Effects of Ion Pair Separation on Inversion
and Rotation of Sulfur-, Selenium-, and
Silicon-Stabilized Organolithium Reagents **
Most vinyl,['] cyclopropyl,[*' and a - a l k o ~ y [ ~lithium
]
reagents are configurationally stable. Other types racemize
too rapidly for trapping experiments[4".'1 or require very low
temperatures to avoid rapid eq~ilibration.[~"'
61 Advances in
the utility of chiral organolithium reagents depend on detailed structural i n f ~ r m a t i o nand
~ ~ ]mechanistic understanding of their racemization. One limiting mechanism for
organolithium reagent epimerization is a bimolecular assoAn alternative ion
ciative process through intermediate l.[4b1
pair dissociation process via 2 has been frequently proposed
since polar solvents usually accelerate racernization,"", 6a1
but this has not been directly testable since the ion pair
status of most organolithium reagents is not known. Our
hexamethylphosphoric triamide (HMPA) titration techthis problem by allowing the unn i q ~ e ' ~ '*]. ~addresses
.
equivocal identification of contact and separated ion pairs.
The correlation between ion pair structure and inversion
barriers for several organolithium reagents can now be
established.
,\A
2 [*
LI
L,
\
I+%-y
I Li
1
3c
3' P
SiMep
1 Carbanion-C
By Hans J. Reich* and Robert R. Dykstra
LI
l
L
We selected compounds 3I9] and 4 for detailed study.
These and related compounds are valuable reagents["I that
are easily prepared and stable in polar solvents. The presence
L
(III(IIIIJlllj111/1111//111/1(
5 4 3 2 1 0 -1
0.5
-6
-80
-101
~I
[**l
This work was supported by the U.S. National Science Foundation.
AngeM-. Chetn. Int. Ed. Engt. 1993, 32. N o . 10
(0VCH
27.5
27.0
26.5
26.0
- 8
R-//Lih; quintet in the 'Li NMR spectra and absence of
coupling between lithium and carbon. Formation of separated ion pairs did not result in the loss of diastereotopicity of
the SiMe, group even though increases in lJC-H
from 136 to
148 Hz and lJc.sifrom 85 to 106 Hz144 indicate significant
planarization at the carbanion center.
Variable-temperature 13C NMR spectra of 3 in THF,
THFjl HMPA, and THF/3HMPA are shown in Figure 2. In
THF, coalescence is observed at - 100 "C and in THF/
IHMPA at - 112 "C. The slight lowering of the barrier on
j
z
,
0 HMPA
[*] Prof. Dr. H. J. Reich. R. R. Dykstra
Department of Chemistry. University of Wisconsin
Madison. W1 53706 (USA)
Telefax: Int. code + (608)265-4534
-0.5
-6
Fig. 1. 'T, 'Li, and "P N M R spectra of an HMPA titration of 3 in a 3:2
mixture ofTHF/ether at - 121 "C. The insets show spectra from a 13C-enriched
sample (h = HMPA, // = solvent-separated ion pair).
T["Cl
of several NMR-active nuclei ('H, 13C, 7Li, 29Si, and 77Se)
facilitates careful NMR investigations. Figure 1 shows the
results of an HMPA titration of 3. The 7Li, 31P,and I3C
data, as well as other spectra not shown, indicate that in
= 18.5 Hz) with a pyramiT H F 3 is a contact ion pair ('JL,-,
dal carbanion center ( lJC-H
= 132 Hz,'', ''I 'J,,,
= 75 Hz).19]
The dimethyl(pheny1)silyl group served as a probe and indicated the configurational equilibration of 3; below - 100 "C
3 is chiral on the NMR timescale. The adduct 3 . 1 HMPA is
also a pyramidal ('JC-H
= 136 Hz, 'Jc.si= 85 Hz), chiral
contact ion pair (lJc-L,
= 14.4 Hz). After the addition of
three equivalents of HMPA, 3 is a fully solvent-separated ion
pair, as shown by the characteristic R-//Lih: quartet and
0.0
-113
1.0 HMPA
3.0 HMPA
L
u
LAL
m
1
(
1
1
1
1
1
(
1
1
1
1
)
)
1
1
1
1
1
1
1
1
1
1
1
)
1
5 4 3 2 1 0 - 1
5 4 3 2 1 0 - I
-6
-6
4 3 2 1 0-1
-6
Fig. 2. Variable-temperature "C N M R spectra of 3 in a 3:2 mixture of THF/
ether after addition of 0. 1, and 3 equiv of HMPA.
coordination to HMPA (AG* = 8.0 to 7.8 kcalmol-' for 3,
8.2 to 7.8 kcalmol-' for 4) would seem to indicate that inversion of the pyramidal structure is accelerated by the strong
donor ligand HMPA which loosens C-Li coordination. One
might, therefore, expect an even lower barrier for the separated ion pair. The separated ion pair, however, racemizes
more slowly (approximately 1/20 as fast at -101 "C)
Verlagsgesellschufi mbH. 0-69451 Weinheim, 1993
0570-OR3319311010-1469$ 10.00+ .25,'0
I469
with a coalescence temperature above -80°C (AG* =
9.1 kcalmol-' for 3,9.5 kcalmol-' for 4). The separation of
the ion pair (formation of 2) cannot be the rate-determining
step for the racemizaton of 3 and 4. These data cannot easily
be explained by an inversion process but fit well for rotation.
Similar conclusions have been reached by R. W. Hoffmann
et al. in a study of arylseleno- and aryltelluro-substituted
alkyllithium reagents.["] Barriers to rotation in carbanions
have origins in stabilizing n-cs' interactions when the carbanion lone pair is aligned with the S-Ph or Si-R bonds, and
destabilizing repulsions between lone pairs when the carbanion is not aligned.['31Such interactions should be stronger in
a free carbanion than in a contact ion pair. When the carbanion C atom rotates away from the favored orientation, the
charge density on the C atom must increase. This extra
charge is more effectively stabilized by lithium in a contact
rather than in a separated ion pair.
The racemization of 3 in T H F is independent of concentration (0.04 to 0.64 M). In addition, the presence of 1 mol
equiv of the separated ion pair lithium f l ~ o r e n i d e [had
~ ~ ]no
effect on the racemization rate, so the reaction does not
involve intermediate 1. Since lithium fluorenide did lower
the coalescence temperature for loss of Li-C coupling from
-92 to -112 "C, Li-Li exchange may be intermolecular.
The complex behavior of 3 and 4 prompted us to examine
a system in which there is no ambiguity between inversion
and rotation. The organolithium reagent 5 is achiral in the
absence of rotational barriers even if inversion is slow. We
studied this compound, as well as the related PhS(PhSe)CLiSiMe,Ph (6). Both are largely separated ion pairs in
T H F as well as in THF/HMPA, and both show diastereotopic SiMe, groups up to at least -20 "C. For 5 in a 1 : 3
mixture of THF/ether the SiMe, signals coalescence at
-12 "C (AG* = 13.3 kcalmol-I), whereas for 6 in a 3:2
mixture of THF/ether coalescence is reached at 7 "C
(AG' = 14.3 kcalmol-I). The measured barrier probably
corresponds to rotation about the C-S bond, which converts the chiral conformation of 5 to the achiral.
achiral
chiral
These findings provide a basis for understanding one odd
facet of the behavior of 3: the racemization barrier of the
separated ion pair increased as more HMPA was added
( T , = - 98 "C at 2 equiv of HMPA, - 79 "C at 3 equiv, and
-75 "C at 6 equiv). This suggests that the racemization of
the separated ion pair R - / / L i h i might actually proceed via
the contact ion pair. If the rotation barrier for the separated
ion pair 3 is as high as those measured for 5 and 6, then a
mechanism involving recoordination of lithium to carbon
(accompanied by loss of one or more coordinated HMPA
molecules) explains the rate-retarding effect of high HMPA
concentrations. Clearly, the racemization is mechanistically
complex. A coupled inversion- rotation process fits our
data: the inversion is more significant for the contact ion
pair in THF, and rotation dominates for the (probably planar) solvent-separated ion pair in THF/HMPA. We envision
a mechanism in which lithium coordination to S or Se allows
the lithium to move from one face of the carbanion to the
1470
f , VCH ~ ~ r l u ~ r ~ e . r i ~ inhH.
N r ~ / iD-69451
~i/f
Wenihcim, 1993
other without separation of the ion pair. Coupled rotationinversion processes have been extensively studied for sulfenamides, which are isoelectronic with phenythiocarbanSubstantial rotation barriers have been reported for lithiated sulfones,[6d. in which lithium coordinates to oxygen,
but not for lithiated sulfides, selenides, or silanes.[121However, the strong preference of Li for the equatorial position in
dithianes["] is equivalent to high rotation barriers in acyclic
systems. The observation of a diastereotopic CH, group in
the tetramethylethylenediamine (TMEDA) adduct of
Ph,PCH,Li (Act = 7 kcalmol-'),[''1 interpreted as indicating a barrier to inversion, implies that there is a significant
rotation barrier associated with C-P bonds as well. The
relatively high epimerization barriers of x-lithio selenides,L4".w
amines,r6b1and ethers[31take on a
new perspective with this information.
Received: May 12, 1993 [Z 6081 IE]
German version: Angeu. Chem. 1993, 105, 1489
~~
[l] a) E. J. Panek. B. L. Neff, H. Chu, M. G . Panek, J. Am. Chem. Soc. 1975,
97. 3996; R. Knorr, E. Lattke, Tcwuh~~drw~
Lett. 1977. 3969; b) D.
Seyferth. L. G. Vaughan. J. A m . Chem. Soc. 1964.86.883; H . Neumann.
D.Seebach, Chem. Ber. 1978, 111. 2785.
[2] a ) G . Kobrich, D . Merkel. K. Imkampe, Chem. Eer. 1973, 106. 2017;
b) D. E. Appleqnist. A. H. Peterson, J. A m . Chem. Soc. 1961, 83. 862;
H M. Walborsky. F. J. Impastato. A. E Young, ibid. 1964. 86, 3283.
[3] W. C. Still. C. Sreekumar. J. A m . Chem. Soc. 1980. 102, 1201; D. Hoppe,
F. Hintze, P. Tehben. A ~ R P I IC/irm.
'.
1990, 102, 1457: Angew. C/zem. h i .
Ed Engl. 1990. 29, 1422.
[41 For previous papers see: a) H. J. Reich, M. D. Bowe, J. A m . Chew. SOC.
1990, 112. 8994; b ) a bimolecular mechanism has been reported for a
cyclohexyllithium derivative: H. J. Reich, M. A. Medina, M. D. Bowe,
hid 1992, 114. I 1003; c ) H. J. Reich. D. P. Green, ibid. 1989. 111, 8729;
d ) H. J. Reich. J. P. Borst, ibid. 1991. 113. 1835; e) H.J. Reich, R. R. Dykstra. ihid. 1993. 115, 7041.
[5] P. G . McDougdl, B. D. Condon, Terruhedrun Lett. 1989, 30, 789.
[6] a) Selenides: R. W. Hoffmann, M. Julius, K. Oltmann, Tetruhedron Lerr.
1990, 31. 7419; A. Krief, G . Evrard, E. Badaoui, V. de Beys, R. Dieden,
ihid. 1989, 30, 5635; b) amines: W. H. Pearson, A. C . Lindbeck, J A m .
Chem. Soc. 1991, 113. 8546; S. T. Kerrick, P. Beak, hid. 1991. 113,9708:
c ) bromides: R. W. HOffmdnn. T. Ruhland, M. Bewersdorf, J. Chem. SOC.
Ciiern. Commun. 1991. 195; d) sulfones: H.-J. Gais, G . Hellmann, J. A m .
Chmi. Sue. 1992, 114. 4439.
[7] W. N. Setzer. P. von R. Schleyer, Adv. Orgunomet. Chein. 1985.24.353; R.
Amstutz. T. Laube, W. B. Scbweizer, D. Seebach, J. D. Dunitz, H e h . Chim.
Aclu 1984.67.224;G . Boche, Angex.. Chem. 1989,101,286; Angen. Chem.
Inf. Ed. Engl. 1989, 28, 277.
[8] P. R. Rdithby, D. Reed, R. Snaith, D. S. Wright, Angebr. Chem. 1991, f03,
1040; Angew. Chrm. h t . Ed. Engl. 1991, 30, 1011
[9] H.-J. Gais, J. Vollhardt. C. Kriiger. Angew. Clief77.1988, 100,1108; Angew.
Chem. Ini. Ed. Engl. 1988, 27, 1092.
[lo] a) E. Schaumann, C . Friese, Terruhedron Lett. 1989, 30, 7033; b) D. J.
Ager, M. B. East, J. Org. Chem. 1986, 51, 3983; J. Otera, Y. Niibo. H.
Nozaki, ibid. 1989, 54. 5003; M. R. Myers, T. Cohen, ihid 1989. 54. 1290.
[I I] W Zarges, M. Marsch. K. Harms, W.Koch, G. Frenking, G . Boche, Chem.
Brr. 1991, 124, 543; G . Chassaing. A . Marquet, Terruhedron 1978, 34,
1399; D. Seehach, J. Gabriel, R. Hassig, Heh. C h m Aera 1984, 67, 1083.
[12] T Ruhland. R. Dress, R. W. Hoffmann, Angew Chem. 1993, 105, 1487;
Angew. Chem. Int. Ed. Engl. 1993, 32, 1467.
[I31 J.-M. Lehn, G. Wipff, Helv. Chiin. Actu 1977. 60, 1239; F. Bernardi, A.
Bottoni, G. S. Valli, A. Venturini, Guzr. Chim. Irul. 1990, 120. 301; P. von
R. Schleyer, T. Clark, A. J. Kos, G . W Spitmagel, C . Rohde, D. Arod,
K . N. Houk, N. G. Rondan, J A m . Chem. SOC.1984, 106, 6467.
1141 M. Raban. D. Kost, Trfruherlron 1984, 40, 3345.
[15] H:J. Gais, G . Hellmann. H. J. Lindner, Ang. Chem. 1990, 102.96; Angew.
Chrm. Int. E d Engl. 1990. 29. 100.
[16] E. L. Eliel. A. A . Hartmann, A. G. Abatjoglou. J A m . Chem. Soc. 1974,
94, 1807.
1171 G . Frdenkel. W. R. Winchester. P. G . Williard, Orgunomclullic~1989, 8 ,
2308.
057fJ-0833/93/1010-1470 B I 0 00+ .25/0
Angeu.. Cliem. 1 m Ed. Engl. 1993, 32. N o . 10