2] (6)
with thermal ellipsoids drawn to the 50% probability level and all H
atoms omitted for clarity.
Figure 9. ORTEP drawing of [Cu2(C(OBu-t)2)2(μ-Cl)2] (4) with
thermal ellipsoids drawn to the 50% probability level and all H atoms
omitted for clarity.
[Cu2(C(OBu-t)2)2(μ-OBu-t)2] (6), and [Cu4(CF3)2-
(C(OBu-t)2)2(μ3-OBu-t)2] (7).42 Figures 8ꢀ12 display struc-
tures and ORTEP drawings of 4-7 that are all bis(t-butoxy)-
carbene Cu(I) species and rare examples of alkoxycarbene
derivatives of copper.43
The results described above indicate unambiguously that the
CuCF3 decomposition is governed by the formation of a Cu(I)
difluorocarbene species that undergo extremely facile44 nucleo-
philic displacement of the fluorines with t-butoxide. The alkali
metal cation present in the solution apparently plays a crucial role
in the decomposition process by binding to the fluoride to form
thermodynamically stable MF (Scheme 3). This scheme also
readily accounts for the aforementioned fact that the CuCF3
products made with NaOR are noticeably less stable than those
derived from KOR because “harder” Na+ has a higher affinity for
fluoride than K+ and is therefore more electrophilic toward the
CuCF3 complex. Furthermore, the decomposition stops after all
of the potassium cations are consumed and precipitated out in
the form of KF. As can be seen from the stoichiometry shown in
Scheme 3, three potassium cations decompose one CuCF3
moiety, which accords with the cessation of the decomposition
after the originally observed nearly quantitative yield of the
CuCF3 drops by one-third to ca. 65% (Table 2). At this point,
t-butoxide is no longer available for the Cu-promoted reaction
with PhI to give t-BuOPh. Scheme 3 also explains the faster
CuCF3 decomposition at higher concentrations that speed up
the abstraction of fluoride by the alkali metal cation.
Figure 12. ORTEP drawing of [Cu4(CF3)2(C(OBu-t)2)2(μ3-OBu-t)2]
(7) with thermal ellipsoids drawn to the 50% probability level and all H
atoms omitted for clarity.
Scheme 3
CF3Cu(I) complex. This bond length is comparable to that
in [(bathophenanthroline)Cu(CF3)] (1.907(9) Å)13e and
much shorter than in [(TMS-IPr)Cu(CF3)] (1.967(6) Å),14a
[(SIiPr)Cu(CF3)] (2.022(4) Å),14a [(Ph3P)3Cu(CF3)] (2.018(7),
2.025(7), and 2.031(10) Å),14d and [(phen)Cu(PPh3)(CF3)]
(1.985(1) Å).14d This trend suggests that the μ3-OBu-t ligand
stabilizing the CuCF3 fragment in 7 is, quite predictably, a con-
siderably weaker electron donor than any of the ligands in the other
structurally characterized CuCF3 complexes. In spite of the lack
of strongly donating ligands, however, our fluoroform-derived
A comment is due on the structure of 7 bearing a CuCF3
fragment (Figure 12). The CuꢀCF3 bond in 7 (1.8908(16) Å) is
one of the shortest ever reported for a structurally characterized
20907
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
CuCF3 compounds are efficient trifluoromethylating reagents
(see below).
Scheme 4
Stabilization of the CuCF3 Reagents. We reasoned that the
spontaneous CuCF3 decomposition might be minimized or
suppressed altogether by treatment of the freshly prepared
solution with an acid HX that would react with the formally
present “one extra equivalent of t-BuOK” to give t-BuOH and
KX. Precipitation of the latter would remove from the solution
the potassium ions that abstract fluorines from the CF3 ligand on
Cu (Scheme 3). As a result of a series of tests using various acidic
compounds, it was found that Et3N 3HF (TREAT HF)45 and
3
Py nHF (70% HF) provide excellent stabilization to the CuCF3
3
solutions. Addition of TREAT HF in the amount of 1/3 mol per
1 mol of Cu after the cupration step resulted in precipitation of
KF and produced stable solutions of the CuCF3 reagent. Upon
treatment with Et3N 3HF, the 19F NMR signal from the CuCF3
3
shifted slightly upfield to ꢀ26.3 ppm from its original value of
ꢀ24.2 ppm, and a new minor singlet resonance appeared in the
CuCF3 region at ꢀ30.4 ppm, which is assigned to [Cu(CF3)2]ꢀ
on the basis of the literature data.8,14,41b In the 13C NMR
spectrum of a solution similarly prepared and stabilized with
TREAT HF in DMF-d7, the CF3 ligand on Cu of the main CuCF3
species resonated as a quartet at 150.7 ppm (JCꢀF = 350 Hz).
In a typical experiment, TREAT HF was added within 5ꢀ15
min after the cupration. Precipitation of KF began instanta-
neously. Quantitative 19F NMR analysis of the resultant solution
showed that the total yield of the trifluoromethylcopper species
produced was 95%. This yield was calculated on the basis of the
CuꢀCF3 bonds formed, with 85% contribution from a mono-
CF3-ligated Cu species (ꢀ26.3 ppm) and 10% from [Cu(CF3)2]ꢀ
(ꢀ30.4 ppm). The molar percentage of the [Cu(CF3)2]ꢀ was,
however, 5% because the Cu atom in this anion bears two CF3
groups. On storage of the solution at room temperature, a minor
drop in the original yield of 95% to 93% and 91% was detected
after 24 h and 3 days, respectively. An aliquot of the same solution
that was stored at ꢀ35 °C for 8 days showed no sign of decom-
position. Characterization of “ligandless” CuCF3 compounds, that is,
those lacking ligands such as NHCs,14a,b phen14c,d and its deriva-
tives,13e or Ph3P,14d is nontrivial.8 As Wiemers and Burton46 con-
cluded in their classical study, CuCF3 is “an elusive and complex
species”. Nonetheless, as shown below, our CF3H-derived CuCF3
can be used as an efficacious trifluoromethylating reagent.
Trifluoromethylation Reactions with CF3H-Derived
CuCF3. The subject of this paper is the new reaction of direct
cupration of fluoroform. A detailed description of various trifluoro-
methylation reactions with our CuCF3 reagents, as well as our
studies of their structure and the mechanism of the cupration, are
beyond the scope of this publication and will be reported separately.
However, to demonstrate the efficiency and versatility of the
reagents, a summary of selected trifluoromethylation reactions using
our fluoroform-derived CuCF3 solutions is presented in Scheme 4.
Our reagents can be used to synthesize other CuCF3 com-
plexes and CF3-derivatives of other metals. For instance, addition
of [IPrH]+ Clꢀ (1 equiv) to a freshly prepared CuCF3 solution in
DMF resulted in instantaneous formation of [(IPr)Cu(CF3)]14a
in quantitative yield. An important starting material for the synthesis
of trifluoromethyl palladium(II) aryls, [(tmeda)Pd(Ph)(CF3)],15b
was produced quantitatively in the reaction of the TREAT HF-
stabilized CuCF3 reagent with [(tmeda)Pd(Ph)(I)]. Organic elec-
trophiles can also be trifluoromethylated with our reagents. 1,1,1-
trifluoroethane was formed from MeI and the CuCF3 at room
temperature. A series of aryl iodides bearing electron-donating and
electron-withdrawing substituents on the ring underwent smooth
trifluoromethylation with the stabilized CuCF3 reagent at 50 °C
in 70ꢀ99% yield. Even 2-bromopyridine was converted to 2-trifluo-
romethylpyridine under these conditions, albeit in 30% yield.
Importantly, these reactions neither require promotion with addi-
tional costly ligands such as phenanthroline, nor produce C2F5-
derivatives as side products that most frequently emerge from
decomposition of various CuCF3 species to difluorocarbene.1ꢀ8
’ CONCLUSIONS
We have discovered direct cupration of fluoroform, the most
attractive CF3 source for trifluoromethylation reactions. The
cupration reaction employs only readily available and inexpensive
reagents, CuCl and t-BuOK in a 1:2 ratio, and furnishes valuable
CuCF3 compounds in nearly quantitative yield. The cupration
occurs within seconds at room temperature and is not mediated by
ꢀ
CF3 or CF2, which accounts for its remarkably high selectivity.
After the formation, the CuCF3 solutions can be efficiently
stabilized with TREAT HF to produce CuCF3 reagents that do
not decay significantly for at least 1ꢀ2 days at room temperature.
These reagents can be used for a variety of reactions to transfer the
CF3 group to another metal/element and for trifluoromethylation
of organic electrophiles.
Our method compares favorably with the one employing
zincation.13d The zincation reaction employs a more costly
metalating agent, zinc bis-2,2,6,6-tetramethylpiperidide, and is
efficient for higher 1H-perfluoroalkane substrates but consider-
ably less so for CF3H. At the temperatures required for Rf transfer
from Zn to Cu (60ꢀ90 °C), the just formed CF3Cu partially
decomposes to give C2F5Cu. As a result, the organic trifluoro-
methylated product is contaminated with the pentafluoroethyl
derivative and needs to be isolated and purified by unconven-
tional means. In the single trifluoromethylation example re-
ported in that work,13d preparative HPLC was required to
obtain the product in pure form (51% yield).
It is also worth noting that our method generates much less
inorganic byproducts than the Swarts-type process8,10 that con-
sumes 3 equiv of Cl2 and coproduces 6 equiv of HCl (chlorine
waste) per each equivalent of ArCF3 made. For larger scale
20908
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
applications of our method, dialkoxycuprate solutions made from
CuCl and t-BuOK could be filtered prior to the cupration step in
order to remove and recycle the precipitated KCl. After the
cupration and stabilization with an HF source, the KF produced
could also be separated by filtration and recycled. We also believe
that a continuous process might be possible on the basis of our
technology. Importantly, trifluoromethylation reactions with our
CuCF3 reagents readily occur in the absence of additional ligands
such as phenanthroline and are not complicated by side-formation
of C2F5 derivatives.
for 30 min at room temperature, the sealed tube was brought out and
evacuated on a vacuum line to ∼1 mmHg.47 At vigorous stirring, fluoroform
was introduced to ∼50 psi, after which the pressure dropped to 5ꢀ10 psi
within a few seconds. 19F NMR analysis of the sample 10ꢀ15 min after the
reaction indicated that a CuCF3 species (ꢀ24.2 ppm) was produced in 95%
yield. Decomposition of the product was monitored by 19F NMR (Table 2).
As the decomposition occurred, the singlet 19F NMR signal from CuCF3 was
slightly shifting upfield from ꢀ24.2 to ꢀ25.6 ppm. Note that the cupra-
tion reaction may be performed in a standard flask by bubbling CF3H
through the dialkoxycuprate solution. However, the developed procedure
in FischerꢀPorter tubes provides better protection of the reagents and
products from accidental exposure to air.
’ EXPERIMENTAL SECTION
Synthesis and Isolation of K(DMF)[Cu(OBu-t)2] (1). A mix-
ture of CuCl (0.50 g; 5.05 mmol) and t-BuOK (1.18 g; 10.54 mmol) in
DMF (5 mL) was vigorously stirred for 30 min at room temperature.
Filtration of the mixture produced solid KCl that was identified by
powder X-ray diffraction after washing with ether (5 mL). The ether
washing and the filtrate were combined and kept at ꢀ35 °C for 20 h. The
white needle-shaped crystals were separated by filtration, washed with
cold ether (2 ꢁ 0.5 mL), and dried under vacuum. The yield was 1.043 g
(64%). The structure of K(DMF)[Cu(OBu-t)2] was established by
single-crystal X-ray diffraction. Elemental analysis could not be per-
formed because of the extreme air- and moisture-sensitivity of the
compound. 1H NMR (C6D6, 25 °C), δ: 7.70 (s, 1H, DMF), 2.41
(s, 3H, DMF), 1.94 (s, 3H, DMF), 1.60 (s, 18H, t-Bu). 13C NMR (C6D6,
25 °C), δ: 162.9 (s, CH, DMF), 69.5 (s, C, t-Bu), 37.3 (s, CH3, t-Bu),
35.7 (s, CH3, DMF), 31.3 (s, CH3, DMF).
Synthesis and Isolation of Na(DMF)2[Cu(OBu-t)2] (2). (a)
DMF (2 mL) was added, at stirring, to a mixture of CuCl (51 mg;
0.51 mmol) and t-BuOK (109 mg; 1.13 mmol). The copper salt quickly
dissolved and a large amount of needle-shaped crystals precipitated out
within 5ꢀ10 s to produce a thick suspension. Gently heating the mixture
to ∼50ꢀ55 °C resulted in dissolution of the white needles. The cloudy
solution was filtered warm to remove insoluble NaCl. The solid-free,
colorless filtrate produced uniformly shaped, very thin and extremely air-
and moisture sensitive white needle crystals on cooling to room
temperature. The structure of the product Na(DMF)2[Cu(OBu-t)2]
was established by single-crystal X-ray diffraction. (b) A mixture of CuCl
(253 mg; 2.56 mmol), t-BuOK (508 mg; 5.29 mmol), and DMF
(10 mL) was stirred at 50ꢀ55 °C for 30 min and filtered warm. The
solid NaCl was washed with ether (5 mL). The ether washing and the
filtrate were combined and kept at ꢀ35 °C for 1.5 h. The white solid was
separated by filtration, washed with ether (0.5 mL), hexanes (15 mL),
and dried under vacuum. The yield was 883 mg (91% if calculated for
Na(DMF)2[Cu(OBu-t)2]). However, 1H NMR indicated that the DMF
to t-BuO ratio in the solid was <1. 1H NMR (THF-d8, 25 °C), δ: 7.92
(s, 1.5H, DMF), 2.88 (s, 4.5H, DMF), 2.77 (s, 4.5H, DMF), 1.18 (s, 18H,
t-Bu). 13C NMR (THF-d8, 25 °C), δ: 162.6 (s, CH, DMF), 68.8 (s, C,
t-Bu), 36.9 (br s, CH3, t-Bu), 36.1 (s, CH3, DMF), 31.2 (s, CH3, DMF).
Preparation of CuCF3 from Fluoroform in NMP. Stabiliza-
tion Effect. A 3-oz FischerꢀPorter tube equipped with a pressure
gauge, a needle valve, and a Teflon-coated magnetic stir-bar was charged
with CuCl (0.50 g; 5.05 mmol), t-BuOK (1.20 g; 10.71 mmol), NMP
(10 mL), and PhF (187 μL; 2.00 mmol; internal standard). After the
mixture was stirred for 30 min at room temperature, the sealed tube was
brought out and quickly evacuated on a vacuum line to ∼1 mmHg.47 At
vigorous stirring, fluoroform was introduced to ∼50 psi, after which the
pressure dropped to 5ꢀ10 psi within a few seconds. 19F NMR analysis of
the sample 10ꢀ15 min after the reaction indicated that a CuCF3 species
(ꢀ24.9 ppm) was produced in 96% yield. Within 5ꢀ10 min, a 5-mL
aliquot of this solution was withdrawn and treated with TREAT HF
(135.5 μL; 0.83 mmol). Decomposition of the CuCF3 in the unstabi-
lized and stabilized solutions was monitored by 19F NMR. Without the
stabilization, the yield of CuCF3 dropped to 76% and 68% after 4 and 14 h,
All chemicals were purchased from Aldrich (CuCl, CuBr, MeONa,
MeOK, EtOK, EtONa, t-BuONa, TREAT HF, Py nHF (70% HF)),
3
Alfa Aesar (CuI, t-BuOK), and Apollo Scientific (CF3H). All manipula-
tions were performed under argon in a glovebox, unless noted otherwise.
Anhydrous DMF (Alfa Aesar or from an MBraun SPS) and NMP (Alfa
Aesar) were used without additional purification. Benzene, toluene,
hexanes, ether, and THF were distilled from Na/OCPh2. All solvents,
internal standards for quantitative 19F NMR analysis, and liquid organic
halide reagents were stored over activated 4 Å molecular sieves in a
glovebox. Glass pressure reaction vessels (FischerꢀPorter tubes) were
purchased from Andrews Glass Co., Inc. 1H, 13C, and 19F NMR spectra
were recorded on Bruker Avance 400 Ultrashield and Bruker Avance 500
Ultrashield NMR spectrometers. Quantitative 19F NMR analyses were
carried out with D1 = 5 s. X-ray powder diffraction analysis was
performed on a Bruker D8 Advance powder diffractometer using a
transmission configuration. An Agilent Technologies 7890A chromato-
graph equipped with a 5975C MSD unit was used for GC-MS analysis.
Reaction of CF3H with CuCl, t-BuOK, phen (1:1:1), and PhI
in DMF. A mixture of CuCl (33 mg; 0.33 mmol), anhydrous 1,10-
phenanthroline (60 mg; 0.33 mmol), potassium t-butoxide (39 mg;
0.35 mmol), and DMF (2 mL) was stirred for 1 h. 4,40-difluorobiphenyl
(16.5 mg; 0.09 mmol) was added as an internal standard and a 0.6 mL
aliquot of the resultant cloudy solution was placed in a standard 5-mm
glass NMR tube. The tube was sealed with a rubber septum and CF3H
(20 mL; 0.89 mmol) was bubbled through the solution via a syringe
needle. After the addition of CHF3 iodobenzene (100 μL; 0.90 mmol)
was added and the tube was placed in the probe of an NMR spectrometer
for monitoring at 25 °C. Theresults are summarizedin Figure 1, displaying
the formation of a CuCF3 species and its transformation to PhCF3 in
∼45ꢀ50% yield, as calculated on the amount of CuCl used.
Cupration Experiments in NMR Tubes. (a) A mixture of CuCl
(50 mg; 0.51 mmol), t-BuOK (118 mg; 1.05 mmol), DMF (1 mL), and
PhF (50 μL; 0.53 mmol; internal standard) was stirred for 30 min. A
0.6 mL aliquot of the resultant cloudy solution was placed in a 5-mm
NMR tube that was then capped with a rubber septum and brought out.
Fluoroform gas (15 mL; 0.67 mmol) was bubbled through the solution
via a syringe needle. The reaction was complete within seconds. 19F
NMR analysis of the sample 10ꢀ15 min after the reaction indicated that
a CuCF3 species (ꢀ24.0 ppm) was produced in 90ꢀ95% yield (e.g., see
Figure 2). No other 19F NMR-detectable products were observed.
(b) Similar results were obtained in NMP. (c) Repeating the experiment
with CuCl in the presence of styrene (2 equiv) produced identical
results; neither gem-difluorophenylcyclopropane nor any other side
products were detected. (d) No formation of 1,1-difluoro-2-methyl-2-
phenylcyclopropane was observed in a repeat using CuI as the Cu(I)
source and α-methylstyrene (2 equiv) as a CF2 trap.
Preparation of CuCF3 from Fluoroform in DMF. Stability
Studies. A 3-oz FischerꢀPorter tube equipped with a pressure gauge, a
needle valve, and a Teflon-coated magnetic stir-bar was charged with CuCl
(0.50 g; 5.05 mmol), t-BuOK (1.19 g; 10.62 mmol), DMF (10 mL), and
PhCF3 (123μL; 1.00 mmol; internal standard). After the mixture was stirred
20909
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Table 3. Aromatic Trifluoromethylation Reactions with CF3H-Derived CuCF3 Reagents at 50 °Ca
a See the Experimental Section for specifics.
respectively. In contrast, the total yield of CuCF3 (ꢀ26.9 ppm)
including 3% of [Cu(CF3)2]ꢀ (ꢀ31.0 ppm) in the stabilized solution
after 4 and 14 h was 92% and 90%, respectively.
equipped with a pressure gauge, a needle valve, and a Teflon-coated
magnetic stir-bar was charged with CuCl (0.50 g; 5.05 mmol), t-BuOK
(1.18 g; 10.54 mmol), DMF (10 mL), and PhF (187 μL; 2.00 mmol;
internal standard). After the mixture was stirred for 30 min at room
temperature, the sealed tube was brought out and quickly evacuated on a
vacuum line to ∼1 mmHg.47 At vigorous stirring, fluoroform was
introduced to ∼50 psi, after which the pressure dropped to 5ꢀ10 psi
within a few seconds. After 15 min, a solution of TREAT HF (272 μL; 1.67
mmol) in DMF (1 mL) was added. By 19F NMR, the yield of the CuCF3
in the stabilized solution was 98% (including 3% of [Cu(CF3)2]ꢀ;
ꢀ31.0 ppm). This solution was used in a series of trifluoromethylation
reactions. For each aromatic trifluoromethylation reaction, to 0.6 mL of
this solution (0.21 mmol of CuCF3) placed in a 5-mm NMR tube was
added an organic halide (0.32 mmol) and the mixture was heated at
50 °C (oil bath). Results of these experiments are summarized in
Table 3. In another experiment, 0.1 mL of this solution (0.035 mmol
of CuCF3) was added to [(tmeda)Pd(Ph)(I)] (15 mg; 0.035 mmol) in
DMF (0.5 mL). After 5.5 h at room temperature, full conversion to
[(tmeda)Pd(Ph)(CF3)]15b was observed by19F NMR(δ = ꢀ19.8 ppm).
In another experiment, to 0.6 mL of the CuCF3 solution (0.21 mmol)
was added CH3I (80 μL, 6-fold excess). After 1.5 days at room
temperature, full conversion of the CuCF3 reagent was observed by
19F NMR. The yield of CF3CH3 (ꢀ59.6 ppm, q, JFꢀH = 13 Hz) present
in the liquid phase was 28%. The total yield could not be determined
because of the distribution of the 1,1,1-trifluoroethane product (bp =
ꢀ47 °C) between the solution and the gas phase.
Preparation of CuCF3 from Fluoroform in DMF. Stabiliza-
tion Effect. A 3-oz FischerꢀPorter tube equipped with a pressure
gauge, a needle valve, and a Teflon-coated magnetic stir-bar was charged
with CuCl (0.50 g; 5.05 mmol), t-BuOK (1.18 g; 10.54 mmol), DMF
(10 mL), and PhF (187 μL; 2.00 mmol; internal standard). After the
mixture was stirred for 30 min at room temperature, the sealed tube was
brought out and quickly evacuated on a vacuum line to ∼1 mmHg.47 At
vigorous stirring, fluoroform was introduced to ∼50 psi, after which the
pressure dropped to 5ꢀ10 psi within a few seconds. After 5 min, a solution of
TREAT HF (272 μL; 1.67 mmol) in DMF (1 mL) was added. A 1-mL
aliquot was withdrawn and placed in a freezer at ꢀ35 °C. By 19F NMR, the
yield of the CuCF3 in the stabilized solution stored at room temperature
under argon was 94%, 92%, 90%, and 89% (including 5% of [Cu(CF3)2]ꢀ,
ꢀ30.4 ppm) after 0.2, 13, 60, and 80 h, respectively. In the 1-mL sample of
the same stabilized solution stored at ꢀ35 °C the yield was 93% after 80 h.
Stabilization of the CuCF3 Reagent with TREAT HF in
Excess. A 3-oz FischerꢀPorter tube equipped with a pressure gauge,
a needle valve, and a Teflon-coated magnetic stir-bar was charged with
CuCl (0.50 g; 5.05 mmol), t-BuOK (1.18 g; 10.54 mmol), DMF
(10 mL), and PhF (187 μL; 2.00 mmol; internal standard). After the
mixture was stirred for 30 min at room temperature, the sealed tube was
brought out and quickly evacuated on a vacuum line to ca. 1 mmHg.47 At
vigorous stirring, fluoroform was introduced to ∼50 psi, after which the
pressure dropped to 5ꢀ10 psi within a few seconds. After 5 min, a
solution of TREAT HF (326 μL; 2.00 mmol) in DMF (1 mL) was
added. By 19F NMR, the yield of the CuCF3 in the stabilized solution
stored at room temperature under argon was 97%, 92%, and 89%
(including 3% of [Cu(CF3)2]ꢀ, ꢀ30.4 ppm) after 0.2, 10, and 38 h,
respectively. This experiment shows that use of TREAT HF in 20% excess
provides neither extra stabilization, nor destabilization to the CuCF3 reagent.
Trifluoromethylation Reactions with Stabilized CuCF3
Reagents Derived from CF3H. A 3-oz FischerꢀPorter tube
Reaction of CF3H-Derived CuCF3 with IPrH+ Clꢀ. A mixture
of CuCl (50 mg; 0.51 mmol), t-BuOK (118 mg; 1.05 mmol), DMF-d7
(1.3 mL), and PhF (187 μL; 2 mmol; internal standard) was stirred for
30 min. A 0.65 mL aliquot of the resultant cloudy solution was placed in a
5-mm NMR tube that was then capped with a rubber septum and
brought out. Fluoroform gas (15 mL; 0.67 mmol) was bubbled through
the solution via a syringe needle. 19F NMR analysis of the sample 10ꢀ15
min after the reaction indicated that a CuCF3 species (ꢀ23.0 ppm)
was formed in >95% yield. Treatment of this solution with IPrH+ Clꢀ
20910
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
Table 4. Crystallographic Data for Complexes 1ꢀ7
1
2
3
4
5
6
7
Formula
C11H25Cu1-
K1N1O3
321.96
C14H32Cu1-
N2Na1O4
378.95
C72H164Cu6-
I2K8N2O20
2409.95
C18H36Cl2Cu2O4
C18H36Br2Cu2O4
C26H54Cu2O6
C28H54Cu4F6O6
Formula weight
Crystal size (mm)
Crystal color
Temp (K)
Crystal system
Space group
A (Å)
B (Å)
C (Å)
α (deg)
514.45
603.37
589.77
854.87
0.30 ꢁ 0.03 ꢁ 0.03 0.30 ꢁ 0.06 ꢁ 0.03 0.10 ꢁ 0.10 ꢁ 0.10 0.30 ꢁ 0.20 ꢁ 0.10 0.20 ꢁ 0.20 ꢁ 0.10 0.40 ꢁ 0.20 ꢁ 0.02 0.40 ꢁ 0.30 ꢁ 0.30
colorless
100
monoclinic
P21/n
10.1264(16)
6.3330(10)
25.079(4)
90
colorless
100
monoclinic
P21/c
6.202(2)
19.855(8)
16.646(7)
90
yellow
100
colorless
100
colorless
100
monoclinic
P21/c
11.3451(10)
21.5829(19)
11.1733(10)
90
yellow
100
yellow
100
trigonal/hexagonal monoclinic
R 3c
17.754(2)
17.754(2)
62.184(7)
90
monoclinic
P21/c
11.8480(10)
8.5446(7)
16.2194(14)
90
monoclinic
P21/n
11.8758(3)
10.8214(3)
14.3329(3)
90
P21/c
11.4005(9)
21.0308(17)
11.1890(9)
90
β (deg)
96.011(7)
90
1599.5(4)
4
1.337
1.623
29.93
64609
3987
95.483(14)
90
2040.3(13)
4
1.234
1.107
26.91
20546
2679
90
120
16974(4)
6
1.415
2.003
24.76
28860
2221
115.130(2)
90
2428.8(3)
4
1.407
1.989
29.99
65084
5937
114.292(4)
90
2493.7(4)
4
1.607
4.928
33.36
23735
6535
109.051(4)
90
1552.1(2)
2
1.262
1.403
33.58
42123
5446
96.8830(10)
90
1828.69(8)
2
1.553
2.360
39.57
26332
7677
γ (deg)
V (Å3)
Z
F (g/cm3)
µ (mmꢀ1
)
θmax (deg)
Reflec. measured
Unique
reflections obv.
Absorpt. correct.
Trans. min/max
Parameters
[Rint = 0.0320]
TWINABS
0. 67/1.00
161
[Rint = 0.0887]
[Rint = 0.0554]
[Rint = 0.0341]
SADABS
0. 81/0.98
256
[Rint = 0.0523]
SADABS
0.69/1.00
247
[Rint = 0.0582]
SADABS
0.82/0.98
209
[Rint = 0.0330]
SADABS
0.84/1.00
236
SADABS
0.81/0.98
209
SADABS
0.98/1.00
250
R1/wR2 [I>2σ(I)] 0.0450/0.1100
R1/wR2 [all data] 0.0545/0.1148
Goodness-of-fit (F2) 1.125
Peak/hole (e/Å3)
0.575/-0.664
0.0456/0.0785
0.0961/0.0923
1.007
0.0576/0.1536
0.0872/0.1722
1.016
0.0266/0.0665
0.0326/0.0700
1.044
0.0437/0.1005
0.0686/0.1104
1.040
0.0868/0.2317
0.0945/0.2367
1.225
0.0362/0.0981
0.0513/0.1078
1.025
0.372/-0.351
1.556/-1.838
0.560/-0.334
1.321/-0.710
1.499/-1.425
1.386/-1.150
(105 mg; 0.25 mmol) in DMF-d7 (0.3 mL) produced [(IPr)Cu(CF3)]14a
quantitatively. H NMR (DMF-d7, 25 °C), δ: 7.98 (s, 2H, IPr), 7.60
(t, 2H, J = 7.7 Hz, arom CH), 7.48 (d, 4H, J = 7.7 Hz, arom CH), 2.64
(m, 4H, J = 7.0 Hz, i-Pr CH), 1.32 (d, 12H, J = 7.0 Hz, i-Pr CH3), 1.27
(d, 12H, J = 7.0 Hz, i-Pr CH3). Signals from t-BuOH were also observed in
the 1H NMR spectrum at 4.23 ppm (s, 2H, OH) and at 1.19 ppm (s, 18H,
CH3). 19F NMR (DMF-d7, 25 °C), δ: ꢀ30.7 (s, CuCF3).
using the XP program. Missing atoms were subsequently located from
difference Fourier synthesis and added to the atom list. Least-squares
refinement on F2 using all measured intensities was carried out using
SHELXTL. All non-hydrogen atoms were refined, including anisotropic
displacement parameters. A summary of the crystallographic data is pre-
sented in Table 4. The twin crystal of 1 was processed with SAINT
(simultaneous data processing) and TWINABS for absorption correction.
The structure was refined using the overlapping reflections (crystal ratio =
90:10). High-symmetry 3 crystallizes in the trigonal/hexagonal space group
R 3c. The center of the cluster (I1) is located on a 3 axis and only 1/6 of the
complex was refined and the rest generated by symmetry operations. The
outer-sphere iodine with an occupation of 1/6 is disordered in three
positions around another 3 axis. Two different DMF molecules attached
to the potassium atoms 1 and 2 are disordered in different orientations
surrounding the iodide anion. Complexes 4 and 5 are isostructural, except
one of the t-Bu groups is disordered in two orientations in 4(ratio of disorder
= 70:30). Dimer 6 (monoclinic, space group P21/c) was refined for a
disordered model with two different positions and an occupancy ratio of
87:13. As both disordered positions exhibited Ci-symmetry, the correspond-
ing half-molecules were refined. The less occupied position that is shifted in
respect to the principal position by 0.5 axis units is likely because of the
presence of a second smaller crystal forming a symmetry related twin with
the main crystal. This position appeared as an electron density shadow
overlapping the main structure. After the refinement using this model, the R1
value lowered from 12.3% to 8.7%. The crystal of 7 had Ci-symmetry and the
corresponding half molecule was refined. The CF3 group in 7is disordered in
two positions with occupancy factors of 62% and 38%.
1
Single Crystal X-ray Structure Determination. Crystal
Handling and Mounting. The extremely air- and moisture-sensitive
crystals of 1 and 2 were picked up using 200 μm MiTeGen Capton
MicroMounts (M1-L19ꢀ200) and mounted on Crystalcap Copper
Magnetic holders (Hampton Research) inside a glovebox, then quickly
dipped into liquid nitrogen. The crystals were then directly mounted on
the goniometer at 100 K using Hampton Research Cryo Tools: a CrystalCap
Copper magnetic handling tool, a Crystalwand vial clamp for sample
manipulation, and a Cryotong for crystal transfer at cryotemperatures. Air-
sensitive crystals of 3ꢀ7 were coated with perfluoropolyether inside a
glovebox for protection from air prior to mounting on the goniometer.
Data Collection. Crystal structures of 1, 3, 4, 5, and 7 were deter-
mined using a Bruker-Nonius diffractometer equipped with an APEX II
4K CCD48 area detector, a FR591 rotating anode with MoKα radiation,
Montel mirrors as monochromator, a Kappa 4-axis goniometer, and an
Oxford Cryosystem Plus low temperature device (T = ꢀ173 °C). Crystal
structures of 2 and 6 were determined using an APEX DUO Kappa 4-axis
goniometer equipped with an APEX II 4K CCD area detector, two
Microfocus E025 IμS X-ray sources (MoKα and CuKα), Quazar MX
multilayer optics, and an Oxford Cryosystem Plus low temperature
device (T = ꢀ173 °C). Full-sphere data collection was used with ω and j
scans. The following programs were used: APEX-2 (data collection),48a
Bruker Saint V/.60A (data reduction),48b and SADABS or TWINABS
(absorption correction).48c
’ ASSOCIATED CONTENT
S
Supporting Information. Complete ref 41b (PDF) and full
b
details of crystallographic studies for complexes 1ꢀ7 (CIF). This ma-
Structure Solution and Refinement. The structures were solved
using direct methods as implemented in SHELXTL48d and visualized
terial is available free of charge via the Internet at http://pubs.acs.org.
20911
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
’ AUTHOR INFORMATION
(18) Rozen, S.; Hagooly, A. e-EROS Encyclopedia of Reagents for
Organic Synthesis, http://www3.interscience.wiley.com/cgi-bin/mrwhome/
104554785/HOME, 2001.
(19) Symons, E. A.; Clermont, M. J. J. Am. Chem. Soc. 1981, 103, 3127.
(20) Folleas, B.; Marek, I.; Normant, J.-F.; Saint-Jalmes, L. Tetra-
hedron 2000, 56, 275.
Corresponding Author
vgrushin@iciq.es
(21) (a) Shono, T.; Ishifune, M.; Okada, T.; Kashimura, S. J. Org.
Chem. 1991, 56, 2. (b) Folleas, B.; Marek, I.; Normant, J.-F.; Saint-
Jalmes, L. Tetrahedron Lett. 1998, 39, 2973. (c) Russell, J.; Roques, N.
Tetrahedron 1998, 54, 13771. (d) For an overview of synthetic methods
based on deprotonation of CF3H, see: Langlois, B. R.; Billard, T. ACS
Symp. Ser. 2005, 911, 57.
’ ACKNOWLEDGMENT
We thank Prof. Vladimir I. Bakhmutov for selected NMR
data treatment and Fedor M. Miloserdov for valuable comments.
The ICIQ Foundation and Consolider Ingenio 2010 (Grant
CSD2006-0003) are thankfully acknowledged for support.
(22) (a) Dolbier, W. R., Jr.; Wojtowicz, H.; Burkholder, C. R. J. Org.
Chem. 1990, 55, 5420. (b) Dolbier, W. R. Guide to Fluorine NMR for
Organic Chemists; Wiley & Sons: Hoboken, NJ, 2009.
(23) These are the averaged values obtained from the two rate
constants, 41.7 and 83.3 sꢀ1, and ΔGq = 15.2 and 14.8 kcal molꢀ1, as
derived from the line-shape analysis. While both the 1H and 19F NMR
spectra (Figures 2 and 3) could be quickly recorded for a freshly
prepared sample, the 13C NMR experiment took hours to observe the
CuCF3 resonance (Figure 4). As a result of partial decomposition (see
text) during the long data acquisition, the resonances from the two
nonequivalent t-BuO groups displayed deviation from the original 1:1
integral ratio to ∼2:1. It is also likely that the t-BuO groups of the
carbene ligand C(OBu-t)2 emerging from the decomposition are also
involved in the exchange via degenerate nucleophilic substitution at the
Cu-coordinated carbene carbon. For details, see the subsection entitled
“Cuprating reagents based on CuCl, CuBr, CuI, and various MOR (M =
K, Na). Stability and decomposition of the CuCF3 reagents.”
(24) (a) Shapiro, I. O.; Bogachev, Yu. S.; Bulatova, N. P.; Shapet’ko,
N. N.; Shatenshtein, A. I. Zh. Obsh. Khim. 1990, 60, 244. (b) Chisholm,
M. H.; Drake, S. R.; Naiini, A. A.; Streib, W. E. Polyhedron 1991,
10, 337.
(25) (a) It has been reported that a solution containing [t-BuOCu]4
and t-BuOH (1:4) in C6D6 displays one singlet resonance at 1.16
ppm.25b (b) Geerts, R. L.; Huffman, J. C.; Folting, K.; Lemmen, T. H.;
Caulton, K. G. J. Am. Chem. Soc. 1983, 105, 3503.
(26) Tsuda, T.; Hashimoto, T.; Saegusa, T. J. Am. Chem. Soc. 1972,
94, 958.
(27) Whitesides, G. M.; Sadowski, J. S.; Lilburn, J. J. Am. Chem. Soc.
1974, 96, 2829.
(28) Greiser, T.; Weiss, E. Chem. Ber. 1976, 109, 3142.
(29) Lemmen, T. H.; Goeden, G. V.; Huffman, J. C.; Geerts, R. L.;
Caulton, K. G. Inorg. Chem. 1990, 29, 3680.
’ REFERENCES
(1) For selected monographs, see: (a) Clark, J. H.; Wails, D.;
Bastock, T. W. Aromatic Fluorination; CRC Press: Boca Raton, FL,
1996. (b) Kirsch, P. Modern Fluoroorganic Chemistry; Wiley-VCH:
Weinheim, Germany, 2004. (c) Uneyama, K. Organofluorine Chemistry;
Blackwell: Oxford, U. K., 2006. (d) Ojima, I. Fluorine in Medicinal
Chemistry and Chemical Biology; Wiley-Blackwell: Chichester, U. K.,
2009.
(2) Burton, D. J.; Yang, Z. Y. Tetrahedron 1992, 48, 189.
(3) McClinton, M. A.; McClinton, D. A. Tetrahedron 1992, 32, 6555.
(4) Burton, D. J.; Lu, L. Top. Curr. Chem. 1997, 193, 45.
(5) Schlosser, M. Angew. Chem., Int. Ed. 2006, 45, 5432.
(6) Roy, S.; Gregg, B. T.; Gribble, G. W.; Le, V.-D.; Roy, S.
Tetrahedron 2011, 67, 2161.
(7) Qing, F.-L.; Zheng, F. Synlett 2011, 1052.
(8) Tomashenko, O. A.; Grushin, V. V. Chem. Rev. 2011, 111, 4475.
(9) For selected reviews, see: (a) Kirsch, P.; Bremer, M. Angew.
Chem., Int. Ed. 2000, 39, 4216. (b) Hird, M. Chem. Soc. Rev. 2007,
36, 2070. (c) Berger, R.; Resnati, G.; Metrangolo, P.; Weber, E.; Hulliger,
J. Chem. Soc. Rev. 2011, 40, 3496.
(10) Filler, R. Adv. Fluorine Chem. 1970, 6, 1.
(11) (a) McLoughlin, V. C. R.; Thrower, J. U.S. Patent 3408411,
1968. (b) McLoughlin, V. C. R.; Thrower, J. Tetrahedron 1969, 25, 5921.
(12) (a) Kobayashi, Y.; Kumadaki, I. Tetrahedron Lett. 1969, 4095.
(b) Sato, K.; Tarui, A.; Omote, M.; Ando, A.; Kumadaki, I. Synthesis
2010, 1865.
(13) (a) Inoue, M.; Araki, K. Jpn. Patent JP 2009-234921, 2009.
(b) Oishi, M.; Kondo, H.; Amii, H. Chem. Commun. 2009, 1909.
(c) Knauber, T.; Arikan, F.; R€oschenthaler, G.-V.; Goossen, L. J. Chem.—
Eur. J. 2011, 17, 2689. (d) Popov, I.; Lindeman, S.; Daugulis, O. J. Am.
Chem. Soc. 2011, 133, 9286. (e) Weng, Z.; Lee, R.; Jia, W.; Yuan, Y.;
Wang, W.; Feng, X.; Huang, K.-W. Organometallics 2011, 30, 3229.
(14) (a) Dubinina, G. G.; Furutachi, H.; Vicic, D. A. J. Am. Chem. Soc.
2008, 130, 8600. (b) Dubinina, G. G.; Ogikubo, J.; Vicic, D. A.
Organometallics 2008, 27, 6233. (c) Morimoto, H.; Tsubogo, T.;
Litvinas, N. D.; Hartwig, J. F. Angew. Chem., Int. Ed. 2011, 50, 3793.
(d) Tomashenko, O. A.; Escudero-Adan, E. C.; Martinez Belmonte, M.;
Grushin, V. V. Angew. Chem., Int. Ed. 2011, 50, 7655.
(15) (a) Grushin, V. V.; Marshall, W. J. J. Am. Chem. Soc. 2006,
128, 12644. (b) Grushin, V. V.; Marshall, W. J. J. Am. Chem. Soc. 2006,
128, 4632. (c) Grushin, V. V. Acc. Chem. Res. 2010, 43, 160.
(16) Cho, E. J.; Senecal, T. D.; Kinzel, T.; Zhang, Y.; Watson, D. A.;
Buchwald, S. L. Science 2010, 328, 1679. Also, see: Samant, B. S.;
Kabalka, G. W. Chem. Commun. 2011, 47, 7236.
(17) (a) Pd-catalyzed trifluoromethylation via cyclopalladation with
an electrophilic CF3 source17b and stoichiometric reactions of Ar-CF3
reductive elimination from Pd(IV)17cꢀe were also recently reported.
These transformations, however, are unrelated to the area of nucleophilic
trifluoromethylation of aryl halides because they do not involve Pd(0)
that is required for Ar-X bond activation. (b) Wang, X.; Truesdale, L.;
Yu, J.-Q. J. Am. Chem. Soc. 2010, 132, 3648. (c) Ball, N. D.; Kampf, J. W.;
Sanford, M. S. J. Am. Chem. Soc. 2010, 132, 2878. (d) Ye, Y.; Ball, N. D.;
Kampf, J. W.; Sanford, M. S. J. Am. Chem. Soc. 2010, 132, 14682. (e) Ball,
N. D.; Gary, J. B.; Ye, Y.; Sanford, M. S. J. Am. Chem. Soc. 2011, 133, 7577.
(30) Hakansson, M.; Lopes, C.; Jagner, S. Inorg. Chim. Acta 2000,
304, 178.
(31) Cornforth, J.; Sierakowski, A. F.; Wallace, T. W. J. Chem. Soc.,
Chem. Commun. 1979, 294.
(32) Coult, R.; Fox, M. A.; Gill, W. R.; Herbertson, P. L.; MacBride,
J. A. H.; Wade, K. J. Organomet. Chem. 1993, 462, 19.
(33) Van der Sluys, L. S.; Miller, M. M.; Kubas, G. J.; Caulton, K. G.
J. Am. Chem. Soc. 1991, 113, 2513.
(34) Meyer, C.; Grutzmacher, H.; Pritzkow, H. Angew. Chem., Int.
Ed. Engl. 1997, 36, 2471.
(35) Orlov, N. A.; Bochkarev, L. N.; Nikitinsky, A. V.; Kropotova, V.
Yu.; Zakharov, L. N.; Fukin, G. K.; Khorshev, S. Ya. J. Organomet. Chem.
1998, 560, 21.
(36) (a) Pistoia, G.; Pecci, G.; Scrosati, B. Ric. Sci. 1967, 37, 1167.
(b) Varlamova, T. M.; Gerasimova, G. V.; Mushtakova, S. P. Russ. J. Phys.
Chem. 2006, 80, 1671.
(37) Purdy, A. P.; George, C. F. Polyhedron 1995, 14, 761.
(38) (a) Fiaschi, P.; Floriani, C.; Pasquali, M.; Chiesi-Villa, A.; Guastini,
C. J. Chem. Soc., Chem. Commun. 1984, 888. (b) Fiaschi, P.; Floriani, C.;
Pasquali, M.; Chiesi-Villa, A.; Guastini, C. Inorg. Chem. 1986, 25, 462.
(39) Tye, J. W.; Weng, Z.; Giri, R.; Hartwig, J. F. Angew. Chem., Int.
Ed. 2010, 49, 2185.
(40) (a) In the 1H NMR spectra of DMF-d7 solutions of CuX (X =
Cl, Br, I) and t-BuOM (M = K, Na) (1:2) only one singlet from the
20912
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913

Journal of the American Chemical Society
ARTICLE
t-BuO group was observed. Although exact structures of 1-3 in solu-
tion are unknown and might be different from the solid state ones, all
three cuprates likely remain aggregated in solution. Note that t-BuOM
(M = K, Na) are clusters and aggregates in the solid state and in solution
(see refs 40bꢀ40d for selected reports), exhibiting conductance in
DMF that is too small for reliable measurements.40e (b) Schmidt, P.;
Lochmann, L.; Schneider, B. J. Mol. Struct. 1971, 9, 403. (c) Kissling,
R. M.;Gagne, M. R.J. Org. Chem. 2001, 66, 9005. (d) Nekola, H.; Olbrich, F.;
Behrens, U. Z. Anorg. Allg. Chem. 2002, 628, 2067. (e) Kolthoff, I. M.;
Chantooni, M. K., Jr. Anal. Chem. 1979, 51, 1301.
(41) (a) DMI (1,3-dimethyl-2-imidazolidinone) has been used to
stabilize CuCF3 reagents prepared from CuBr, KF, and CF3SiMe3 in
DMF.41b In one of our experiments, we observed a slightly higher yield
(∼80%) of the CuCF3 product in the cupration reaction with CuCl and
t-BuOK (1:2) in 4:1 (v/v) DMF/DMI at [Cu] = 1.0 M, as compared
with ca. 70ꢀ75% yield obtained in a similar experiment in pure DMF.
However, the CuCF3 products in both solutions decomposed at about
the same rate. (b) Kuett, A.; et al. J. Org. Chem. 2008, 73, 2607.
(42) The formation of [Cu2(C(OBu-t)2)2(μ-Cl)2] (4) and
[Cu2(C(OBu-t)2)2(μ-Br)2](5) isolated from the experiments employ-
ing CuCl and CuBr, respectively, is explained by the fact that although
KCl and KBr are poorly soluble in DMF,36 small quantities of these
halides still remain in the solution after the formation of 1 and 2. Both 4
and 5 are less soluble than their t-BuO analogue 6 and hence crystallize
out more readily.
(43) (a) Barluenga, J.; Lopez, L. A.; Lober, O.; Tomas, M.; Garcia-
Granda, S.; Alvarez-Rua, C.; Borge, J. Angew. Chem., Int. Ed. 2001,
40, 3392. (b) Sivasankar, C.; Baskaran, C.; Samuelson, A. G. J. Chem. Sci.
2006, 118, 237. (c) Bellemin-Laponnaz, S. Polyhedron 2010, 29, 30.
(44) For reviews, see: (a) Brothers, P. J.; Roper, W. R. Chem. Rev.
1988, 88, 1293. (b) Torrens, H. Coord. Chem. Rev. 2005, 249, 1957.
(c) Hughes, R. P. Eur. J. Inorg. Chem. 2009, 4591.
(45) For general reviews of TREAT HF, see: (a) McClinton, M. A.
Aldrichim. Acta 1995, 28, 31. (b) Haufe, G. J. Prakt. Chem. 1996, 338, 99.
(46) Wiemers, D. M.; Burton, D. J. J. Am. Chem. Soc. 1986, 108, 832.
(47) To avoid evaporation and losses of the volatile external
standard (PhCF3 or PhF) the mixture was kept under vacuum for no
longer than 15 s. In a control experiment, a DMF (10 mL) solution was
prepared, containing PhF at the concentration used in the cupration
reactions and CF3COOK as a nonvolatile 19F NMR internal standard.
The solution was then placed in the FischerꢀPorter tube used for the
cupration reactions and kept under vacuum (ca. 1 mmHg) for 15 s. No
detectable loss of PhF was detected by 19F NMR analysis. After the
solution was kept under vacuum on the same line for 10 min, 19F NMR
analysis indicated ca. 10% loss of PhF.
(48) (a) Data collection with APEX II, versions v1.0-22, v2009.1-0, and
v2009.1-02 (2007); Bruker AXS Inc.: Madison, WI, U.S.A. (b) Data
reduction with Bruker SAINT, versions V.2.10 (2003), V/.60A, and V7.60A
(2007); Bruker AXS Inc., Madison, WI, U.S.A. (c) Blessing, R. H. Acta
Cryst. A 1995, 51, 33; SADABS, V.2.10(2003), V2008, and V2008/1
(2001); TWINABS, version 2008/4; Bruker AXS Inc.: Madison, WI,
U.S.A. (d) Sheldrick, G. M. Acta Cryst. A 2008 A64, 112; SHELXTL,
versions V6.12 and 6.14.
20913
dx.doi.org/10.1021/ja2081026 |J. Am. Chem. Soc. 2011, 133, 20901–20913