Solution structure and stereochemistry of alkyl- and silyl-substituted allenyl-propargyllithium reagents

Article abstract of DOI:10.1021/ja991719d

Analysis of ¹³C chemical shifts and Li-C couplings showed that allenyllithium (5A-Li) was a mixture of monomer and dimer in THF, both with an allenyl structure. Similarly, the metalation products of 2-butyne (6A-Li), 4-methylpentyne (9A-Li), an

Full text of DOI:10.1021/ja991719d

VOLUME 121, NUMBER 42
OCTOBER 27, 1999
© Copyright 1999 by the
American Chemical Society
Solution Structure and Stereochemistry of Alkyl- and Silyl-Substituted
Allenyl-Propargyllithium Reagents
Hans J. Reich,* Johnathan E. Holladay, Tamara G. Walker, and Jennifer L. Thompson
Contribution from the Department of Chemistry, UniVersity of Wisconsin, Madison, Wisconsin 53706
ReceiVed May 24, 1999
Abstract: Analysis of 13C chemical shifts and Li-C couplings showed that allenyllithium (5A-Li) was a mixture
of monomer and dimer in THF, both with an allenyl structure. Similarly, the metalation products of 2-butyne
(6A-Li), 4-methylpentyne (9A-Li), and 4,4-dimethylpentyne (21A-Li) in THF, as well as several 1, 3-dialkyl
(32A-Li, 33A-Li, 34A-Li) propargyl-allenyllithiums formed by metalation or Li/Sn exchange were all
monomeric lithioallenes in THF. Compound 6A-Li was shown to have less than 5% and 21A-Li less than 3%
of the propargyllithium isomers (6P-Li, 21P-Li) present from analysis of residual broadening of the propargyl
carbon by Li-C coupling. The reagent prepared by metalation of dicyclopropylacetylene (8P-Li) has a propargyl
structure, but two related reagents with a cyclobutane spanning the 3,3-positions (37A-Li and 39A-Li) had
allenyl structures. Several triorganosilyl-substituted reagents were also investigated. Those with silyl groups
at the allenyl position (28A-Li, 31A-Li) are allenyllithiums, those with silyl groups at the propargyl position
(22-Li to 26P-Li) showed chemical shifts intermediate between those of allenyl and propargyl isomers, and
the shifts were strongly temperature-dependent under some conditions. These compounds are probably
equilibrating mixtures of allenyl and propargyllithiums or equilibrating mixtures of unsymmetrically π-complexed
structures, with barriers to interconversion (∆Gq-150) below 4 kcal/mol. Several of the organolithium reagents
studied had diastereotopic carbon signals (SiMe2, CMe2, or CPh2 groups), which allowed determination of
barriers to configurational inversion of the chiral allenyl fragment. Barriers from 6.1 kcal/mol (26A-Li in
dimethyl ether) to 14.5 kcal/mol (39A-Li in 3:2 THF/ether) were measured.
Introduction
lithium and the carbanionic fragment. We can consider four
limiting monomeric structures: localized allenyl (1), localized
propargyl (2), delocalized π-bonded (3), or separated ion pair
An understanding of the factors that influence the regiose-
lectivity of reactions of allenyl-propargyllithium (A/P-Li)
reagents1a,2-4 requires information about the solution structure
of the reagent and the nature of the interaction between the
son, B. O¨ .; Dykstra, R. R.; Phillips, N. H. J. Am. Chem. Soc. 1998, 120,
7201-7210. (i) Reich, H. J.; Dykstra, R. R. Angew. Chem., Int. Ed. Engl.
1993, 32, 1469-1470. Reich, H. J.; Dykstra, R. R. J. Am. Chem. Soc. 1993,
115, 7041-7042. (j) Reich, H. J. J. Chem. Educ.: Software Ser. D 1996, 3,
2. Reich, H. J.; Medina, M. A.; Bowe, M. D. J. Am. Chem. Soc. 1992, 114,
11003-11004. (k) Sikorski, W. H.; Sanders, A. W.; Reich, H. J. Magn.
Reson. Chem. 1998, 36, S118-S124.
(2) (a) Epsztein, R. In ComprehensiVe Carbanion Chemistry; Buncel,
E., Durst, T., Eds.; Elsevier: New York, 1984; Part B, Chapter 3. (b)
Yamamoto, H. In ComprehensiVe Organic Synthesis; Fleming, I., Trost,
B., Eds.; Pergamon Press: Oxford, U.K., 1991; Vol. 2, p 81. (c) Schuster,
H. F.; Coppola, G. M. Allenes in Organic Synthesis; Wiley: New York,
(1) (a) Reich, H. J.; Holladay, J. E. J. Am. Chem. Soc. 1995, 117, 8470-
8471. (b) Reich, H. J.; Holladay, J. E. Angew. Chem., Int. Ed., Engl. 1996,
35, 2365-2367. (c) Reich, H. J.; Reich, I. L.; Yelm, K. E.; Holladay, J. E.;
Gschneidner, D. J. Am. Chem. Soc. 1993, 115, 6625-6635. (d) Reich, H.
J.; Holladay, J. E.; Mason, J. D.; Sikorski, W. H. J. Am. Chem. Soc. 1995,
117, 12137-12150. (e) Reich, H. J.; Shah, S. K.; Gold, P. M.; Olson, R.
E. J. Am. Chem. Soc. 1981, 103, 3112-3120. (f) Reich, H. J.; Kulicke, K.
J. J. Am. Chem. Soc. 1996, 118, 273-274. (g) Reich, H. J.; Borst, J. P.;
Dykstra, R. R.; Green, D. P. J. Am. Chem. Soc. 1993, 115, 8728-8741. (h)
Reich, H. J.; Green, D. P.; Medina, M. A.; Goldenberg, W. S.; Gudmunds-
10.1021/ja991719d CCC: $18.00 © 1999 American Chemical Society
Published on Web 10/08/1999

9770 J. Am. Chem. Soc., Vol. 121, No. 42, 1999
Reich et al.
(4, SIP). In the discussions to follow we will identify the three
carbons as CA (allenyl), CC (central), and CP (propargyl).
of the bis-diethyl ether solvate 12 shows a delocalized structure
of the propargyl type, with close lithium contacts to all three
carbons of the π-system.9 In contrast, a closely related THF
solvate 13 in which the phenyl group on CA bears a chelating
ortho-methoxy substituent, has a dimeric structure with con-
siderable allenic character (the chelated structure does not appear
to survive in ether solution).9 Two bisacetylenic reagents 14
and 15 showing cyclic head-to-tail dimer structures, with
multicenter contact between lithium and carbon of the π-pro-
pargyl type, have been reported.10a
When we began our work, there was only scattered informa-
tion available on the structures of allenyl-propargyllithium
reagents. Proton NMR studies were often ambiguous. For
example, the metalation product of 2-butyne gives chemical
shifts of δ 2.14 for CH2 and δ 1.82 for CH3, consistent with an
assignment of structure 6P-Li, although the structure is actually
allenic.5 The metalation product of 3,3-dimethylallene was
correctly assigned structure 7A-Li on the basis of a δ 4.70 signal
for the allenyl proton.3 More definitive information is provided
by 13C NMR spectra. A key early study of the metalation product
of allene showed 13C NMR signals consistent only with the
allenyl structure 5A-Li, based on the strong downfield shift
(196.4 ppm) of the central carbon, similar to those of allene
itself (208.5 ppm).6 The 13C NMR spectrum of 4-methyl-1,2-
pentadienyllithium (9-Li) was interpreted in terms of an
equilibrating A/P structure.7 We have re-examined these reagents
as well as a number of others and report the results below.
Infrared spectra have also been diagnostic. Absorptions at
1850-1900 cm-1 were assigned to allenyl and >2000 cm-1 to
propargyllithium reagents. Allenyllithium has an absorption at
1895 cm-1 11,12
.
The metalation product of dicyclopropylacety-
lene 8P-H was identified as a propargyllithium reagent (8P-
Li) on the basis of an IR absorption at 2160 cm-1 13
Zinc, boron,
.
titanium, and aluminum allenyl and propargyl structures have
also been identified by their IR absorptions.14,15a
In connection with synthetic and mechanistic studies of the
lithium reagents prepared from 1,4-bis(trimethylstannyl)-2-
butyne (16) and 2,3-bis(trimethylstannyl)-1,3-butadiene (18), we
encountered complex regiochemical effects that required precise
knowledge of solution structure of the reagents involved. We
were able to characterize the monolithium reagent, prepared
from 16 by Li/Sn exchange, as the allenyl 17A-Li and not the
propargyl isomer 17P-Li on the basis of well-defined Li-C
coupling to a single carbon and a strongly downfield shifted
central carbon analogous to the parent allenyllithium.6 On the
basis of similar considerations, the monolithium reagent prepared
from 18 has a vinyllithium structure 19 and not the vinyli-
deneallyllithium structure 20.1c
On the basis of a very extensive 13C NMR and UV study of
mono, bis, and tris phenylated A/P-Li lithium reagents (includ-
ing 10, 11, 12, and 13), Dem′yanov and co-workers conclude
that the electronic structure is delocalized, with no separate
existence of allenyl and propargyllithiums.8 Most show 13C shifts
that are more propargylic than allenic in nature, in particular,
reagents with two aryl groups at the propargyl carbon CP such
as 10 and 12. On the other hand, a compound with alkyl groups
at CP (11), has a greater tendency to be a contact ion pair and
shows chemical shifts that are allenic in nature.
The data summarized above do not provide a sufficient basis
for assigning structures to the many known hydrocarbon- and
heteroatom-substituted A/P-Li reagents. We have reported
preliminary results1a,b of spectroscopic studies on such reagents
with alkyl, silyl, phenoxy, alkoxy, carbamoyloxy, phenylthio,
(9) Dem′yanov, P.; Boche, G.; Marsch, M.; Harms, K.; Fyodorova, G.;
Petrosyan, V. Liebigs Ann. 1995, 457-460.
(10) (a) Setzer, W. N.; Schleyer, P. v. R. AdV. Organomet. Chem. 1985,
24, 353-451. Schleyer, P. v. R. Pure Appl. Chem. 1984, 56, 151-162. (b)
Bauer, W.; Winchester, W. R.; Schleyer, P. v. R. Organometallics 1987, 6,
2371-2379. (c) Lambert, C.; Schleyer, P. v. R.; Wu¨rthwein, E.-U. J. Org.
Chem. 1993, 58, 6377. (d) Jemmis, E. D.; Chandrasekhar, J.; Schleyer, P.
v. R. J. Am. Chem. Soc. 1979, 101, 2848.
X-ray crystal structures have provided some additional
information about allenyl-propargyllithiums, although none are
simple alkyl-substituted derivatives. An X-ray crystal structure
1984. (d) Moreau, J.-L. In The Chemistry of Ketenes, Allenes, and Related
Compounds; Patai, S., Ed.; Wiley: New York, 1980; p 363. (e) Klein, J. In
The Chemistry of the Carbon-Carbon Triple Bond; Patai, S., Ed.; Wiley:
New York, 1978; p 343.
(11) Jaffe, F. J. Organomet. Chem. 1970, 23, 53-62.
(12) (a) Priester, W.; West, R.; Chwang, T. L. J. Am. Chem. Soc. 1976,
98, 8413-8421. (b) West, R.; Carney, P. A.; Mineo, I. C. J. Am. Chem.
Soc. 1965, 87, 3788-3789. (c) West, R.; Jones, P. C. J. Am. Chem. Soc.
1969, 91, 6156-6161. (d) Klein, J.; Becker, J. Y. J. Chem. Soc., Chem.
Commun. 1973, 576-577.
(3) Creary, X. J. Am. Chem. Soc. 1977, 99, 7632-7639.
(4) Huynh, C.; Linstrumelle, G. J. Chem. Soc., Chem. Commun. 1983,
1133-1134.
(13) (a) Ko¨brich, G.; Merkel, D.; Imkampe, K. Chem. Ber. 1973, 106,
2017-2024. (b) Ko¨brich, G.; Merkel, D.; Thiem, K. W. Chem. Ber. 1972,
105, 1683-1693. (c) Ko¨brich, G.; Merkel, D. Liebigs Ann. Chem. 1972,
761, 50-66.
(14) Ishiguro, M.; Ikeda, N.; Yamamoto, H. J. Org. Chem. 1982, 47,
2225-2227. Furuta, K.; Ishiguro, M.; Haruta, R.; Ikeda, N.; Yamamoto,
H. Bull. Soc. Chem. Jpn. 1984, 57, 2768-2776.
(15) (a) Pearson, N. R.; Hahn, G.; Zweifel, G. T. J. Org. Chem. 1982,
47, 3364-3366. (b) Zweifel, G.; Backlund, S. J.; Leung, T. J. Am. Chem.
Soc. 1978, 100, 5561-5562.
(5) Klein, J.; Becker, J. Y. Tetrahedron 1972, 28, 5385.
(6) van Dongen, J. P. C. M.; van Dijkman, H. W. D.; de Bie, M. J. A.
Recl. TraV. Chim. Pays-Bas 1974, 93, 29-32.
(7) Suzuki, M.; Morita, Y.; Noyori, R. J. Org. Chem. 1990, 55, 441-
449.
(8) Dem′yanov, P. I.; Styrkov, I. M.; Krut′ko, D. P.; Vener, M. V.;
Petrosyan, V. S. J. Organomet. Chem. 1992, 438, 265. Dem′yanov, P. I.;
Krut′ko, D. P.; Borzov, M. V.; Luk′yanov, E. V.; Petrosyan V. S. Russ.
Chem. Bull. 1997, 46, 1939-1947.

Substituted Allenyl-Propargyllithium Reagents
J. Am. Chem. Soc., Vol. 121, No. 42, 1999 9771
pyridylthio, and phenylseleno substituents at CP and report here
the full details on those with alkyl and silyl substituents. Our
study has involved several series of closely related structures
to probe substituent effects. We have also examined reagents
in solvents of different donor power (ether, dimethyl ether, THF,
HMPA). This has allowed us to establish the mode of dimer-
ization in nonpolar solvents and to compare contact and
separated ion pair structures in polar solvents. Some of the
reagents examined are chiral, and our study has provided
information about the racemization process by DNMR studies.
The principal tool we have used is variable temperature 13C,
6Li, and 7Li NMR spectroscopy, but some compounds were also
examined by IR spectroscopy.
The cyclobutylidene allene 37 was prepared using a method
reported by Bailey, involving an intramolecular addition of an
alkyllithium to an acetylene.18 Compound 39 was prepared using
the procedure of Harada.19 The final reaction gave, in addition
to 39A-H, a significant fraction of the product of hydride (rather
than tert-butyl) transfer, which was removed by performing a
partial metalation of the mixture with tert-butyllithium and
carbonating the product (which left unreacted 39A-H). Both
37A-H and 39A-H could be metalated using tert-butyllithium
to give 37A-Li and 39A-Li.
Results
Synthesis of Allenyl-Propargyllithium Reagents. Allenyl-
lithium (5A-Li) was prepared by metalation of allene.6 The
hydrocarbon-derived lithium reagents with the allenyl/propargyl
grouping at the end of the chain (6A-Li, 9A-Li, 21A-Li) were
prepared according to literature procedures by metalation of the
appropriate methyl alkyl acetylenes.15a,16 Silylation of these gave
compounds 22P-H to 27P-H, precursors to the lithium reagents
22-Li to 27-Li with silyl groups at CP. The silylation gave only
acetylenic products,16 provided that ether (rather than THF) was
used as solvent. The nonterminal reagent 8P-Li was prepared
as reported13 by metalation of dicyclopropylacetylene 8P-H.
Solvents and NMR Spectral Measurements. The principal
goal of the present work was to establish the structures of a
series of A/P-Li reagents in commonly used ethereal solvents.
Since THF freezes at -109 °C, we frequently used mixtures of
THF, diethyl ether (mp -116 °C), and/or dimethyl ether (mp
-138.5 °C) for NMR studies (most solutions also contained
pentane from the n-BuLi or t-BuLi used to prepare the reagents).
Diethyl ether-THF mixtures are usable to -135 °C, THF/
dimethyl ether mixtures to -145 °C, and ternary THF/ether/
dimethyl ether or binary dimethyl ether/pentane mixtures down
to below -160 °C. Dimethyl ether is only a slightly weaker
donor solvent than is THF, and diethyl ether is substantially
weaker.1f,20a
The lithium reagent 28-Li with trimethylsilyl substitution at
CA was prepared following a literature procedure17 by metalation
of the 1-trimethylsilylpropyne. The reagent 31A-Li was prepared
by Li/Se exchange of 30, which was prepared by sequential
methylation and silylation of the dianion formed from phenyl
propargyl selenide (29).1e The same lithium reagent was also
prepared by metalation of 1-phenyldimethylsilyl-1,2-butadiene.
Several of the lithium reagents (32, 33, 34) were prepared
by Li/Se or Li/Sn exchange, following the procedure illustrated
for compound 33.1d This route can often be used for the
preparation of either the propargyl- (kinetic product of the final
stannylation) or allenylstannane (thermodynamic stannylation
product).
Allenyllithium. The 13C NMR spectrum of the metalation
product of allene has been interpreted in terms of an allenyl
structure.6 We have re-examined allenyllithium and found that
in 2.5:1:1 Me2O/THF/pentane the single set of three signals
(18) Bailey, W. F.; Aspris, P. H. J. Org. Chem. 1995, 60, 754-757.
(19) Harada, T.; Katsuhira, T.; Osada, A.; Iwasaki, K.; Maejima, K.;
Oku, A. J. Am. Chem. Soc. 1996, 118, 11377-11390.
(20) (a) Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1995, 117, 9863-
9874. (b) Wanat, R. A.; Collum, D. B.; Van Duyne, G.; Clardy, J.; DePue,
R. T. J. Am. Chem. Soc. 1986, 108, 3415-3422.
(16) Despo, A. D.; Chiu, S. K.; Flood, T.; Peterson, D. J. J. Am. Chem.
Soc. 1980, 102, 5122-5123.
(17) (a) Corey, E. J.; Kirst, H. A. Tetrahedron Lett. 1968, 5041-5043.
(b) Corey, E. J.; Ru¨cker, C. Tetrahedron Lett. 1982, 23, 719-722. (c) Corey,
E. J.; Yu, C.-M.; Lee, D.-H. J. Am. Chem. Soc. 1990, 112, 878-879.

9772 J. Am. Chem. Soc., Vol. 121, No. 42, 1999
Reich et al.
Figure 1. 13C NMR spectra of CA for 0.26 5A-Li in 2.5:1:1 Me2O/
THF/pentane at -150 °C.
Figure 2. 13C signals of 0.16 M 21-Li in 9:1 Me2O/pentane at -140
°C. 7Li-13C coupling constants: for CA, J ) 24 Hz, for CP, J < 0.4 Hz.
(43.8, 87.3 and 197.5 ppm) at -80 °C decoalesces into two
sets at around -110 °C. The relative areas of the pairs of signals
are concentration-dependent, so they cannot represent a pair of
isomeric dimers such as a 4-center and head-to-tail dimer (40).
Both species are clearly allenyl, based on the downfield shift
of the central and propargyl carbons, as well as the well defined
tert-butyl at CA showed a larger fraction of allenyllithium than
did the methyl (see below).
We can estimate the maximum fraction of the propargyl-
lithium isomer 21P-Li present as follows. The equilibration
between 21A-Li and 21P-Li must be intramolecular since C-Li
coupling is observed. Thus the propargyl carbon CP would be
broadened by C-Li coupling in proportion to the fraction of
21P-Li present. The line width of CP (3.5 Hz) is only 1.2 Hz
greater than that of the central carbon CC (Figure 2); thus JC-Li
for the propargyl carbon is <0.4 Hz (J ) 1/3v1/2). If we estimate
that the full JC-Li for 21P-Li is g16 Hz,22 the line width of the
fully coupled CP carbon would be >48 Hz (3J). Thus there can
be no more than 3% of propargyllithium isomer 21P-Li present.
Compound 6-Li in 2:1 THF/ether showed an excess broadening
of 2 Hz for Cp so that there can be no more than 5% of 6P-Li
in the equilibrium mixture. This analysis presumes that the two
isomers are in rapid equilibrium on the NMR time scale.
1,3-Dialkyl-Substituted Allenyllithiums. We have previ-
ously reported on the 1,3-dialkylallenyllithium 32-Li.1d We
report here some additional information about this compound
and data for 33-Li and 34-Li. All are allenyllithiums with well-
defined 13C-7Li coupling to CA and negligible broadening of
CP.
13C-7Li coupling of the allenyl carbon (CA, 88.8 ppm, JC-Li
)
27 Hz, and 87.2 ppm, JC-Li ) 16 Hz, Figure 1). The 1:1:1:1
quartet identified the species favored at lower concentration as
a monomer 5A-Li, and the 1:2:3:4:3:2:1 septet must be a cyclic
7
oligomer (5A-Li)n (CA coupled equally to two Li). Unfortu-
nately, accurate measurement of the concentration dependence
was not possible since a precipitate (identified as 1-propynyl-
lithium) formed as spectra were measured. The results of one
successful dilution on a homogeneous sample gave an aggrega-
tion state difference of two between the monomer and the second
species. Thus allenyllithium apparently dimerizes in the normal
4-center fashion of most localized organolithium reagents.
The ratio of monomer to dimer was temperature-dependent.
At -140 °C a 0.26 M solution of 5-Li is an approximately 2:1
molar ratio of monomers to dimers. The fraction of monomer
increases on warming until the ratio is 2:3 at -80 °C (estimated
from the chemical shift of the averaged signal). Thus the
monomer is enthalpically favored but entropically disfavored.
Alkyl-Substituted Allenyllithiums. Alkyl substitution at
carbanion centers generally has a destabilizing effect, and
placing an alkyl group on CA is expected to destabilize the
allenyl isomer. Both the methyl- (6-Li)5 and isopropyl- (9-Li)-
substituted7 compounds have been previously studied by NMR
spectroscopy. Our examination of these structures using low-
temperature 13C NMR spectra in THF solutions showed 6A-Li
and 9A-Li to have the C-Li coupling and downfield central
carbon (CC) chemical shifts characteristic of allenyllithium
structures.
Aggregation. In contrast to the parent allenyllithium 5A-Li,
which is a mixture of monomer and dimer in solvents containing
THF, the alkylated allenyllithiums 6A-Li, 9A-Li, 21A-Li, 32A-
Li, 33A-Li, and 34A-Li showed no indications of dimerization.
In all cases the lithium-bearing carbon was coupled to a single
lithium. This is consistent with the general observation that the
extent of aggregation decreases with increasing alkyl substitution
at carbanion centers.10b,23,24a However, in the weaker solvent
diethyl ether, both 32A-Li and 33A-Li dimerize. All of 13C
signals and the 6Li signals are doubled, and the ratio is
independent of concentration between 0.64 and 0.01 M (6Li
NMR, signals at 0.92 and 0.97 ppm). The C-Li NMR signal
of 32A-6Li is a multiplet (Figure 3) best interpreted as the
superposition of two 1:2:3:2:1 quintets (offset by 2J) arising
from the meso and dl diastereomers of the dimer, as also seen
A tert-butyl substituent at CA should maximally destabilize
the allenyl structure, not only because of the “normal” alkyl
effect but also because the bond angle changes caused by
rehybridization at CA should result in increased strain (B-
strain),21 and because of steric inhibition of solvation of the
lithium cation. Nevertheless, 21-Li has the characteristic chemi-
cal shifts of an allenyllithium. We have been able to test this
putative “tert-butyl effect” with the pair of silyl-substituted
compounds 23-Li and 27-Li and found that the compound with
(22) We have measured JC-Li for two authentic propargyllithium reagents
prepared by metalation of phenyl 2-butynyl sulfide, 16 Hz, and phenyl
2-butynyl selenide, 18 Hz.1a
(23) McGarrity, J. F.; Ogle, C. A. J. Am. Chem. Soc. 1985, 107, 1805-
1810.
(24) (a) Seebach, D.; Hassig, R.; Gabriel, J. HelV. Chim. Acta 1983, 66,
308-337. (b) Seebach, D.; Siegel, H.; Gabriel, J.; Ha¨ssig, R. HelV. Chim.
Acta 1980, 63, 2046-2053.
(21) For an example of a tert-butyl substituent that increased the fraction
of propargyl isomer in an allenylzinc reagent, see ref 19.

Substituted Allenyl-Propargyllithium Reagents
J. Am. Chem. Soc., Vol. 121, No. 42, 1999 9773
6
Figure 3. 13C NMR signal of Li-labeled CA of (32A-Li)2 in 3.7:1
Figure 4. 13C signals of 0.32 M 26A-Li in 12:1 Me2O/pentane at -145
°C.
ether/pentane at -120 °C. Superimposed on the signal for CA is a
simulation of a 5:3 mixture of two 1:2:3:2:1 quintets (J(13C-6Li) ) 6
Hz).
for 1-methoxy-1-lithio-3-tert-butylallene.10c The assignment as
two superimposed chemically shifted quintets separated by 2J
was confirmed by measuring the spectra at a different field
strength. A variable temperature study of (32A-Li)2 in ether
showed the two sets of signals coalesced between -50 and -46
°C. The isopropyl methyls were diastereotopic and gave four
signals at low temperature, which also coalesced in this
temperature range.
Silyl-Substituted Allenyl-Propargyllithiums. A variety of
silyl-substituted allenyl-propargyllithium reagents have been
prepared and utilized in synthetic organic chemistry.14,17,25a-c
We have examined several reagents with a silyl group at either
CP or CA.
Figure 5. 13C chemical shifts in Me2O (b), 2,5-dimethyltetrahydrofuran
(0), and Me2O-8HMPA (4) of CC, CA, and CP of 23-Li.
The presence of a carbanion-stabilizing group at either
terminus of an A/P-Li reagent might be expected to favor
localization of charge and lithium coordination at that carbon.
This is in fact what has been observed. For example, the
metalation products of phenyl 2-butynyl sulfide and selenide
are rare examples of lithium reagents with a propargyl structure.1a
The consequences of silyl substitution tend to follow this pattern,
but the data is more ambiguous than for the S and Se analogues.
Triorganosilyl Substituents at CA. Compounds 28A-Li and
31A-Li show 13C NMR properties consistent with assignment
as allenyllithium reagents. The chemical shifts were only slightly
temperature-dependent (∆δ ∼1.0 ppm between -50 and -150
°C). Well-defined C-Li coupling was observed for CA.
Compound 28A-Li showed strong IR absorptions for an allenyl
stretch (1868, 1886 cm-1) and weaker absorptions at a position
expected for a propargyllithium (2026 cm-1). Addition of excess
HMPA to a dimethyl ether solution of 28A-Li resulted in
complete conversion to a SIP.1g The central carbon (CC) moved
almost 27 ppm upfield on ionization, indicating the charge-
localizing effect of lithium coordination.
coupling to 7Li, and the line widths of CP and CC are the same,
so there is no indication of any association between Li and CP
(Figure 4).
The other CP-silyl-substituted reagents have more ambiguous
NMR properties. Compound 23A-Li has chemical shifts inter-
mediate between those of propargyl and allenyllithiums. In some
solvents such as dimethyl ether and THF these shifts show
temperature dependence too large to be caused by normal
changes in solvation,26 aggregation,27 or conformation, but
instead, they are likely to reflect substantial changes in the
electronic structure of the lithium reagent (Figure 5). The
reagents 22-Li, 24-Li, and 25-Li show similar behavior. Figure
6 compares the chemical shift behavior of these with authentic
allenyllithium (6A-Li) and propargyllithium (41P-Li) reagents.
The most economical explanation is that these lithium reagents
are mixtures of the allenyl and propargyl isomers in rapid
equilibrium, a conclusion supported by the IR spectrum of 23-
Li, which has absorptions at both 2179 and 1894 cm-1. A
bridged structure (3) should show only a single intermediate
absorption.
The IR spectrum of 23-Li has absorptions at both 2179 and
1894 cm-1, consistent with formulation as an equilibrium
mixture of A and P isomers. A bridged structure (3) should
show only a single intermediate absorption.
Effect of Ion Pair Separation on Silyl-Substituted Re-
agents. We have examined the effect of HMPA on all of the
silyl-substituted A/P-Li reagents. This experiment provides
information about the strength of association between lithium
and the carbanion and some insight into the effect of lithium
coordination on the electronic structure of the carbanion. A
typical HMPA titration is presented in Figure 7. With up to 1
Triorganosilyl Substituents at CP. We examined a family
of compounds with substituents at CP. Of the compounds
studied, the tert-butyldimethylsilyl analogue 26-Li shows the
simplest behavior. Although the CC chemical shift is somewhat
upfield of that observed for the hydrocarbon allenyllithiums,
the other features are those of a true allenyllithium. CA shows
(26) Jackman, L. M.; Scarmoutzos, L. M. J. Am. Chem. Soc. 1987, 109,
5348-5355.
(27) The monomer and dimer of allenyllithium differ by 1.5 ppm (CA),
6.9 ppm (CC), and 1.3 ppm (CP). The PhLi ipso carbon varies from 174.2
ppm for the ether tetramer and 187.1 ppm for the ether dimer to 196.7 for
the THF monomer.1h
(25) (a) Hartley, R. C.; Lamothe, S.; Chan, T. H. Tetrahedron Lett. 1993,
34, 1449-1452. (b) Wang, K. K.; Gu, Y. G.; Liu, C. J. Am. Chem. Soc.
1990, 112, 4424-4431. Wang, K. K.; Wang, Z.; Sattsangi, P. D. J. Org.
Chem. 1996, 61, 1516-1518. (c) Daniels, R. G.; Paquette, L. A. Tetrahedron
Lett. 1981, 22, 1579-1582.

9774 J. Am. Chem. Soc., Vol. 121, No. 42, 1999
Reich et al.
Figure 6. Temperature dependence of CC in A/P-Li reagents (Me2O
solvent) with alkyl substitution at CA and proton (6A-Li), silyl (22-Li
to 27-Li), or thio (41P-Li) substitution at CP.
Figure 8. Comparison of CIP and SIP structures for triorganosilyl-
substituted A/P-Li reagents.
SIP has propargyl-like chemical shifts (see Figure 5), consistent
with charge localization at the Si-bearing carbon. This is true
even for the tert-butyldimethylsilyl compound 26-Li, for which
the CIP is strongly allenic; here lithium coordination signifi-
cantly changes the structure.
Barrier to Interconversion of Propargyl and Allenyllith-
iums. For many σ-allyl or σ-A/P-Li organometallics the
position of electrophilic substitution is determined by the
position of the organometallic group, usually in an SE2N
sense.15b,17c,31 For others the interconversion between isomers
is faster than the rate of substitution, and the Curtin-Hammett
principle applies. This can be detected either by a lack of
correlation between product ratios and organometallic regio-
isomer ratios25b or by the observation of product ratios that vary
with electrophile.3,15a One of the goals of the current study was
to measure the energy barrier to interconversion of pairs of
isomeric allenyl and propargyllithium reagents to help establish
whether the site of metalation has an influence on product
regiochemistry. This requires reagents in which both isomers
are present in significant amount, so that DNMR techniques
can be used to measure interconversion rates, or place an upper
or lower limit on the rate.
The silyl-substituted reagents 22-Li to 26-Li are suitable
substrates. The most sensitive indicator of incipient decoales-
cence between the A and P signals would be the CC signal
because of the large chemical shift difference (∼70 ppm, 6300
Hz) between the two isomers. Although excess broadening was
typically seen for CC, in none of these reagents was decoales-
cence of the A and P isomer NMR signals detected even at the
lowest temperatures (-160 °C) accessible. The three allenyl
signals of the trimethylsilyl-substituted analogue 22-Li, chosen
because it appears to be an almost 1:1 mixture of A and P
structures, are shown in Figure 9. CA shows some signs of
splitting, and CP shows excess broadening at -150 °C, probably
from residual 13C-7Li J coupling.28 CC gave excess broadening
of ∼15 Hz, which allows estimation of a rate constant of at
least 106 s-1 and ∆Gq < 3.6 kcal/mol for A/P isomerization
(we assume that the chemical shifts of CC in the A and P isomers
are 165 and 98 ppm; with a shift of 142 ppm this means A/P )
66/34). This barrier is also an upper limit on the free-energy
difference between localized A and P structures such as 1 and
2 and bridged structures such as 3.
Figure 7. 7Li and 31P NMR spectra of an HMPA titration of 23-Li in
Me2O at -150 °C.
equiv of HMPA, only the mono-HMPA CIP was present. The
lithium signals of the 23-Li(Me2O)n and the 23-Li(HMPA)1
species are unusually broad, and the 31P signal shows the
characteristic broadening of a T1-relaxed P-Li-coupled 1:1:1:1
quartet.28-30 The broadening is not due to chemical exchange
6
(T2) effects, since the natural abundance Li triplet is clearly
resolved in the center of the broad doublet. The relatively short
7
T1 of the Li nucleus in these compounds explains why it has
been difficult to resolve 13C-7Li coupling. The observation of
only a mono-HMPA CIP at 0.5 equiv of HMPA rules out a
temperature-dependent CIP/SIP equilibrium as an explanation
for the chemical shift behavior of 22-Li to 25-Li (Figure 6).
All of the silyl-substituted lithium reagents were fully
converted to SIP’s on addition of 4-6 equiv of HMPA (Figure
8). The strength of C-Li association is similar to that of
lithiomethanes bearing two carbanion-stabilizing groups such
as Me3Si(PhS)CHLi or (PhS)2CHLi.1i The free carbanion of the
(28) NMR signals of nuclei coupled to a 7Li nucleus with a short T1
value show a characteristic peak shape in which the outer two pairs of the
1:1:1:1 quartet first collapse to a broad doublet.29,30 Frequently the natural
abundance 6Li signals (1:1:1 triplet) can be seen in the valley between the
peaks (for a 31P example see Figure 7, for a 13C example see ref 24b).
(29) Gillespie Bacon, J.; Gillespie, R. J.; Quail, J. W. Can. J. Chem.
1963, 41, 3063-3069.
(30) Fraenkel, G.; Subramanian, S.; Chow, A. J. Am. Chem. Soc. 1995,
117, 6300-6307.
(31) Marshall, J. A.; Perkins, J. J. Org. Chem. 1994, 59, 3509.

Substituted Allenyl-Propargyllithium Reagents
J. Am. Chem. Soc., Vol. 121, No. 42, 1999 9775
Figure 9. 13C NMR spectrum of 22-Li in 5:2:1:1 Me2O/Et2O/THF/
pentane at -150 °C.
Alkyl-substituted A/P-Li reagents are likely to have a higher
barrier to A/P isomerization than silyl-substituted ones such as
22-Li because of charge stabilization by the silyl group.
Racemization of related lithium reagents, a process related to
the 1,3-shift, has a 5 kcal higher barrier in alkyl (33A-Li)
compared with a silyl analogue (26A-Li). Simple alkyl substitu-
tion seems unlikely to provide a system in which both isomers
are significantly present (6-Li and 21-Li are optimally substi-
tuted, but are >95% allenyl). We turned to angle strain to alter
the balance between the A and P isomers. The lithium reagent
8-Li has been reported to be a propargyllithium on the basis of
its IR and 1H NMR spectra.13 On the basis of the chemical shifts
of CP (-7.5 ppm) and CC (99.8 ppm), about 40 and 90 ppm
upfield of those found in typical allenyllithiums, we can confirm
this assignment. Thus the increase in ring strain on going from
formally an sp3 to an sp2 cyclopropyl carbon, approximately
13 kcal/mol,32 is sufficient to overwhelm the normal preference
for an allenyllithium structure.
Figure 10. Variable temperature 13C NMR spectra of 26A-Li in Me2O.
The left stack shows CA.
line shape analysis were obtained (see Figure 11).1j,33 The
coalescence temperature is slightly concentration-dependent
(0.13 M to 0.5 M) but is not significantly affected by the
presence of Li+ species (0.16 M 33A-Li, 0.32 M of the separated
ion pair lithium fluorenide1g). The lithium-carbon coupling in
33A-Li is dynamically averaged by intermolecular exchange
already at -90 °C, some 50 °C below the coalescence of the
methyl groups.
The cyclobutylidene allenyllithium 39A-Li shows diaster-
eotopic phenyl groups. Coalescence of ipso carbons (δ 152.3,
152.9) occurred at 25 °C. The data for racemization of this
lithium reagent are not sufficiently accurate for measurement
of activation parameters, but ∆Gq ) 14.5 kcal/mol.
We have also measured the inversion barrier for the silyl-
substituted reagents 26A-Li and 31A-Li, detected by coales-
cence of the SiMe2 groups. The SiMe2 and C-Li regions of
the 13C NMR spectra of 26A-Li are shown in Figure 10, together
with simulations. Only the five lowest temperatures were used
in the analysis (See Figure 11). There is no appreciable
broadening of the Li-C-coupled multiplet even well after
coalescence of the SiMe2 group. The inversion of 31A-Li is
also an intramolecular process.
A lighter touch is needed, so we examined two cyclobutyl-
idene systems, in which the propargyl isomer should be favored
by approximately the difference in strain between cyclobutane
and methylenecyclobutane (1.4 kcal/mol). Compound 37-Li
shows chemical shifts nearly identical to those of other
allenyllithium reagents. Even 39-Li, in which we have maxi-
mally destabilized the allenyl isomer by combining cyclobutane
angle strain effect with the (presumed) tert-butyl effect, is
unambiguously allenic. We have thus been unable to measure
the barrier to interconversion of alkyl or silyl-substituted A and
P isomers.
Discussion
The thermodynamic stability differences between allenyl and
propargyl structures in the absence of perturbing substituents
are small. Methylacetylene actually has a more favorable heat
of formation (∆Hf°298 ) 44.3 kcal/mol) than allene (45.9 kcal/
mol).34 A free-energy comparison using the position of the
equilibrium for the Cope rearrangement of 41 and 43 gives a
preference of 0.93 kcal/mol at 198 °C for the allenyl isomer.35
Configurational Stability of Allenyl-Propargyllithiums. We
have examined the configurational stability in ethereal solvents
of several of the lithium reagents discussed above. Compounds
32A-Li and 33A-Li bear isopropyl groups that lose their
diastereotopic nonequivalence when the lithium reagents race-
mize on the NMR time scale. For 32A-Li in 1:1 THF/ether the
diastereotopic 13C methyl signals (∆δ ) 0.1 ppm), peaks merge
at about -60 °C, probably due to a convergence of chemical
shifts, rather than a true coalescence. For 33A-Li the peak
separation is larger (∼1 ppm). One of the signals is coincident
with a THF signal, but a coalescence can be detected between
-60 and -20 °C. In 1:1 Me2O/Et2O the isopropyl methyl
signals are free of solvent interferences and spectra suitable for
The preference of elements more electropositive than carbon
for the allenyl position (bonding to CA) is more pronounced.
Trimethylallenylsilane and trimethylpropargylsilane equilibrate
at 555 °C to give 86.1% of the allenyl isomer, ∆G°555 ) 2.9
(32) The strain of methylenecyclopropane is 13 kcal/mol higher than
that of cyclopropane: Greenberg, A.; Liebman, J. F. Strained Organic
Molecules; Academic Press: New York, 1978; pp 66, 94.
(33) See Supporting Information.
(34) Stull, D. R.; Prophet, H. JANAF Thermochemical Tables, 2nd ed.;
National Bureau of Standards, Washington, D.C., 1971.

9776 J. Am. Chem. Soc., Vol. 121, No. 42, 1999
Reich et al.
Placement of anion-stabilizing groups at CP gives systems in
which the natural propensity for allenyl structures conflicts with
the presumed effect of the substituents. For a single substituent
like PhS (41P-Li), 2-PyS (45P-Li), and PhSe or double
substitution by phenyl (10, 12),8 the carbanion-stabilizing effect
dominates since propargyllithium structures result.1a For silyl
substitution (23-Li) the NMR results indicate that intermediate
structures are usually present. For a phenyl group at CP (11,
46A-Li), the structure seems to be allenic.8,40
13C Chemical Shifts and Couplings. The allenyl-propargyl
nature of these lithium reagents is well defined by the chemical
shifts of the central and lithium bearing carbon atoms, as
indicated by the agreement between the shifts and other data
(JC-Li, IR). In a number of these reagents, the lithium cation
was localized on a single carbon to the limits of detection by
the Li-C coupling criterion. Nevertheless, there was significant
evidence for delocalization of charge to the nonmetalated carbon.
Thus, for allenyllithium 5A-Li the terminal carbon (CP) is shifted
upfield by 30 ppm from that in allene.6 This is similar to that
found in a substantially localized allyllithium reagent (47,
probably an aggregate) in pentane solvent.41a
Figure 11. Barriers to configurational inversion of 26A-Li (∆Gq
-138
) 6.1 kcal/mol, ∆Hq ) 6.8 ( 0.4 kcal/mol, ∆Sq ) -5 ( 3 eu in Me2O),
31A-Li (∆Gq
) 9.1 kcal/mol, 5:3:2 Me2O/THF/Et2O), 32A-Li
-93
(∆Gq-65 > 9.6 kcal/mol, THF), 33A-Li (∆Gq-43 ) 10.8 kcal/mol, ∆Hq
) 11.2 ( 0.2 kcal/mol, ∆Sq ) 12 ( 3 eu in 1:1 Me2O/Et2O), and
39A-Li (∆Gq g 14.5 kcal/mol in 3:2 THF/ether).
25
kcal/mol.36 There are examples of propargyl stannanes and
boranes where “complete” equilibration to the allenyl isomer
was observed.1d,31
A PES study showed gas-phase acidities (∆G°acid) of 373 kcal/
mol for allene and 375 kcal/mol for the CH3 of methylacety-
lene.37 These authors concluded that C3H3 in the gas phase has
an allenyl structure (dC-H bent, CH2 planar) rather than
propargyl (CH2 pyramidal, CH linear). These results are
consistent with several high-level computational studies of the
anion, which have predicted an allenyl-like structure, with no
local minimum for the propargyl structure.9,10c-d,38 For the
lithium reagent an allenyl rather than propargyl structure was
also found, with significant distortion from linearity to accom-
modate lithium contacts at both ends of the anion as in 3.
We have examined a sufficiently wide range of alkyl-
substituted A/P-Li reagents to show that allenyl structures are
favored in solution by a significant margin. Since 8-Li has a
propargyl structure and 37-Li has an allenyl structure, the
difference between A and P isomers is less than the extra strain
of a methylene cyclopropane (about 14 kcal/mol) but more than
the destabilization of a methylenecyclobutane (1.4 kcal/mol) plus
the destabilization of a partially delocalized carbanion center
by a methyl or tert-butyl group (>1 kcal/mol).39 When carban-
ion-stabilizing groups (Ph (11A-Li),8 MeO (44A-Li),10c SiR3
(28A-Li), SAr1a) are placed at CA only allenyllithium structures
are seen.
Many lithium reagents having Li on an sp2-hybridized carbon
(such as phenyllithium,1h,41b vinyllithium6) show strong down-
field shifts of the Li-bearing carbon. The allenyllithium reagents
also show this effect, although it appears to be smaller (13 ppm)
than that for aryl and vinyllithium reagents The origin of these
shifts is related to the favorable HOMO-LUMO relationship
at sp2 carbanionic carbons.41b,42 Such effects are largely absent
at sp3 centers (such as CP in a propargyllithium reagent) as a
result of orbital symmetry effects (overlap between filled and
unoccupied orbitals after a 90° rotation).
Structures of Allenyl-Propargyllithium Reagents with
Intermediate 13C Chemical Shifts. Five of the A/P-Li reagents
investigated (22-Li, 23-Li, 24-Li, 25-Li, and 29-Li) have
intermediate chemical shift values for CC. We have considered
three explanations: (1) a CIP/SIP equilibrium; (2) a bridged
structure such as 3; (3) and an equilibrium between allenyl and
propargyllithium reagents.
Although the temperature dependence is as expected for a
CIP/SIP equilibrium,43 the observation of unambiguously con-
tacted mono-HMPA complexes for 23-Li rules out the presence
of sufficient SIP species to cause the chemical shift changes
observed. Lithium reagents that are partially ion-separated in
THF or other ether solvents will quantitatively form a SIP on
addition of even 1 equiv of HMPA (e.g., LiI1g).
Computational studies of several A/P-Li reagents have
predicted global minima with structures such as 3, and several
(35) Owens, K. A.; Berson, J. A. J. Am. Chem. Soc. 1990, 112, 5973-
5985.
(40) Maercker A.; Fischenich, J. Tetrahedron 1995, 51, 10209-10218.
(41) (a) Fraenkel, G.; Halasa, A. F.; Mochel, V.; Stumpe, R.; Tate, D. J.
Org. Chem. 1985, 50, 4563-4565. (b) Jones, A. J.; Grant, D. M.; Russell,
J. G.; Fraenkel, G. J. Phys. Chem. 1969, 73, 1624-1626.
(42) Berger, S.; Fleischer, U.; Geletneky, C.; Lohrenz, J. C. W. Chem.
Ber. 1995, 128, 1183-1186.
(43) Buncel, E.; Menon, B. C.; Colpa, J. P. Can. J. Chem. 1979, 57,
999-1005. Buncel, E.; Menon, B. J. Org. Chem. 1979, 44, 317-320.
Hogen-Esch, T. E.; Smid, J. J. Am. Chem. Soc. 1966, 88, 307-318.
(36) Slutsky, J.; Kwart, H. J. Am. Chem. Soc. 1973, 95, 8678-8685.
(37) Oakes, J. M.; Ellison, G. B. J. Am. Chem. Soc. 1983, 105, 2969-
2975. Robinson, M. S.; Polak, M. L.; Bierbaum, V. M.; DePuy, C. H.;
Lineberger, W. C. J. Am. Chem. Soc. 1995, 117, 6766-6778.
(38) Wilmshurst, J. K.; Dykstra, C. E. J. Am. Chem. Soc. 1980, 102,
4668.
(39) For related behavior of some allyltitanium derivatives, see Kasatkin,
A.; Sato, F. Angew. Chem., Int. Ed. Engl. 1996, 35, 2848-2849.

Substituted Allenyl-Propargyllithium Reagents
J. Am. Chem. Soc., Vol. 121, No. 42, 1999 9777
carbanion orbital at CA results in stronger C-Li bonding and
reduces the level of lithium solvation as compared to that of
the formally sp3-hybridized propargyl carbanion carbon CP.
Higher temperatures would favor the isomer with the lower
degree of lithium solvation. Stronger C-Li covalency at CA
would also explain the tendency for more allenyl character in
the contact vs the separated ion pairs seen in most of the systems
studied. The absence of temperature dependence for 23-Li in
2,5-dimethyltetrahydrofuran and ether solution suggests that
either the A/P equilibrium has no enthalpic component in these
solvents or perhaps here symmetrically bridged structures such
as 3 are present.
When methyl groups on silicon are replaced by the more
electron-withdrawing phenyl groups, the fraction of propargyl
isomer increases (compare 22-Li to 23-Li, 24-Li and 25-Li);
consistent with the hypothesis that carbanion-stabilizing groups
(such as Ph, PhS, PhSe, or R3Si) at CP favor the propargyl
isomer. On the other hand, substitution of methyl on silicon by
tert-butyl results in a reagent (26-Li) with predominant or
exclusive allenyl character. This effect can be rationalized on
the basis of solvation effects: the larger SiMe2But group hinders
solvation of the lithium in the propargyl isomer compared with
SiMe3.
Configurational Stability of Allenyl-Propargyllithiums.
Only scattered information is available about the stereodynamics
of A/P-Li reagents. An early report concerns 8P-Li, in which
the 1H NMR AA′BB′ pattern of the cyclopropyl group in
benzene at room temperature coalesced and became an A4
pattern as THF was added. Okamura and co-workers metalated
diastereomeric allenes (49 and its isomer) and found that
deuteration gave identical products.44 However, treatment of
sulfoxide 48 and its isomer with MeLi to give 49, a reaction
that probably goes through an allenyllithium intermediate,
proceeded stereospecifically with retention of configuration,
implying some stereochemical lifetime of the lithium reagent
before protonation occurs.45
Figure 12. Thermodynamic parameters for the A/P equilibration of
22-Li to 25-Li.
crystal structures with this type of bonding have been reported
(12, 14, 15).10a,9 Since the high-level calculations were per-
formed without appropriate solvents on Li, it is not clear whether
such structures should also be present in solution. Some evidence
for bridged structures has been obtained for 1-methoxyallenyllith-
ium,10c in which 6Li-1H HOESY cross-peaks were detected for
the protons on the CdCH2 terminus in the THF-solvated dimer.
Since this structure lacks a proton at CA (as do most of our
structures), it is hard to evaluate the degree of bridging actually
present from the qualitative observation of a cross-peak. C-Li
coupling is seen only for CA. Our failure to detect 13C-7Li
coupling in all except one of the CP-silyl-substituted reagents
(although such coupling can be routinely observed in “normal”
A and P organolithium reagents) suggests a qualitatively
different type of structure, perhaps 3.
Some direct evidence against a bridged structure and in favor
of equilibrating isomers for these reagents comes from IR
spectra. Although not all features of the pertinent region between
1800 and 2200 cm-1 are fully explained in all of these reagents,
compound 41P-Li, which is clearly propargyl in its 13C chemical
shifts, shows absorption only at ∼2150 cm-1. Reagent 6A-Li,
on the other hand, shows only a band at ∼1890 cm-1
.
Compound 23-Li has absorptions at both 2179 and 1894 cm-1
in THF. A bridged structure should show a single intermediate
absorption.
Origin of the Temperature-Dependent Chemical Shifts for
22-Li-26-Li. We feel that an equilibrium between allenyl and
propargyllithium reagents best explains the intermediate 13C
chemical shifts observed in those solvents where the shifts are
temperature-dependent. Why do low temperatures favor the
propargyl isomer?
A thermodynamic analysis of the equilibrium between A and
P structures was performed, using an estimated δ of 165 ppm
for CC in the allenyl (23A-Li) and 98 ppm for the propargyl
isomer (23P-Li) (Figure 12). A principal source of uncertainty
in the thermodynamic parameters is the estimation of the CC δ,
which ranges from 191.5 ppm for monomeric allenyllithium
5A-Li to 162 ppm for 26A-Li. For each compound the propargyl
isomer is enthalpically favored, whereas the allenyl isomer is
entropically favored.
The observed effects can be summarized as follows: (1)
higher temperatures favor the allenyl isomer; (2) solvents less
polar than THF or Me2O such as ether and 2,5-dimethyltet-
rahydrofuran favor the propargyl isomer and the temperature
dependence is lost; (3) SIP’s show propargyl-like 13C chemical
shifts; (4) substitution of Me by Ph on silicon favors the
propargyl isomer; and (5) substitution of Me by t-Bu on silicon
favors the allenyl isomer.
Both the in situ metalation-trimethylsilylation of 5046 and the
KOtBu/DOtBu isotope exchange of 5147 occurred with retention
of configuration. Thus the heteroatom-substituted lithium or
potassium carbanion intermediates have substantial configura-
tional stability: an estimate of 22 kcal/mol was made for
inversion of the chloroallenyl anion formed from 51. It is
expected from analogies with inversion of heterosubstituted
lithium reagents and nitrogen pyramidal inversion that RO- and
Cl-substituted reagents would show exceptionally high barriers.
Our stereodynamic measurements were in part a byproduct
of structural studies and so were not taken under uniform
conditions of solvent and concentration, nor do the compounds
To explain the temperature dependence, we suggest a
difference in solvation between the A and P isomers. Even a
fractional decrease in the average number of solvent molecules
in the A isomer would explain the small ∆H° and ∆S° changes.
Perhaps the more highly directional, formally sp2-hybridized,
(44) van Kruchten, E. M. G. A.; Haces, A.; Okamura, W. H. Tetrahedron
Lett. 1983, 24, 3939-3942.

9778 J. Am. Chem. Soc., Vol. 121, No. 42, 1999
Reich et al.
form a pleasing series. We report the results (Figure 11) because
there is little data available on racemization barriers for A/P-
Li reagents. Our measurement of the inversion barrier for 33A-
Li is the first information on the configurational stability of an
alkyl-substituted allenyllithium. The barrier of 10.7 kcal/mol
in 1:1 Me2O/Et2O is close to that calculated for the inversion
barrier of naked allenyl anion (9 kcal/mol).38 Although the
barrier to configurational inversion of organolithium reagents
is often (but not always1i*,48) lowered by going to more strongly
coordinating solvents,49,13a the inversion/diastereomer intercon-
version barrier of 32A-Li in ether, where the principal aggregate
appears to be a dimer, is not dramatically different from that in
THF.
The Li-C coupling of 33A-Li disappears at approx 50 °C
lower temperature than the coalescence of the methyl groups,
so there are fast intermolecular exchange processes of the C-Li
bond (probably through transient dimers), which must occur
with predominant retention of configuration at carbon. We thus
cannot determine whether the racemization of 33A-Li is intra
or intermolecular.
This behavior is in contrast to that of 26A-Li and 31A-Li,
for which the configurational inversion is intramolecular since
the Li-C-coupled multiplet survives even after coalescence of
the SiMe2 group (Figure 10). Although direct comparison of
the silyl and alkyl-substituted systems is problematic because
the intra- or intermolecular nature of the racemization in the
latter is unknown, it is certain that the barrier to intramolecular
inversion of 33A-Li is at least as high as the one measured
(∆Gq ) 10.8 kcal/mol). The substantial lowering in ∆Gq and
∆Hq on replacing an alkyl by a silyl must result principally
from loosening of C-Li coordination, a consequence of the
carbanion-stabilizing effect of the silyl group. It is interesting
that 31A-Li shows substantially higher configurational stability
than does 26A-Li. Presumably silyl substitution at CA in 31A-
Li reinforces the natural localization of charge at CA, whereas
in 26A-Li the silicon substituent tends to equalize charges on
CA and CP.
monomeric in THF or dimethyl ether; 32A-Li and 33A-Li are
dimeric in ether.
Ring-strain effects result in a propargyl structure when a three-
membered ring spans CP (8P-Li). However, similar 4-ring
structures (37A-Li, 39A-Li) are still allenyllithiums.
All of the A/P-Li reagents investigated with diagnostic
diastereotopic groups (23-Li, 24-Li, 26A-Li, 32A-Li, 33A-Li,
and 39A-Li) underwent slow racemization on the NMR time
scale at low temperatures, with free energies of activation
barriers ranging from 6.1 to 14.5 kcal/mol.
Experimental Section
General. All reactions involving lithium reagents were run in oven
(125 °C) or flame dried flasks under a dry nitrogen atmosphere.
Tetrahydrofuran (THF) and ether were freshly distilled from sodium
benzophenone ketyl. Methyl ether (boiling point -20 °C) was freshly
distilled from alkyllithium reagents. Hexamethylphosphoric triamide
(HMPA) was distilled from CaH2 at reduced pressure and stored under
N2 over molecular sieves. Common lithium reagents were titrated with
n-PrOH or i-PrOH in THF or ether using 1,10-phenanthroline as
indicator.50 Reported reaction temperatures are those of the cooling bath
with -78 °C being maintained by a dry ice/ethanol bath and -20 °C
maintained by a freezer. Bath temperatures are reported for Kugelrohr
distillations.
Precautions. HMPA has long been recognized to have low to
moderate acute toxicity in mammals.51a Inhalation exposure to HMPA,
however, has been shown to induce nasal tumors in rats.51b,c Adequate
precautions must be taken to avoid all forms of exposure to HMPA.
Organoselenium and organotin compounds are toxic and should be
handled with care in a fume hood.
NMR Spectroscopy. Multinuclear low-temperature NMR experi-
ments were run on a wide-bore AM-360 spectrometer at 90.56 MHz
(13C), 139.962 MHz (7Li), 52.998 MHz (6Li), or 145.784 MHz (31P) or
an AM-500 spectrometer at 125.6 MHz (13C) with the spectrometer
unlocked (the field was generally very stable, and only occasionally
did an experiment have to be abandoned due to a field shift). Gaussian
multiplication (GM) was performed using the Gaussian broadening
parameter (GB) equal to the duration of the FID and the line broadening
parameter (LB) equal to -(digital resolution)/GB.
Various techniques were utilized to make unambiguous 13C chemical
shift assignments including the following: DEPT 90, DEPT 135,
QUATD, and 1H-coupled 13C. QUATD is a pulse sequence that
dephases all carbons with one-bond 1H coupling leaving only quaternary
carbons, and additionally provides long-range 1H decoupling. QUATD
proved to be a most important tool since two of the three carbons of
the allenyl-propargyl framework of many of the compounds studied
are quaternary carbons. Additionally 13C-7Li couplings usually had
improved signal-to-noise when one of these techniques was employed.
General Procedure for NMR Spectroscopy of Lithium Reagents.
Samples of the lithium reagent in THF or other solvents (total 4.0 mL
for 10 mm tubes, 0.5 mL for 5 mm tubes) were prepared in thin-walled
NMR tubes, which were oven-dried, fitted with a septum, and flushed
with N2. The outside top portions of the tubes were lightly greased to
make a better seal for the septa, which were held securely by Parafilm.
Silicon grease was placed on the septa tops to seal punctures, and the
tubes were stored at -78 °C until the experiment was performed. The
field was adjusted to a CDCl3 lock sample, the spectrometer unlocked,
and the probe cooled (-90 to -155 °C). Temperatures were measured
using the internal 13C shift thermometer tris(trimethylsilyl)methane1k
or a calibrated RTD (resistance temperature device) accurate to 0.1
°C. The reading was taken 5-10 min after the RTD element had been
Summary. Alkyl-substituted A/P-Li reagents show a pro-
nounced tendency for localized structures, with lithium pre-
dominantly coordinated to the allenyl carbon CA (5A-Li, 6A-
Li, 9A-Li, 21A-Li, 32A-Li, 33A-Li, 34A-Li, 37A-Li, 39A-
Li). Placement of an anion-stabilizing group like triorganosilyl
(28A-Li, 31A-Li) at CA reinforces this tendency. Placement of
anion-stabilizing substituents at CP reduces the preferences for
allenyl structures and may lead either to pure propargyl
compounds (41P-Li), or to intermediate structures (22-Li to 26-
Li). Several of these show a pronounced temperature depen-
dence of their chemical shifts, suggesting that they are mixtures
of allenyl and propargyl isomers in rapid equilibrium on the
NMR time scale. Since decoalescence could not be achieved
even below -150 °C, ∆Gq for A/P interconversion must be
less than 4 kcal/mol. Placement of silyl groups at either CA or
CP leads to easier ion separation, so that SIPs are formed when
more than 1 equiv of HMPA is added. All of the A/P-Li
reagents except the parent (5A-Li) and possibly 8P-Li are
(45) Neef, G.; Eder, U.; Seeger, A. Tetrahedron Lett. 1980, 21, 903-
906.
(46) Hoppe, D.; Gonschorrek, C. Tetrahedron. Lett. 1987, 28, 785-788.
(47) le Noble, W. J.; Chiou, D.-M.; Okaya, Y. J. Am. Chem. Soc. 1978,
100, 7743-7744.
(50) Watson, S. C.; Eastham, J. F. J. Organomet. Chem. 1967, 9, 165-
168.
(51) (a) Kimbrough, R. D.; Gaines, T. B. Nature 1966, 211, 146-147.
Shott, L. D.; Borkovec, A. B.; Knapp, W. A., Jr. Toxicol. Appl. Pharmacol.
1971, 18, 499. (b) J. Natl. Cancer Inst. 1982, 68, 157. (c) Mihal, C. P., Jr.
Am. Ind. Hyg. Assoc. J. 1987, 48, 997. American Conference of Govern-
mental Industrial Hygienists: TLVs-Threshold Limit Values and Biological
Exposure Indices for 1986-1987; ACGIH: Cincinnati, OH, Appendix A2,
p 40.
(48) Maercker, A.; Geuss, R. Angew. Chem., Int. Ed. Engl. 1971, 10,
270-271.
(49) Curtin, D. Y.; Koehl, W. J., Jr. J. Am. Chem. Soc. 1962, 84, 1967-
1973. Knorr, R.; Lattke, E. Tetrahedron Lett. 1977, 3969-3972. Hoell,
D.; Schnieders, C.; Mu¨llen, K. Angew. Chem., Int. Ed. Engl. 1983, 22, 243-
244. Chong, J. M.; Park, S. B. J. Org. Chem. 1992, 57, 2220-2222.

Substituted Allenyl-Propargyllithium Reagents
J. Am. Chem. Soc., Vol. 121, No. 42, 1999 9779
lowered into the probe. (The reading fluctuated within -0.1 °C at this
point.) The sample was inserted and the spectrometer tuned. Since
deuterated solvents were not used, the magnet was shimmed on the
free induction decay (FID) signal of a restricted acquisition of C-3 of
THF, CH2 of ether, or the carbon signal of methyl ether. The probe
was then tuned to the nucleus of interest, and spectra were measured.
Referencing 13C, 7Li, and 31P NMR Spectra. 13C spectra were
referenced to solvent peaks (previously referenced to tetramethylsilane).
The following values were used: THF (C-2 ) 67.96 ppm), methyl
ether (60.08), and ether (CH2 ) 66.51). The 7Li and 31P chemical shift
references were sealed capillary tubes containing LiCl (0.3 M MeOH,
δ 0.00) and PPh3 (1 M THF, δ -6.00), respectively. These tubes were
oven-dried before use.
128.31 (p-SePh), 137.10 (i-SiPh), 135.84 (o-SiPh), 128.84 (m-SiPh),
129.29 (p-SiPh). MS M+ calcd C18H20SeSi 344.0499; exp 344.0506.
13C NMR Spectroscopy of Allenyllithium (5A-Li) in Methyl
Ether. Allene gas (-0.3 mL) and methyl ether (3.5 mL) were
condensed into a 10 mm NMR cooled to -78 °C. t-BuLi (0.4 mL)
was added to dry the reagent and solvent. The contents were distilled
at room temperature into a second 10 mm NMR tube at -78 °C. t-BuLi
(0.38 mL, 1.70 M, 0.64 mmol) was added, and the solution was warmed
to -20 °C for 20 min. A white precipitate formed immediately, and
the solution slowly developed a yellow color. The white solid remained
even after the addition of 2 mL of THF; upon standing it settled to the
bottom of the NMR tube and did not interfere with spectroscopy. An
additional 0.64 mL of t-BuLi was added, and volume was adjusted to
form a 0.27 M solution with a Me2O/THF ratio of 2.5:1. Two sets of
signals were present (Figure 1). Coupling JC-Li in the monomer could
be resolved at -140 °C (J ) 26 Hz). Coupling in both monomer (JC-Li
) 27 Hz) and dimer (JC-Li ) 16 Hz) was resolved at -150 °C.
Assignments were verified using DEPT 90, 135, and QUATD experi-
ments. In a variable temperature experiment the signals of monomer
(δ 41.2, 88.7, 191.7) and dimer (δ 44.8, 87, 198.5) began to broaden
at -130 °C and were fully coalesced at -80 °C (δ 43.8, 87.3, 197.6).
Following the experiment, the solution was quenched with PhMe2SiCl.
Aqueous workup followed by removal of solvent resulted in 178 mg
(95% yield) of a 95:5 ratio of propargyl/allenyl silanes 5P-SiMe2Ph
and 5A-SiMe2Ph. 1H NMR of dimethylphenylpropargylsilane, 5P-
SiMe2Ph (200 MHz, CDCl3) δ: 0.40 (s, 9H), 1.72 (d, J ) 2.9 Hz,
2H), 1.86 (t, J ) 2.9 Hz, 1H), 7.33-7.42 (m, 3H), 7.51-7.60 m, 2
H).57 The NMR sample also contained 25% of PhMe2SiOSiMe2Ph.
The solid (which formed in all experiments in THF or Me2O) was
identified by decanting the solution, adding fresh THF and PhMe2-
SiCl. The product was identified as 1-dimethylphenylsilyl-1-propyne.
13C NMR (75.5 MHz, CDCl3) δ -0.77 (SiCH3), 3.72 (CH3), 81.78
(C), 104.80 (C), 127.80 (CH), 129.25 (CH), 133.61 (CH), 137.46 (C).
Sample Experimental Procedure for Metalation and Spectros-
copy of 2-Alkynes: 13C NMR of 3-Lithio-2,2-dimethyl-3,4-penta-
diene (21A-Li), 0.64 M in Methyl Ether and Ether. t-BuLi (0.38
mL, 1.8 M, 0.70 mmol) was added to a -78 °C solution of
2,2-dimethyl-3-pentyne (21P-H, 86 µL, 0.64 mmol) in methyl ether
(3.6 mL) in a 10 mm NMR tube. The solution was warmed to -20 °C
for 20 min and then cooled back to -78 °C. 13C NMR identified a
clean solution of 21A-Li. The line widths for CC and CP were 3.53
and 4.77 Hz (LB ) 1 Hz), and CA was coupled to lithium (δ 114.6,
1:1:1:1 q, J ) 24 Hz at -140 °C, Figure 2). Assignments were verified
Starting Materials. Commercially available materials were used
except as follows: n-Bu6Li was synthesized from 1-chlorobutane and
6LiE.52 Dicyclopropylacetylene (8P-H),13b 3,3-diphenylcyclobutanone,53
and the trimethylstannyl and/or methylseleno procursors to lithium
reagents 32-Li, 33-Li, and 34-Li1d were prepared by literature
procedures.
General Procedure for the Synthesis of Propargyl Silanes:
Synthesis of 1-Phenyldimethylsilyl-2-butyne, 23P-H.55,56 To an oven-
dried, nitrogen-purged flask with a magnetic stir bar and septum was
added ether (70 mL), 2-butyne (1.6 mL, 20 mmol), and t-BuLi (10.7
mL, 1.78 M, 19 mmol) at -78 °C. The flask was warmed to -20 °C
for 1 h and cooled to -78 °C, and PhMe2SiCl (2.5 mL, 19 mmol) was
added. The solution was warmed to 0 °C and diluted in hexane (70
mL), washed with water (3 × 20 mL) and brine (20 mL), and dried
over Na2SO4. The solvent was removed by rotary evaporation to give
2.475 g (13.1 mmol, 69% yield) of 23P-H as a clear colorless liquid,
which was purified by flash chromatography (5% ether/hexane as
eluent) or kugelrohr distillation (60-80 °C bath temperature, 0.25
mmHg). The reaction must be run in ether; reactions in THF resulted
in up to 20% allene product. 1H NMR (200 MHz, CDCl3) δ: 0.36 (s,
6H), 1.64 (q, J ) 2.86 Hz, 2H), 1.76 (t, J ) 2.86 Hz, 3H), 7.32-7.36
(m, 3H), 7.52-7.57 (m, 2H). 13C NMR (90.56 MHz, CDCl3) δ: -3.48
(CH3), 3.56 (CH3), 6.22 (CH2), 74.51 (C), 75.87 (C), 127.75 (CH),
129.21 (CH), 133.55 (CH), 137.96 (C). IR: v 2224, 2096. MS M+
calcd 188.1021; exp 188.1026.56b
3-Phenylseleno-1-dimethylphenylsilyl-1-butyne (30).1e A solution
of lithium diisopropylamide (2.2 mmol) was prepared by addition of
n-BuLi (1 mL, 2.2 mmol, 2.21 M) to a solution of freshly distilled
diisopropylamine (0.3 mL, 2.2 mmol) in THF (10 mL) at -78 °C.
Propargyl phenyl selenide (138 µL, 1.0 mmol) was added to the solution
at -78 °C, followed in 10 min by MeI (65.4 µL, 1.0 mmol). The
solution was stirred at -78 °C for 10 min, and PhMe2SiCl (176 µL,
1.0 mmol) was added. After 10 min, the reaction mixture was allowed
to warm to -20 °C for 40 min. Ether/hexane (1:1, 30 mL) was added,
and the reaction mixture was washed with H2O (5 × 15 mL), NH4Cl
solution (2 × 15 mL), and brine (2 × 15 mL) and dried over Na2SO4.
Siloxane was removed by preparative TLC, eluting the plate 4× with
distilled hexane as solvent. The plate was allowed to dry after each
elution. The product 30 was a light yellow oil (180.5 mg, 53% yield).
1H NMR (300 MHz, CDCl3) δ: 0.35 (s, 6 H), 1.63 (d, J ) 4.8 Hz, 3
H), 3.95 (q, J ) 4.8 Hz, 1 H), 7.25-7.70 (m, 10 H). 13C NMR (75
MHz, CDCl3) δ: -0.82 (SiMe), 22.33 (CH3), 25.47 (CH), 86.29 (C-
2), 109.16 (C-1), 129.18 (i-SePh), 133.68 (o-SePh), 127.80 (m-SePh),
1
7
using DEPT 135, QUATD, and H-coupled experiments. 21A-Li Li
NMR δ: 0.63. 13C NMR (90.56 MHz) δ: 43.93 (CP), 114.6 JC-Li
)
24 Hz (CA), 178.26 (CC); the tert-butyl signal assignment were
ambiguous. 21P-H (90.56 MHz, THF -125 °C) δ: 3.18 (CP), 74.15
and 89.99 (CA and CC). In ether 21P-H partially metalated in the
presence of t-BuLi to give a single allene species. 13C-7Li coupling
was not resolved. (21A-Li)2 δ: 46.8 (CP), 114.8 (CA), 176.2 (CC); the
tert-butyl signals were not distinguishable from solvent signals.
One of the NMR solutions was trapped with PhMe2SiCl (0.12 mL,
0.73 mmol). An aqueous workup afforded a 66% yield of a 6:1 molar
ratio of 27P-H and phenyldimethylsiloxane.
Sample Experimental Procedure for Spectroscopic Studies of
1-Triorganosilyl-2-butynes: 13C NMR of 1-Lithio-1-dimethylphen-
ylsilyl-2-butyne/3-Lithio-1-dimethylphenylsilyl-1,2-butadiene (23-
Li), 0.16 M in Methyl Ether. t-BuLi (0.4 mL, 1.7 M, 0.67 mmol)
was added to a -78 °C solution of 23P-H (133 µL, 0.64 mmol) in
methyl ether (3.3 mL), or other solvent. The NMR tube was warmed
to -20 °C for 20 min then cooled back to -78 °C. 13C spectra were
obtained at the following temperatures: -155, -145, -121, -95, -65,
and -35 °C (see Table 1 and Figures 5 and 6). Chemical shift
assignments were verified using QUATD, DEPT 135, DEPT 90, and
1H-coupled experiments. Lithium coupling was not observed, although
(52) Fraenkel, G.; Fraenkel, A. M.; Geckle, M. J.; Schloss, F. J. Am.
Chem. Soc. 1979, 101, 4745-4747. Fraenkel, G.; Pramanik, P. J. Chem.
Soc., Chem. Commun. 1983, 1527-1529. Seebach, D.; Ha¨ssig, R.; Gabriel,
J. HelV. Chim. Acta 1983, 66, 308-337.
(53) Maercker, A.; Berkulin, W. Chem. Ber. 1990, 123, 185-192.
(54) Crossland, R. K.; Servis, K. L. J. Org. Chem. 1970, 35, 3195-
3196.
(55) Iyer, S.; Liebeskind, L. S. J. Am. Chem. Soc. 1987, 109, 2759-
2770.
(56) (a) Compound 22P-H has been previously prepared by silylation
of 2-butynylmagnesium bromide (Pornet, J.; Kolani, N.; Mesnard, D.;
Miginiac, L. J. Organomet. Chem. 1982, 236, 177-187.) and methylation
of 3-trimethylsilyl-1-lithiopropyne (Nativi, C.; Taddei, M. J. Org. Chem.
1988, 53, 820-826). (b) Compound 23P-H has also been prepared by
reaction of iodomethyldimethylphenylsilane with 1-propynyllithium: Flem-
ing, I.; Lawrence, N. J. J. Chem. Soc., Perkin Trans. 1 1992, 3309-3326.
7
CP did broaden (line width 40 Hz, residual Li coupling) at -150 °C.
A variable concentration study was performed in 1:1 THF/pentane
in a 5 mm NMR tube. Dilutions were accomplished by adding 500 µL
(57) Majetich, G.; Hull, K.; Casares, A. M.; Khetani, V. J. Org. Chem.
1991, 56, 3958-3973.

9780 J. Am. Chem. Soc., Vol. 121, No. 42, 1999
Table 1. Variable Temperature 13C NMR δ for 23-Lia
Reich et al.
Table 2. Variable Temperature 13C NMR Data for 23-//
+
Li(HMPA)4 in 8 Equiv of HMPA
T (°C) -35
-65
-95
-121
-145
-155
T (°C)
-35
9.19
115.78
65.96
9.69
-65
7.75
113.58
64.75
9.81
-95
7.35
113.02
64.55
9.86
-120
-155
CP
CC
CA
22.24 19.90 17.16 12.72
140.44 139.10 135.60 126
74.41 73.22 72.02 70.18 68.03
7.2
117.43
4.1
109.7b
CP
CC
CA
Me
SiMe2
ipso
7.00
112.57
64.55
9.94
2.49
151.45
6.68
112.23
64.61
10.00
2.56
67.00
8.53
1.83/0.85
147.76
134.96
127.57
127.74
methyl 11.62 11.39 10.93 10.11
9.06
SiMe2
ipso
0.73
0.72
0.84
1.01
1.48/0.97
2.08
151.57
2.34
151.82
2.41
151.68
147.22 147.21 147.26 147.39 147.61
151.15
ortho 134.64 134.66 134.70 134.78 134.88
meta
para
127.38 127.38 127.41 127.47 127.53
127.49 127.53 127.60 127.65 127.71
1
0.1.8:1 (A/P) ratio. H NMR (200 MHz, CDCl3) 23P-D δ: 0.36 (s,
9H), 1.64 (q, peak interference, 1H), 1.75 (d, J ) 2.74 Hz, 3H), 7.28-
7.38 (m, 3H), 7.50-7.57 (m, 2H). 23A-D δ: 0.35 (s, 9H), 1.61 (d, J
) 4.0 Hz, JD-H -0.6 Hz, 3H), 5.02 (q, J ) 4.0 Hz, JD-H -0.6 Hz,
1H), 7.28-7.38 (m, 3H), 7.50-7.57 (m, 2H). 13C NMR (90.56 MHz,
CDCl3) 23P-D δ: -3.49 (SiCH3), 3.54 (CH3), 5.90 (JC-D ) 19 Hz,
CP), 74.51 and 75.85 (CA, CC), 127.74 (ortho), 129.20 (para), 133.63
(meta), 137.93 (ipso). 23A-D δ: -2.31 (SiCH3), 12.95 (CH3), 78.2
(JC-D ) 18 Hz, CA), 80.51 (CP), 127.93 (ortho), 129.01 (para), 133.53
(meta), 138.69 (ipso), 212.12 (CC). IR: 1932 (s), 2100 (w) 2220 (w).
a Measured in parts per million.
of 1:1 THF/pentane with a microliter syringe then removing 0.5 mL
of the solution with a 1 mL syringe. Between 0.64 and 0.08 M at -90
°C, CA changed from 12.1 to 13.9 ppm, CC from 125.9 to 130.3 ppm,
CP from 69.0 to 70.2, and the CH3 signal from 9.9 to 10.5 ppm.
Experiments were performed in 2,5-dimethyltetrahydrofuran as
described above and are reported in Figure 5.
An HMPA titration was performed during one experiment by
removing the sample from the probe and placing the NMR tube into a
-78 °C bath then adding small aliquots of HMPA through the septum
and replacing the NMR tube into the probe. The reagent fully separated
in the presence of 4 equiv of HMPA to form a propargyllithium
separated ion 23-//Li(HMPA)4+ (Figure 8). In the presence of 8 equiv
of HMPA, the chemical shifts of 23-//Li(HMPA)4+ did not move
significantly over a temperature range of -155 to -35 °C. (see Table
2 and Figure 5). The separated ion was stable to proton transfer during
the course of the experiment. 7Li NMR (139.96 MHz, Me2O) δ: 0.50
ppm (Me2O), 0.01 ppm (4-HMPA, quintet); referenced to LiCl in
methanol (0 ppm). The C-H coupling constant (1JC-H) for CP was 138
Hz for the CIP at -69 °C and 142 Hz for the separated ion.
Acknowledgment. We thank the National Science Founda-
tion for support of this research and Wayne Goldenberg and
Rui Tang for the synthesis of 31P-Se and 31A-Li.
Supporting Information Available: Preparative procedures
for precursors, reproductions of 13C and 1H NMR spectra, and
chemical shift data for 6A-Li, 8P-Li, 9A-Li, 22-Li, 23-Li, 25-
Li, 26-Li, 27-Li, 28A-Li, 31A-Li, 32A-Li, 33A-Li, 34A-Li,
37A-Li, and 39A-Li; IR spectra of 6A-Li, 23-Li, and 41P-Li.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Several of the NMR experiments were quenched with CH3OD. A
3:2 THF/ether solution gave a 76% yield of 23A-D and 23P-D in a
JA991719D