The steric course of S(E)2 reactions of unstabilized α- aminoorganolithiums: Distinguishing between SET and polar mechanisms

Article abstract of DOI:10.1021/ja9923235

The competing mechanisms that determine the steric course of electrophilic aliphatic substitution reactions (S(E)2) of configurationally stable, unstabilized α-aminoorganolithiums are compared. The steric course of the reaction of lithiated pyrrolidines and piperidines with electrophiles is variable, such as when comparisons are made between two electrophiles and a single organolithium, or between two organolithiums and a single electrophile. The possible pathways considered are single electron transfer (SET) and competing polar substitutions (S(E)2ret vs S(E)2inv). Catalysis of organolithium racemization by the electrophile was eliminated as a possible source of racemic products. When the products are completely racemic, our evidence suggests that SET is the most likely mechanism; when polar pathways are operative, stereoselectivities vary from 75 to 100%, and may be invertive or retentive at the carbanionic carbon, depending on the electrophile.

Full text of DOI:10.1021/ja9923235

3344
J. Am. Chem. Soc. 2000, 122, 3344-3350
The Steric Course of SE2 Reactions of Unstabilized
R-Aminoorganolithiums: Distinguishing between SET and Polar
Mechanisms
Robert E. Gawley,* Eddy Low, Qianhui Zhang, and Roy Harris
Contribution from the Department of Chemistry, UniVersity of Miami, Coral Gables, Florida 33124-0431
ReceiVed July 6, 1999. ReVised Manuscript ReceiVed February 4, 2000
Abstract: The competing mechanisms that determine the steric course of electrophilic aliphatic substitution
reactions (SE2) of configurationally stable, unstabilized R-aminoorganolithiums are compared. The steric course
of the reaction of lithiated pyrrolidines and piperidines with electrophiles is variable, such as when comparisons
are made between two electrophiles and a single organolithium, or between two organolithiums and a single
electrophile. The possible pathways considered are single electron transfer (SET) and competing polar
substitutions (SE2ret vs SE2inv). Catalysis of organolithium racemization by the electrophile was eliminated
as a possible source of racemic products. When the products are completely racemic, our evidence suggests
that SET is the most likely mechanism; when polar pathways are operative, stereoselectivities vary from 75
to 100%, and may be invertive or retentive at the carbanionic carbon, depending on the electrophile.
Functionalized organolithiums such as those having an oxygen
retention in eq 2! Both Beak and Clayden have found divergent
behavior in benzylic lithiums having a chirality axis as well a
chirality center on the lithium-bearing carbon, such as the
example in eq 3.14,19 All of these examples are benzylic
organolithiums, and an invertive transition state will obviously
be stabilized by resonance. Clearly, the steric course of an SE2
reaction is not limited in the way an SN2 reaction is, but the
factors that determine steric course are not readily apparent.
or nitrogen on the metal-bearing carbon continue to grow in
importance as reactivity in electrophilic substitutions is explored
(see refs 1-14 for reviews). One often puzzling aspect of such
studies is the steric course of the reaction, which may proceed
with retention or inversion, depending on the electrophile (the
suffixes “inv” and “ret” may be used to distinguish these
possibilities15). For example, Hoppe showed that the configu-
rationally stable R-oxyorganolithium shown in eq 1 reacted with
esters and methanol with retention, while reacting with acid
chlorides and acetic acid with inversion.16 Hammerschmidt
reported that a similar R-oxyorganolithium reacts with a
trialkylstannyl chloride with inversion.17 Beak found steric
dichotomy with the R-aminoorganolithium in eq 2, which
reacted with acid chlorides with retention and carbon dioxide
with inversion.18 Note also the dichotomy between these two
systems: acid chlorides react with inversion in eq 1 and with
To study the steric course of an electrophilic substitution,
one must know the absolute configuration of the organolithium,
and it must be configurationally stable under the conditions of
the reaction. Several years ago, we reported that 2-lithiopyrro-
lidines and -piperidines are among the most stable of function-
alized organolithiums,20,21 which makes them ideal candidates
for stereochemical studies. They also have considerable synthetic
potential,12 in that they react with a broad range of electrophiles
(unlike many functionalized organolithiums) to give substitution
products in good to excellent yields,22,23 and undergo stereo-
selective [2,3]-sigmatropic ylide rearrangements.24 In exploratory
studies of the steric course of these reactions, we found that
several trends emerged, as summarized in Scheme 1.22 With
these lithiated heterocycles, the steric course took three broadly
defined paths: complete retention, varying degrees of inversion,
and complete racemization, depending on the electrophile and
the size of the heterocyclic ring. Note that the lithium in these
compounds is not chelated by a carbonyl, as it is in the examples
of eqs 1-3, although it is probably bridged to the nitrogen.25
(1) Beak, P.; Reitz, D. B. Chem. ReV. 1978, 78, 275.
(2) Beak, P.; Zajdel, W. J.; Reitz, D. B. Chem. ReV. 1984, 84, 471.
(3) Meyers, A. I. Aldrichim. Acta 1985, 18, 59.
(4) Highsmith, T. K.; Meyers, A. I. In AdVances in Heterocyclic Natural
Product Synthesis; Pearson, W. H., Ed.; JAI: Greenwich, CT, 1991; Vol.
1, p 95.
(5) Gawley, R. E.; Rein, K. In ComprehensiVe Organic Synthesis.
SelectiVity, Strategy, and Efficiency in Modern Organic Chemistry; Trost,
B. M., Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 3, p 65.
(6) Gawley, R. E.; Rein, K. S. In ComprehensiVe Organic Synthesis;
Schreiber, S., Ed.; Pergamon: Oxford, 1991; Vol. 1, p 459.
(7) Meyers, A. I. Tetrahedron 1992, 48, 2589.
(8) Hoppe, D.; Hintze, F.; Tebben, P.; Paetow, M.; Ahrens, H.;
Schwerdtfeger, J.; Sommerfeld, P.; Haller, J.; Guarnieri, W.; Kolczewski,
S.; Hense, T.; Hoppe, I. Pure Appl. Chem. 1994, 66, 1479.
(9) Beak, P.; Basu, A.; Gallagher, D. J.; Park, Y. S.; Thayumanavan, S.
Acc. Chem. Res. 1996, 29, 552.
(10) Yus, M. Chem. Soc. ReV. 1996, 155.
(11) Hoppe, D.; Hense, T. Angew. Chem., Int. Ed. Engl. 1997, 36, 2282.
(12) Gawley, R. E. Curr. Org. Chem. 1997, 1, 71.
(13) Kessar, S. V.; Singh, P. Chem. ReV. 1997, 97, 721.
(14) Clayden, J. Synlett 1998, 810.
(15) Gawley, R. E. Tetrahedron Lett. 1999, 40, 4297.
(16) Carstens, A.; Hoppe, D. Tetrahedron 1994, 50, 6097.
(17) Hammerschmidt, F.; Hanninger, A.; Vo¨llenkle, H. Chem. Eur. J.
1997, 3, 1728.
(18) Faibish, N. C.; Park, Y. S.; Lee, S.; Beak, P. J. Am. Chem. Soc.
1997, 119, 11561.
(19) Thayumanavan, S.; Lee, S.; Liu, C.; Beak, P. J. Am. Chem. Soc.
1994, 116, 9755.
(20) Gawley, R. E.; Zhang, Q. J. Am. Chem. Soc. 1993, 115, 7515.
(21) Gawley, R. E.; Zhang, Q. Tetrahedron 1994, 50, 6077.
(22) Gawley, R. E.; Zhang, Q. J. Org. Chem. 1995, 60, 5763.
(23) Gawley, R. E.; Campagna, S. In ECHET96. Electronic Conference
on Heterocyclic Chemistry; Rzepa, H., Snyder, J., Eds.; Royal Society of
Chemistry: London, 1997 (CD-ROM).
(24) Gawley, R. E.; Zhang, Q.; Campagna, S. J. Am. Chem. Soc. 1995,
117, 11817.
10.1021/ja9923235 CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/25/2000

Reactions of Unstabilized R-Aminoorganolithiums
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3345
Scheme 1. Trends in the Steric Course of Electrophilic
Substitutions of Lithiated N-Methyl and N-Allyl Pyrrolidines
and N-Methyl Piperidines22,23
Scheme 2. Three Limiting Possibilities for Aliphatic
Electrophilic Substitutionsa
Note also that the trends we observed for these “unstabilized”
R-aminoorganolithiums differ from those cited above. For
example, both esters and acid chlorides proceed with retention
of configuration. It is also interesting that the examples of
configurational inversion that we reported in 199522 and 199723
(Scheme 1), which were proven for six examples, are distinct
from those shown in eqs 1-3 in that the transition state for
inversion at the carbanionic carbon is not mesomerically
stabilized. In contrast to the examples in Scheme 1, most, if
not all, nonbenzylic R-amino and R-oxy organolithiums react
with retention, making these pyrrolidino and piperidino lithiums
unique.
With the examples in eqs 1-3, Scheme 1, and others in the
literature, the factors that determine the steric course are not
clear, although hypotheses have been offered. For example,
Hoppe suggests that (in his system, eq 1) a combination of anion
geometry, electrophile coordinating ability, and electrophile
LUMO energy come into play.16 Beak proposes that, for the
R-aminoorganolithium in eq 2, relatively unreactive electrophiles
and/or electrophiles that coordinate to the lithium react with
retention, while fast-reacting and/or noncoordinating electro-
philes react with inversion.18
In this paper, we examine several subsets of data generalized
in Scheme 1 in which the stereoselectivity is variable, or in
which the steric course changes as a function of the electrophile,
and examine the reasons for the differences. We interpret these
results as competition between the three mechanistic pathways
shown in Scheme 2: single electron transfer (SET) reactions
a X is a heteroatom substituent such as N, O, S, Se, etc.
that involve radical intermediates and form racemic products,
and polar bimolecular electrophilic substitution reactions (SE2)
that may proceed with net retention (SE2ret) or inversion (SE-
2inv) of configuration at the metal-bearing carbon. A key
question arises when racemic, or partly racemic products are
obtained: which pathway(s) are competing, SET or nonselective
polar (SE2ret vs SE2inv) substitutions? In this work, both the
starting organolithiums and the reaction products have only one
stereocenter. This is a conscious choice, whose aim is to avoid
diastereomeric bias in transition states. We are currently
evaluating the steric course of diastereoselective reactions, and
will report these findings in due course.
Results and Discussion
(25) (a) Boche, G.; Marsch, M.; Harbach, J.; Harms, K.; Ledig, B.;
Schubert, F.; Lohrenz, J. C. W.; Ahlbrecht, H. Chem. Ber. 1993, 126, 1887.
(b) Becke, F.; Heinemann, F. W.; Ru¨ffer, T.; Wiegeleben, P.; Boese, R.;
Bla¨ser, D.; Steinborn, D. J. Organomet. Chem. 1997, 548, 205.
Carbonyl Electrophiles. For 1 and 2, reactions with carbon
dioxide, benzaldehyde, acetone, and cyclohexanone proceed with

3346 J. Am. Chem. Soc., Vol. 122, No. 14, 2000
Gawley et al.
Scheme 3. Three Routes for Production of Racemic
Benzophenone Adduct
100% retention of configuration, within the limits of detection
(eq 4). On the other hand, benzophenone gives addition products
that are completely racemic (eq 5).22 All of these reactions go
in yields ranging from 50 to 92%. The benzophenone addition
differed from the other reactions in that the reaction mixture
turned blue upon addition of the benzophenone, and ESR
analysis of aliquots of the reaction mixture revealed the presence
of benzophenone ketyl in the reaction mixture.22 This situation
is reminiscent of earlier observations of benzophenone ketyl in
organolithium additions, where presence and absence of the ketyl
was correlated with the presence and absence of stereoselectivity
in carbonyl addition reactions, and an SET mechanism was
inferred.26 In the addition of 1 and 2 to benzophenone, mixtures
of 1,2- (3/4) and 1,6-addition products (5/6) were obtained, and
both were racemic. Through an oversight, we neglected to
mention this fact in our earlier paper.22 We have since observed
varying amounts of these products, the appearance of which
seems to depend on the concentration of organolithium.
It could be argued that the presence of radicals in a reaction
mixture does not necessarily place them on the reaction
coordinate. To address this issue, we consider all the possible
mechanisms for producing racemic products, and then seek to
eliminate all but one.27 Scheme 3 lists three possibilities: (a)
single electron transfer, SET, involves oxidation of the orga-
nolithium to a radical, which couples with the benzophenone
ketyl with no selectivity; (b) benzophenone catalyzes the
racemization of the organolithium, and the racemic organo-
lithium adds by a polar, SE2ret mechanism; and (c) the
organolithium adds by competing SE2ret and SE2inv pathways.
Of these possibilities, competing polar pathways seem to be
the least likely possibility, since all other carbonyl electrophiles
add to these organolithiums with 100% retention (SE2ret). The
most likely alternatives are SET and racemization of the
organolithium, whereby racemization of the organolithium
precedes a polar SE2ret addition. Since organolithiums 1 and 2
are configurationally stable at -80 °C for over an hour in THF-
TMEDA,21 the benzophenone would have to catalyze its
racemization for this mechanism to be operative. Racemization
could occur if SET to the radical/ketyl pair were reversible, or
if a lithium atom transfer reaction took place between the
heterocyclic radical and the organolithium.
The catalytic possibilities were tested by treating the orga-
nolithiums 1 and 2 with a substoichiometric amount of ben-
zophenone, waiting 1 h, then quenching the reaction mixture
with cyclohexanone, which adds with 100% retention. The
product of the reaction contained both cyclohexanone and
benzophenone adducts, and analysis of the enantiomer ratio of
the cyclohexanone adduct was used to determine the degree of
racemization of the organolithium (eq 6). Enantiomer ratios were
determined by NMR analysis of the crude reaction mixture in
the presence of mandelic acid or Mosher acid as a chiral
solvating agent.28 This experiment has been conducted in both
the pyrrolidine and the piperidine series; the results of the
pyrrolidine experiment are shown in Figure 1. The enantiomer
ratio of the N-Boc-stannylpyrrolidine precursor to 1 was
determined by removal of the Boc group, acylation with benzoyl
chloride, and SFC analysis on a Whelk-O Pirkle column.
Comparison with authentic samples of the racemic benzophe-
none and cyclohexanone adducts revealed that the benzophenone
adduct was racemic as expected, and that the enantiomer ratio
(er) of the cyclohexanone adducts matched that of the starting
stannane within the limits of experimental error. Therefore,
benzophenone did not catalyze the racemization of the orga-
nolithium. Having reasonably eliminated two of the three
possibilities, SET is left as the most likely explanation for the
(26) Rein, K. S.; Chen, Z.-H.; Perumal, P. T.; Echegoyen, L.; Gawley,
R. E. Tetrahedron Lett. 1991, 32, 1941.
(27) “Eliminate all other factors, and the one which remains must be
the truth.” -Sherlock Holmes, in: Conan Doyle, A. In The Sign of the
Four. Edited with an Introduction by Christopher Roden; Roden, C., Ed.;
Oxford University Press: Oxford, 1993; p 8.
(28) Benson, S. C.; Cai, P.; Coilon, M.; Tokles, M. A.; Snyder, J. K. J.
Org. Chem. 1988, 53, 5335.

Reactions of Unstabilized R-Aminoorganolithiums
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3347
Scheme 4. Three Possibilities for Nonselective Alkylation
of 1
of configuration, while the same reaction with 1 (96:4 er)
afforded 9 having a 75:25 er (eq 7); this corresponds to 78%
inversion with 1. Although the major trend is toward SE2inv,
there is considerable retention involved as well. N-Allylpyrro-
lidine 11 affords an 87% yield of alkylated product 12 in 88:12
er with the same electrophile, corresponding to 92% inversion
(eq 8).23 The steric course of the reactions shown in eqs 7 and
8 was proven for 9 and 10 by independent synthesis, and 12 by
Pirkle analysis.22,23 Again, the three most reasonable pathways
leading to racemic product are SET, organolithium racemization,
and competing polar pathways (Scheme 4).
Radical probes have long been used to test SET mecha-
nisms,35 including alkylation of lithiated heterocycles.36,37 We
chose the most commonly used probe, hexenyl bromide, to test
for SET in this system. Alkylation of 1 having an er of 96:4 to
97:3 with hexenyl bromide in THF at -78 °C afforded a 70%
yield of hexenyl coupled product (13) with an er of 73:27 (75-
76% inversion, which corresponds to 54% racemization) (eq
9). No significant amounts of cyclopentylmethyl coupled product
were detected (e10% by GC-MS). This negatiVe result does
Figure 1. N-Methyl region of the 300 MHz 1H NMR pyrrolidine
addition products 3 and 5, in the presence of Mosher acid as chiral
solvating agent. (a) Paired N-methyl signals of diastereomeric salts of
rac-3; (b) reaction mixture of eq 6; (c) paired N-methyls of diastere-
omeric salts of rac-5.
mechanistic course of these additions. This conclusion is
supported by the appearance of the substitution products 5 and
6 (eqs 5 and 6) in the reaction mixture. Previous reports have
shown that some organolithium29,30 and organomagnesium31-34
reagents add to aryl ketones by an SET radical mechanism to
give mixtures of 1,2-, 1,4-, and 1,6-addition products.
Unactivated Alkyl Halide Electrophiles. The reaction of 1
with alkyl halides is significantly less stereoselective than
reactions with 2.22 For example, reaction of 1-bromo-3-
phenylpropane with 2 (g99:1 er) yields 10 with 100% inversion
(29) Yamataka, H.; Kawafuji, Y.; Nagareda, K.; Miyano, N.; Hanafusa,
T. J. Org. Chem. 1989, 54, 4706.
(30) Olah, G. A.; Wu, A.; Farooq, O. Synthesis 1991, 1179.
(31) Ashby, E. C.; Bowers, J. R., Jr. J. Am. Chem. Soc. 1981, 103, 2242.
(32) Yamataka, H.; Matsuyama, T.; Hanafusa, T. J. Am. Chem. Soc. 1989,
111, 4912.
(33) Holm, T. Acta Chem. Scand. 1983, B37, 567.
(34) Ashby, E. C. Pure Appl. Chem. 1980, 52, 545.
(35) Ingold, K. U. Acc. Chem. Res. 1980, 13, 317.
(36) Meyers, A. I.; Edwards, P. D.; Rieker, W. F.; Bailey, T. R. J. Am.
Chem. Soc. 1984, 106, 3270.
(37) Gawley, R. E.; Hart, G. C.; Bartolotti, L. J. J. Org. Chem. 1989,
54, 175.

3348 J. Am. Chem. Soc., Vol. 122, No. 14, 2000
Gawley et al.
not proVe the absence of an SET process, since the rates of the
polar and radical coupling reactions have not been measured.
Nevertheless, this same radical probe has been used to demon-
strate the presence of SET processes in at least two previous
studies of coupling reactions of lithiated pyrrolidines and
piperidines, one of which was done in this lab.36,37 In light of
these precedents, we believe that these observations show that
SET oxidation of 1 does not occur to an extent sufficient to
account for the 54% racemization found in the production of
13.
In a second experiment, half an equivalent of hexenyl bromide
was added; after 1 h, the reaction was quenched with cyclo-
hexanone (eq 10). The cyclohexanone adduct, 7, so produced
had the same er as the starting stannane (within experimental
error), eliminating the possibility of racemization of the orga-
nolithium as a possible route to partial racemization. The
Competing Mechanisms. It appears that in electrophilic
substitutions of 1 and 2, the products are completely racemic
when SET is operative. Why does SET intervene in some of
these processes and not others? Eberson has applied Marcus
theory to organic processes,38 and notes that whether a polar
mechanism or an SET mechanism is followed depends on the
energy difference between the oxidation potential of a nucleo-
phile and the reduction potential of an electrophile. This theory
has been used by Arnett to support a polar process for the aldol
addition of lithium pinacolonate to benzaldehyde,39 and by
Bordwell for anionic substitutions on alkyl halides.40 The
oxidation potentials of 1 and 2 are not known. The oxidation
potential of benzyllithium is -1.43 V (THF/HMPA).41 The
reduction potential of benzophenone is -1.87 V (LiClO4/
THF),42 and for benzaldehyde it is -1.94 V (LiClO4/THF).39
There is considerable variance in these numbers, depending on
solvent and supporting electrolyte, however. If the oxidation
potentials of organolithiums 1 and 2 are in the same ballpark
as benzyllithium (probably more negative), the differences in
redox potentials are not large enough to draw any conclusions
about electron transfer rates. All we can say at this point is that
SET competes with polar pathways when the electrophile is
easily reduced, such as benzophenone. The addition of both 1
and 2 to benzaldehyde is retentive at the metal-bearing carbon,
indicating a polar addition, while benzophenone adds by an SET
mechanism. Since benzophenone is more easily reduced than
benzaldehyde, relative oxidation potentials of a number of
functionalized organometallics can be approximated by compar-
ing the steric course of additions to benzophenone and benzal-
dehyde. Pyrrolidine 1 and piperidine 2, which are oxidized by
benzophenone but not benzaldehyde, fall between lithiated
piperidinooxazolines and lithiated tetrahydroisoquinolyloxazo-
lines, which are oxidized by both benzaldehyde and benzophe-
none, and lithiated BOC-pyrrolidines or magnesiated tetrahy-
droisoquinolyl oxazolines or pivalamides, which are not oxidized
by either.43
conclusion is that, for alkylation of lithiopyrrolidine 1, SET is
not involved, and the polar processes, SE2ret and SE2inv,
compete more effectively than in the piperidine series (2). Most
likely, this dichotomy is due to steric factors, either in the
monomer or an aggregate (vide infra).
Activated Alkyl Halides. Benzyl bromide and tert-butyl
bromoacetate, like benzophenone, afford good to excellent yields
of completely racemic products 14-17 from 1 and 2 (eq 11).22
To probe the possibility of SET in this case (Scheme 4), we
required a radical probe that is “activated” similar to a benzylic
or allylic halide. For this we chose cyclopropylmethyl bromide.
From data summarized by Ingold,35 it can be calculated that
the cyclopropylmethyl radical has a half-life of only ∼1 µs at
-80 °C, suffering ring-opening to butenyl radical. When 1 was
allowed to react with cyclopropylmethyl bromide at -78 °C,
GC-MS analysis showed a mixture containing butenyl-coupled
product 18 and cyclopropylmethyl-coupled product 19 in an
approximate 2:1 ratio, along with additional compounds tenta-
tively identified as dimers 20 (eq 12). The two alkylated
pyrrolidines 18 and 19 were prepared independently as proof
of structure. The butenyl compound 18, obtained as shown in
eq 12, was racemic as indicated by CSP-GC; the er of 19 could
not be determined. The independently synthesized racemic 19
was not resolved on CSP-GC, and NMR analysis was compli-
cated by the presence of extra signals due to the N-methylpyr-
rolidinyl dimers, which could not be removed chromatographi-
cally. These dimers were not noticed in alkylations reported
previously,22 perhaps because they are not as prevalent when
the reaction is run in the presence of TMEDA. Nevertheless, a
radical process undoubtedly produces the butenylated product,
and may also produce the dimers, so it appears that SET plays
a major role in the stereorandom coupling of activated alkyl
halides.
Since both SE2ret and SE2inv processes are observed, both
must be allowed by orbital symmetry.44 Since the predominant
route is SE2inv, we will assume that this is the preferred mode
of reaction of these R-aminoorganolithiums, and look for reasons
(38) Eberson, L. Electron-Transfer Reactions in Organic Chemistry;
Springer: Berlin, 1987.
(39) Arnett, E. M.; Palmer, C. A. J. Am. Chem. Soc. 1990, 112, 7354.
(40) Bordwell, F. G.; J. A. Harrelson, J. J. Org. Chem. 1989, 54, 4893.
(41) Jaun, B.; Schwarz, J.; Breslow, R. J. Am. Chem. Soc. 1980, 102,
5741.
(42) Palmer, C. A.; Ogle, C. A.; Arnett, E. M. J. Am. Chem. Soc. 1992,
114, 5619.
(43) Gawley, R. E. In AdVances in Asymmetric Synthesis; Hassner, A.,
Ed.; JAI: Greenwich, CT, 1998; Vol. 3, p 77.
(44) Pearson, R. G. Acc. Chem. Res. 1971, 4, 152.

Reactions of Unstabilized R-Aminoorganolithiums
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3349
Experimental Section
Preparation of stannylpyrrolidines and piperidines and transmetala-
tions to 1 and 2 were described previously.21,22 Methods of er
determination and products of the reactions with all electrophiles except
cyclopropylmethyl bromide were reported in 1995.22
Procedure for Testing the Catalytic Pathways. The reaction with
hexenyl bromide and cyclohexanone is typical. (S)-N-Methyl-2-(tri-
butylstannyl)pyrrolidine (0.63 g, 1.69 mmol) was dissolved in 6 mL
(0.28M) of anhydrous THF. This solution was cooled to -78 °C and
BuLi (1.5 M in hexanes, 1.4 mL, 1.24 equiv) was added to the reaction
by dropwise addition via syringe over 3 min. The reaction was stirred
at -78 °C for 20 min. Neat 6-bromo-1-hexene (115 µL, 0.5 equiv)
was added dropwise over 1 min to the reaction mixture and stirred for
1 h at -78 °C. Then neat cyclohexanone (90 µL, 0.5 equiv) was added
dropwise over 1 min to the reaction mixture and again stirred for 1 h.
The reaction was quenched at -78 °C with 2 mL of 2 M HCl. The
reaction was warmed to room temperature and concentrated in vacuo.
The biphasic mixture was extracted with ether (5 × 2 mL) to remove
the stannane byproducts. The aqueous layer was then basified with 4
pellets of KOH. The basified solution was extracted with ether (5 × 2
mL). The combined organic layers were dried over anhydrous K2CO3,
filtered, and concentrated to a light yellow oil, 0.17 g. GC-MS analysis
showed 6 peaks: 2 were dimers (3 and 4%), N-methyl 2-(cyclopen-
tylmethyl)pyrrolidine (3%), N-methyl-2-(5-hexenyl)pyrrolidine (32%),
N-methyl-2-(1-hydroxycyclohexyl)pyrrolidine (51%), and tetrabutyl-
stannane (7%). Retention times and MS data for the last 3 products
corresponded to authentic samples.22 NMR analysis of the reaction
mixture (5.8 mg) with (S)-(+)-O-acetylmandelic acid (6.0 mg) dissolved
in 0.76 mL of CDCl3 indicated the enantiomers of the cyclohexanone
adduct were present in a 92:8 ratio.
Figure 2. Competing transition structures for SE2 reactions of 1 and
2.
for the lower selectivity based on steric grounds. Assuming the
lithium in 1 and 2 bridges the nitrogen, as was found in the
solid-state structure of R-(dimethylamino)benzyllithium,25 and
as suggested by theory,25,45 we can predict the probable
conformations of 1 and 2, and in turn offer the following
explanation of the steric contribution to the stereochemical
dichotomy.
Lithiopiperidine 1 can adopt two half-chair conformations,
one of which is completely unhindered for an SE2inv reaction,
since the lithium is in the pseudoaxial conformation (Figure
2a, right). For lithiopyrrolidine 2, envelope conformations
probably predominate, with the one illustrated in Figure 2b likely
being the most stable, since it places the methylene “flap” anti
to the solvated cation. Here, an SE2inv approach is discouraged
due to the C-4â-hydrogen shown. This steric interaction may
be enough to allow the SE2ret mechanism to compete. This
explanation, which is illustrated for monomers of 1 and 2, may
be complicated by the aggregation states, which are currently
under study.
Conclusion. Lithioheterocycles 1 and 2 appear to react with
electrophiles that are not easily reduced by a polar mechanism,
and with easily reduced electrophiles by an SET mechanism.
Carbonyl electrophiles that are not easily reduced react exclu-
sively by an SE2ret pathway, and unactivated alkyl halides by
a predominantly or exclusively SE2inv pathway, depending on
ring size. Lithiopyrrolidine 1 reacts with unactivated alkyl
halides with about 75% SE2inv and 25% SE2ret, with steric
factors postulated to play a role when SE2inv is less favorable;
lithiopiperidine 2 reacts with unactivated alkyl halides exclu-
sively by an SE2inv mechanism. With electrophiles that are
easily reduced, such as benzophenone and activated halides, SET
appears to be the predominant mechanism in both heterocyclic
systems. In these systems, the steric course of the reaction can
be used as an indicator of the mechanism: complete racem-
ization is observed when SET is the most likely mechanism,
whereas both invertive and retentive polar mechanisms may
occur when SET is less likely.
Reaction of 2-Lithio-N-methylpyrrolidine with Cyclopropyl-
methyl Bromide. (S)-N-Methyl-2-(tributylstannyl)pyrrolidine (0.50 g,
1.34 mmol) was dissolved in 10.5 mL (0.13 M) of anhydrous THF
and cooled to -78 °C. A solution of BuLi (1.25 M in hexanes, 1.1
mL, 1 equiv) was added dropwise to the cooled solution over 1 min.
The reaction was stirred for 20 min, then neat cyclopropylmethyl
bromide (154 µL, 1.2 equiv) was added over 1 min. The reaction
mixture was stirred for 1 h, then quenched at low temperature with 2
mL of 2 M HCl. The reaction mixture was allowed to reach room
temperature and the solvent was removed in vacuo. The biphasic
mixture was extracted with diethyl ether (5 × 2 mL), and the organic
layers were discarded. The aqueous layer was then basified with 4
pellets of KOH and extracted with ether (5 × 2 mL). The combined
organic layers were dried over anhydrous potassium carbonate, filtered,
and concentrated to a light yellow oil. Crude yield 0.10 g, 54%. GC-
MS analysis indicated 4 products: two dimers (12% and 11%),
N-methyl-2-(cyclopropylmethyl)pyrrolidine (27%), and N-methyl-2-(3-
butenyl)pyrrolidine (51%). The identities of the last two products were
confirmed by comparison to authentic samples.
N-Methyl-2-(3-butenyl)pyrrolidine. N-Methyl-2-(tributylstannyl)-
pyrrolidine (0.86 g, 2.30 mmol) was dissolved in 12 mL (0.2 M) of
anhydrous THF. The solution was cooled to -78 °C and equilibrated
for 10 min. Then BuLi (1.55 M in hexanes, 1.95 mL, 1.3 equiv) was
added dropwise by syringe into the cooled solution over 3 min to give
a light yellow solution that gradually deepens and lightens over a 20
min induction period. The reaction was quenched after 20 min with
4-bromobutene (316 µL, 1.3 equiv). The reaction was then stirred for
1 h at -78 °C and quenched at this temperature with 2 mL of 2 M
HCl. The reaction was then allowed to reach room temperature and
concentrated in vacuo. The biphasic mixture was transferred into a
disposable test tube and extracted with diethyl ether (5 × 2 mL). The
aqueous layer was placed in an ice bath and basified with 4 KOH
pellets. This process gave a biphasic mixture and was extracted with
ether (5 × 2 mL). The combined organic layers were dried over
potassium carbonate, filtered, and concentrated in vacuo without heating
and gave a light yellow oil (0.11 g, 33%). GCMS analysis of the crude
product indicated 3 products: the N-methyl-2-(3-butenyl)pyrrolidine
(61%) and 2 dimers (20% and 17%). The crude material was
chromatographed through silica and gradient eluted with 1 to 5%
methanol in dichloromethane and gave 15 mg of N-methyl-2-(3-
(45) Schleyer, P. v. R.; Clark, T.; Kos, A. J.; Spitznagel, G. W.; Rohde,
C.; Arad, D.; Houk, K. N.; Rondan, N. G. J. Am. Chem. Soc. 1984, 106,
6467.

3350 J. Am. Chem. Soc., Vol. 122, No. 14, 2000
Gawley et al.
1
butenyl)pyrrolidine. H NMR (400 MHz, CDCl3): δ 5.86-5.76 (m,
4.7 (CH2, cyclopropyl), 3.8 (CH2, cyclopropyl). Anal. Calcd for C13H23-
1H) (RCHdCH2), 5.03-4.91 (m, 2H) (RCHdCH2), 3.06 (dt, 1H, J )
2 Hz, 7.5 Hz) (CH2(R)CHN), 2.30 (s, 3H), 2.1-1.88 (m, 5H), 1.81-
1.61 (m, 3H), 1.48-1.38 (m, 2H). 13C NMR (100 MHz, CDCl3): δ
138.7 (RCHdCH2), 114.4 (RCHdCH2), 66.0, 57.2, 40.3 (NCH3), 32.8,
30.9, 30.6, 21.8.
NO2: C, 69.29; H, 10.29. Found: C, 68.54; H, 10.21.
N-Methyl-2-(cyclopropylmethyl)pyrrolidine. Method A: N-Boc-
2-(cyclopropylmethyl)pyrrolidine (0.249 g, 1.1 mmol) was dissolved
in 8 mL of anhydrous ether. The solution was cooled to -78 °C and
allowed to equilibrate for 10 min. DIBAL-H (4.4 mL, 4.0 equiv, 1.0
M in cyclohexane) was added dropwise to the ether solution. The
cooling bath was removed and the reaction mixture was allowed to
reach room temperature. The reaction was stirred for 75 h then cooled
to -78 °C and quenched with saturated sodium potassium tartrate (5
mL). The reaction mixture was stirred overnight to give two clear layers.
The two layers were separated and the aqueous layer was extracted 3
times with 10 mL of ether. The combined organic layer was dried over
anhydrous MgSO4, filtered, and concentrated in vacuo with no heating
to give a light yellow oil. The oil was chromatographed through silica
using a mixture of ethyl acetate, hexanes, and ethanol (1:5:0.5) to give
a colorless liquid. Yield 40 mg, 26%.
Independent Synthesis of N-Methyl-2-(cyclopropylmethyl)pyr-
rolidine. N-Boc-2-allylpyrrolidine. N-Boc-pyrrolidine (2.53 g, 14.8
mmol) and 5.8 mL (2.6 equiv) of TMEDA were dissolved into 30 mL
of anhydrous ether. The solution was cooled to -78 °C and allowed
to equilibrate for 30 min. Then s-BuLi (1.13 M in cyclohexane, 17.0
mL, 1.13 equiv) was added dropwise over a period of 7 min. The
reaction was stirred for 1 h at -78 °C. The reaction mixture was then
quenched with allyl bromide (1.28 mL, 1.5 equiv) by dropwise addition
via syringe. The dry ice acetone bath was removed and the reaction
was allowed to reach room temperature. Stirring was continued for 10
h at room temperature. The reaction mixture was quenched with 10
mL of water and extracted with 3 × 30 mL of ether. The organic layers
were combined and dried over anhydrous MgSO4, filtered, and
concentrated to give a light yellow liquid. The crude material was
chromatographed through silica with 3% EtOAc in hexanes to give a
Method B: N-Boc-2-(cyclopropylmethyl)pyrrolidine (1.03 g, 4.5
mmol) was dissolved in 5 mL of anhydrous THF. The solution was
cooled to -78 °C and allowed to equilibrate for 10 min. DIBAL-H
(18.3 mL, 4.0 equiv, 1.0 M in cyclohexane) was added dropwise to
the THF solution. The cooling bath was removed and the reaction
mixture was allowed to reach room temperature. The reaction was
refluxed for 10 h then cooled to -78 °C and quenched with saturated
sodium potassium tartrate (15 mL). The reaction mixture was stirred
overnight to give two clear layers. The two layers were separated and
the aqueous layer was extracted 3 times with 10 mL of ether. The
combined organic layers were dried over anhydrous MgSO4, filtered,
and concentrated in vacuo with no heating to give a light yellow oil.
The oil was slowly vacuum distilled at room temperature. The distillate
was collected in a tube cooled to -78 °C (dry ice, 2-propanol). The
product is a clear colorless liquid. Yield 0.32 g, 50.2%. 1H NMR (400,
MHz, CDCl3): δ 3.00 (dt, 1H, J ) 2 Hz, 7 Hz), 2.24 (s, 3H), 2 (m,
3H) (RCHN(Me)CH2ax, CH2CH2eq), 1.6 (m, 2H) (CH2CH2CH2N), 1.50
(m, 1H) (CH2axCH2CH2N), 1.34 (m, 1H) (cyclopropyl-CH2), 1.20 (m,
1H) (cyclopropyl-CH2), 0.60 (m, 1H) (CH, cyclopropyl), 0.38 (d, 2H,
J ) 8 Hz) (CH2, cyclopropyl), 0 (m, 2H) (CH2, cyclopropyl). 13C NMR
(100, MHz, CDCl3): δ 66.7, 57,3, 40.5 (NCH3), 39.0, 31.0, 21.8, 8.4
(CH, cyclopropyl), 4.8 (CH2, cyclopropyl), 4.3 (CH2, cyclopropyl).
Anal. Calcd for C9H17N: C, 77.63; H, 12.31. Found: C, 75.97; H,
12.18.
1
colorless oil. Yield 1.521 g, 49%. H NMR (400, MHz, CDCl3): δ
5.71 (m, 1H), (RCHdCH2), 5.00 (m, 2H), (RCHdCH2), 3.78 (br. d,
1H), (CH2(allyl)CHN), 3.26 (br. d, 2H), 2.45 (br. d, 1H), 2.09 (m, 1H),
(CH2dCHCH), 1.8 (m, 4H), 1.44 (s, 9H), (CH3). 13C NMR (100, MHz,
CDCl3): δ 154.5, (CdO),135.3, (RCHdCH2), 116.9, (RCHdCH2),
79.0, (OC(CH3)3), 56.7, 46.7-46.3, 39.0-38.2, 30.1-29.2, 23.6-22.9.
Anal. Calcd for C12H21NO2: C, 68.21; H, 10.02. Found: C, 67.94; H,
10.12.
N-Boc-2-(cyclopropylmethyl)pyrrolidine. Following the procedure
of Suda,46 an Aldrich diazomethane generator was charged with a
solution of KOH (0.744 g, 2 equiv, in 80% EtOH). An ether solution
(35 mL) of Diazald (2.84 g, 2 equiv) was added dropwise while being
heated in a 65-70 °C water bath. Diazomethane was condensed with
a dry ice-acetone coldfinger. The receiver flask was also cooled in a
dry ice-acetone bath. After complete addition of the Diazald solution,
10 mL of ether was added dropwise to the generator. The distillate
was collected until it was colorless. Neat N-Boc-2-allylpyrrolidine (1.33
g, 6.63 mmol) was added to the (-78 °C) diazomethane solution by
syringe. The syringe was rinsed twice with 5 mL of ether. The reaction
mixture was then warmed to 0 °C in an ice water bath. Pd(OAc)2 (31.0
mg, 2 mol %) was added to the solution in one portion. The reaction
was stirred for 30 min to give a yellow solution with a brown precipitate.
The reaction mixture was filtered through a 3 cm pad of Celite, then
concentrated in vacuo to give a yellow oil. The above process was
repeated twice with the yellow oil. NMR spectral analysis of the crude
oil indicated the absence of alkene protons. The crude oil was column
chromatographed with 3% EtOAc in hexanes to give a colorless oil.
Acknowledgment. We gratefully acknowledge support of
this work by the NIH (GM 56271). Acknowledgment is also
made to the donors of the Petroleum Research Fund, adminis-
tered by the American Chemical Society, for support of this
work (32984-AC1). NMR and MS instrumentation was provided
by grants from the NSF and NIH shared instrumentation
programs. We are grateful to Professor Dennis Curran for a
helpful discussion on radical probes and to summer undergradu-
ate research fellows Ruth Figueora and Dana Cairo for technical
assistance.
1
Yield 1.37 g, 97%. H NMR (400, MHz, CDCl3): δ 3.81 (br q, 1H,
J ) 5.1 Hz) (CH2(R)CHN), 3.30 (m, 2H) (CH2CH2N), 1.97-1.73 (m,
4H), 1.58-1.36 (m, 1H) (CH, cyclopropyl), 0.60-0.55 (m, 1H) (CH2,
cyclopropyl), 0.44-0.35 (m, 2H) (CH2, cyclopropyl). 13C NMR (100,
MHz, CDCl3): δ 154.5 (CdO), 78.8 (OC(CH3)3), 57.5 (CH3), 46.5-
46.0, 39.3-38.3, 30.5-29.8, 28.5, 23.8-23.0, 7.8 (CH, cyclopropyl),
(46) Suda, M. Synthesis 1981, 714.
JA9923235