A mechanistic and structural investigation of the (-)-sparteine mediated asymmetric benzylic lithiation substitution reactions of N-Boc-N-(p-methoxyphenyl)benzylamine

Article abstract of DOI:10.1021/ja972225o

Mechanistic and structural studies show the high enantioenrichments in the products from lithiation substitutions of N-Boc-N-(p-methoxyphenyl)benzylamine (1) by n-BuLi/(-)-sparteine (6) arise from an enantioselective deprotonation of 1 to provide configurationally stable (R)-2/6. NMR spectroscopy establishes that ¹³C, ⁶Li labeled (R)-2/6 and (S)-2/6 are monomeric with lithium complexed to the benzylic position, the carbonyl of the Boc group and (-)-sparteine. Deprotonations of the tertiary protons in (R)- and (S)-N-Boc-N-(p-methoxyphenyl)-α-methylbenzylamine ((R)-8 and (S)-8) with n-BuLi/TMEDA provide (R)-9/TMEDA and (S)-9/TMEDA, respectively, with high enantioenrichments. Absolute configurations assigned to (R)-2 and (R)-1-d₁ allow analysis of the electrophile dependent stereochemistry of the reactions of these configurationally stable organolithium intermediates.

Full text of DOI:10.1021/ja972225o

VOLUME 119, NUMBER 48
DECEMBER 3, 1997
© Copyright 1997 by the
American Chemical Society
A Mechanistic and Structural Investigation of the (-)-Sparteine
Mediated Asymmetric Benzylic Lithiation Substitution
Reactions of N-Boc-N-(p-methoxyphenyl)benzylamine
Neil C. Faibish, Yong Sun Park, Steven Lee, and Peter Beak*
Contribution from the Department of Chemistry, UniVersity of Illinois at Urbana-Champaign,
Urbana, Illinois 61801
ReceiVed July 7, 1997X
Abstract: Mechanistic and structural studies show the high enantioenrichments in the products from lithiation-
substitutions of N-Boc-N-(p-methoxyphenyl)benzylamine (1) by n-BuLi/(-)-sparteine (6) arise from an enantioselective
6
deprotonation of 1 to provide configurationally stable (R)-2/6. NMR spectroscopy establishes that 13C, Li labeled
(R)-2/6 and (S)-2/6 are monomeric with lithium complexed to the benzylic position, the carbonyl of the Boc group
and (-)-sparteine. Deprotonations of the tertiary protons in (R)- and (S)-N-Boc-N-(p-methoxyphenyl)-R-
methylbenzylamine ((R)-8 and (S)-8) with n-BuLi/TMEDA provide (R)-9/TMEDA and (S)-9/TMEDA, respectively,
with high enantioenrichments. Absolute configurations assigned to (R)-2 and (R)-1-d1 allow analysis of the electrophile
dependent stereochemistry of the reactions of these configurationally stable organolithium intermediates.
Introduction
mono- and disubstituted products 3 and 5.2a In this sequence 1
is deprotonated with n-BuLi/(-)-sparteine (6) to generate an
organolithium complex 2, which reacts with alkylating and
carbonyl electrophiles to provide the new carbon-carbon bonds
of 3 with high enantioenrichments. The tertiary carbanion 4,
which can be generated subsequently from 3 with the achiral
base n-BuLi/TMEDA at -78 °C, reacts with electrophiles
stereoselectively to produce 5 also with high enantioenrichment.
We now report investigations of the pathways and structures
involved in these asymmetric lithiation-substitution sequences.
The formation of enantioenriched products 3 can arise from
either of two limiting reaction pathways under the influence of
a chiral diamine ligand L*.8 One pathway is asymmetric
deprotonation, in which one of the prochiral benzylic hydrogens
(4) Klein, S.; Marek, I.; Poisson, J.-F.; Normant, J.-F. J. Am. Chem. Soc.
1995, 117, 8853.
(5) Muci, A. R.; Campos, K. R.; Evans, D. A. J. Am. Chem. Soc. 1995,
117, 9075.
(6) (a) Voyer, N.; Roby, J. Tetrahedron Lett. 1995, 36, 6627. (b)
Schlosser, M.; Limat, D. J. Am. Chem. Soc. 1995, 117, 12342.
(7) Park, Y. S.; Beak, P. J. Org. Chem. 1997, 62, 1574.
(8) Beak, P.; Basu, A.; Gallagher, D. J.; Park, Y. S.; Thayumanavan, S.
Acc. Chem. Res. 1996, 29, 552 and references cited therein.
Methods for asymmetric carbon-carbon bond formation are
of considerable current interest. Understanding the pathways
of these reactions should provide a basis for rational improve-
ment of the methodology. Asymmetric lithiation-substitution
sequences with (-)-sparteine as the chiral ligand have been
shown to provide products with high enantiomeric ratios (ers).1-8
We have reported enantioselective sequences beginning with
N-Boc-N-(p-methoxyphenyl)benzylamine (1) which afford the
X Abstract published in AdVance ACS Abstracts, November 1, 1997.
(1) (a) Paulsen, H.; Graeve, C.; Hoppe, D. Synthesis 1996, 141. (b)
Hoppe, D.; Marsch, M.; Harms, K.; Boche, G.; Hoppe, I. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 2158. (c) 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.
(2) (a) Park, Y. S.; Boys, M. L.; Beak, P. J. Am. Chem. Soc. 1996, 118,
3757 (b) Gross, K. M. B.; Jun, Y. M.; Beak, P. J. Org. Chem. 1997, in
press. (c) Weisenburger, G. A.; Beak, P. J. Am. Chem. Soc. 1996, 118,
12218. (d) Basu, A.; Beak, P. J. Am. Chem. Soc. 1996, 118, 1575. (e) Wu,
S.; Lee, S.; Beak, P. J. Am. Chem. Soc. 1996, 118, 715. (f) Thayumanavan,
S.; Basu, A.; Beak, P. J. Am. Chem. Soc. 1997, 119, 8209.
(3) Tsukazaki, M.; Tinkl, M.; Roglans, A.; Chapell, B. J.; Taylor, N. J.;
Snieckus, V. J. Am. Chem. Soc. 1996, 118, 685.
S0002-7863(97)02225-7 CCC: $14.00 © 1997 American Chemical Society

11562 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
Faibish et al.
Scheme 1
is preferentially abstracted by the chiral base RLi/L* from the
substrate 1 to provide a configurationally stable species 2. The
organolithium intermediate 2 then undergoes stereoselective
electrophilic substitution to provide the enantioenriched product
3. The second pathway, which we refer to as asymmetric
substitution, involves the formation of a racemic organolithium
species rac-2 either initially or by racemization of 2. Com-
plexation of rac-2 with L* and reaction of the complex with an
electrophile provides the enantioenriched product 3. In the
asymmetric deprotonation pathway the stereoinformation which
is transferred to the products is introduced in the initial step,
whereas in asymmetric substitutions the asymmetric induction
leading to the products occurs in a post-deprotonation step.8
Scheme 2
as shown in Table 2. If the same organolithium intermediate 2
is obtained from lithiodestannylation as from deprotonation, this
result demonstrates that the racemic organolithium intermediate,
rac-2, is not transformed into enantioenriched products in the
presence of (-)-sparteine under the reaction conditions.
Results and Discussion
Reaction Pathways. Treatment of 1 with 1.2 equiv of
n-BuLi/6 in toluene at -78 °C provided after alkylation with
methyl triflate (S)-8 in 87% yield with a 97:3 er. The absolute
configuration of (S)-8 is assigned after oxidative cleavage of
the p-methoxyphenyl group with ceric ammonium nitrate
(CAN)9 by comparison of CSP-HPLC retention times with that
of the authentic compound. The tin derivative (S)-7 was
prepared in 97% yield with a 95:5 er using the same methodol-
ogy as described for (S)-8.2a The absolute configuration
assigned to (S)-7 was determined by X-ray crystallography.10
Reaction of the enantioenriched tin compound (S)-7 with
n-BuLi/6 provides (R)-8 in 81% yield and with a 95:5 er.2a Thus,
a convenient route to either enantiomer of 8 is available via
either the deprotonation-methylation of 1 or via the transmeta-
lation-methylation of (S)-7. These results are consistent with
an asymmetric deprotonation pathway as shown in Scheme 1.
The assignments of absolute configuration of (R)-2/6 and (S)-
2/6 are also obtained from these transformations (Vide infra).
To further distinguish the two limiting pathways we generated
the racemic intermediate rac-2 via tin-lithium exchange under
conditions which exposed it to the chiral ligand. Transmeta-
lation of racemic 7 with n-BuLi/6 in toluene at -78 °C followed
by reaction with methyl triflate afforded racemic 8 in 83% yield
The two pathways can be distinguished also by submitting
the racemic analogue rac-1-d1 to the standard lithiation-
substitution protocol and comparing the results to the reaction
of 1.2e,6b,11,12 The synthesis of rac-1-d1 was accomplished by
treating 1 with n-BuLi/TMEDA at -78 °C. After 2 h, the
reaction mixture was quenched with MeOD to furnish rac-1-d1
with 99% d1 according to FIMS, which was also confirmed by
1H NMR. The pertinent comparisons are summarized in Table
1.
For asymmetric deprotonation, (S)-1-d1 is termed the matched
enantiomer as it has the proton in the favored position for
abstraction with n-BuLi/6. The enantiomer (S)-1-d1 should react
with n-BuLi/6 and methyl triflate essentially as does 1 to give
(S)-8-d1 as shown in Scheme 2. However, for the mismatched
enantiomer, (R)-1-d1, the deuterium is in the favored position
for abstraction with n-BuLi/6, and its removal would be inhibited
(11) (a) Hoppe, D.; Paetow, M.; Hintze, F. Angew. Chem., Int. Ed. Engl.
1993, 32, 394. (b) Rein, K.; Pappas, M. G.; Anklekar, T. V.; Hart, G. C.;
Smith, G. A.; Gawley, R. A. J. Am. Chem. Soc. 1989, 111, 2211. (c) Eliel,
E. L.; Hartmann, A. A.; Abatjoglou, A. G. J. Am. Chem. Soc. 1974, 96,
1807. (d) Curtin, D. Y.; Kellom, D. B. J. Am. Chem. Soc. 1953, 75, 6011.
(12) Gallagher, D. J.; Du, H.; Long, S. A.; Beak, P. J. Am. Chem. Soc.
1996, 118, 11391.
(9) Kronenthal, D. R.; Han, C. Y.; Taylor, M. K. J. Org. Chem. 1982,
47, 2765.
(10) We have previously deposited the X-ray data for (S)-7.7

Substitution Reactions of N-Boc-N-(p-methoxyphenyl)benzylamine
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11563
Table 2. Transmetalation of 7 to Produce 8
Table 1. Reactions of 1 and rac-1-d1 with n-BuLi/6 in Toluene To
Provide (S)-8 and (S)-8-d1
er of
reaction
entry reactant
ligand
time (h) yielda (%) er of Productb
1
2
3
4
5
racemic (-)-sparteine
10
0.5
10
0.5
10
83
73
81
75
79
racemic
95:5
95:5
70:30
56:44
95:5
95:5
95:5
95:5
(-)-sparteine
(-)-sparteine
TMEDA
TMEDA
a Isolated yield. b The error in er’s is judged to be (5%.
Y
reactant
yield (%)
S:R ratioa
d1a (%)
Y ) H
Y ) D (99% d1)
a The percent error is (5%.
1
100
88
98:2
68:32
rac-1-d1
89
Configurational Stability and Reactivity. The data and the
sequence in Scheme 1 show that the complexes of 2/6 are
configurationally stable under the reaction conditions. The
configurational stability of 2 and 2/6 under different conditions
was investigated by subjecting the enantioenriched stannane
(S)-7 to tin-lithium exchange in the presence of (-)-sparteine
and TMEDA as shown in Table 2. Treatment of (S)-7 with
n-BuLi/6 for 0.5 and 10 h and subsequent reaction with methyl
triflate provided (R)-8 in 73% and 81% yields with 95:5 er,
respectively, as shown in entries 2 and 3. However, reaction
of (S)-7 in the presence of the achiral ligand TMEDA gave (R)-8
in 70:30 and 56:44 er, respectively, after the same transmeta-
lation periods as shown by entries 4 and 5. In these lithiod-
estannylations we assume that tin-lithium exchange is occurring
through a retentive pathway.15,16 These results further demon-
strate that 6 is a more effective ligand in maintaining configura-
tions of these organolithium species than is TMEDA.2f,17-19
The configurationally stablility of 2/6 provides an opportunity
to determine if the diastereomeric complexes react with elec-
trophiles with sufficiently different activation energies to affect
the enantiomeric ratio. In this variant of the Hoffmann test,
reaction with a deficiency and an excess of the electrophile could
reveal that possibility.20
Separate reactions were conducted in which an excess (1.5
equiv) and a deficiency (0.1 equiv) of methyl triflate was added
to (R)-2 after a 2 h lithiation period with n-BuLi/6 to provide
(S)-8 as shown in Table 3. In these experiments the solvent
methyl tert-butyl ether was used in order to provide a more
observable change in er. In the reference reaction with excess
electrophile, the methyl triflate was allowed to react with (R)-2
for 1 h before the reaction was quenched with MeOD.
However, in the reaction with a deficiency of electrophile, the
methyl triflate was allowed to react with (R)-2 for only 2 min
before quenching with MeOD. This protocol was followed
since the er of the initial (S)-8 generated is of interest. In
addition, we were concerned that (S)-8 could undergo further
lithiation, especially for the experiment using a deficiency of
electrophile where there is a large excess of (R)-2 relative to
by a large deuterium isotope effect at -78 °C.2e,6b,11a,13 As
shown in Scheme 2, the mismatched enantiomer may follow
three different paths in lithiation-methylation. If the proton is
abstracted (Route A), the product (S)-8-d1 would have an eroded
er relative to the product from the reaction of 1 and have a net
deuterium incorporation of the reactant rac-1-d1. If the
deuterium is removed (Route B), the product (S)-8-d1 would
have a decreased deuterium content, but the enantioenrichment
would be comparable to that from the reaction of 1. If both
the deuterium kinetic isotope effect and the enantioselectivity
are sufficiently high to prevent reaction (Route C), the product
(S)-8-d1 would be produced in decreased yield relative to that
from 1 but shows similar deuterium content to that of rac-1-d1
and enantioenrichment similar to that from 1. In this case
recovered 1-d1 would be enriched in (R)-1-d1.
Lithiation of 1 with n-BuLi/6 and subsequent reaction with
methyl triflate after 2 h provided (S)-8 in quantitative yield and
98:2 er. Reaction of rac-1-d1 (99% d1) under identical condi-
tions provided (S)-8-d1 (89% d1) in 88% yield and 68:32 er.
The observed decrease in er and deuterium incorporation in the
product indicates that the reaction proceeds through an asym-
metric deprotonation pathway in which Route A dominates but
Route B is available. The deuterium kinetic isotope effect is
more important than the inherent enantioselectivity in this case.
The conclusion that the asymmetric induction occurs through
asymmetric deprotonation was further tested by submitting the
enantioenriched (R)-1-d1 (97% d1) to the lithiation-substitution
protocol. The absolute configuration of (R)-1-d1 is based upon
the assumption that the reaction of (R)-2 with MeOD occurs
with retention (Vide infra).14 Treatment of (R)-1-d1 with
n-BuLi/6 and reaction with methyl triflate afforded (R)-8 in 75%
yield (78% d1) and 60:40 er. A similar result was observed
when carbon dioxide was used as electrophile. These results
further support the asymmetric deprotonation pathway. The
deuterium in (R)-1-d1 is in the favored position for asymmetric
deprotonation, and thus, due to the deuterium kinetic isotope
effect, the formation of (R)-2 is inhibited with the result that
(S)-2 is formed leading to (R)-8-d1 with a 60:40 er. This
rationale is consistent with an invertive reaction pathway for
reaction of 2 with methyl triflate (Vide infra).
(15) (a) Still, W. C.; Sreekumar, C. J. Am. Chem. Soc. 1980, 102, 1201.
(c) Reich, H. J.; Borst, J. P.; Coplien, M. B.; Phillips, N. H. J. Am. Chem.
Soc. 1992, 114, 6577. (b) Aggarawal, V. K. Angew. Chem., Int. Ed. Engl.
1994, 33, 175.
(16) In the only case that has been reported in which tin-lithium takes
place with any inversion, retention still predominates. Clayden, J.; Pink, J.
H. Tetrahedron Lett. 1997, 38, 2565.
(17) The configurational stability of the carbanion in the absence of ligand
could not be determined since the lithiodestannylation did not proceed with
n-BuLi alone.
(18) An alternative possibility is that, in the presence of TMEDA, 2 may
react nonstereoselectively with methyl triflate.
(19) Pearson and Chong have also observed for carbamate and urea
stabilized R-aminoorganolithiums the addition of HMPA and TMEDA,
respectively, significantly diminish configurational stability. Pearson, W.
H.; Lindbeck, A. C.; Kampf, J. W. J. Am. Chem. Soc. 1993, 115, 2622.
Chong, J. M.; Park, S. B. J. Org. Chem. 1992, 57, 2220.
(20) In order to test for this possibility we used a variant of the Hoffmann
test,21 which we have termed the “poor man’s” Hoffmann test since it does
not require a chiral enantioenriched electrophile.22 Since (R)-2 is configu-
rationally stable, a reaction with a deficiency of electrophile could provide
an enantiomeric ratio different from that with an equivalent of electrophile.
(13) Kaiser, B.; Hoppe, D. Angew. Chem., Int. Ed. Engl. 1995, 34, 323.
(14) Hammerschmidt and Hoppe have found also that MeOD and AcOD
react with retention of configuration in reactions of oxygen dipole stabilized
carbanions which are similar to 2. Hammerschmidt, F.; Hanninger, A. Chem.
Ber. 1995, 128, 1069.

11564 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
Faibish et al.
Table 3. Reactions of (R)-2 in t-BuOMe with an Excess or
Deficiency of MeOTf To Yield (S)-8
absent in the subsequent generation of an equimolar mixture of
the diastereomeric complexes (Vide infra).
equiv of MeOTf
yielda (%)
S:R
1.5
93
2.4
92:8
96:4
0.1b
a The yield is relative to 1. b Trapping time was 2 min.
methyl triflate. The 1H NMR and field ionization mass spectra
of (S)-8 for both experiments showed no deuterium incorporation
(i.e. <2%), and, therefore, we concluded that in both cases
further lithiation of the product does not occur. The products
of these reactions show a difference in er of 92:8 and 96:4 for
(S)-8, a difference which is only slightly beyond experimental
error. This result could be taken optimistically to indicate the
major diastereomer (R)-2/6 reacts faster with methyl triflate than
does the minor diastereomer (S)-2/6. However, the results also
show there is not an opportunity for substantial enantioenrich-
ment by using a deficiency of methyl triflate as the electrophile.
NMR Spectroscopic Studies. The work of Fraenkel23 and
Seebach24 provides the basis for the use of isotopic dipolar
couplings in organolithium complexes to assign structure to the
The structure of 2-13C,6Li/6 was also investigated by 13C
NMR spectroscopy. In addition to two peaks at 54.4 and 53.9
ppm which we assign to the two rotamers of 1-13C, the spectrum
contained two strong absorptions at 71.1 and 71.9 ppm in a
90:10 ratio which we assign to the benzylic carbons of the
intermediates (R)- and (S)-2-13C,6Li. Unfortunately, as shown
6
in Figure 1b the coupling to Li is only partially resolved. In
support of the 6Li NMR data, the 13C NMR spectrum indicates
a monomeric organolithium intermediate. In order to insure
the measured integrals represented an accurate ratio of the two
species, the 13C NMR spectrum was acquired with a 100 s delay
between pulses.28 Further evidence of the assignments in the
13C NMR spectrum was provided by the acquisition of a proton
coupled 13C NMR spectrum. The spectrum showed two
doublets for the two benzylic carbons at 71.1 and 71.9 ppm
indicative of two methine groups, 13CH6Li. After completion
of the NMR spectroscopic study, the sample was allowed to
react with methyl triflate to provide (S)-8-13C in 91:9 er.
In order to generate an equimolar mixture of (R)- and (S)-
2-13C,6Li, the racemic trimethyltin labeled compound rac-7-
13C was prepared.2a Transmetalation in toluene-d8 at -78 °C
6
intermediates in this sequence. A Li and 13C NMR spectro-
scopic investigation of the structures of the lithiated intermedi-
ates (R)- and (S)-2 was carried out with the appropriately labeled
reactants 1-13C and n-Bu6Li.25-27 Compound 1-13C was pre-
pared as described previously using labeled R-13C benzyl
bromide.2a
Generation of the organolithium intermediate 2-13C,6Li was
accomplished by reaction of 1-13C with 1.1 equiv of n-Bu6Li/6
in toluene-d8 at -78 °C. As shown in Figure 1a, the 6Li NMR
spectrum contained two doublets at 1.29 and 1.18 ppm in an
89:11 ratio with each showing 1J(6Li, 13C) ) 3.7 Hz. The data
are consistent with the monomeric complexes (R)- 2-13C,6Li/6
6
with 1.5 equiv of n-Bu6Li/6 provided the Li and 13C NMR
6
and (S)-2-13C,6Li/6 in which each Li atom is bonded to only
spectra shown in Figure 1c,d. The 6Li NMR shows two doublets
centered at 1.39 ppm (1J(6Li,13C) ) 3.9 Hz) and 1.28 ppm (1J(6-
Li,13C) ) 3.7 Hz) in a 51:49 ratio. There is also a weak broad
absorption at 0.5 ppm which remains unassigned. The 13C NMR
spectrum contained, in addition to unreacted rac-7-13C, two
benzylic lithiated carbon absorptions at 71.9 and 71.2 ppm in a
one 13C atom. By imposing a 30 s delay between pulses to
account for differences in T1, the ratio of the two diastereomeric
complexes is established to be 91:9 as shown in Table 4.28 Two
additional weak peaks observed at 0.81 and 1.66 ppm were
(21) (a) Hoffmann, R. W.; Lanz, J.; Metternich, R.; Tarara, G.; Hoppe,
D. Angew. Chem., Int. Ed. Engl. 1987, 26, 1145. (b) Hirsch, R.; Hoffmann,
R. W. Chem. Ber. 1992, 125, 975.
(22) Basu, A.; Gallagher, D. J.; Beak, P. J. Org. Chem. 1996, 61, 5718.
(23) Fraenkel, G; Fraenkel, A. M.; Geckle, M. J.; Schloss, F. J. Am.
Chem. Soc. 1979, 101, 4745.
6
50:50 ratio. As evident from Figure 1d, the one-bond Li/13C
coupling is marginally resolved. Consistent with our previous
assignments, the NMR data support the monomeric intermedi-
ates (R)- and (S)-2-13C,6Li with both the Boc group and (-)-
sparteine simultaneously complexing to provide the normally
encountered tetracoordinated lithium. The data indicate that the
same organolithium intermediate is obtained from either the
deprotonation of 1 or the lithiodestannylation of 7.
(24) (a) Seebach, D.; Siegel, H.; Gabriel, J.; Ha¨ssig, R. HelV. Chim. Acta
1980, 63, 2046. (b) Seebach, D.; Ha¨ssig, R.; Gabriel, J. HelV. Chim. Acta
1983, 66, 308.
(25) For some recent examples using 6Li and 13C NMR spectroscopy in
the characterization of organolithium compounds, see: (a) Lichtenfeld, C.
M.; Ahlbrecht, H. Tetrahedron 1996, 52, 10025. (b) Sorger, K. Schleyer,
P.v.R.; Stalke, D. J. Am. Chem. Soc. 1996, 118, 1086. (c) Fraenkel, G.;
Martin, K. V. J. Am. Chem. Soc. 1995, 117, 10336. (d) Bauer, W.;
Griesinger, C. J. Am. Chem. Soc. 1993, 115, 10871.
The structure of 2-13C,6Li was also investigated by carrying
out a similar study in ether-d10.29 As shown in Figure 2a, the
6Li NMR spectrum of (R)- and (S)-2-13C,6Li contained three
peaks centered at 1.89 ppm. The above data indicate two
partially overlapping doublets (∆δ ) 3.9 Hz, 1J(6Li,13C) ) 3.9
Hz) in a 71:29 ratio. A minor unassigned broad peak appears
at approximately 0.8 ppm.
6
7
(26) For some recent examples using Li or Li NMR spectroscopy in
the characterization of lithium amides, see: (a) Lucht, B.L.; Bernstein, M.
P.; Remenar, J. F.; Collum, D. B. J. Am. Chem. Soc. 1996, 118, 10707. (b)
Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1996, 118, 3529. (c) Lucht,
B. L.; Collum, D. B. J. Am. Chem. Soc. 1996, 118, 2217. (d) Hilmersson,
G.; Ahlberg, P.; Davidsson, O¨ J. Am. Chem. Soc. 1996, 118, 3539. (e)
Hilmersson, G.; Davidsson, O¨ J. Org. Chem. 1995, 60, 7660. (f) Reich, H.
J.; Kulicke, K. J. J. Am. Chem. Soc. 1996, 118, 273.
The lithiated intermediates were also studied by 13C NMR
spectroscopy. The spectrum possessed two peaks at 54.6 and
(27) For a review of the advantages of 6Li versus 7Li NMR spectroscopy,
see: Gu¨nther, H. In AdVanced Applications of NMR to Organometallic
Chemistry; Gielen, M., Willem, R., Wrackmeyer, B., Eds.; John Wiley &
Sons: New York, 1996; Chapter 9, pp 247-290.
(29) We also conducted an NMR investigation in THF-d8; however, the
6Li spectrum showed a large singlet and several minor absorptions.
Apparently, the increased ligating ability of THF alters the structure of the
organolithium intermediate. This phenomenon may be the cause of the
reduced enantioselectivity in THF.2a
(28) A similar delay was used in all subsequent spectra.

Substitution Reactions of N-Boc-N-(p-methoxyphenyl)benzylamine
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11565
6
6
Figure 1. (a) 73.6 MHz Li and (b) 125.7 MHz 13C NMR spectra of 1-13C and n-Bu6Li in toluene-d8 (0.15 M) at -78 °C. (c) 73.6 MHz Li and
(d) 125.7 MHz 13C NMR spectra of rac-7-13C and n-Bu6Li in toluene-d8 (0.10 M) at -78 °C.
and toluene-d8 are 16.7 and 17.4 ppm, respectively.32 These
Table 4. Ratio of Diastereomeric Complexes Observed by NMR
and upon Trapping with Methyl Triflate (-78 °C)
values are consistent with a pyramidal benzylic carbon atom
with some degree of delocalization into the benzene ring.33
The NMR data reveal that the organolithium intermediate is
monomeric, and the resulting diastereomeric complexes are
configurationally stable consistent with an asymmetric depro-
tonation mechanism. In toluene-d8, the observed ratio of
diastereomeric complexes agrees well with the enantiomeric
ratio of the products obtained on reaction with methyl triflate.
However, in ether-d10 an increase in er in the products relative
to the ratio of complexes observed by NMR may be a result of
some background asymmetric substitution
Tertiary Benzyllithium Analogues. We have investigated
the lithiation-substitution at the tertiary benzylic carbons of
(R)-8 and (S)-8 with the results shown in Scheme 3 and Table
5.2a The reactant (S)-8 was obtained by subjecting 1 to the
standard lithiation-substitution protocol and carrying out a
(R)-2-13C,6Li:(S)-2-13C,6Li
ether-d10 toluene-d8
71:29 89:11
6Li NMR
13C NMR
73:27
81:19
90:10
91:9
trap with MeOTf
53.9 ppm which are assigned to the two rotamers of 1-13C. The
spectrum also contained two strong absorptions at 71.5 and 70.4
ppm in a 73:27 ratio which we assign to the benzylic carbons
of the intermediates (R)- and (S)-2-13C,6Li. Each peak showed
coupling to Li (1J(6Li,13C) ) 3.9 Hz for both absorptions) as
6
indicated by the 1:1:1 triplet in Figure 2b. The proton coupled
13C NMR spectrum showed two doublets of triplets for the
13
benzylic carbons at 71.5 and 70.4 ppm indicative of two
-
CH6Li absorptions. As shown in Table 4, reaction with methyl
triflate provided (S)-8-13C in 81:19 er. Hence, in toluene-d8
and ether-d10, the 6Li and 13C NMR spectra are fully consistent
with the monomeric structures (R)- 2-13C,6Li/6 and (S)-2-13C,6-
Li/6.
(32) Interestingly, we observed almost an identical chemical shift change
for our previous study on the dilithio-N-methyl amide shown below in THF-
d8.12
The chemical shifts of the benzylic resonance in the 13C NMR
specrtra of 2-13C,6Li relative to the unlithiated starting material
1-13C can be used to estimate the hybridization of the lithiated
benzylic stereogenic center.12,30 Previous studies by Peoples
and Grutzner showed that the formation of the planar (7-
phenylnorbornyl)potassium caused a 33.9 ppm downfield shift,
while the pyramidal species (7-phenylnorbornyl)lithium caused
only a 10.1 ppm downfield shift relative to unmetalated
7-phenylnorbornane.31 The averaged downfield shifts for the
13C signals of (S)-2-13C,6Li/6 and (R)-2-13C,6Li/6 in ether-d10
(33) For a large number of organolithium compounds, an empirical rule
1
has emerged which usually correctly predicts the value of J(6Li,13C) (see
eq 1).34 Thus despite differences in hybridization,35 this rule predicts a value
of approximately 17 Hz for monomers, 8.5 Hz for dimers, and 5.7 Hz for
static tetramers or hexamers. Benzyllithiums represent a specific class of
compounds which possess significantly lower coupling constants than that
predicted by eq 1.25c,36 For example, Fraenkel and Martin found for the
parent monomeric benyllithium with 0.005 M TMEDA in THF-d8, a
coupling constant of 3.8 Hz.25c Fraenkel proposed that benyllithiums possess
increased ionic character and thus, as derived by Karplus, Grant, and
1
1
Litchman37 for J(13C,1H), the magnitude of J(6Li,13C) decreases
(30) Lutz, G. P.; Wallin, A. P.; Kerrick, S. T.; Beak, P. J. Org. Chem.
1991, 56, 4938.
(31) Peoples, P. R.; Grutzner, J. B. J. Am. Chem. Soc. 1980, 102, 4709.
1J(6Li, 13C) ≈ 17/n Hz
(1)
6
where n ) number of equivalently coupled Li or 13C nuclei.

11566 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
Faibish et al.
6
Figure 2. (a) 73.6 MHz Li and (b) 125.7 MHz 13C NMR spectra of 1-13C and n-Bu6Li in ether-d10 (0.15 M) at -78 °C.
Scheme 3
10 was obtained in 61% yield and 95:5 er with retention. The
absolute configuration of 10 is based on comparison of optical
rotation to the N-benzoyl amides of previously assigned
authentic enantiomers.39
Reaction of (S)-8 of 99:1 er with n-BuLi/TMEDA and
subsequent reaction with allyl triflate provided (S)-11 in 96%
yield and 98:2 er with inversion.2a The enantiomer (R)-11 was
prepared in 92% yield and 99:1 er from (R)-8 using the same
reaction sequence. The absolute configuration of 11 is based
on comparison to the Mosher amides of previously assigned
authentic enantiomers.40 With (S)-8 as the substrate, TMEDA
as the ligand, and carbon dioxide as the electrophile, (R)-12
was obtained in 84% yield with 70:30 er, while with (-)-
sparteine as the ligand (R)-12 was obtained in 81% yield and
99:1 er with inversion. The absolute configuration of 12 was
assigned by comparing the CSP-HPLC retention times of the
corresponding methyl ester with that of (S)-10. This result
shows that the electrophilic substitution of the tertiary lithium
intermediate (S)-9/TMEDA with CO2 is highly stereoselective
in the presence of sparteine but not in the presence of TMEDA.
We also observed that, in the presence of TMEDA, allyl bromide
and allyl iodide react with (S)-9/TMEDA nonstereoselectively
to afford (S)-11 in 67:33 and 53:47 er, respectively.2b
An alternative approach which provides both enantiomers at
the quaternary center is to change the order of the substitution
reactions. For example, compound (S)-13 of 99:1 er was
obtained in 89% yield by deprotonation-ethylation of (S)-8 in
toluene. The enantiomer (R)-13 was obtained in 47% yield
and 93:7 er by deprotonation-methylation of (S)-14 in toluene
which was obtained from 1 by a deprotonation-ethylation se-
quence.41
Table 5. Lithiation-Substitution at the Tertiary Benzylic Carbons
of (R)-8 and (S)-8
reactant (er)
E
product (er, yield)
(S)-8 (97:3)
(S)-8 (97:3)
(S)-8 (99:1)
(S)-8 (99:1)
(S)-8 (99:1)a
(R)-8 (99:1)
MeOD
(S)-8-d1 (97:3, 80%)
(S)-10 (95:5, 61%)
(S)-11 (98:2, 96%)
(R)-12 (70:30, 84%)
(R)-12 (99:1, 81%)
(R)-11 (99:1, 92%)
ClCO2CH3
allyl triflate
CO2
CO2
allyl triflate
The capability of generating either enantiomer of 9 was
investigated. An enantioenriched sample (99:1 er) of (S)-8 was
deprotonated with n-BuLi/TMEDA and then allowed to react
with trimethyltin chloride to provide (S)-15. Lithiodestanny-
lation of (S)-15 was then carried out with n-BuLi/TMEDA to
generate (R)-9/TMEDA and subsequently reacting with MeOD.
The transmetalation was allowed to proceed for 8 h and provided
(R)-8-d1 with 99:1 er. Since (R)-8-d1 was obtained with the
same enantioenrichment as the initial substrate, (R)-9/TMEDA
and (S)-9/TMEDA must be configurationally stable under these
a n-BuLi/6 was used.
single recrystallization to afford (S)-8 in 99:1 er. The reactant
(R)-8 was synthesized from (S)-7 via a tin-lithium exchange-
lithiation to provide (R)-8 in 99:1 er after a single recrystalli-
zation.
When (S)-8 of 97:3 er was treated with 1.2 equiv of n-BuLi/
TMEDA in toluene at -78 °C for 8 h followed by addition of
MeOD, (S)-8-d1 (98%-d1) was obtained in 80% yield with an
er of 97:3 with retention of configuration.14,38 With (S)-8 as
the reactant and methyl chloroformate as the electrophile, (S)-
(37) (a) Karplus, M.; Grant, D. M. Proc. Natl. Acad. U.S.A. 1959, 45,
1269. (b) Grant, D. M.; Litchman, W. M. J. Am. Chem. Soc. 1965, 87,
3994. (c) Litchman, W. M.; Grant, D. M. J. Am. Chem. Soc. 1967, 89,
2228.
(38) When CH3COOD was used as the electrophile, (S)-8-d1 (99%-d1)
was obtained in 78% yield with a 92:8 er.
(39) Obrecht, D.; Bohdal, U.; Broger, C.; Bur, D.; Lehmann, C.; Ruffieux,
R.; Schonholzer, P.; Spiegler, C.; Mu¨ller, K. HelV. Chim. Acta. 1995, 78,
563.
(40) Hua, D. H.; Miao, S. W.; Chen, J.-S.; Iguchi, S. J. Org. Chem.
1991, 56, 4.
(34) Bauer, W.; Schleyer, P. v. R. In AdVances in Carbanion Chemistry;
Snieckus, V., Ed.; JAI Press: Greenwich, CT, 1992; Vol. 1, p 89.
(35) This unusual behavior has recently been concluded to arise from a
simultaneous increase in s-character and decrease in lithium-carbon
covalency. Lambert, C.; Schleyer, P. v. R. Angew. Chem., Int. Ed. Engl.
1994, 33, 1129.
(36) (a) Hoell, D.; Lex, J.; Mu¨llen, K. J. Am. Chem. Soc. 1986, 108,
5983. (b) Ruland, T.; Hoffmann, R. W.; Schade, S.; Boche, G. Chem. Ber.
1995, 128, 551.

Substitution Reactions of N-Boc-N-(p-methoxyphenyl)benzylamine
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11567
Table 6. Kinetic Resolution of (R)-8
n-BuLi/(-)-sparteine
extent of
er (yield) of
er (yield) of
(R)-12
(equiv)
reactiona (%)
(R)-8
0.5
0.8
30
83
75:25 (55%) 87:13 (22%)
81:19 (12%) 57:43 (66%)
a The extent of reaction was determined by proton NMR of the crude
reaction mixture.
reaction conditions. Corresponding tertiary lithiated dipole-
stabilized R-oxygen carbanions have been shown by Hoppe to
be configurationally stable.44
Figure 3. Energy diagram for asymmetric deprotonation.
that the proton would be removed by the organolithium
associated with the carbonyl oxygen in the preequilibrium
complex.8,44 With these assignments, the stannylations of 2 and
9 must proceed with inversion.
We achieved a moderate kinetic resolution of rac-8 with
n-BuLi/6. Lithiation of 8 with 0.5 equiv of n-BuLi/6 and
reaction with CO2 provided (R)-8 and (R)-12 in 55% and 22%
yield and 75:25 and 87:13 er, respectively, as shown in Table
6. With 0.8 equiv of n-BuLi/6 provided (R)-8 and (R)-12 in
12% and 66% yield and with 81:19 and 57:43 ers, respectively.
Reactions of the benzylic lithiated species (R)-2/6, (S)-2/6,
(R)-9/TMEDA, and (S)-9/TMEDA with electrophiles can occur
with retention or inversion.7,43,44 Both stereochemistries have
been observed for reactions of a number of R-heteroatom
organolithiums species with electrophiles.42,43,44 The reaction
pathways are shown in Scheme 4 for 1 and (S)-8. On the basis
of above assignments, electrophilic substitutions of (R)-2/6 and
(S)-9/TMEDA proceed with inversion with methyl triflate,
trimethyltin chloride, and carbon dioxide, while MeOD and
methyl chloroformate proceed with retention.45 The pattern that
reactions of highly reactive and/or non-lithium coordinating
electrophiles proceed with inversion and less reactive and/or
lithium coordinating electrophiles proceed with retention is
consistent with previous suggestions.2f,44
Reaction Pathway. Our experimental work establishes that
the lithiation-substitutions of 1 at the benzylic position proceed
by the pathway of asymmetric deprotonation. The lithiated
intermediate (R)-2/6 is a configurationally stable monomeric
species with the carbonyl oxygen complexed to the lithium. The
same species 2 is generated by deprotonation as by destanny-
lation from the appropriate precursors. The absolute configu-
ration of (S)-7 is established, and lithiation-deuteration se-
quences at the tertiary benzylic carbon of (R)-8 and (S)-8 give
net retention.
From well established precedents which show that tin-lithium
exchange proceeds with retention, the configuration of (R)-2
and (S)-2 can be assigned with confidence.15,16 Hence, the chiral
base n-BuLi/6 preferentially removes the pro-R proton from 1
to give (R)-2/6. The assignment of retention of configuration
in the lithiation of 1 and 8 is consistent with the expectation
The energetics of the lithiation-methylation can be analyzed
as shown in Figure 3 for the conversion of 1 to 3. In the
asymmetric deprotonation the pro-R hydrogen of 1 is prefer-
entially abstracted with a ∆∆Gq ) 1.5 kcal/mol to provide (R)-
2/6 and (S)-2/6 which are configurationally stable in a ratio of
98:2. With methyl triflate as the electrophile, the major
diastereomer (R)-2/6 reacts faster than does the minor diaste-
reomer (S)-2/6 with the difference in free energy between
transition structures estimated as 0.28 kcal/mol at -78 °C.
Summary
The present work provides analysis of an asymmetric
deprotonation-substitution pathway which provides products
with high enantioselectivities in replacements of prochiral
methylene protons with carbon-carbon bonds. Either enanti-
omer of products with one or two new carbon-carbon bonds
can be obtained by proper choice of the electrophile or by use
of lithiation-substitution-lithiation sequences. Application of
this methodology and extensions to other systems are under
further study.
(41) The loss of enantioenrichment and low conversion in toluene was
improved by using ether as the solvent to give (R)-12 in 81% yield and
97:3 er.
(42) The reaction of the secondary lithium intermediate (R)-2 with ClCO2-
Me and CO2 also provided the opposite configuration with ClCO2Me
proceeding through a retentive pathway and CO2 proceeding through an
invertive pathway.
(43) For examples of both invertive and retentive electrophilic trapping
of organolithium intermediates, see: (a) Thayumanavan, S.; Lee, S.; Liu,
C.; Beak, P. J. Am. Chem. Soc. 1994, 116, 9755. (b) Derwig, C.; Hoppe,
D. Synthesis 1996, 149. (c) Gawley, R. E.; Zhang, Q. J. Org. Chem. 1995,
60, 5763.
(45) For the reaction with MeOD, we suggest that complexation with
(44) Carstens, A.; Hoppe, D. Tetrahedron, 1994, 50, 6097.
(R)-2/6 and (S)-9/TMEDA occurs prior to deuterium transfer.

11568 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
Faibish et al.
mixture was allowed to slowly warm to room temperature. Workup
consisted of addition of water, extraction of the aqueous layer with
diethyl ether, extraction of the combined diethyl ether extracts with
saturated NH4Cl solution, drying over anhydrous MgSO4, filtration,
and concentration in Vacuo. The crude mixture purified by chroma-
tography to give the pure product.
Scheme 4
Removal of the p-Methoxyphenyl Group by CAN. The concen-
trated crude product mixture was dissolved in CH3CN-H2O (4:1; ca.
0.05 M), and CAN (ceric ammonium nitrate, 2.5 equiv) was added at
0 °C. After stirring at 0 °C for 0.5 h, the mixture was diluted with
diethyl ether, poured into water, and extracted with diethyl ether. The
extracts were combined, dried over MgSO4, filtered through a pad of
Celite, and concentrated in Vacuo. The crude mixture was further
purified by chromatography to give a pure product.
N-Boc-N-(p-methoxyphenyl)-r-trimethylstannylbenzylamine ((S)-
7). From 600 mg of 1 was obtained 887 mg (97%) of 7 as a white
solid: mp 84-85 °C; [R]22 ) -113.2 (c 0.1, CHCl3); 1H NMR
D
(CDCl3, 300 MHz) δ 7.32-6.77 (m, 9H, Ph + PhOMe), 4.06 (s, 1H,
CHPh), 3.78 (s, 3H, PhOCH3), 1.49 (s, 9H, -C(CH3)3), 0.14 (s, 2J
(
119SnCH) ) 26.4 Hz, 9H, Sn(CH3)3); 13C NMR (CDCl3, 75 MHz) δ
157.3, 156.6, 144.2, 137.9, 128.4, 128.2, 127.0, 124.3, 113.5, 80.0,
59.1, 55.2, 28.3, -6.8; EIMS (70 eV) m/z (rel intensity) 462 (5), 421
(5), 406 (30), 376 (54), 346 (4), 330 (7), 256 (24), 212 (100), 196
(68), 165 (100), 57 (99). Anal. Calcd for C22H31NO3Sn: C, 55.49;
H, 6.56; N, 2.94. Found: C, 55.48; H, 6.81; N, 2.85.
The enantiomeric ratio of 7 was determined to be 95:5 on a Chiralcel
OD column (The major enantiomer had a retention time of 8.7 min,
and the minor enantiomer had a retention time of 9.5 min.).
N-Boc-N-(p-methoxyphenyl)-r-methylbenzylamine ((S)-8). From
619 mg of 1 was obtained 562 mg (87%) of 8 as a white solid: mp
95-96 °C; H NMR (CDCl3, 300 MHz) δ 7.31-6.72 (m, 9H, Ph +
Experimental Section
1
Preparation of NMR Samples. To a solution of (-)-sparteine (≈
0.16 mmol) in 0.6 mL of toluene-d8 or ether-d10 (previously distilled
over CaH2) at -78 °C was added n-Bu6Li. The reaction mixture was
stirred for 30 min and transferred via cannula to a solution of 1-13C or
rac-7-13C (≈ 0.14 mmol) in 0.3 mL of toluene-d8 or ether-d10. After
5 min, the resulting solution was transferred via cannula to a septum-
capped 5 mm NMR tube (dried in oven at 120 °C overnight) immersed
in a -78 °C bath.
PhOMe), 5.70 (brs, 1H,CHPh), 3.76 (s, 3H, PhOCH3), 1.44 (d, J )
7.2 Hz, 3H, CHCH3), 1.38 (brs, 9H, -C(CH3)3); 13C NMR (CDCl3, 75
MHz) δ 158.0, 155.4, 142.2, 131.8, 130.8, 128.0, 127.5, 127.0, 113.3,
79.8, 55.2, 54.4, 28.3, 17.9; EIMS (70 eV) m/z (rel intensity) 327 (12,
M+), 271 (42), 227(10), 212 (23), 167 (100), 123 (170, 105 (37), 57
(29). Anal. Calcd for C20H25NO3: C, 73.37; H, 7.70; N, 4.28.
Found: C, 73.20; H, 7.68; N, 4.21.
After removal of the p-methoxyphenyl group, the enantiomeric ratio
of (S)-8 was determined to be 97:3 in favor of the S enantiomer by
chiral HPLC analysis of the corresponding 3,5-dinitrobenzoyl derivative
on a (S)-N-naphthylleucine column. The absolute configuration was
assigned by comparison of CSP-HPLC retention time with authentic
material prepared from commercially available (S)-R-methyl-benzyl-
amine (the S-enantiomer (major) had a retention time of 8.0 min, and
the R-enantiomer (minor) had a retention time of 5.7 min). After one
recrystallization from n-hexane, (S)-8 of 99:1 er was obtained in 81%
yield.
N-Boc-N-(p-methoxyphenyl)-r-ethyl-benzylamine ((S)-14). From
239 mg of 1 was obtained 239 mg (92%) of 14 as a colorless oil: 1H
NMR (CDCl3, 300 MHz) δ 7.62-6.25 (m, 9H, Ph + PhOMe), 5.45
(brs, 1H,CHPh), 3.74 (s, 3H, PhOCH3), 1.83 (m, 2H, CHCH2CH3),
1.39 (brs, 9H, -C(CH3)3), 1.01 (t, J ) 7.0 Hz, 3H, CHCH3); 13C NMR
(CDCl3, 75 MHz) δ 158.0, 155.6, 140.8, 131.6, 130.8, 128.3, 127.9,
127.1, 113.2, 79.7, 61.2, 55.1, 28.3, 24.4, 11.3; EIMS (70 eV) m/z (rel
intensity) 341 (11, M+), 285 (22), 240 (6), 212 (43), 167 (100), 134
(7), 123 (8), 119 (14), 91 (42), 57 (48). HRMS calcd for C21H27NO3
(M+ + 1) 341.1983, found 341.1991.
After removal of the p-methoxyphenyl group, the enantiomeric ratio
of 14 was determined to be 97:3 er in favor of the S enantiomer by
chiral HPLC analysis of the corresponding 3,5-dinitrobenzoyl derivative
on a (S)-N-naphthylleucine column. The absolute configuration was
assigned by conversion to a commercially available amino acid (The
S-enantiomer (major) had a retention time of 7.3 min, and the
R-enantiomer (minor) had a retention time of 5.3 min.).
Acquisition of NMR Spectra. The 6Li and 13C NMR spectra were
recorded at -78 °C on a Varian UI500WB spectrometer operating at
73.6 MHz for 6Li (delay ) 30 s) and 125.7 MHz for 13C (delay ) 100
6
6
s). The Li NMR spectra were referenced to a solution of LiBr in
THF-d8 and the 13C spectra to the residual solvent peaks in toluene-d8
(137.5 ppm) or ether-d10 (65.3 ppm). Due to the relatively small
difference in resonance between 6Li and 2H, the 6Li NMR spectra were
run unlocked. This protocol improved spectral quality.
Preparation of 6Li-Butyllithium. 6Li metal (rod) 2.90 g (482
mmol) in a drybox was hammered flat, cut into small pieces, and loaded
into a Schlenk flask. The flask was removed from the drybox, and 80
mL of pentane and 25.2 mL (241 mmol) of 1-chlorobutane were then
added. The reaction was sonicated overnight at ≈ 40 °C and filtered
through Celite using stardard Schlenk apparatus to yield a yellow
solution of n-Bu6Li.
Enantiomeric Purity Analyses. Analytical HPLC was performed
for both racemic and enantioenriched compounds using a Rainin HPXL
Solvent Delivery System attached to a Rheodyne pump, a Rainin
Dynamax absorbance detector (254 nm), and a computer equipped with
Dynamax MacIntegrator. Either a Chiralcel OD column (0.5 mL/min
flow rate, 5% MTBE in hexane), a (S)-N-naphthylleucine column (2.0
mL/min flow rate, 20% i-PrOH in hexane), a Whelk-O column (1.0
mL/min flow rate, 0.75% i-PrOH in hexane), or a Chiralpak AD column
(0.8 mL/min flow rate, 1.5% i-PrOH in hexane) was used for the
separation of enantiomers.
General Procedure for the Asymmetric Syntheses of N-Boc-N-
(p-methoxyphenyl)-r-Substituted Benzylamines: Syntheses of (S)-
7, (S)-8, (R)-1-d1, and (S)-14. To a solution of (-)-sparteine (1.2 equiv)
in toluene (ca. 0.1 M) at -78 °C was added n-BuLi (1.2 equiv). The
reaction mixture was stirred for 30 min at -78 °C, and then a solution
of 1 (1.0 equiv) in toluene (ca. 0.2 M) was transferred to the above
solution at -78 °C. The resulting reaction mixture was stirred at -78
°C for 4 h, and then an electrophile (1.2 equiv) in toluene (ca. 0.2 M)
was added after precooling. After stirring for 3 h at -78 °C, this
General Procedure for the Tin-Lithium Exchange Reaction with
n-BuLi/Ligand: Preparation of (R)-8 from (S)-7 and Preparation
of (R)-8-d1 from 15. To a solution of 1.2 equiv of (-)-sparteine (or
TMEDA) in toluene (ca. 0.1 M) at -78 °C was added n-BuLi (1.2
equiv). The reaction mixture was stirred for 30 min at -78 °C, and
then a solution of 7 (1.0 equiv) in toluene (ca. 0.2 M) was transferred
to the above solution at -78 °C. The resulting reaction mixture was

Substitution Reactions of N-Boc-N-(p-methoxyphenyl)benzylamine
J. Am. Chem. Soc., Vol. 119, No. 48, 1997 11569
stirred at -78 °C for 4 h, and then methyl triflate (1.2 equiv) in toluene
(ca. 0.2 M) was added after precooling. After stirring for 3 h at -78
°C, this mixture was allowed to slowly warm to room temperature.
Workup consisted of addition of water, extraction of the aqueous layer
with diethyl ether twice, extraction of the combined diethyl ether
extracts with saturated NH4Cl solution, drying over anhydrous MgSO4,
filtration, and concentration in Vacuo. The crude product was further
purified by chromatography to give a pure product.
(S)-10, the amine was treated with benzoyl chloride and Et3N in THF
to provide methyl-(S)-N-benzoyl-2-methyl-2-phenylglycinate: 1H NMR
(CDCl3, 500 MHz) δ 7.84-7.26 (m, 11H, 2Ph + NH), 3.74 (s, 3H,
OMe), 2.18 (s, 3H, CMe). The absolute configuration of (S)-10 was
determined by optical rotation of the amide by comparison to authentic
compound: [R]22 ) +20.1 (c 0.02, CHCl3).
D
N-Boc-N-(p-methoxyphenyl)-r-allyl-r-methyl-benzylamine ((S)-
11) from (S)-8. From 197 mg of (S)-8 (99:1 er) was obtained 11 as
212 mg of a colorless oil in 96% yield: 1H NMR (CDCl3, 300 MHz)
δ 7.51-6.71 (m, 9H, Ph + PhOMe), 5.56 (m, 1H, CHdCH2), 5.03
(m, 2H, CHdCH2), 3.81 (s, 3H, PhOCH3), 2.77 (brt, 1H, CH2-
CHdCH2), 2.51 (dd, J ) 6.0 Hz, 13.1 Hz, 1H, CH2CHdCH2), 1.37
(s, 3H, NCCH3), 1.11 (brs, 9H, -C(CH3)3); 13C NMR (CDCl3, 75 MHz)
δ 158.2, 148.5, 134.4, 133.9, 131.4, 127.9, 125.8, 125.2, 125.1, 118.5,
113.7, 79.9, 63.7, 55.3, 45.3, 28.0, 27.1; EIMS (70 eV) m/z (rel
intensity) 367 (2, M+), 326 (1), 270 (2), 226 (26), 167 (100), 108 (9),
57 (17).
From 84 mg of 7 was obtained 84 mg (81%) of (R)-8 as a white
1
solid: mp 95-96 °C; H NMR (CDCl3, 300 MHz) δ 7.31-6.72 (m,
9H, Ph + PhOMe), 5.70 (brs, 1H, CHPh), 3.76 (s, 3H, PhOCH3), 1.44
(d, J ) 7.2 Hz, 3H, CHCH3), 1.38 (brs, 9H, -C(CH3)3).
After removal of the p-methoxyphenyl group, the enantiomeric ratio
of (R)-8 was determined to be 95:5 in favor of the R enantiomer by
chiral HPLC analysis of the corresponding 3,5-dinitrobenzoyl derivative
on a (S)-N-naphthylleucine column. The absolute configuration was
assigned by comparison of CSP-HPLC retention time with authentic
material prepared from commercially available (S)-R-methyl-benzyl-
amine (The R-enantiomer (major) had a retention time of 5.5 min, and
the S-enantiomer (minor) had a retention time of 7.6 min.). After one
recrystallization from n-hexane, (R)-8 of 99:1 er was obtained in 80%
yield.
The enantiomeric ratio of 11 was determined to be 98:2 using
racemic material as a standard by 1H-NMR. The absolute configuration
was assigned by comparison to the Mosher amides of previously
1
assigned authentic enantiomers (300 MHz H-NMR using chiral shift
reagent (+)-Eu(hfc)3 derivatives in benzene-d6; PhOCH3, (S) δ 3.58,
(R) δ 3.68).
Transmetalation of 15 To Provide (R)-8-d1. To a solution of 15
(152 mg, 0.310 mmol) and TMEDA (0.056 mL, 0.372 mmol) in toluene
(5.3 mL) at -78 °C was added n-BuLi (0.248 mL, 0.372 mmol, 1.5
M). After 1 h, CH3OD (3 mL) was added, and this mixture was allowed
to warm to ambient temperature. This mixture was washed with 4
mL of NH4Cl (aqueous), the two layers were separated, and the aqueous
layer was extracted with Et2O (2 × 5 mL). The combined organic
layers were dried (MgSO4), filtered, and concentrated. The remainder
was filtered through a short plug of silica with CH2Cl2 and concentrated.
Purification by preparative HPLC (1-5% EtOAc in hexane; 10.0 to
12.5 mL/min; 25 cm × 21.4 mm i.d.) yielded (R)-8-d1 as a white solid
(56 mg, 57%). mp ) 101-103 °C. The enantiomeric excess was
determined to be 99:1 er in favor of the R enantiomer by CSP HPLC
(Whelk-0 column; 0.75% i-PrOH in hexane; 1.0 mL/min; retention time
for the R enantiomer: 19 min). The deuterium incorporation was
N-Boc-N-(4-methoxyphenyl)-r-allyl-r-methyl-benzylamine ((R)-
11) from (R)-8. The product (R)-11 of 99:1 er was prepared in 92%
yield from (R)-8 of 99:1 er by the same procedure for the preparation
of (S)-11. After removing p-methoxyphenyl group and Boc group of
(R)-11, the amine was treated with (R)-Mosher acid chloride and Et3N
in THF. The absolute configuration of (R)-11 was determined by 1H-
NMR of the (R,R)-Mosher amide by comparison to the authetic
diastereomers: 1H NMR (CDCl3, 300 MHz) δ 7.55-7.22 (m, 10H,
2Ph), 5.56 (m, 1H, CHdCH2), 5.14 (m, 2H, CHdCH2), 3.45 (s, 3H,
OMe), 2.90-2.60 (m, 2H, CH2CHdCH2), 1.76 (s, 3H, NCCH3).
N-Boc-N-(p-methoxyphenyl)-2-methyl-2-phenylglycine ((R)-12)
from (S)-8. From 128 mg of 8 was obtained 122 mg (84%) of (R)-12
as a colorless oil: 1H NMR (CDCl3, 300 MHz) δ 10.73 (brs, 1H,
CO2H), 7.70-6.81 (m, 9H, Ph + PhOMe), 3.78 (s, 3H, CH3O), 1.54
(s, 3H, CMe), 1.41 (s, 9H, C(CH3)3); 13C NMR (CDCl3, 125 MHz) δ
176.6, 158.7, 157.0, 141.1, 132.3, 130.7, 128.0, 127.4, 126.1, 113.8,
82.1, 68.8, 55.4, 28.2, 28.1; HRMS calcd for C21H25NO5 371.1733,
found 371.1731.
1
determined to be >95% by H NMR.
General Procedure for the Preparation of r,r-Disubstituted
N-Boc Benzylamines. To a solution of 1.2 equiv of TMEDA in toluene
(ca. 0.1 M) at -78 °C was added n-BuLi (1.2 equiv). The reaction
mixture was stirred for 10 min at -78 °C, and then a solution of
R-substituted benzylamine (1.0 equiv) in toluene (ca. 0.2 M) was
transferred to the above solution at -78 °C. The resulting reaction
mixture was stirred at -78 °C for 4 h, and then allyl triflate (1.2 eq)
in toluene (ca. 0.2 M) was added after precooling. After being stirred
for 3 h at -78 °C, this mixture was allowed to slowly warm to room
temperature. Workup consisted of addition of water, extraction of the
aqueous layer with diethyl ether twice, extraction of the combined
diethyl ether extracts with saturated NH4Cl solution, drying over
anhydrous MgSO4, filtration, and concentration in Vacuo. The crude
product was further purified by chromatography to give a pure
disubstituted product.
N-Boc-N-(p-methoxyphenyl)-2-methyl-2-phenylglycine Methyl
Ester ((S)-10) from (S)-8. From 276 mg of 8 (97:3 er) was obtained
197 mg (61%) of 10 as a colorless oil: 1H NMR (CDCl3, 400 MHz)
δ 7.57-6.74 (m, 9H, Ph + PhOMe), 3.82 (s, 3H), 3.76 (s, 3H), 1.57
(s, 3H, CMe)1.39 (s, 9H, C(CH3)3); 13C NMR (CDCl3, 100 MHz) δ
173.8, 158.3, 155.1, 132.6, 130.9, 127.6, 127.0, 114.1, 113.3, 80.6,
67.8, 55.2, 52.4, 28.0, 26.5; HRMS calcd for C22H27NO5 385.1889,
found 385.1888.
In order to prepare the ester for the CSP-HPLC analysis, (R)-12
was refluxed for 2 h in MeOH containing excess thionyl chloride to
give the corresponding Boc deprotected methyl ester: 1H NMR (CDCl3,
300 MHz) δ 7.58-6.39 (m, 9H, Ph + PhOMe), (brs, 1H), 3.69 (s,
3H), 3.68 (s, 3H), 1.89 (s, 3H, CMe). The enantiomeric ratio of (R)-
12 was determined to be 70:30 (99:1 with 6) in favor of the R
enantiomer by chiral HPLC. (Chiralpak-AD column; the R-enantiomer
(major) had a retention time of 19.0 min, and the S enantiomer (minor)
had a retention time of 18.1 min.).
N-Boc-N-(p-methoxyphenyl)-r-ethyl-r-methyl-benzylamine ((S)-
13) from (S)-8. From 244 mg of (S)-8 was obtained 237 mg (89%)
of 13 as a colorless oil: 1H NMR (benzene-d6, 300 MHz) δ 7.50-
6.67 (m, 9H, Ph + PhOMe), 3.24 (s, 3H, PhOCH3), 1.87 (m, 1H,
NCCHaHb), 1.53 (m, 1H, NCCHaHb), 1.51 (s, 3H, CH3), 1.09 (brs, 9H,
-C(CH3)3), 0.59 (t, J ) 7.4 Hz, 3H, CH2CH3); 13C NMR (CDCl3, 100
MHz) δ 158.1, 155.1, 148.1, 134.0, 131.2, 127.7, 125.6, 125.3, 113.7,
79.7, 64.5, 55.3, 35.7, 27.9, 24.6, 9.2; EIMS (70 eV) m/z (rel intensity)
355 (1, M+), 254 (2), 223 (11), 167 (100), 133 (13), 123 (16), 91 (40),
57 (29); HRMS calcd for C22H29NO3 355.2147, found 355.2151.
The enantiomeric ratio of (S)-13 was determined to be 99:1 using
In order to prepare the Boc deprotected methyl ester for the CSP-
HPLC analysis, 10 was treated with 20% trifluoroaceticacid in
methylene chloride at room temperature. 1H NMR (CDCl3, 300 MHz)
δ 7.58-6.39 (m, 9H, Ph + PhOMe), 4.78 (brs, 1H), 3.69 (s, 3H), 3.68
(s, 3H), 1.89 (s, 3H, CMe). The enantiomeric ratio of 10 was
determined to be 95:5 in favor of the S enantiomer by chiral HPLC
using racemic material as a standard (Chiralpak-AD column; 1.5%
isopropyl alcohol in hexane; 0.8 mL/min; the S enantiomer (major)
had a retention time of 18.0 min, and the R enantiomer (minor) had a
retention time of 19.0 min.).
racemic material as a standard by 1H-NMR. The absolute configuration
1
was assigned by analogy to the formation of (R)-11 (300 MHz H-
NMR using chiral shift reagent (+)-Eu(hfc)3 derivatives in benzene-
d6; PhOCH3, (S) δ 3.62, (R) δ 3.67).
N-Boc-N-(p-methoxyphenyl)-r-ethyl-r-methyl-benzylamine ((R)-
13) from (S)-14. From 194 mg of (S)-14 was obtained 164 mg (81%)
of (R)-13 as a colorless oil: 1H NMR (benzene-d6, 300 MHz) δ 7.50-
6.69 (m, 9H, Ph + PhOMe), 3.24 (s, 3H, PhOCH3), 1.87 (m, 1H,
NCCHaHb), 1.55 (m, 1H, NCCHaHb), 1.51 (s, 3H, CH3), 1.09 (brs, 9H,
-C(CH3)3), 0.59 (t, J ) 7.4 Hz, 3H, CH2CH3).
After removal of the p-methoxyphenyl group and the Boc group of

11570 J. Am. Chem. Soc., Vol. 119, No. 48, 1997
Faibish et al.
The enantiomeric ratio of (R)-13 was determined to be 97:3 using
CH3OD was added and this mixture was allowed to warm slowly to
ambient temperature. Workup consisted of addition of water, extraction
of the aqueous layer with diethyl ether, extraction of the combined
diethyl ether extracts with saturated NH4Cl solution, drying over
anhydrous MgSO4, filtration, and concentration in Vacuo. The crude
mixture purified by column chromatography using EtOAc:petroleum
ether (1:10) as eluent to give (S)-8 in 93% yield. The enantiomeric
ratio was determined to be 92:8 by CSP HPLC (Whelk-O column; the
S enantiomer (major) had a retention time of 21 min and the R
enantiomer (minor) had a retention time of 19 min.). The deuterium
incorporation was determined to be 1.1% by FIMS. The reaction with
0.1 equiv of MeOTf was carried out in a similar manner, except the
reaction with 0.1 equiv of MeOTf was quenched with CH3OD after 2
min to provide (S)-8 in 2.4% yield (1.6% d1, 96:4 er).
Preparation of rac-1-d1. To a solution of TMEDA (0.30 mL, 1.98
mmol) in toluene (6.0 mL) at -78 °C was added n-BuLi (1.37 mL,
1.98 mmol). After 30 min, a solution of 1 (516 mg, 1.65 mmol) in
toluene (4.0 mL) at -78 °C was added. After 2 h, MeOD (0.33 mL,
8.23 mmol) was added. The resulting reaction mixture was then
allowed to slowly warm to room temperature. Workup consisted of
addition of water, extraction of the aqueous layer with diethyl ether,
extraction of the combined diethyl ether extracts with saturated NH4Cl
solution, drying over anhydrous MgSO4, filtration, and concentration
in Vacuo. The crude mixture purified by column chromatography using
EtOAc:petroleum ether (1:10) as eluent to give rac-1-d1 in 100% yield.
The deuterium incorporation was determined to be 99% d1 by FIMS.
Reaction of 1 and rac-1-d1 with n-BuLi. To a solution of 6 (0.14
mL, 0.506 mmol) in toluene (6.0 mL) at -78 °C was added n-BuLi
(0.42 mL, 0.607 mmol). After 30 min, a solution of rac-1-d1 (159
mg, 0.506 mmol) in toluene (4 mL) at -78 °C was added. After 2 h,
methyl trifluoromethanesulfonate (0.069 mL, 0.607 mmol) was added.
Workup consisted of addition of water, extraction of the aqueous layer
with diethyl ether, extraction of the combined diethyl ether extracts
with saturated NH4Cl solution, drying over anhydrous MgSO4, filtration,
and concentration in Vacuo. The crude mixture purified by column
chromatography using EtOAc:petroleum ether (1:10) as eluent to give
(S)-8-d1 in 89% yield. The enantiomeric ratio was determined to be
68:32 by CSP HPLC (Whelk-O column; the S enantiomer (major) had
a retention time of 21 min, and the R enantiomer (minor) had a retention
time of 19 min.). The deuterium incorporation was determined to be
89% by FIMS. The reaction with 1 was carried out in a similar manner
to provide (S)-8 in 100% yield.
racemic material as a standard by 1H-NMR. The absolute configuration
1
was assigned by analogy to the formation of (R)-11 (300 MHz H-
NMR using chiral shift reagent (+)-Eu(hfc)3 derivatives in benzene-
d6; PhOCH3, (S) δ 3.39, (R) δ 3.44).
N-Boc-N-(p-methoxyphenyl)-r-methyl-r-trimethylstannyl-ben-
zylamine (15) from (S)-8. To a solution of (S)-8 (145 mg, 0.443 mmol;
99:1 er) and TMEDA (0.080 mL, 0.532 mmol) in toluene (7.5 mL) at
-78 °C was added n-BuLi (0.355 mL, 0.532 mmol). After 8 h,
trimethyltin chloride (0.532 mL, 0.532 mmol, 1.0 M solution in
hexanes) was added. After ca. 3 h at -78 °C, this mixture was allowed
to warm to ambient temperature over ca. 4 h and then left at ambient
temperature for ca. another 7 h. This mixture was washed with 5 mL
of NH4Cl (aqueous), the two layers were separated, and the aqueous
layer was extracted with Et2O (2 × 5 mL). The combined organic
layers were dried (MgSO4), filtered, and concentrated. The remainder
was filtered through a short plug of silica with CH2Cl2 and concentrated.
Purification by preparative HPLC (1% EtOAc in hexane; 10.0 mL/
min; 25 cm × 21.4 mm i.d.) yielded 15 as a clear, colorless oil (163
mg, 75%): 1H-NMR (CDCl3, 500 MHz) δ -0.04 (t, J ) 25.3, 9H,
Me3Sn), 1.36 (s, 9H, C(CH3)3), 1.46 (s, 3H, CH3CN), 3.73 (s, 3H,
OCH3), 6.65-7.29 (m, 9H, Ar). 13C-NMR (CDCl3, 125 MHz) δ -5.4
(t, J ) 176.6 Hz, Me3Sn); 22.7 (CH3-CN); 28.3 (C(CH3)3); 55.2 (CH3O);
56.6 (NCSn); 80.1 (C(CH3)3); 113.0, 124.2, 124.8, 127.98, 128.01,
116
129.7, 147.8, 157.4 (Ar); 157.7 (CdO); HRMS calcd for C23H33NO3
-
Sn (M+) 487.1473, found 487.1477 (0.4 mDa).
Reaction of (R)-1-d1 with n-BuLi/6. To a solution of 6 (0.13 mL,
0.576 mmol) in toluene (6.0 mL) at -78 °C was added n-BuLi (0.40
mL, 0.576 mmol). After 30 min, a solution of (R)-1-d1 (151 mg, 0.480
mmol; 97% d1) in toluene (4.0 mL) at -78 °C was added. After 1.5
h, methyl trifluoromethanesulfonate (0.065 mL, 0.576 mmol) was
added. The resulting reaction mixture was then allowed to slowly warm
to room temperature. Workup consisted of addition of water, extraction
of the aqueous layer with diethyl ether, extraction of the combined
diethyl ether extracts with saturated NH4Cl solution, drying over
anhydrous MgSO4, filtration, and concentration in Vacuo. The crude
mixture purified by column chromatography using EtOAc:petroleum
ether (1:10) as eluent to give (R)-8-d1 in 75% yield. The enantiomeric
ratio was determined to be 60:40 by CSP HPLC (Whelk-O column;
the S enantiomer (major) had a retention time of 21 min, and the R
enantiomer (minor) had a retention time of 19 min.). The deuterium
incorporation was determined to be 78% by FIMS.
Reaction of 1 with 1.5 and 0.1 Equiv of MeOTf in t-BuOMe To
Provide (S)-8. To a solution of 6 (0.13 mL, 0.553 mmol) in t-BuOMe
(6.0 mL) at -78 °C was added n-BuLi (0.38 mL, 0.553 mmol). After
30 min, a solution of 1 (144 mg, 0.460 mmol) in t-BuOMe (4.0 mL)
at -78 °C was added. After 2 h, methyl trifluoromethanesulfonate
(0.078 mL, 0.690 mmol) was added. After 1 h at -78 °C, 0.25 mL of
Acknowledgment. We are grateful to the National Institutes
of Health (GM-18874) and the National Science Foundation
(NSF-95-26355) for the support of this work.
JA972225O