Structure and reactivity of lithium diisopropylamide solvated by polyamines: Evidence of monomer- and dimer-based dehydrohalogenations

Article abstract of DOI:10.1021/ja9741279

⁶Li and ¹⁵N NMR spectroscopic studies show that hexane solutions of LDA containing <1.0 equiv of N,N,N',N',N'-pentamethyldiethylenetriamine (PMDTA) per lithium contain a mixture of unsolvated LDA oligomers, monosolvated open dimer, and monosolvated monomer. At > 1.0 equiv of PMDTA per lithium, monomer is the dominant species. Addition of PMDTA to LDA in toluene affords open dimer at low [PMDTA] and a mixture of LDA monomer and benzyllithium (resulting from toluene deprotonation) at high [PMDTA]. The results are compared and contrasted with previous investigations of LDA solvated by N,N,N',N'-tetramethylethylenediamine (TMEDA) and (±)-trans- N,N,N',N'-tetramethylcyclohexanediamine (TMCDA). The reactivities of LDA solvated by TMEDA, TMCDA, and PMDTA were probed by investigating the dehydrohalogenation of (±)-2-exo-bromonorbornane. All three ligands afford qualitatively similar behavior: (1) a maximum reactivity at low ligand concentrations ascribed to monosolvated LDA dimers and (2) ligand- concentration-independent rates at high ligand concentrations ascribed to monosolvated LDA monomers. Structure and rate differences in hexane and toluene solutions are noted.

Full text of DOI:10.1021/ja9741279

J. Am. Chem. Soc. 1998, 120, 4081-4086
4081
Structure and Reactivity of Lithium Diisopropylamide Solvated by
Polyamines: Evidence of Monomer- and Dimer-Based
Dehydrohalogenations
Julius F. Remenar and David B. Collum*
Contribution from the Department of Chemistry, Baker Laboratory, Cornell UniVersity,
Ithaca, New York 14853-1301
ReceiVed December 4, 1997
Abstract: Li and 15N NMR spectroscopic studies show that hexane solutions of LDA containing <1.0 equiv
of N,N,N′,N′′,N′′-pentamethyldiethylenetriamine (PMDTA) per lithium contain a mixture of unsolvated LDA
oligomers, monosolvated open dimer, and monosolvated monomer. At >1.0 equiv of PMDTA per lithium,
monomer is the dominant species. Addition of PMDTA to LDA in toluene affords open dimer at low [PMDTA]
and a mixture of LDA monomer and benzyllithium (resulting from toluene deprotonation) at high [PMDTA].
The results are compared and contrasted with previous investigations of LDA solvated by N,N,N′,N′-
tetramethylethylenediamine (TMEDA) and (()-trans-N,N,N′,N′-tetramethylcyclohexanediamine (TMCDA). The
reactivities of LDA solvated by TMEDA, TMCDA, and PMDTA were probed by investigating the
dehydrohalogenation of (()-2-exo-bromonorbornane. All three ligands afford qualitatively similar behavior:
6
(1) a maximum reactivity at low ligand concentrations ascribed to monosolvated LDA dimers and (2) ligand-
concentration-independent rates at high ligand concentrations ascribed to monosolvated LDA monomers.
Structure and rate differences in hexane and toluene solutions are noted.
enetriamine (PMDTA).6
-10
NMR spectroscopic studies reveal
Introduction
monomeric and dimeric LDA-PMDTA solvates as well as
several important consequences of using toluene as the “inert”
cosolvent. Rate studies of an LDA-mediated dehydrohaloge-
There is a considerable body of literature describing reactions
1-3
of organolithium reagents and related lithium salts.
From
the relatively small number of structural and rate studies and
enormous number of empirical observations have emerged a
series of dictums that describe organolithium reactivities. Three
that have been reiterated frequently are as follows: (1) strong
donor solvents (ligands) promote conversion of aggregates to
monomers or ion pairs; (2) monomers are more reactive than
aggregates; and (3) strong donor solvents enhance reactivity.
In fact, any one of these is often implied to be a corollary of
the other two. Over the past few years we have challenged
10
nation (eq 1) reveal competing elimination mechanisms based
upon LDA monomers as well as evidence of dimer-based
pathways as first proposed by Schlosser and co-workers.1
Contrary to conventional wisdom, the polyamines afford
optimum elimination rates via the putative dimer-based path-
ways.
1,12
3
these ideas as simplistic and, at times, fundamentally flawed.
We describe herein investigations of lithium diisopropylamide
,5
(
(
LDA)4 solvated by N,N,N′,N′-tetramethylethylenediamine
3
TMEDA), (()-trans-N,N,N′,N′-tetramethylcyclohexanedi-
6-10
amine (TMCDA),
and N,N,N′,N′′,N′′-pentamethyldiethyl-
Results
(
1) Seebach, D. Angew. Chem., Int. Ed. Engl. 1988, 27, 1624. For
Solution Structures: LDA/PMDTA. We previously re-
ported the structures of LDA-solvated TMEDA and TMCDA
extensive additional references to general reviews of organolithium
chemistry, see ref 2.
(
2) Reviews of lithium amide structural studies: (a) Gregory, K.;
(9) Remenar, J. F.; Lucht, B. L.; Kruglyak, D.; Romesberg, F. E.;
Gilchrist, J. H.; Collum, D. B. J. Org. Chem. 1997, 62, 5748.
(10) The LDA-mediated elimination of (()-2-exo-bromonorbornane has
been studied previously: Remenar, J. F.; Collum, D. B. J. Am. Chem. Soc.
1997, 119, 5573.
Schleyer, P. v. R.; Snaith, R. AdV. Inorg. Chem. 1991, 37, 47. (b) Mulvey,
R. E. Chem. Soc. ReV. 1991, 20, 167. (c) Collum, D. B. Acc. Chem. Res.
1
993, 26, 227.
(
3) Collum, D. B. Acc. Chem. Res. 1993, 26, 227.
(
4) Lithium dialkylamides are important reagents in synthetic organic
(11) Matsuda, H.; Hamatani, T.; Matsubara, S.; Schlosser, M. Tetrahe-
dron 1988, 44, 2865.
chemistry. For extensive leading references, see refs 7-10.
5) For leading references to other mechanistic, spectroscopic, and
crystallographic studies of LDA, see ref 8.
6) For leading references to TMCDA and PMDTA in organolithium
chemistry, see refs 7-9.
7) For previous spectroscopic investigations of lithium amides solvated
(
(12) For selected applications and mechanistic investigations of lithium
amide-mediated 1,2-eliminations, see: (a) Mass, R. J.; Rickborn, B. J. Org.
Chem. 1986, 51, 1992. (b) Kopka, I. E.; Nowak, M. A.; Rathke, M. W.
Synth. Commun. 1986, 16, 27. (c) Evans, K. L.; Prince, P.; Huang, E. T.;
Boss, K. R.; Gandour, R. D. Tetrahedron Lett. 1990, 31, 6753. (d) Reisdorf,
D.; Normant, H. Organomet. Chem. Synth. 1972, 1, 375.
(13) Bernstein, M. P.; Romesberg, F. E.; Fuller, D. J.; Harrison, A. T.;
Williard, P. G.; Liu, Q. Y.; Collum, D. B. J. Am. Chem. Soc. 1992, 114,
5100.
(
(
by polydentate ligands, see: Lucht, B. L.; Bernstein, M. P.; Remenar, J. F.
J. Am. Chem. Soc. 1996, 118, 10707. Also, see refs 8 and 9.
(8) Remenar, J. F.; Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1997,
1
19, 5567.
S0002-7863(97)04127-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/22/1998

4082 J. Am. Chem. Soc., Vol. 120, No. 17, 1998
Remenar and Collum
Chart 1
Figure 1. Representative NMR spectra of samples containing 0.10 M
6
15
6
[
Li, N]LDA in 2:1 toluene-pentane mixtures. (A) Li NMR spectrum
6
with 2.0 equiv of PMDTA at -95 °C; (B) Li NMR spectrum with 0.5
equiv of PMDTA at -115 °C; (C) partial 15N NMR spectrum with 0.5
equiv of PMDTA at -115 °C.
to be dimer 4 and monomer 5, respectively.8,13 The structure
of LDA solvated by PMDTA has not been described.
Addition of 0.5 equiv of PMDTA (per Li) to 0.1 M solutions
6
15
14
of [ Li, N]LDA in toluene affords substantial concentrations
of open dimer 7, unsolvated cyclic oligomers,1 monomer 6,
and an impurity shown to be PMDTA-solvated benzyllithium
indicate that the benzyllithium concentrations would remain
several orders of magnitude below detection limits. However,
the Li singlet at δ 0.85 ppm is observed only in toluene
solutions of LDA/PMDTA, increases in proportion to the toluene
concentration (using hexane cosolvent), and remains at constant
intensity over a range of [PMDTA] (1.5-10 equiv per Li). The
3,14
6
8
(Figure 1, eq 2). Monomer 6 and benzyllithium 8 are the
1
5
15
N NMR spectra contained no new N resonances, arguing
6
against a new LDA-derived fragment. The Li singlet appears
substantially more slowly in toluene-d8 than toluene, yet reaches
the same maximum intensity. A compelling assignment of 8
came from the C NMR spectrum of LDA/PMDTA in toluene.
The three most readily observable aromatic carbon resonances
at 159.7, 116.6, and 105.6 ppm are similar to those reported
for related benzyllithium derivatives. The Li and C NMR
spectra of benzyllithium prepared independently from n-BuLi/
PMDTA in toluene confirm the assignment. Last, addition of
diisopropylamine to the toluene solution of benzyllithium/
13
only observable forms at >1.0 equiv of PMDTA. Monomer 6
6
1
is characterized by a Li doublet (δ 0.86 ppm, JLi-N ) 8.5
Hz) and an 15N triplet (δ 89.4 ppm, JLi-N ) 8.1 Hz). The
assignment of open dimer 7 also follows directly from the
coupling patterns. The Na-Lib-Nc-Lid connectivity in 7 is
1
19
6
13
6
1
readily discerned from a Li doublet (Lid, δ 0.19 ppm, JLi-N
6
1
6
)
)
1
5.1 Hz), a Li doublet-of-doublets (Lib, δ 2.86 ppm, JLi-N
PMDTA affords a Li spectrum that is indistinguishable from
15
1
10.3 and 5.4 Hz), an N triplet (Na, δ 99.7 ppm, JLi-N )
that derived from LDA/PMDTA in toluene.
15
1
0.0 Hz), and an N quintet (Nc, δ 75.7 ppm, JLi-N ) 5.3
Substantially different results emerge when hexane is used
as the cosolvent. Addition of 0.5 equiv of PMDTA (per Li) to
1
5
Hz). The connectivities were confirmed by single-frequency
1
6
3
8
2
17
6
15
decoupling. The η (tridentate) rather than η (bidentate)
PMDTA ligation in 6 and 7 is inferred from the dramatically
different results obtained with TMEDA (see below).
The assignment of the PMDTA-solvated benzyllithium (8)
0.1 M solutions of [ Li, N]LDA in hexane affords unsolvated
LDA cyclic oligomers, monomer 6, and a partially solvated
1
3
(18) The relative pKA’s of carbon acids and amines have been studied
previously: Fraser, R. R.; Mansour, T. S. J. Org. Chem. 1984, 49, 3442.
Chevrot, C.; Perichon, J. Bull. Soc. Chem. Fr. 1977, 421. Bordwell, F. G.;
Bartmess, J. E.; Drucker, G. E.; Margolin, Z.; Matthews, W. S. J. Am. Chem.
Soc. 1975, 97, 3226. Streitwieser, A., Jr.; Ciula, J. C.; Krom, J. A.; Thiele,
G. J. Org. Chem. 1991, 56, 1074. Streitwieser, A., Jr. In ComprehensiVe
Carbanion Chemistry; Buncel, E., Durst, T., Eds.; Elsevier: New York,
1980; Chapter 7. Ahlbrecht, H.; Schneider, G. Tetrahedron 1986, 42, 4729.
Arnett, E. M.; Moe, K. D. J. Am. Chem. Soc. 1991, 113, 7068. Fraser, R.
R.; Mansour, T. S. Tetrahedron Lett. 1986, 27, 331. Arnett, E. M.; Moe,
K. D. J. Am. Chem. Soc. 1991, 113, 7068. Fraser, R. R.; Mansour, T. S.;
Savard, S. J. Org. Chem. 1985, 50, 3232. Fraser, R. R.; Bresse, M.; Mansour,
T. S. J. Am. Chem. Soc. 1983, 105, 7790.
1
8
was elusive at first since literature pKA’s in other solvents
(
14) Kim, Y.-J.; Bernstein, M. P.; Galiano-Roth, A. S.; Romesberg, F.
E.; Williard, P. G.; Fuller, D. J.; Harrison, A. T.; Collum, D. B. J. Org.
Chem. 1991, 56, 4435.
15) For extensive references to the chemistry of lithium amide open
dimers, see ref 8.
16) Gilchrist, J. H.; Harrison, A. T.; Fuller, D. J.; Collum, D. B. J. Am.
Chem. Soc. 1990, 112, 4069.
17) Jutzi, P.; Schl u¨ ter, E.; Kr u¨ ger, C.; Pohl, S. Angew. Chem., Int. Ed.
Engl. 1983, 22, 994. Reich, H. J.; G u¨ dmundsson, B. O¨ . J. Am. Chem. Soc.
996, 118, 6074.
(
(
(
1
(19) Fraenkel, G.; Martin, K. V. J. Am. Chem. Soc. 1995, 117, 10336.

Structure and ReactiVity of LDA SolVated by Polyamines
oligomer that we believe to be open dimer 7 in rapid degenerate
J. Am. Chem. Soc., Vol. 120, No. 17, 1998 4083
2
0
exchange. Addition of 2.0 equiv of PMDTA affords 6 and 7
in a 6:1 molar ratio. Monomer 6 is readily characterized by a
6
1
15
Li doublet (δ 1.58 ppm, JLi-N ) 8.7 Hz) and N triplet (δ
1
9
0.0 ppm, JLi-N ) 8.6 Hz). Dimer 7 manifests a very broad
6
1
Li triplet at -50 °C (δ 2.16 ppm, JLi-N ) 5.1 Hz). Dropping
the probe temperature causes further broadening of the resonance
consistent with a coalescence, but the slow exchange limit
21
showing two discrete lithiums of 7 could not be obtained. The
partial solvation (<1.0 solvent per Li) is inferred from the
disappearance of 7 at higher [PMDTA].
We concluded the structural studies by determining the
relative binding constants of PMDTA and TMCDA to the LDA
monomer. Competition of 2.0 equiv of PMDTA and 2.0 equiv
of TMCDA in toluene at -80 °C afforded a 1:9 mixture of
monomers 5 and 6 (eq 3) corresponding to Keq ) 16 and ∆G°
Figure 2. Plot of kobsd vs [TMEDA] in toluene cosolvent for the
elimination of (()-exo-2-halonorbornane (0.004 M) by LDA (0.10 M).
The (+) symbol represents data obtained at 0 °C with (()-exo-2-
chloronorbornane and the (b) symbol represents data obtained at -40
°C with (()-exo-2-bromonorbornane.
)
1.1 kcal/mol ((10%). This is similar to the 1.5 kcal/mol
preference of PMDTA for the lithium hexamethyldisilazide
7
(LiHMDS) monomer.
Rate Studies. (a) General. We extended previously re-
ported rate studies of the LDA-mediated dehydrobrominations
of (()-2-exo-bromonorbornane (1, eq 1)10 to include LDA
coordinated by TMEDA, TMCDA, and PMDTA. The LDA
14
used to prepare stock solutions was isolated and recrystallized.
The LDA concentrations (0.025-0.40 M) were maintained high
relative to 1 (0.004 M) to ensure pseudo-first-order conditions.
These conditions also precluded formation of substantial
concentrations of mixed aggregates arising from incorporation
22
of the LiBr byproduct.
The ligand concentrations were
adjusted by using hydrocarbon cosolvent. The rates were
monitored via GC analysis of quenched aliquots following the
1
0
decrease of 1 relative to an internal undecane standard. The
decay of 1 displayed first-order behavior over >3 half-lives for
all LDA-ligand combinations. Substantial kinetic isotope
effects (kobsd(H)/kobsd(D) ) 1.8-4.1) determined by comparing 1
with the 3,3-dideuterio analogue (1-d2)23 confirmed a rate-
limiting proton abstraction in all instances.
Figure 3. Plot of kobsd vs [PMDTA] in hexanes cosolvent for the
elimination of (()-exo-2-bromonorbornane (0.004 M) by LDA (0.10
M) at -40 °C. The (+) symbol represents data that were not included
in the fit.
LDA orders in the range 0.65-0.85. Further deconvolution of
the rate data for the LDA/TMEDA-mediated dehydrohaloge-
nation proved difficult due to the temperature-dependent des-
1
0
olvation of LDA dimer 4 noted previously. Fortunately, rate
studies with LDA/TMCDA and LDA/PMDTA provide a clearer
mechanistic picture.
(
b) LDA/TMEDA. Plots of the pseudo-first-order rate
constant (kobsd) vs [TMEDA] in hexane are shown in Figure 2.
At high [TMEDA] and low temperature (-40 °C), a zeroth-
order TMEDA dependence is apparent. A rate maximum at
(c) LDA/PMDTA. A plot of kobsd vs [PMDTA] for the
LDA-mediated elimination of 1 is illustrated in Figure 3. The
zeroth-order PMDTA dependence at high [PMDTA]sconditions
affording monomer 6sattests to an elimination mechanism
requiring one PMDTA per Li in the transition structure. A plot
of kobsd vs [LDA] in 3.0 M PMDTA (Figure 4) with least-squares
fit to eq 4 affords an LDA order of 1.02 ( 0.03 consistent with
a monomeric LDA fragment in the rate-limiting transition
structure (eq 5).
0
.75 equiv of TMEDA in conjunction with no measurable rate
in the absence of TMEDA attests to a mechanism with a
transition structure bearing a lower TMEDA:lithium ratio than
the reactants (<1.0 equiv of TMEDA per Li). Dehydrohalo-
genations at -20 and 0 °C display more gradual saturation
behavior. (Alkyl chloride 2 was employed at 0 °C to maintain
observable rates.) Determinations of the reaction order in LDA
at several TMEDA concentrations and temperatures afforded
(20) The degenerate site exchange of lithium amide open dimers has
been most carefully investigated for TMEDA-solvated lithium 2,2,4,6,6-
9
pentamethylpiperidide.
(
21) We generally find that lithium amide and other organolithium
dynamic exchange processes observable by NMR spectroscopy are con-
siderably faster in hexane than in toluene.
(
amide mixed aggregation, see: Romesberg, F. E.; Collum, D. B. J. Am.
Chem. Soc. 1994, 116, 9198.
At low [PMDTA] we observe an approximate inverse-first-
24
order PMDTA dependence, consistent with a second pathway
requiring a desolvation. Monitoring kobsd vs [LDA] with 2.0
equiv of PMDTA in hexane (1.0 equiv of uncoordinated
22) For an extensive bibliography and leading references to lithium
(
23) Brown, H. C.; Krishnamurthy, S. J. Am. Chem. Soc. 1972, 94, 7159.
Creary, X.; Hatoum, H. N.; Barton, A.; Aldridge, T. E. J. Org. Chem. 1992,
7, 1887. Bartsch, R. A.; Lee, J. G. J. Org. Chem. 1991, 56, 212.
(24) A fit of kobsd vs [PMDTA] (0.1-3.5 M) to the expression kobsd ) k′
n
5
+ k′′[PMDTA] affords n ) -0.74 ( 0.06. (See Figure 3.)

4084 J. Am. Chem. Soc., Vol. 120, No. 17, 1998
Remenar and Collum
Figure 4. Plot of kobsd vs [LDA] in 3.0 M PMDTA with hexanes
cosolvent for the LDA-mediated elimination of (()-exo-2-bromonor-
bornane (0.004 M) at -40 °C.
Figure 6. Plot of kobsd vs [LDA] in 2.0 M TMCDA with toluene
cosolvent for the LDA-mediated elimination of (()-exo-2-bromonor-
bornane (0.004 M) at -40 °C.
the rate at low [TMCDA] is qualitatively consistent with a
second mechanism requiring a desolvation.
Discussion
LDA-Polyamine Solution Structures. NMR spectroscopic
5
investigations described previously and those described herein
reveal the structure of LDA-ligand complexes to be highly
dependent upon the ligand and its concentration. For example,
TMEDA affords exclusively disolvated dimer 4. The TMEDA
ligands on 4 are weakly coordinating, displaying an appreciable
tendency to dissociate near ambient temperatures to afford a
mixture of unsolvated cyclic oligomers. TMCDA shows a
substantially higher affinity for LDA, affording exclusively
8
Figure 5. Plot of kobsd vs [TMCDA] in toluene cosolvent for the
elimination of (()-exo-2-bromonorbornane (0.004 M) by LDA (0.10
M) at -40 °C.
monomer 5 at all TMCDA concentrations. PMDTA affords
monomer 6 at >1.0 equiv per Li and predominantly open dimer
when only 0.5 equiv of PMDTA per Li is added.
7
The spectroscopic investigations of LDA/PMDTA mixtures
PMDTA) affords an LDA order of 1.22 ( 0.12. Under these
conditions LDA exists as a mixture of monomer 6 and up to
revealed interesting hydrocarbon cosolvent effects. While open
dimer 7 is observable at 0.5 equiv of PMDTA in toluene,
monomer 6 becomes the exclusive form at g1.0 equiv. Similar
behaviors were observed when cumene and mesitylene are used.
In contrast, limited concentrations of open dimer 7 persist well
beyond 1.0 equiv of PMDTA when hexane is employed.
Pronounced stabilizations of lithium amide monomers by toluene
15% open dimer 7. The isotope effects determined by measur-
ing the reaction rates for 1 and 1-d2 are kH/kD ) 1.8 ( 0.2 at
both 0.10 M and 3.0 M PMDTA.
The dehydrohalogenations display measurably different rates
in toluene than in hexane. The maximum kobsd at <1.0 equiv
of PMDTA is substantially less pronounced in toluene, and the
saturation behavior is more readily established. The reaction
rates at high [PMDTA] are approximately 3 times faster in
hexane. The detectable formation of benzyllithium 8 (see
above) does not appear to accelerate the elimination, consistent
with the greater kinetic basicity of amides relative to alkyllithi-
ums.25
7
-9,26
and related aromatic hydrocarbons
have been tentatively
2
7
ascribed to the large solvent quadrupoles. The hydrocarbon
cosolvents also influence reactivity (see below).
A second consequence of using toluene as the cosolvent is
the appearance of appreciable concentrations of PMDTA-
solvated benzyllithium 8 (eq 2). Toluene is not metalated by
LDA in the presence of TMCDA, TMEDA, or a host of other
8
polydentate ligands. The formation of 8 from LDA/PMDTA
(
d) LDA/TMCDA. A plot of kobsd for the LDA/TMCDA-
mediated elimination of 1 as a function of [TMCDA] (Figure
) is qualitatively similar to that observed for LDA/PMDTA.
is interesting given that pKA measurements suggest such a
metalation to be endothermic by as much as 6.0 kcal/mol.18
Although the literature pKA’s were measured in THF or DMSO
solutions, one might assume that solvation effects on the two
lithium salts would largely cancel. Clearly, this is not correct.
PMDTA is a good ligand for LDA relative to TMEDA,
TMCDA, and related polydentate ligands. It would appear,
therefore, that PMDTA imparts a disproportionately large
stabilization to benzyllithium 8.
5
The zeroth-order TMCDA dependence and first-order LDA
dependence (1.10 ( 0.10 at 2.0 M TMCDA; Figure 6) implicate
a transition structure based upon a monosolvated LDA mono-
mer. The isotope effect determined by measuring the reaction
rates for 1 and 1-d2 at 2.0 M TMCDA is kH/kD ) 3.3 ( 0.3,
confirming the rate limiting proton transfer. A maximum in
(
25) Trecourt, F.; Mallet, M.; Marsais, F.; Queguiner, G. J. Org. Chem.
LDA-Mediated Dehydrohalogenations: Monomer- vs
Dimer-Based Mechanisms. Rate studies of the dehydrohalo-
genation of (()-2-exo-bromonorbornane (1, eq 1) revealed
mechanistic similarities for LDA/TMEDA, LDA/TMCDA, and
LDA/PMDTA. Although the facile desolvation of TMEDA-
1
988, 53, 1367. Veracini, S.; Gau, G. NouV. J. Chim. 1977, 1, 419.
(
(
26) Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1996, 118, 2217.
27) Kumpf, R. A.; Dougherty, D. A. Science 1993, 261, 1708.
Dougherty, D. A. Science 1996, 271, 163. We thank Professor Dougherty
for helpful discussions.

Structure and ReactiVity of LDA SolVated by Polyamines
J. Am. Chem. Soc., Vol. 120, No. 17, 1998 4085
solvated dimer 413 precluded a detailed rate study, the other
two ligands proved more mechanistically tractable. At elevated
ligand concentrations, the elimination rates are independent of
ligand concentration and linearly dependent on the LDA
concentrations, consistent with rate-limiting transition structures
such as 9 and 10 containing monosolvated monomeric LDA
fragments. Similar results were obtained with LDA solvated
by vicinal amino ethers.10
and TMCDA is not reflected in the nearly equal elimination
rates, suggesting that ground state and transition state solvation
effects cancel.
2
9
Which pathway affords the highest reaction rates under
optimal conditions? Conventional wisdom seems to suggest
that optimum rates will be obtained by adding strong ligands
to promote the monomer-based pathway. In this particular rate
study, the maximum rates are obtained by promoting the putative
dimer-based pathway at low concentrations of the inferior ligand.
Similar conclusions were drawn from rate studies of LDA-
3
0
mediated imine metalations.
Some comment on the role of the hydrocarbon cosolvent is
warranted. In several investigations we have noted a surprising
stabilization of lithium amide monomers by toluene and related
aromatic hydrocarbons eVen in the presence of excess donor
7
-9,26
ligands.
For instance, monomer 6 was found to be
stabilized relative to open dimer 7 in toluene compared to
hexane. However, toluene also causes a measurable (up to
3
-fold) suppression of both the monomer- and dimer-based
eliminations when compared to hexane. While the rate inhibi-
tions are not large, we feel they are remarkable for a change
from one lipophilic cosolvent to another. Beak reported 5-fold
changes in the absolute stereocontrol for a reaction of alkyl-
lithium/sparteine complexes by simply substituting toluene for
31
hexane. What makes aromatic solvents so special? As noted
2
7
by Dougherty, the large quadrupoles of benzenoids may be
important. Exactly how the highly quadrupolar aromatic
solvents influence organolithium structure and reactivity remains
unclear and an important issue.
We feel that the high elimination rates at low ligand
concentrations support the hypothesis first put forth by Schlosser
and co-workers that dehydrohalogenations can proceed via open
11
dimers. The complete absence of elimination when no ligand
is added indicates that solvation is required at the rate limiting
transition structure. The rate maxima at <1.0 equiv of ligand
per Li (or, in other words, the inverse dependencies of the rates
on the ligand concentrations) implicate rate limiting transition
structures containing <1.0 equiv of ligand per Li. We infer,
therefore, that the rate limiting transition structures contain 0.5
equiv of ligand per Li. Bolstered by spectroscopically observ-
able open dimer 7, we reiterate Schlosser’s hypothesis by
invoking open dimer transition structures such as 11. Evidence
Summary and Conclusions
Investigations of the structure and reactivity of polyamine-
solvated LDA led to the following general observations and
conclusions:
1
. Treatment of LDA with excess TMEDA, TMCDA, and
PMDTA affords (respectively) dimer 4, monomer 5, and
monomer 6. Treatment of LDA with <1.0 equiv of PMDTA
affords substantial concentrations of open dimer 7, while
TMEDA and TMCDA afford 4 and 5 (respectively) along with
unsolvated oligomers. Toluene and several other aromatic
hydrocarbons appear to stabilize PMDTA-solvated monomer 6
relative to open dimer 7 as found in previous investigations of
amine-solvated lithium amides.
implicating the importance of open dimers has been accumulat-
ing rapidly.15 Whether the bromide ion interacts with one of
28
the two lithiums is not obvious. Interestingly, investigations
of the dehydrohalogenation with LDA solvated by vicinal amino
2
. Toluene solutions of LDA/PMDTA show appreciable
ethers (MeOCH2CH2NR2) did not provide evidence of such a
dimer-based mechanism.10
concentrations of PMDTA-solvated benzyllithium 8 in equilib-
rium with monomer 6, despite pKA measurements in other
solvents suggesting the metalation of toluene by LDA to be
highly endothermic. The solvent-dependent relative acidities
of diisopropylamine and toluene are ascribed to an unusually
high stabilization of benzyllithium by PMDTA. The results
highlight the limitations of extrapolating pKA’s from one solvent
to another.
Influence of Solvation and Aggregation on Reactivity.
Precisely how solvation and aggregation influence organolithium
reactivity is poorly understood. For example, a perennial issue
is whether monomers or aggregates are more reactive. From a
simplistic thermochemical perspective, the least stable reactant
should be the most reactive. Consequently, forcing the forma-
tion of monomers by adding monomer-stabilizing ligands should
attenuate their reactivity. Indeed, the relative reaction rates via
the monomer pathway (at 3.0 M ligand concentration) are as
follows: TMEDA, 1.0; TMCDA 0.17; PMDTA, 0.19. The
highest reactiVity is obserVed for the ligand that affords dimeric
rather than monomeric LDA in the ground state. Furthermore,
the approximate 1.0 kcal/mol difference in binding of PMDTA
3
. Rate studies of LDA-mediated dehydrohalogenations (eq
1
) reveal qualitatively similar behavior for three polyamines.
At high ligand concentrations, the eliminations are found to
proceed via transition structures based upon monosolvated
monomers (e.g., 9 and 10) akin to those noted in related rate
10
studies. At low ligand concentrations, more efficient pathways
suggested to involve lithium amide open dimers (e.g., 11)
(
28) We investigated the stereochemistry of the elimination using (()-
10
3
-exo-deuterio-2-exo-bromonorbornane as described previously. However,
(30) Bernstein, M. P.; Collum, D. B. J. Am. Chem. Soc. 1993, 115, 789.
Bernstein, M. P.; Collum, D. B. J. Am. Chem. Soc. 1993, 115, 8008.
(31) Wu, S.; Lee, S.; Beak, P. J. Am. Chem. Soc. 1996, 118, 715.
(32) Kofron, W. G.; Baclawski, L. M. J. Org. Chem. 1976, 41, 1879.
(33) Hall, P.; Gilchrist, J. H.; Harrison, A. T.; Fuller, D. J.; Collum, D.
B. J. Am. Chem. Soc. 1991, 113, 9575.
all ligands and ligand concentrations afforded approximately 70-80% syn-
exo elimination, affording little mechanistic insight.
(
29) There are a number of reports where ostensibly weaker solvent-
lithium interactions lead to increased overall reaction rates. For an extensive
bibliography, see ref 10.

4086 J. Am. Chem. Soc., Vol. 120, No. 17, 1998
Remenar and Collum
dominate. Detection of the dimer-based pathway supports a
proposed elimination mechanism proffered by Schlosser and
co-workers.11
a second time. A series of oven-dried, nitrogen-flushed 5 mL
serum vials (5-10 per rate constant) fitted with stir bars were
charged with the LDA stock solution and brought to the desired
temperature (( 0.2 °C) with use of a constant-temperature bath
fitted with a National Bureau of Standards thermometer. The
(()-2-exo-bromonorbornane was added as a 0.047 M stock
solution in the appropriate ligand-toluene mixture containing
undecane (0.047 M) as a GC standard. The vessels were
periodically quenched with 1:1 H2O-THF at intervals chosen
to ensure an adequate sampling of each of the first three half-
lives. The quenched aliquots were extracted into Et2O, and the
extracts were analyzed via capillary GC. Analysis using
Hewlett-Packard GC fitted with an autoinjector and a 60 m DB-5
column (J & W Scientific) provided excellent reproducibility.
The eliminations were monitored by following the decrease of
the (()-2-exo-bromonorbornane (1) relative to the internal
undecane standard. Following the formation of the norbornene
afforded equivalent rate constants ((10%). Rate constants were
determined by numerical integration with a convergence protocol
using the Scientist distributed by MicroMath. Nonlinear least-
squares fits to the integral form of the rate laws afforded
numerically indistinguishable results. The reported errors
correspond to one standard deviation. The observed rate
constants were shown to be reproducible within (5%.
4
. While higher organolithium reactivities are often associ-
ated with strongly coordinating polydentate ligands and ac-
companying deaggregations, the studies described herein reveal
optimal rates from the putative dimer-based pathways.
Experimental Section
Reagents and Solvents. All amines and hydrocarbons were
distilled by vacuum transfer from blue or purple solutions
containing sodium benzophenone ketyl. The hydrocarbon stills
6
contained 1% tetraglyme to dissolve the ketyl. Li metal (95.5%
enriched) was obtained from Oak Ridge National Laboratory.
6
6
15
The [ Li]ethyllithium used to prepare the [ Li, N]LDA was
1
4
prepared and purified by the standard literature procedure.
6
15
[
Li, N]LDA was prepared and isolated as an analytically pure
14
solid as described previously. The diphenylacetic acid used
to check solution titers32 was recrystallized from methanol and
sublimed at 120 °C under full vacuum. Air- and moisture-
sensitive materials were manipulated under argon or nitrogen
with use of standard glovebox, vacuum line, and syringe
techniques.
NMR Spectroscopic Analyses. Samples for spectroscopic
analyses were prepared by using a sample preparation protocol
described in detail elsewhere.33 Standard Li, N, and C NMR
6
15
13
Acknowledgment. We thank the National Institutes of
Health for direct support of this work and Hoffmann-La Roche
for a Roche Bioscience Fellowship to J.F.R. We also acknowl-
edge the National Science Foundation Instrumentation Program
(CHE 7904825 and PCM 8018643), the National Institutes of
Health (RR02002), and IBM for support of the Cornell Nuclear
Magnetic Resonance Facility.
spectra were recorded on a Varian XL-400 spectrometer
operating at 58.84, 40.52, and 100.58 MHz (respectively) or
on a Varian Unity 500 spectrometer operating at 73.57, 58.84,
6
15
13
and 125.76 MHz (respectively). The Li, N, and
C
6
resonances are referenced to 0.3 M [ Li]LiCl/MeOH at -100
C (0.0 ppm), neat Me2NEt at -100 °C (25.7 ppm), and the
°
toluene methyl resonance at -100 °C (20.4 ppm), respectively.
Kinetics. For a kinetic run corresponding to a single rate
constant, a relatively concentrated (0.5-0.8 M) stock solution
of LDA in a ligand-toluene solution was prepared and titrated
to determine the precise concentration. The solution was diluted
to a concentration appropriate for the particular series and titrated
Supporting Information Available: 6Li and 15N NMR
spectra of LDA/PMDTA and rate data for the dehydrohaloge-
nations (17 pages, print/PDF). See any current masthead page
for ordering information and Web access instructions.
JA9741279