Synthetic, structural, and mechanistic aspects of an amine activation process mediated at a zwitterionic Pd(II) center

Article abstract of DOI:10.1021/ja046415s

A zwitterionic palladium complex [[Ph₂BP₂]Pd(THF) ₂][OTf] (1) (where [Ph₂BP₂] = [Ph ₂B(CH₂PPh₂)₂]-) reacts with trialkylamines to activate a C-H bond adjacent to the amine N atom, thereby producing iminium adduct complexes [Ph₂BP₂]Pd(N,C: η²-NR₂CHR′). In all cases examined the amine activation process is selective for the secondary C-H bond position adjacent to the N atom. These palladacycles undergo facile β-hydride elimination/olefin reinsertion processes as evident from deuterium scrambling studies and chemical trap studies. The kinetics of the amine activation process was explored, and β-hydride elimination appears to be the rate-limiting step. A large kinetic deuterium isotope effect for the amine activation process is evident. The reaction profile in less polar solvents such as benzene and toluene is different at room temperature and leads to dimeric {[Ph₂BP₂]Pd} ₂ (4) as the dominant palladium product. Low-temperature toluene-d₈ experiments proceed more cleanly, and intermediates assigned as [Ph₂BP₂]Pd(NEt₃)(OTF) and the iminium hydride species [[Ph₂BP₂]Pd(H)(Et ₂N=CHCH₃)][OTf] are directly observed. The complex (Ph₂-SiP₂)Pd(OTf)₂ (14) was also studied for amine activation and generates dimeric [(Ph₂SiP₂)Pd] ₂[OTf]₂ (16) as the dominant palladium product. These collective data are discussed with respect to the mechanism of the amine activation and, in particular, the influence that solvent polarity and charge have on the overall reaction profile.

Full text of DOI:10.1021/ja046415s

Published on Web 11/10/2004
Synthetic, Structural, and Mechanistic Aspects of an Amine
Activation Process Mediated at a Zwitterionic Pd(II) Center
Connie C. Lu and Jonas C. Peters*
Contribution from the DiVision of Chemistry and Chemical Engineering, Arnold and Mabel
Beckman Laboratories of Chemical Synthesis, California Institute of Technology,
Pasadena, California 91125
Received June 17, 2004; E-mail: jpeters@caltech.edu
Abstract: A zwitterionic palladium complex [[Ph2BP2]Pd(THF)2][OTf] (1) (where [Ph2BP2] ) [Ph2B(CH2PPh2)2]-)
reacts with trialkylamines to activate a C-H bond adjacent to the amine N atom, thereby producing iminium
adduct complexes [Ph2BP2]Pd(N,C:η2-NR2CHR′). In all cases examined the amine activation process is
selective for the secondary C-H bond position adjacent to the N atom. These palladacycles undergo facile
â-hydride elimination/olefin reinsertion processes as evident from deuterium scrambling studies and chemical
trap studies. The kinetics of the amine activation process was explored, and â-hydride elimination appears
to be the rate-limiting step. A large kinetic deuterium isotope effect for the amine activation process is
evident. The reaction profile in less polar solvents such as benzene and toluene is different at room
temperature and leads to dimeric {[Ph2BP2]Pd}2 (4) as the dominant palladium product. Low-temperature
toluene-d8 experiments proceed more cleanly, and intermediates assigned as [Ph2BP2]Pd(NEt3)(OTf) and
the iminium hydride species [[Ph2BP2]Pd(H)(Et2NdCHCH3)][OTf] are directly observed. The complex (Ph2-
SiP2)Pd(OTf)2 (14) was also studied for amine activation and generates dimeric [(Ph2SiP2)Pd]2[OTf]2 (16)
as the dominant palladium product. These collective data are discussed with respect to the mechanism of
the amine activation and, in particular, the influence that solvent polarity and charge have on the overall
reaction profile.
Harman group,4 who provided spectroscopic data to support the
generation of an Os(II) η2-iminium hydride complex via
Introduction
â-hydride elimination upon reduction of an Os(III) amine
complex. Murahashi also invoked η2-iminium hydrides of Pd-
(II) as intermediates in trialkylamine exchange reactions cata-
lyzed by palladium black at 200 °C (NR1 R2 f NR1 + NR2
Exploiting transition metals to functionalize tertiary amines
via initial C-H bond activation is a growing area of synthetic
interest.1 For instance, Murai and co-workers described a
catalytic process for the carbonylation of tertiary amines at a
position adjacent to an amine N atom using a rhodium(I)
precursor.2 This system requires amine substrates with pendant
directing groups to chelate the metal catalyst. More recently,
Murahashi et al. reported a ruthenium catalyst for the cyanation
of tertiary amines without the requirement of directing groups.3
Iminium adducts of ruthenium, while not observed, were
proposed as key intermediates. Precedent for this type of
reactivity had been previously demonstrated at osmium by the
2
3
3
+ NR1 R2 + NR1R2 , where R1 * R2).5 Another example of a
2
2
catalytic transformation of tertiary amines comes from Gold-
man’s group with the report that iridium(III) dihydride precur-
sors can mediate the dehydrogenation of tertiary amines to
produce enamines in the presence of a H2-acceptor (i.e., NEt3
+ alkene f Et2NCHdCH2 + alkane).6
Previous work from our lab has concerned mechanistic
aspects of aromatic C-H bond activations mediated by a
zwitterionic platinum(II) center.7 With an interest in extending
this effort, we turned to developing a related palladium system.
In this report a reactive Pd(II) complex is described, [[Ph2BP2]-
Pd(THF)2][OTf] (1), where [Ph2BP2] represents the bis(phos-
phino)borate ligand [Ph2B(CH2PPh2)2]-. This system stoichi-
ometrically activates a variety of trialkylamines at a secondary
(1) (a) Murahashi, S.-I. Angew. Chem., Int. Ed. Engl. 1995, 34, 2443-2465.
(b) Doye, S. Angew. Chem., Int. Ed. 2001, 40, 3351-3353.
(2) (a) Chatani, N.; Asaumi, T.; Ikeda, T.; Yorimitsu, S.; Ishii, Y.; Kakiuchi,
F.; Murai, S. J. Am. Chem. Soc. 2000, 122, 12882-12283. For Rh-catalyzed
carbenoid insertions into C-H bonds R to N, see: (b) Davies, H. M. L.;
Hansen, T.; Hopper, D. W.; Panaro, S. A. J. Am. Chem. Soc. 1999, 121,
6509-6510. (c) Davies, H. M. L.; Venkataramani, C.; Hansen, T.; Hopper,
D. W. J. Am. Chem. Soc. 2003, 125, 6462-6468. (d) Davies, H. M. L.;
Venkataramani, C. Angew. Chem., Int. Ed. 2002, 41, 2197-2199. For a
related example of imine activation, see: (e) Sakaguchi, S.; Kubo, T.; Ishii,
Y. Angew. Chem., Int. Ed. 2001, 40, 2534-2536.
(4) (a) Barrera, J.; Orth, S. D.; Harman, W. D. J. Am. Chem. Soc. 1992, 114,
7316-7318. (b) Orth, S. D.; Barrera, J.; Rowe, S. M.; Helberg, L. E.;
Harman, W. D. Inorg. Chim. Acta 1998, 270, 337-344.
(5) (a) Murahashi, S.-I.; Hirano, T.; Yano, T. J. Am. Chem. Soc. 1978, 100,
348-350. (b) Murahashi, S.-I.; Yoshimura, N.; Tsumiyama, T.; Kojima,
T. J. Am. Chem. Soc. 1983, 105, 5002-5011.
(3) (a) Murahashi, S.-I.; Komiya, N.; Terai, H.; Nakae, T. J. Am. Chem. Soc.
2003, 125, 15312-15313. For other Ru examples, see: (b) Chatani, N.;
Asaumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am. Chem.
Soc. 2001, 123, 10935-10941. (c) Chatani, N.; Fukuyama, T.; Tatamidani,
H.; Kakiuchi, F.; Murai, S. J. Org. Chem. 2000, 65, 4039-4047.
(6) Zhang, X.; Fried, A.; Knapp, S.; Goldman, A. S. Chem. Commun. 2003,
2060-2061.
(7) (a) Thomas, J. C.; Peters, J. C. J. Am. Chem. Soc. 2001, 123, 5100-5101.
(b) Thomas, J. C.; Peters, J. C. J. Am. Chem. Soc. 2003, 125, 8870-8888.
9
15818
J. AM. CHEM. SOC. 2004, 126, 15818-15832
10.1021/ja046415s CCC: $27.50 © 2004 American Chemical Society

Aspects of an Amine Activation Process
A R T I C L E S
at 80 °C for 12 h, and the former was used without further purification.
The reagent NaCN (99%) was dried under vacuum with heating at 80
°C. tert-Butyl isocyanide was dried over 3 Å molecular sieves and used
without further purification. Elemental analyses were performed by
Desert Analytics, Tucson, AZ. A Varian Mercury-300 NMR spectrom-
eter and a Varian Inova-500 NMR spectrometer were used to record
Figure 1. Relationship between amine-to-iminium ion and alcohol-to-
ketone oxidation reactions.
1H, 13C, 19F, and 31P NMR spectra unless otherwise stated. 1H and 13
C
NMR chemical shifts were referenced to residual solvent. Proton peaks
were assigned based on TOCSY-1D and 31P-decoupled 1H NMR
studies. 31P NMR chemical shifts are reported relative to an external
standard of 85% H3PO4. 19F NMR chemical shifts are reported relative
to an external standard of hexafluorobenzene. IR measurements were
obtained with a KBr solution cell using a Bio-Rad Excalibur FTS 3000
spectrometer controlled by Bio-Rad Merlin Software (v. 2.97) set at 4
cm-1 resolution. X-ray diffraction experiments were performed at the
Beckman Institute Crystallographic Facility on a Bruker Smart 1000
CCD diffractometer.
C-H position adjacent to the amine N atom. Formal proton
loss accompanies the activation process, and structurally unusual
iminium adduct complexes of palladium are obtained.
This reaction system is particularly well suited to mechanistic
examination, and we have therefore undertaken such a study.
We attempt to address the reaction’s overall selectivity and
suggest a plausible operational mechanism. Our data indicate
that the processes may be mechanistically germane to the mode
of palladium-catalyzed alcohol oxidations8 (Figure 1) and also
to the in situ reduction of Pd(II) precursors by trialkyamines9
in Pd(0)-catalyzed reactions, as, for example, in the Heck
reaction.10 These processes are also of potential relevance to
â-hydride elimination processes of palladium amides11 and to
â-hydride eliminations that generate imines from secondary
amines.12
[[Ph2BP2]Pd(THF)2][OTf] (1). A THF solution of AgOTf (72.9 mg,
71 µmol) was added to a THF suspension of {[Ph2BP2]Pd(µ-Cl)}2
(0.200 g, 142 µmol). The reaction was stirred for 7 h and filtered
through glass wool. The bright yellow filtrate was dried in vacuo for
1
9 h to give a yellow foam-powder (0.252 g, 92% yield). H NMR
(C6D6, 500 MHz) δ 7.41 (dd, J ) 7.0 and 12.0 Hz, 8H, Ho of PPh2),
7.23 (br d, J ) 7.0 Hz, 4H, Ho of BPh2), 7.09 (t, J ) 7.5 Hz, 4H, Hm
of BPh2), 7.04 (tt, J ) 1.0 and 7.5 Hz, 2H, Hp of BPh2), 6.97 (dd, J )
6.5 and 8.0 Hz, Hp of PPh2), 6.89 (app t, 8H, Hm of PPh2), 3.44 (m,
8H, O(CH2CH2)2), 1.95 (br dd, J ) 5.0 and 14.5 Hz, 4H, CH2PPh2),
1.26 (m, 8H, O(CH2CH2)2); 13C NMR (d6-benzene, 75 MHz) δ 160.3
(br, Cipso of BPh2), 133.1 (app t, J ) 5.4 Hz, Co of PPh2), 132.6 (Co of
BPh2), 131.9 (Cp of PPh2), 129.8 (d, J ) 55.8 Hz, Cipso of PPh2), 129.0
(app t, J ) 5.4 Hz, Cm of PPh2), 127.7 (Cm of BPh2), 124.0 (Cp of BPh2),
118.7 (d, J ) 319 Hz, CF3), 69.9 (O(CH2CH2)2), 25.9 (O(CH2CH2)2),
18.6 (br, CH2PPh2); 31P NMR (C6D6, 121 MHz) δ 48.2; 19F NMR (C6H6,
282 MHz) δ 973.5. Anal. Calcd for C47H50BF3O5P2PdS: C, 58.61; H,
5.23; N, 0. Found: C, 58.66; H, 5.18; N, 0.08.
Experimental Section
General Considerations. All manipulations were carried out using
standard Schlenk or glovebox techniques under a dinitrogen atmosphere.
Unless otherwise noted, solvents were deoxygenated and dried by
thorough sparging with N2 gas, followed by passage through an
activated alumina column. Diethyl ether, tetrahydrofuran, petroleum
ether, and benzene were typically tested with a standard purple solution
of sodium benzophenone ketyl in tetrahydrofuran in order to confirm
effective oxygen and moisture removal. Deuterated solvents were
purchased from Cambridge Isotope Laboratories, Inc. The solvents were
dried over activated 3 Å molecular sieves and degassed by freeze-
pump-thaw cycles prior to use. The preparation of Et2NC(H)dCH2,13
diisopropyl(2-propenyloxymethyl)amine,14 (COD)PdCl2,15 Ph2SiP2,7 and
[NBu4][Ph2BP2]16 was carried out following literature procedures.
Amines and o-phenylpyridine were purchased from Aldrich, dried over
CaH2, distilled or vacuum transferred, and stored over activated 3 Å
molecular sieves. The reagents (PhCN)2PdCl2 and AgOTf were
purchased from Strem; the latter was dried under vacuum with heating
[NBu4][[Ph2BP2]PdCl2] (2). A THF solution (10 mL) of (COD)-
PdCl2 (0.600 g, 2.10 mmol) was added dropwise to a stirring suspension
of [NBu4][Ph2BP2] (1.694 g, 2.10 mmol) in THF (13 mL). The reaction
mixture immediately became homogeneous and turned burnt orange.
After stirring for 30 min, all volatiles were removed in vacuo. The
resulting orange oil was taken up in benzene to precipitate the product
as a light yellow crystalline solid. The solids were collected on a frit
and washed with benzene and petroleum ether to give light yellow 2
(0.410 g, 40% yield). 1H NMR (d6-acetone, 500 MHz) δ 7.56 (m, 8H,
Ho of PPh2), 7.24 (app t, J ) 7.0 Hz, 4H, Hp of PPh2), 7.11 (td, J )
1.5 and 7.5 Hz, 8H, Hm of PPh2), 6.85 (br d, J ) 6.5 Hz, 4H, Ho of
BPh2), 6.70 (app t, J ) 7.5 Hz, 4H, Hm of BPh2), 6.65 (tt, J ) 1.5 and
7.5 Hz, 2H, Hp of BPh2), 3.35 (m, 8H, N(CH2CH2CH2CH3)4), 1.76 (br
d, J ) 10.5 Hz, 4H, CH2PPh2), 1.73 (m, 8H, N(CH2CH2CH2CH3)4),
1.38 (sextet, J ) 7.5 Hz, 8H, N(CH2CH2CH2CH3)4), 0.95 (t, J ) 7.5
Hz, 12H, N(CH2CH2CH2CH3)4); 13C NMR (d6-acetone, 125 MHz) δ
163.3 (br, Cipso of BPh2), 135.5 (dd, J ) 4.3 and 55.6 Hz, Cipso of PPh2),
134.5 (m, Co of PPh2), 132.7 (Co of BPh2), 129.8 (Cp of PPh2), 127.7
(m, Cm of PPh2), 126.7 (Cm of BPh2), 122.6 (Cp of BPh2), 59.3 (t, J )
2.4 Hz, N(CH2CH2CH2CH3)4), 24.4 (N(CH2CH2CH2CH3)4), 21.6 (br,
CH2PPh2), 20.3 (N(CH2CH2CH2CH3)4), 13.9 (N(CH2CH2CH2CH3)4);
31P NMR (d6-acetone, 121 MHz) δ 35.46. Anal. Calcd for C54H70BCl2-
NP2Pd: C, 65.96; H, 7.18; N, 1.42. Found: C, 65.83; H, 7.60; N, 1.63.
(8) For mechanistic considerations pertaining to these types of reactions, see:
(a) Nishimura, T.; Onoue, T.; Ohe, K.; Uemura, S. J. Org. Chem. 1999,
64, 6750-6755. (b) Mueller, J. A.; Sigman, M. S. J. Am. Chem. Soc. 2003,
125, 7005-7013. (c) Jensen, D. R.; Schultz, M. J.; Mueller, J. A.; Sigman,
M. S. Angew. Chem., Int. Ed. 2003, 42, 3810-3813. (d) Steinhoff, B. A.;
Stahl, S. S. Org. Lett. 2002, 4, 4179-4181. (e) Steinhoff, B. A.; Fix, S.
R.; Stahl, S. S. J. Am. Chem. Soc. 2002, 124, 766-767. (f) Mueller, J. A.;
Jensen, D. R.; Sigman, M. S. J. Am. Chem. Soc. 2002, 124, 8202-8203.
(g) ten Brink, G.-J.; Arends, I. W. C. E.; Sheldon, R. A. AdV. Synth. Catal.
2002, 344, 355-369. (h) Trend, R. M.; Stoltz, B. M. J. Am. Chem. Soc.
2004, 126, 4482-4483. (i) Noronha, G.; Henry, P. M. J. Mol. Catal. A:
Chem. 1997, 120, 75-87. (j) For a review on this topic, see: Muzart, J.
Tetrahedron 2003, 59, 5789-5816. (k) Steinhoff, B. A.; Guzei, I. A.; Stahl,
S. S. J. Am. Chem. Soc. 2004, 126, 11268-11278.
(9) See: Trzeciak, A. M.; Ciunik, Z.; Zio´łkowski, J. J. Organometallics 2002,
21, 132-137 and references therein.
(10) (a) Tsuji, J. Palladium Reagents and CatalystssInnoVations in Organic
Synthesis; John Wiley & Sons: New York, 1995; pp 125-128. (b)
Beletskaya, I. P.; Cheprakov, A. V. Chem. ReV. 2000, 100, 3009-3066
and references therein.
(11) For a review, see: (a) Hartwig, J. F. Angew. Chem., Int. Ed. 1998, 37,
2046-2067. See also: (b) Hartwig, J. F.; Richards, S.; Baran˜ano, D.; Paul,
F. J. Am. Chem. Soc. 1996, 118, 3626-3633.
{[Ph2BP2]Pd(µ-Cl)}2 (3). A THF solution of PdCl2(NCPh)2 (1.4856
g, 3.834 mmol) was added rapidly to a stirring THF solution of
[NBu4][Ph2BP2] (3.0900 g, 3.834 mmol). The orange-yellow solution
was filtered through Celite and dried in vacuo. The residue was washed
thoroughly with Et2O to give pale yellow solids, which by 31P NMR
was a mixture of compounds 2 and 3. The solids were dissolved in
THF (150 mL) in a 300 mL of RBF with a stir bar, and NaBPh4 (1.3123
g, 3.834 mmol) was added in the dark. After stirring for 8 h in the
(12) (a) Strieter, E. R.; Blackmond, D. G.; Buchwald, S. L. J. Am. Chem. Soc.
2003, 125, 13978-13980. (b) Guram, A. S.; Rennels, R. A.; Buchwald, S.
L. Angew. Chem., Int. Ed. Engl. 1995, 34, 1348-1350.
(13) Laban, G.; Mayer, R. Z. Chem. 1967, 7, 12-13.
(14) Arenz, T.; Frauenrath, H.; Raabe, G.; Zorn, M. Liebigs Ann. Chem. 1994,
931-942.
(15) Drew, D.; Doyle, J. R. Inorg. Synth. 1990, 28, 346-349.
(16) Peters, J. C.; Thomas, J. C. Inorg. Synth. 2004, 34, 8-14.
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15819

A R T I C L E S
Lu and Peters
dark, yellow solids precipitated from the brown solution. The first crop
of solids were collected and washed sparingly with THF and a 1:1
THF/Et2O solution. To remove NaCl, the solids were dissolved in CH2-
Cl2 and filtered through Celite. The bright yellow solution was dried
in vacuo to give yellow crystalline solids. A second crop of product
was collected from the concentrated brown filtrate at -30 °C. The crops
(m, 14H, aryl H), 6.97-7.00 (m, 12H, aryl H), 2.97 (app hextet, J )
6.0 Hz, 1H, PdNCHCH3), 2.46 (septet d, J ) 3.6 and 6.9 Hz, 1H,
N(CHH′Me)Et′, 2.17-2.45 (m, 7H, N(CHH′Me)C′HH′Me, CH2PPh2,
and C′H2P′Ph2), 0.75 (td, J ) 8.1 and 10.2 Hz, 3H, PdNCHCH3), 0.47
(m, 6H, N(CH2Me)C′H2Me); 13C NMR (CDCl3, 125 MHz) δ 163.4
(br, Cipso and C′ipso of BPh2), 140.7 (d, J ) 26.1 Hz, Cipso of PPh2),
139.2 (d, J ) 34.3 Hz, C′ipso of PPh2), 139.1 (d, J ) 28.0 Hz, Cipso of
P′Ph2), 138.3 (d, J ) 37.8 Hz, C′ipso of P′Ph2), 133.6 (Co of BPh2),
133.2 (d, J ) 13.1 Hz, Co of PPh2), 133.1 (Co′ of BPh2), 132.8 (d, J )
12.7 Hz, Co′ of PPh2), 132.34 (d, J ) 13.2 Hz, Co of P′Ph2), 132.31 (d,
J ) 11.6 Hz, Co′ of P′Ph2), 128.8 (d, J ) 2.3 Hz, Cp of PPh2), 128.62
(d, J ) 1.9 Hz, Cp′ of PPh2), 128.59 (d, J ) 1.5 Hz, Cp of P′Ph2),
128.2 (d, J ) 2.0 Hz, Cp′ of P′Ph2), 127.95 (d, J ) 6.2 Hz, Cm of
PPh2), 127.8-127.9 (m, Cm′ of PPh2, Cm and Cm′ of P′Ph2), 126.4 (Cm
of BPh2), 126.1 (Cm′ of BPh2), 122.73 (Cp of BPh2), 122.66 (Cp′ of BPh2),
64.3 (dd, J ) 13.5 and 57.9 Hz, PdN(CHMe)), 52.2 (s, N(CH2Me)),
45.6 (s, N(C′H2Me)), 20.6 (br m, CH2PPh2 and C′H2PPh2), 16.1 (app
1
were combined to give 1.911 g of 3 (73% yield). H NMR (CDCl3,
300 MHz) δ 7.30-7.18 (m, 12H, Hp and Ho of PPh2), 7.07 (t, J ) 7.8
Hz, 8H, Hm of PPh2), 6.91-6.82 (m, 10H, aryl H’s of BPh2), 1.79 (br
d, J ) 10.5 Hz, 4H, CH2PPh2); 13C NMR (CDCl3, 125 MHz) δ 161.4
(br, Cipso of BPh2), 133.3 (m, Co of PPh2), 132.1 (Co of BPh2), 131.5
(d, J ) 56.0 Hz, Cipso of PPh2), 130.7 (Cp of PPh2), 128.1 (m, Cm of
PPh2), 126.8 (Cm of BPh2), 123.0 (Cp of BPh2), 17.7 (br, CH2PPh2); 31
P
NMR (CDCl3, 121.5 MHz) δ 43.07. Anal. Calcd for C76H68B2Cl2P4-
Pd: C, 64.71; H, 4.86; N, 0.00. Found: C, 64.80; H, 4.83; N, <0.05.
[Ph2BP2]Pd(o-phenylpyridine)OTf (5). To a bright yellow THF
solution of 1 (75.0 mg, 78.0 µmol) was added o-phenylpyridine (11.1
µL, 84.0 µmol). The solution immediately turned off-white and slightly
turbid. Petroleum ether (5 mL) was added to precipitate solids. The
white solids were collected and washed liberally with Et2O. The solids
were dried in vacuo to give 69 mg of a pale yellow powder (91% yield).
Complex 5 is fluxional in solution at room temperature. 1H NMR
(CDCl3, 300 MHz, -40 °C) δ 8.35 (d, J ) 4.8 Hz, 1H, Ho of pyr),
7.69-7.52 (m, 12H, aryl), 7.32-7.07 (m, 10H, aryl), 6.84 (t, 2H, J )
6.4 Hz, Hp of BPh2), 6.72-6.67 (m, 12H, aryl), 6.24 (br d, J ) 7.5 Hz,
2H, aryl), 2.55 (d, J ) 13.5 Hz, 1H, CHH′PPh2), 1.93 (d, J ) 14 Hz,
1H, CHH′PPh2), 1.65 (d, J ) 14 Hz, 1H, CHH′P′Ph2), 1.53 (d, J )
13.8 Hz, 1H, CHH′P′Ph2); 31P NMR (CDCl3, 121 MHz, -40 °C) δ
46.3 (d, J ) 20 Hz), 35.3 (d, J ) 20 Hz); 19F NMR (CDCl3, 282 MHz,
-40 °C) δ -81.6. Anal. Calcd for C50H43BF3NO3P2PdS: C, 61.65; H,
4.45; N, 1.44. Found: C, 61.80; H, 4.21; N, 1.54.
t,
J ) 3.5 Hz, PdN(CHMe)), 13.1 (s, N(CH2Me)), 12.5 (s,
N(CH2Me′)): 31P NMR (C6D6, 121 MHz) δ 32.2 (d, J ) 35.7 Hz),
15.7 (d, J ) 35.7 Hz). Anal. Calcd for C44H48BNP2Pd: C, 68.63; H,
6.28; N, 1.82. Found: C, 68.82; H, 5.99; N, 1.88.
[Ph2BP2]Pd(N,C:η2-NCy2CHCH3) (8). Compound 1 (200.0 mg,
207.86 µmol) and NEtCy2 (0.96 mL, 4.17 mmol) were dissolved in
THF. After stirring for 3 h, the solution was dried in vacuo. The residue
was washed with petroleum ether, benzene, and acetonitrile. The
remaining pink powder was dissolved in CH2Cl2 and filtered through
Celite (69 mg, 38% yield). Single crystals suitable for X-ray diffraction
studies were grown from vapor diffusion of petroleum ether into a CH2-
Cl2 solution of 8 at -30°C. 1H NMR (CDCl3, 300 MHz) δ 7.11-7.36
(m, 24H, aryl H), 6.80-6.93 (m, 6H, aryl H), 3.38 (app quintet d, J )
1.1 and 2.9 Hz, 1H, PdNCHCH3), 2.79 (app q, J ) 4.9 Hz, 1H,
NCH(CH2)5), 1.08-2.07 (m, 25H, aliphatic H’s), 0.96 (td, J ) 5.4 and
9.0 Hz, 3H, PdNCHCH3); 13C NMR (CDCl3, 125 MHz) δ 164.2 (br,
Cipso and C′ipso of BPh2), 141.6 (d, J ) 23.9 Hz, Cipso of PPh2), 139.6
(d, J ) 34.6 Hz, C′ipso of PPh2), 137.7 (d, J ) 28.8 Hz, Cipso of P′Ph2),
137.2 (d, J ) 38.2 Hz, C′ipso of P′Ph2), 133.96 (d, J ) 13.4 Hz, Co of
PPh2), 133.4 (d, J ) 13.2 Hz, Co′ of PPh2), 133.1 (Co of BPh2), 132.7
(Co′ of BPh2), 131.88 (d, J ) 10.7 Hz, Co of P′Ph2), 131.85 (d, J )
11.2 Hz, Co′ of P′Ph2), 129.0 (d, J ) 2.0 Hz, Cp of PPh2), 128.9 (d, J
) 1.9 Hz, Cp′ of PPh2), 128.4 (d, J ) 2.4 Hz, Cp of P′Ph2), 128.0 (d,
J ) 9.2 Hz, Cm of PPh2), 127.9 (d, J ) 1.5 Hz, Cp′ of P′Ph2), 127.72
(d, J ) 8.8 Hz, Cm′ of PPh2), 127.70 (d, J ) 10.1 Hz, Cm of P′Ph2),
127.68 (d, J ) 8.0 Hz, Cm′ of P′Ph2), 126.2 (s, Cm of BPh2), 126.1 (s,
Cm′ of BPh2), 122.3 (Cp of BPh2), 122.2 (Cp′ of BPh2), 64.5 (dd, J )
13.4 and 56.7 Hz, PdN(CHMe)), 63.8 (N(CH(CH2)5)), 59.9 (d, J )
3.1 Hz, N(C′H(C′H2)5)), 36.5 (Cy), 35.4 (Cy), 35.2 (Cy), 32.71 (Cy),
26.7 (Cy), 26.4 (Cy), 26.22 (Cy), 26.18 (Cy), 25.5 (2 overlapping s,
Cy), 20.4-21.9 (br m, CH2PPh2 and C′H2PPh2), 15.7 (d, J ) 4.3 Hz,
PdN(CHMe)); 31P NMR (CDCl3, 121 MHz) δ 31.8 (d, J ) 31.1 Hz),
13.2 (d, J ) 31.1 Hz). Anal. Calcd for C52H60BNP2Pd: C, 71.12; H,
6.89; N, 1.59. Found: C, 70.77; H, 6.65; N, 1.36.
[Ph2BP2]Pd(N,C:η2-NCH3CH(CH2)3) (9). Compound 9 was pre-
pared with N-methylpyrrolidene using a procedure identical to that used
for compound 7 (67 mg, 82% yield). Single crystals suitable for X-ray
diffraction studies were grown from vapor diffusion of petroleum ether
into a THF solution of 9 at -30 °C. 1H NMR (C6D6, 300 MHz) δ 7.63
(br t, J ) 6.3 Hz, 4H, aryl H), 7.38 (dd, J ) 8.7 and 9.6 Hz, 2H, Ho
of PPhPh′), 7.11-7.30 (m, 12H, aryl H’s), 6.96-6.99 (m, 12 H, aryl
H’s), 2.86 (m, 1H, PdN(Me)CH(CH2)3), 2.75 (m, 1H, aliphatic H of
pyrrolidene), 2.32-2.25 (m, 2H, CHH′PPh2), 2.14-2.10 (m, 2H,
C′HH′P′Ph2), 2.07 (d, J ) 3.3 Hz, 3H, NMe), 1.71-1.87 (m, 2H,
aliphatic H’s of pyrrolidene), 1.56 (m, 1H, aliphatic H of pyrrolidene),
1.21-1.37 (m, 2H, aliphatic H’s of pyrrolidene); 13C NMR (CDCl3,
125 MHz) δ 163.6 (br, Cipso and C′ipso of BPh2), 140.2 (d, J ) 26.0 Hz,
[Ph2BP2]Pd(N,C:η2-NiPr2CHCH3) (6). Compound 6 was prepared
with NiPr2Et using a procedure identical to that used for compound 7
1
(68.0 mg, 82% yield). H NMR (C6D6, 300 MHz) δ 7.32-7.50 (m,
14H, aryl H), 7.00-7.14 (m, 16H, aryl H), 3.38 (app quintet d, J )
1.8 and 6.9 Hz, 1H, PdNCHCH3), 2.95 (septet, J ) 6.0 Hz, 1H,
N(CHMe2)C′HMe2), 2.74 (septet, J ) 6.0 Hz, 1H, N(CHMe2)C′HMe2),
2.18-2.42 (m, 4H, CH2PPh2 and C′H2P′Ph2), 0.88 (d, J ) 6.9 Hz, 3H,
N(CHMe2)C′HMe2), 0.835 (d, J ) 6.3 Hz, 3H, N(CHMe2)C′HMe2),
0.73 (td, J ) 9.6 and 15.9 Hz, 3H, PdNCHCH3), 0.50 (d, J ) 5.4 Hz,
6H, N(CHMe2)C′HMe2); 13C NMR (CDCl3, 125 MHz) δ 163.8 (br,
Cipso and C′ipso of BPh2), 141.5 (d, J ) 24.3 Hz, Cipso of PPh2), 139.1
(d, J ) 35.4 Hz, C′ipso of PPh2), 137.9 (d, J ) 29.3 Hz, Cipso of P′Ph2),
137.1 (d, J ) 40.5 Hz, C′ipso of P′Ph2), 133.9 (d, J ) 31.5 Hz, Co of
PPh2), 133.3 (d, J ) 12.7 Hz, Co′ of PPh2), 133.0 (Co of BPh2), 132.6
(Co′ of BPh2), 132.1 (d, J ) 11.6 Hz, Co of P′Ph2), 132.0 (d, J ) 10.8
Hz, Co′ of P′Ph2), 129.0 (d, J ) 2.0 Hz, Cp of PPh2), 128.9 (d, J ) 1.9
Hz, Cp′ of PPh2), 128.4 (d, J ) 2.3 Hz, Cp of P′Ph2), 128.1 (d, J ) 1.5
Hz, Cp′ of P′Ph2), 128.0 (d, J ) 9.2 Hz, Cm of PPh2), 127.7-127.9 (m,
Cm′ of PPh2, Cm and Cm′ of P′Ph2), 126.2 (Cm of BPh2), 126.1 (Cm′ of
BPh2), 122.3 (Cp of BPh2), 122.2 (Cp′ of BPh2), 62.6 (dd, J ) 13.8 and
56.7 Hz, PdN(CHMe)), 54.6 (N(CHMe2)2), 50.4 (N(C′HMe2)2),
26.1 (N(CHMe2)2), 24.8 (N(CHMe′2)2), 23.4 (N(C′HMe2)2), 21.2
(N(C′HMe′2)2), 20.5-22 (br m, CH2PPh2 and C′H2PPh2), 15.9 (dd, J
) 2 and 5.3 Hz, PdN(CHMe)); 31P NMR (CDCl3, 121 MHz) δ 32.1
(d, J ) 32.0 Hz), 14.1 (d, J ) 32.0 Hz). Anal. Calcd for C46H52BNP2-
Pd: C, 69.23; H, 6.57; N, 1.76. Found: C, 68.62; H, 6.77; N, 1.77.
[Ph2BP2]Pd(N,C:η2-NEt2CHCH3) (7). Compound 1 (100.0 mg,
103.9 µmol) and NEt3 (290 µL, 2.081 mmol) were dissolved in 8 mL
of THF. The reaction mixture turned from yellow to a burnt orange.
After 3 h, all volatiles were removed in vacuo, and 10 mL of petroleum
ether was added to precipitate a peach-colored powder and wash away
the excess amine. The solids were dissolved in 15 mL of benzene and
flashed through a silica plug. The pale pink filtrate was dried in vacuo
1
to give light pink crystalline solids (78.0 mg, 98% yield). H NMR
Cipso of PPh2), 139.4-138.8 (m, C′ipso of PPh2, Cipso and C′ipso of P′Ph2),
(C6D6, 300 MHz) δ 7.62 (br d, J ) 6.9 Hz, Ho of BPh2), 7.11-7.41
133.6 (Co of BPh2), 133.2 (Co′ of BPh2), 133.0 (d, J ) 13.1 Hz, Co of
9
15820 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004

Aspects of an Amine Activation Process
A R T I C L E S
PPh2), 132.8 (d, J ) 13.7 Hz, Co′ of PPh2), 132.2 (d, J ) 13.1 Hz, Co
of P′Ph2), 131.9 (d, J ) 11.4 Hz, Co′ of P′Ph2), 128.9 (Cp of PPh2),
128.6 (Cp′ of PPh2), 128.5 (Cp of P′Ph2), 128.3 (Cp′ of P′Ph2), 128.0-
127.9 (m, Cm and Cm′ of PPh2, Cm and Cm′ of P′Ph2), 126.3 (Cm of BPh2),
126.2 (Cm′ of BPh2), 122.8 (Cp and Cp′ of BPh2), 69.5 (dd, J ) 14.6 and
61.8 Hz, PdN(Me)CH), 60.3 (N(CH2)), 45.5 (d, J ) 3.8 Hz, NMe),
31.6 (CH2 of pyrrolidene), 26.1 (CH2 of pyrrolidene), 20.2 (br m, CH2-
PPh2), 17.9 (br m, C′H2PPh2); 31P NMR (C6D6, 121 MHz) δ 32.7 (d,
J ) 39.5 Hz), 16.9 (d, J ) 39.5 Hz). Anal. Calcd for C43H44BNP2Pd:
C, 68.50; H, 5.88; N, 1.86. Found: C, 68.53; H, 5.96; N, 2.06.
[Ph2BP2]Pd(N,C:η2-NCH3CH(CH2)4) (10). Compound 10 was
prepared with N-methylpiperidene using a procedure identical to that
used for compound 7. Yield 54.1 mg, 69%. 1H NMR (C6D6, 300 MHz)
δ 7.67 (br d, 2H, J ) 6.9 Hz, Ho of BPhPh′), 7.61 (br d, 2H, J ) 6.0
Hz, Ho′ of BPhPh′), 7.31-7.39 (m, 8H, Ho and Ho′ of PPhPh′ and
P′PhPh′), 7.25 (t, J ) 6.9 Hz, 2H, Hp and Hp′ of BPhPh′), 7.10-7.14
(m, 4H, Hm and Hm′ of BPhPh′), 6.95-6.98 (m, 12 H, Hm, Hm′, Hp, and
6.5 Hz, 3H, N(CHMe2)2), 0.91 (d, J ) 6.0 Hz, 3H, N(CHMe2)2), 0.57
(d, J ) 6.5 Hz, 3H, N(CHMe2)2), 0.53 (d, J ) 7.0 Hz, 3H, N(CHMe2)2),
0.47 (t, J ) 7.0 Hz, 3H, PdNCHCH2CH3); 13C NMR (CDCl3, 125 MHz)
δ 164.1 (br, Cipso and C′ipso of BPh2), 140.7 (d, J ) 25.0 Hz, Cipso of
PPh2), 139.1 (d, J ) 28.0 Hz, C′ipso of PPh2), 138.3 (dd, J ) 2.0 and
37.0 Hz, Cipso of P′Ph2), 138.1 (d, J ) 36.2 Hz, C′ipso of P′Ph2), 133.5
(d, J ) 12.8 Hz, Co of PPh2), 133.1 (Co of BPh2), 132.9 (d, J ) 12.3
Hz, Co′ of PPh2), 132.7 (Co′ of BPh2), 132.63 (d, J ) 11.1 Hz, Co of
P′Ph2), 132.62 (d, J ) 11.9 Hz, Co′ of P′Ph2), 128.8 (d, J ) 2.3 Hz, Cp
of PPh2), 128.7 (d, J ) 1.9 Hz, Cp′ of PPh2), 128.6 (d, J ) 1.5 Hz, Cp
of P′Ph2), 128.2 (d, J ) 1.6 Hz, Cp′ of P′Ph2), 128.0 (d, J ) 9.3 Hz, Cm
of PPh2), 127.8-127.7 (m, Cm′ of PPh2, Cm and Cm′ of P′Ph2), 126.2
(Cm of BPh2), 126.1 (Cm′ of BPh2), 122.3 (Cp of BPh2), 122.2 (Cp′ of
BPh2), 70.2 (dd, J ) 14.1 and 56.5 Hz, PdN(CHEt)), 54.1 (N(CHMe2)2),
50.7 (N(C′HMe2)2), 26.1 (N(CHMe2)2), 26.0 (N(CHMe′2)2), 23.4
(N(C′HMe2)2), 23.1 (PdN(CHCH2Me)), 22.0 (br q, CH2PPh2), 21.5
(N(C′HMe′2)2), 20.3 (br q, C′H2PPh2), 15.3 (dd, J ) 7.8 and 8.5 Hz,
PdN(CHCH2Me)); 31P NMR (C6D6, 121 MHz) δ 32.3 (d, J ) 30.2
Hz), 13.7 (d, J ) 30.2 Hz). Anal. Calcd for C47H54BNP2Pd: C, 69.51;
H, 6.70; N, 1.72. Found: C, 69.23; H, 6.67; N, 1.42.
Hp′ of PPhPh′ and P′PhPh′), 2.98 (m, 1H, PdN(Me)CH(CH2)4), 2.29-
2.42 (m, 3H, CHH′PPh2 and CHH′P′Ph2), 2.11 (br d, J ) 12.3 Hz, 1H,
CHH′P′Ph2), 2.04 (d, J ) 3.3 Hz, 3H, NMe), 1.82 (app q, J ) 12. 3
Hz, 1H, aliphatic H of piperidene), 1.71 (m, 1H, aliphatic H of
piperidene), 1.07-1.15 (m, 4H, aliphatic H’s of piperidene), 0.86-
0.88 (m, 2H, aliphatic H’s of piperidene); 13C NMR (CDCl3, 125 MHz)
δ 163.4 (br, Cipso and C′ipso of BPh2), 140.7 (d, J ) 25.9 Hz, Cipso of
PPh2), 139.5 (d, J ) 28.2 Hz, C′ipso of PPh2), 139.1-139.5 (m, Cipso
and C′ipso of P′Ph2), 133.7 (Co of BPh2), 133.1 (Co′ of BPh2), 132.9 (d,
J ) 13.1 Hz, Co of PPh2), 132.2 (m, Co′ of PPh2, Co and Co′of P′Ph2),
128.8 (Cp of PPh2), 128.6 (Cp′ of PPh2), 128.5 (Cp of P′Ph2), 128.3 (Cp′
of P′Ph2), 127.9-128.0 (br m, Cm and Cm′ of PPh2, Cm and Cm′ of P′Ph2),
126.4 (Cm of BPh2), 126.1 (Cm′ of BPh2), 122.8 (Cp and Cp′ of BPh2),
66.9 (dd, J ) 13.8 and 60.4 Hz, PdN(Me)CH), 51.4 (N(CH2)), 49.7
(NMe), 22.6 (CH2 of piperidene), 22.4 (CH2 of piperidene), 20.7 (br
m, CH2PPh2), 17.8 (br m, C′H2PPh2), 16.6 (CH2 of piperidene); 31P
NMR (C6D6, 121 MHz) δ 32.7 (d, J ) 39.5 Hz), 16.9 (d, J ) 39.5
Hz). Anal. Calcd for C44H46BNP2Pd: C, 68.81; H, 6.04; N, 1.82.
Found: C, 67.93; H, 6.39; N, 1.41.
[Ph2BP2]Pd(C(NtBu)H)(CNtBu) (13). Compound 7 (32.0 mg, 41.6
µmol) and tert-butylisocyanide (15 µL, 248.5 µmol) were dissolved in
benzene. The reaction mixture turned bright yellow. After 1 h, all
volatiles were removed in vacuo, and the residue was taken up in a
benzene/petroleum ether solution. After storing at -30 °C, microcrys-
talline solids precipitated. The solids were collected and washed with
petroleum ether and diethyl ether (27.9 mg, 80% yield). IR(CH2Cl2,
1
KBr): ν(HCdNtBu) ) 1608 cm-1, ν(CtN) ) 2195 cm-1; H NMR
(C6D6, 300 MHz) δ 8.75 (dd, J ) 38.8 and 27.5 Hz, 1H, Pd(CNtBu)-
H), 7.57-7.56 (m, 8H, aryl H’s), 7.31 (t, J ) 9.0 Hz, 4H, aryl H’s),
7.21-7.08 (m, 6H, aryl H’s), 6.95-6.87 (m, 12H, aryl H’s), 2.28 (d,
J ) 14.7 Hz, 2H, CH2PPh2), 2.13 (d, J ) 13.8 Hz, 2H, CH′2PPh2),
0.95 (s, 9H, Pd(CNtBu)H), 0.58 (s, 9H, Pd(CNtBu)); 13C NMR (CDCl3,
125 MHz) δ 179.3 (dd, J ) 8.6 and 131 Hz, Pd(CNtBu)H), 163.4 (v
br, Cipso of BPh2), 137.4 (d, J ) 35.6 Hz), 133.0 (dd, J ) 10 and 25
Hz), 132.8, 132.4, 130.5, 129.0 (dd, J ) 2 and 69 Hz), 127.8 (dd, J )
11 and 34 Hz), 126.7, 126.4, 123.0, 122.7, 29.9, 29.2, 19 (br m), 16
(br m); 31P NMR (C6D6, 121 MHz) δ 23.6 (d, J ) 55.8 Hz), 13.1 (d,
J ) 55.8 Hz). Anal. Calcd for C48H53BNP2Pd: C, 68.87; H, 6.38; N,
3.35. Found: C, 69.19; H, 6.64; N, 3.06.
[Ph2BP2]Pd(N,C:η2-NiPr2CH2) (11). Compound 1 (100.4 mg, 104.4
µmol) and diisopropyl(2-propenyloxymethyl)amine (357.4 mg, 2.088
mmol) were dissolved in THF. The reaction mixture turned from yellow
to orange-red immediately but faded to pale yellow after stirring for 5
min. After 1 h, all volatiles were removed in vacuo, and petroleum
ether was added to precipitate the powder and wash away the excess
amine. The solids were dissolved in benzene and flashed through a
silica plug. The filtrate was dried in vacuo to give pale yellow solids
(Ph2SiP2)Pd(OTf)2 (14). A THF solution of AgOTf (455.0 mg, 1.77
mmol) was added dropwise to a THF solution of Ph2Si(CH2PPh2)2-
PdCl2 (671.2 mg, 0.886 mmol). The reaction turned bright yellow and
cloudy. The reaction mixture was filtered through Celite, and the filtrate
was concentrated in vacuo to a thick yellow residue. Upon addition of
petroleum ether and vigorous stirring, 14 was isolated as a yellow
1
(63.7 mg, 78% yield). H NMR (C6D6, 300 MHz) δ 7.55 (br d, J )
6.9 Hz, Ho of BPh2), 7.32-7.44 (m, 8H, aryl H’s), 7.08-7.18 (m, 6H,
aryl H’s), 6.98-7.02 (m, 12H, aryl H’s), 2.61 (m, 2H, N(CHMe2)2),
2.51 (app t, J ) 5.1 Hz, 2H, PdNCH2), 2.29 (m, 4 H, CH2PPh2), 0.70
(d, J ) 6.9 Hz, 6H, N(CHMeMe′)2), 0.59 (d, J ) 6.0 Hz, 6H,
N(CHMeMe′)2); 13C NMR (C6D6, 75 MHz) δ 163.8 (v br, Cipso of BPh2),
140.6 (d, J ) 28.7 Hz, Cipso of PPh2), 139.8 (d, J ) 38.3 Hz, Cipso of
P′Ph2), 133.9 (Co of BPh2), 133.4 (d, J ) 12.6 Hz, Co of PPh2), 133.0
(d, J ) 12.0 Hz, Co of P′Ph2), 128.3-129.1 (m, Cp and Cm of PPh2 and
P′Ph2), 127.0 (Cm of BPh2), 123.2 (Cp of BPh2), 54.5 (d, J ) 2.0 Hz,
PdN(CHMe2)2), 50.1 (dd, J ) 56 and 13 Hz, NCH2), 24.1 (N(CH-
MeMe′)2), 21.7 (N(CHMeMe′)2), 20.3 (br m, CH2PPh2 and CH2P′Ph2);
31P NMR (C6D6, 121 MHz) δ 31.5 (d, J ) 39.3 Hz), 16.3 (d, J ) 39.3
Hz). Anal. Calcd for C45H50BNP2Pd: C, 68.93; H, 6.43; N, 1.79.
Found: C, 68.66; H, 6.73; N, 1.64.
1
powder in 98% yield (854 mg). H NMR (d6-acetone, 300 MHz) δ
7.89 (dd, J ) 7.5 and 12.9 Hz, 8H, Ho of PPh2), 7.67 (br t, J ) 7.5 Hz,
4H, Hp of PPh2), (td, J ) 2.1, 7.5 Hz, 8H, Hm of PPh2), 7.38-7.33 (m,
6H, Ho and Hp of SiPh2), 7.20 (t, J ) 7.5 Hz, 4H, Hm of SiPh2), 3.01
(dd, J ) 6.9 and 17.0 Hz, 4H, CH2SiPh2); 13C NMR (CDCl3, 125 MHz)
δ 135.0 (Co of SiPh2), 134.5 (m, Co of PPh2), 134.3 (Cp of PPh2), 133.0
(m, Cipso of PPh2), 131.2 (Cp of SiPh2), 130.4 (m, Cm of PPh2), 129.0
(Cm of SiPh2), 127.4 (d, J ) 59.9 Hz, CF3), 6.6 (m, CH2PPh2) (Cipso of
SiPh2 was not clearly visible); 31P NMR (d6-acetone, 121 MHz) δ 36.4;
19F NMR (d6-acetone, 282 MHz) δ -74.3. Anal. Calcd for C40H34F6O6P2-
PdS2Si: C, 48.76; H, 3.48; N, 0. Found: C, 48.36; H, 3.46; N, <0.05.
(Ph2SiP2)PdCl2 (15). A THF solution of PdCl2(NCPh)2 (398.3 mg,
1.038 mmol) was added rapidly to a stirring THF solution of Ph2Si-
(CH2PPh2)2 (603.0 mg, 1.038 mmol). After stirring for 1 h, the solution
was filtered through Celite and dried in vacuo. The residue was washed
thoroughly with Et2O/CH3CN and Et2O/THF (>4:1) and dried under
vacuum overnight to give a pale lime-yellow powder (768.3 mg, 89%).
1H NMR (CDCl3, 300 MHz) δ 7.62 (dd, J ) 7.5 and 12.3 Hz, 8H, Ho
of PPh2), 7.39 (dd, J ) 6.9 and 8.4 Hz, 4H, Hp of PPh2), 7.34-7.25
(m, 10H, Hm of PPh2 and Hp of SiPh2), 7.12 (t, J ) 7.5 Hz, 4H, Hm of
[Ph2BP2]Pd(N,C:η2-NiPr2CHCH2CH3) (12). Compound 12 was
prepared with NiPr2 Pr using a procedure identical to that used for
n
compound 7 (65.8 mg, 78% yield). 1H NMR (C6D6, 500 MHz) δ 7.37-
7.51 (m, 14H, aryl H), 6.96-7.19 (m, 16H, aryl H), 3.19 (m, 1H,
PdNCHEt), 2.86 (d septet, J ) 3.0 and 6.5 Hz, 1H, N(CHMe2)C′HMe2),
2.81 (septet, J ) 6.5 Hz, 1H, N(CHMe2)C′HMe2), 2.19-2.38 (m, 4H,
CH2PPh2 and C′H2P′Ph2), 1.28 (m, 2H, PdNCHCH2Me), 1.00 (d, J )
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15821

A R T I C L E S
Scheme 1
Lu and Peters
SiPh2), 7.20 (d, J ) 6.9 Hz, 4H, Ho of SiPh2), 2.11 (dd, J ) 5.4 and
15.0 Hz, 4H, CH2SiPh2); 13C NMR (CDCl3, 75 MHz) δ 133.8 (Co of
SiPh2), 133.7 (app t, J ) 6.0 Hz, Co of PPh2), 132.9 (m, Cipso of PPh2),
131.6 (Cipso of SiPh2), 131.5 (Cp of PPh2), 130.2 (Cp of SiPh2), 128.4
(app t, J ) 6.0 Hz, Cm of PPh2), 128.3 (Cm of SiPh2), 9.4 (dd, J ) 13.1
and 15.8 Hz, CH2PPh2); 31P NMR (CDCl3, 121 MHz) δ 22.66. Anal.
Calcd for C38H34Cl2P2PdSi: C, 60.21; H, 4.52; N, 0. Found: C, 59.34;
H, 4.52; N, <0.05.
NiPr2C(O)CH3 with LiAlD4 under standard conditions.17 The amine was
extracted with petroleum ether and filtered through a silica plug to give
spectroscopically pure NiPr2CD2CH3. The reaction was carried out in
triplicate, giving kobs ) 9.36, 9.65, and 10.57 × 10-5 s-1
.
Rate Dependence on Amine Concentration. Seven kinetic runs
were conducted in which the concentration of NiPr2Et was varied from
0.16 to 1.56 M (3 to 30 equiv relative to 1). The total volume of the
samples was kept constant at 0.6 mL. The following values for kobs
(×10-4 s-1) were obtained: 1.21 (3 equiv of amine), 1.80 (5 equiv of
amine), 3.03 (10 equiv of amine), 4.43 (15 equiv of amine), 5.49 (20
equiv of amine), 6.96 (25 equiv of amine), and 8.41 (30 equiv of amine).
Activation Parameters (∆Hq, ∆Sq) for the Overall Amine Activa-
tion. A series of reactions was conducted over a 35 °C temperature
range (0- 35.3 °C). The NMR probe temperatures were measured using
an anhydrous MeOH standard. The following values for kobs[THF]i2/
[NiPr2Et]i (×10-2 s-1) were obtained: 0.37 (0.0 °C), 0.94 (7.5 °C),
1.76 (13.3 °C), 1.82 (13.3 °C), 2.81 (17.1 °C), 5.41 (23.1 °C), 9.16
(28.4 °C), and 18.41 (35.3 °C). The data provided a satisfactory linear
fit in the Eyring plot, from which the activation parameters were
calculated, ∆Hq ) 17.9 ( 0.2 kcal/mol, ∆Sq ) -4 ( 1 eu. Regression
analyses of the data were performed using Microsoft Excel, and the
activation parameters are reported with 95% confidence.
VT-NMR Studies. Complex 1 (20.0 mg, 20.8 µmol) was dissolved
in 0.6 mL of toluene-d8, transferred to a J-Young NMR tube, and chilled
inside a glovebox cold-well. Triethylamine (145 µL, 1.04 mmol) was
then added to the cold solution. The tube was inserted into an NMR
probe at -50 °C. As the temperature was slowly raised to 0 °C, the
reaction was monitored by 1H and 31P NMR spectroscopy. Conversion
to the palladacycle 7 was relatively clean (∼80% crude yield by 31P
NMR spectroscopy).
[((Ph2SiP2)Pd)2][OTf]2 (16). Diisopropylethylamine (54 µL, 0.310
mmol) was added dropwise to a THF solution (8 mL) of 14 (100.5
mg, 0.102 mmol). Within 5 min the solution turned from yellow to
burnt orange. The reaction was monitored by 31P NMR. After 20 h all
volatiles were removed in vacuo. After washing with Et2O, the residue
was redissolved in CH3CN and filtered through Celite. Et2O was added
to the solution to precipitate orange crystalline solids. The solids were
collected, washed with Et2O/CH3CN (9:1), and dried in vacuo to afford
1
59.3 mg of 16 (70% yield). H NMR (CDCl3, 300 MHz) δ 7.8-7.6
(m, 8H, aryl H’s), 7.4-7.0 (m, 44H, aryl H’s), 6.9 (d, 8H, aryl H’s),
2.6 (d, 4H, CH2SiPh2), 1.7 (d, 4H, C′H2Si′Ph2); 13C NMR (CDCl3, 125
MHz) δ 134.9 (Cp of PPh2), 134.2-134.1 (m, Co of PPh2 and Cp of
P′Ph2), 133.3 (d, J ) 11 Hz, Cm of PPh2), 133.1 (m, Cipso of PPh2),
131.5 (Cipso of SiPh2), 130.6 (d, J ) 11 Hz, Cm of P′Ph2), 129.9 (Cp of
SiPh2), 129.3 (m, Cipso of P′Ph2), 128.8 (m, Co of P′Ph2), 128.3 (Cm of
SiPh2), 108.0 (br m, CF3), 13.8 (CH2SiPh2), 12.5 (C′H2SiPh2); 19F NMR
(CDCl3, 282 MHz) δ -74.7; 31P NMR (CDCl3, 121 MHz) δ 22.2 (d,
J ) 50 Hz), -0.2 (d, J ) 50 Hz). Anal. Calcd for C78H68F6O6P4Pd2S2-
Si2: C, 56.02; H, 4.10; N, 0. Found: C, 55.93; H, 3.83; N, <0.05.
[(Ph2SiP2)Pd(2-phenylpyridine)OTf][OTf]. o-Phenylpyridine was
added dropwise to a THF solution of 14. This compound was generated
in situ for the purpose of this experiment and has not been isolated in
crystalline form. 31P NMR (THF, 121 MHz) δ 37.1 (d, J ) 14.5 Hz),
22.8 (d, J ) 14.5 Hz); 19F NMR (THF, 282 MHz) δ -75.3, -75.9.
Standard Kinetic Protocol: Measurement of Reaction Rates.
Complex 1 (30.0 mg, 31.2 µmol) was dissolved in 0.6 mL of THF,
transferred to a J-Young NMR tube, and chilled inside a glovebox cold-
well below -50 °C. Diisopropylethylamine (109 µL, 626 µmol) was
then added to the cold solution. The tube was inserted into an NMR
probe at 23 °C, and the temperature was allowed to equilibrate for 10
min. The decay of complex 1 was monitored by 31P{1H} NMR
spectroscopy against an internal integration standard, which consisted
of a sealed capillary tube containing a 1.6 M solution of Ph3PdO in
CH2Cl2. Spectra were taken at intervals of ca. 3 min for ca. 3.5 half-
lives. Besides the integral standard, the only observable peaks were
due to 1 and the palladacycle 6. The data were satisfactorily fit to an
exponential decay, and the first-order rate constant was obtained from
the ln(1) versus time plot. The reaction was carried out in triplicate,
giving kobs ) 5.63, 5.49, and 5.76 × 10-4 s-1. This protocol was also
used in determining KIE data, measuring the rate dependence on amine
concentration and generating the Eyring plot.
Results
Synthesis and Structural Characterization of [[Ph2BP2]-
Pd(THF)2][OTf] (1) and [Ph2BP2]Pd(THF)(OTf) (1′). A
reactive [Ph2BP2]Pd(II) system was required to carry out the
C9H bond activation studies described herein. A labile triflate
derivative, [Ph2BP2]Pd(OTf)(L), was targeted, and the following
protocol (Scheme 1) afforded the species of interest. Mixing
[Ph2BP2][NBu4] and PdCl2(NCPh)2 in THF afforded a mixture
of two products, [[Ph2BP2]PdCl2][NBu4] (2) and {[Ph2BP2]Pd-
(µ-Cl)}2 (3). The mixture of 2 and 3 was converted completely
to {[Ph2BP2]Pd(µ-Cl)}2 (3) by addition of a slight excess of
NaBPh4 to the reaction mixture (Scheme 1). Crystals of 3 were
grown from THF/CH2Cl2 solutions at -30 °C in 73% overall
yield (based upon PdCl2(NCPh)2). Dimeric 3 was then converted
to the more reactive species, [[Ph2BP2]Pd(THF)2][OTf] (1), by
(17) March, J. AdVanced Organic ChemistrysReactions, Mechanisms, and
Structure; John Wiley & Sons: New York, 1992; p 448.
KIE Studies. NiPr2CD2CH3 was prepared by reducing the amide
9
15822 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004

Aspects of an Amine Activation Process
A R T I C L E S
complex in the presence of NiPr2Et, which served as a formal
H+ acceptor to produce [HNiPr2Et][OTf].20 In the present case,
adding an excess of NiPr2Et to a benzene solution of 1 at 25 °C
indeed generated a stoichiometric equivalent of [HNiPr2Et][OTf].
However, the major palladium-containing species generated
(>70% of the mixture by 31P NMR spectroscopy) was the
previously reported palladium(I) dimer, {[Ph2BP2]Pd}2 (4, eq
1).21 A second diamagnetic species was also evident (31P NMR),
but none of the anticipated aryl transfer product, [Ph2BP2]Pd-
(Ph)(THF), was detected.22 Several observations from the
literature suggest that [Ph2BP2]Pd(Ph)(THF) may be unstable.
For example, it is known that biphenyl and other biaryls can
be liberated from palladium(II) aryl species.23 Biphenyl has also
been derived directly from a benzene C-H activation process
mediated by palladium(II).24 Related aryl coupling processes
are known for phosphine-supported platinum(II) systems.25,7b
In the present [Ph2BP2]Pd(II) system, however, benzene-derived
biphenyl could not be identified (1H NMR, GC9MS).
Figure 2. (A) Displacement ellipsoid representation (50%) of complex 1′.
Selected bond distances (Å) and angles (deg): Pd-O1, 2.157(2); Pd-O2,
2.155(2); Pd-P1, 2.2378(8); Pd-P2, 2.2309(9); O1-Pd-O2, 86.52(7); P1-
Pd-P2, 89.16(3); O1-Pd-P2, 92.33(6); O2-Pd-P1, 91.90(5). (B) Dis-
placement ellipsoid representation (50%) of complex 5. Selected bond
distances (Å) and angles (deg): Pd-O1, 2.195(2); Pd-N, 2.139(3); Pd-
P1, 2.2718(9); Pd-P2, 2.2503(9); O1-Pd-N, 89.85(9); P1-Pd-P2, 88.43-
(3), O1-Pd-P1, 88.00(6); N-Pd-P2, 93.87(7). Hydrogen atoms and
solvent molecules were omitted for clarity.
addition of AgOTf in THF. As indicated by its formula, complex
1 exists in THF solution predominantly as a bis(solvento) cation,
which is consistent with the following: (a) a single resonance
(s, δ ) 48.2 ppm) at 23 °C in the 31P NMR spectrum, (b)
1
integrations consistent with 2 equiv of THF in the H NMR
spectrum, and (c) combustion analysis data. Nonetheless, an
X-ray diffraction analysis of a single-crystal grown from a
purified sample in THF at -30 °C revealed a zwitterionic,
mono-THF adduct, ([Ph2BP2]Pd(THF)(OTf)) (1′) (Scheme 1,
Figure 2A), and an unbound THF molecule. The THF coligands
of 1 are appreciably labile and in rapid exchange with the
competitive triflate ligand. This exchange process, 1 H 1′ +
THF, was observed by 31P VT-NMR spectroscopy in CD2Cl2
solution. A broad signal is evident at 47 ppm, indicative of
fluxional behavior at ambient temperature. Upon cooling the
sample to -85 °C, two signals become fully resolved with a
peak separation of 916 Hz. These peaks coalesce at -42.3 °C,
and a barrier of 16.2 kcal/mol can be calculated for the THF
and triflate exchange at -42.3 °C.18
The crystal structure of 1′ shows an essentially square planar
complex with bond angles that deviate only slightly from their
ideal values. The O(1)-Pd-O(2) and O(1)-Pd-P(2) bond
angles are 86.52(7)° and 92.33(6)°, respectively. The Pd-O
bond distances for each of the oxygen donors are identical
(2.157(2) and 2.155(2) Å for the THF and triflato donors,
respectively). Complex 1′ has few structurally authenticated
congeners in the palladium literature. The most comparable
structures are those of the complexes [(dppp)Pd(H2O)(X)][X]
(where X ) OTf and OTs) reported separately by Stang and
Toniolo.19 The Pd-O bond distances in 1′, and also its Pd-P
bond distances (2.2378(8) and 2.2309(9) Å), compare favorably
with these examples.
To further explore whether 1 might activate aryl C-H bonds,
o-phenylpyridine was tested as a substrate, our explanation being
that a favorable chelate stabilization might encourage metalation
of the arene ring. Several platinum(II) and palladium(II) halide
and pseudo-halide precursors are known to react with o-
phenylpyridine in the presence of a sacrificial base to produce
cyclometalated products.26 Exposure of 1 to either stoichiometric
or excess o-phenylpyridine afforded a single product, the adduct
complex [Ph2BP2]Pd(o-phenylpyridine)(OTf) (5), which was
verified by X-ray crystallography. Complex 5 is quite hindered,
as can be gleaned from its solid-state crystal structure (Figure
2B). At room temperature in solution, its 1H NMR resonances
are broad, presumably due to hindered rotation about the Pd-N
bond. These peaks sharpen upon cooling to -40 °C, and the
expected four resonances for the ligand methylene protons
become well resolved. Relative to 1′, the Pd-OTf interaction
of 5 is modestly elongated (2.195(2) versus 2.155(2) Å), also
true of the Pd-P bond distances (2.25-2.27 versus 2.23-2.34
Å). These relative bond lengths reflect the steric congestion
suffered by 5, which nonetheless is unexpectedly robust. For
example, no reaction occurred when 5 was incubated in THF
at 80 °C in the presence of either excess o-phenylpyridine or
excess K2CO3.
(20) Harkins, S. B.; Peters, J. C. Organometallics 2002, 21, 1753-1755.
(21) Lu, C. C.; Peters, J. C. J. Am. Chem. Soc. 2002, 124, 5272-5273.
(22) We note that [Ph2BP2]Pd(Ph)(CH3CN) can be generated in solution and
spectroscopically identified by reaction of {[Ph2BP2]Pd(µ-Cl)}2 with
PhMgBr in the presence of acetonitrile.
Reactivity of 1 with Arene Substrates. The possibility for
complex 1 to mediate aryl C-H bond activation was briefly
examined by dissolving 1 in benzene solution in the presence
of the sterically encumbered amine base NiPr2Et. Our lab had
observed phenyl transfer from benzene to a platinum(II) triflate
(23) Yagyu, T.; Hamada, M.; Osakada, K.; Yamamoto, T. Organometallics 2001,
20, 1087-1101.
(24) Ackerman, L. J.; Sadighi, J. P.; Kurtz, D. M.; Labinger, J. A.; Bercaw, J.
E. Organometallics 2003, 22, 3884-3890.
(25) Konze, W. V.; Scott, B. L.; Kubas, G. J. J. Am. Chem. Soc. 2002, 124,
12550-12556.
(18) Gasparro, F. P.; Kolodny, N. H. J. Chem. Educ. 1977, 54, 258-261.
(19) (a) Stang, P. J.; Cao, D. H.; Poulter, G. T.; Arif, A. M. Organometallics
1995, 14, 1110-1114. (b) Benetollo, F.; Bertani, R.; Bombieri, G.; Toniolo,
L. Inorg. Chim. Acta 1995, 233, 5-9.
(26) (a) Constable, E. C.; Thompson, A. M. W. C.; Leese, T. A.; Reese, D. G.
F.; Tocher, D. A. Inorg. Chim. Acta 1991, 182, 93-100. (b) Wong-Foy,
A. G.; Henling, L. M.; Day, M.; Labinger, J. A.; Bercaw, J. E. J. Mol.
Catal. A: Chem. 2002, 189, 3-16.
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15823

A R T I C L E S
Lu and Peters
Activation of Trialkylamines by Complex 1. Treating 1 with
20 equiv of NiPr2Et in THF solution at ambient temperature
resulted in the gradual consumption of starting material and the
rise of a single new palladium product, along with a stoichio-
metric equivalent of [HNiPr2Et][OTf]. The new palladium
species exhibited resonances in the 31P NMR spectrum that
compared well to a minor product that was observed when 1
was exposed to NiPr2Et in benzene solution (vide supra). Single
crystals of this species were grown from a THF/petroleum ether
solution, and an X-ray diffraction study revealed it to be the
three-membered palladacycle [Ph2BP2]Pd(N,C:η2-NiPr2CHCH3)
(6, eq 2). This formulation is consistent with the complex’s
spectral data in solution. For example, the 13C{1H} NMR
spectrum features a signature resonance at 62.6 ppm (dd, 2JC-Pcis
To induce reactivity of primary C-H bonds, the reactivity
of 1 with NCy2Me and NiPr2Me was examined. A steric bias
might in these cases be expected to direct the palladium center
toward the methyl position. However, the predominant byprod-
uct (70-80%) for these substrates was the familiar palladium-
(I) dimer, {[Ph2BP2]Pd}2 (4) (31P NMR).
Addition of diisopropyl(2-propenyloxymethyl)amine to 1
effected formal C-O rather than C-H bond cleavage to
generate [Ph2BP2]Pd(N,C:η2-NiPr2CH2) (11, eq 3) in good yield.
Allyl ethers are known to react with Pd(0) sources to generate
palladium allyl products via formal C-O cleavage.27 The high-
yield isolation of 11 shows quite clearly that [Ph2BP2]Pd(N,C:
η2-NR2CH2) are viable products of amine activation, and it is
therefore puzzling that a similar product is not formed when 1
is treated with NiPr2Me (vide infra). To determine the fate of
the allyl ether group, an NMR-scale reaction was monitored in
THF-d8 with 3 equiv of diisopropyl(2-propenyloxymethyl)-
amine. A 1H NMR spectrum of the crude product mixture
suggested the presence of an aldehyde byproduct which could
be isolated by vacuum transfer and clearly identified as acrolein
by comparison of its 1H NMR spectrum to that of an authentic
sample.
2
) 13.8 Hz, JC-Ptrans ) 56.7 Hz) due to the coupling of an
organometallic carbon atom to two nonequivalent phosphorus
nuclei.
The scope and selectivity of the amine activation process was
explored (Scheme 2). In general, the activation process was
found to proceed for trialkylamine substrates that feature
secondary C-H bonds adjacent to the N atom. Palladacycle
products that feature three-membered rings appear to be formed
exclusively, with no apparent tendency to form larger metal-
lacycles. To test the reaction’s selectivity, the amine NCy2Et
was chosen as a substrate as it contains both tertiary and
secondary C-H bonds that can undergo activation. Treatment
of 1 with excess NCy2Et in THF resulted in clean conversion
to only one palladacycle, [Ph2BP2]Pd(N,C:η2-NCy2CHMe) (8),
derived from formal cleavage of a secondary C-H bond of the
ethyl group. The substrates N-methylpyrrolidene and N-meth-
ylpiperidene were also examined to compare the reactivity of
primary versus secondary C-H bonds. Again, activation at the
secondary C-H position occurred exclusively, affording the
bicyclic palladacycles [Ph2BP2]Pd(N,C:η2-NMeCH(CH2)3) (9)
and [Ph2BP2]Pd(N,C:η2-NMeCH(CH2)4) (10), respectively.
Scheme 2
Molecular Structures of the Palladacycles 8 and 9. X-ray
diffraction studies of the palladacycles were pursued to gain
insight into the structural nature of the Pd-N-C core. Crystals
of 6 and 11 suffered from disorder of the iminium group,
R2CdNiPr2, which sits in two orientations. Derivatives 8 and 9
provided more reliable structural parameters (Figure 3, Table
1). In each structure the bond angles around palladium are
largely distorted from 90°, though all four atoms coordinated
to palladium lie in a well-defined plane. Specifically, for 8 and
9, the N-Pd-CR angles are acute (39.7(1)° and 38.3(2)°,
9
15824 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004

Aspects of an Amine Activation Process
A R T I C L E S
Figure 3. Displacement ellipsoid representations (50%) of [Ph2BP2]Pd(N,C:η2-NCy2CHMe) (8) and [Ph2BP2]Pd(N,C:η2-NMeCH(CH2)3) (9). Hydrogen
atoms have been omitted for clarity.
Table 1. Selected Bond Distances (Å) and Angles (°) for
Complexes 8 and 9
In this latter view a palladium(II) center is formally coordinated
to an alkyl ligand with a â-amine donor occupying the fourth
site.
complex 8
complex 9
Pd-C(39)
2.037(3)
2.180(3)
2.3641(9)
2.2549(9)
1.438(4)
1.502(4)
1.484(4)
93.60(3)
122.31(7)
104.04(9)
39.7(1)
Pd-C(43)
2.048(4)
2.104(3)
2.3354(8)
2.2945(8)
1.361(5)
1.460(5)
1.420(6)
93.49(3)
120.9(1)
107.6(2)
38.3(2)
Close examination of the structural parameters of 8 and 9
Pd-N
Pd-N
does not unambiguously distinguish between the two resonance
forms. The Pd-CR distances observed for 8 and 9 are unusually
short for cis-bis(phosphine)Pd complexes. Their respective
distances of 2.037(3) and 2.048(4) Å are just below both the
minimum Pd-C(sp3) and Pd-C(sp2-alkene) bond distances
reported for palladium supported by two phosphines in cis
relation. The N-CR bond lengths for complexes 8 and 9 are
between those of free iminium cations (ca. 1.32 Å) and N-C
single bonds (ca. 1.50 Å) but are also quite different from one
another (1.438(4) Å for 8 and 1.361(5) Å for 9). The latter N-C
bond distance is the shortest thus far reported for mononuclear
metallaaziridines of related structure (range 1.392-1.548 Å,
mean of 1.440 Å, SD ) 0.03 Å, Cambridge Structural
Database29). The bond angles around the nitrogen atom are
nearly 120° for complex 8,30 suggesting the nitrogen atom is
perhaps better formulated as sp2- rather than sp3-hybridized.
Probing the Reversibility of Palladacycle Formation. To
probe whether the formation of the palladacycles [Ph2BP2]Pd-
(N,C:η2-NR2CHR′) might be reversible, we examined whether
deuterium would scramble into the palladacycle aliphatic
positions upon incubation in the presence of a deuterated
ammonium salt. Exposure of 1 to iPr2NCH2CH2CH3 proceeded
in similar fashion to provide the expected palladacycle [Ph2-
BP2]Pd(N,C:η2-iPr2NCHCH2CH3) (12). A homogeneous solu-
tion of complex 12 and 15 equiv of [DNiPr2CD2Et][OTf] was
stirred at 25 °C for several days. Complex 12 was stable under
these conditions: no degradation was evident in the 31P NMR
spectrum of the reaction solution. Upon examination of the 2H
NMR spectrum, however, it was evident that deuterium had
scrambled into the R- and â-carbon positions of the iminium
ligand in an approximate ratio of 1:5, respectively (eq 4).
Pd-P(1)
Pd-P(1)
Pd-P(2)
Pd-P(2)
N-C(39)
N-C(43)
N-C(41)
N-C(39)
N-C(47)
N-C(40)
P(1)-Pd-P(2)
P(1)-Pd-N
P(2)-Pd-C(39)
N-Pd-C(39)
C(39)-N-C(41)
C(39)-N-C(47)
C(41)-N-C(47)
P(1)-Pd-(2)
P(1)-Pd-N
P(2)-Pd-C(43)
N-Pd-C(43)
C(43)-N-C(39)
C(43)-N-C(40)
C(39)-N-C(40)
115.8(3)
117.9(3)
119.3(2)
117.4(4)
108.0(4)
117.0(4)
Figure 4. Limiting resonance forms for the palladacycles.
respectively) and the P(1)-Pd-N angles (122.3° and 120.9°)
are significantly larger than the P(2)-Pd-CR angles (104.0°
and 107.6°).
Palladacycles akin to 8 and 9 have not to our knowledge been
described. However, metallaaziridines featuring other transition
metals are well established, though they are formed from
different methods of preparation.28 For example, a family of
nickel aziridines (PR3)(X)Ni(N,C:η2-CH2NMe2) that are struc-
turally related to the palladacycles described here were first
reported nearly three decades ago.28d Two limiting resonance
forms are to be considered with respect to the Pd-N-C core
(Figure 4).28a,d The first emphasizes a Pd(0) form of the complex
with an anionic {[Ph2BP2]Pd}- fragment coordinated by an
iminium cation (RHCdNR2+). This formulation is analogous
to the many olefin complexes of palladium(0) that are known.
The other limiting resonance form assigns the alkylamine
functionality as a three-electron LX-type organometallic ligand.
(27) See: Yamamoto, T.; Akimoto, M.; Saito, O.; Yamamoto, A. Organome-
tallics 1986, 5, 1559-1567 and references therein.
In a related experiment we examined whether iminium ligand
substitution would proceed in the presence of an ammonium
(28) For pertinent examples, see: (a) Crawford, S. S.; Knobler, C. B.; Kaesz,
H. D. Inorg. Chem. 1977, 16, 3201-3207. (b) Matsumoto, M.; Nakatsu,
K.; Tani, K.; Nakamura, A.; Otsuka, S. J. Am. Chem. Soc. 1974, 96, 6777-
6778. (c) Abel, E. W.; Rowley, R. J.; Mason, R.; Thomas, K. M. J. Chem.
Soc., Chem. Commun. 1974, 72. (d) Sepelak, D. J.; Pierpont, C. G.;
Barefield, E. K.; Budz, J. T.; Poffenberger, C. A. J. Am. Chem. Soc. 1976,
98, 6178-6185.
(29) Allen, F. H. Acta Crystallogr. 2002, B58, 380-388.
(30) For complex 9, the bond angles around nitrogen are also near 120° with
the exception of the pyrrolidene ring angle, which is restrained by the ring.
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15825

A R T I C L E S
Lu and Peters
salt and a trialkylamine. The palladacycle [Ph2BP2]Pd(N,C:η2-
NiPr2CHCH3) 6 was incubated with 15 equiv of [HNEt3][OTf]
and 15 equiv of NEt3 in THF at 80 °C. Under these conditions,
none of the palladacycle [Ph2BP2]Pd(N,C:η2-NEt2CHCH3) 7
could be detected after several days (eq 5). If 6 could undergo
protonation to produce [Ph2BP2]Pd(NiPr2Et2)(OTf) at some
reasonable rate, then we would expect a detectable degree of
exchange to occur (i.e., [Ph2BP2]Pd(NiPr2Et)(OTf) + NEt3 f
[Ph2BP2]Pd(NEt3)(OTf) + NiPr2Et).
Figure 5. Eyring plot for the reaction of 1 with 20-fold NiPr2Et from 273.2
to 308.5 K, ∆Hq ) 17.9 ( 0.2 kcal/mol, ∆Sq ) -4 ( 1 eu, ∆Gq ) 19.0
( 0.3 kcal/mol at 23.1 °C.
Addition of tert-Butyl Isocyanide and Sodium Cyanide to
[Ph2BP2]Pd(N,C:η2-NEt2CHCH3) (7). The addition of excess
tert-butyl isocyanide to 7 resulted in a single new palladium
complex, [Ph2BP2]Pd(C(H)dNtBu)(CNtBu) (13), and concomi-
tant expulsion of the enamine Et2NC(H)dCH2 (eq 6). Complex
13 is characterized by two intense vibrations in its infrared
spectrum (ν(HCdNtBu) ) 1608 cm-1, ν(CtNtBu) ) 2195
cm-1; CH2Cl2/KBr) and a signature 1H NMR resonance at 8.75
ppm for the iminoformyl proton Pd(C(H)dNtBu).31 The release
of the enamine Et2NC(H)dCH2 was confirmed by comparing
the crude reaction spectrum with that of an independently
prepared sample.
Figure 6. Rate dependence on [NiPr2Et]. Reaction conditions: [1] ) 0.052
M; [NiPr2Et] is varied between 0.16 and 1.56 M at 296 K in THF (0.6 mL
total solution volume).
1 and the product 6 were the only two species observed. The
rate constant, kobs ) 5.6(1) × 10-4 s-1, was obtained as the
average of three independent runs, and ∆Gq was calculated to
be 19.0 ( 0.3 kcal/mol at 23.1 °C. Perhaps a more meaningful
constant is k′, where kobs ) k′[NiPr2Et]/[THF]2 (vide infra). By
assuming constant concentrations of NiPr2Et and THF, a value
of k′ ) 5.6(1) × 10-2 s-1 is obtained ([NiPr2Et]i ) 0.88 M and
[THF]i ) 9.33 M). The temperature dependence of the reaction
rate was explored between 0 and 35.3 °C. An Eyring plot of
these data is provided in Figure 5, from which the activation
parameters ∆Hq ) 17.9 ( 0.2 kcal/mol and ∆Sq ) -4 ( 1 eu
can be extrapolated. The effect of the amine concentration [Ni-
Pr2Et] on the reaction rate was also examined by monitoring
the rate of decay of 1 in the presence of 3-30 equiv of NiPr2Et
at 23 °C in THF. The rate data for consumption of 1 versus
[NiPr2Et] are plotted in Figure 6 and indicate a rate that is first
order in [NiPr2Et].
Addition of cyanide anion to 7 effected the release of the
iminium ligand as a functionalized amine product via generation
of a new C-C bond at the R position (eq 7). Treatment of 7
with 5 equiv of NaCN in methanol/THF produced Et2NCH-
(CN)CH3, confirmed by GC-MS and 1H NMR analysis follow-
ing workup. A 1H NMR spectrum of the reaction solution
immediately following the addition in CD3OD/THF-d8 estab-
lished that the consumption of 7 is very rapid at room
temperature and that the Et2NCH(CN)CH3 byproduct is released
quantitatively. The palladium product(s) of the reaction have
not been identified.
The relative rate of the reaction between 1 (0.044 M) and
the substrate NiPr2Et or NiPr2CD2Me (0.88 M) was examined
in THF. A dramatic attenuation in rate was observed for the
deuterated amine substrate (kobs ) 9.9(6) × 10-5 s-1), providing
a kinetic deuterium isotope effect (KDIE), krel(kH/kD), of 5.9.
For the case of NiPr2CD2Me, a small amount of an intermediate
Kinetic Data. The decay of complex 1 (0.044 M) in the
presence of 20 equiv of NiPr2Et was monitored by 31P NMR
spectroscopy and exhibited clean first-order decay in THF at
23 °C. The reaction rate was conveniently determined by
integrating the 31P NMR resonance of complex 1 versus an
internal standard of Ph3PdO, sealed as a CH2Cl2 solution within
a capillary tube. During the reaction course the starting material
species was detected in the 31P NMR spectrum. This species
2
featured a pair of doublets at 28.5 and 39.4 ppm with JPP
)
62 Hz and can be tentatively assigned as the amine adduct
complex, [Ph2BP2]Pd(NiPr2CD2Me)(OTf), based upon the simi-
larity of its chemical shift to [Ph2BP2]Pd(o-phenylpyridine)-
(OTf) (5). We presume that [Ph2BP2]Pd(NiPr2CD2Me)(OTf)
accumulates due to the attenuated C-D activation rate.
(31) (a) Ciriano, M.; Green, M.; Gregson, D.; Howard, J. A. K.; Spencer, J. L.;
Stone, F. G. A.; Woodward, P. J. Chem. Soc., Dalton Trans. 1979, 8, 1294-
1300. (b) Kayaki, Y.; Shimizu, I.; Yamamoto, A. Bull. Chem. Soc. Jpn.
1997, 70, 917-927. (c) Veya, P.; Floriani, C.; Chiesi-Villa, A.; Rizzoli,
C. Organometallics 1994, 13, 441-450.
9
15826 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004

Aspects of an Amine Activation Process
A R T I C L E S
1
2
Inspection of the H and H NMR spectra of the reaction
between 1 and NiPr2CD2Me after the complete consumption of
1 revealed that the palladacycle product 6 contained deuterium
at the R- and â-carbon positions in a nearly statistical ratio of
1:4 (eq 8). The ammonium salt byproducts, [DNiPr2CD2CH3]+
and [HNiPr2CD2CH3]+, showed no evidence for deuterium
scrambling into the â-carbon position. The extent of deuterium
scrambling into the alkyl chain was next investigated using the
BP2]Pd}2 (4) as the major product and only a small amount of
the palladacycle 6 (or 7). At low temperature (toluene, -35 to
0 °C), however, the palladacycle products dominate. The case
of NEt3 was examined in detail because of its slightly simplified
1H NMR spectrum in the aliphatic region. A toluene-d8 solution
of 1 in the presence of 50 equiv of NEt3 was examined from
-50 to 0 °C by both 1H and 31P NMR spectroscopy (see Figure
7). In the low-temperature regime (-50 °C), two broad
resonances are observed in the 31P NMR spectrum (42.7 ppm,
45.2 ppm) due to slow exchange between 1 and 1′. As noted
previously, a similar spectrum is observed for 1 in CD2Cl2 at
-50 °C in the absence of amine. At -35 °C the two signals
have coalesced into a single resonance that is centered at 45
ppm, indicating that 1 and 1′ are in fast exchange. As the sample
is further warmed to -20 °C, precursor 1 decays and four new
doublets arise that represent two independent species. The more
downfield pair of doublets (45.0 and 29.5 ppm, JPP ) 25.9 Hz)
is relatively similar in chemical shift to the resonances observed
for the o-phenylpyridine adduct complex 5 (46.3 and 35.3 ppm).
By analogy, this intermediate is proposed to be the amine adduct
complex [Ph2BP2]Pd(NEt3)(OTf). An alternative and perhaps
equally reasonable formulation for this intermediate would be
to suggest that an agostic C-H bond from the amine ligand
fills the fourth site and that the triflate anion is outer sphere.
However, examination of the 1H NMR spectrum between -60
and -20 °C provided no evidence for an agostic C-H proton.
i
N-propylamine substrate Pr2NCD2CH2CH3. In this case deu-
terium incorporation occurred only at the R- and â-carbon
positions and in a statistical ratio of 1:2. No scrambling into
the γ-carbon position was detected (eq 9).
Observation of Intermediates by NMR Spectroscopy
Using NEt3 in Toluene. Intermediates during the amine
activation reaction were difficult to study because of their low
concentrations in THF. Unfortunately, THF is the only solvent
that mediates the quantitative conversion of 1 and amine to the
corresponding palladacycle product. A reasonable compromise
was found by studying the activation reaction at low temperature
in toluene. Recall that in benzene or toluene the treatment of 1
with NiPr2Et (or NEt3) at room temperature generates {[Ph2-
2
The other pair of doublets (32.5 and 18.0 ppm, JPP ) 36.0
Hz), which corresponds to another intermediate, is more
complex in the proton-coupled 31P NMR spectrum (Figure 7,
inset a). The doublet centered at 18 ppm splits into a doublet
of doublets with a rather large secondary coupling constant of
∼200 Hz. The 1H NMR spectrum shows a doublet of doublets
3
centered at -8.62 ppm (3JHP (cis) ) 30.5 Hz and JHP (trans)
Figure 7. Reaction profile followed by 31P{1H} NMR spectroscopy of 1 and 50 equiv of NEt3 in d8-toluene. The temperature was gradually warmed from
-50 to 0 °C. (a) Inset shows splitting of peak upon removing proton decoupling. Spectra and inset correspond to the scale shown at the bottom.
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15827

A R T I C L E S
Lu and Peters
) 200.4 Hz). This resonance simplifies to a singlet when the
31P nuclei are decoupled. These data collectively intimate a
square planar “[Ph2BP2]Pd(H)(L)” hydride intermediate. As
discussed below, this intermediate is proposed to be the hydride
iminium adduct [[Ph2BP2]Pd(H)(Et2NdCHCH3)][OTf].
The two detectable intermediates decayed as the concentration
of palladacycle 7 increased. As the reaction proceeded at -20
°C, the intermediates maintained a ∼1:1 ratio until they were
consumed. Integration of the final 31P NMR spectrum acquired
at 0 °C established that 7 had been generated in ∼80% yield.
As shown in Figure 7, the palladium dimer {[Ph2BP2]Pd}2 4
was present in only trace quantity (doublets at ∼10 and 27 ppm).
The majority of the impurities comprise the resonances between
44 and 47 ppm and are not readily assigned.
Comparison to the System [(Ph2SiP2)Pd(THF)2][OTf]2.
The lack of precedent for the palladaziridines described herein
was surprising given their thermal stability and the fact that
five- and six-membered palladacycles are rather common.32 To
inquire as to whether the choice of the anionic [Ph2BP2]
phosphine ligand was serendipitous, a palladium system sup-
ported by the neutral bis(phosphine) ligand Ph2Si(CH2PPh2)2
(abbreviated as Ph2SiP2) was studied. The ligand Ph2SiP2 is a
structurally faithful, neutral analogue of [Ph2BP2] that gives rise
to reaction profiles very similar to the more familiar bis-
(diphenylphosphino)propane (dppp) ligand.7 The desired precur-
sor complex, (Ph2SiP2)Pd(OTf)2 (14), was conveniently prepared
in two steps in 87% overall yield by reaction of Ph2SiP2 with
PdCl2(NCPh)2 to generate (Ph2SiP2)PdCl2 (15), followed by
treatment with 2 equiv of AgOTf to give 14.
Similar to [[Ph2BP2]Pd(THF)2][OTf] (1), (Ph2SiP2)Pd(OTf)2
(14) coordinated o-phenylpyridine in THF but did not effect
arylmetalation in the temperature range between 25 and 80 °C.
Moreover, like 1, complex 14 is quite reactive toward trialkyl-
amine substrates. Indeed, complex 14 was fully consumed in
less than 1 h at room temperature, releasing a stoichiometric
equivalent of [HNiPr2Et][OTf] when exposed to just 3 equiv of
NiPr2Et in THF. Complete consumption of 1 on a similar time
scale requires 20 equiv of NiPr2Et. Despite these promising
signs, the major palladium-containing product in the case of
14 is the dimeric palladium(I) complex [{(Ph2SiP2)Pd}2][OTf]2
(16) (eq 10). A second set of NMR signals (31P and 13C NMR)
was also present, albeit in much lower concentration relative
to 16, that may be attributed to [(Ph2SiP2)Pd(N,C:η2-NiPr2-
CHCH3)][OTf]. The 13C NMR spectrum showed a characteristic
resonance at 47 ppm, roughly the chemical shift expected for
the Pd-bound carbon atom of the iminium ligand. Other tertiary
amines provided similar results, but in no case were the
secondary reaction products isolated or more thoroughly char-
acterized due to their low concentrations and the difficulty in
separating them from dimeric 16. Gentle warming of a solution
that presumably contains [(Ph2SiP2)Pd(N,C:η2-NiPr2CHCH3)]-
[OTf] does not lead to its conversion to 16.
BP2] and commercially available (PhCN)2PdCl2, complex 1 is
readily prepared in three steps and in an overall isolated yield
of 67%. The THF and triflate ligands of 1 are highly labile.
Consequently, complex 1 reacts as an unsaturated Pd(II) center
with two open cis sites. The reactivity of 1 with trialkylamines
is marked in its rapidity at room temperature. Moreover, the
reaction conditions and workup are fairly general for the various
amine substrates examined. The palladacycles are produced
cleanly and isolable in high yields (typically 80%). Extremely
bulky amines, however, result in lower isolated yields (e.g., 38%
yield for [Ph2BP2]Pd(N,C:η2-NCy2CHMe) (8)). The decreased
yields are likely the result of the difficulty in separating the
palladacycles from the ammonium salt byproduct due to the
decreased solubility associated with the increased rigidity of
the more encumbered palladacycles.
The amine activation process is selective for secondary C-H
bonds. The bias against tertiary C-H bonds is presumably of
steric origin, while the bias against primary C-H bonds is
consistent with C-H bond dissociation energies (methylene CH2
< primary CH3). While the focus of this paper centers on amine
activation exclusively, the transformations described may prove
to be more general. One clue that this might be the case comes
from the curious reaction that converts 1 to 11 upon addition
of diisopropyl(2-propenyloxymethyl)amine (eq 3). A mecha-
nistic sequence that would account for this transformation
invokes coordination of an O-atom lone pair to Pd(II) followed
by a â-hydride elimination step. The oxonium intermediate
formed could then rearrange to 11 along with the release of
acrolein (Scheme 3). Comparing C-H bond dissociation ener-
gies, the allylic C-H bond is expected to be weaker and thus
is preferentially cleaved. The proposed scheme is consistent with
the observation that acrolein is a byproduct of the reaction.
Reactivity of the Palladacycles. The observation that
deuterium can scramble from a deuterated ammonium salt into
the aliphatic positions of the neutral palladacycles (see eq 4)
suggests that the palladacycles are not static species in solution
but are perhaps capable of dynamic â-hydride elimination/olefin
reinsertion processes akin to those of cationic palladium alkyl
species such as Brookhart’s (R-diimine)Pd(alkyl)+ systems.33
As shown in Scheme 4, facile chain walking would allow
deuterium to wash into the R- and â-carbon positions, though
we must stress that the direct observation of these intermediates
has not been realized. Curiously, deuterium does not wash into
the γ-carbon position. This discrepancy may arise from differ-
ences in the stability of the palladium hydride intermediates
Discussion
Access to the chemistry described herein required a synthetic
route to [NBu4][Ph2BP2]Pd(THF)2][OTf] (1). Starting with [Ph2-
(32) (a) Canty, A. J.; Patel, J.; Skelton, B. W.; White, A. H. J. Organomet.
Chem. 2000, 607, 194-202. (b) Dunina, V. V.; Gorunova, O. N.; Averina,
E. B.; Grishin, Y. K.; Kuz’mina, L. G.; Howard, J. A. K. J. Organomet.
Chem. 2000, 603, 138-151. (c) Ryabov, A. D. Chem. ReV. 1990, 90, 403-
424.
(33) (a) Tempel, D. J.; Johnson, L. K.; Huff, R. L.; White, P. S.; Brookhart, M.
J. Am. Chem. Soc. 2000, 122, 6686-6700. (b) Shultz, L. H.; Tempel, D.
J.; Brookhart, M. J. Am. Chem. Soc. 2001, 123, 11539-11555.
9
15828 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004

Aspects of an Amine Activation Process
A R T I C L E S
Scheme 3
Scheme 4
that lie along the path leading to deuterium incorporation
(Scheme 4). Incorporation of deuterium at the R and â positions
is expected to proceed via a palladium enamine hydride
species,34 whereas incorporation at the γ position is expected
to proceed through a palladium allylamine hydride species. It
may be that this latter species is conformationally disfavored,
and hence deuterium scrambling traverses exclusively along the
lower three paths shown in Scheme 4. It is also plausible to
suggest that electronic factors would play a role in destabilizing
a Pd allylamine hydride species. In the Pd enamine hydride,
the N-atom lone pair should increase the π-basicity of the olefin
ligand. This would not be true of an allylamine hydride species.
Although palladium enamine hydride complexes are invoked
as plausible intermediates in Scheme 4, they cannot be directly
observed. It appears, however, that they can be trapped. For
instance, addition of excess tert-butylisocyanide to [Ph2BP2]-
Pd(N,C:η2-NEt2CH2) (7) liberates enamine and traps the inferred
hydride to generate the complex [Ph2BP2]Pd(C(H)dNtBu)-
(CNtBu) (13) (eq 6).
The reactivity of 7 with NaCN to generate a new C-C bond
is consistent with nucleophilic attack of cyanide at an electro-
philic iminium ion bound to Pd(0) (eq 7). Though we have not
thoroughly explored the reactivity of these palladacycles, it is
apparent that the Pd-N-C core shows reactivity consistent with
both of its limiting resonance forms: as a Pd(0) iminium adduct
(NaCN) and as a Pd(II) alkyl complex with an amine donor
(â-hydride elimination, deuterium scrambling).
Mechanism. Three limiting mechanisms for the amine C-H
activation process are presented in Scheme 5. Common to each
path is an amine adduct of 1. The amine adduct A is designated
with a triflate ligand coordinated to palladium, though THF and
triflate are likely in rapid exchange in THF solvent. Intermediate
A presumably rearranges to species B, whereby an intra-
molecular CR-H bond replaces the donor ligand (OTf or THF)
(34) A Pd enamine hydride intermediate has been previously proposed, see:
Murahashi, S.-I.; Watanabe, T. J. Am. Chem. Soc. 1979, 101, 7429-7430.
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15829

A R T I C L E S
Scheme 5
Lu and Peters
in the site adjacent to the amine donor. At the formal C-H
bond-breaking step, the proposed paths diverge. Path I invokes
â-hydride elimination to generate the palladium iminium hydride
complex C, which is then deprotonated by an exogenous base
to give the palladacycle product. Another possibility, path II,
suggests that the agostic proton in B is directly deprotonated.
Alternatively, the agostic C-H bond may undergo oxidative
addition to a Pd(IV) alkyl hydride complex D, which like C is
subsequently deprotonated (path III).
Proposed intermediate B is common to each of the three paths.
While no direct spectroscopic evidence for B has been obtained,
its intermediacy to either a â-hydride elimination process (path
I) or an oxidative addition step (path III) seems reasonable to
infer. Moreover, an agostic C-H proton coordinated to a
positively charged palladium center should have markedly
increased acidity, which would facilitate its direct deprotonation
by a Brønsted base (path II). A wealth of mechanistic informa-
tion has been collected for cationic P2Pd(II) alkyl complexes
by Brookhart and co-workers. In some favorable cases an agostic
C-H proton coordinated to a cationic palladium(II) center can
be directly observed (e.g., (dippp)Pd(CH2CH2-µ-H)+).35 The
topological analogy between inferred B and P2Pd(R)+ systems
is apparent (Figure 8).
The spectroscopic observation of an intermediate palladium
hydride species at low temperature in d8-toluene solution is
evidence against path II. A possible counterargument to this
assertion is that this intermediate lies along a nonproductive
path and is in equilibrium with B. If path II is operative,
however, then the deprotonation step would have to be rate
determining to give rise to the large KDIE observed. The rate
law of the reaction would then be expected to show a second-
order dependence on amine concentration, assuming exogenous
amine to be the base. Observation of a first-order dependence
of rate on amine concentration therefore appears to be incon-
sistent with path II. Moreover, the small magnitude of ∆Sq is
inconsistent with a bimolecular rate-limiting step and therefore
appears to be incompatible with path II.
Experimentally discriminating between paths I and III is less
straightforward. Both paths invoke palladium hydride intermedi-
ates, both are expected to exhibit a primary KDIE, and both
could exhibit a first-order dependence on amine. The prevalence
of â-hydride elimination processes in Pd(II) complexes of
similar structure to B favors path I. Given the lack of a well-
defined Pd(II/IV) redox couple associated with a C-H bond
cleavage step, there is no compelling reason to invoke such a
scenario here (i.e., path III).36
In the context of path I, the work of Murahashi et al. is
relevant. These authors proposed the intermediacy of an η2-
iminium hydride of Pd(II), Pd(H)(RHCdNR1R2)+, in trialkyl-
amine redistribution reactions catalyzed by palladium black at
200 °C.5 Formal oxidative insertion of Pd(0) into a secondary
C-H bond adjacent to an amine N atom was invoked as the
key bond-breaking step. In the present system, we suggest
â-hydride elimination from Pd(II) to be the C-H bond-cleaving
step. An additional system that appears to be mechanistically
related to the present case comes from Harman and co-workers,
who provided spectroscopic data to suggest that one-electron
reduction of an Os(III) amine complex induces â-hydride
elimination to generate an Os(II) η2-iminium hydride complex.4
The mechanistic data presented here is interesting to compare
with that available for Pd-catalyzed alcohol oxidation systems.
The KDIE value we report for the amine activation process (5.9)
is large by comparison to KDIE’s that have been reported
elsewhere for catalytic alcohol oxidation (values range between
1.3 and 2.5).8b,d,f,g,i Sigman’s group reported a notable exception
in which the KDIE for an N-heterocyclic carbene-supported Pd-
(II) system was reported to be quite large in magnitude (6.8),8c
(35) (a) Ledford, J.; Schultz, C. S.; Gates, D. P.; White, P. S.; DeSimone, J.
M.; Brookhart, M. Organometallics 2001, 20, 5266-5276. (b) Shultz, L.
H.; Brookhart, M. Organometallics 2001, 20, 3975-3982.
(36) For examples of Pt(II/IV) redox couples associated with C-H bond
cleavage, see: (a) Wick, D. D.; Goldberg, K. I. J. Am. Chem. Soc. 1997,
119, 10235. (b) Jensen, M. P.; Wick, D. D.; Reinartz, S.; White, P. S.;
Templeton, J. L.; Goldberg, K. I. J. Am. Chem. Soc. 2003, 125, 8614-
8624. (c) Stahl, S. S.; Labinger, J. A.; Bercaw, J. E. Angew. Chem., Int.
Ed. 1998, 37, 2180. (d) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417,
507-514.
Figure 8. Comparison of proposed intermediate B and (dippp)Pd(CH2-
CH2-µ-H)+.
9
15830 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004

Aspects of an Amine Activation Process
A R T I C L E S
Table 2. Summary of Observations for Various Reaction
Scheme 6
Conditions
crude
intermediates
product distributiona
A
C
palladacycle
{[PhBP2]Pd}2 (4)
THF, 23 °C
toluene/benzene, 23 °C
toluene, -20 °C
quantitative
minor
observed observed major
major
trace
a On the basis of 31P NMR spectroscopy.
constants for â-hydride elimination steps are few in number.
The most related case is Harman’s osmium system, which is
reported to undergo â-hydride elimination at a rate of 6 × 10-2
s-1 at 25 °C.4 The rate constant k′ reported here (k′ ) 5.6(1) ×
10-2 s-1) is of similar magnitude.
Reaction Profile as a Function of Solvent and Tempera-
ture. The reaction profile changes dramatically upon varying
the solvent and temperature (Table 2). The variation in product
distribution between low and ambient temperatures in toluene
may suggest that {[PhBP2]Pd}2 (4) is the thermodynamic
product and that the palladacycles are the kinetic products.
Nonetheless, the palladacycles are thermally quite stable. They
show no proclivity to convert to 4 at high temperatures upon
isolation (up to 90 °C). We presume that complex 4, like the
palladacycles, is formed from the Pd iminium hydride inter-
mediate (C). Owing to the low polarity of toluene, the transition
state for deprotonation of C must be sufficiently raised that the
competitive loss of iminium triflate dominates the reaction
profile. Such a step would generate “[Ph2BP2]PdH”, which in
the absence of a strong donor is known to be quite unstable
and leads to the generation of {[Ph2BP2]Pd}2 (4).21 For the case
of the more polar solvent THF, the barrier to deprotonation is
presumably very low relative to the barrier for iminium
dissociation. Consequently, complex 1 funnels cleanly into the
palladacycle product in THF. In toluene, intermediates A and
C appear to be in equilibrium. In this case, â-hydride elimination
cannot be rate limiting.37 Instead, deprotonation becomes both
rate limiting and product determining in toluene.
and Goldman reported a large KDIE (7) for the catalytic
dehydrogenation of tertiary amines by an iridium system.6 At
present it is difficult to say what isotope-dependent factors
collectively contribute to the magnitudes of the various kinetic
isotope effects that have been recorded, though this is an
interesting topic for further study. While few activation param-
eters are available for Pd-catalyzed alcohol oxidation, Sigman
reported values of ∆Hq ) 20.25 ( 0.89 kcal/mol and ∆Sq )
-5.4 ( 2.7 eu for a sparteine-supported system under conditions
where â-hydride elimination is rate determining. These param-
eters compare favorably with those we report here (∆Hq ) 17.9
( 0.2 kcal/mol and ∆Sq ) -4 ( 1 eu).
Rate Law in THF. The proposed mechanism for the amine
activation process in THF is represented in Scheme 6 as
elementary steps. All steps are drawn as reversible except for
the deprotonation step, which appears to be irreversible based
upon the lack of overall iminium exchange in the palladacycles
(eq 5). Inclusion of intermediate B would not alter the rate law
and has therefore been excluded.
The following rate law is derived by treating 1 H 1′ as a fast
equilibrium with a constant Keq and applying the steady-state
approximation to intermediates A and C:
Reaction Profile as a Function of Charge on Palladium.
Solvent is not the only factor contributing to the observed
product distribution. The electrophilicity of the palladium species
also seems to play a role in determining how competitive
iminium release from intermediate C will be relative to
deprotonation. For example, when (Ph2SiP2)Pd(OTf)2 14, the
(Ph2SiP2)-supported analogue of 1, is exposed to NiPr2Et, an
analogous Pd iminium hydride intermediate may be invoked,
but the iminium ligand is presumably more labile in this case.
Thus, a pathway leading to a dimeric Pd(I) product, [(Ph2SiP2)-
Pd]2[OTf]2 16, would be favored. Several previous studies have
shown that the anionic [Ph2BP2] ligand is appreciably more
electron releasing than (Ph2SiP2).7,21 It is therefore reasonable
to anticipate that a π-bonded iminium ligand in the adduct
complex, [(Ph2SiP2)Pd(H)(R2NdCHCH3)]2+, would be more
labile than such a ligand in the complex [[Ph2BP2]Pd(H)(R2Nd
CHCH3)]+ (C). The anionic charge of the [Ph2BP2] chelate thus
appears to play an important role with respect to the generation
of these unusual palladacycle species.
rate )
Keqk1k2k3[1][NiPr2Et]2
k-1k3[THF]2[NiPr2Et] + k-1k-2[THF]2 + k2k3[NiPr2Et][THF]
(11)
In the limit where THF is in vast excess and k3[NiPr2Et] .
k-2, the rate equation simplifies to
Keqk1k2[1][NiPr2Et]
rate )
and
k-1[THF]2
Keqk1k2[NiPr2Et]
k-1[THF]2
kobs
)
(12)
The reduced rate law is in agreement with the experimental data
wherein the reaction rate is first order in both [NiPr2Et] and
[1]. Moreover, the rate law can be applied to the kinetic studies
performed in THF with 20-fold amine for which kobs was
measured. The rate constant k′ (k′ ) Keqk1k2/k-1) comprises two
equilibrium constants and the rate of the â-hydride elimination
step (k2). Future studies would be needed to ascertain Keq and
k1/k-1 and to obtain a value for k2. Reports of the isolated rate
(37) We studied the product distribution for the reaction of 1 and Et2NCD2CD3
in toluene to obtain a KIE of 2.0. This primary KIE does not help to
establish whether â-hydride elimination is rate determining in toluene. For
instance, a primary KIE would also be consistent with deprotonation of
the Pd-H as the rate-limiting step.
9
J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004 15831

A R T I C L E S
Lu and Peters
In summary, iminium adducts of palladium supported by a
bis(phosphino)borate ligand are readily generated from â-hy-
dride elimination of amines at a coordinatively unsaturated Pd-
(II) center. It is rare to isolate or even observe intermediates
relevant to the in situ reduction of Pd(II) by amines or the Pd-
(II)-catalyzed oxidation of alcohols. Thus, the isolable Pd(0)-
iminium adducts and the mechanistic observations accompa-
nying this study may provide important mechanistic information
regarding the nature of these processes. Moreover, this study
also alludes to an attractive opportunity for incorporating Pd-
(II) species in the presence of an oxidant within the broader
synthetic context of new methods for the catalytic functional-
ization of tertiary amines.
Acknowledgment. This work was supported by the NSF
(CHE-0132216), the ACS Petroleum Research Fund, and the
Dreyfus Foundation. The authors thank Theodore Betley, Larry
Henling, and Dr. Michael Day for crystallographic assistance.
C.C.L. thanks the Department of Defense for a Graduate
Research Fellowship.
Supporting Information Available: Complete crystallo-
graphic details for 1′, 5, 8, and 9 (PDF, CIF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA046415S
9
15832 J. AM. CHEM. SOC. VOL. 126, NO. 48, 2004