Isotope effects and the nature of stereo- and regioselectivity in hydroaminations of vinylarenes catalyzed by palladium(II)-diphosphine complexes

Article abstract of DOI:10.1021/ol049137a

(Equation Presented) The hydroamination of styrene with aniline catalyzed by phosphine-ligated palladium triflates exhibits a substantial ¹³C isotope effect at the benzylic carbon. This supports rate-determining nucleophilic attack of amine on a η³-phenethyl palladium complex. Deuterium exchange observations and predicted isotope effects based on DFT calculations support this mechanism. Selectivity in these reactions is determined by the facility of palladium displacement after reversible hydropalladation of the alkene.

Full text of DOI:10.1021/ol049137a

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
ORGANIC
LETTERS
2004
Vol. 6, No. 14
2469-2472
Isotope Effects and the Nature of
Stereo- and Regioselectivity in
Hydroaminations of Vinylarenes
Catalyzed by Palladium(II)−Diphosphine
Complexes
Loan K. Vo and Daniel A. Singleton*
Department of Chemistry, Texas A&M UniVersity, P.O. Box 30012,
College Station, Texas 77842
singleton@mail.chem.tamu.edu
Received May 13, 2004
ABSTRACT
The hydroamination of styrene with aniline catalyzed by phosphine-ligated palladium triflates exhibits a substantial 13C isotope effect at the
benzylic carbon. This supports rate-determining nucleophilic attack of amine on a η3-phenethyl palladium complex. Deuterium exchange
observations and predicted isotope effects based on DFT calculations support this mechanism. Selectivity in these reactions is determined
by the facility of palladium displacement after reversible hydropalladation of the alkene.
Due to the ubiquity of the amine functionality in natural
products, there has been considerable interest in the develop-
ment of transition metal-catalyzed hydroaminations of al-
kenes and alkynes.1,2 One highly promising approach to the
hydroamination of olefins, reported by Hartwig and co-
workers, is the addition of anilines to vinylarenes catalyzed
by a combination of Pd(OCOCF3)2, a chelating bis-phosphine
such as DPPF, and a triflate ion source such as TfOH. This
system exhibits high Markovnikov selectivity and high yields
(eq 1). Enantioselective reactions can be achieved with
BINAP as ligand.3
(1) (a) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. J. Am. Chem. Soc.
1988, 110, 8729. (b) Johnson, J. S.; Bergman, R. G. J. Am. Chem. Soc.
2001, 123, 2923. (c) Ackermann, L.; Bergman, R. G. Org. Lett. 2002, 4,
1475. (d) Heutling, A.; Doye, S. J. Org. Chem. 2002, 67, 1961. (e) Li, Y.;
Marks, T. J. J. Am. Chem. Soc. 1996, 118, 9295. (f) Tian, S.; Arredondo,
V. M.; Stern, C. L.; Marks, T. J. Organometallics 1999, 18, 2568. (g)
Douglass, M. R.; Ogasawara, M.; Hong, S.; Metz, M. V.; Marks, T. J.
Organometallics 2002, 21, 283. (h) Beller, M.; Trauthwein, H.; Eichberger,
M.; Breindl, C.; Muller, T. L. Eur. J. Inorg. Chem. 1999, 7, 1121. (i)
Utsunomiya, M.; Kuwano, R.; Kawatsura, M.; Hartwig, J. F. J. Am. Chem.
Soc. 2003, 125, 5608.
Future progress in broadening the range of effective sub-
strates for this reaction and increasing control over selectivity
may depend on a detailed understanding of the mechanism.
A recent advance was Hartwig’s isolation of η3-benzyl
palladium complexes under reaction conditions.4 In stoichio-
metric reactions, these complexes are converted to hydroami-
nation products prior to reaction of external arene. This
(2) (a) Muller, T. E.; Beller, M. Chem. ReV. 1998, 98, 675. (b) Nobis,
M.; Driessen-Holscher, B. Angew. Chem., Int. Ed. 2001, 40, 3983.
(3) Kawatsura, M.; Hartwig, J. F. J. Am. Chem. Soc. 2000, 122, 9546.
(4) Nettekoven, U.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 1166.
10.1021/ol049137a CCC: $27.50 © 2004 American Chemical Society
Published on Web 06/18/2004

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
supports their catalytic intermediacy. From this and other
observations, Hartwig proposed the mechanism of Scheme
1. In this mechanism, the Pd(II) hydride 4 effects hydropal-
Scheme 1
Figure 1. 13C KIEs (k12 /k13 ) for hydroamination of styrene with
C
aniline catalyzed by Pd(COCOCF3)2, DPPF, and TfOH in benzene
at 90 °C. Standard deviations (n ) 6) in the last digit are shown in
parentheses.
In a catalytic reaction, the observed KIEs need not reflect
the rate-limiting or turnover-limiting step for the catalytic
cycle, but the KIEs do reflect the first irreversible step
between free substrate and product. Qualitatively, the large
(≈1.030) 13C KIE observed at the R-olefinic carbon indicates
a major σ-bonding change at this carbon in this step. An
SN2 reaction, for example, might typically exhibit an isotope
effect of this magnitude.6 The formation of an alkene-metal
π-complex such as the formation of 5 would be expected to
exhibit much smaller, approximately equal 13C KIEs at the
olefinic carbons.7 The â-olefinic 13C KIE, at ≈1.006, is
smaller than would be expected for an olefin insertion step,
such as the conversion of 5 to 6. For example, Landis
observed a 13C KIE of 1.017 at the analogous carbon in a
Ziegler-Natta polymerization.8 Within the confines of the
Hartwig mechanism of Scheme 1, the 13C KIEs strongly
support nucleophilic attack on 6 as the selectivity-determining
first irreversible conversion of styrene. Since 6 is the apparent
resting-state of the catalyst, this step would also be the rate-
determining step.
If this is the case, the η3-benzyl ligand in 6 would have to
be in rapid equilibrium with free styrene. To test this, a
hydroamination reaction of styrene employing 65 mol % of
partially deuterated aniline (78% deuterium content, prepared
by equilibration of aniline with D2O) was taken to 32%
conversion (eq 2). Analysis of recovered unreacted styrene
revealed approximately 25% deuterium in each of the olefinic
positions (see 8-10), but no detectable deuterium (<1%)
was found in the aromatic ring. From the 2H NMR, the ratio
of 8:9:10 (ignoring multiply deuterated isomers) was 1.00:
1.01:1.06. On the basis of calculated equilibrium isotope
effects for styrene, the ratio of 8:9:10 at equilibrium would
be 1.00:1.01:1.07.9,10 From this, it appears that the exchange
of deuterium into the olefinic positions of styrene had reached
equilibrium.11 This experiment cannot rule out H/D equili-
bration occurring via a side reaction that is unrelated to the
catalytic cycle, but the formation of 8 and 9 supports the
conclusion from KIEs that incorporation of styrene into the
η3 complex is reversible. In addition, the formation of 10
ladation of the vinyl arene via π-complex 5 to afford η3-
benzyl complex 6. The product is then formed by external
nucleophilic attack of amine on the η3 complex.
The observations above in stoichiometric reactions sug-
gested that formation of the η3 complex is irreversible, in
which case the hydroamination’s regio- and stereoselectivity
would be determined by selectivity in the hydropalladation
process. On the other hand, rate observations for the forward
reaction of a η3 complex versus the overall catalytic reaction
were consistent with the observed reaction enantioselectivity,4
suggesting that nucleophilic attack on the η3 complex, not
its formation, determines the selectivity. In addition, Hartwig
made the interesting observation that an isolated η3-benzylic
complex, presumably the major stereoisomer formed, af-
forded the catalytic reaction’s minor enantiomeric product.
This seems unusual if the selectivity is determined at the
stage of formation of the η3 complex.
With the goal of deciphering the nature of the selectivity-
determining step under synthetic conditions, we first deter-
mined 13C kinetic isotope effects (KIEs) for the hydroami-
nation of styrene with aniline. The complete set of 13C KIEs
were determined at natural abundance by NMR methodol-
ogy.5 Four hydroaminations of ≈5 M styrene in benzene with
limiting aniline catalyzed by 2 mol % Pd(OCOCF3)2, 3 mol
% DPPF, and 20 mol % TfOH at 90 °C were taken to
77-90% conversion. Under these conditions, the Markovni-
kov hydroamination product is formed in 92% isolated yield.
The unreacted styrene was recovered by chromatography
followed by fractional distillation and then analyzed by 13
C
NMR. Changes in isotopic composition relative to the
original styrene were determined using the aromatic para
carbon as a standard, assuming that its isotopic composition
does not change. From the changes in isotopic composition,
the 13C KIEs were calculated as previously described.5 The
isotope effects obtained from four independent reactions are
summarized in Figure 1.
(5) Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117, 9357.
(6) Lee, J. K.; Bain, A. D.; Berti, P. J. J. Am. Chem. Soc. 2004, 126,
3769.
(7) Frantz, D. E.; Singleton, D. A. J. Am. Chem. Soc. 2000, 122, 3288.
(8) Landis, C. R.; Rosaaen, K. A.; Uddin, J. J. Am. Chem. Soc. 2002,
124, 12062.
2470
Org. Lett., Vol. 6, No. 14, 2004

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
suggests that the regioisomerically opposite hydropalladation
of styrene also occurs reversibly.
11 (the model for 4), but it would also be accessible by
simple proton transfer to the Pd(0)-styrene complex 12. The
energetics of 11 suggest that it is unlikely to be formed from
protonation of a Pd(0)-bis-phosphine by an ammonium salt,
but the catalytic cycle could readily shunt the high-energy
11 via 12. There is virtually no barrier to formation of the
η3 complex 15 from 13: after inclusion of zero-point energy,
the transition structure for olefin insertion (14) is lower in
energy than 13. The highest predicted barrier in the cycle is
the conversion of 15 to 17 via transition structure 16. This
is true also when estimates of entropy are taken into
account: the predicted free energy barrier for the cycle versus
standard-state reactants is 28.6 kcal/mol. This barrier is
consistent with a reaction that takes several hours at 100
°C. Product 17, after the nucleophilic substitution step, is
actually higher in energy than the η3 complex 15, but it is
downhill and presumably rapid for 17 to dissociate into the
protonated product and the Pd(0)-bis-phosphine complex
to continue the cycle. The free-energy barrier going backward
from 15 is predicted to be 16.6 kcal/mol, much lower than
the barrier going forward.12 Overall, the calculations predict
that the η3 complex would be the resting point in the catalytic
cycle with reversible incorporation of styrene to that point.
This is in excellent qualitative agreement with the 13C KIEs
and deuterium incorporation observations.
Theoretical calculations were employed to provide a more
detailed interpretation of the isotope effects and to explore
the complete catalytic cycle. A problem in this system is
the difficulty of accurately accounting for the charge-charge
interaction of cationic complexes with the triflate counterion.
This problem was made calculationally tractable by leaving
out the triflate entirely but including a PCM solvent model.
Comparison with experimental observations such as KIEs
will be used to assess whether the resulting structures are
reasonably representative of the solution chemistry.
The model catalytic cycle of Figure 2 was obtained in
B3LYP calculations employing an SDD basis set and core
potential on Pd, with a 6-31+G** basis set on other atoms.
The π-complex 13 is predicted to be readily formed from
13C KIEs for transition structures 14 and 16 (Figure 3)
were predicted from the scaled vibrational frequencies9 by
Figure 3. Predicted 13C KIEs(k12 /k13 ) at 90 °C. (a) KIEs if
C
C
hydropalladation is irreversible, based on transition structure 14.
(b) KIEs if hydropalladation is reversible and nucleophilic attack
of the η3 complex is the first irreversible step for styrene, based on
transition structure 16.
the method of Bigeleisen and Mayer,10 including a one-
dimensional tunneling correction.13 13C KIE predictions have
proven to be highly accurate in reactions not involving
hydrogen transfer, so long as the calculation accurately
(9) The calculations used the program QUIVER (Saunders, M.; Laidig,
K. E.; Wolfsberg, M. J. Am. Chem. Soc. 1989, 111, 8989). Becke3LYP
frequencies were scaled by 0.9614.
(10) (a) Bigeleisen, J.; Mayer, M. G. J. Chem. Phys. 1947, 15, 261. (b)
Wolfsberg, M. Acc. Chem. Res. 1972, 5, 225. (c) Bigeleisen, J. J. Chem.
Phys. 1949, 17, 675.
1
(11) The loss of integration in the H NMR for the olefinic hydrogens
was also consistent with complete equilibration, but the uncertainty in this
measurement is relatively large.
Figure 2. Calculated model catalytic cycle for the hydroamination
of styrene in B3LYP calculations with an SDD basis set on Pd and
6-31+G** on all other atoms. The energies shown (in kcal/mol)
include zpe and a PCM solvent model and are relative to 15 if
going forward from 15 through the cycle. The overall cycle is
downhill by 7.2 kcal/mol.
(12) A referee suggested that the observation of forward-going η3
complex in stoichiometric reactions (ref 4) could be the result of use of
100 equiv of aniline. If so, then the barriers for forward and back reactions
from 15 are much closer in energy than the calculations predict. Notably,
however, the concentration of aniline in the KIE experiments here is roughly
similar to that in stoichiometric reactions in ref 4.
(13) Bell, R. P. The Tunnel Effect in Chemistry; Chapman & Hall:
London, 1980; pp 60-63.
Org. Lett., Vol. 6, No. 14, 2004
2471

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
depicts the mechanism and transition-state geometry.14 The
hydrogen transfer associated with transition structure 14
makes the predicted KIEs for this step less quantitatively
reliable, but the qualitative pattern of predicted isotope effects
should be correct.
If hydropalladation of styrene is irreversible, medium-sized
KIEs are predicted (Figure 3a) for both of the olefinic
carbons of styrene, with very small KIEs for the aromatic
carbons. This pattern does not fit with the experimental KIEs
in Figure 1. In contrast, for nucleophilic attack on the η3
complex, a large 13C KIE is predicted at the R-olefinic carbon
along with small but significant KIEs at the ortho and ipso
aromatic carbons (Figure 3b). Considering the differences
between the calculational model and the actual experimental
system, the agreement between the predicted KIEs of Figure
3b and experiment is excellent.15
The regiochemically opposite hydropalladation of styrene
suggested by the formation of 10 was also modeled theo-
retically. The π-complex 19 leading to the reverse hydro-
palladation is very similar in energy to 13 in the regular
catalytic cycle. Although olefin insertion in 19 is less
exothermic than the formation of η3 complex 15, there is
little barrier predicted for the insertion process via 20 to
afford 21. This is consistent with the formation of 8 and 9
versus 10 at roughly similar rates. The electron-deficient Pd
in 21 attempts to gain an extra ligand through an agostic
interaction with an adjacent C-H bond. However, 21 is still
much less stable than the alternative 16e- 15, and the re-
versed hydropalladation is predicted to be readily reversible.
16. The high energy of this transition state accounts for the
regiospecificity of the reaction.
The enantioselectivity observed in reactions employing
BINAP-type ligands should reflect factors influencing
the relative energy of diastereomeric transition states for
nucleophilic attack on the η3 complex. A qualitative model
for this transition state was built by combining transi-
tion structure 16 and a molecular mechanics model of
(R)-BINAP-Pd (with parameters chosen to approximate
Hartwig’s crystal structure4 of a η3 complex). At the stage
of the η3 complex, there was no obvious steric preference
among the possible diastereomeric structures; this is in line
with Hartwig’s observation of multiple η3 complexes present
under the reaction conditions.4 However, the η3 complex
must twist as it turns into a η2-benzene complex in the
process of nucleophilic displacement. In the complex leading
to the major product (Figure 4a), this twist appears to relieve
Figure 4. Model for understanding the enantioselectivity in
hydroaminations mediated by BINAP-type ligands. (a) Transition
state leading to the major enantiomer. (b) Transition state leading
to the minor enantiomer.
The mechanistic picture that emerges from the combination
of experimental observations and calculations provides
insight into the selectivity in these reactions. The regiochem-
istry of the overall hydroamination is not derived from
regioselectivity in the hydropalladation process but rather
reflects the relative energy of transition states for nucleophilic
displacement on a η3 complex versus a primary alkyl Pd
complex. Our model for the latter process was attack of NH3
on 21; a transition structure located for this displacement
(see Supporting Information) was 11.8 kcal/mol higher than
steric interactions. In the complex leading to the minor
enantiomeric product, the necessary twist of the benzyl group
seems to increase steric interactions. It should be emphasized
that this is currently only a qualitative model, but it does
account for the direction of the observed enantioselectivity.
In summary, a combination of kinetic isotope effects,
deuterium exchange observations, and theoretical calculations
supports a consistent mechanism in which the key selectivity-
determining step is a nucleophilic displacement on a η3 com-
plex. This suggests ways to gain further control of selectivity
in these reactions, and we plan to pursue such studies.
(14) (a) Beno, B. R.; Houk, K. N.; Singleton, D. A. J. Am. Chem. Soc.
1996, 118, 9984. (b) Meyer, M. P.; DelMonte, A. J.; Singleton, D. A. J.
Am. Chem. Soc. 1999, 121, 10865. (c) DelMonte, A. J.; Haller, J.; Houk,
K. N.; Sharpless, K. B.; Singleton, D. A.; Strassner, T.; Thomas, A. A. J.
Am. Chem. Soc. 1997, 119, 9907. (d) Singleton, D. A.; Merrigan, S. R.;
Liu, J.; Houk, K. N. J. Am. Chem. Soc. 1997, 119, 3385.
(15) The experimental KIE of ≈1.006 at the â-carbon of styrene is
recognizably in the opposite direction of the predicted KIE of 0.999 at this
carbon. This discrepancy may be the result of small amounts of polymer-
ization during the reaction or the distillative reisolation of the unreacted
styrene, as polystyrene was always observed in the pot residue. Polymer-
ization of styrene would contribute to the observed â-carbon KIE but have
little effect on the other carbons. For 13C KIEs for free-radical polymeri-
zation, see: Singleton, D. A.; Nowlan, D. T., III.; Jahed, N.; Matyjaszewski,
K. Macromolecules 2003, 36, 8609.
Acknowledgment. We thank the NIH (Grant GM-45617)
and The Robert A. Welch Foundation for research support
and the NSF (CHE-0077917) for NMR instrumentation.
Supporting Information Available: Experimental pro-
cedures, as well as energies and geometries of all calculated
structures. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL049137A
2472
Org. Lett., Vol. 6, No. 14, 2004