Multiple hydrogen/lithium interconversions at the same benzene nucleus: Two at the most [3]

Full text of DOI:10.1021/ja0032733

3822
J. Am. Chem. Soc. 2001, 123, 3822-3823
Multiple Hydrogen/Lithium Interconversions at the
Same Benzene Nucleus: Two at the Most
Manfred Schlosser,* Laurence Guio, and Fre´de´ric Leroux
Section de Chimie (BCh), UniVersite´
CH-1015 Lausanne, Switzerland
ReceiVed September 5, 2000
Every first-year student is advised not to draw formulas
carrying two like charges in close vicinity. Electrostatic repulsion,
so goes the argument, would make such structures energetically
exorbitant and hence unrealistic. This plausible rule of thumb is
confirmed by recent computational work at the MP2, MP4, and
B3LYP levels of theory. Proton abstraction from a “naked” phenyl
anion by another phenyl anion is a highly endothermal process,
requiring reaction enthalpies of 112, 96, and 91 kcal/mol at ortho,
meta, and para positions.1 However, the situation changes
profoundly when the phenyl anion-promoted deprotonation of
phenyllithium is examined. This time the process is exothermal
at ortho and para positions by approximately 10 and 5 kcal/mol,
respectively.1 In other words, the chelating and π-coordinating
properties of the lithium atom render phenyllithium more acidic
than benzene is!
With this as a background it should be intriguing to assess
computationally (i.e., under gas phase conditions) and experi-
mentally (in solution) the relative basicities of the three phen-
ylenedilithiums which at least formally result when one phenyl-
lithium cannibalizes another one. In this context recent work by
Bickelhaupt et al.2 deserves attention. 1,3,5-Trilithiobenzene
proved to be too basic to be generated from 1,3,5-tribromobenzene
by permutational halogen/metal exchange even when tert-butyl-
lithium was employed as the reagent. 5-Bromo-1,3-phenylene-
dilithium was obtained almost exclusively under optimized
conditions.2 1,3,5-Trilithiobenzene (79%) did form along with 1,3-
phenylenedilithium (13%) and phenyllithium (7%) as byproducts
when the halogen in 1,3,5-tribromobenzene was replaced reduc-
tively using the 4,4′-di-tert-butylbiphenyl/lithium “radical-anion”.2
The reluctance of benzene nuclei to tolerate a total of three
lithium substituents casts doubt on reports published by H. Gilman
et al.3,4 more than a quarter of a century ago and regularly quoted
in the scientific literature since then. The authors claimed to have
discovered an unprecedented temperature effect. If performed at
-75 °C, the reaction between 1,3,5-trifluorobenzene and a slight
excess of tert-butyllithium afforded 1,3,5-tri(tert-butyl)benzene
(1) in up to 72% yield.3,4 Under quite similar conditions, but at
-115 °C and after subsequent treatment with chlorotrimethyl-
silane, 1,3,5-trifluoro-2,4,6-tris(trimethylsilyl)benzene (2) was
isolated as the sole product (95-96%).3,4 The trilithiated species
3 was assumed to be its direct precursor.
and fast trilithiation at a temperature as low as -115 °C, fluorine
being known to act only as a moderately strong exchange-
accelerating neighboring group.7
A reinvestigation conformed our suspicion. When we attempted
to trap the trilithiated 1,3,5-trifluorobenzene (3) at -115 °C with
electrophiles other than chlorotrimethylsilane (e.g., dimethyl
sulfate and carbon dioxide), only monosubstituted derivatives 4
were obtained. After consecutive treatment of 1,3,5-trifluoroben-
zene in tetrahydrofuran at -100 °C, in the absence or presence
of potassium tert-butoxide and N,N,N′,N′′,N′′-pentamethyldieth-
ylenetriamine (PMDTA), with tert-butyllithium (3 equiv, 2 h) and
with phenylacetylene-ω-d the recovered substrate contained
mainly (g90%) a single deuterium atom and had incorporated a
second or third isotopic label in trace amounts at best (e3 and
e1%, respectively). Repetition of this reaction at -75 °C, using
3.0 or 6.0 equiv of tert-butyllithium, afforded 1,3,5-tri-tert-
butylbenzene (1) in 37 and 72%. Both samples of the hydrocarbon
were essentially dideuterated. When the reaction mixture (made
with 6.0 equiv of tert-butyllithium) was poured on dry ice rather
(6) Nucleophilic displacement of p-fluorine atoms from cyano or oxazoline
substituted arenes: Richtzenhain, H. Ber. Dtsch. Chem. Ges. 1944/46, 77/79,
1-6; Chem. Ber. 1948, 81, 92-97; Meyers, A. I.; Mihelich, E. D. J. Am.
Chem. Soc. 1975, 97, 7383-7385.
(7) Snieckus, V. Chem. ReV. 1990, 90, 879-933.
(8) The structural assignment is supported by 1H- and 13C NMR spectra
and elementary analysis.
(9) Krizan, T. D.; Martin, J. C. J. Am. Chem. Soc. 1983, 105, 6155-6157;
see also: House, H. O.; Czuba, L. J.; Gall, M.; Olmstead, H. D. J. Org. Chem.
1969, 34, 2324-2336.
(10) Lipshutz, B. H.; Wood, M. R.; Lindsley, C. W. Tetrahedron Lett. 1995,
36, 4385-4388.
(11) We have monitored the reaction between chlorotrimethylsilane (0.5
M) and butyllithium (0.5 M in diethyl ether at -75 °C and in tetrahydrofuran
at -100 °C) and chlorotrimethylsilane (0.5 M) and tert-butyllithium (0.5 M
in tetrahydrofuran at -100 °C) in the time interval ranging from 0 to 720
min and found approximate half-lives of 1 h, e1 min, and .1 day,
respectively.
(12) Authentic material7 prepared by consecutive treatment of 1-bromo-
3,5-difluorobenzene in tetrahydrofuran with tert-butyllithium, cuprous iodide
tributylphosphine complex, again tert-butyllithium and 1-nitronaphthalene
[method: Bergbreiter, E. D.; Whitesides, G. M. J. Am. Chem. Soc. 1974, 96,
4937-4944; Bergbreiter, E. D.; Reichert, O. M. J. Organomet. Chem. 1977,
125, 119-124].
These results are mysterious for two reasons. The authors tacitly
assume the fluoride displacement by tert-butyl groups to obey
the standard nucleophilic addition/nucleofugal elimination mode
although this mechanistic pattern is restricted to heavily strained5
or acceptor-activated6 haloarenes as substrates. Equally incom-
prehensible is how 1,3,5-trifluorobenzene could undergo smooth
(13) Authentic material prepared from 3,5-di-tert-butylaniline [Burgers, J.;
van Hartingsveldt, W.; van Keulen, J.; Verkade, P. E.; Visser, H.; Wepster,
B. M. Recl. TraV. Chim. Pays-Bas 1956, 75, 1327-1342] by the Balz-
Schiemann reaction [method: Roe, A. Org. React. 1949, 5, 193-228].
(14) Crowther, G. P.; Sundberg, R. J.; Sarpeshkar, A. M. J. Org. Chem.
1984, 49, 4657-4663; Slocum, D. W.; Hayes, G.; Kline, N. Tetrahedron Lett.
1995, 36, 8175-8178.
(15) Cabiddu, S.; Contini, L.;. Fattuani, C.; Floris, C.; Gelli, G. Tetrahedron
1991, 47, 9279-9288.
(16) Maercker, A. In Houben-Weyl: Methoden der organischen Chemie;
Hanack, M., Ed.; Thieme: Stuttgart, 1993; Vol. E 19d, pp 448-566, see pp
462-466.
* Author for correspondence. Telephone: ++41/21/692 39 51. Fax: ++41/
21/692 39 65. E-mail: manfred.schlosser@ico.unil.ch.
(1) Bachrach, S. M.; Hare, M.; Kass, S. R. J. Am. Chem. Soc. 1998, 120,
12646-12649.
(2) Rot, N.; Bickelhaupt, F. Organometallics 1997, 16, 5027-5031.
(3) Dua, S. S.; Gilman, H. J. Organomet. Chem. 1974, 64, C1 - C2.
(4) Howells, R. D.; Gilman, H. Tetrahedron Lett. 1974, 15, 1319-1320.
(5) Bickelhaupt, F.; Jenneskens, L. W.; Klamer, J. C.; de Wolf, W. H. J.
Chem. Soc., Chem. Commun. 1984, 733-735.
10.1021/ja0032733 CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/28/2001

Communications to the Editor
J. Am. Chem. Soc., Vol. 123, No. 16, 2001 3823
On the basis of these findings, a consistent picture of the events
at -100 and -75 °C can be drawn. Chlorotrimethylsilane is a
treacherous electrophile since it reacts very slowly with bulky
bases such as lithium diisopropyl amide and lithium 2,2,6,6-
tetramethylpiperidide, coexisting with them at low temperatures
over hours.9,10 The same is the case with tert-butyllithium and,
to some extent, even butyllithium as we have realized now.11 Thus,
a silylation step can and must be interposed between the first,
second, and third hydrogen/lithium interconversion or, in other
words, the threefold electrophilic substitution is mediated by
nothing but monometalated intermediates (6 - 8). At -75 °C,
the dilithiated trifluorobenzene 9 gradually begins to emerge but
does not survive. A cascade of alternating lithium fluoride
eliminations and tert-butyllithium additions to the thus generated
dehydroarenes is instantaneously unleashed. Each dilithiated
intermediate being less stable than its precursor (12 < 11 < 10
< 9), the sequence does not stop before all three fluorine atoms
are displaced, that is the dilithiated tri-tert-butylbenzene (12) is
attained. No trace of 1-tert-butyl-3,5-difluorobenzene12 nor of 1,3-
di-tert-butyl-5-fluorobenzene13 was detected in the neutralized
reaction mixture.
One should beware of unwarranted generalizations. If not
trilithiated species, dilithiated ones can be generated and char-
acterized by electrophilic substitution provided that the metalation-
promoting neighboring groups are solidly attached to the aromatic
ring and not nucleofugally as labile as halogens. The double
lithiation of 1,4-dimethoxybenzene14 and 1,3,5-trimethoxyben-
zene15,16 does indeed occur quite effectively.
Acknowledgment. This work was financially supported by the
Schweizerische Nationalfonds zur Fo¨rderung der wissenschaftlichen
Forschung, Bern (Grant 20-55′303-98) and the Bundesamt fu¨r Bildung
und Wissenschaft, Bern (Grant 97.0083 linked to the TMR project
FMRXCT970129). This publication is dedicated to Professor Rolf
Huisgen on the occasion of his 80th birthday (in June, 2000). He has
taught intellectual and experimental rigor to generations of scholars.
Supporting Information Available: Preparation and characterization
of authentic samples for comparison [trimethyl(2,4-trifluorophenyl)silane;
(2,4,6-trifluoro-1,3-benzenediyl)-1,3-bis(trimethylsilane); (2,4,6-trifluoro-
1,3,5-benzenetriyl)-1,3,5-tris(trimethylsilane), 2; 1-tert-butyl-3,5-difluo-
robenzene; 1,3-di(tert-butyl)-5-fluorobenzene; 1,3,5-tri(tert-butyl)benzene,
1; 2,4,6-trifluorobenzoic acid, 4; 2,4,6-tri(tert-butyl)isophthalic acid, 5];
details concerning the reactions between 1,3,5-trifluorobenzene and tert-
butyllithium; kinetic investigation of the condensation between chloro-
trimethylsilane and butyllithium or tert-butyllithium (PDF). This material
is made available free of charge via the Internet at http://pubs.acs.org.
than quenched with phenylacetylene-ω-d, pure 2,4,6-tri-tert-
butylbenzene-1,3-dicarboxylic acid (5) (mp > 250 °C dec)8 was
isolated in 68% yield.
JA0032733