Energetics and mechanism of alkylamine N-H bond cleavage by palladium hydroxides: N-H activation by unusual acid-base chemistry

Article abstract of DOI:10.1021/om970856l

The dimeric palladium hydroxide complex [(PPh₃)(Ph)Pd(μ-OH)]₂ underwent proton exchange reactions to cleave the N-H bonds of both alkyl and aryl primary amines to generate dimeric palladium amido complexes. This acid-base chemistry was reversible and provided thermodynamic information on the relative bond energies of the bridging amido and bridging hydroxo ligands. The mechanism of the N-H activation reaction was studied in both the forward and reverse directions to gain a detailed picture of the reaction pathway. Our data indicate that the reaction proceeds via cleavage of the hydroxide dimer by association of 2 equiv of amine followed by reversible proton transfer to create the amido ligand and a bound water ligand. Reassociation of the amido monomer with one of the original hydroxide monomers with displacement of water and amine gives the final amido hydroxo dimer product. Microscopic reversibility leads to the prediction that the reaction of water with the amido complexes is first order in product amine and is, therefore, autocatalytic. This unusual kinetic behavior was observed.

Full text of DOI:10.1021/om970856l

5706
Organometallics 1997, 16, 5706-5715
En er getics a n d Mech a n ism of Alk yla m in e N-H Bon d
Clea va ge by P a lla d iu m Hyd r oxid es: N-H Activa tion by
Un u su a l Acid -Ba se Ch em istr y
Michael S. Driver and J ohn F. Hartwig*
Department of Chemistry, Yale University, P.O. Box 208107,
New Haven, Connecticut 06520-8107
Received October 2, 1997X
The dimeric palladium hydroxide complex [(PPh3)(Ph)Pd(µ-OH)]2 underwent proton
exchange reactions to cleave the N-H bonds of both alkyl and aryl primary amines to
generate dimeric palladium amido complexes. This acid-base chemistry was reversible and
provided thermodynamic information on the relative bond energies of the bridging amido
and bridging hydroxo ligands. The mechanism of the N-H activation reaction was studied
in both the forward and reverse directions to gain a detailed picture of the reaction pathway.
Our data indicate that the reaction proceeds via cleavage of the hydroxide dimer by
association of 2 equiv of amine followed by reversible proton transfer to create the amido
ligand and a bound water ligand. Reassociation of the amido monomer with one of the
original hydroxide monomers with displacement of water and amine gives the final amido
hydroxo dimer product. Microscopic reversibility leads to the prediction that the reaction
of water with the amido complexes is first order in product amine and is, therefore,
autocatalytic. This unusual kinetic behavior was observed.
In tr od u ction
tions, the amination of olefins,32 and recently reported
aryl halide aminations.33-37
The catalytic chemistry of soluble transition metal
complexes with alkylamine substrates is not well-
developed.1 One reason for this scarcity of catalysis
with amines is the absence of reactive transition metal
amido complexes prepared directly from alkylamine
substrates. N-H bond cleavage processes that generate
metal amido complexes are well-known for early transi-
tion metal complexes,2,3 for alkali metals,4-6 and for
organolithium reagents.4,6 However, early transition
metal amides are often very stable, and alkali metal
reagents are rare as catalysts.
The acid-base chemistry of alkali metal amides has
been studied by a number of groups and is fundamental
(12) Bakos, J .; Toty, I.; Heil, B.; Marko, L. J . Organomet. Chem.
1985, 279, 23.
(13) Bakos, J .; Toth, I.; Heil, B.; Szalontai, G.; Parkanyi, L.; Fulop,
V. J . Organomet. Chem. 1989, 370, 263.
(14) Bakos, J .; Orosz, A.; Heil, B.; Laghmari, M.; Lhoste, P.; Sinou,
D. J . Chem. Soc., Chem. Commun. 1991, 1684.
(15) Becalski, A. G.; Cullen, W. R.; Fryzuk, M. D.; J ames, B. R.;
Kang, G. J .; Rettig, S. J . Inorg. Chem. 1991, 30, 5002-5008.
(16) Burk, M. J .; Feaster, J . E. J . Am. Chem. Soc. 1992, 114, 6266.
(17) Chan, Y. N. C.; Osborn, J . A. J . Am. Chem. Soc. 1990, 112,
9400.
(18) Cullen, W. R.; Fryzuk, M. D.; J ames, B. R.; Kutney, J . P.; Kang,
G. J .; Herb, G.; Thorburn, I. S.; Spogliarich, R. J . Mol. Catal. 1990,
62, 243.
(19) Imine hydrogenation catalyzed by rhodium and iridium com-
plexes would be expected to occur through an amide complex. However,
mechanistic data is scarce. For a direct observation of imine insertion
that produces an amide complex potentially similar to those in catalytic
systems, see: Fryzuk, M. D.; Piers, W. E. Organometallics 1990, 9,
986-98.
(20) For a recent review, see: J ames, B. R. Chem. Ind. 1995, 62,
167-180.
(21) Kang, G. J .; Cullen, W. R.; Fryzuk, M. D.; J ames, B. R.; Kutney,
J . P. J . Chem Soc., Chem. Commun. 1988, 1466.
(22) Longley, C. J .; Goodwin, T. J .; Wilkinson, G. Polyhedron 1986,
5, 1625.
The only soluble late transition metal complexes that
cleave the NH bonds in alkylamines to form an amido
complex by any reaction have involved Os clusters.7-9
Even the oxidative addition of the NH bond of ammonia
to late transition metal complexes is rare.10 Despite the
paucity of isolated amido complexes, they can act as
viable intermediates in catalysis and are, most likely,
intermediates in imine11-29 and nitrile30,31 hydrogena-
X Abstract published in Advance ACS Abstracts, December 1, 1997.
(1) Roundhill, D. M. Chem. Rev. 1992, 92, 1-27.
(2) Hillhouse, G. L.; Bercaw, J . E. J . Am. Chem. Soc. 1984, 106,
5472.
(23) Ng, Y.; Chan, C.; Meyer, D.; Osborm, J . A. J . Chem. Soc.m
Chem. Commun. 1990, 869.
(3) Walsh, P. J .; Hollander, F. J .; Bergman, R. G. Organometallics
1993, 12, 3705-3723.
(24) Spindler, F.; Pugin, B.; Blaser, H. U. Angew. Chem., Int. Ed.
Engl. 1990, 29, 558.
(25) Vastag, S.; Bakos, J .; Toros, S.; Takach, N. E.; King, R. B.; Heil,
B.; Marko, L. J . Mol. Catal. 1984, 22, 283.
(4) Volhardt, K. P. C. Organic Chemistry; W. H. Freeman
Company: New York, 1987.
&
(5) Hancock, E. M.; Cope, A. C. Organic Syntheses 1945, 25, 25.
(26) Willoughby, C. A.; Buchwald, S. L. J . Am. Chem. Soc. 1992,
114, 7562.
(27) Willoughby, C. A.; Buchwald, S. L. J . Org. Chem. 1993, 58,
7627.
(28) Willoughby, C. A.; Buchwald, S. L. J . Am. Chem. Soc. 1994,
116, 11703-11714.
(29) Willoughby, C. A.; Buchwald, S. L. J . Am. Chem. Soc. 1994,
116, 8952.
(30) Armor, J . N. Inorg. Chem. 1978, 17, 203-213.
(31) Yoshida, T.; Okano, T.; Otsuka, S. J . Chem. Soc., Chem.
Commun. 1979, 870.
(32) Casalnuovo, A. L.; Calabrese, J . C.; Milstein, D. J . Am. Chem.
Soc. 1988, 110, 6738-6744.
(6) Heaney, H. Chem. Rev. 1962, 62, 81-97 and references therein.
(7) Lin, Y.; Mayr, A.; Knobler, C. B.; Kaesz, H. D. J . Organomet.
Chem. 1984, 272, 207-229.
(8) Bryan, E. G.; J ohnson, B. F. G.; Lewis, J . J . Chem. Soc., Dalton
Trans. 1977, 1328-1330.
(9) The following paper includes the N-H activation of benzylamine,
but the other amines in this work are arylamines, see: Glueck, D. S.;
Winslow, L. J .; Bergman, R. G. Organometallics 1991, 10, 1462-1479.
(10) Casalnuovo, A. L.; Calabrese, J . C.; Milstein, D. Inorg. Chem.
1987, 26, 971-973.
(11) Amrani, Y.; Lecomte, L.; Simou, D.; Bakos, J .; Toth, I.; Heil, B.
Organometallics 1989, 8, 542.
S0276-7333(97)00856-X CCC: $14.00 © 1997 American Chemical Society

Alkylamine N-H Bond Cleavage by Pd Hydroxides
Organometallics, Vol. 16, No. 26, 1997 5707
Sch em e 1
to many organic reactions.6,38,39 The identity of the
solvent and alkali metal has a large effect on the
aggregation number and the resulting basicity. In
contrast, the acid-base chemistry of late transition
metal complexes with alkylamines and the acid-base
chemistry of late transition metal alkylamides is much
less explored. Since it is well-known that the identity
of the metal can influence the basicity of a polar reagent,
it is important to understand how the late transition
metals and their accompanying ligands affect the acid-
base chemistry of the amido group. It is this type of
information that will assist in developing transition
metal catalyzed chemistry involving amido groups and
amine substrates such as arylation of amines, oxidation
of amines, and formation of amines by hydrogenation.
We report here the unusual chemistry of hydroxide
and alkylamide complexes of palladium toward alky-
lamines and water. Amines are deprotonated by pal-
ladium hydroxide complexes to form palladium amides.
This reaction is, of course, exactly the opposite of what
occurs with alkali metal reagents. Thus, the ion-pair
acidity of alkylamines is, at least formally, greater than
that of water in THF solvent when palladium is the
counterion bound to the amide or hydroxide. A detailed
account of the thermodynamics of this acid-base chem-
istry and of the mechanism by which the N-H bond is
cleaved is reported, and it shows an unusual dependence
of the rate on amine concentration in the forward direc-
tion and autocatalytic behavior in amine for the reverse.
anti-isomers. The two isomers of 2a were detected by
1
31P and H NMR spectroscopy. At room temperature,
the ratio of the syn- and anti-isomers was approximately
1:2. The ratio of syn-isomer increased as a THF solution
of 2a was cooled, and crystallization from THF/Et2O at
-25 °C gave crystals of 2a as predominantly (>90%)
the syn-isomer. After several hours in aromatic sol-
vents, a 1:2 equilibrium mixture of the two isomers
was reestablished. Crystalline 2a isomerized to an
equilibrium mixture of isomers within 5 min at -20 °C
in methylene chloride or THF. In contrast, only the syn-
isomer that contained both phosphine ligands trans
to the anilido group was observed for 2,6-dimethylanil-
ide, 2b.
Reaction of 1 in neat isobutylamine at room temper-
ature for 15 min gave bis(amide) 3c. Apparently,
branching at the â-position gives mono(amido) com-
plexes, but branching at the γ-position gives bis(amides).
Reaction of 1 equiv of isobutylamine converted only half
of the starting material 1 to the amide 3c. No mixed
palladium isobutylamido hydroxo dimer was observed.
It was not necessary to use neat amine to observe
complete conversion of 1, but high amine concentrations
reduced the formation of side products. The mixed
dimers 3a and 3b possessed a syn geometry with the
phosphine ligands trans to the amido group, while the
bis(amido) dimer 3c had an anti-geometry. The geom-
etry of these complexes was determined by 1H, 13C, and
31P NMR spectroscopy as well as X-ray diffraction in
the case of 3a .
Resu lts
Rea ction of 1 w ith Secon d a r y Am in es. Reactions
of 1 with secondary alkyl- or arylamines did not result
in the formation of isolable palladium amido complexes.
Reaction of 10 equiv of diethylamine with a THF
solution of 1 at 70 °C or reaction in neat diethylamine
at room temperature gave PPh3O as the only product
observed by 31P NMR spectroscopy along with a black
Pd-containing precipitate. Reaction of 1 with either
diisopropylamine or N-methylaniline resulted in similar
decomposition of 1 forming PPh3O and a black Pd
precipitate.
P alladiu m Am ido Com plexes by NH Bon d Cleav-
a ge: Rea ction s of 1 w ith P r im a r y Am in es. The
addition of excess primary amines to the PPh3-ligated
palladium aryl hydroxide dimer 140,41 led to the complete
consumption of 1, elimination of water, and formation
of palladium amides at either room temperature or
slightly above room temperature in 58-88% isolated
yields as outlined in Scheme 1. The reactions occurred
in 68-97% yields by 31P NMR spectrometry, with an
internal standard, even for amines that generate pal-
ladium amides with â-hydrogens.
Rea ction of 1 w ith P yr r ole. The addition of 5 equiv
of pyrrole to a THF solution of 1 containing excess PPh3
formed the monomeric palladium pyrrolyl complex
trans-(PPh3)2Pd(Ph)(NC4H4) (4), as shown in Scheme 2,
in 91% isolated yield. The identity of this complex was
confirmed by preparing it independently from KNC4H4
and trans-(PPh3)2Pd(Ph)I.
Complex 2a was formed as a mixture of syn- and anti-
isomers. Hydroxo dimer 1 is also a mixture of syn- and
(33) Louie, J .; Hartwig, J . F. Tetrahedron Lett. 1995, 36, 3609.
(34) Guram, A. S.; Rennels, R. A.; Buchwald, S. L. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1348.
(35) Louie, J .; Paul, F.; Hartwig, J . F. Organometallics 1996, 15,
2794-2805.
(36) Driver, M. S.; Hartwig, J . F. J . Am. Chem. Soc 1996, 118, 7217-
7218.
(37) Wolfe, J . P.; Wagaw, S.; Buchwald, S. L. J . Am. Chem. Soc.
1996, 118, 7215-7216.
(38) Gregory, K.; Schleyer, P. v. R.; Snaith, R. Adv. Inorg. Chem.
1991, 37, 47 and references therein.
(39) Collum, D. B. Acc. Chem. Res. 1993, 26, 227-234 and references
therein.
Rea ction of Th iol To F or m P a lla d iu m Th iola to
Com p lexes. As expected, the more acidic SH bond of
thiols also reacted with the dimeric palladium hydrox-
ides to generate palladium thiolato complexes. As
shown in Scheme 2, reaction between 2.5 equiv of tert-
butylthiol and tolylpalladium hydroxo complex 1b at
room temperature formed the bis(thiolato)palladium
(40) Gurshin, V. V.; Alper, H. Organometallics 1993, 12, 1890-1901.
(41) We have modified the procedure for the synthesis of 1 as
described in the Experimental Section.
1
complex 5 in 49% isolated yield. Comparison of the H

5708 Organometallics, Vol. 16, No. 26, 1997
Driver and Hartwig
Ta ble 2. Selected In tr a m olecu la r Dista n ces for
{P d 2(P P h 3)2P h 2(µ-OH)(µ-NH-t-Bu )}(3a )a
Pd1-Pd2
Pd1-P1
Pd1-O1
Pd1-N1
Pd1-C1
2.905(1)
2.253(2)
2.142(5)
2.083(5)
1.989(7)
Pd2-P2
Pd2-O1
Pd2-N1
Pd2-C25
N1-C49
O1-H53
2.251(2)
2.140(4)
2.086(6)
2.015(7)
1.49(1)
0.798
a
Distances are in angstroms. Estimated standard deviations
in the least significant figure are given in parentheses.
Ta ble 3. Selected In tr a m olecu la r An gles for
{P d 2(P P h 3)2P h 2(µ-OH)(µ-NH-t-Bu )}(3a )a
Pd2-Pd1-P1
Pd2-Pd1-O1
Pd2-Pd1-N1
Pd2-Pd1-C1
P1-Pd1-O1
P1-Pd1-N1
P1-Pd1-C1
O1-Pd1-N1
O1-Pd1-C1
N1-Pd1-C1
Pd1-Pd2-P2
Pd1-Pd2-O1
Pd1-Pd2-N1
Pd1-Pd2-C25
130.17(6)
47.3(1)
45.9(2)
122.9(2)
97.6(1)
176.1(2)
89.7(2)
79.4(2)
170.2(2)
93.1(2)
133.07(6)
47.3(1)
45.8(1)
P2-Pd2-O1
P2-Pd2-N1
P2-Pd2-C25
O1-Pd2-N1
O1-Pd2-C25
N1-Pd2-C25
Pd1-O1-Pd2
Pd1-N1-Pd2
Pd1-N1-C49
Pd2-N1-C49
Pd1-O1-H53
Pd2-O1-H53
Pd1-N1-H41
Pd2-N1-H41
C49-N1-H41
99.3(1)
178.7(2)
91.5(2)
79.3(2)
169.1(2)
89.8(3)
85.4(2)
88.3(2)
120.5(4)
118.7(4)
120.41
126.05
109.17
109.17
109.17
F igu r e 1. ORTEP drawing of {Pd2(PPh3)2Ph2(µ-OH)(µ-
NH-t-Bu)}(3a ). Hydrogen atoms are omitted for clarity.
Sch em e 2
124.0(2)
a
Angles are in degrees. Estimated standard deviations in the
least significant figure are given in parentheses.
planes is 56.91°. As discussed recently,44 some bridging
hydroxo and amido complexes are bent and others are
planar with the bent structures having dihedral angles
ranging from 32-45°. The two halves of the core appear
to be chemically, though not crystallographically, equiva-
lent, and the two Pd-O and Pd-N distances are
indistinguishable and are 2.14 and 2.08 Å, respectively.
Despite the larger size of nitrogen and the presence of
an N-bound alkyl group, the Pd-N distances are shorter
than the Pd-O distances. As demonstrated below, the
shorter distance is not a result of a stronger Pd-N
bond.45
N-H Bon d -F or m in g Rea ction s. The proton ex-
change reactions of 1 with primary amines to generate
palladium amido complexes and H2O are reversible.
Heating a THF solution of the alkyl- or arylamido
complexes 2a ,b and 3a -c with 100 equiv of H2O at 70
°C for 1.5 h gave complete consumption of the amido
complex and regenerated 1 as the only species observed
by 31P NMR spectroscopy. The proton exchange reac-
tion with thiols, however, was not reversible. Heating
a THF solution of 5 with an excess of water at 70 °C for
several hours led to complete decomposition of 5 with
no formation of 1. GC/MS indicated the formation of a
variety of sulfur-containing organic compounds.
Th er m od yn a m ic An a lysis of th e NH Bon d Clea v-
a ge. As discussed above, 1 reacts reversibly with
primary amines to give palladium amides and water.
This reversibility allowed for measurement of the equi-
librium constants, Keq, of these proton exchange reac-
tions and calculation of ∆H and ∆S for the reactions.46
The rapid reaction of 1 with aniline to generate 2a and
H2O (eq 1) was used to determine the temperature
Ta ble 1. Da ta for X-r a y Str u ctu r e of
{P d 2(P P h 3)2P h 2(µ-OH)(µ-NH-t-Bu )} (3a )
empirical formula
fw
C52H51N1O1P2Pd2‚C2.6H1.6Cl1.6
1070.29
cryst color/habit
cryst syst
colorless plate
triclinic
lattice params:
a ) 12.765(3) Å
b ) 14.029(5) Å
c ) 15.176(4) Å
R ) 94.65(2)°
â ) 107.81(2)°
γ ) 93.30(2)°
2569(3) Å3
volume
space group
Z value
P1h (No. 2)
2
radiation
Mo KR (λ ) 0.710 69 Å)
temperature
residuals: R, Rw
goodness of fit indicator
-93 °C
0.040; 0.056
3.16
and 31P NMR spectra of 5 to previously prepared
palladium thiolato dimers42 confirmed the identity of 5
and indicated that the complex has a syn-geometry.
Rea ction of 1 w ith Alcoh ols a n d P er oxid es.
Reaction between 1 and alcohols or peroxides did not
provide clean chemistry. No reaction between 1 and
either tert-butyl alcohol or tert-butyl hydroperoxide
occurred at room temperature, while reaction at 70 °C
for 2 h gave a black Pd precipitate and PPh3O.
X-r a y Cr ysta l Str u ctu r e of 3a . An X-ray quality
crystal of 3a was obtained by slow crystallization from
CH2Cl2/Et2O at -35 °C. An ORTEP diagram of 3a is
provided in Figure 1 with selected distances (Å) and
angles (deg) given in Tables 2 and 3.43 The four-
membered ring core is markedly puckered. The dihe-
dral angle between the O-Pd1-N and O-Pd2-N
(44) For a well-characterized example and related references, see:
Ruiz, J .; Martinez, M. T.; Vicente, C.; Garcia, G.; Lopez, G.; Chaloner,
P.; Hitchcock, P. B. Organometallics 1993, 12, 4321.
(45) For a recent discussion of bond lengths and strengths, see:
Ernst, R. D.; Freeman, J . W.; Stahl, L.; Wilson, D. R.; Arif, A. M.;
Nuber, B.; Ziegler, M. L. J . Am. Chem. Soc. 1995, 117, 5075.
(42) Louie, J .; Hartwig, J . F. J . Am. Chem. Soc. 1995, 117, 11598.
(43) X-ray data for 3a was previously submitted as Supporting
Information, see: Driver, M. S.; Hartwig, J . F. J . Am. Chem. Soc. 1996,
118, 4206-4207.

Alkylamine N-H Bond Cleavage by Pd Hydroxides
dependence of Keq. Data were obtained by monitoring
1 + H2NPh h 2a + H2O
Organometallics, Vol. 16, No. 26, 1997 5709
∆H ) -3.6 kcal/mol ∆S ) 1.2 eu
(1)
the relative concentrations of 1 and 2a in a THF solution
by 31P NMR spectroscopy between -20 and 40 °C.
Below -20 °C, material precipitated from solution. The
graph of ln(Keq) vs -1/T (Figure S1)47 was linear and
resulted in a ∆H° of -3.6 ( 0.4 kcal‚mol-1 and ∆S° of
1.2 ( 0.9 eu, indicating a small entropic contribution
to ∆G°. The equilibrium constant for the reaction of 1
with sec-butylamine (eq 2) was measured in a similar
fashion and was found to be 12, which corresponds to
F igu r e 2. Plot of 1/kobs vs [H2O]. R ) 0.99.
∆G298° of -1.5 kcal‚mol-1
.
Keq ) 12
1 + H2N-sec-Bu y
z 3b + H2O
∆G ) -1.5 kcal/mol
(2)
Mech a n ism of NH Bon d Clea va ge by 1. The
cleavage of the NH bond of an amine to generate a
reactive metal-amido complex is an important step for
observing transition metal reaction chemistry with
amine substrates. The proposed catalytic cycle for the
addition of aniline across an olefin involves activation
of an NH bond to generate a reactive Ir amido com-
plex.32 More recently, it has been proposed that pal-
ladium alkoxide complexes may activate the NH bond
of amines to generate the reactive palladium amido
complexes involved in the catalytic amination of aryl
halides.48 A better understanding of how the NH bond
of an amine is activated to generate reactive amido
intermediates will aid in the development of improved
processes for the utilization of amine substrates.
F igu r e 3. Plot of ln{[3b]t/[3b]0} vs time for the reaction
of 3b with H2O in the absence of H2N-sec-Bu. The graph
deviates from linearity because kobs of eq 8 changes as a
result of the increasing [H2N-sec-Bu] over the course of the
reaction.
To gain a better understanding of the chemistry of
amines and of the amido complexes, we studied the
reverse of the N-H bond cleavage, reaction of 3b with
H2O to form 1 and free amine. Study of the reaction of
3b with H2O will reveal information about steps of the
N-H bond cleavage process that occur after the rate-
determining step. Kinetic data for the reaction of 3b
with H2O were again obtained by 1H NMR spectroscopy
in THF-d8 solvent, this time at 70 °C and under
conditions that are pseudo-first-order in 3b. Plots of
ln[3b] vs time (Figure S4) were linear over 3 half-lives,
indicating first-order behavior in [3b]. In contrast to
the N-H bond cleavage process, the reaction rates were
independent of [D2O], even though the transformation
involved reaction of 3b with D2O. Monitoring the
We began our investigation of the mechanism of the
NH activation reaction by studying the reaction of 1
with sec-butylamine. Kinetic data for the reaction of 1
with sec-butylamine were obtained by monitoring the
1
decay of the rapidly equilibrating isomers of 1 by H
NMR spectroscopy in THF-d8 solvent at 25 °C. The
reactions were conducted under conditions that were
pseudo-first-order in [1]. Plots of ln[1] vs time (Figure
S2) were linear over 3 half-lives, and reaction rates were
independent of [1]0, indicating a first-order dependence
on [1]. A graph of ln(kobs) vs ln[sec-butylamine] (Figure
S3) was linear with a slope of 2.0 ( 0.1, demonstrating
a second-order dependence on [sec-butylamine]. This
reaction order is surprising considering that only one
amido ligand is incorporated into the product 3b.
Reaction rates were independent of added PPh3 (130-
390 mM). Another set of reactions were conducted with
added H2O. The graph of 1/kobs vs [H2O] in Figure 2
was linear with a positive slope and a non-zero y-
intercept. These data indicate that the reaction is
inhibited by added H2O but that the reaction order in
water is not simply inverse first order. Finally, reac-
tions using deuterium-labeled 1-(µ-OD)2 with D2N-sec-
butyl revealed an inverse isotope effect, kH/kD ) 0.7 (
0.1.
reaction at varied [sec-butylamine] and graphing ln(kobs
)
vs ln[sec-butylamine] (Figure S5) resulted in a line with
a slope of 1.0 ( 0.1, indicating that the reaction is first
order in [sec-butylamine]. Since the reaction is first
order in amine while amine is a product of the reaction,
this reaction of 3b with water is autocatalytic. Indeed,
reactions of 3b with water in the absence of added
amine showed autocatalytic behavior. A plot of ln[3b]
vs time (Figure 3) deviates from linearity as a result of
the changing concentration of amine product. We have
not been able to ascertain the initiation process that
generates amine.
Discu ssion
Rever sibility of th e P r oton Tr a n sfer Rea ction s
of 1 w ith Am in es. The reactivity of 1 with free amines
to produce palladium amido complexes and water stands
in marked contrast to the reactivity predicted by pKa
values and the reactivity of main group and early metal
(46) The concentrations of water and amine were calculated based
on their known starting concentrations and the stoichiometry of the
reaction depicted in eq 1.
(47) Figures S1-S5 are provided as Supporting Information.
(48) Mann, G.; Hartwig, J . F. J . Am. Chem. Soc. 1996, 118, 13109-
13110.
49
amides with water. In organic solvents, the pKa of
H2O was measured to be approximately 31 while that
(49) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456-463.

5710 Organometallics, Vol. 16, No. 26, 1997
Driver and Hartwig
of ammonia was estimated to be 41, indicating H2O is
much more acidic than ammonia and presumably alkyl-
amines. This suggests that H2O should react with the
amido anion to give amine and hydroxide. In the
reactions of main group and early transition metal
amides with H2O, this is indeed the case. Alkali-metal
amides react violently with H2O to form the correspond-
ing hydroxide and free amine. Also, the early metal
amido complex Cp*2Hf(NH2)H reacts rapidly with H2O
to give Cp*2Hf(OH)H and NH3.2
However, late transition metal organometallic com-
plexes are often more stable toward water; similarly,
late metal amido complexes appear to be more stable
toward water than early metal amides. The addition
of diarylamines to Cp*(PMe3)2Ru(OH) established an
equilibrium that favored the hydroxide complex,50 while
addition of diarylamines to (DPPE)Pt(Me)(OH) favored
formation of the amido complexes, as observed for the
reaction of the alkylamines with the palladium hydroxo
dimer (vide supra).44 Apparently, metal amido com-
plexes become increasingly more stable toward reaction
with H2O as one moves to the right in the transition
metal series.
F igu r e 4. H-X vs relative LnM-X bond strengths in
kcal‚mol-1. Amido ligand bond strengths (() calculated for
2a and 3b superimposed on values for (DPPE)Pt(Me)X (9),
Cp*(PMe3)2Ru-X (O), and Cp*2(OCMe3)ThX (3) from refs
50 and 52 with LnM-OH bond strength arbitrarily as-
signed to 0 kcal‚mol-1
.
was independent of the identity of the metal fragment
(Figure 4), in the absence of M-X multiple bonding.52
Many amido and alkoxo ligands bridge between two
metal centers. Thus, it is important to understand how
bridging interactions affect the relative bond strengths.
Th er m od yn a m ic An a lysis of P r oton Tr a n sfer
Rea ction s of 1 w ith P r im a r y Am in es. The revers-
ibility of the reactions of 1 with primary amines allowed
for the measurement of the thermodynamic parameters
for the proton transfer reactions and the relative Pd-
(µ-OH)-Pd and Pd-(µ-NHR)-Pd bond strengths. The
entropic term of the proton exchange reaction was small,
∆S° ) 1.2 eu, as would be expected considering the
equal numbers of molecules on either side of eq 1. Using
∆Hrxn of 3.6 kcal‚mol-1 along with the known 119
kcal‚mol-1 O-H bond strength of water and the 88
kcal‚mol-1 N-H bond strength of aniline,51 we calcu-
lated that the bridging Pd-(µ-OH)-Pd interaction is
stronger than the bridging Pd-(µ-NHAr)-Pd interac-
Perhaps surprisingly, one can see that the bridging
amido bond strengths lie on the linear correlation
between LnM-X and H-X bond strengths established
from the Bryndza and Bercaw study and Marks study
(Figure 4). For this analysis, the bridging hydroxide
interaction was assigned as 0 in our case, and the
hydroxide bond strength was designated 0 in the
Bryndza and Marks studies. One might expect the
more basic N electron pair to be a stronger donor to the
Pd center than the O lone pair electrons, making the
relative bond strengths for a combination of the two
bridging interactions substantially different than the
relative strengths for the terminal ligands. The close
correlation of bridging amido bond strengths to the
terminal values demonstrates otherwise. To the extent
that the bonds of the bridging group can be dissected
into one covalent and one dative bond, the dative
interaction of the amido group is either not substantially
stronger than that of the hydroxo group or the stronger
dative interaction is accompanied by weakening of the
covalent interaction.
Mech a n ism of P r oton Exch a n ge Rea ction of 1
w ith sec-Bu tyla m in e. Potential mechanisms for the
reaction of 1 with sec-butylamine are shown in Schemes
3 and 4. Pathway A consists of initial reversible dimer
cleavage to form two three-coordinate species, followed
by coordination of amine. After coordination of amine,
reversible proton transfer occurs to form the amido
ligand and water, which dissociates to form a three-
coordinate amido species. This amido species then
associates with a three-coordinate hydroxo complex to
give the product 3b. There is a significant problem with
this mechanism. We have shown previously that three-
coordinate PPh3-ligated palladium alkyl amido com-
plexes undergo reductive elimination faster than dimer-
ization.54 Although the dimerization of a three-coordinate
amido complex is slightly different from trapping by the
three-coordinate hydroxo complex, one would expect to
see some reductive elimination products; no N-alkyl
tion by approximately 27 kcal‚mol-1
.
The reaction of 1 with sec-butylamine was also
reversible, and the Keq was measured to be 12, corre-
sponding to a ∆G298° ) -1.5 kcal‚mol-1. Since ∆S for
the similar reaction involving aniline was small, one can
assume that ∆G298° = ∆H° in this case as well. From
the ∆Hrxn of 1.5 kcal‚mol-1, the 119 kcal‚mol-1 O-H bond
strength of water, and the presumably similar N-H
bond strength of sec-butylamine to the 100 kcal‚mol-1
N-H bond strength in H2NMe,51 we concluded that the
bridging Pd-(µ-OH)-Pd interaction is stronger that the
bridging Pd-(µ-NHR)-Pd interaction by roughly 17
kcal‚mol-1
.
The relative M-X bond strengths for various ligands
have previously been measured for the systems Cp*-
(PMe3)3Ru-X and (DPPE)MePt-X,50 Cp*2(OCMe3)Th-
X,52 and Cp*(PMe3)(H)Ir-X53 with a detailed discussion
and comparison presented by Bryndza and Bercaw.50
In these studies, a linear correlation between the
relative LnM-X bond strengths and H-X bond strengths
was found where X was a terminal, covalent ligand. This
linear relationship appeared to be a general trend that
(50) Bryndza, H. E.; Fong, L. K.; Paciello, R. A.; Tam, W.; Bercaw,
J . E. J . Am. Chem. Soc. 1987, 109, 1444.
(51) McMillen, D. F.; Golden, D. M. Ann. Rev. Phys. Chem. 1982,
33, 493-532.
(52) Bruno, J . W.; Marks, T. J .; Morss, L. R. J . Am. Chem. Soc. 1983,
105, 6824-6832.
(53) Buchanan, J . M.; Stryker, J . M.; Bergman, R. G. J . Am. Chem.
Soc. 1986, 108, 1537.

Alkylamine N-H Bond Cleavage by Pd Hydroxides
Organometallics, Vol. 16, No. 26, 1997 5711
Sch em e 3
determining step. Most important, the predicted order
in [amine] and [H2O] are different from the ones
observed, ruling out the dissociative mechanism of
pathway C.
Paths D depict two kinetically related mechanisms
in which 2 equiv of amine are involved in the N-H
cleavage process. In the top mechanism of path D, 2
equiv of amine react with dimeric 1 by an associative
process to generate 2 equiv of a monomer with a
terminal hydroxo ligand and a coordinated amine on the
same complex. In the bottom mechanism, 1 equiv of
amine is used to generate a terminal hydroxo ligand in
a bimetallic intermediate. The subsequent step in these
mechanisms would involve reversible proton transfer
between one of the amines, either coordinated or
uncoordinated, and one of the hydroxo ligands to form
coordinated or free water and an amido ligand. In the
case of the top path D the amido aquo intermediate
would combine with one of the amino hydroxo interme-
diates to give a dinuclear µ-OH complex, while in the
case of the bimetallic mechanism, the same dinuclear
µ-OH intermediate would be formed by exchange of the
hydroxo ligand with amine. This exchange may or may
not involve coordination of amine. In the final step, the
amido ligand replaces the coordinated amine to form
3b. Both mechanisms in path D would show a first-
order dependence on [1], a second-order dependence
on [amine], and an order in [H2O] that depends on the
water concentration. The rate expressions for paths
D are given in eq 6; the order in [H2O] for this
mechanism is best revealed by inverting eq 6 to generate
eq 7. Equation 7 shows that a plot of 1/kobs vs [H2O]
arylamine products were observed. Pathway A would
follow the rate expression in eq 3, which predicts a first-
order dependence on 1, a first-order dependence on
[amine], and an inverse-first-order dependence on
[H2O]. In contrast to this prediction, the N-H bond
-d[1]
) kobs[1] where
dt
K1K2K3K4k5[H2N-sec-Bu]
kobs
)
(3)
[H2O]
cleavage process showed a second-order dependence on
[amine], rigorously ruling out pathway A.
Pathway B depicts direct deprotonation of the amine
substrate to form a coordinated H2O. The water ligand
would then be displaced by the amide anion to form the
product 3b. This pathway would give rise to the rate
expression in eq 4, which includes a first-order depen-
dence on 1, a first-order dependence on [amine], and a
zero-order dependence on [H2O]. This mechanism is
-d[1]
) kobs[1] where
dt
K12K13k14k15[H2N-sec-Bu]2
kobs
)
)
or
k-14[H2O] + k15
-d[1]
) kobs[1] where
dt
K16k17k15[H2N-sec-Bu]2
(6)
(7)
k-17[H2O] + k15
k6k7[H2N-sec-Bu]
kobs
)
(4)
k-6 + k7
k-14[H2O]
K12K13k14k15[H2N-sec-Bu]2
1
kobs
+
ruled out by the observed second-order dependence on
[amine] and the inhibition of the reaction by added
water.
1
K12K13k14[H2N-sec-Bu]2
Pathway C involves reversible phosphine dissociation
from 1 to create a position for amine coordination. After
amine coordination, proton transfer would occur to form
the amido ligand and water. In the final step, phos-
phine would reassociate with the Pd complex to displace
the water and form 3b. This mechanism would follow
the rate expression in eq 5, which predicts a first-order
dependence on [1], a first-order dependence on [amine],
and a zero-order dependence on [H2O]. There is no
k-17[H2O]
1
or
+
K16K17k15[H2N-sec-Bu]2 K16k17[H2N-sec-Bu]2
will be linear with a non-zero y-intercept. This behavior
is exactly what was observed in Figure 2. The complex
order in [H2O] shows that the mechanism involves
reversible formation of water followed by replacement
of the amine ligand by the amide to form 3b.
In addition, the reversible proton transfer step will
provide a kinetic isotope effect that originates from the
proton transfer equilibrium. The value of KH/KD for the
proton transfer and, therefore, the kH/kD for the overall
reaction is likely to be <1 as a result of the higher νOH
than νNH. An inverse isotope effect, kH/kD ) 0.7, was
observed for the reaction of 1-(µ-OD)2 with D2N-sec-Bu.
All of our kinetic data are consistent with the proton
transfer reaction of 1 with sec-butylamine proceeding via
-d[1]
) kobs[1] where
dt
kobs ) K8K9K10k11[H2N-sec-Bu]
(5)
phosphine dependence as a result of both dissociation
and association occuring before or as part of the rate-
(54) Driver, M. S.; Hartwig, J . F. J . Am. Chem. Soc. 1995, 117,
4708-4709.

5712 Organometallics, Vol. 16, No. 26, 1997
Driver and Hartwig
Sch em e 4. P a th D
amine-facilitated generation of a terminal hydroxo ligand
by one of the two kinetically indistinguishable mecha-
nisms in path D.
oxidative addition of acidic N-H bonds or those acti-
vated by chelation has been observed a number of
times,32,55-60 the oxidative addition of the less acidic NH
bonds of alkylamines and ammonia has only been
observed in a few cases and most of these are with early
metals.2 The reaction of ammonia with (PEt3)2Ir(C2H4)2-
Cl to give a dimeric iridium amido complex is the only
report of the oxidative addition of ammonia to a late
transition metal system.10 The only reports of the
oxidative addition of alkylamines have occurred with
Os clusters.7,8 Reactions of late transition metal com-
plexes with the NH bonds of activated amines has also
occurred by proton exchanges,44,50 but such σ-bonded
ligand exchanges with alkylamines are even more
unusual than oxidative addition of alkylamine N-H
bonds. Presumably, the absence of proton transfer
chemistry with alkylamines is due to their weak Brøn-
sted acidity and strong Lewis basicity.
The principle of microscopic reversibility states that
a reaction must follow the same mechanism in both the
forward and reverse directions. Since the reaction of 1
with sec-butylamine to produce 3b and H2O is revers-
ible, we studied the reaction in the reverse direction in
order to add further support for our mechanistic conclu-
sions, since the involvement of two amines in the N-H
cleavage process is counterintuitive. For the reverse
reaction, addition of water to 3b, the mechanisms shown
in path D of Scheme 4 would result in the rate
expressions shown in eq 8, assuming that k-14 or k-17
is irreversible. In order for the reaction to proceed, a
-d[3b]
) kobs[3b] where
dt
The σ-bonded ligand exchange is formally acid-base
chemistry, analogous to the deprotonation of an amine
by an alkali metal reagent. However, two features of
the palladium chemistry are remarkable. First, the
thermodynamics of the acid-base chemistry are the
opposite of what would be expected from pKa data.
Streitwieser has shown that the ion-pair acidities of
metal amides are highly dependent on the identity of
the metal.61 For example, the pK’s of the lithium and
cesium ion pairs of diphenylamide differ by 5 pK units
at 19.05 and 24.20, respectively. Some of this difference
can be explained by the aggregation number, but the
type of ion pair and the strength of the metal amide
interaction are also important. In the N-H bond
cleavage chemistry described in this work, the counter-
cation is the palladium fragment and the amide and
hydroxide are transition metal complexes rather than
alkali metal complexes. By the definition of ion-pair
acidities, the amine is actually more acidic than water
in these organic solvents when a palladium fragment
serves as the counterion. Second, the acid-base chem-
istry occurs with mild reagents. Both the palladium
hydroxo complex and amide products are air stable. This
property stands in contrast to the typical reagents that
react with amine N-H bonds such as organolithium
k-14k-15[H2N-sec-Bu][H2O]
kobs
)
or
k-14[H2O] + k15
k-17k-15[H2N-sec-Bu][H2O]
(8)
k-17[H2O] + k15
high [H2O] is necessary. Under these conditions, the
k-14[H2O] or k-17[H2O] term in the denominator is much
larger than k15 and the rate expression simplifies to the
one shown in eq 9. This expression predicts a first-order
-d[3b]
) kobs[3b] where
dt
kobs ) k-15[H2N-sec-Bu]
(9)
dependence on [3b], a first-order dependence on [amine],
and no dependence on [H2O]. Indeed, our kinetic data
showed these reaction orders. Since the rate expression
includes a first-order term in the product amine, the
amine produced over the course of the reaction will
catalyze the addition of water to 3b. In other words,
the involvement of the extra amine in the forward
reaction should give rise to autocatalytic behavior in
amine for the reverse reaction, and this autocatalytic
behavior was observed. In this case, the extra amine
appears to be used to generate a terminal amido or
hydroxo ligand that is more basic than the bridging
ligand and can deprotonate the water or amine sub-
strate.
(55) Samat, A.; Sala-Pala, J .; Guglielmetti, R.; Guerchais, J . Nouv.
J . Chem. 1978, 2, 13-14.
(56) Ladipo, F. T.; Merola, J . S. Inorg. Chem. 1990, 29, 4172-4173.
(57) Hsu, G. C.; Kosar, W. P.; J ones, W. D. Organometallics 1994,
13, 385-396.
(58) J ones, W. D.; Dong, L.; Myers, A. W. Organometallics 1995,
14, 855-861.
(59) Fornies, J .; Green, M.; Spencer, J . L.; Stone, F. G. J . Chem.
Soc., Dalton Trans. 1977, 1006-1009.
Con clu sion s
(60) Park, S.; Hedden, D.; Roundhill, D. M. Organometallics 1986,
5, 2151.
The N-H bond cleavage chemistry of amines with
late transition metals is not well-studied. Although the
(61) Krom, J . A.; Petty, J . T.; Streitwieser, A. J . Am. Chem. Soc.
1993, 115, 8024-8030.

Alkylamine N-H Bond Cleavage by Pd Hydroxides
Organometallics, Vol. 16, No. 26, 1997 5713
isomer). Anal. Calcd for C54H47NOP2Pd2‚Et2O (confirmed by
integrated 1H NMR spectra): C, 64.81; H, 5.35; N, 1.30.
Found: C, 64.52; H, 5.66; N, 1.30.
reagents or alkali metals, all of which are highly air
sensitive and which are typically incapable of being used
as catalysts. Thus, the palladium chemistry offers a
pathway for formation of a metal amido species that is
mild and that offers a means for catalytic modification
of the amine substrate.
P r ep a r a tion of syn -(P P h 3)(C6H5)P d (µ-OH)(µ-NH-2,6-
(CH3)2C6H3)P d (C6H5)(P P h 3) (2b). Into a screw-capped test
tube was weighed 186 mg (0.201 mmol) of 1. The solid
material was dissolved in 8 mL of THF, and H2N-2,6-
(CH3)2C6H3 (0.25 mL, 2.01 mmol) was added to the test tube.
The test tube was sealed, and the reaction was stirred at 70
°C for 1 h. A 31P{1H} NMR spectrum of the reaction mixture
indicated complete consumption of the starting material. The
reaction mixture was filtered through Celite, concentrated and
layered with Et2O. The THF/Et2O solution was cooled at -35
°C to yield 120 mg (58%) of white crystalline material. 1H
NMR (C6D6): δ -4.11 (s, 1H), 1.12 (s, 3H), 3.29 (t, J ) 5.0 Hz,
1H), 4.43 (s, 3H), 6.51-6.61 (m, 6H), 6.80 (t, J ) 7.0 Hz, 2H),
6.92-7.09 (m, 21H), 7.42 (d, J ) 6.0 Hz, 2H), 7.50-7.61 (m,
12H). 31P{1H} NMR (THF): δ 27.5. 13C{1H} NMR (C6D6): δ
17.9 (s), 24.6 (s), 118.9 (s), 122.2 (s), 126.2 (s), 127.1 (s), 128.3
(d, J ) 10.3 Hz), 128.9 (s), 129.2 (s), 130.0 (s), 131.4 (s), 131.8
(d, J ) 43.1 Hz), 134.5 (d, J ) 11.9 Hz), 136.3 (d, J ) 3.7 Hz),
149.9 (d, J ) 7.3 Hz), 151.3 (t, J ) 2.8 Hz). IR (cm-1, KBr):
3622 (w, νOH), 3326 (w, νNH). Anal. Calcd for C56H51NOP2-
Pd2: C, 65.38; H, 5.00; N, 1.36. Found: C, 65.03; H, 5.41; N,
1.20.
This result is important for the further development
of transition metal catalyzed chemistry of amines. For
example, our laboratory and Buchwald’s have developed
a catalytic process for the amination of aryl halides in
the presence of alkoxide base.33,34 Our work on proton
exchange reactions presented here along with studies
conducted on isolated palladium alkoxide complexes48
have shown that the alkoxide complexes (DPPF)Pd(Ar)-
(O-t-Bu) undergo proton transfer reactions with free
amines to generate reactive palladium amido complexes,
which then reductively eliminate coupled amine prod-
uct. These results suggest that alkoxide complexes may
be intermediates in the conversion of aryl halide com-
plexes to arylamido complexes in the amination chem-
sitry and, more generally, that alkoxide complexes may
be used for the production of amido complexes by N-H
activation.
P r ep a r a tion of syn -(P P h 3)(C6H5)P d (µ-OH)(µ-NH-t-Bu )-
P d (C6H5)(P P h 3) (3a ). Into a screw-capped test tube was
weighed 103 mg (0.111 mmol) of 1. H2N-t-Bu (8 mL) was
added to the test tube. The test tube was sealed and heated
at 70 °C with stirring for 1 h. The reaction mixture became
Exp er im en ta l Section
Gen er a l Meth od s. Unless otherwise noted, all reactions
and manipulations were performed in an inert atmosphere
glovebox or by using standard Schlenk techniques. Benzene,
toluene, THF, ether, and pentane solvents were distilled from
sodium/benzophenone prior to use. Amines were distilled from
CaH2, and triphenylphosphine was sublimed prior to use in
kinetic studies. Both H2O and D2O were degassed with N2(g)
before use in kinetic studies. All other chemicals were used
as received from commercial suppliers.
homogeneous.
A
31P{1H} NMR spectrum of the reaction
mixture indicated complete consumption of the starting mate-
rial. The reaction mixture was filtered through Celite and
concentrated. The THF solution was layered with Et2O and
cooled at -35 °C to yield 68.5 mg (63%) of white crystalline
material. 1H NMR (C6D6): δ -3.10 (s, 1H), 1.54 (t, 6.0 Hz,
1H), 1.64 (s, 9H), 6.83-6.94 (m, 24H), 7.57-7.66 (m, 16H). 31P-
{1H} NMR (THF): δ 26.2. 13C{1H} NMR (C6D6) δ: 36.7 (s),
59.1 (s), 122.3 (s), 127.1 (s), 128.3 (apparent t, J ) 5.0 Hz),
129.9 (s), 132.2 (d, J ) 42.2 Hz), 134.7 (apparent t, J ) 6.2
Hz), 138.5 (s), 150.7 (d, J ) 10.4 Hz). IR (cm-1, KBr): 3619
(w, νOH). Anal. Calcd for C52H51NOP2Pd2‚THF‚0.5Et2O (con-
firmed by integrated 1H NMR spectra): C, 63.92; H, 5.92; N,
1.29. Found: C, 63.84; H, 5.96; N, 1.39.
P r ep a r a tion of syn -(P P h 3)(C6H5)P d (µ-OH)(µ-NH-sec-
Bu )P d (C6H5)(P P h 3) (3b). Into a screw-capped test tube was
weighed 102 mg (0.110 mmol) of 1. The solid material was
dissolved in 8 mL of THF, and H2N-sec-Bu (55 µL, 0.550 mmol)
was added to the test tube. The test tube was sealed, and the
reaction was stirred at room temperature for 5 h. A 31P{1H}
NMR spectrum of the reaction mixture indicated complete
consumption of the starting material. The reaction mixture
was filtered through Celite. The THF solution was concen-
trated and layered with Et2O. The THF/Et2O solution was
cooled at -35 °C to yield 77.5 mg (72%) of white crystalline
material. 1H NMR (C6D6): δ -3.25 (s, 1H), 0.63 (t, J ) 7.4
Hz, 3H), 1.43 (m, 1H), 1.57 (apparent hept, J ) 6.9 Hz, 1H),
1.78 (d, J ) 6.3 Hz, 3H), 1.96 (apparent hept, J ) 6.6 Hz, 1H),
3.68 (m, 1H), 6.78-6.93 (m, 24H), 7.54-7.62 (m, 16H). 31P-
{1H} NMR (THF): δ 27.3 (s), 27.1 (s). 13C{1H} NMR (C6D6): δ
11.0 (s), 26.2 (s), 35.1 (s), 58.2 (s), 122.0 (s), 122.0 (s), 126.8
(s), 126.9 (s), 128.1 (d, J ) 12.1 Hz), 128.1 (d, J ) 12.1 Hz),
129.6 (s), 129.62 (s), 132.0 (d, J ) 41.8 Hz), 132.2 (d, J ) 42.1
Hz), 134.3 (d, J ) 11.7 Hz), 134.3 (d, J ) 11.7 Hz), 137.3 (d, J
) 2.4 Hz), 137.7 (d, J ) 2.8 Hz), 150.2 (d, J ) 9.7 Hz), 151.5
(d, J ) 10.1 Hz). IR (cm-1, KBr): 3611 (w, νOH), 3301 (w, νNH).
Anal. Calcd for C56H51NOP2Pd2: C, 63.68; H, 5.24; N, 1.43.
Found: C, 63.50; H, 5.32; N, 1.36.
P r ep a r a tion of [(P P h 3)(C6H5)P d (µ-OH)]2 (1). Into a 20
mL vial was weighed 150 mg (0.180 mmol) of trans-(PPh3)2-
Pd(C6H5)(I).62 The solid material was dissolved in 12 mL of
THF with stirring. CsOH (400 mg, 2.7 mmol) was added to
the vial as a solid. The reaction was stirred at room temper-
ature for 2 h.
A
31P{1H} NMR spectrum of the reaction
mixture indicated complete consumption of the starting mate-
rial. The reaction mixture was filtered through a medium-
fritted funnel containing Celite, and the resulting clear
solution was concentrated under vacuum. Pentane was added
to precipitate the product. The product was collected on a
medium-fritted funnel to give 137 mg (83%) of white solid. The
31P{1H} and 1H NMR spectra matched those previously
published.40
P r ep a r a tion of (P P h 3)(C6H5)P d (µ-OH)(µ-NHC6H5)P d -
(C6H5)(P P h 3) (2a ). Into a vial was weighed 128 mg (0.138
mmol) of 1. The solid material was dissolved in 8 mL of THF,
and 63 µL (0.693 mmol) of aniline was added to the reaction
solution. The reaction was stirred at room temperature for
30 min.
A
31P{1H} NMR spectrum of the reaction mixture
indicated complete consumption of the starting material and
the formation of 2a as a mixture of syn- and anti-isomers. The
reaction solution was filtered through Celite, concentrated, and
layered with Et2O. The THF/Et2O solution was cooled at -20
°C to yield 121 mg (88%) of amber crystalline 2a as predomi-
nately (>90%) the syn-isomer. 1H NMR: (C6D6, syn-isomer)
δ -3.40 (s, 1H), 2.75 (t, J ) 4.2 Hz, 1H), 6.63-6.78 (m, 5H),
6.81-6.89 (m, 20H), 7.19-7.22 (m, 6H), 7.44-7.50 (m, 12H),
7.62 (d, J ) 7.8 Hz, 2H); (C6D6, anti-isomer) δ -1.53 (d, J )
3.4 Hz, 1H), 1.55 (d, J ) 4.8 Hz, 1H). 31P{1H} NMR (THF): δ
34.1, 31.5 (anti-isomer), 29.8 (syn-isomer). IR (cm-1, KBr):
3603 (m, νOH), 3322 (w, νNH syn-isomer), 3308 (w, νNH anti-
P r epar ation of a n ti-[(P P h 3)(C6H5)P d(µ-NH-isoBu )]2 (3c).
Into a screw-capped test tube was weighed 186 mg (0.201
mmol) of 1. H2N-isoBu (10 mL) was added to the test tube.
The test tube was sealed, and the reaction was stirred at room
(62) Fitton, P.; J ohnson, M. P.; McKeon, J . E. J . Chem. Soc., Chem.
Commun. 1968, 6-7.

5714 Organometallics, Vol. 16, No. 26, 1997
Driver and Hartwig
temperature for 1 h, over which time the reaction mixture
became homogeneous.
Kin etic An a lysis of th e Rea ction of 1 w ith H 2N-sec-
Bu . Dependence on [H2N-sec-Bu]: [(PPh3)(C6H5)Pd(µ-OH)]2
was weighed into a vial (21.7 mg or 50.3 mg) and dissolved in
THF-d8 (1.3 mL or 1.8 mL). An aliquot of the standard
solutions (0.50 and 0.45 mL) was placed into a screw-capped
NMR tube. The tube was sealed with a cap containing a PTFE
septum and removed from the drybox. The appropriate
amount of H2N-sec-Bu (4.5-20.5 µL) was added to the NMR
tube by syringe right before data acquisition began. Reaction
A
31P{1H} NMR spectrum of the
reaction mixture indicated complete consumption of the start-
ing material. The reaction mixture was filtered through
Celite, concentrated, and layered with Et2O. The THF/Et2O
solution was cooled at -35 °C to yield 121 mg (61.5%) of white
crystalline material. 1H NMR (C6D6): δ -0.6 (m, 2H), 0.13 (d,
J ) 6.3 Hz, 6H), 0.68 (d, J ) 6.6 Hz, 6H), 1.83 (m, J ) 4.0 Hz,
2H), 2.63 (m, J ) 3.8 Hz, 2H), 3.62 (m, 2H), 6.83-7.07 (m,
24H), 7.52-7.69 (m, 16H). 31P{1H} NMR (THF) δ 29.4. 13C-
{1H} NMR (C6D6): δ 20.0 (s), 21.4 (s), 34.3 (s), 60.9 (s), 122.1
(s), 126.9 (s), 128.5 (apparent t, J ) 4.6 Hz), 129.9 (s), 132.7
(d, J ) 42.2 Hz), 134.6 (apparent t, J ) 6.0 Hz), 138.3 (s), 157.2
(d, J ) 11.0 Hz). IR (cm-1, KBr): 3336 (w, νNH), 3304 (w, νNH).
Anal. Calcd for C56H60N2P2Pd2: C, 64.93; H, 5.84; N, 2.70.
Found: C, 64.64; H, 5.89; N, 2.67.
P r ep a r a tion of tr a n s-(P P h 3)2P d (C6H5)(NC4H4) (4). Into
a vial was weighed 120 mg (0.130 mmol) of 1 with 170 mg
(0.649 mmol) of PPh3. The solids were dissolved in 8 mL of
THF and 45 µL (0.649 mmol) of pyrrole. The reaction was
stirred for 2 h. The 31P{1H} spectrum indicated complete
consumption of the starting material. The reaction solution
was filtered through Celite, concentrated, and layered with
Et2O. Cooling the THF/Et2O solution at -35 °C gave 183 mg
(91%) of light orange crystals. 1H NMR (C6D6): δ 6.36-6.48
(m, 3H), 6.52 (m, 2H), 6.65 (m, 2H), 6.91-7.05 (m, 18H), 7.11
(d, J ) 7.0 Hz, 2H), 7.31-7.38 (m, 12H). 31P{1H} NMR (THF):
δ 20.6. Anal. Calcd for C46H39NP2Pd: C, 71.37; H, 5.08; N,
1.81. Found: C, 71.13; H, 5.01; N, 1.77.
1
rates were monitored by H NMR spectroscopy at 25 °C using
an automated program that collected single-pulse experiments
with at least 1 min intervals between data acquisitions. The
resonance for the hydroxyl proton and an aromatic resonance
were used to monitor the concentration of 1 throughout the
course of the reaction. The following pseudo-first-order rate
constants were obtained at different concentrations of H2N-
sec-Bu (k, [H2N-sec-Bu]): 4.3 × 10-5 s-1, 0.089 M; 1.2 × 10-4
s-1, 0.15 M; 3.1 × 10-4 s-1, 0.27 M; 4.3 × 10-4 s-1, 0.30 M; 1.2
× 10-3 s-1, 0.45 M.
Dep en d en ce on [H2O]: [(PPh3)(C6H5)Pd(µ-OH)]2 was
weighed into a vial (41.2 mg, 0.0446 mmol) and dissolved in
2.5 mL of THF-d8. A 0.56 mL aliquot of the standard solution
was placed into a screw-capped NMR tube. The tube was
sealed with a cap containing a PTFE septum and removed
from the drybox. H2N-sec-Bu (10 µL, 0.10 mmol) and the
appropriate amount of H2O (3-14 µL) were added to the NMR
tube by syringe right before data acquisition began. Data were
acquired as described above. The following pseudo-first-order
rate constants were obtained at different concentrations of H2O
(k, [H2O]): 7.4 × 10-4 s-1, 0.30 M; 6.6 × 10-4 s-1, 0.49 M; 4.8
× 10-4 s-1, 0.69 M; 3.9 × 10-4 s-1, 0.88 M; 3.5 × 10-4 s-1, 1.0
M; 3.0 × 10-4 s-1, 1.4 M.
P r ep a r a tion of syn -[(P P h 3)(p-CH3C6H4)P d (µ-S-t-Bu )]2
(5). Into a screw-capped test tube was weighed 100 mg (0.10
mmol) of [(PPh3)(p-CH3C6H4)Pd(µ-OH)]2. The solid material
was dissolved in 8 mL of THF, and HS-t-Bu (21 mg, 0.24 mmol)
was added to the test tube. The test tube was sealed, and the
reaction was stirred at room temperature for 1 h. The reaction
Syn th esis of D2N-sec-Bu : H2N-sec-Bu (8 mL) was placed
into a 25 mL screw-capped test tube. D2O (8 mL) was added
to the test tube. The mixture was shaken vigorously. Enough
NaCl was added to the tube to saturate the D2O, and the amine
was extracted with n-Bu2O (2 × 5 mL). The ether layer was
then mixed with additional D2O (2 × 5 mL). The ether layer
was then collected, and the deuterated amine was isolated by
mixture turned dark red.
A
31P{1H} NMR spectrum of the
reaction mixture indicated complete consumption of the start-
ing material. The reaction mixture was filtered through
Celite, concentrated, and layered with Et2O. The THF/Et2O
solution was cooled at -35 °C to yield 56 mg (49%) of white
crystalline material. 1H NMR (C6D6): δ 1.21 (s, 9H), 1.47 (s,
9H), 2.05 (s, 6H), 6.62 (d, J ) 7.4 Hz, 4H), 6.94-7.05 (m, 18H),
7.53 (d, J ) 7.4 Hz, 4H), 7.62-7.68 (m, 12H). 13C{1H} NMR
(C6D6): δ 20.8 (s), 34.6 (s), 38.3 (s), 45.5 (s), 47.7 (s), 127.9 (t,
J ) 4.7 Hz), 129.5 (s), 131.1 (s), 133.0 (d, J ) 43.6 Hz), 135.2
(t, J ) 5.3 Hz), 139.6 (s), 147.3 (t, J ) 3.0 Hz). 31P{1H} NMR
(THF): δ 21.8.
distillation under N2(g).
A
1H NMR spectrum of the amine
showed no resonance for the amino protons.
Syn th esis of [(P P h 3)(C6H5)P d (µ-OD)]2. [(PPh3)(C6H5)Pd-
(µ-OH)]2 (125 mg) was weighed into a 20 mL vial and dissolved
in 5 mL of THF. D2O (10 mL) was added to the vial. The
mixture was stirred for 8 h. The product was collected on a
medium-fritted funnel. The 1H NMR spectrum showed no
resonance for the hydroxo protons.
Mea su r em en t of Isot op e E ffect for t h e R ea ct ion of
[(P P h 3)(C6H5)P d (µ-OD)]2 w ith D2N-sec-Bu . 1-µ-(OD)2 (5.8
mg, 0.0063 mmol) was weighed into a small vial and dissolved
in 0.45 mL of THF-d8. This solution was transferred to a
screw-capped NMR tube. The tube was sealed with a cap
containing a PTFE septum and removed from the drybox. D2N-
sec-Bu (15 µL of a 5.2 M solution in n-Bu2O, 0.078 mmol) was
added to the NMR tube by syringe. Data were collected as
described above. The psuedo-first order-rate constant obtained
were 1.9 × 10-4 and 2.1 × 10-4 s-1 resulting in kH/kD ) 0.7
(.1.
Rea ction of 3a w ith H2O. 3a (8.6 mg, 0.0088 mmol) was
weighed into a small vial and dissolved in 0.5 mL of THF. The
solution was transferred to a screw-capped NMR tube. The
tube was sealed with a cap containing a PTFE septum and
removed from the drybox. H2O (16 µL, 0.89 mmol) was added
to the reaction by syringe. The NMR tube was heated at 70
°C for 1.5 h. The 31P{1H} NMR spectrum indicated complete
conversion to 1.
Va r ia ble-Tem p er a tu r e Deter m in a tion of Keq for Rea c-
tion of 1 w ith H2NC6H5. Complex 1 (18.1 mg, 0.020 mmol)
was dissolved in 0.5 mL of a mixture of THF with a small
amount of toluene-d8 and transferred to a screw-capped NMR
tube containing a PTFE septum. H2NPh (1.8 µL, 0.020 mmol)
and H20 (8.8 µL, 0.49 mmol) were added to the NMR tube by
syringe. The NMR tube was placed into the NMR probe cooled
at -20 °C, and the sample was shimmed. 31P{1H} NMR
spectra were taken at 10 °C intervals up to 40 °C. The sample
was allowed to equilibrate for a minimum of 40 min and was
reshimmed at each temperature before data were acquired.
31P{1H} NMR spectra were acquired with a minimum 30 s
pulse delay (T1 ) 3.6 s for 1).
Mea su r em en t of Keq for Rea ction of 1 w ith H2N-sec-
Bu . Complex 1 (11 mg, 0.012 mmol) was weighed into a small
vial along with a small amount (<2 mg) of trimethoxybenzene
to act as an internal standard. The solids were dissolved in
0.6 mL of anhydrous THF-d8 and transferred to a screw-capped
NMR tube. The NMR tube was sealed with a cap containing
a PTFE septum and removed from the drybox.
A
1H NMR
Kin et ic An a lysis of t h e R ea ct ion of 3b w it h H 2O.
Dependence on [H2N-sec-Bu]: Complex 3b (30.4 mg, 0.0311
mmol) was dissolved in 2.5 mL of THF-d8. A 0.50 mL aliquot
of the stock solution was placed in a screw-capped NMR tube.
The tube was sealed with a cap containing a PTFE septum
and removed from the drybox. D2O (110 µL, 5.50 mmol) and
spectrum was taken. Water (5.0 µL, 0.28 mmol) and H2N-
sec-Bu (2.4 µL, 0.024 mmol) were added to the NMR tube by
syringe. The reaction was left at room temperature for 12 h.
The equilibrium concentrations were determined by integra-
tion of the 1H NMR spectrum. Keq was calculated to be 12.

Alkylamine N-H Bond Cleavage by Pd Hydroxides
Organometallics, Vol. 16, No. 26, 1997 5715
the appropriate amount of H2N-sec-Bu (3.1-18.5 µL) were
septum. The NMR tube was removed from the drybox, and
10.0 µL (0.500 mmol) of D2O was added to the tube by syringe.
The reaction was monitored by 1H NMR spectroscopy at 70
°C. Data were acquired as described above.
added to the NMR tube by syringe right before data acquisition
1
began. Reaction rates were monitored by H NMR spectros-
copy at 70 °C. Data were acquired as described above. The
following pseudo-first-order rate constants were obtained at
different concentrations of H2N-sec-Bu (k, [H2N-sec-Bu]): 5.2
× 10-4 s-1, 0.050 M; 7.8 × 10-4 s-1, 0.075 M; 7.7 × 10-4 s-1
,
Ack n ow led gm en t. This work was generously sup-
ported by the Department of Energy (Grant No. DE-
RG02-96ER14678). We also gratefully acknowledge
support from a DuPont Young Professor Award, a Union
Carbide Innovative Recognition Award, a National
Science Foundation Young Investigator Award, a Drey-
fus Foundation New Faculty Award, and a Camile
Dreyfus Teacher-Scholar Award for support for this
work. J .F.H. is a fellow of the Alfred P. Sloan Founda-
tion. We thank Alpha/Aesar for donation of palladium
chloride.
0.099 M; 1.0 × 10-4 s-1, 0.12 M; 1.4 × 10-3 s-1, 0.15 M; 1.3 ×
10-3 s-1, 0.17 M; 1.5 × 10-3 s-1, 0.20 M; 2.2 × 10-3 s-1, 0.25
M; 3.3 × 10-3 s-1, 0.29 M.
Dep en d en ce on [H2O]: 3b (28.4 mg, 0.0290 mmol) was
dissolved in 2.0 mL of THF-d8. A 0.40 mL aliquot of the stock
solution was placed in a screw-capped NMR tube. The tube
was sealed with a cap containing a PTFE septum and removed
from the drybox. H2N-sec-Bu (5.8 µL, 0.057 mmol) and the
appropriate amount of D2O (68-137 µL) were added to the
NMR tube by syringe right before data acquisition began.
Reaction rates were monitored by 1H NMR spectroscopy at 70
°C. Data were acquired as described above. The following
pseudo-first-order rate constants were obtained at different
concentrations of H2O (k, [H2O]): 1.6 × 10-3 s-1, 7.31 M; 1.6
× 10-3 s-1, 10.2 M; 1.7 × 10-3 s-1, 12.7 M.
Su p p or tin g In for m a tion Ava ila ble: Plots of kinetic data
for the reaction of sec-butylamine with 1 and the reaction of
3b with H2O and sec-butylamine (3 pages). Ordering informa-
tion is given on any current masthead page.
Au toca ta lytic Beh a vior : 3b (19.8 mg, 0.0202 mmol) was
dissolved in 0.50 mL of THF-d8 and transferred into an NMR
screw-capped NMR tube sealed with a cap containing a PTFE
OM970856L