Platinum-catalyzed direct amination of allylic alcohols under mild conditions: Ligand and microwave effects, substrate scope, and mechanistic study

Article abstract of DOI:10.1021/ja9046075

Transition metal-catalyzed amination of allylic compounds via a π-allylmetal intermediate is a powerful and useful method for synthesizing allylamines. Direct catalytic substitution of allylic alcohols, which forms water as the sole coproduct, has recently attracted attention for its environmental and economical advantages. Here, we describe the development of a versatile direct catalytic amination of both aryl- and alkyl-substituted allylic alcohols with various amines using Pt-Xantphos and Pt-DPEphos catalyst systems, which allows for the selective synthesis of various monoallylamines, such as the biologically active compounds Naftifine and Flunarizine, in good to high yield without need for an activator. The choice of the ligand was crucial toward achieving high catalytic activity, and we demonstrated that not only the large bite-angle but also the linker oxygen atom of the Xantphos and DPEphos ligands was highly important. In addition, microwave heating dramatically affected the catalyst activity and considerably decreased the reaction time compared with conventional heating. Furthermore, several mechanistic investigations, including ¹H and ³¹P{¹H} NMR studies; isolation and characterization of several catalytic intermediates, Pt(xantphos)Cl₂, Pt(η²-C₃H₅OH)(xantphos), etc; confirmation of the structure of [Pt(η³-allyl)(xantphos)]OTf by X-ray crystallographic analysis; and crossover experiments, suggested that formation of the π-allylplatinum complex through the elimination of water is an irreversible rate-determining step and that the other processes in the catalytic cycle are reversible, even at room temperature.

Full text of DOI:10.1021/ja9046075

Published on Web 09/17/2009
Platinum-Catalyzed Direct Amination of Allylic Alcohols under
Mild Conditions: Ligand and Microwave Effects, Substrate
Scope, and Mechanistic Study
Takashi Ohshima,* Yoshiki Miyamoto, Junji Ipposhi, Yasuhito Nakahara,
Masaru Utsunomiya, and Kazushi Mashima*
Department of Chemistry, Graduate School of Engineering Science, Osaka UniVersity,
Toyonaka, Osaka 560-8531, Japan
Received June 6, 2009; E-mail: ohshima@chem.es.osaka-u.ac.jp; mashima@chem.es.osaka-u.ac.jp
Abstract: Transition metal-catalyzed amination of allylic compounds via a π-allylmetal intermediate is a
powerful and useful method for synthesizing allylamines. Direct catalytic substitution of allylic alcohols,
which forms water as the sole coproduct, has recently attracted attention for its environmental and
economical advantages. Here, we describe the development of a versatile direct catalytic amination of
both aryl- and alkyl-substituted allylic alcohols with various amines using Pt-Xantphos and Pt-DPEphos
catalyst systems, which allows for the selective synthesis of various monoallylamines, such as the biologically
active compounds Naftifine and Flunarizine, in good to high yield without need for an activator. The choice
of the ligand was crucial toward achieving high catalytic activity, and we demonstrated that not only the
large bite-angle but also the linker oxygen atom of the Xantphos and DPEphos ligands was highly important.
In addition, microwave heating dramatically affected the catalyst activity and considerably decreased the
reaction time compared with conventional heating. Furthermore, several mechanistic investigations, including
1H and 31P{1H} NMR studies; isolation and characterization of several catalytic intermediates, Pt(xantphos)Cl2,
Pt(η2-C3H5OH)(xantphos), etc; confirmation of the structure of [Pt(η3-allyl)(xantphos)]OTf by X-ray crystal-
lographic analysis; and crossover experiments, suggested that formation of the π-allylplatinum complex
through the elimination of water is an irreversible rate-determining step and that the other processes in the
catalytic cycle are reversible, even at room temperature.
Scheme 1. Preparation of Allylamine 4 from Allylic Alcohol 1
Introduction
Transition metal-catalyzed transformation of allylic com-
pounds via a π-allylmetal intermediate is one of the most
powerful and useful methods for carbon-carbon and carbon-
heteroatom (N, O, and S) bond formation.1 It is widely utilized
for the synthesis of various natural and unnatural compounds.
The reaction usually employs activated allylic alcohol deriva-
tives 2, in particular, allylic halides, carboxylates, and carbon-
ates, because of the poor reactivity of the parent allylic alcohols
1 (Scheme 1). The use of such activated substrates, however,
causes the formation of more than stoichiometric amounts of
unwanted chemical waste in both the preactivation and the allylic
substitution steps. Thus, a direct catalytic substitution of allylic
alcohols, which forms water as the sole coproduct, has recently
attracted much attention from an environmental and economical
point of view.2
Allylamines are ubiquitous in various biologically active
compounds,3 such as the antifungal drug Naftifine4a,b and the
calcium channel blocker Flunarizine,4c,d and are highly useful
substrates for many types of reactions, such as asymmetric
isomerization5 and ring-closing metathesis.6 Direct catalytic
substitutions of allylic alcohols with amine (1 f 4) toward the
(1) For reviews, see: (a) Tsuji, J. Transition Metal Reagents and Catalysis;
Wiley-VCH: Weinheim, Germany, 2000. (b) Trost, B. M.; Lee, C. In
Catalytic Asymmetric Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-VCH:
New York, 2000. (c) Trost, B. M.; Crawley, M. L. Chem. ReV. 2003,
103, 2921.
(4) (a) Petranyi, G.; Ryder, N. S.; Stu¨tz, A. Science 1984, 224, 1239. (b)
Kanno, H.; Taylor, R. J. K. Tetrahedron Lett. 2002, 43, 7337. (c)
Olesen, J. J. Neurol 1991, 238, S23. (d) Mar´ın, M. T.; Margarit, M. V.;
Salcedo, G. E. Il Farmaco 2002, 57, 723.
(2) For review, see: (a) Muzart, J. Tetrahedron 2005, 61, 4179. (b) Muzart,
J. Eur. J. Org. Chem. 2007, 3077. (c) Tamaru, Y. Eur. J. Org. Chem.
2005, 2647.
(5) For reviews, see: (a) Akutagawa, S.; Tani, K. In Catalytic Asymmetric
Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-VCH: New York, 2000. (b)
Noyori, R. Asymmetric Catalysis in Organic Synthesis; Wiley-VCH:
New York, 1994; Chapter 3, p 95.
(3) For reviews, see: (a) Cheikh, R. B.; Chaabouni, R.; Laurent, A.; Mison,
P.; Nafti, A. Synthesis 1983, 685. (b) Johannsen, M.; Jørgensen, K. A.
Chem. ReV. 1998, 98, 1689. (c) Trost, B. M.; Crawley, M. L. Chem.
ReV. 2003, 103, 2921.
(6) For reviews, see: (a) Handbook of MetathesisGrubbs, R. H., EdWiley-
VCH: Weinheim, Germany, 2003; Vols 1-3. (b) Nicolaou, K. C.;
Bulger, P. G.; Sarlar, D. Angew. Chem., Int. Ed. 2005, 44, 4490.
9
10.1021/ja9046075 CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 14317–14328 14317

A R T I C L E S
Ohshima et al.
development of a highly efficient method of synthesizing
allylamines in an environmentally benign manner have been
studied by many research groups.2 Most of the previous catalytic
systems, however, require rather severe reaction conditions or
the addition of considerable amounts of an activator, such as
As2O3,7 B2O3,8 SnCl2,9 PPh3-DEAD,10 Ti(O-i-Pr)4,11 BPh3,12
BF3•Et2O,13 CO2,14 BEt3,15 and 1-AdCO2H,16 including an
efficient enantioselective variant,17 to increase the leaving ability
of the hydroxy group. Recently, palladium-catalyzed direct
aminations of allylic alcohols without the use of an activator
were developed by the research groups of Ozawa and Yoshi-
fuji,18 Ikariya,19 Shinokubo and Oshima,20 le Floch,21 and
Breit,22 leading to the development of a highly atom economical
synthetic process for allylamines.23 The substrate generality of
those reactions, however, remains limited. In particular, the
reaction with aliphatic primary amines results only in the
formation of the diallylation products19,20 because of the higher
nucleophilicity of the monoallylation product compared to that
of the starting substrate. In 2007, we reported a new direct
catalytic amination of allylic alcohols24 promoted by the
combination of Pt(cod)Cl2 and a large bite-angle ligand,
DPEphos25 or Xantphos.26 This platinum catalysis proceeded
with high monoallylation selectivity of both aromatic (up to
>88:1) and aliphatic (up to 30:1) primary amines. Later, le Floch
et al. reported that the reactivity of the platinum-catalyzed
amination was improved by using a catalytic system of [Pt(η3-
allyl)(dpp-xantphos)]PF6 (DPP-Xantphos: 4,5-bis(2,5-diphenyl-
1H-phosphol-1-yl)-9,9-dimethyl-9H-xanthene) and NH4PF6 (20
mol %) as an activator.27 These platinum-catalyzed reactions,
however, have much room for improvement in terms of substrate
generality of the allylic alcohol because only allyl alcohol and
R- or γ-aryl-substituted allylic alcohols afford a high yield and
the reaction of alkyl-substituted allylic alcohol results in
moderate yield because of side reactions, such as ꢀ-H elimina-
tion. Herein, we report a highly versatile direct substitution
reaction of both aryl- and alkyl-substituted allylic alcohols with
various amines catalyzed by Pt-Xantphos and Pt-DPEphos
complexes without the need for an additional activator. We
observed drastic ligand effects, and the large bite-angle as well
as the linker oxygen atom of the Xantphos and DPEphos ligands
were essential to achieve high catalytic activity. Microwave
irradiation efficiently accelerated the reaction without a loss of
selectivity, making it possible to lower the reaction temperature
to 50 °C and shorten the reaction time to 1 h for arylamines
and 2 h for alkylamines. In addition, the usefulness of the present
direct catalysis was demonstrated by a one-step synthesis of
the biologically active compounds Naftifine and Flunarizine in
95% and 94% yields, respectively. Finally, on the basis of
(7) Lu, X.; Lu, L.; Sun, J. J. Mol. Catal. 1987, 41, 245.
(8) Lu, X.; Jiang, X.; Tao, X. J. Organomet. Chem. 1988, 344, 109.
(9) (a) Masuyama, Y.; Takahara, J. P.; Kurusu, Y. J. Am. Chem. Soc.
1988, 110, 4473. (b) Masuyama, Y.; Kagawa, M.; Kurusu, Y. Chem.
Lett. 1995, 1121.
(10) Lumin, S.; Falck, J. R.; Capdevila, J.; Karara, A. Tetrahedron Lett.
1992, 33, 2091.
(11) (a) Itoh, K.; Hamaguchi, N.; Miura, M.; Nomura, M. J. Chem. Soc.,
Perkin Trans. 1 1992, 2833. (b) Satoh, T.; Ikeda, M.; Miura, M.;
Nomura, M. J. Org. Chem. 1997, 62, 4877. (c) Yang, S.-C.; Hung,
C.-W. J. Org. Chem. 1999, 64, 5000. (d) Yang, S.-C.; Tsai, Y.-C.;
Shue, Y.-J. Organometallics 2001, 20, 5326. (e) Shue, Y.-J.; Yang,
S.-C.; Lai, H.-C. Tetrahedron Lett. 2003, 44, 1481.
1
several mechanistic investigations, including H and 31P{1H}
NMR studies, isolation of several catalytic intermediates,
confirmation of the structure of [Pt(η3-allyl)(xantphos)]OTf by
X-ray crystallographic analysis, and crossover experiments, we
present a plausible catalytic cycle in which (1) all catalytic
intermediates in the cycle are in rapid equilibrium even at room
temperature; (2) among them, [Pt(η3-allyl)(xantphos)]X is the
most stable species; and (3) the formation of the π-allylplatinum
complex through the elimination of water, which is activated
by in situ generated ammonium salt, is an irreversible rate-
determining step.
(12) Stary, I.; Stara´, I. G.; Kocovsky, P. Tetrahedron Lett. 1993, 34, 179.
(13) Tsay, S.; Lin, L. C.; Furth, P. A.; Shum, C. C.; King, D. B.; Yu,
S. F.; Chen, B.; Hwu, J. R. Synthesis 1993, 329.
(14) Sakamoto, M.; Shimizu, I.; Yamamoto, A. Bull. Chem. Soc. Jpn. 1996,
69, 1065.
(15) (a) Kimura, M.; Tomizawa, T.; Horino, Y.; Tanaka, S.; Tamaru, Y.
Tetrahedron Lett. 2000, 41, 3627. (b) Kimura, M.; Horino, Y.; Mukai,
R.; Tanaka, S.; Tamaru, Y. J. Am. Chem. Soc. 2001, 123, 10401. (c)
Kimura, M.; Futamata, M.; Shibata, K.; Tamaru, Y. Chem. Commun.
2003, 234.
(16) Yang, S.-C.; Hsu, Y.-C.; Gan, K.-H. Tetrahedron 2006, 62, 3949.
(17) Yamashita, Y.; Gopalarathnam, A.; Hartwig, J. F. J. Am. Chem. Soc.
2007, 129, 7508.
(18) (a) Ozawa, F.; Okamoto, H.; Kawagishi, S.; Yamamoto, S.; Minami,
T.; Yoshifuji, M. J. Am. Chem. Soc. 2002, 124, 10968. (b) Ozawa,
F.; Ishiyama, T.; Yamamoto, S.; Kawagishi, S.; Murakami, H.;
Yoshifuji, M. Organometallics 2004, 23, 1698.
Results and Discussion
Reaction Using Arylamines. The study was initiated by
evaluating solvent effects using cinnamyl alcohol (1a) and
aniline (3a) as representative substrates at 110 °C or the solvent
reflux temperature (Table 1, entries 1-7). To clarify the solvent
(19) Kayaki, Y.; Koda, T.; Ikariya, T. J. Org. Chem. 2004, 69, 2595.
(20) Kinoshita, H.; Shinokubo, H.; Oshima, K. Org. Lett. 2004, 6, 4085.
(21) (a) Piechaczyk, O.; Doux, M.; Ricard, L.; Le Floch, P. Organometallics
2005, 24, 124. (b) Thoumazet, C.; Gru¨tzmacher, H.; Deschamps, B.;
Ricard, L.; le Floch, P. Eur. J. Inorg. Chem. 2006, 3911. (c)
Piechaczyk, O.; Thoumazet, C.; Jean, Y.; le Floch, P. J. Am. Chem.
Soc. 2006, 128, 14306. (d) Mora, G.; Deschamps, B.; van Zutphen,
S.; Le Goff, X. F.; Ricard, L.; le Floch, P. Organometallics 2007, 26,
1846.
28
effects, reactions assisted by Pt(cod)Cl2 (1.0 mol %) and
DPEphos (2.0 mol %) in various solvents were stopped after
3 h and the yield of allylamine 4aa was determined by gas
chromatographic analysis. Among the solvents examined, di-
oxane (entry 3, 72%), which was used in an earlier study,24
(22) Usui, I.; Schmidt, S.; Keller, M.; Breit, B. Org. Lett. 2008, 10, 1207.
(23) Lewis acid-catalyzed direct substitution of allylic, propargylic, and
benzylic alcohols with electron-deficient nitrogen nucleophiles were
reported. For representative examples, see: (a) Nishibayashi, Y.;
Milton, M. D.; Inada, Y.; Yoshikawa, M.; Wakiji, I.; Hidai, M.;
Uemura, S. Chem.sEur. J. 2005, 11, 1433. (b) Terrasson, V.; Marque,
S.; Georgy, M.; Campagne, J.-M.; Prim, D. AdV. Synth. Catal. 2006,
348, 2063. (c) Qin, H.; Yamagiwa, N.; Matsunaga, S.; Shibasaki, M.
Angew. Chem., Int. Ed. 2007, 46, 409. (d) Guo, S.; Song, F.; Liu, Y.
Synlett 2007, 964. (e) Reddy, C. R.; Madhavi, P. P.; Reddy, A. S.
Tetrahedron Lett. 2007, 48, 7169. (f) Sreedhar, B.; Reddy, P. S.;
Reddy, M. A.; Neelima, B.; Arundhathi, R. Tetrahedron Lett. 2007,
48, 8174. (g) Jana, U.; Maiti, S.; Biswas, S. Tetrahedron Lett. 2008,
49, 858. (h) Wang, G.-W.; Shen, Y.-B.; Wu, X.-L. Eur. J. Org. Chem.
2008, 4367. (i) Huang, W.; Shen, Q.-S.; Wang, J.-L.; Zhou, X.-G.
Chin. J. Chem. 2008, 26, 729.
(24) Utsunomiya, M.; Miyamoto, Y.; Ipposhi, J.; Ohshima, T.; Mashima,
K. Org. Lett. 2007, 9, 3371.
(25) (a) Dube, G.; Selent, D.; Taube, R. Z. Chem. 1985, 25, 154. (b)
Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M. Organometallics 1995, 14, 3081. (c) Caporali,
M.; Mu¨ller, C.; Staal, Bastiaan, B. P.; Tooke, D. M.; Spek, A. L.; van
Leeuwen, P. W. N. M. Chem. Commun. 2005, 3478.
(26) For a review, see: (a) Kamer, P. C. J.; van Leeuwen, P. W. N. M.;
Reek, J. N. H. Acc. Chem. Res. 2001, 34, 895. (b) Hillebrand, S.;
Bruckmann, J; Kru¨ger, C.; Haenel, M. W. Tetrahedron Lett. 1995,
36, 75.
(27) Mora, G.; Piechaczyk, O.; Houdard, R.; Me´zailles, N.; Le Goff, X.-
F.; le Floch, P. Chem.sEur. J. 2008, 14, 10047.
(28) See the Supporting Information for details.
9
14318 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Amination of Allylic Alcohols
A R T I C L E S
Table 1. Effects of Solvents and Reaction Temperature on
Pt-Catalyzed Direct Amination with Arylamine 3aa
Table 2. Effects of Diphosphine Ligands on Pt-Catalyzed Direct
Amination with Arylamine 3aa
entry
solvent
temp
time (h)
yield of 4aab
entry
ligand
bite-angle (ꢀ)c
yield of 4aad
1
2
3
4
5
6
7
8
9
10
toluene
110 °C
3
3
3
3
3
3
3
23
23
78
63%
38%
72%
22%
79%
63%
57%
<1%
89%
52%
1,2-dimethoxyethane reflux (82 °C)c
1
2
3
C6H4-1,2-(PPh2)2
DPPP
BINAP
SEGPHOS
6-DPPon
Trost ligand
DPEphos
87°
nde
dioxane
DMA
DMF
DMSO
d
dioxane
DMF
DMF
reflux (101 °C)c
110 °C
110 °C
110 °C
110 °C
50 °C
96°
nde
97°
nde
4
98°
nde
5f
6
104°
107°
107°
108°
108°
111°
112°
40%
19%
71%
80%
nde
7
8
9
10
11
50 °C
rt
Xantphos
POP ligand
Floriani ligand
t-Bu-Xantphos
8%
a 2.0 mmol scale, 0.4 mL of solvent was used. b Determined by GC
analysis. c Boiling point of pure solvent. d Solvent-free conditions.
nde
a 2.0 mmol scale, 0.4 mL of DMF was used. b Phosphine to platinum
c
ratio (P/Pt). Bite-angle of [Pt(η3-allyl)(diphosphine)]+ optimized with
and DMF (entry 5, 79%) gave good results, in sharp contrast
to the unsatisfactory result obtained with the closely related
solvent DMA (entry 4, 22%). Interestingly, a marked difference
between dioxane (entry 8, <1%) and DMF (entry 9, 89%) was
observed at a lower temperature (50 °C) with a prolonged
reaction time (23 h). Using DMF as the solvent, the direct
amination of 1a proceeded even at room temperature, although
the yield was moderate even after 78 h (entry 10, 52%). Further,
platinum catalysis proceeded under solvent-free conditions with
reasonably good reactivity (entry 7, 57%), presenting great
possibilities for the development of a more efficient and
environmentally friendly catalytic synthetic process of allylamines.
The great success in lowering the reaction temperature to 50
°C led us to examine the ligand effects under the optimized
conditions (Table 1, entry 9). We previously found that the large
bite-angle ligands DPEphos and Xantphos are highly effective
for platinum catalysis;24,29 therefore, we tested various large
bite-angle diphosphine ligands (Table 2). Due to the lack of
information about the bite-angles of Pt(η3-allyl)(diphosphine)
complexes, the bite-angles (ꢀ) shown in Table 2 were obtained
by geometry optimization using the B3LYP30 hybrid-density
functional theory (LANL2DZ for Pt and 6-31G(d,p) for oth-
ers).28 Although 1,2-bis(diphenylphosphino)benzene (ꢀ ) 87°),
DPPP (ꢀ ) 96°), BINAP (ꢀ ) 97°), and SEGPHOS (ꢀ ) 98°)
ligands, whose bite angles are smaller than 100°, did not promote
the reaction (entries 1-4), larger bite-angle ligands such as
6-DPPon,22 which in situ dimerizes in aprotic solvent through
hydrogen bonding and acts as a bidentate ligand (ꢀ ) 104°),31
and Trost ligand (ꢀ ) 107°),32 showed moderate reactivities to
afford allylamine 4aa in 40% (entry 5) and 19% (entry 6) yields,
B3LYP function. d Determined by GC analysis. e Not detected in the
reaction mixture. f 4 mol % of ligand was used to keep P/Pt to 4.
respectively. The use of DPEphos (ꢀ ) 107°)25 and Xantphos
(ꢀ ) 108°)26 greatly improved catalyst activity with Xantphos
giving the best result (entry 8, 80% yield). On the other hand,
the use of the POP ligand (ꢀ ) 108°)33 and t-Bu-Xantphos (ꢀ
) 112°)34 afforded no product (entries 9 and 11) despite their
large
bite-angle.
These
findings
clearly
indi-
cated that, in addition to a large bite-angle, an efficient ligand
should be composed of fully aryl- or alkenyl27-substituted
phosphines. Moreover, the very low reactivity when using the
Floriani ligand (ꢀ ) 111°)35 (entry 10), which is a methylene-
linked variant of DPEphos, strongly suggests the importance
of the linker oxygen in DPEphos and Xantphos. The effect of
the linker oxygen is discussed in the last section of this
manuscript.
(29) For reviews on the influence of the bite-angle of bidentate phosphane
ligands, see: (a) van Haaren, R. J.; Oevering, H.; Coussens, B. B.;
van Strijdonck, G. P. F.; Reek, J. N. H.; Kamer, P. C. J.; van Leeuwen,
P. W. N. M. Eur. J. Inorg. Chem. 1999, 1237. (b) Oestreich, M. Eur.
J. Org. Chem. 2005, 783. (c) Johns, A. M.; Utsunomiya, M.; Incarvito,
C. D.; Hartwig, J. F. J. Am. Chem. Soc. 2006, 128, 1828.
(30) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(33) (a) Sacconi, L.; Gelsomini, J. Znorg. Chem. 1968, 7, 291. (b) Greene,
P. T.; Sacconi, L. J. Chem. Soc. A 1970, 866. (c) Dapporto, P.; Sacconi,
L. J. Am. Chem. Soc. 1970, 92, 4133. (d) Dapporto, P.; Sacconi, L.
J. Chem. Soc. A 1971, 1914. (e) Thewissen, D. H. M. W.; Timmer,
K.; Noltes, J. G.; Marsman, J. W.; Laine, R. M. Inorg. Chim. Acta
1985, 97, 143.
(31) For a review of self-assembling ligand, see: (a) Breit, B. Angew. Chem.,
Int. Ed 2005, 44, 6816. (b) Breit, B.; Seiche, W. J. Am. Chem. Soc.
2003, 125, 6608. (c) Chevallier, F.; Breit, B. Angew. Chem., Int. Ed.
2006, 45, 1599. (d) Weis, M.; Waloch, C.; Seiche, W.; Breit, B. Angew.
Chem., Int. Ed. 2007, 46, 3037. (e) Smejkal, T.; Breit, B. Organo-
metallics 2007, 26, 2461.
(34) Mispelaere-Canivet, C.; Spindler, J.-F.; Perrio, S.; Beslin, P. Tetra-
hedron 2005, 61, 5253.
(35) Lesueur, W.; Solari, E.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. Inorg.
Chem. 1997, 36, 3354. Structural similarities between DPEphos and
Floriani ligand were observed not only in computationally optimized
structures of Pt(η3-allyl)(diphosphine) complexes shown in Table 2
but also in crystal structure of Pd(diphosphine)Cl2 complexes (see also
ref 25c).
(32) For a review, see: (a) Trost, B. M.; Machacek, M. R.; Aponick, A.
Acc. Chem. Res. 2006, 39, 747. (b) Trost, B. M.; Van Vranken, D. L.
Angew. Chem., Int. Ed. Engl. 1992, 31, 228.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14319

A R T I C L E S
Ohshima et al.
Scheme 2. Effects of Microwave Irradiation on Pt-Catalyzed Direct
Amination
Table 3. Effects of Ligand to Platinum Ratio on Pt-Catalyzed
Direct Amination with Arylamine 3ba
entry Pt(cod)Cl2x (mol %) Xantphos y (mol %) yield of 4abb yield of 5abb TONc
1
2
3
4
1.0
1.0
1.0
1.0
0.5
0.1
3.0
2.0
1.5
1.0
0.5
0.1
ndd
ndd
5%
8%
8%
7.5%
<1%
76%
80%
84%
85%
57%
86
96
100
200
570
Organic synthesis assisted by microwave heating has grown
rapidly since the pioneering work of Gedye in 1986.36 In many
organic transformations, microwave heating dramatically de-
creases reaction times and increases product yields in compari-
son with conventional heating methods; therefore, we examined
the reaction under microwave irradiation conditions (Scheme
2). Using a CEM Discover single-mode microwave reactor, the
reaction mixture was heated at 50 °C and full conversion was
achieved within 1 h. Conventional heating, however, required
4 h for completion. Reaction temperature was monitored using
a standard IR temperature monitoring system and controlled by
microwave irradiation power (standard mode) with an air flow
cooling system (20 psi). When polar DMF was used as the
solvent, 5-6 W of power was sufficient to maintain the reaction
temperature at 50 °C. Although the microwave reactor included
a standard IR temperature monitoring system, the reaction
temperature was also checked with an alcohol thermometer
immediately after turning off the irradiation,37 revealing that
the error in the IR temperature monitoring system was within
5 °C. These results suggest the existence of nonthermal
microwave effects.
5
6e
a 2.0 mmol scale, 0.4 mL of DMF was used. b Isolated yield.
c Turnover number of the platinum catalyst. d Not detected in the
reaction mixture. e 4.0 mmol scale, 0.8 mL of DMF was used.
Table 4. Pt-Catalyzed Direct Amination with Various Arylamines 3a
Finally, optimization of the ligand to platinum ratio was
performed using para-methoxyaniline (3b) under microwave
irradiation conditions. As shown in Table 3, the presence of
the extra ligand prevented the reaction (entry 1), and optimal
results were obtained with a Xantphos to platinum ratio of 1:1
(entry 4), showing that the active catalyst species would be a
coordinately unsaturated Pt-xantphos 1:1 complex. Under the
optimal reaction conditions, the use of only 0.5 mol % of the
catalyst completed the reaction within 1 h (entry 5), and a
turnover number of 570 was achieved when the catalyst loading
was reduced to 0.1 mol % (entry 6).
We then investigated the scope and limitations of different
arylamines 3 (Table 4). Under our previously reported reaction
conditions (1.0 mol % of Pt(cod)Cl2, 2.0 mol % of DPEphos,
dioxane reflux), the reaction of anilines bearing an electron-
withdrawing group gave only moderate yield due to their low
nucleophilicity. For example, 4-fluoroaniline (3c) and 4-chlo-
roaniline (3d) gave the corresponding monoallylation products
4ac and 4ad in 55% and 41% yield, respectively, even after
42 h. Under the present optimized conditions, the reactions of
aniline (3a) and various para-substituted anilines 3c-h with
electron-withdrawing groups were efficiently promoted by only
a 2.0 mmol scale, 0.4 mL of DMF was used. b Isolated yield. c Not
detected in the reaction mixture d Reaction time was 2 h.
0.5 mol % of the catalyst to provide the desired monoallylation
product 4 in high yield and high monoallylation selectivity
(entries 1-7). Reaction of para-iodoaniline (3f) was the only
ineffective case with 3% of the product 3af and recovery of
the unreacted 3f, probably due to oxidative addition of 3f onto
the Pt(0) complex (entry 5). Sterically more congested ortho-
chloroaniline (3i) prevented the diallylation reaction and
(36) For recent reviews, see: (a) Kappe, C. O. Angew. Chem., Int. Ed. 2004,
43, 6250. (b) Hayes, B. L. Aldrichimica Acta 2004, 37, 66. (c) Roberts,
B. A.; Strauss, C. R. Acc. Chem. Res. 2005, 38, 653. (d) Herrero,
M. A.; Kremsner, J. M; Kappe, C. O. J. Org. Chem. 2008, 73, 36.
(37) Because traditional techniques for temperature measurements cannot
be applied in microwave heating, the reaction temperature was
measured using an alcohol thermometer immediately after turning off
the microwave irradiation.
9
14320 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Amination of Allylic Alcohols
A R T I C L E S
Scheme 3. Pt-Catalyzed Direct Amination with Alkylamine 6a in
DMF
Table 6. Effects of Ligand to Platinum Ratio on Pt-Catalyzed
Direct Amination with Alkylamine 6ba
entry
ligand
x (mol %)
heating
time (h) yield of 7abb yield of 8abb
Table 5. Effects of Solvents on Pt-Catalyzed Direct Amination with
Alkylamine 6aa
1
2
3
4
5
6
7
8
Xantphos
Xantphos
DPEphos
DPEphos
Xantphos
Xantphos
DPEphos
DPEphos
1.0
2.0
1.0
2.0
1.0
2.0
1.0
2.0
microwave
microwave
microwave
microwave
conventional
conventional
conventional
conventional
4
4
81%
13%
32%
85%
10%
ndc
4%
ndc
ndc
3%
ndc
ndc
ndc
ndc
3
3
24
24
24
24
trace
16
entry
solvent
yield of 7aab
1
2
3
4
5
toluene
1,2-dichloroethane
CHCl3
DMSO
CH3CN
19%
35%
19%
15%
14%
a 2.0 mmol scale, 0.4 mL of toluene was used. b Isolated yield. c Not
detected in the reaction mixture.
Scheme 4. Direct Amination with Alkylamine 6b using
Pt(xantphos)2 and Pt(dpephos)2
a 2.0 mmol scale, 0.4 mL of solvent was used. b Determined by GC
analysis.
predominantly afforded the monoallylation product 4ai (entry
7, 96%). Moreover, secondary amine 3j was also a good
substrate for this transformation (entry 8, 96%).
Reaction Using Alkylamines. We next examined the direct
substitution of allylic alcohol with alkylamines, which is more
challenging. When benzylamine (6a) was used as a nucleophile,
the conditions optimized for arylamines resulted in an unsat-
isfactory low yield of 7aa (34%) and 8aa (16%) along with
unexpected byproduct 9, which is an allylation product of
dimethylamine (Scheme 3). In the reaction of arylamines, this
byproduct 9 was not detected at all, suggesting that dimethy-
lamine would not be generated by simple thermal decomposition
of DMF but was derived by platinum-catalyzed transamidation
of DMF in the presence of highly nucleophilic benzylamine
(6a).
To avoid such an undesirable side reaction, we re-examined
the direct amination of 1a using benzylamine (6a) as the
nucleophile in several polar and nonpolar solvents other than
DMF (Table 5). Under the standard-mode microwave irradiation
conditions (70 °C, ∼60 W), the reactions in all tested solvents
resulted in full conversion within 1 h, making it difficult to
precisely evaluate the solvent effects. Thus, the maximum
irradiation power was decreased to 15 W with minimal air flow
cooling to keep the reaction temperature at 70 °C. Among the
tested solvents, 1,2-dicloroethane (entry 2) was best. In the case
of chlorinated solvents (entries 2 and 3), however, the HCl salt
of 6a was formed, indicating that the chlorinated solvents
decomposed under the reaction conditions. Thus, we used
toluene (entry 1) as the solvent for further investigation of the
reaction with alkylamines. For practical operation to keep the
reaction temperature at 50 °C, the toluene solvent conditions
required 50-60 W of power (standard mode) with standard air
flow cooling.
decomposition of the Pt(0) species.38 To stabilize the Pt(0)
species in the reaction mixture, we performed the reaction with
an extra amount of the ligand (2.0 equiv to platinum), although
we were concerned that it would generate a less active
18-electron Pt(0) complex Pt(diphosphine)2 (Table 6). Surpris-
ingly, in spite of the structural similarities of Xantphos and
DPEphos, each ligand gave completely different results (entries
1-4). In the case of Xantphos, the use of 2.0 equiv of ligand
to platinum substantially prevented the amination reaction (entry
2, 13%) with the formation of yellow precipitate, which was
the Pt(xantphos)2 complex. The reason for the low reactivity
was attributed to the insolubility of this complex in toluene.39
Moreover, this isolated 18-electron Pt(0) complex did not
promote the reaction of 1a with 6b at all (Scheme 4). On the
other hand, 2.0 equiv of DPEphos to platinum greatly improved
the yield from 32% (entry 3) to 85% (entry 4) without
decomposition of the platinum complex. These results clearly
indicate the high solubility of the in situ generated Pt(dpephos)2
complex and the easy dissociation of one DPEphos ligand or
monodissociation of the bidentate bisphosphine from the 18-
electron Pt(dpephos)2 complex to yield the coordinately unsatur-
ated active Pt(0) complex. In fact, the Pt(dpephos)2 complex,
which was prepared by the NaBH4 reduction of the Pt(dpep-
(38) Because the reaction with a less congested alkylamine such as 6a
formed more black precipitates than that with a more congested
alkylamine such as 6b, decomposition of the Pt(0) species may be
induced by the coordination of alkylamine to platinum. See also the
Supporting Information.
(39) Pt(xantphos)2 complex, synthesized by the reaction of Pt(dba)2 with
the Xantphos ligand (4.0 equiv), was almost insoluble, not only in
toluene but also in other solvents such as CHCl3, THF, and DMF.
Even after 3 h irradiation in toluene, direct amination of 1a
with 6a resulted in a moderate yield of 7aa (58% yield)
accompanied by the formation of black precipitates due to partial
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14321

A R T I C L E S
Ohshima et al.
Table 7. Pt-Catalyzed Direct Amination with Various Alkylamines
Substrate Scope of Allylic Alcohols. Direct amination of alkyl-
substituted allylic alcohols remained an unsolved target. Because
microwave irradiation in DMF greatly improved the catalyst
activity and lowered the reaction temperature to 50 °C, we
anticipated that under such mild conditions the reaction of the
alkyl-substituted allylic alcohols could be catalyzed without
undergoing side reactions such as ꢀ-H elimination. Under the
optimized conditions for arylamine (1.0 mol % of Pt(cod)Cl2,
1.0 mol % of Xantphos, DMF, microwave irradiation, 60 °C,
standard mode), we performed direct amination of ally alcohol
(1b) and various aryl- and alkyl-substituted allylic alcohols 1a-h
with aniline (3a) (1.5 equiv) (Table 8). These studies also
provided several new insights into the mechanism. The reaction
of 1b (entry 1) proceeded smoothly in a manner similar to
γ-phenyl-substituted allylic alcohol 1a (entry 2). Despite the
moderate yield of γ-methyl-substituted 1c under the previous
reaction conditions (43% yield, dioxane, reflux, 18 h), the
present reaction conditions allowed for the full conversion of
1c to the allylated products 4ca (77%), 4ca′ (10%), and 5ca
(6%) without proceeding to the undesired ꢀ-H elimination (entry
4). When R-substituted allylic alcohols 1a′ (entry 3) and 1c′
(entry 5) were used, the obtained regio-, stereo-, and monoal-
lylation selectivities were almost the same as those using
γ-substituted allylic alcohols 1a (entry 2) and 1c (entry 4),
respectively, strongly suggesting that the reaction proceeds via
π-allylplatinum intermediates. ꢀ-Methyl-substituted allylic al-
cohol 1d (entry 6), cyclic allylic alcohol 1f (entry 9), and R,R-
disubstituted allylic alcohol 1g′ (entry 12) were also good
substrates for the direct catalytic amination reaction to give the
desired products in high yield. On the other hand, the reactions
of R,γ-disubstituted allylic alcohol 1e (entry 7) and γ,γ-
disubstituted allylic alcohols 1g (entry 10) and 1h (entry 13)
resulted in low to moderate yield with partial decomposition of
the platinum species. Fortunately, conventional heating instead
of microwave irradiation greatly improved the yield of 4ea (entry
8, 91%), 4ga (entry 11, 85%), and 4ha (entry 14, 70%), though
these reactions required a longer reaction time (20-48 h).28
Based on the fact that R,R-disubstituted allylic alcohol 1g′ reacts
much faster than γ,γ-disubstituted allylic alcohol 1g, the
reactivity of the platinum catalysis is more strongly affected
by steric congestion around the CdC double bond in allylic
alcohol than that around the hydroxy group, showing the
importance of η2-coordination of the CdC double bond to the
platinum atom prior to the generation of the π-allyl intermedi-
ates. It is worth noting that all reactions of various allylic
alcohols proceeded in quite high monoallylation selectivity. To
the best of our knowledge, this is the first time that such broad
substrate generality (both aryl- and alkyl-substituted allylic
alcohols and both arylamines and alkylamines) and high
monoallylation selectivity were achieved in this type of direct
reaction.
6a
a 2.0 mmol scale, 0.4 mL of toluene was used. b Isolated yield. c 3.0
equiv of alkylamine 6 was used. d Solvent-free conditions.
hos)Cl2 complex with the DPEphos ligand, also catalyzed the
reaction in moderate yield (Scheme 4). The same tendencies as
shown in entries 1-4 were observed in the conventional heating
reactions (entries 5-8), but the obtained yields of the product
7ab were quite low even after 24 h, again suggesting the
existence of nonthermal microwave effects.
With the efficient reaction conditions for alkylamines (Table
6, entry 4) in hand, we performed the reaction with various
primary and secondary alkylamines (Table 7). Although 3.0
equiv of amine was used when sterically less congested primary
alkylamines were employed (entries 1-5), monoallylation
products 7 were obtained in high yield along with the formation
of only small amounts of diallylation products 8 (entries 1-8).
To our knowledge, such high monoallylation selectivity was
only achieved by platinum catalysts bearing large bite-angle
diphosphine ligands.24,27 The reactions proceeded under solvent-
free conditions (entries 3 and 5), resulting in high volumetric
productivity and low waste generation. Moreover, direct ami-
nation with secondary alkylamines 6h-k proceeded quite
efficiently to afford the desired products 7ah-ak, including the
antifungal drug Naftifine (7aj)4a,b and the calcium channel
blocker Flunarizine (7ak),4c,d in excellent yields (entries 9-12).
As compared with the previously reported conventional heating
conditions in dioxane (reflux 101 °C, 18 h),24 we succeeded in
lowering the reaction temperature to 50-60 °C and decreasing
the reaction time to 2-6 h while maintaining a high yield and
high monoallylation selectivity.
Mechanistic Studies. Based on the following intensive
1
mechanistic studies, including H and 31P{1H} NMR studies,
the isolation of several catalytic intermediates, confirmation of
the structure of [Pt(η3-allyl)(xantphos)]OTf by X-ray crystal-
lographic analysis, and crossover experiments, a plausible
catalytic cycle of the platinum-catalyzed direct amination of
allylic alcohol is depicted in Scheme 5.
1. Generation of Pt(0) Species from Pt(cod)Cl2 (A f B).
When Pt(0) species, such as Pt(PPh3)4 and Pt(dba)2, were used
as a catalyst precursor, only low to moderate catalyst activities
were observed28 probably due to the existence of extra ligand,
which would prevent the formation of coordinatively unsaturated
9
14322 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Amination of Allylic Alcohols
A R T I C L E S
Table 8. Pt-Catalyzed Direct Amination of Various Allylic Alcohols 1a
a 2.0 mmol scale, 0.4 mL of DMF was used. b Isolated yield. c Reaction temperature was 50 °C. d 0.5 mol % of Pt(cod)Cl2 and 0.5 mol % of
Xantphos were used. e E/Z mixture of 1c (E/Z ) 97/3) was used. f 4ca was obtained as an E/Z mixture (10/1). g Reaction temperature was 80 °C.
h Conventional heating conditions. i 4ha was obtained as an E/Z mixture (5.8/1). j Not detected in the reaction mixture.
Scheme 5. Proposed Catalytic Cycle of the Platinum-Catalyzed Direct Amination of Allylic Alcohola,b
a Substituents of allylic alcohol 1 are omitted for clarity. b Structurally characterized complexes are boxed.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14323

A R T I C L E S
Ohshima et al.
Scheme 6. Reaction of Pt(cod)Cl2 with DPEphos, Amine, and
Allylic Alcohol
Scheme 7. Preparation of Pt(xantphos)Cl2 (A),
Pt(η2-C3H5OH)(xantphos) (B), and [Pt(η3-allyl)(xantphos)]OTf (DOTf
)
active Pt(0) species. In sharp contrast, Pt(cod)Cl2 was found to
be the best precursor.28 Our first question to be addressed was
which compound reduced Pt(II) to Pt(0): phosphine ligand,
amine, or allylic alcohol. Phosphine ligand can reduce Pd(OAc)2
with water and amine to generate Pd(0) species and phosphine
oxide. Thus, we first examined this possibility; a mixture of
Pt(cod)Cl2, 2.0 equiv DPEphos, and water in DMF was heated
at 110 °C in the presence or absence of aniline (3a) (Scheme
6). In both cases, however, only the formation of Pt(dpephos)Cl2
(A) was observed with the remaining extra DPEphos ligand,
and neither the Pt(0) species nor phosphine oxide was detected.
In contrast, when Pt(cod)Cl2 and DPEphos were heated in the
presence of excess cinnamyl alcohol (1a), trace amounts of
cinnamaldehyde (9a) and dicinnamyl ether (10a), which may
be produced by nucleophilic addition of 1a to the π-allylplati-
num complex, were detected. The addition of Et3N significantly
increased the yield of these products (9a: 33%, 10a: 34%).
Although only a small amount of Pt(0) species, assignable to
Pt(η2-PhC3H4OH)(xantphos) (B) (vide infra), was detected by
31P{1H} NMR, the results of the control reactions shown in
Scheme 6 suggest that the Pt(II) species is reduced to the Pt(0)
species by allylic alcohol 1 through ꢀ-H elimination, which is
different from Wacker-type mechanism reported by Atwood et
al.40 We later found that we observed only a small proportion
of Pt(0) species because the Pt(0) is easily reoxidized to the
Pt(II) species in the presence of a chloride ligand (vide infra).
2. Syntheses, Characterization, and Catalyst Activities of Pt
Complexes Pt(η2-C3H5OH)(xantphos) (B) and [Pt(η3-allyl)(xant-
phos)]X (D). We next examined the synthesis of putative
intermediates Pt(η2-C3H5OH)(L) (B) and [Pt(η3-allyl)(L)]X (D)
where L is DPEphos or Xantphos. The results using Xantphos
are summarized in Scheme 7.41 Reaction of Pt(cod)Cl2 with 1.0
equiv Xantphos in CHCl3 at room temperature led to the
formation of a Pt(II) complex, Pt(xantphos)Cl2 (A) (31P{1H}
1
NMR δ 6.60, s with Pt satellite, JP-Pt ) 3694 Hz), in 97%
yield. NaBH4 reduction of A in the presence of allyl alcohol
(1b) afforded a Pt(0) complex, Pt(η2-C3H5OH)(xantphos) (B),
in 94% yield. In the 31P{1H} NMR spectrum, two different
phosphorus signals appeared as two doublets with Pt satellites
2
1
at δ 19.8 (d with Pt satellite, JP-P ) 51.4 Hz, JP-Pt ) 3687
2
1
Hz) and 18.1 (d with Pt satellite, JP-P ) 51.4 Hz, JP-Pt
)
3508 Hz), suggesting the presence of two magnetically different
phosphorus atoms. Moreover, significant high field shifts of
alkenyl protons in 1H NMR [δ 3.43-3.33 (m, 1 H, CH2dCH-
CH2), 2.61-2.50 (m, 1 H, CHHdCH-CH2), 2.50-2.34 (m, 1
H, CHHdCH-CH2)] also support a η2-coordination of 1b to
platinum metal. A π-allyl complex, [Pt(η3-allyl)(xantphos)]OTf
(DOTf), was synthesized in 84% yield by the reaction of
Pt(xantphos)Cl2, allyltributylstannane (1.0 equiv), and AgOTf
(1.0 equiv) and identified by 31P{1H} NMR (δ 1.24 ppm, 1JP-Pt
1
) 4179 Hz), H NMR, 19F NMR, FAB Mass, and elemental
analysis. The structure of this complex was further confirmed
by X-ray crystallographic analysis (Figure 1) of crystals, which
were produced by slow diffusion of hexane into the saturated
CH2Cl2 solution of the complex at room temperature. The bite-
angle of the crystal (107°) is almost the same as that obtained
by DFT calculations (Table 2, 108°).
Having putative intermediates Pt(η2-C3H5OH)(xantphos) (B)
and [Pt(η3-allyl)(xantphos)]OTf (DOTf) in hand, the catalyst
activities of these complexes were investigated in the reaction
of 1b and 3a. As shown in Scheme 8, both complexes promoted
the reaction to give the product, suggesting that these complexes
are involved in the catalytic cycle. Reactivities of these
complexes, however, differed from that of the
Pt(cod)Cl2-Xantphos system. Under conventional heating
conditions at 50 °C, the Pt(0) complex B required 24 h to reach
full conversion, while the reactions using Pt(cod)Cl2-Xantphos
and DOTf completed in 10 and 2 h, respectively. Higher catalyst
(40) Helfer, D. S.; Phaho, D. S.; Atwood, J. D. Organometallics 2006, 25,
410.
(41) In a similar way, Pt(dpephos)Cl2 complex A (31P{1H} NMR δ 1.97,
1
s with Pt satellite, JP-Pt ) 3796 Hz) and Pt(η2-C3H5OH)(dpephos)
2
complex B (31P{1H} NMR δ 25.7 (d with Pt satellite, JP-P ) 46.6
Hz, 1JP-Pt ) 3596 Hz), 24.9 (d with Pt satellite, 2JP-P ) 46.6 Hz, 1JP-Pt
) 3596 Hz)) were synthesized in 89% and 58% yields, respectively.
9
14324 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Amination of Allylic Alcohols
A R T I C L E S
Scheme 9. Generation of [Pt(η3-allyl)(xantphos)]X (D) from
Pt(η2-C3H5OH)(xantphos) (B)
E f D, Scheme 5, counterclockwise direction). The formation
of 3a could not be confirmed due to overlapping of the signals
of the aromatic protons in H NMR.
Figure 1. ORTEP drawing of [Pt(η3-allyl)(xantphos)]OTf (DOTf) with
thermal ellipsoids at the 50% probability level. The hydrogen atoms,
counteranions, and solvent are omitted for clarity. Selected bond lengths
(Å) and angles (°): Pt1-P1 2.3302(19), Pt1-P2 2.325(2), Pt1-C1 2.195(8),
Pt1-C2 2.142(9), Pt1-C3 2.182(8), Pt1-O1 3.506(4), C1-C2 1.336(14),
C2-C31.344(15);P1-Pt1-P2107.04(6),C2-Pt1-P1123.6(3),C2-Pt1-P2
122.3(3).
1
The existence of this rapid retro reaction (C f F f E f D)
was further supported by the following crossover experiments
(Scheme 10). We performed direct amination of cinnamyl
alcohol (1a) with 4-methoxyaniline (3b) in the presence of
N-allylaniline (4ba). Employing both condition A (Pt(cod)Cl2-
Xantphos) and condition B ([Pt(η3-allyl)(xantphos)]OTf), we
detected not only simple coupling product 4ab but also
significant amounts of crossover products 4aa and 4bb. Thus,
it was confirmed that the retro reaction of complex C, giving
complex D, 1a, and 3a, proceeded under the reaction conditions.
4. Reactivity Performance of [Pt(η3-allyl)(xantphos)]OTf
(DOTf) (D f E f F f C f D). To obtain more insight into the
mechanism, we next examined the reactivity of complex DOTf
Scheme 8. Catalytic Activities of Pt(η2-C3H5OH)(xantphos) (B) and
[Pt(η3-allyl)(xantphos)]OTf (DOTf
)
1
toward nucleophile aniline (3a) (Scheme 11). H and 31P{1H}
NMR spectroscopic monitoring of a mixture of DOTf and 12.0
equiv of 3a in DMF-d7 indicated that almost no reaction
occurred even after 6 h of heating at 50 °C. In contrast, when
allyl alcohol (1b) was added to this mixture with 3a, clear
conversion of 1b to the allylated products 4ba (62%) and 5ba
(38%) was detected by 1H NMR, though only the starting DOTf
was detected on 31P{1H} NMR spectroscopy. Recently, le Floch
and coauthors reported that, in the case of [Pt(η3-allyl)(dpp-
xantphos)]PF6, addition of benzylamine (6a) led to the formation
activities of Pt(cod)Cl2-Xantphos and DOTf than B may be
attributed to the generation of ammonium salts H2NR1 ·HCl and
H2NR1 ·HOTf, which could potentially act as more efficient
activators of the hydroxyl group to accelerate the elimination
of hydroxide through a hydrogen bonding of ammonium salts
with hydroxy group of allylic alcohol (C′ f D).
of
a
new complex, tentatively assigned as Pt(η2-
H2CdCHCH2NH2Bn+PF6 )(dpp-xantphos) (E).27 In sharp con-
trast to their results, our results shown in Schemes 9-11 suggest
that (1) complexes D, E, F, and C were in rapid equilibrium,
even at room temperature, (2) among them, D was the most
stable species even in the presence of excess amounts of
nucleophile 3, and (3) the formation of D from C through the
elimination of water (C f D) was the rate-determining step of
this catalytic cycle and required a reaction temperature of 50
°C.
-
3. Generation of [Pt(η3-allyl)(xantphos)]X (D) from Pt(η2-
C3H5OH)(xantphos) (B) (B f C f D). To elucidate the formation
of the π-allylplatinum complex D from the Pt(0) complex Pt(η2-
C3H5OH)(xantphos) (B), HOTf and HCl salts of N-allylaniline
(4ba) were added to a solution of B in DMF-d7, and these
reactions were monitored by 1H and 31P{1H} NMR spec-
troscopies (Scheme 9). The 31P{1H} NMR spectrum indicated
that the addition of 2.0 equiv 4ba·HOTf salt to the Pt(0)
complex B at room temperature led to the exclusive formation
of [Pt(η3-allyl)(xantphos)]OTf (DOTf) within 5 min (Figure 2).
Similarly, full conversion of B to [Pt(η3-allyl)(xantphos)]Cl (DCl)
was achieved within 5 min by the addition of 2.0 equiv 4ba·HCl
(Figure 3). Their 1H NMR spectra, however, revealed the
presence of free allyl alcohol (1b), but not free N-allylaniline
(4ba) or water, strongly suggesting that π-allylplatinum com-
plexes D were not generated through the elimination of water
from allyl alcohol (1b) coordinated to the Pt(0) atom (C f D,
Scheme 5, clockwise direction) but rationally through the
elimination of aniline 3a from N-allylaniline (4ba) (C f F f
Finally, we investigated the possibility of the retro reaction
of DOTf to C (D f C, Scheme 5, counterclockwise direction).
Treatment of D with allylamine 4aa and water in DMF-d7 at
60 °C for 20 h resulted in no reaction, probably due to the low
nucleophilicity of water to the π-allylplatinum complex D. Based
on these facts, we concluded that the formation of complex D
from complex C through the elimination of water is an
irreversible rate-determining step.
5. Effects of Counteranion (X-) on Stability of Pt(η3-allyl)(x-
antphos)]X (D). When we examined the formation of [Pt(η3-
allyl)(xantphos)]Cl (DCl) from Pt(η2-C3H5OH)(xantphos) (B) by
the addition of the HCl salt of N-allylaniline (4ba) (Figure 3),
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14325

A R T I C L E S
Ohshima et al.
Figure 2. 31P{1H} NMR of platinum complexes in DMF-d7: (a) Pt(η2-C3H5OH)(xantphos) (B) and (b) B + 4ba·HOTf (2 equiv), after 5 min.
Figure 3. 31P{1H} NMR of platinum complexes in DMF-d7: (a) Pt(η2-C3H5OH)(xantphos) (B); (b) B + 4ba·HCl (2 equiv), after 5 min, (c) after 3 h, (d)
after 45 h.
we found that [Pt(η3-allyl)(xantphos)]Cl (DCl) was slowly
converted to Pt(xantphos)Cl2 (A) and the ratio of complexes
DCl and A reached 10:1 after 45 h at 50 °C, even under
thoroughly degassed conditions. Such an oxidation reaction42
was not observed when 4ba·HOTf was added to the complex
B. In addition, 31P{1H} NMR measurement in DMF-d7 revealed
that the addition of 4ba·HCl (2.0 equiv) gradually converted
DOTf to Pt(xantphos)Cl2 (A) (DOTf: A ) 25:75, after 18 h of
heating at 50 °C). As shown in Scheme 6, the reduction of
Pt(cod)Cl2 with Xantphos by allylic alcohol 1a resulted in the
formation of only 4% of Pt(0) species together with 96% of
Pt(II) complex A, despite the fact that 1a was oxidized to the
corresponding aldehyde in 33%. This is due to the easy
reoxidation property of DCl. The resulting Pt(II) species again
requires another equivalent of allylic alcohol 1a to participate
in the catalytic cycle. These results, taken together with the fact
that DOTf has higher catalyst activity than the
Pt(cod)Cl2-Xantphos system (Scheme 8), indicate that the use
of DOTf leads to the development of a more efficient catalytic
system.
6. Further Investigation of Ligand Effects of DPEphos
and Xantphos. As far as we examined, only platinum complexes
bearing a diphosphine ligand with a large bite-angle have
efficient catalytic activity. Although the high monoallylation
selectivity of Pt-DPEphos and Pt-Xantphos complexes can be
explained by their large bite-angle as previously proposed,24 it
is difficult to understand their high catalyst activities on direct
amination simply based on their large bite-angle. On the basis
of DFT calculations, le Floch and coauthors proposed that
ligands with a large bite-angle effectively prevent the formation
of platinum hydride complexes and therefore the formation of
an inactive bicyclic aminopropyl complex.27 The results shown
in Table 2, however, clearly demonstrated that not only the large
bite-angle but also the linker oxygen in DPEphos and Xantphos
are very important for achieving the high catalyst performance.
Based on the X-ray crystallographic analysis of DOTf (Figure
1), there is no direct interaction between the linker oxygen atom
(42) Although the precise mechanism is not clear, the oxidation of [Pt(η3-
allyl)(xantphos)]Cl (DCl) or complex C by the molecular oxygen
remaining even after degassing may lead to the formation of
Pt(xantphos)Cl2 (A).
9
14326 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Amination of Allylic Alcohols
A R T I C L E S
Scheme 10. Crossover Experimentsa
Table 9. Pt-catalyzed Direct Amination Using Various
Monophosphine Ligandsa
entry
ligand
yield of 4aac
1
2
3
4
5
6
7
8
9
PPh3
P(C6H4-2-Me)Ph2
P(C6H4-4-Me)Ph2
P(C6H4-2-OMe)Ph2
P(C6H4-3-OMe)Ph2
P(C6H4-4-OMe)Ph2
P(C6H3-2,6-(OMe)2)Ph2
P(C6H4-2-CH2OMe)Ph2
P(C6H4-2-NMe2)Ph2
41%
ndd
18%
69%
33%
11%
45%
2%
ndd
a 2.0 mmol scale, 0.4 mL of DMF was used. b Phosphine to platinum
ratio (P/Pt). c Determined by GC analysis. d Not detected in the reaction
mixture.
substituted ligands led to lower yields (entries 2 and 3). On the
other hand, ortho-methoxy substituted ligand P(C6H4-2-
OMe)Ph2 afforded the product 4aa in 69% yield (entry 4),
whereas, very interestingly, meta- and para-methoxy substituted
ligands decreased the yield to 33% (entry 5) and 11% (entry
6), respectively.44 These results indicate that catalyst activity
does not directly correlate with the electronic effects of the ligand.
Introducing one more methoxy group at the other ortho-position
somewhat decreased the yield to 45% (entry 7), and replacing a
methoxy group with a methoxymethyl group (entry 8) and
dimethylamino group (entry 9) resulted in almost no reaction. The
exact reason for such dramatic positive effects of the oxygen atom
at the ortho-position of triarylphosphine ligands on catalyst activity
remains unclear, but the effects may be due to an increase in the
leaving ability of the hydroxy group through hydrogen bonding.
These findings provide highly useful information for the develop-
ment of further efficient ligands including chiral variants.
a Results under condition A (left) and condition B (right).
Scheme 11. Reactivity of [Pt(η3-allyl)(xantphos)]OTf (DOTf) with
Amine and Allylic Alcohol
Conclusion
of Xantphos and the platinum atom (Pt1-O1 3.51 Å).43 Thus,
we shifted our attention to the electronic effects of the linker
oxygen, which may enhance the electron density of the
phosphines. To closely examine the electronic effects of the
ligand on the catalyst activities, various electron-rich mono-
phosphine ligands were used (Table 9). As compared with the
benchmark ligand PPh3 (entry 1, 41%), the use of methyl
We successfully developed a versatile direct catalytic amination
of allylic alcohols with amines using Pt-Xantphos and Pt-DPEphos
catalysts as a useful method to synthesize various allylamines in
an environmentally friendly manner. We demonstrated that both
the large bite-angle as well as the linker oxygen atom of the
Xantphos and DPEphos ligands were essential to achieve high
catalytic activity. Moreover, under microwave heating the reaction
reached completion in a much shorter period of time compared to
conventional heating, suggesting the existence of nonthermal
microwave effects. Under the optimized conditions, the direct
substitution reaction of both aryl- and alkyl-substituted allylic
alcohols with a wide variety of aryl- and alkylamines proceeded
efficiently to afford the desired monoallylamines in high yield. The
usefulness of the present catalysis was successfully demonstrated
by one-step synthesis of the biologically active compounds Naftifine
and Flunarizine from allylic alcohol without the use of an activator.
To the best of our knowledge, this is the first report of such broad
substrate generality and high monoallylation selectivity for this type
of direct reaction. Finally, several mechanistic investigations,
including 1H and 31P{1H} NMR studies, isolation and characteriza-
tion of several catalytic intermediates, confirmation of the structure
of [Pt(η3-allyl)(xantphos)]OTf by X-ray crystallographic analysis
and crossover experiments, suggested that (1) the reduction of
Pd(cod)Cl2 with allylic alcohol and amine H2NR1 generate the
(43) Weller and Willis reported that linker oxygen of DPEphos acted as a
hemilabile ligand, increasing stability of the active Rh catalyst, see:
(a) Moxham, G. L.; Randell-Sly, H. E.; Brayshaw, S. K.; Woodward,
R. L.; Weller, A. S.; Willis, M. C. Angew. Chem., Int. Ed. 2006, 45,
7618. (b) Moxham, G. L.; Randell-Sly, H. E.; Brayshaw, S. K.; Weller,
A. S.; Willis, M. C. Chem.sEur. J. 2008, 14, 8383.
(44) Our preliminary DFT calculations (LANL2DZ for Pt and 6-31G for
others) suggested that P-O-P angle of {[Ph2(2-MeO-C6H4)P]2Pt(η3-
allyl)}+ is in the 102° to 106° range (the P-O-P angle of [(Ph3P)2Pt(η3-
allyl)]+ obtained by DFT optimizations (102°) is almost the same as
that of crystal (101°)45). The direct amination of 1a with 3a using
Pt(cod)Cl2 (0.5 mol %) and P(C6H4-2-OMe)Ph2 (1 mol %, P/Pt ) 2
or 2 mol %, P/Pt ) 4) in DMF at 50 °C for 1 h with microwave
irradiation (standard mode) afford the product 4aa in only 25% (1
mol % of ligand) and 4% (2 mol % of ligand) yields, respectively,
with recovery of the starting materials. These results indicate the partial
decomposition of Pt[P(C6H4-2-OMe)Ph2]n complex under microwave
irradiation conditions and confirm the advantage of bidentate DPEphos
and Xantphos ligands over monodentate Ph2P(C6H4-2-OMe) ligand.
(45) Huang, T.-M.; Hsu, R.-H.; Yang, C.-S.; Chen, J.-T.; Lee, G.-H.; Wang,
Y. Organometallics 1994, 13, 3657.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14327

A R T I C L E S
Ohshima et al.
active Pd(0) species along with the ammonium salt H2NR1 ·HCl
and the aldehyde, (2) Pt(η2-C3H5OH)(xantphos) (B) and [Pt(η3-
allyl)(xantphos)]OTf (DOTf) are involved in the catalytic cycle, (3)
the ammonium salt H2NR1 ·HCl acts as more efficient activatiors
of the hydroxyl group to accelerate the elimination of hydroxide
through a hydrogen bonding, (4) formation of the π-allylplatinum
complex D through the elimination of water (C f D) is an
irreversible rate-determining step, and (5) the other processes in
the catalytic cycle are reversible, even at room temperature. Further
mechanistic studies, especially on the ligand effects, further
investigation of the scope of the syntheses of highly functionalized
bioactive natural and unnatural compounds, and application to
enantioselective variants are ongoing in our group.
CHHdCH), 2.50-2.34 (m, 1H, CHH ) CH2), 1.47 (s, 3H, CH3),
1.35 (s, 3H, CH3), 0.96 (dd, J ) 8.0, 3.5 Hz, 1H, OH); 31P{1H}
NMR (121 MHz, C6D6, 35 °C): δ 19.8 (d with Pt satellite, 2JP-P
)
51.4 Hz, 1JPt-P ) 3687 Hz), 18.1 (d with Pt satellite, 2JP-P ) 51.4
Hz, 1JP-Pt ) 3508 Hz); 31P{1H} NMR (121 MHz, DMF-d7, 35 °C):
2
1
δ 19.5 (d with Pt satellite, JP-P ) 52.1 Hz, JPt-P ) 3679 Hz),
2
1
17.6 (d with Pt satellite, JP-P ) 51.1 Hz, JPt-P ) 3490 Hz); IR
(KBr, ν/cm-1) 3419, 3053, 2967, 2922, 1434, 1408, 1230, 1096,
744, 693, 534, 512; MS (FAB) m/z 815 ([M-OH+H]+), 773 (M
-C4H5OH); mp. 217.2-219.4 °C (decompose); Anal. Calcd. for
C42H38O2P2Pt: C, 60.65; H, 4.60. Found: C, 60.31; H, 4.43.
Synthesis of [Pt(η3-allyl)(xantphos)]OTf (DOTf). Allyltributyl-
stannane (55 mg, 0.165 mmol, 1.1 equiv) was added to a suspension
of Pt(xantphos)Cl2 (136 mg, 0.161 mmol) in THF (20 mL) at room
temperature. This solution was added to AgOTf (42 mg, 0.161 mmol,
1.0 equiv) in a Schlenk tube. After being stirred at the same temperature
for 5 h, the mixture was concentrated under reduced pressure. The
residue was extracted with CH2Cl2 and the filtrate was concentrated.
The resulting solid was washed with CHCl3/hexane (1/10) (10 mL ×
5) to remove ClSnBu3 salt and then dried under reduced pressure to
afford [Pt(xantphos)(η3-allyl)]OTf as a white solid (130 mg, 0.135
mmol, 84%); 1H NMR (300 MHz, CDCl3, 35 °C): δ 7.65 (d, J ) 7.7
Hz, 2H, aromatic), 7.50-6.98 (m, 20H, aromatic), 6.61 (t, J ) 8.3
Hz, 2H, aromatic), 5.75-5.45 (m, 1H, allylic CH), 3.59-3.48 (br,
2H, allylic CH2), 3.06-2.82 (m, 2H, alllylic CH2), 1.81 (s, 3H, CH3),
1.59 (s, 3H, CH3); 31P{1H} NMR (121 MHz, CDCl3, 35 °C): δ 1.24
(s with Pt satellite, 1JP-Pt ) 4179 Hz); 31P{1H} NMR (121 MHz, DMF-
Experimental Section
General Procedure for the Pt-Catalyzed Direct Amination
Using Arylamines (Table 4). To a solution of Pt(cod)Cl2 (3.8 mg,
0.010 mmol, 0.5 mol %) and Xantphos (5.9 mg, 0.010 mmol, 0.5
mol %) in DMF (0.4 mL) were added cinnamyl alcohol (1a, 265.4
mg, 2.0 mmol) and aniline (3a, 278.4 mg, 3.0 mmol, 1.5 equiv).
The reaction was heated using a CEM Discover single-mode
microwave reactor (standard mode) for 1 h. The resulting mixture
was directly purified by flash column chromatography (silica gel,
hexane/EtOAc ) 45/1) to afford 4aa as a colorless oil (387.2 mg,
88%) and 5aa as a colorless oil (38.8 mg, 6%).
General Procedure for the Pt-Catalyzed Direct Amination
Using Alkylamines (Table 7). To a solution of Pt(cod)Cl2 (7.6
mg, 0.020 mmol, 1.0 mol %) and DPEphos (21.6 mg, 0.040 mmol,
2.0 mol %) in toluene (0.4 mL) were added cinnamyl alcohol (1a,
265.0 mg, 2.0 mmol) and cyclohexylamine (6e, 295.4 mg, 3.0
mmol, 1.5 equiv). The reaction was heated using CEM Discover
single-mode microwave reactor (standard mode) for 2 h. The resulting
mixture was directly purified by flash column chromatography (silica
gel, hexane/EtOAc/Et3N ) 20/5/1) to afford 7ae as a colorless oil
(357.8 mg, 85%) and 8ae as a colorless oil (48.2 mg, 7%).
1
d7, 35 °C): δ 0.85 (s with Pt satellite, JPt-P ) 4210 Hz); 19F NMR
(282 MHz, CDCl3, 35 °C); δ -79.1; IR (KBr, ν/cm-1) 3069, 3055,
2965, 2925, 1436, 1413, 1266, 1223, 1150, 1097, 1030, 753, 695, 637;
MS (FAB) m/z 814 ([M-OTf]+), 773; mp. 257.0-257.8 °C (decom-
pose); Anal. Calcd for C43H37F3O4P2SPt ·CHCl3: C, 48.79; H, 3.54;
Found: C, 48.70; H, 3.31.
Structure Determination and Refinement. The structure was
solved by direct methods (SIR-92) and expanded using Fourier
techniques (DIRDIF-99). The structure was refined on F2 by full-matrix
least-squares methods, using SHELXL-97. The non-hydrogen atoms
were anisotropically refined and all H-atoms were isotropically refined
using the riding model. The function minimized was [Σω(Fo2 - Fc2)2]
(ω ) 1/[σ2(Fo2) + (0.0399P)2 + 0.0149P]), where P ) (Max(Fo2,0)
+ 2Fc2)/3 with σ2 (Fo2) from counting statistics. The function R1 and
ωR2 were (Σ||Fo| - |Fc||)/Σ|Fo| and [Σω(Fo2 - Fc2)2/Σ(ωFo4)]1/2,
respectively, and R1 and wR2 are 0.0616 and 0.1612 for 9521
reflections with I > 2.0σ(I), respectively. All hydrogen atoms for each
complex were placed in calculated positions (C-Hsp2 ) 0.95 Å, Uiso(H)
) 1.2*Ueq(C) Å2, Csp3-H(CH2) ) 0.98 Å, Uiso(H) ) 1.2Ueq(C) Å2)
and kept fixed. All calculations ware performed using the Crystal-
Structure crystallographic software package except for refinement,
which was performed using SHELXL-97.
Synthesis of Pt(xantphos)Cl2 (A). Pt(cod)Cl2 (120 mg, 0.32
mmol) and Xantphos (186 mg, 0.32 mmol, 1.0 equiv) were
dissolved in CHCl3 (20 mL). After being stirred at room temperature
for 1.5 h, the solvent was removed under reduced pressure. The
residue was washed with hexane/CH2Cl2 (10/1) (10 mL × 3) and
dried to afford Pt(xantphos)Cl2 as a white solid (264 mg, 0.31 mmol,
97%); 1H NMR (300 MHz, CDCl3, 35 °C): δ 7.62 (d, J ) 7.7 Hz,
2H), 7.48-7.34 (m, 10 H), 7.26-7.19 (m, 2H), 7.16-7.08 (m, 4H),
7.06-6.97 (m, 8H), 1.86 (s, 6H, CH3); 31P{1H} NMR (121 MHz,
1
CDCl3, 35 °C): δ 6.60 (s with Pt satellite, JP-Pt ) 3694 Hz);
31P{1H} NMR (121 MHz, DMF-d7, 35 °C): δ 7.14 (s with Pt
1
satellite, JP-Pt ) 3669 Hz); MS (FAB) m/z 773 ([M-Cl2]+); IR
(KBr, ν/cm-1) 3054, 3019, 2970, 2924, 2857, 1480, 1434, 1403,
1220, 1096, 753, 740, 706, 692, 533; mp >300 °C (measurement
limit); Anal. Calcd for C39H32Cl2OP2Pt4•CH2Cl2: C, 51.68; H, 3.69;
Found: C, 52.04; H, 3.35.
Acknowledgment. This work was supported by a Grant-in-Aid
for Science Research in a Priority Area (No. 20037040, “Chemistry
of Concerto Catalysis”) and Grant-in-Aid for Scientific Research (B)
(21390003) from the Ministry of Education, Culture, Sports, Science
and Technology, Japan, and Mitsubishi Chemical Corporation Founda-
tion. Some preliminary calculations were performed on the IBM p570
server in Mitsui chemicals Inc. through the courtesy of its computa-
tional science department.
Synthesis of Pt(η2-C3H5OH)(xantphos) (B). A solution of
Pt(xantphos)Cl2 (244 mg, 0.29 mmol) and allyl alcohol (1b, 142
mg, 2.5 mmol, 8.5 equiv) in THF (4.5 mL) was stirred at room
temperature for 30 min. Then, an aqueous solution (2.0 mL) of
NaBH4 (55 mg, 1.5 mmol, 5.0 equiv) was added to the solution.
After being stirred at the same temperature for 5 h, the mixture
was concentrated under reduced pressure and the resulting solid
was extracted with CH2Cl2. The filtrate was concentrated under
reduced pressure and the resulting brown solid was washed with
hexane/CH2Cl2 (10/1) (5 mL × 2) to give Pt(xantphos)(η2-C3H5OH)
as a brown solid (239 mg, 0.27 mmol, 94%); 1H NMR (300 MHz,
C6D6, 35 °C): δ 7.58-7.44 (m, 8H, aromatic), 7.25-7.05 (m, 8H,
aromatic), 7.00-6.85 (m, 6H, aromatic), 6.82-6.70 (m, 4H,
aromatic), 4.06-3.93 (m, 1H, CHHOH), 3.65-3.49 (m, 1H,
CHHOH), 3.44-3.33 (m, 1H, CH2dCHCH2), 2.61-2.50 (m, 1H,
Supporting Information Available: Experimental procedures,
characterization of the products with 1H and 13C NMR spectra, 1H
and 31P{1H} NMR studies, computational data, other detailed
results and discussion (PDF); X-ray crystallographic data of DOTf
(CIF). This material is available free of charge via the Internet at
http://pubs.acs.org.
JA9046075
9
14328 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009