Regioselectivity in the palladium/(S)-BINAP(S)-catalyzed asymmetric allylic amination: Reaction scope, kinetics, and stereodynamics

Article abstract of DOI:10.1021/om050537r

Studies defining the scope of the reaction and understanding the factors controlling the unusual regioselectivity observed in the Pd/{BINAP(S)}-catalyzed allylic amination reaction are presented. Most of the unsymmetrically substituted allylic substrates tested show regioselectivity that is the reverse of that which would normally be expected in Pd-catalyzed systems. Several cationic η³-allylic complexes containing a Pd{BINAP(S)} fragment have been synthesized in an attempt to correlate this behavior with structural features of the ligand and the solution structure of the allylpalladium complex. Notably, allylic amines with an α-quaternary carbon center may be prepared with high regioselectivity with this system. Spin-saturation transfer studies (SST) were performed on two (κ²-P,S)Pd-;r-allyl complexes in order to determine relative rates for Pd-η³- η¹-η³-allyl interconversions and the hemilabile behavior of the BINAP(S) ligand.

Full text of DOI:10.1021/om050537r

5076
Organometallics 2005, 24, 5076-5083
Regioselectivity in the Palladium/
(S)-BINAP(S)-Catalyzed Asymmetric Allylic Amination:
Reaction Scope, Kinetics, and Stereodynamics
J. W. Faller* and Jeremy C. Wilt
Department of Chemistry, Yale University, New Haven, Connecticut 06520
Received June 28, 2005
Studies defining the scope of the reaction and understanding the factors controlling the
unusual regioselectivity observed in the Pd/{BINAP(S)}-catalyzed allylic amination reaction
are presented. Most of the unsymmetrically substituted allylic substrates tested show
regioselectivity that is the reverse of that which would normally be expected in Pd-catalyzed
systems. Several cationic η3-allylic complexes containing a Pd{BINAP(S)} fragment have
been synthesized in an attempt to correlate this behavior with structural features of the
ligand and the solution structure of the allylpalladium complex. Notably, allylic amines with
an R-quaternary carbon center may be prepared with high regioselectivity with this system.
Spin-saturation transfer studies (SST) were performed on two (κ2-P,S)Pd-π-allyl complexes
in order to determine relative rates for Pd-η3-η1-η3-allyl interconversions and the hemilabile
behavior of the BINAP(S) ligand.
Scheme 1. Preparation of
Introduction
[(η3-Allyl)Pd(S-BINAP(S))]SbF6 (2)
The transition-metal-catalyzed allylic substitution
reaction has become a powerful synthetic tool in recent
years, and the amount of literature on the subject
continues to grow.1,2 Nevertheless, this reaction remains
of considerable interest because significant challenges
still exist. Impressive progress has been made with Pd
in terms of increasing the scope and stereoselectivity
with which symmetrically disubstituted allylic sub-
strates undergo this reaction.3-7 Additionally, Ir-,8-13
Rh-,14-16 and Mo17-catalyzed allylic substitutions have
provided very high regioselectivity for nucleophilic
attack at the substituted allylic terminus when the
other terminus is unsubstituted. Considerably less
attention has been devoted to the Pd-catalyzed reaction
of unsymmetrical 1,3-disubstituted allylic com-
pounds.18-27 These types of substrates present the
possibility of producing regioisomeric products depend-
ing on which of the allylic termini undergoes substitu-
tion. During the course of our studies on various Pd-
catalyzed allylic substitution reactions with (S)-BINAP(S)
1 (Scheme 1) as a ligand, we have found several
instances in which the regiochemical outcome of the
reaction was the reverse of that which Pd normally gives
rise to.28-30 We were interested in defining the factors
(1) Trost, B. M.; VanVranken, D. L. Chem. Rev. 1996, 96, 395-422.
(2) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921-2943.
(3) Evans, D. A.; Campos, K. R.; Tedrow, J. S.; Michael, F. E.; Gagne,
M. R. J. Org. Chem. 1999, 64, 2994-2995.
(4) Vonmatt, P.; Loiseleur, O.; Koch, G.; Pfaltz, A.; Lefeber, C.;
Feucht, T.; Helmchen, G. Tetrahedron: Asymmetry 1994, 5, 573-584.
(5) Sennhenn, P.; Gabler, B.; Helmchen, G. Tetrahedron Lett. 1994,
35, 8595-8598.
(18) You, S. L.; Zhu, X. Z.; Luo, Y. M.; Hou, X. L.; Dai, L. X. J. Am.
Chem. Soc. 2001, 123, 7471-7472.
(19) Hayashi, T.; Kishi, K.; Yamamoto, A.; Ito, Y. Tetrahedron Lett.
1990, 31, 1743-1746.
(6) Sprinz, J.; Helmchen, G. Tetrahedron Lett. 1993, 34, 1769-1772.
(7) Evans, D. A.; Campos, K. R.; Tedrow, J. S.; Michael, F. E.; Gagne,
M. R. J. Am. Chem. Soc. 2000, 122, 7905-7920.
(8) Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 15164-
15165.
(20) Watson, I. D. G.; Styler, S. A.; Yudin, A. K. J. Am. Chem. Soc.
2004, 126, 5086-5087.
(21) Blacker, A. J.; Clarke, M. L.; Loft, M. S.; Williams, J. M. J.
Org. Lett. 1999, 1, 1969-1971.
(22) Faller, J. W.; Sarantopoulos, N. Organometallics 2004, 23, 2179.
(23) Boele, M. D. K.; Kamer, P. C. J.; Lutz, M.; Spek, A. L.; de Vries,
J. G.; van Leeuwen, P.; van Strijdonck, G. P. E. Chem.-Eur. J. 2004,
10, 6232-6246.
(9) Takeuchi, R.; Ue, N.; Tanabe, K.; Yamashita, K.; Shiga, N. J.
Am. Chem. Soc. 2001, 123, 9525-9534.
(10) Kiener, C. A.; Shu, C. T.; Incarvito, C.; Hartwig, J. F. J. Am.
Chem. Soc. 2003, 125, 14272-14273.
(11) Janssen, J. P.; Helmchen, G. Tetrahedron Lett. 1997, 38, 8025-
8026.
(24) van Haaren, R. J.; Goubitz, K.; Fraanje, J.; van Strijdonck, G.
P. F.; Oevering, H.; Coussens, B.; Reek, J. N. H.; Kamer, P. C. J.; van
Leeuwen, P. Inorg. Chem. 2001, 40, 3363-3372.
(25) Acemoglu, L.; Williams, J. M. J. Adv. Synth. Catal. 2001, 343,
75-77.
(12) Garcia-Yebra, C.; Janssen, J. P.; Rominger, F.; Helmchen, G.
Organometallics 2004, 23, 5459-5470.
(13) Lipowsky, G.; Miller, N.; Helmchen, G. Angew. Chem., Int. Ed.
2004, 43, 4595-4597.
(26) Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 1999, 121, 4545-
4554.
(14) Evans, P. A.; Robinson, J. E.; Moffett, K. K. Org. Lett. 2001, 3,
3269-3271.
(27) Pretot, R.; Pfaltz, A. Angew. Chem., Int. Ed. 1998, 37, 323-
325.
(15) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2003, 125, 8974-
8975.
(28) Faller, J. W.; Wilt, J. C.; Parr, J. Org. Lett. 2004, 6, 1301-
1304.
(16) Evans, P. A.; Lawler, M. J. J. Am. Chem. Soc. 2004, 126, 8642-
8643.
(29) Faller, J. W.; Wilt, J. C. Tetrahedron Lett. 2004, 45, 7613-7616.
(30) Faller, J. W.; Wilt, J. C. Org. Lett. 2005, 7, 633-636.
(17) Belda, O.; Moberg, C. Acc. Chem. Res. 2004, 37, 159-167.
10.1021/om050537r CCC: $30.25 © 2005 American Chemical Society
Publication on Web 09/07/2005

Pd/(S)-BINAP(S)-Catalyzed Allylic Amination
Organometallics, Vol. 24, No. 21, 2005 5077
Table 1. Amination of Alkyl-Monosubstituted
Allylic Carbonates
Table 2. Amination of Unsymmetrical
Phenyl-Substituted Substrates
Table 3. Amination of 3,3-Alkyl Disubstituted
Substrates
that contribute to these unusual regioselectivities in the
context of Pd-catalyzed allylic amination. Substrate
scope, ligand architecture, and the solid and solution
structure of some of the intermediate Pd-π-allyl com-
plexes were examined. These studies include the mea-
surement of η3-η1-η3 and allylic termini interconver-
sion processes by spin-saturation transfer experiments.
Catalysis Results
As a starting point, we chose to examine the effects
of using different regio- and stereoisomers of butenyl
ethyl carbonates 3-5 (Table 1). Not unexpectedly, sub-
strates 3-5 gave very similar results, which presumably
arises from rapid equilibration of the isomeric interme-
diate Pd-η3-crotyl species with respect to attack of ben-
zylamine. In the great majority of Pd-catalyzed allylic
substitution reactions, nucleophilic attack on the less sub-
stituted, more sterically accessible, allylic terminus is pre-
ferred. What is unusual about these results when using
2 as catalyst is the extremely high regioselectivity in
favor of attack on the more substituted allylic terminus.
Upon increasing the size of R to i-Pr, however, the
substitution pattern now resembles what one expects
from most Pd-catalyzed allylic substitution reactions,
heavily favoring the linear achiral product 8b.
with other 3,3-disubstituted starting materials 15 and
16 gave the branched products 18a and 19a, respec-
tively, in greater than 98:2 regioisomeric ratios. This
methodology could possibly be effective in the creation
of chiral quaternary centers with pendant amine and
terminal alkene functionality.
Discussion
To investigate the origin of the selectivity in these
systems, an evaluation of a number of factors must be
When unsymmetrically substituted allylic substrates
9-11 incorporating Ph are employed in this reaction,
we also find that amination occurs exclusively at the
less substituted terminus (Table 2). These results are
in accord with what is observed when substrate 6 is
used, i.e., a large R group giving rise to the traditional
regioselectivity observed with Pd.
The selectivity we observed for substitution on the
more substituted allylic position with small alkyl sub-
stituents encouraged us to attempt the preparation of
quaternary carbon centers from 3,3-disubstituted allylic
starting materials (Table 3). Previously, we reported
that amination of a 3,3-dimethyl-substituted allyl, 14,
with 2 produced 17a and 17b in a 91:9 ratio. When
using the related [(η3-allyl)Pd{BIPHEP(S)}]SbF6 com-
plex 20, based on BIPHEP(S), 21, as the catalyst (Figure
1), this ratio was further increased to >99:1 in favor of
17a. Extension of this methodology using 20 as catalyst
Figure 1. ORTEP diagram of [(η3-allyl)Pd{BIPHEP(S)}]-
SbF6 (20). Pd1-C1 ) 2.194(9)Å; Pd1-C3 ) 2.244(10) Å;
Pd-S1-P2 ) 100.28(9)°. Major conformation of allyl is
shown.

5078 Organometallics, Vol. 24, No. 21, 2005
Faller and Wilt
Figure 2. Differing rate constants for cis vs trans and RH
vs HH nucleophilic attack.
considered. Assuming that the relevant intermediate is
a cationic [(η3-allyl)Pd(κ2-P,S)] species, we decided to
undertake a thorough investigation of several of these
types of complexes, in the solid state as well as in
solution. Thus, the critical transition states should
involve attack on the face of the allyl opposite of the
metal. The regiochemistry of the amination product
should be controlled by the relative importance of the
electronic and steric effects of the substituent on the
terminus of the allyl versus the relative reactivity of the
allylic termini trans or cis to phosphorus. When the
ligand is a P,X-heterobidentate, there is a preference
for attack trans to the phosphorus for symmetrically
substituted allyls. This preference can be restated as a
ratio of rate constants for attack trans or cis to phos-
phorus as kt:kc > 20:1 (Figure 2).31-37
Figure 3. Regio- and stereochemical model for allylic
aminations.
Scheme 2. Preparation of
[(η3-Crotyl)Pd(S-BINAP(S))]SbF6 (22)
For a homobidentate P,P-ligand there is a preference
for attack on the unsubstituted terminus for monosub-
stituted allyls. That is, the rate constant ratio is kHH
>
substitution occurs trans to the highest trans-effect
donor, then we have a working model for prediction of
the regiochemistry of the products when starting from
unsymmetrically substituted allylic starting materials
(Figure 3). To test this hypothesis, the crotyl complex
22 was synthesized (Scheme 2) in order to determine
the preference for the positioning of the methyl group
with respect to the P,S donor set. Although X-ray
kMeH, which is ∼7:3 for BINAP, but >20:1 for many
others; this can be attributed to a steric effect to the
approach of the nucleophile (Figure 2).31,38,39 One should
note, however, that for the case of two methyl substit-
uents on the same terminus, the steric effect can be
potentially overcome by the modified bonding of that
terminus to the metal owing to adjustment of the
relative nucleophilicity of the termini yielding kMeMe
kHH (see for example refs 38, 40-43).
>
1
quality crystals of 22 could not be obtained, H NMR
spectroscopy reveals that it exists as >90% of the
To summarize briefly, the regioselectivity of these
reactions will be determined by a combination of these
factors: the positioning of the R group relative to the
donor atoms, the relative rate of attack on termini trans
to P and S, and the balance of steric and electronic
effects controlling the rates of attack on the two allylic
termini.
We expect that, owing to the large/small nature of the
P,S donor set, the larger R group should prefer to be
trans to P owing to less severe steric repulsions that
would be present were it cis to P. If we take into account
that the attack of the nucleophile in Pd-catalyzed allylic
diastereomers with the methyl group trans to P, as
1
indicated by the magnitude of H-31P couplings. For-
tunately, crystallization of the 1,1-dimethylallyl complex
23 occurs readily, and the structure is shown in Figure
4. The 1H NMR spectrum in CD2Cl2 solution shows that
the doubly substituted allylic terminus is trans to P,
consistent with the structure observed in the solid state.
Additionally, η3-cinnamyl complex 24 was synthesized,
and in the solid state it exists as a ∼2:1 mixture of
(31) Akermark, B.; Hansson, S.; Krakenberger, B.; Vitagliano, A.;
Zetterberg, K. Organometallics 1984, 3, 679-682.
(32) Tu, T.; Zhou, Y. G.; Hou, X. L.; Dai, L. X.; Dong, X. C.; Yu, Y.
H.; Sun, J. Organometallics 2003, 22, 1255-1265.
(33) Delbecq, F.; Lapouge, C. Organometallics 2000, 19, 2716-2723.
(34) Sprinz, J.; Kiefer, M.; Helmchen, G.; Huttner, G.; Walter, O.;
Zsolnai, L.; Reggelin, M. Tetrahedron Lett. 1994, 35, 1523-1526.
(35) Brown, J. M.; Hulmes, D. I.; Guiry, P. J. Tetrahedron 1994,
50, 4493-4506.
(36) Szabo, K. J. Organometallics 1996, 15, 1128-1133.
(37) Constantine, R. N.; Kim, N.; Bunt, R. C. Org. Lett. 2003, 5,
2279-2282.
(38) Oslob, J. D.; Akermark, B.; Helquist, P.; Norrby, P. O. Orga-
nometallics 1997, 16, 3015-3021.
(39) Tenaglia, A.; Heumann, A. Angew. Chem., Int. Ed. 1999, 38,
2180-2184.
(40) Akermark, B.; Zetterberg, K. J. Am. Chem. Soc. 1984, 106,
5560-5561.
(41) Akermark, B.; Vitagliano, A. Organometallics 1985, 4, 1275-
1283.
(42) Akermark, B.; Zetterberg, K.; Hansson, S.; Krakenberger, B.;
Vitagliano, A. J. Organomet. Chem. 1987, 335, 133-142.
(43) Akermark, B.; Hansson, S.; Krakenberger, B.; Zetterberg, K.;
Vitagliano, A. Chem. Scr. 1987, 27, 525-526.
Figure 4. ORTEP diagram of [(η3-1,1-dimethylallyl)Pd-
(S-BINAP(S))]SbF6 (23). Pd1-C1 ) 2.098(6) Å; Pd1-C3 )
2.268(7) Å; Pd-S1-P2 ) 106.17(8)°.

Pd/(S)-BINAP(S)-Catalyzed Allylic Amination
Organometallics, Vol. 24, No. 21, 2005 5079
Scheme 3. Attack on Enantiotopic π-Allyl Faces
trapped selectively, regardless of its relative concentra-
tion under the reaction conditions.
While the selective trapping of a particular regioiso-
mer gives rise to the observed regiochemistry when
starting from an unsymmetrically substituted subtrate,
the selective addition to one of the enantiotopic allylic
faces gives rise to the observed enantioselectivity (Scheme
3). When Curtin-Hammett conditions are met, the
rapidly interconverting allylic faces provide the means
to transform chiral racemic substrates into enantiopure
products. When these conditions are not met, the
consequence is a “memory effect”. A “memory effect”
may be conveniently defined as an incomplete scram-
bling of regio- and stereochemical information in the
allylic starting material.22,25,44-46 These effects often
result from the rate of nucleophilic attack being faster
than (or approximately equal to) the rates of Pd-π-allyl
interconversions. We have noted “memory effects” in our
previous work with catalyst 2; we therefore desired to
quantify the rates of these Pd-π-allyl interconversions
in order to obtain a better understanding of our observed
regio- and stereochemistry in the allylic amination
reaction.
The viability of our allylic amination reactions operat-
ing under Curtin-Hammett conditions was examined
by performing variable-temperature 1H NMR and spin-
saturation studies on some [(η3-allyl)Pd(P,S)]SbF6 com-
plexes.47 Generally one expects that Curtin-Hammett
conditions will be met owing to rapid η3-η1-η3 equili-
bration and interconverting T-shaped intermediates.48,49
One should note, however, that the barrier to intercon-
version and/or preference for a specific T-shaped inter-
mediate may be increased for heterobidentate ligands
owing to the general difficulty of placing two high trans
effect ligands trans to each other (e.g., η1-allyl and P).
Finally, one must consider the possibility that the
heterobidentate ligand may be hemilabile. An intercon-
version of this kind would proceed through decoordina-
tion of the S atom followed by rearrangement of the
T-shaped intermediate without apparent allyl rotation.
This would allow for the equivalent of a π-allyl pseu-
dorotation, which has the effect of switching the allylic
termini with respect to their position relative to the P,S
donor set (Figure 6). The processes that lead to allylic
termini interconversion are especially important in
understanding the regioselectivity of the Pd-catalyzed
allylic aminations presented here.
Figure 5. ORTEP diagram of [(η3-cinnamylallyl)Pd(S-
BINAP(S))]SbF6 (24). Pd1-C1 ) 2.136(5) Å; Pd1-C3 )
2.308(10) Å; Pd-S1-P2 ) 104.74(6)°. Major conformation
found in crystal is shown.
diastereomers, both of which have the Ph group trans
to P (Figure 5). In CD2Cl2 solution, complex 24 exists
as a 96:4 regioisomeric mixture, with the isomers having
Ph trans to P predominating. These complexes clearly
demonstrate the role of the relative steric effects pre-
sented by the P,S arrangement in determining the
preferred regiochemistry of intermediates of the sub-
stituted complexes if we assume that nucleophilic attack
occurs trans to P.
However, there is clearly more involved in regiocon-
trol than simple steric effects; although 3-5 and 14-
16 are preferentially attacked at the more substituted
position, i-Pr-substituted 6 and phenyl-containing sub-
strates 9-11 are attacked at the less substituted
position exclusively. This reversal of regioselectivity is
most likely due to a change in the relative reactivity of
the allylic termini of the intermediate Pd-π-allyl species
involved. As the R group gets larger, based on the afore-
mentioned ligand effects, we one would expect that a
larger proportion of the Pd-η3-allyl would have that
large R group trans to P, which predisposes the allylic
fragment for substitution on the more substituted end.
However, as the size of R increases, the steric hindrance
toward nucleophilic substitution on that terminus also
increases. The relative importance of these factors may
impart a greater reactivity for nucleophilic substitution
on the minor isomer (unsubstituted terminus). Since
these reactions are presumably operating under Cur-
tin-Hammett conditions, even a small percentage of a
minor isomer can give rise to a major percentage of the
product if that minor isomer is sufficiently more reac-
tive.
The cinnamyl substrate 10 proceeded with addition
to the less substituted terminus. This implies that the
minor regioisomer of 24 (methylene trans to P) is much
more reactive than the major regioisomer (Ph trans to
P). However, in order for the reaction to proceed
exclusively through the minor regioisomer, it must be
occurring under Curtin-Hammett conditions; that is,
the regio- and stereoisomers of the allylic moiety must
be interconverting at a greater rate than the reaction
is occurring. This allows the more reactive isomer to be
Complex 2 and its achiral analogue 32 (Scheme 4)
were examined with NMR line broadening and SST
(44) Lloyd-Jones, G. C.; Stephen, S. C. Chem.-Eur. J. 1998, 4, 2539-
2549.
(45) Lloyd-Jones, G. C.; Stephen, S. C.; Murray, M.; Butts, C. P.;
Vyskocil, S.; Kocovsky, P. Chem.-Eur. J. 2000, 6, 4348-4357.
(46) Lloyd-Jones, G. C. Synlett 2001, 161-183.
(47) Ogasawara, M.; Takizawa, K.; Hayashi, T. Organometallics
2002, 21, 4853-4861.
(48) Bosnich, B.; Mackenzie, P. B. Pure Appl. Chem. 1982, 54, 189-
195.
(49) Mackenzie, P. B.; Whelan, J.; Bosnich, B. J. Am. Chem. Soc.
1985, 107, 2046-2054.

5080 Organometallics, Vol. 24, No. 21, 2005
Faller and Wilt
Figure 6. η3-Allyl pseudorotation (1) vs P,S-ligand hemi-
Figure 8. SST study of 2 (62 °C): kη -η -η ) 1.60 s-1; khemi
) 2.19 s-1. [Note owing to the proximity of the resonance
at 2.7 δ there is some inadvertent saturation of the
resonance.]
3
1
3
lability (2).
Scheme 4. Preparation of 32
conversion to the other enantiomer of the complex. This
process would not cause exchange between sites B and
C.
The observation of saturation in B and C must occur
via a slower process. This could occur via interconver-
sion of a T-shaped intermediate of either an η1-allyl or
a κ1-form of the hemilable ligand P,S-ligand (Figure 7
shows the exchanges for the hemilabile ligand; an η1-
allyl would provide a direct path from A to C). Direct
exchange between B and C could also occur via forma-
tion of an η1-allyl cis to S, but saturation transfer was
not observed there. The decrease in intensity at C can
be attributed to a high degree of saturation at D and
this magnetization then being transferred on to C. Thus,
a decrease in B and C upon irradiation at A is consistent
with hemilability of the P,S-ligand.
experiments in order to determine both the barriers to
η3-η1-η3 exchange and allylic termini exchange (π-allyl
pseudorotation or ligand hemilability). Several features
are noted on heating and irradiation of the allyl
resonances of 32 (Figure 7): (1) there are four reso-
nances observed for the terminal protons, and those for
the two enantiomeric complexes are equivalent; (2) the
anti resonances (A, B) are distinguished by the greater
coupling to the central proton and the resonances trans
to P by greater coupling to 31P; (3) resonances cis to S
(A,D) show greater line broadening than those cis to P;
(4) irradiation at C produces no significant saturation
transfer to B, but an increase in intensity owing to an
NOE; (5) irradiation at A produces a large decrease in
the intensity at A and smaller intensity decreases in B
and D. This demonstrates that an expected η3-η1-η3
interconversion is the fastest process, as it interconverts
anti (A) and syn (D) protons cis to P and results in
Thus, we interpret the results for complex 32 as η3-
η1-η3 exchange at a rate of ∼3.5 s-1 and allylic termini
interconversion at a rate from P,S-ligand hemilability
of ∼0.1 s-1 at 72 °C (Figure 7).
Owing to the chirality of BINAP(S) and the enantio-
meric nature of the binding of the allyl moiety to an
unsymmetrical chelate, the allyl resonances are ren-
dered diastereotopic in complex 2. If the exchanges are
as outlined for 32, then there will be two sets of
resonances that exchange among themselves. Quite
unexpectedly, the BINAP(S) complex 2 showed prefer-
ential transfer of saturation from the terminus trans
to S to the terminus trans to P. This implies that
BINAP(S) is more prone to hemilability than 31. A
saturation transfer experiment shows that allylic ter-
mini interconversion occurs ∼1.4 times faster than η3-
η1-η3 exchange at 62 °C (Figure 8). The observation that
the allylic termini exchange rate is faster than η3-η1-
η3 exchange is extremely unusual; the η3-η1-η3 ex-
change is usually the faster process. The result of this
rapid exchange is that the regioselectivity of the reaction
most likely depends on the relative reactivity of the
allylic termini to an approximately equal extent as the
steric effects presented by 1, placing the allylic amina-
tion reaction under Curtin-Hammett control. It is likely
that these rates of exchange are increased under the
reaction conditions owing to the presence of other
ligands (amines and counterions).
Figure 7. SST study on 32 (72 °C), kη -η -η ) 3.52 s-1
khemi ) 0.14 s-1
;
3
1
3
.

Pd/(S)-BINAP(S)-Catalyzed Allylic Amination
Organometallics, Vol. 24, No. 21, 2005 5081
insights into the allylic amination reaction catalyzed by
2 and its congeners will prove useful in understanding
future results and potentially in new catalyst design.
Experimental Section
General Procedures. All synthetic manipulations were
carried out using standard Schlenk techniques under inert
atmosphere. Dichloromethane and thf were distilled over
appropriate drying agents prior to use. Ethyl acetate was dried
over MgSO4 and degassed with a stream of N2 before use.
Allylpalladium chloride dimer, (η3-1-phenylallyl)palladium
chloride dimer, and (η3-1,1-dimethylallyl)palladium chloride
dimer were prepared according to their literature procedures.50
S-BINAP(S) 128 and catalyst 228 were prepared as previously
described. (Z)-Crotyl alcohol, 3-methyl-2-buten-1-ol, benzyl-
amine, (η3-crotyl)palladium chloride dimer, S-BINAP, BI-
PHEP, 1,2-bis(diphenylphosphino)benzene, and NaSbF6 were
purchased and used as received. All spectra were recorded on
either a Bruker 400 or 500 MHz spectrometer. Spectral
characterization and optical rotation data of product amines
7a,18 12,51 13,52 17a,53 28,54 29,7 and 307 were consistent with
their respective literature data. Enantiomeric excesses were
determined by 1H NMR chiral shift experiments using (-)-
MTPA as previously described.30 Optical rotations were mea-
sured on a Perkin-Elmer model 341 polarimeter at 589 nm
and 25 °C, using a 1 dm path length. Spin-saturation transfer
experiments were performed using standard techniques.55-57
Preparation of Allylic Carbonates. To a stirred solution
of the appropriate alcohol (10 mmol), pyridine (3.2 mL, 40
mmol), and DMAP (10 mg) in thf (60 mL) at 0 °C was added
ethyl chloroformate (2.6 mL, 27 mmol) dropwise. The resulting
cloudy mixture was stirred for 12 h at RT. Brine (60 mL) and
Et2O (20 mL) were added, and the organic layers were
separated, washed successively with 10% HCl and brine, dried
over MgSO4, and concentrated to leave the crude product.
Column chromatography over silica gel (10% EtOAc/90%
hexanes) provided the pure allylic carbonates as clear, colorless
to pale yellow oils. Yields were generally 75-95%.
Figure 9. Preferential saturation of diastereomeric anti
protons trans to P in complex 2 at δ 2.33.
Additionally, when complex 2 undergoes its allylic
termini interconversion, the rate constants of the ir-
radiated anti proton trans to S going to the two
diastereomeric anti protons trans to P are different, as
evidenced by the differing amounts of signal decrease
in the spin-saturation transfer experiment. This obser-
vation implies that the allylic termini exchange mech-
anism occurs mainly by ligand hemilability rather than
by a mechanism through an η1-allyl mechanism.
Because complex 2 exists as a 52:48 ratio of exo/endo
diastereomers in solution, it is clear that the energetic
difference between the two diastereomeric η3-allyl posi-
tions is negligible. Therefore, if the allylic termini
exchange was taking place via an η1-allyl mechanism,
we would expect that the two diastereomeric positions
would be saturated equally, since the incipient allyl
moiety should have no significant preference for which
diastereomer it forms. However, they are saturated in
an approximately 2:1 ratio with a preference for the
opposite diastereomer (Figure 9). This is strong evidence
that the allylic terminus exchange that is observed is
due to the hemilability of 1 on Pd. We believe that the
origin of this hemilability may lie in the chelate ring
size that the BINAP(S) ligand forms on complexation
with Pd. As both BINAP(S) and ligand 31 present the
same steric and electronic arrangement to Pd (both are
bis(triaryl)phosphine monosulfides), the only notable
difference lies in the fact that 1 forms a nine-membered
chelate ring versus the six-membered chelate ring
formed with 31.
Since ligand 1 is hemilabile on Pd, it is possible that
the true catalyst under our reaction conditions may be
one in which 1 is monodentate, with an amine, solvent
molecule, or cation filling the vacant coordination site.
In conclusion, a variety of factors are responsible for
the regioselectivity observed in the Pd/BINAP(S)-
catalyzed allylic amination reaction. The steric effects
of 1 serve to direct the sterically larger allylic terminus
trans to P. Owing to the preference for nucleophilic
attack trans to P, this should lead to the more substi-
tuted product predominantly. Since under some condi-
tions these reactions may be under Curtin-Hammett
control, the ratio of products may be unrelated to the
thermodynamic ratio observed for the equilibrating
substrates. For example with cinnamyl derivatives, it
appears that the more reactive species is that with the
less-substituted terminus trans to P and the reaction
proceeds exclusively through the minor Pd-η3-allyl
isomer. The dynamic processes that lead to the allylic
interconversion processes were examined by spin-
saturation transfer techniques, and the hemilability of
1 was confirmed. The rate of allylic termini exchange
was found to be faster than the rate of η3-η1-η3
exchange in complex 2, which is highly unusual. These
(Z)-2-Buten-1-yl ethyl carbonate (5): from (Z)-2-buten-
1
1-ol. H NMR (400 MHz, C6D6): δ 5.54 (1H, dt, J ) 6.8, 11.0
Hz); 5.41 (1H, dq, J ) 7.0, 11.0 Hz); 4.57 (2H, d, J ) 6.8 Hz);
3.90 (2H, q, J ) 7.0 Hz); 1.35 (3H, d, J ) 7.0 Hz); 0.90 (3H, t,
J ) 7.0 Hz). 13C NMR (125 MHz, C6D6): δ 156.3, 130.4, 125.2,
64.2, 63.6, 14.8, 13.5. Anal. Calcd for C7H12O3: C, 58.32; H,
8.39. Found: C, 58.25; H, 8.46.
(E)-4-Methyl-2-penten-1-yl ethyl carbonate (6): from
(E)-4-methyl-2-penten-1-ol.58 1H NMR (400 MHz, C6D6): δ 5.53
(1H, dd, J ) 6.2, 15.5 Hz); 5.42 (1H, dt, J ) 6.8, 15.5 Hz); 4.47
(2H, d, J ) 6.2 Hz); 3.91 (2H, q, J ) 7.1 Hz); 2.01 (1H, sept.,
J ) 6.8 Hz); 0.91 (3H, t, J ) 7.1 Hz); 0.79 (6H, d, J ) 6.8 Hz).
13C NMR (100 MHz, C6D6): δ 156.2, 143.8, 121.9, 68.9, 64.2,
31.6, 22.6, 14.8. Anal. Calcd for C9H16O3: C, 62.76; H, 9.36.
Found: C, 62.63; H, 9.56.
3-Methyl-2-buten-1-yl ethyl carbonate (14): from 3-meth-
1
yl-2-buten-1-ol. H NMR (400 MHz, C6D6): δ 5.35 (1H, t, J )
7.2 Hz); 4.56 (2H, d, J ) 7.2 Hz); 3.92 (2H, q, J ) 7.0 Hz);
(50) Auburn, P. R.; Mackenzie, P. B.; Bosnich, B. J. Am. Chem. Soc.
1985, 107, 2033-2046.
(51) Rehn, S.; Ofial, A. R.; Mayr, H. Synthesis 2003, 12, 1790-1796.
(52) Ohta, T.; Sasayama, H.; Nakajima, O.; Kurahashi, N.; Fujii,
T.; Furukawa, I. Tetrahedron: Asymmetry 2003, 14, 537-542.
(53) Davies, S. G.; Fox, J. F.; Jones, S.; Price, A. J.; Sanz, M. A.;
Sellers, T. G. R.; Smith, A. D.; Teixeira, F. C. J. Chem. Soc., Perkin
Trans. 1 2002, 1757-1765.
(54) Pak, C. S.; Kim, T. H.; Ha, S. J. J. Org. Chem. 1998, 63, 10006-
10010.
(55) Forsen, S.; Hoffman, R. A. J. Chem. Phys. 1964, 40, 1189.
(56) Forsen, S.; Hoffman, R. A. J. Chem. Phys. 1963, 39, 2892.
(57) Faller, J. W. Det. Org. Struct. Phys. Methods 1973, 5, 75-97.
(58) Green, M. P.; Prodger, J. C.; Hayes, C. J. Tetrahedron Lett.
2002, 43, 6609-6611.

5082 Organometallics, Vol. 24, No. 21, 2005
Faller and Wilt
1.44 (3H, s); 1.40 (3H, s); 0.91 (3H, t, J ) 7.0 Hz). 13C NMR
(125 MHz, C6D6): δ 156.5, 139.7, 119.8, 64.9, 64.1, 26.2, 18.4,
14.9. Anal. Calcd for C8H14O3: C, 60.74; H, 8.92. Found: C,
60.52; H, 9.01.
3-Ethyl-2-penten-1-yl ethyl carbonate (15): from 3-ethyl-
2-penten-1-ol.59 1H NMR (500 MHz, C6D6): δ 5.35 (1H, t, J )
7.2 Hz); 4.65 (2H, d, J ) 7.2 Hz); 3.92 (2H, q, J ) 7.0 Hz);
1.89 (2H, q, J ) 7.6 Hz); 1.81 (2H, q, J ) 7.4 Hz); 0.91 (3H, t,
J ) 7.0 Hz); 0.82 (3H, t, J ) 7.4 Hz); 0.80 (3H, t, J ) 7.6 Hz).
13C NMR (125 MHz, C6D6): δ 156.5, 150.5, 117.7, 64.7, 64.1,
29.9, 24.5, 14.9, 14.1, 13.0. Anal. Calcd for C10H18O3: C, 64.49;
H, 9.74. Found: C, 64.37; H, 9.79.
2-Cyclohexyliden-1-yl ethyl carbonate (16): from 2-cy-
clohexylidene ethanol.60 1H NMR (500 MHz, C6D6): δ 5.35 (1H,
t, J ) 7.2 Hz); 4.62 (2H, d, J ) 7.2 Hz); 3.92 (2H, q, J ) 7.0
Hz); 1.99 (2H, m); 1.89 (2H, m); 1.32 (6H, m); 0.91 (3H, t, J )
7.0 Hz). 13C NMR (125 MHz, C6D6): δ 156.5, 147.5, 116.5, 64.2,
64.1, 37.8, 29.7, 29.1, 28.5, 27.4, 14.9. Anal. Calcd for
C11H18O3: C, 66.64; H, 9.15. Found: C, 66.53; H, 9.30.
Product Amine Characterization Data. Benzyl-4-
methyl-2-pentenylamine (8b): 1H NMR (400 MHz, C6D6):
δ 7.29 (2H, d, J ) 7.4 Hz); 7.18 (2H, t, J ) 7.4 Hz); 7.09 (1H,
t, J ) 7.4 Hz); 5.55 (2H, m); 3.60 (2H, s); 3.08 (2H, d, J ) 4.0
Hz); 2.17 (1H, m); 0.93 (6H, d, J ) 6.8 Hz). 13C NMR (125 MHz,
C6D6): δ 142.0, 139.7, 129.1, 129.0, 127.6, 126.9, 54.2, 52.1,
31.8, 23.3.
trans to S, 1H, d, J ) 12.6 Hz). 31P NMR (162 MHz, CD2Cl2):
δ 41.3 (s, PdS), 41.2 (s, PdS), 26.4 (s, P), 24.0 (s, P). 13C NMR
(125 MHz, CD2Cl2, allylic region only): δ 120.2 (central, d, JPC
) 6.5 Hz); 119.4 (central, d, JPC ) 6.5 Hz); 80.3 (trans to P, d,
JPC ) 32.1 Hz); 78.7 (trans to P, d, JPC ) 32.1 Hz); 66.3 (trans
to S, d, JPC ) 3.0 Hz); 64.9 (trans to S, d, JPC ) 3.0 Hz). Anal.
Calcd for C39H33F6P2PdSSb: C, 49.95; H, 3.55. Found: C,
49.69; H, 3.81.
BIPHEP(S) (21). To a solution of BIPHEP (678 mg, 1.3
mmol) in thf (20 mL) was added sulfur (29 mg, 0.91 mmol),
and the resulting cloudy solution was stirred until clear. The
solvent was removed under reduced pressure, and the residue
was subjected to column chromatography over silica gel (3:1
CH2Cl2/hexanes) to provide the pure product in 27% yield
based on sulfur (136 mg). Compound 11 can be recrystallized
1
from CH2Cl2/hexanes. H NMR (500 MHz, CD2Cl2): δ 7.73-
7.64 (4H, m); 7.46 (1H, t, J ) 7.2 Hz); 7.36 (3H, m); 7.32-7.17
(12H, m); 7.12 (2H, t, J ) 7.8 Hz); 7.05 (1H, t, J ) 7.7 Hz);
7.05 (1H, t, J ) 7.7 Hz); 6.93 (1H, dd, J ) 7.8, 11.6 Hz); 6.72
(1H, t, J ) 7.8 Hz); 6.51 (1H, m); 6.40 (1H, m). 31P NMR (202
MHz, CD2Cl2): δ 42.0 (PdS); -14.7 (P). Anal. Calcd for
C36H28P2S: C, 77.96; H, 5.09. Found: C, 77.49; H, 4.90.
[(η3-Crotyl)Pd(S-BINAP(S))]SbF6 (22): yield 77%. Com-
plex 8 exists predominantly as a 55:45 ratio of exo/endo
diastereomers with Me trans to P. Major diastereomer: (400
MHz, CD2Cl2): δ 8.34-6.02 (32 H, aryl region); 5.12 (central
H, ddd, J ) 6.4, 6.9, 12.8 Hz); 4.05 (1H, anti to central, trans
to P, ddq, JHH ) 6.4, 12.8 Hz, JPH ) 10.2 Hz); 3.43 (1H, syn to
central, trans to S, d, J ) 6.9 Hz); 2.32 (1H, anti to central,
Benzyl-3-ethyl-2-pentenylamine (18a): 1H NMR (400
MHz, C6D6): δ 7.37 (2H, d, J ) 7.4 Hz); 7.20 (2H, t, J ) 7.4
Hz); 7.11 (1H, t, J ) 7.4 Hz); 5.51 (1H, dd, J ) 11.0, 17.7 Hz);
5.08 (1H, d, J ) 11.0 Hz); 5.03 (1H, d, J ) 17.7 Hz); 3.47 (2H,
s); 1.33 (4H, q, J ) 7.4 Hz); 0.76 (6H, t, J ) 7.4 Hz). 13C NMR
(100 MHz, C6D6): δ 145.7, 142.7, 129.2, 129.1, 127.6, 114.4,
60.2, 47.1, 28.8, 8.2.
trans to S, d, J ) 12.0 Hz); 1.81 (3H, dd, JHH ) 6.4 Hz, JPH
)
9.8 Hz). Minor diastereomer: (400 MHz, CD2Cl2): δ 8.34-6.02
(32 H, aryl region); 5.12 (central H, ddd, J ) 6.4, 6.9, 12.8
Hz); 4.34 (1H, anti to central, trans to P, ddq, JHH ) 6.4, 12.8
Hz, JPH ) 10.2 Hz); 3.76 (1H, syn to central, trans to S, d, J )
6.9 Hz); 2.17 (1H, anti to central, trans to S, d, J ) 12.0 Hz);
1.65 (3H, dd, JHH ) 6.4 Hz, JPH ) 9.8 Hz). 31P NMR (both major
diastereomers, 162 MHz, CD2Cl2): δ 45.5 (PdS, d, JPP ) 4 Hz);
44.6 (PdS, d, JPP ) 4 Hz); 25.5 (P, d, JPP ) 4 Hz); 23.9 (P, d,
JPP ) 4 Hz). 13C NMR (both major diastereomers, 125 MHz,
CDCl3, allylic region only): δ 118.3 (central C, d, JPC ) 4.7
Hz); 116.9 (central C, d, JPC ) 4.7 Hz); 102.1 (trans to P, d,
JPC ) 28.0 Hz); 98.2 (trans to P, d, JPC ) 28.0 Hz); 61.3 (trans
to S, d, JPC ) 2.8 Hz); 58.2 (trans to S, d, JPC ) 2.8 Hz); 17.9
(Me, d, JPC ) 4.2 Hz); 17.8 (Me, d, JPC ) 4.2 Hz). Anal. Calcd
for C48H39F6P2PdSSb: C, 54.80; H, 3.74. Found: C, 54.85; H,
3.72.
Benzyl-1-vinylcyclohexylamine (19a): 1H NMR (400
MHz, C6D6): δ 7.38 (2H, d, J ) 7.4 Hz); 7.20 (2H, t, J ) 7.4
Hz); 7.11 (1H, t, J ) 7.4 Hz); 5.55 (1H, dd, J ) 10.9, 17.7 Hz);
5.07 (1H, d, J ) 10.9 Hz); 4.98 (1H, d, J ) 17.7 Hz); 3.51 (2H,
s); 1.62-1.14 (10H, m). 13C NMR (125 MHz, C6D6): δ 146.5,
142.9, 129.2, 129.1, 127.5, 113.2, 56.7, 47.2, 36.1, 27.1, 22.7.
Anal. Calcd for C15H21N: C, 83.67; H, 9.83; N, 6.50. Found:
C, 83.34; H, 9.83; N, 6.37.
General Procedure for Syntheses of (η3-allyl)Pd(Li-
gand)SbF6 Complexes 20, 22, 23, 24, 31, and 33. To a flame-
dried Schlenk flask were added the appropriate allylpalladium
chloride dimer (1 equiv), the appropriate ligand (2 equiv), and
NaSbF6 (2.5 equiv). The contents were placed under inert
atmosphere, and CH2Cl2 was added. The resulting cloudy
solution was stirred for 16 h at RT. The solution was then
filtered through a pad of Celite, evaporated to dryness, and
triturated with Et2O to leave the desired compound as a yellow
powder. Yields and crystallization details (where appropriate)
are given below with the individual complexes.
[(η3-1,1-Dimethylallyl)Pd(S-BINAP(S))]SbF6 (23): yield
90%. X-ray quality yellow crystals were grown from slow
evaporation of a CH2Cl2/Et2O solution. Major diastereomer:
1H NMR (400 MHz, CD2Cl2): δ 8.20-6.46 (31 H, aryl region);
6.04 (1H, aryl, d, J ) 8.6 Hz); 4.88 (1H, central allylic H, dd,
J ) 7.5, 13.0 Hz); 3.57 (1H, syn to central, trans to S, d, J )
7.5 Hz); 2.24 (1H, anti to central, trans to S, d, J ) 13.0 Hz);
1.72 (3H, syn to central, d, JPH ) 9.8 Hz); 1.19 (3H, anti to
central, d, JPH ) 6.3 Hz). 31P NMR (162 MHz, CD2Cl2): δ 45.3
(PdS, d, JPP ) 4 Hz); 26.6 (P, d, JPP ) 4 Hz). 13C NMR (125
[(η3-Allyl)Pd(BIPHEP(S))]SbF6 (20): yield 86%. X-ray
quality crystals were grown from CH2Cl2/Et2O. In solution,
complex 10 exists as a 1:1 ratio of diastereomers. 1H NMR (400
MHz, CD2Cl2): δ 8.26-6.29 (28H, aryl region); 5.51 (central,
1H, dddd, J ) 7.0, 7.6, 12.6, 14.0 Hz); 5.18 (central, 1H, dddd,
J ) 7.0, 7.6, 12.6, 14.0 Hz); 4.93 (syn to central, trans to P,
1H, dd, JPH ) 7.0 Hz, JHH ) 7.6 Hz); 4.58 (syn to central, trans
to P, 1H, dd, J ) JPH ) 7.0 Hz, JHH ) 7.6 Hz); 4.00 (syn to
central, trans to S, 1H, d, J ) 7.0 Hz); 3.91 (syn to central,
trans to S, 1H, d, J ) 7.0 Hz); 3.65 (anti to central, trans to P,
1H, dd, JPH ) 10.2 Hz, JHH ) 14.0 Hz); 2.92 (anti to central,
trans to P, 1H, dd, JPH ) 10.2 Hz, JHH ) 14.0 Hz); 2.72 (anti
to central, trans to S, 1H, d, J ) 12.6 Hz); 2.35 (anti to central,
MHz, CDCl3, allylic region only): δ 117.4 (quat. C, d, JPC
)
25.5 Hz); 112.0 (central C, d, JPC ) 4.7 Hz); 57.6 (-CH2, d,
JPC ) 2 Hz); 27.4 (Me, d, JPC ) 4.8 Hz); 22.3 (Me, d, JPC ) 5.8
Hz). Minor diastereomer: 1H NMR (400 MHz, CD2Cl2):
δ
8.20-6.46 (31 H, aryl region); 5.81 (1H, aryl, d, J ) 8.6 Hz);
4.72 (1H, central allylic H, dd, J ) 7.5, 13.0 Hz); 3.00 (1H,
syn to central, trans to S, d, J ) 7.5 Hz); 2.84 (1H, anti to
central, trans to S, d, J ) 13.0 Hz); 1.93 (3H, syn to central, d,
JPH ) 10.0 Hz); 1.19 (3H, anti to central, d, JPH ) 6.6 Hz). 31
P
NMR (162 MHz, CD2Cl2): δ 47.3 (PdS, d, JPP ) 6 Hz); 24.5
(P, d, JPP ) 6 Hz). 13C NMR (125 MHz, CDCl3, allylic region
only): δ 116.3 (quat. C, d, JPC ) 25.5 Hz); 110.1 (central C, d,
(59) Alhassan, S. S.; Cameron, R. J.; Curran, A. W. C.; Lyall, W. J.
S.; Nicholson, S. H.; Robinson, D. R.; Stuart, A.; Suckling, C. J.;
Stirling, I.; Wood, H. C. S. J. Chem. Soc., Perkin Trans. 1 1985, 1645-
1659.
(60) Srikrishna, A.; Rao, M. S.; Gharpure, S. J.; Babu, N. C. Synlett
2001, 1986-1988.
JPC ) 4.7 Hz); 56.8 (-CH2, d, JPC ) 2 Hz); 27.5 (3H, d, JPC
)
4.8 Hz); 21.8 (3H, d, JPC ) 5.8 Hz). Anal. Calcd for C49H41F6P2-
PdSSb: C, 55.21; H, 3.88. Found: C, 55.13; H, 3.85.

Pd/(S)-BINAP(S)-Catalyzed Allylic Amination
Organometallics, Vol. 24, No. 21, 2005 5083
[(η3-1-Phenylallyl)Pd(S-BINAP(S))]SbF6 (24): yield 82%.
X-ray quality yellow crystals were grown from a CH2Cl2/Et2O
solution. Complex 24 exists as a 52:44:4 ratio of diastereomers
in solution. 1H NMR (52% isomer, 400 MHz, CD2Cl2): δ 8.24-
6.50 (aryl region, 36 H); 5.85 (aryl, 1H, d, J ) 8.6 Hz); 5.84
(central, 1H, ddd, J ) 6.8, 12.8, 13.0 Hz); 5.06 (trans to P, 1H,
dd, J ) 10.7, 13.0 Hz); 3.97 (syn to central, trans to S, 1H, d,
J ) 6.8 Hz); 2.41 (anti to central, trans to S, 1H, d, J ) 12.8
subjecting branched N-(1-phenyl-2-propenyl)benzylamine8 to
our standard reaction conditions, we observed partial isomer-
ization (50%) to the linear product 12 over a 24 h period. This
rearrangement could reflect the greater thermodynamic stabil-
ity of the linear product with respect to the conjugation of the
double bond into the aromatic ring. The rate of isomerization
would suggest that isomerization of a kinetic branched isomer
could account for only part of the linear isomer. This control
experiment, however, does not have an excess of ethyl carbon-
ate, which could accelerate the process.
Structure Determination and Refinement. Crystals
were obtained by slow diffusion of diethyl into methylene
chloride solution of complexes 20, 23, and 24. Data were
collected on a Nonius KappaCCD (Mo KR radiation) diffrac-
tometer at -100 °C and were not specifically corrected for
absorption other than the inherent corrections provided by
Scalepack.64 The structures were solved by direct methods
(SIR92)65 and refined on F for all reflections.66 Non-hydrogen
atoms were refined with anisotropic displacement parameters.
Hydrogen atoms were included at calculated positions. Rel-
evant crystal and data parameters are presented in Table 1.
Structure Determination of 20. This structure determi-
nation was straightforward, and the compound had crystal-
lized in a triclinic cell with Z ) 2, indicating a space group of
P1h. There was a disorder with two orientations of the allyl in
a ratio of 54:46. The major conformer is shown in Figure 1.
Note that the Pd to terminal carbon atom distances are the
average of the two conformers.
1
Hz). H NMR (44% isomer, 400 MHz, CD2Cl2): δ 8.24-6.50
(aryl region, 36 H); 6.06 (aryl, 1H, d, J ) 8.6 Hz); 5.67 (central,
1H, ddd, J ) 6.8, 12.8, 13.0 Hz); 4.78 (trans to P, 1H, dd, J )
10.7, 13.0 Hz); 3.55 (syn to central, trans to S, 1H, d, J ) 6.8
Hz); 2.52 (anti to central, trans to S, 1H, d, J ) 12.8 Hz). 31P
NMR (both major isomers, 162 MHz, CD2Cl2): δ 45.5 (PdS,
d, J ) 4.0 Hz); 43.7 (PdS, d, J ) 4.0 Hz); 25.3 (P, d, J ) 4.0
Hz); 25.1 (P, d, J ) 4.0 Hz). 13C NMR (both major isomers,
allyl region only, 125 MHz, CD2Cl2): δ 111.6 (central, d, JPC
) 6.9 Hz); 111.5 (central, d, JPC ) 6.9 Hz); 103.3 (trans to P,
d, JPC ) 27.5 Hz); 96.5 (trans to P, d, JPC ) 27.5 Hz); 63.3
(trans to S, d, JPC ) 2.8 Hz); 59.0 (trans to S, d, JPC ) 2.8 Hz).
Anal. Calcd for C53H41F6P2PdSSb: C, 57.14; H, 3.71. Found:
C, 57.18; H, 3.68.
1,2-Bis(diphenylphosphino)benzene monosulfide (31):
prepared in an identical manner to BIPHEP(S) (21); yield 65%
1
with respect to S; recrystallized from CH2Cl2/Et2O. H NMR
(400 MHz, CDCl3): δ 7.81-7.73 (aryl, 4H); 7.42-7.26 (aryl,
8H); 7.23-7.06 (aryl, 8H); 7.04-6.97 (aryl, 4H). 31P NMR (162
MHz, CDCl3): δ 45.0 (PdS, d, JPP ) 31.7 Hz); -16.5 (P, d, JPP
) 31.7 Hz). 13C NMR (100 MHz, CDCl3): δ 142.8-128.4 (aryl,
30 C). Anal. Calcd for C30H24P2S: C, 75.23; H, 5.06. Found:
C, 74.87; H, 4.80.
Structure Determination of 23. The data showed an
orthorhombic cell with absences consistent with P21. There
were no disorder problems with this structure, as shown in
Figure 4.
[(η3-Allyl)Pd(1,2-bis(diphenylphosphino)benzene mono-
sulfide)]SbF6 (32): yield 93%. X-ray quality yellow crystals
were grown from a CH2Cl2/hexanes solution. 1H NMR (500
MHz, CDCl3): δ 7.79 (aryl, 1H, app. t, J ) 7.6 Hz); 7.62 (aryl,
1H, app. t, J ) 7.6 Hz); 7.57-7.21 (aryl, 18H); 7.11 (aryl, 1H,
d, J ) 7.7 Hz); 7.09 (aryl, 1H, d, J ) 8.2 Hz); 6.99 (aryl, 1H,
d, J ) 7.7 Hz); 6.97 (aryl, 1H, d, J ) 8.2 Hz); 5.64 (dddd, J )
7.0, 7.3, 12.7, 13.8 Hz); 4.92 (syn to central, trans to P, 1H,
dd, J ) 7.3, 10.0 Hz); 3.71 (anti to central, trans to P, 1H, dd,
J ) 10.0, 13.8 Hz); 3.64 (syn to central, trans to P, 1H, d, J )
7.0 Hz); 2.85 (anti to central, trans to S, 1H, d, J ) 12.7 Hz).
31P NMR (162 MHz, CDCl3): δ 42.4 (PdS, d, JPP ) 47.4 Hz);
15.1 (P, d, JPP ) 47.4 Hz). 13C NMR (125 MHz, CD2Cl2): δ
120.2 (central, JPC ) 4.8 Hz); 78.0 (trans to P, JPC ) 28.4 Hz);
70.3 (trans to S, JPC ) 2.8 Hz). Anal. Calcd for C33H29F6P2-
PdSSb: C, 45.99; H, 3.39. Found: C, 45.94; H, 3.33.
Structure Determination of Orthorhombic 24. The
data showed an orthorhombic cell with absences consistent
with P212121. An endo:exo disorder of the cinnamyl ligand was
found with both conformers having the phenyl group trans to
P, but the geometry superimposed the phenyl groups to some
extent. The phenyl group was modeled with a single rigid
group and partial occupancies (67:33) for two of the allyl
carbons. This approximation will lead to some significant error
in the Pd-C distances. Neverthless, the overall trend of longer
distances trans to P are evident. This structure was included
to demonstrate that, although some metrical data may be
suspect, the geometrical preference for the substituted termi-
nus being trans to P is retained.
Acknowledgment. We thank the donors of the
Petroleum Research Fund administered by the
American Chemical Society for support of our work
(PRF#43212).
General Procedure for Catalytic Allylic Amination.
These reactions (Tables 1-3) were performed as previously
reported30 at the indicated temperatures.
Stability of Amine Products under Reaction Condi-
tions. Since amination at more substituted termini of allyls
has the potential of rearranging to linear products,61-63 we
tested for potential conversion to thermodynamic products.
This is particularly relevant to the formation of linear product
from cinnamyl reactions. It was found that, upon subjecting
amine product 7a to our standard reaction conditions, no
isomerization or racemization was observed. However, upon
Supporting Information Available: Crystallographic
data, 13C NMR data, and rate constant calculations for SST
experiments. This material is available free of charge via the
Internet at http://pubs.acs.org.
OM050537R
(64) Minor, W., Otwinowski, Z., Eds. HKL2000 (Denzo-SMN) soft-
ware package. Processing of X-ray Diffraction Data Collected in
Oscillation Mode. In Methods in Enzymology, Macromolecular Crystal-
lography; Academic Press: New York, 1997.
(65) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A. J.
Appl. Crystallogr. 1993, 26, 343.
(66) TEXSAN for Windows version 1.06, Crystal Structure Analysis
Package; Molecular Structure Corporation, 1997-1999.
(61) Vitagliano, A.; Akermark, B. J. Organomet. Chem. 1988, 349,
C22-C26.
(62) Backvall, J. E.; Nordberg, R. E.; Zetterberg, K.; Akermark, B.
Organometallics 1983, 2, 1625-1629.
(63) Nilsson, Y. I. M.; Andersson, P. G.; Backvall, J. E. J. Am. Chem.
Soc. 1993, 115, 6609-6613.