A persistent α-fluorocarbanion and its analogues: Preparation, characterization, and computational study

Full text of DOI:10.1002/anie.200901414

Communications
DOI: 10.1002/anie.200901414
Carbanions
A Persistent a-Fluorocarbanion and Its Analogues: Preparation,
Characterization, and Computational Study**
G. K. Surya Prakash,* Fang Wang, Nan Shao, Thomas Mathew, Golam Rasul, Ralf Haiges,
Timothy Stewart, and George A. Olah*
Fluoromethylated organics are of substantial interest in
pharmaceutics, biochemistry, and materials science owing to
their unique properties.[1] Nucleophilic fluoromethylation
reactions have been used as major synthetic strategies for
introducing fluoromethyl building blocks into organic com-
pounds.[2] Recently, a-fluoro(phenylsulfonyl)methane (FSM)
derivatives have been extensively studied as monofluoro-
methyl pronucleophiles in numerous reactions (Scheme 1).[3]
Scheme 1. Nucleophilic monofluoromethylation reactions of an electro-
phile (E+) by a-fluoro(phenylsulfonyl)methane (FSM) derivatives.
Such nucleophilic fluoromethylation reactions involve
Figure 1. The bis(phenylsulfonyl)methide ion (I) and its derivatives.
a-fluorocarbanions (FCAs) with unique characteristics as
the key intermediates. Interestingly, a number of computa-
tional studies on a-FCAs have shown that they are electroni-
cally, thermodynamically, and sterically different from their
parent analogues as a result of the specific electronic effects of
fluorine.[4]
Carbanions usually have pyramidal structures, however
they deviate from this and assume planar structures when in
conjugation with adjacent nitro, carbonyl, or phenylsulfonyl
groups.[4,5] The a-fluorine atom, owing to its high electro-
negativity, can stabilize the carbanions through its strong
electron withdrawing effect. However, repulsive force of the
vicinal anion–lone pair (derived from the adjacent anionic
carbon center and non-bonded 2p lone pairs on the a-fluorine
atom) makes fluorine less stabilizing than chlorine, bromine,
or some other electron-withdrawing groups (Figure 1) as
shown by ready a-fluoride elimination which can lead to
complex products.[6] In fact, fluorine is found to destabilize
a-carbanions in many cases.[6–8] Consequently, the structural
investigation of FCAs is still a challenge. To our knowledge,
a-monofluorocarbanions have not yet been structurally
characterized, despite the reports of many computational
studies on fluorine-bearing carbanions. Herein, we disclose,
for the first time, the preparation, NMR spectroscopic and
X-ray structural characterization, and computational studies
of the persistent a-fluorobis(phenylsulfonyl)methane
(FBSM) anion and its analogues (Figure 1, I–V) as a part of
our investigations of the synthetic and mechanistic aspects of
nucleophilic monofluoromethylations.
FBSM has been found to be an effective reagent for
nucleophilic monofluoromethylations and is superior to
(phenylsulfonyl)monofluoromethane (PhSO2CH2F) in many
cases. Hu et al. pointed out that the additional phenylsulfonyl
group not only stabilizes the formed carbanion, but also
increases the nucleophilicity of the fluorinated carbanion by
its softening ability.[3b] The stabilization of the carbanion by
two phenylsulfonyl groups is supported by the following
observations: a) the parent carbanion I (Figure 1), studied by
Henderson and others,[9] has been readily prepared and
proved to be highly thermostable, b) Dubenko et al.[10] have
reported the preparation of potassium salts of IV and V,
c) later, Ochiai and co-workers[11] reported that brominated
bis(perfluoroalkylsulfonyl) carbanion VI also can be crystal-
lized as a fairly stable solid. Based on these studies, we
decided to prepare the fluorine-bearing carbanion II and
study its stability under suitable conditions.
[*] Prof. Dr. G. K. S. Prakash, F. Wang, N. Shao, Dr. T. Mathew,
Prof. Dr. G. Rasul, Dr. R. Haiges, T. Stewart, 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)
Fax: (+1)213-740-6679
E-mail: gprakash@usc.edu
olah@usc.edu
[**] Support of our work by Loker Hydrocarbon Research Institute is
gratefully acknowledged.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200901414.
5358
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 5358 –5362

Angewandte
Chemie
Several attempts were made to obtain a salt of anion II in
a persistent separable crystalline form. However, none of the
products was stable in the solid state under usual conditions
(Scheme 2). While monitoring the tBuOK mediated depro-
Scheme 2. Preparation of a-fluorobis(phenylsulfonyl)methide ion (II)
under different conditions: A) THF, room temperature, B) DMF, room
temperature. [a] Not stable in the solid state. [b] Not isolated.
[c] Stable at ambient temperature for several weeks under an inert
atmosphere.
tonation of FBSM by NMR spectroscopy (1H, 19F) the signals
for the acidic proton at d = 6.1 ppm and the doublet F signal
at d = À171.9 ppm disappeared. Appearance of a new sharp
singlet at about d = À202 ppm was observed in the 19F NMR
spectrum, indicative of facile deprotonation of FBSM under
basic conditions. However, instability of the carbanionic
Figure 2. X-ray crystal structures of A) bis(phenylsulfonyl)methide (I)
and B) a-fluorobis(phenylsulfonyl)methide (II) anions.[15]
products owing to defluorination restricted their isolation in
structure of II-N is essentially consistent with the properties
of a-sulfonyl carbanions described by Gais and co-workers.[14]
+
À
solid crystalline form. To suppress any probable C F···M
interaction[12] and the resulting defluorination, we decided to
use appropriate bases with countercations that would weakly
coordinate with the fluorine atom. Among the choices, alkyl
ammonium hydroxides were found to be most desirable
because of their diminished interaction with the substrate,
greater commercial availability, and structural variability. We
found that reaction of FBSM with tetrabutylammonium
hydroxide resulted in the formation of the ionic salt II-N
which is stable in the solid state for up to few weeks under
inert atmosphere.
For instance, the average S C17 bond length (1.716 ꢀ) in II-N
À
is noticeably shorter than the corresponding one in sulfone[13]
À
whereas the S O bonds of II-N are slightly longer. Further
NMR spectroscopy studies of II with different counterions
(Li+ (II-Li), Na+ (II-Na), and nBu4N+ (II-N)) in [D6]DMSO
reveal the liquid-phase behavior of the anion (Table 1).
Table 1: NMR chemical shifts of carbanions I and II in combination with
various cations.
Compound
da-F/H
[ppm]
dH [ppm]
dC
x
[ppm][a]
The complex II-N was crystallized from its solution in a
THF/toluene/hexane solvent mixture as fine cubes. X-ray
crystallographic studies indicated that the fluorinated ion pair
II-N is significantly different from its parent and brominated
analogues in its electronic as well as stereochemical nature
(Figure 2). The most remarkable structural characteristic is
that II-N has a pyramidal structure, the critical pyramidaliza-
tion angle (F) around the carbanion center is found to be
49.98, whereas a planar structure is adopted by both I and VI.
Considering that the pyramidalization angle of a standard
tetrahedral structure is 608, it is evident that II-N is strongly
sp3 (pyramidalized) rather than sp2 hybridized. On the other
o-H
m-H
p-H
I-H (HBSM)[b]
5.94 7.87–7.89 7.60–7.64 7.73–7.77
3.65 two multiplets at 7.30–7.34 (2H) and 63.8
7.70–7.74 (3H)
71.8
I-N[b]
II-H (FBSM) À171.9 7.93–7.95 7.71–7.75 7.86–7.90 104.4
II-Li
II-Na
II-N
À202.5 7.15–7.17 7.04–7.08 7.19–7.23 127.2
À202.8 7.17 7.04–7.08 7.21–7.23 127.1
À202.3 7.16–7.18 7.04–7.08 7.19–7.23 127.2
[a] Of the carbanionic carbon atom Cx. [b] HBSM=bis(phenylsulfonyl)-
methane; I-N=I[nBu4N]+.
À
hand, the C17 F1 distance of 1.404 ꢀ is considerably longer
than the typical Csp3 F bond length in fluorohydrocarbons
As expected, the 19F NMR spectroscopy studies showed
an upfield shift of approximately 30 ppm for II compared to
II-H (FBSM). The essentially identical 19F chemical shifts for
all the salts (II-Li, II-Na, and II-N) clearly indicate the
absence of strong interactions between the cations (Li+, Na+,
nBu4N+) and the FBSM anion in [D6]DMSO. More interest-
À
À
(1.32–1.38 ꢀ as in CH3F, CH2F2, CF3H, and CF4). This C F
À
bond in II-N is very close to the “upper limit” of tertiary C F
bonds (1.408 ꢀ).[13] Nevertheless, the C Br bond of VI,
À
À
1.884 ꢀ, is much shorter than the regular C Br bond which is
approximately 1.95–1.98 ꢀ. Besides these two differences, the
Angew. Chem. Int. Ed. 2009, 48, 5358 –5362
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5359

Communications
ingly, the signals of the carbanionic
carbon atom of II in the 13C NMR
spectrum showed a downfield shift
by 22.7 ppm compared to II-H
(FBSM). In contrast, the signals
of the carbanionic carbon atom of
I-N has an upfield shift of approx-
imately 8 ppm compared to that in
I-H (HBSM).
For a better understanding of
the pyramidal nature of II, we
carried out computational studies
on the anions I–V, including opti-
mized conformations, ionization
energies, and charge densities. A
diagram of gas-phase proton affin-
ities of these anions is illustrated
(Figure 3). By defining the heats of
formation of the corresponding
neutral molecules as 0.0 kcalmolÀ1,
it is found that the gas-phase
deprotonation becomes more diffi-
cult as the electronegativity of the
substituent increases (anion II–V).
The diagram shows that Cl and Br
are able to stabilize the anion
(relative to I), however, OCH3
and Fare destabilizing substituents.
It is also discovered that the
Figure 3. Gas-phase proton affinities (ionization energy/electronegativity profile) of carbanions I–V.
“cis-pyramidal”
conformation
(IIb; Figure 3), which agrees well
with the X-ray determined geome-
try, is 2.1 kcalmolÀ1 more stable
than the “cis-planar” conformation
(IIa) at the B3LYP/6-311 + G-
(2d,p)//B3LYP/6-311 + G(2d,p) +
ZPE level.[15] Although the trans-
planar structure IIc is found to be
the most stable conformation of II,
it is only 0.9 kcalmolÀ1 more stable
than the experimentally observed
conformation IIb. In contrast, the
trans-planar structures are favored
when the substituent X is H, Cl, or
Br. Calculations show that with the
methoxy substituent the energy
gap between the pyramidal confor-
mation and the planar conforma-
tion is only 0.2 kcalmolÀ1. In other
words, the methoxy group is the
“critical borderline” substituent
showing a clear gradual transition
to the pyramidal conformation as
an energetically preferred struc-
ture.
Table 2: Structural parameters (bond order, bond lengths, and charges) associated with disulfonyl-
methide anions.
X
Bond order[a]
À
X Cx
Bond length [ꢀ]
Mulliken charge
QS QC
F [8]
[b]
À
À
À
À
S O S Cx X Cx (X Cx)st
QO
QX
x
H
0.811
–
0.765
–
1.466 1.690 1.077 1.088–1.099 À0.561 +0.696 À0.240 +0.097 0.0
1.448 1.670 0.949 1.088–1.099 0.0
1.461 1.713 1.921 1.951–1.983 À0.541 +0.696 +0.093 À0.105 0.0
1.435 1.670 1.884 1.951–1.983 0.0
H[c]
Br
–
–
–
–
Br[d]
Cl
OCH3 0.772
–
–
–
–
1.023
1.461 1.712 1.761 1.783–1.858 À0.536 +0.476 +0.354 À0.084 0.0
1.464 1.715 1.389 1.405–1.458 À0.522 +0.963 À0.326 À0.310 18.8
(O)
À0.103
(OCH3)
F[e]
F[f]
F[g]
0.633
0.644
–
1.458 1.731 1.400 1.312–1.408 À0.529 +0.994 À0.066 À0.295 43.3
1.461 1.699 1.381 1.312–1.408 À0.533 +0.707 À0.107 À0.256 0.0
1.439 1.716 1.404 1.312–1.408
–
–
–
–
49.9
À
[a] Natural bond order. [b] Bond lengths in similar Csp3 X structures are used as the standard (st).
[c] Experimental data. [d] X-ray diffraction data of [(CF3SO2)2CBr]À, taken from Ref. [11]. [e] cis-Pyramidal
conformation. [f] trans-Planar conformation. [g] Experimental data.
stereochemistry of the products derived from the anion.
Opposite partial charges are found on the central carbon (Cx)
The partial atomic charges (Q), bond lengths, and
pyramidalization angles (Table 2) from the computational
studies clearly show the structural and electronic effects the
substituents have on the geometry of the anion and the
and the substituents X when X is H, Cl, and Br, resulting in
Coulombic attractions, whereas negative partial charges on
both Cx and the substituent X (when X = F or OCH3 groups)
undoubtedly cause repulsions. A trend that F and OCH3 bear
5360
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 5358 –5362

Angewandte
Chemie
more negative charges than Cl and Br reveals the important
role played by the electronegativity of the substituents in the
electronic properties of the anions. It is important to mention
that the anions are greatly stabilized by the oxygen atoms on
sulfonyl groups, although the substituents X also have
considerable influence on the structures of the anion. This
influence can be strongly supported by two facts: a) Each
oxygen atom bears more than 0.5 negative charges and the
NAO bond order study shows partial single bond character of
Figure 4. Calculated 13C NMR chemical shifts of the a-C (carbanionic
carbon) in I–V and the corresponding neutral molecules.
À
the S O bonds in all structures. b) The
calculated bond lengths agree extremely
well with the experimentally determined
results.
As mentioned above (from the results
of X-ray crystallographic studies), our cal-
À
culations also show that the Cx X bond
lengths in I, IV–VI are significantly shorter
À
than the standard Cx X bonds, which can be
rationalized by the Coulombic attractions.
À
In contrast, the computational C F bond
distances are found to be considerably
À
longer than the regular C F bonds in
analogous organic compounds. This result
is also consistent with the data obtained
from X-ray crystallographic analysis. In
À
addition, the extremely low C F bond
orders of both IIb (0.633) and IIc (0.644),
À
theoretically support the properties of C F
bonds in FCAs and account for the defluori-
nation observed in the solid state. The
measured chemical shift deviation is in
good agreement with the calculated values
in the idealized gas phase (I-H to I-N and
II-H to II-N). Except in the case of FBSM,
all carbanionic carbons show clear upfield
shifts. Calculations show a gradual increase
in chemical shift deviation (from the neutral
to the carbanionic carbon) with the increas-
ing electronegativity of the substituents
(Dd = Br(À24 ppm) < Cl(À18 ppm) <
Figure 5. Orbital representation of the FBSM anion based on molecular orbital (MO)
theory and bond theory.
OCH3(À12 ppm) < F(+14 ppm), Figure 4).
Probably, the electronegativity of the sub-
stituent plays a major role. The gradual
decline in the pyramidalization angles from 49.98 to 0.08 with
decrease in electronegativity (F > OCH3 > Cl > Br) clearly
shows the gradual transition in hybridization of the Cx from
sp3 to typical sp2.
between the two p orbitals on the phenyl groups, which may
partially cancel out the instability caused by the additional
steric repulsion in the cis conformation. The structure of IIb
can be explained by simple bond theory as well. To “localize”
the lone pairs on the fluorine atom and their relation to the
anionic sp3 orbital on Cx, the anion III (Figure 1) is used as a
model structure since the positions of the lone pairs on the
methoxy oxygen atom can be determined by viewing in the
Based on these results, a detailed structural picture of
FBSM anion can be illustrated (Figure 5). In terms of MO
theory, maximum orbital overlap occurs between one of the
2p orbitals on Cx and a 3p orbital on each of the S atoms since
all three orbitals are parallel to each other. Although one of
the 2p orbitals on the fluorine and the 2p orbital on Cx are in
the same plane, the maximum overlap is not possible because
these two orbitals are not parallel. Similarly, there are also
four 2p orbitals on the two oxygen atoms interacting with the
3p orbitals on S. The sum of these combinations forms p-type
MOs, and the bonding orbital is illustrated as IIb-HOMOÀ33
(Figure 5). Moreover, the IIb-HOMOÀ14 shows the overlap
À
direction of the O CH3 bond. According to the calculation,
the O CH3 bond is roughly periplanar to the anionic sp3
À
orbital on Cx which means the two lone pairs on oxygen are
gauche to the sp3 orbital. Consequently, the lone pairs on the
FBSM anion IIb are supposed to have a similar conformation
to the methoxy oxygen atom in III. The planar structure
causes more electronic repulsion owing to the strong inter-
action of the two lobes of the 2p orbital with all three lone
Angew. Chem. Int. Ed. 2009, 48, 5358 –5362
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5361

Communications
[3] a) T. Fukuzumi, N. Shibata, M. Sugiura, H. Yasui, S. Nakamura,
T. Toru, Angew. Chem. 2006, 118, 5095 – 5099; Angew. Chem. Int.
Ed. 2006, 45, 4973 – 4977; b) C. Ni, Y. Li, J. Hu, J. Org. Chem.
2006, 71, 6829-
pairs on F. Nevertheless, in the pyramidal structure IIb, there
are only two of these lone pairs experiencing similar
repulsion. To some extent, this explanation can be supported
by the MO theory results as well (note that the two lobes of
the 2p orbital in IIb-HOMO have slightly different sizes
where as the two lobes of the 2p orbital in IIc-HOMO have
the same size; Figure 5). The anion IIb can be described as a
resonance hybrid of the various resonance structures
(Figure 5) which rationalizes many of the properties exhibited
by the anion.
In conclusion, the preparation and structural character-
ization of the first persistent FCA that is, a-fluorobis(phe-
nylsulfonyl)methide and related anions with a tetrabutylam-
monium countercation have been achieved. The critical role
of electron-withdrawing substituents in modulating the prop-
erties of bis(phenylsulfonyl)methide anions and the resulting
pyramidalization that occurs with highly electronegative
a-substituents, such as F and OCH3, have been established
by high-level theoretical calculations, further supported by X-
ray crystallographic studies of the a-fluorobis(phenylsulfo-
nyl)methide salt.
6833; c) G. K. S. Prakash, S. Chacko, S. Alconcel, T. Stewart, T.
Mathew, G. A. Olah, Angew. Chem. 2007, 119, 5021 – 5024;
Angew. Chem. Int. Ed. 2007, 46, 4933 – 4936; d) G. K. S. Prakash,
X. Zhao, S. Chacko. F. Wang, H. Vaghoo, G. A. Olah, Beilstein J.
Org. Chem. 2008, 4, 1 – 7; e) G. K. S. Prakash, F. Wang, T.
Stewart, T. Mathew, G. A. Olah, Proc. Natl. Acad. Sci. USA
2009, 106, 4090 – 4094; f) T. Furukawa, N. Shibata, S. Mizuta, S.
Nakamura, T. Toru, M. Shiro, Angew. Chem. 2008, 120, 8171 –
8174; Angew. Chem. Int. Ed. 2008, 47, 8051 – 8054.
[4] K. Uneyama, Organofluorine Chemistry, Blackwell, Oxford,
2006, pp. 17 – 20, and references therein.
[5] a) J. P. Lorand, J. Urban, J. Overs, Q. A. Ahmed, J. Org. Chem.
1969, 34, 4176 – 4178; b) H. J. Castejon, K. B. Wiberg, J. Org.
Chem. 1998, 63, 3937 – 3942; c) D. A. Dixon, T. Fukunaga, B. E.
Smart, J. Am. Chem. Soc. 1986, 108, 4027 – 4031; d) W. B.
Farnham, D. A. Dixon, J. C. Calabrese, J. Am. Chem. Soc.
1988, 110, 2607 – 2611; e) C.-P. Qian, T. Nakai, D. A. Dixon, B. E.
Smart, J. Am. Chem. Soc. 1990, 112, 4602 – 4604.
[6] H. G. Adolph, M. Kamlet, J. Am. Chem. Soc. 1966, 88, 4761 –
4763.
[7] J. Hine, L. G. Mahone, C. L. Liotta, J. Am. Chem. Soc. 1967, 89,
5911 – 5920.
[8] F. Faigl, E. Marzi, M. Schlosser, Chem. Eur. J. 2000, 6, 771 – 777.
[9] D. J. MacDougall, A. R. Kennedy, B. C. Noll, K. W. Henderson,
J. Chem. Soc. Dalton Trans. 2005, 2084 – 2091, and references
therein.
[10] R. G. Dubenko, V. M. Nelpyuev, D. S. Pel’kis, Zh. Org. Khim.
1968, 4, 324 – 328.
Received: March 13, 2009
Revised: April 22, 2009
Keywords: carbanions · density-functional calculations ·
.
fluorine · structure elucidation · substituent effects
[11] M. Ochiai, N. Tada, K. Murai, S. Gota, M. Shiro, J. Am. Chem.
Soc. 2006, 128, 9608 – 9609.
[1] a) P. Kirsh, Modern Fluoroorganic Chemistry, Wiley-VCH,
Weinheim, 2004; b) B. E. Smart, J. Fluorine Chem. 2001, 109,
3 – 11; c) V. A. Soloshonok, Enantiocontrolled Synthesis of
Fluoroorganic Compounds: Stereochemical Challenges and
Biomedical Targets, Wiley-VCH, Chichester, 1999; d) R. E.
Banks, B. E. Smart, J. C. Tatlow, Organofluorine Chemistry:
Principles and Commercial Applications, Plenum, New York,
1994; e) R. Filler, Y. Kobayashi, L. M. Yagupolskii, Organo-
fluorine Compounds in Medicinal Chemistry and Biomedical
Applications, Elsevier, Amsterdam, 1993; f) S. Purser, P. R.
Moore, S. Swallow, J. Gouverneur, Chem. Soc. Rev. 2008, 37, 320.
[2] a) G. K. S. Prakash, A. K. Yudin, Chem. Rev. 1997, 97, 757 – 786;
b) G. K. S. Prakash, J. Hu, Acc. Chem. Res. 2007, 40, 921 – 930;
c) G. K. S. Prakash, S. Chacko, Curr. Opin. Drug Discov. Devel.
2008, 11, 793 – 802.
[12] H. Takemura, S. Nakashima, N. Kon, M. Yasutake, T. Shin-
myozu, T. Inazu, J. Am. Chem. Soc. 2001, 123, 9293 – 9298.
[13] F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G.
Orpen, R. Taylor, J. Chem. Soc. Perkin Trans. 2 1987, S1 – S19.
[14] G. Raabe, H. J. Gais, J. Fleischhauer, J. Am. Chem. Soc. 1996,
118, 4622 – 4630.
[15] CCDC 728306 (FCA II) and CCDC728307 (parent anion I)
contains the supplementary crystallographic data for this paper.
These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif. The details of the X-ray studies and calculations
are included in the supporting information.
5362
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 5358 –5362