Detailed mechanistic information on methane elimination from a methyl(hydrido)platinum(IV) complex. Relevance for the mechanism of methane activation

Article abstract of DOI:10.1021/om9901426

The kinetics of methane reductive elimination from both isomers of the methyl(hydrido)-platinum(IV) complex [Pt(H)(CH₃)₂(BPMA)]⁺ (BPMA = bis(pyridylmethyl)amine) were studied by ¹H NMR spectroscopy. The application of high-pressure techniques enabled the determination of the activation volume for methane elimination (+6 cm³/mol for the cis isomer, ca. +8 cm³/mol for the trans isomer). The low activation volume together with results on the rate and stereochemistry of deuterium incorporation into one methyl group of the cis isomer provides the basis for a suggested mechanism that involves five-coordinate platinum as a steady-state intermediate.

Full text of DOI:10.1021/om9901426

4
156
Organometallics 1999, 18, 4156-4164
Deta iled Mech a n istic In for m a tion on Meth a n e
Elim in a tion fr om a Meth yl(h yd r id o)p la tin u m (IV)
Com p lex. Releva n ce for th e Mech a n ism of Meth a n e
Activa tion
Ulrich Fekl, Achim Zahl, and Rudi van Eldik*
Institute for Inorganic Chemistry, University of Erlangen-N u¨ rnberg,
Egerlandstrasse 1, D-91058 Erlangen, Germany
Received March 1, 1999
The kinetics of methane reductive elimination from both isomers of the methyl(hydrido)-
+
platinum(IV) complex [Pt(H)(CH
3
)
2
(BPMA)] (BPMA ) bis(pyridylmethyl)amine) were
1
studied by H NMR spectroscopy. The application of high-pressure techniques enabled the
determination of the activation volume for methane elimination (+6 cm /mol for the cis
isomer, ca. +8 cm /mol for the trans isomer). The low activation volume together with results
3
3
on the rate and stereochemistry of deuterium incorporation into one methyl group of the cis
isomer provides the basis for a suggested mechanism that involves five-coordinate platinum
as a steady-state intermediate.
2
10
In tr od u ction
The catalytic oxidation of methane to methanol at low
to give an η -methane complex (B), which can inter-
convert into a five-coordinate methyl(hydrido)platinum-
IV) complex (C). Five-coordinate Pt is unstable, but
IV
(
temperatures in homogeneous solution remains an
important challenge. Considerable efforts were invested
in the search for catalytic systems and in the under-
standing of the underlying reaction mechanisms. Im-
pressive progress has been made in the study of the
can be stabilized by an additional ligand to give a
complex of type D. In Scheme 1, the additional ligand
is chloride, but in similar systems nitrogen donors have
instead been used successfully.
We are interested in the initial steps of alkane
activation via oxidative addition and the microscopic
reverse of these steps. It is important to note that
methyl(hydrido)platinum(IV) complexes are less likely
as intermediates in cases where Pt(IV) is not stabilized
sufficiently. A viable pathway for methane activation
II
IV
mechanism of methane activation by Pt /Pt complexes
1
,2
in aqueous solution (Shilov’s reaction ) or other protic
3
solvents. The activation of methane can involve methyl-
(hydrido)platinum(IV) complexes, although they need
not necessarily be involved in all catalytic systems.
These have been postulated as crucial intermediates
2
is then the direct deprotonation of the η -methane
II
both in the protonolysis of some methyl Pt complexes
complex (B) to form the methylplatinum(II) complex (E).
However, alkane activation proceeds via oxidative ad-
dition when Pt(IV) is stabilized strongly. This was
demonstrated in an elegant way by showing that
alkanes can add to labilized Pt(II) complexes to give
alkyl(hydrido)platinum(IV) complexes in a stoichiomet-
and in the activation of methane by several platinum
complexes and were therefore synthesized.4 The mech-
anism of Shilov’s reaction, in accord with Shilov’s
original proposal and the mechanistic conclusions drawn
by Bercaw et al.,3 is shown in Scheme 1, which depicts
a methane activation mechanism involving oxidative
-9
,7
11
ric reaction. A pendant nitrogen donor arm was used
II
addition. A labile solvated Pt species (A) binds methane
for the stabilization of Pt(IV). The use of nitrogen donor
ligands in alkane activating systems is of increasing
importance, as also recently illustrated for the CATA-
*
Corresponding author. Fax: (international) +49(0)9131/852 7387.
E-mail: vaneldik@anorganik.chemie.uni-erlangen.de.
1) Goldshleger, N. F.; Eskova, V. V.; Shilov, A. E.; Shteinman, A.
A. Zh. Fiz. Khim. 1972, 46, 1353.
2) Sen, A.; Lin, M.; Kao, L.-C.; Hutson, A. C. J . Am. Chem. Soc.
992, 114, 6385.
3) For recent and comprehensive review, see: Stahl, S. S.;
Labinger, J . A.; Bercaw, J . E. Angew. Chem., Int. Ed. Engl. 1998, 37,
1
2
LYTICA system, which is based on a bipyrimidine-
Pt(II) complex. We do not address the question in this
study whether the CATALYTICA system involves me-
thyl(hydrido)platinum(IV) complexes or (more likely)
direct deprotonation. The present investigation serves
(
(
1
(
a
2
181.
(
4) Hill, G. S.; Rendina, L. M.; Puddephatt, R. J . Organometallics
2
1
995, 14, 4966.
(10) The η -methane complex is consistently shown throughout this
(
5) De Felice, V.; De Renzi, A.; Panunzi, A.; Tesauro, D. J . J .
Organomet. Chem. 1995, 488, C13.
6) Stahl, S. S.; Labinger, J . A.; Bercaw, J . E. J . Am. Chem. Soc.
995, 117, 9371.
7) Stahl, S. S.; Labinger, J . A.; Bercaw, J . E. J . Am. Chem. Soc.
996, 118, 5961.
8) Canty, A. J .; Dedieu, A.; J in, H.; Milet, A.; Richmond, M. K.
Organometallics 1996, 15, 2845.
9) O’Reilly, S. A.; White, P. S.; Templeton, J . L. J . Am. Chem. Soc.
996, 118, 5684.
paper with a symmetrically (via two hydrogen atoms) bonded methane.
This is an arbitrarily chosen convention. It may also be unsymmetri-
cally bonded, such that basically one C-H bond functions as a donor.
Since rotation of bonded methane is extremely facile and practically
activationless (see ref 42), our conclusions concerning H/D exchange
(
1
1
(
2
(see discussion) do not depend on the exact geometry of η -bonded
(
methane.
(11) Wick, D. D.; Goldberg, K. I. J . Am. Chem. Soc. 1997, 119, 10235.
(12) Periana, R. A.; Taube, D. J .; Gamble, S.; Taube, H.; Satoh, T.;
Fuji, H. Science 1998, 280, 560.
(
1
1
0.1021/om9901426 CCC: $15.00 © 1999 American Chemical Society
Publication on Web 09/04/1999

Methane Elimination from Methyl(hydrido)platinum
Organometallics, Vol. 18, No. 20, 1999 4157
Sch em e 1
as a model study for systems that are known to proceed
via oxidative addition. Since it still remains extremely
difficult to obtain direct mechanistic information on the
initial steps of alkane activation via oxidative addition,
since it is known to eliminate methane on a convenient
time scale. We observed two remarkable facts concern-
ing the formation of the cis and trans forms of 2
(referring to the position of the hydride relative to the
amine nitrogen). First, the protonation to give the cis
isomer 2a is much faster than the protonation to give
the trans isomer 2b. Second, the isomerization cannot
occur intramolecularly, but only in the presence of
unprotonated [Pt(CH3)2(BPMA)] (1). The deprotonation/
protonation mechanism shown in Scheme 2 is suggested
to account for isomerization. A bimolecular reaction
between 1 and 2 is a mechanistic alternative for
isomerization. However, we found no evidence for this
pathway, e.g., no binuclear complex in the reaction
mixture, although methyl(hydrido)platinum(IV) com-
plexes involving a bridging hydride can be very stable.2
The observed isomerization behavior allowed us to study
the reductive elimination of methane from 2a and 2b
independently. We could completely prevent the forma-
tion of the trans isomer 2b: the formation of the cis
isomer 2a is kinetically controlled when 1 is protonated
by a strong acid in dichloromethane or acetone. If an
excess of acid is used (5-10% is sufficient), no trans
isomer 2b can form. This excess acid is consumed within
7
we follow a proposal that the study of methane reduc-
tive elimination can contribute to the understanding of
this type of methane activation mechanism, since both
processes must pass through the same transition state
on the basis of microscopic reversibility.
This paper reports high-pressure kinetic data for
+
methane elimination from [Pt(H)(CH3)2(BPMA)] (BPMA
)
bis(pyridylmethyl)amine), first synthesized by Pud-
1
3
dephatt and co-workers. Earlier studies on related
systems have shown that oxidative addition reactions
are characterized by significantly negative volumes of
activation, whereas reductive elimination is character-
0,21
1
4-19
ized by strongly positive volumes of activation.
We
report a surprisingly low positive volume of activation
+
for methane elimination from [Pt(H)(CH3)2(BPMA)] .
From this value together with results from a deuterium
isotope scrambling study, the conclusion is drawn that
breakage of the bond trans to hydride contributes
significantly to the observed rate constant, but the
reverse (ring closure) reaction cannot be neglected, since
its rate is of the same order of magnitude as the
+
seconds to give [Pt(BPMA)X] (4), which has no further
2
intramolecular proton transfer to give an η -methane
influence on the actual reaction of interest. The net
result is that 1 is completely protonated to give 2a , but
no excess acid is present in the reaction mixture.
However, the use of a small excess of acid at the start
of the reaction guaranteed that no complex 1 is left.
Since the absence of 1 makes isomerization impossible,
the only reaction then observed is the elimination of
methane from 2a to give 3. This can be seen from Figure
complex.
Resu lts
Isom er iza t ion of t h e Com p lex [P t (H )(CH 3)2-
+
(
(
BP MA)] . We selected the complex [Pt(H)(CH3)2-
+
BPMA)] (2) (see Scheme 2) for our mechanistic studies,
1
. On the other hand, addition of a substoichiometric
(
13) J enkins, H. A.; Yap, G. P. A.; Puddephatt, R. J . Organometallics
997, 16, 1946.
14) de Waal, D. J . A.; Gerber, T. I. A.; Louw, W. J .; van Eldik, R.
Inorg. Chem. 1982, 21, 2002.
15) Leipoldt, J . G.; Steynberg, E. C.; van Eldik, R. Inorg. Chem.
987, 26, 3068.
16) Venter, J . A.; Leipoldt, J . G.; van Eldik, R. Inorg. Chem. 1991,
0, 2207.
17) Anhaus, J .; Bajaj, H. C.; van Eldik, R.; Nevinger, L. R.; Keister,
J . B. Organometallics 1989, 8, 2903.
18) D u¨ cker-Benfer, C.; van Eldik, R.; Canty, A. J . Organometallics
994, 13, 2412.
19) Fujita, E.; van Eldik, R. Inorg. Chem. 1998, 37, 360.
amount of acid (e.g., 0.7 equiv) leads to a situation where
a fraction of the platinum is still in the form of complex
1 (30% of the total platinum concentration in our
example). As mentioned above, 1 efficiently mediates
the isomerization between the Pt(IV) isomers. Under
these conditions, the two isomers are in equilibrium
1
(
(
1
(
3
(
(
(20) Hill, G. S.; Puddephatt, R. J . J . Am. Chem. Soc. 1996, 118, 8745.
(21) Hill, G. S.; Vittal, J . J .; Puddephatt, R. J . Organometallics 1997,
16, 1209.
1
(

4
158 Organometallics, Vol. 18, No. 20, 1999
Fekl et al.
Sch em e 2
F igu r e 2. Elimination of methane from an equilibrium
mixture of 2a and 2b. The fast equilibration was achieved
by addition of 0.65 equiv of HBF to 1, such that 35% of
4
1
F igu r e 1. Consecutive H NMR spectra (400 MHz, Pt-H
region) for the elimination of methane from 2a -BF
4
in CD
2
-
Cl at 25 °C, time interval ) 1 h. Note that no trans isomer
2
unprotonated 1 (which mediates the isomerization, see
Scheme 2) was available. The observed rate fits very well
to a single exponential.
2
b (signal at -19.2 ppm) is formed since 10% excess of acid
was used.
Kin et ics of Met h a n e R ed u ct ive E lim in a t ion .
Ra te Con sta n ts, Tem p er a tu r e, a n d P r essu r e De-
p en d en ce. To obtain kinetic data on the reductive
during methane elimination, as shown in Figure 2.
From the integrated NMR signals the equilibrium
constant for isomerization was determined to be K )
1
elimination of methane, at least 15 H NMR spectra
[
2b]/[2a ] ) 0.470(5) at 25 °C in CD2Cl2. The method of
were recorded during at least 3 half-lives of the reaction.
The concentration of the hydride was followed by
integration of the hydride signal, and a clean first-order
behavior was found. Control experiments showed that
the reaction rate observed did not depend on the initial
platinum concentration employed. If intermediates are
involved in these reactions, they must be very reactive
and do not accumulate to a concentration that makes
them observable by NMR.
preventing isomerization by addition of a small excess
of acid also works for mixtures of 2a and 2b. An
equilibrium mixture of 2a and 2b can be prepared as
described above. Addition of an extra amount of acid
such that nothing of 1 is left immediately quenches the
isomerization process. Elimination from the isomer 2a
cis
can then be seen with a rate constant kobs , which is
identical to the rate constant obtained for a solution
containing only 2a . Elimination from 2b occurs as a
We found the rate constants for methane elimina-
tion2 from 2a and 2b to be kobs ) 1.10(2) × 10 and
2
cis
-4
totally independent process, is much slower, and results
trans
trans
-6
-1
in kobs
.
kobs
) 4.83(5) × 10
s
at 25 °C in dichloro-

Methane Elimination from Methyl(hydrido)platinum
Organometallics, Vol. 18, No. 20, 1999 4159
Ta ble 1. Kin etic P a r a m eter s for Meth a n e
can be more difficult and is often subjected to larger
experimental errors than activation volumes. It should,
however, be noted that the activation entropy for
elimination from the cis complex shows the same
peculiarity as the activation volume. A positive value
a
Elim in a tion fr om 2a a n d 2b in CD
2
Cl
2
2
a
2b
-
1
-4
-6
k25 °C/s
(1.10 ( 0.02) × 10
(4.83 ( 0.05) × 10
133 ( 5
#
-1
∆
∆
∆
H /kJ mol
104 ( 3
#
-1
-1
-1
-1
S /J K mol
+28 ( 10
+6.2 ( 0.3
+101 ( 17
of 28 ( 10 J K mol was found, which is very low for
#
3
-1
V /cm mol
+8 ( 1, limited pressure
26
a reductive elimination reaction.
range available24
Deu ter iu m Scr a m blin g a n d Kin etic Deu ter iu m
Effect for th e Cis Com p lex. The incorporation of
a
For further information see supporting Figures S1-S4.
deuterium into the methyl groups of [Pt(D)(CH3)2-
+
(
BPMA)] (2a -P t-D) has been shown by Puddephatt et
1
3
al. to depend on the way the deuterated acid (DCl was
used) is generated. When a stoichiometric amount of
MeCOCl and excess CD3OD were used, deuterium
scrambling was observed, while generation of DCl using
Me3SiCl and D2O did not lead to scrambling. We
preferred to avoid the presence of protic solvents in the
reaction mixture, since it is known that protic hydrogens
can form strong hydrogen bonds with metal-bonded
2
7
hydrides. This could lead to additional pathways for
scrambling. On the other hand, the chloride ion in DCl
could complicate the reaction, since chloride is a rela-
tively good nucleophile and could coordinate to platinum
in some short-lived intermediate species. Therefore we
preferred to protonate N-deuterated 1 with water-free
F3CSO3D in acetone-d6 to give 2a-P t-D (compare Scheme
F igu r e 3. Pressure dependence of the elimination of
methane from 2 in CD Cl . For graphical reasons, 1.0 is
). ∆V is obtained according to δ(ln k)/
δp ) -∆V /RT. The measurements were done at 30.0 °C
2
2
trans
#
added to ln(kobs
#
3
).
cis
trans 24
(k
obs ) and 38.0 °C (kobs
).
Following this procedure, it could be demonstrated
that deuterium scrambling occurs with high selectivity
into the methyl group trans to amine. This is shown in
Figure 4, where the methyl region of the 1H NMR
spectrum after ca. 3 half-lives of 2a can be seen. The
methane-d2. The fact that the trans isomer eliminates
methane much slower by a factor of 20 has important
implications for the mechanism of methane elimination
discussed below. The temperature dependence of these
reactions was studied over the range23 20-40 °C and
CH signal trans to amine has a “twin signal” at higher
3
field strength, which appears only in the deuteration
2
5-38 °C for 2a and 2b, respectively. The calculated
experiments. At high resolution, H-D coupling can be
observed, confirming the assignment as CH2D. Since the
overlap of the CH3 and CH2D signals does not allow
exact integration, the degree of deuteration (discussed
below) was determined from the integration of the
hydride signal. CHD2 signals are not observed. In
contrast to the observation for the trans group, deu-
teration cis to amine occurs only to a negligible extent.
Notably, the methyl group of the elimination product
is deuterated considerably. We observed the formation
of CH4 and CH D but not of CH D under these
thermal activation parameters are summarized in Table
1
1
. Kinetic studies were conducted under pressure up to
50 MPa in order to obtain the activation volume for
the elimination of methane. The dependence of (ln k)
on the applied pressure, which appears to be linear in
the pressure range accessible, is shown in Figure 3,
#
3
-1
which results in ∆V obs ) +6.2(3) cm mol for the cis
3
-1
24
isomer and + 8(1) cm mol for the trans isomer. The
observed low positive activation volumes contrast the
high positive activation volumes usually found for
3
-1
reductive eliminations, e.g., +17 cm mol for elimina-
3
2
2
tion of ethane1 from (phen)Pd(I)(Me)3 or +20 cm mol
8
3
-1
conditions, indicating again that the protonation is
irreversible if excess acid is used. This observation also
suggests that deuterium scrambling, like the elimina-
tion process, occurs intramolecularly. In agreement with
the suggested isomerization mechanism given above,
multiple deuterium incorporation occurs and CH2D2 is
formed (at a later stage in the reaction) only if a
substoichiometric amount of acid is used. In order not
to complicate the kinetic scheme, an excess of acid was
used in the deuterium scrambling studies.
1
6
for elimination of H2 from a ruthenium cluster. It
should further be kept in mind that the molar volume
of liquid methane25 is ca. 40 cm mol . The mechanistic
3
-1
implication of the low activation volume will be dis-
cussed below. The interpretation of activation entropies
(
22) kcis was determined using 5-10% excess of HBF
4
‚OEt
2
(F
3
-
,
CSO H gave identical kinetics), whereas for the measurement of ktrans
3
0
.65 equiv of acid was first added, leading to the formation of an
equilibrium mixture of 2a and 2b. Then the equilibration was “frozen”
by adding an excess of acid. The reaction of 2a and the much slower
It is of interest to compare the deuterium scrambling
rate with the rate of methane elimination. Starting from
the Pt-deuterated complex, two parallel reactions oc-
reaction of 2b were thus followed separately. Alternatively, k
be obtained (but with less accuracy) from the kinetic time trace of an
equilibrium mixture along with the known values of k and K.
23) For the limitations of the temperature range, see the Supporting
Information.
24) The more restricted pressure range for the trans isomer stems
2
could
1
(
(
(26) Although elimination reactions are generally accompanied by
#
from the fact that the glass walls accept protons under high pressure,
which leads to isomerization to give the more reactive cis isomer. An
enhancement in reactivity was observed under pressures higher than
those plotted in Figure 3.
a very positive activation entropy, the ∆S values for alkane elimina-
tion cover a wide range (Gould, G. L.; Heinekey, D. M. J . Am. Chem.
Soc. 1989, 111, 5502), which is indicative of different rate-determining
steps or composite rate laws.
(
25) Terry, M. J .; Lynch, J . T.; Bunclark, M.; Mansell, K. R.;
(27) Crabtree, R. H.; Siegbahn, P. E. M.; Eisenstein, O.; Rheingold,
A. L.; Koetzle, T. F. Acc. Chem. Res. 1996, 29, 348.
Staveley, L. A. K. J . Chem. Thermodyn. 1969, 1, 413.

4
160 Organometallics, Vol. 18, No. 20, 1999
Fekl et al.
Sch em e 3
F igu r e 4. 1H NMR spectrum (300 MHz, CH
the reaction mixture obtained from 2a -P t-D (see Scheme
) after a reaction time of 5 h at 30 °C in acetone-d . The
symbols refer to CH trans to amine (filled circles) in 2a -
P t-D, CH cis to amine in 2a -P t-D (filled diamonds), and
CH in 3 (filled squares). The corresponding symbols of
smaller size designate the Pt satellites. The open symbols
refer to the corresponding CH D groups of the complexes
a -C-D and 3-C-D. It can be seen that 2a -P t-D selectively
incorporates deuterium in its methyl group trans to amine.
The methyl group of the product, complex 3, is deuterated
to considerable extent, implying that the methyl group cis
to amine is eliminated during the reaction and not the
trans group.
region) of
3
F igu r e 5. Concentration profile (as followed by NMR, at
30 °C in acetone-d ) for the reaction of complex 2a -P t-D
6
3
6
(shown as triangles). It undergoes two parallel reactions:
3
deuterium scrambling into the methyl group trans to amine
to give 2a -C-D and reductive elimination to give 3. The
scrambling product 2a -C-D eliminates at a rate very close
to that of 2a -P t-D and gives 3-C-D. The fact that the
scrambling is almost 2 times faster than the elimination
has implications for the suggested mechanism, which is
discussed in the text.
3
3
1
95
2
2
rate constant of 1.3 × 10 s-1 (at 30 °C in acetone) is
obtained compared to when the kinetics of the Pt-H
complex 2a are followed in acetone-H6. From this the
mechanistically important conclusion can be drawn that
k /k is close to 1. From the initial slope, k /k can be
-4
cur: the irreversible elimination of methane and the
essentially irreversible migration of deuterium from Pt
to CH3 trans to amine, shown in Figure 5 (see Experi-
mental Section for more details). It was observed that
H
D
H
D
2
8
estimated to be higher than 0.7 but lower than
1.2.
scr
the rate constant for scrambling (kobs ) is significantly
(28) It might be suggested that preparation of an analogous complex
larger (by a factor of 1.6) than the rate constant for
containing deuterated methyl groups would lead to a better determi-
elim
nation of k /k , since it prevents formation of the Pt-H complex.
methane elimination (kobs ). Noteworthy, the concen-
H
D
However, the secondary isotope effect of six deuterium atoms will lead
to a systematic error that could be as large as the uncertainty of our
determination.
tration-time trace for the reaction of the sum of Pt-H
and Pt-D fits well to a single exponential, and the same

Methane Elimination from Methyl(hydrido)platinum
Organometallics, Vol. 18, No. 20, 1999 4161
Discu ssion
Stea d y-Sta te Tr ea tm en t. First, we consider the
methane elimination from nondeuterated 2a , where
scrambling does not complicate the treatment. The
upper half of Scheme 3 (omit SS2 and do not distinguish
between H and D) resembles that situation. We consider
the possibility that 2a is in a preequilibrium with the
Su ggested Rea ction Mech a n ism . The observation
that the isomers 2a and 2b eliminate methane at
significantly different rates has important implications
for the reaction mechanism. It has been proposed before
that elimination from Pt(IV) and other octahedral d
complexes takes place via a five-coordinate intermediate
6
2
η -methane complex and that the extrusion of methane
from the latter is the rate-determining step (this mech-
2
9,30
and thus is favored when a ligand easily dissociates.
anism would be analogous to the mechanism for ben-
In the system under study, the amine nitrogen cannot
dissociate easily, but the pyridyl residues can dissociate
to form a five-coordinate species. This is shown in
Scheme 3 (for the cis isomer 2a , as structure SS1). The
observation that the trans isomer 2b eliminates meth-
ane by a factor of 20 slower confirms this assumption.
Breakage of the bond trans to a strong σ-donor (like
hydride or methyl) is facile. It is known that the trans
influence of hydride even exceeds the trans influence
3
3
zene elimination from [Tp′Rh(H)(Ph)(N-neopentyl)]).
This would lead to deuterium incorporation cis to amine
prior to elimination, which was not observed. A back
2
reaction can be ruled out for the η -bound methane cis
to amine; thus a rate constant k-3 need not be consid-
ered. The five-coordinate complex is treated as a steady-
state intermediate (SS1), which leads to the expression
for the observed rate constant for methane elimination
(eq 1).
9
,31
of methyl.
In 2a , the pyridyl residue is trans to
hydride, whereas in 2b, it is trans to the weaker σ-donor
methyl. This certainly contributes to the lower reactivity
of 2b. In addition, elimination from complex 2b must
involve a more complicated geometrical rearrangement,
since a methyl group in the position trans to amine
exists in the common product 3 and in 2a , but not in
k k
1 3
elim
kobs
)
(1)
k-1 + k3
In many cases further simplifications of the steady-
state expression are possible. These will be tested in the
following discussion. If k , k , eq 1 would reduce to
2
b, where the hydride is trans to amine. The mechanism
for elimination from 2b is probably similar to elimina-
tion from 2a , as indicated by the similar volume of
activation. However, it might involve an additional
isomerization step. For this reason we discuss the
elimination from 2a in more detail, since it is the
simpler system.
3
-1
2
kobs ) k (k /k ) ) k K
3
1
-1
3
preeq
. The formation of the η -
methane complex would be rate determining following
a rapid preequilibration between 2a and SS1. This
possibility would lead to a kinetic isotope effect in the
34
range between 3 and 5, stemming from the k3 term,
On the basis of the high-pressure kinetic and deute-
rium isotope effect data we propose the mechanism
shown in Scheme 3, which presents a unified view for
deuterium scrambling and methane elimination. It
offers a straightforward explanation for the fact that
deuterium is incorporated selectively trans to amine.
while Kpreeq should not show a significant deuterium
isotope effect. This possibility is ruled out by the
observed isotope effect of close to unity.
On the other hand, if k3 . k-1, eq 1 would reduce to
kobs ) k1. Thus ring opening, that is dissociation of the
pyridine residue, would be rate determining. Although
this possibility would lead to the observed isotope effect
of close to unity and the observed high activation
enthalpy, it is also discarded for the following reasons.
From high-pressure studies on pyridine dissociation
from octahedral complexes it is known that an activa-
2
The formation of an η -bound methane trans to amine
must be reversible, since no fast and irreversible step
3
2
leads to methane elimination. On the other hand, a
2
similar η -bound methane formed cis to amine can be
substituted very rapidly by the uncoordinated pyridyl-
methyl residue. The fact that the scrambling and
elimination rates are of the same order of magnitude
confirms the hypothesis that a common mechanism is
operative. A detailed analysis of the scrambling rate
versus elimination rate will reveal semiquantitative
information on the relative values of k1, k-1, k2, and k3.
It should again be noted that scrambling is observed,
but no isomerization. This implies that the five-
coordinate species SS1 and SS2 are stereochemically
rigid on the time scale of the ring-closure and intra-
molecular proton-transfer reactions they undergo.
3
-1
35-38
-1
tion volume of close to 20 cm mol can be expected,
3
which contrasts the low activation volume of 6 cm mol
observed here.
Furthermore, this simplification is disproved by the
deuterium scrambling results, which are discussed on
the basis of the complete mechanism shown in Scheme
3. Deuterium scrambling is essentially irreversible39
(see Figure 5); omitting k-2 and applying steady-state
conditions leads to eq 2:
(
33) J ones, W. D.; Hessell, E. T. J . Am. Chem. Soc. 1992, 114, 6087.
(29) Goldberg, K. I.; Yan, J .; Breitung, E. M. J . Am. Chem. Soc. 1995,
(34) A k /k of 5 at room temperature is the maximum kinetic
H
D
1
17, 6889.
isotope effect calculated from the Pt-H vibration frequency, whereas
a value close to 3 was reported experimentally: Stahl, S. S.; Labinger,
J . A.; Bercaw, J . E. J . Am. Chem. Soc. 1996, 118, 5961.
(35) Sullivan, T. R.; Stranks, D. R.; Burgess, J .; Haines, R. I. J .
Chem. Soc., Dalton Trans. 1977, 1460.
(
30) Milstein, D. J . J . Am. Chem. Soc. 1982, 104, 5227.
(31) Wilkins, R. G. Kinetics and Mechanism of Reactions of Transi-
tion Metal Complexes, 2nd ed.; VCH: Weinheim, 1991.
32) It is important to note that isomerization to give the trans
(
hydride 2b was not observed under these conditions. Although the
equilibrium is not too unfavorable for interconversion of 2a into 2b,
there is no low-energy path for an intramolecular isomerization.
Obviously, the kinetic barrier for placing the methane hydrogen in SS2
trans to amine is higher than for the migration of the hydrogen into
the cis position. This may be due to the kinetic trans effect of the amine
donor, quite similar to the effect discussed below.
(36) Inamo, M.; Sumi, T.; Nakagawa, N.; Funahashi, S.; Tanaka,
M. Inorg. Chem. 1989, 28, 2688.
(37) Al-Alousy, A.; Alsheri, S.; Burgess, J .; del Mar Graciani, M.;
Moya, M.-L.; Munoz, E.; Rodriguez, A.; Sanchez, F. Trans. Met. Chem.
1993, 18, 179.
(38) Drljaca, A.; Hubbard, C. D.; van Eldik, R.; Asano, T.; Basilevsky,
M. V.; le Noble, W. J . Chem. Rev. 1998, 98, 2167.

4
162 Organometallics, Vol. 18, No. 20, 1999
Fekl et al.
k (k + k )
the right-hand side of eq 2 is larger than 1, since the
d[2a-Pt-D]
1
2
3
-
) [2a-Pt-D]
(2)
observed rate constant for scrambling is larger than the
dt
k-1 + k + k
2 3
scr
observed rate constant for elimination. From kobs
kobs
/
elim
) 1.6 (at 30 °C), it follows that the ratio k2/k3
If k2 or k3 (or both) . k-1, it follows that kobs ) k1.
must be larger than 1.6. We can offer a plausible reason
for why proton (or deuterium) transfer to the CH3 group
trans to amine (k2) is faster than cis to amine (k3).
Amines are stronger σ-donors than pyridyl residues, as
can be seen for example from basicities. The stronger
the donor trans to a methyl group, the weaker the
platinum-carbon bond will be and the faster proton
transfer could be. This seems to be true in our case, as
However, we observed experimentally that the rate
constant for the disappearance of 2a -P t-D is ca. 3 times
larger than the rate constant for methane elimination,
and deuterium scrambling is almost 2 times faster than
methane elimination. It follows that these observations
are not consistent with this proposal.
We are now left with the conclusion that eq 1
describes the observed rate constant for elimination, but
cannot be simplified further. This steady-state expres-
sion is consistent with both the observed kinetic isotope
effect and the activation volume data. To verify the
former statement, a kH/kD of 3 could be expected for k3
1
95
2
1
suggested by the
Pt- H coupling constant of the
methyl groups. An J Pt-H of 67 Hz for CH3 trans to
amine versus 71 Hz for CH3 trans to pyridyl suggests
that the bond trans to amine is indeed weaker. Since
significant bond weakening is required in the transition
state, this possibly can account for the different proton-
transfer rates and the observed ratio k2/k3.
in eq 1. One can easily show that a “dilution” of the
kinetic isotope effect40 occurs due to the contributions
of k1 and k-1. The proton-transfer step associated with
k3 may, of course, be faster than ring closure (k-1), but
of the same order of magnitude. For example, if k3 )
k-1, the observed kH/kD reduces to 1.15.
Equation 1 also accounts for the low positive activa-
tion volume found. Although the ring-opening step
should be accompanied by a very positive volume of
activation, a pronouncedly negative volume of activation
is expected for the proton transfer which reduces the
coordination number of platinum. The observed volume
Con clu sion a n d Ou tlook
4
The present study provides information on the inti-
mate mechanism of the reactions of the methyl(hydrido)-
+
platinum(IV) complex [Pt(H)(CH3)2(BPMA)] . Kinetic
data for deuterium scrambling into the methyl group
trans to amine and reductive elimination of methane
that involves the methyl group cis to amine are pro-
vided. Several mechanistic studies on deuterium scram-
bling and alkane elimination from somewhat related
#
of activation is given by ∆V obs ) -RT[δ(ln k)/δp].
#
#
#
Applied to eq 1, it follows that ∆V obs ) ∆V (k1) + ∆V -
41
rhodium and iridium complexes have been published.
(
k3) + RT{δ[ln(k-1 + k3)]/δp}, and the apparent activa-
The present work, however, gives information on a
platinum system that adds a new feature: The reac-
tivities of two nonequivalent methyl groups are different
and can be separated kinetically. The study presented
here also includes the first high-pressure kinetic data
on this type of reaction. Although the rate law is a
composite function in the present case, and a complete
deconvolution is not possible, mechanistically valuable
semiquantitative information is obtained. From the
relative values of the rate constants for the steps
involved, a qualitative reaction energy profile can be
derived, as shown in Figure 6. Following microscopic
reversibility arguments, this allows conclusions to be
drawn for the methane activation process (the diagram
must then be read from right to left). Interestingly,
methane activation may be regarded as almost ac-
tion volume will depend on the hydrostatic pressure.
The expected nonlinearity of the ln k versus p plot,
however, is in most cases not seen in the pressure range
accessible. Important to note, however, is that the
apparent activation volume will be much lower than
V (k1). The observed low activation volume, therefore,
#
∆
disproves the assumption that k1 is rate determining,
which again shows that k3 cannot be much faster than
k-1.
It is important to discuss the question why the
scrambling rate is higher than the elimination rate.
Substitution of k3 by k2 in eq 1 results in the corre-
sponding expression for the scrambling rate constant
(
see Scheme 3). The fact that scrambling is faster by a
factor of 1.6 than elimination gives a lower limit for the
ratio k2/k3. Using steady-state treatment and the sim-
plification that deuterium incorporation into the methyl
group is irreversible (which is well founded, see Figure
2
complished when the η -methane complex is formed.
The highest activation energy is required for the sub-
stitution of a strong donor ligand by methane. Following
this step, both the interconversion into the five-
coordinate methyl(hydrido)platinum(IV) complex and
ring closure to give a stable six-coordinate complex are
extremely facile, which is also supported by theoretical
5
), it can be shown that eq 3 applies:
scr
elim
k2
kobs
k1 - k
obs
)
(3)
elim
scr
k3
kobs
k1 - kobs
4
2
studies for the interconversion process.
Recently, this concept received direct support from the
demonstration that the generation of a very good leaving
group using a strong Lewis acid enables the addition of
alkanes to Pt(II) to give stable alkyl(hydrido)platinum-
It must be noted that the second term of the product on
(
39) A treatment of the complete system is possible, but unfortu-
nately not simple. We treated the complete system shown in Scheme
using the computer algebra system MAPLE (Maple V.3, student
3
(
IV) complexes, if the ligand has an “open arm” in the
version, 1981-1994, Waterloo Maple Software and University of
Waterloo) The evaluation of [SS1], [SS2], and [SS1′] in terms of the
rate constants is possible. The calculation of the rates for formation of
intermediate (2a -C-D) and product leads to a system of two linear
differential equations that can be solved. However, the solution is so
complex (a multiterm expression that fills several pages) that no
meaningful conclusion could be drawn.
1
1
reactant complex. We now have the means, at least
theoretically, to control the different steps in methane
(41) Mobley, T. A.; Schade, C.; Bergman, R. G. Organometallics
1998, 17, 3574, and references therein.
(40) Griffiths, D. C.; Young, G. B. Organometallics 1989, 8, 875.
(42) Hill, G. S.; Puddephatt, R. J . Organometallics 1998, 17, 1478.

Methane Elimination from Methyl(hydrido)platinum
Organometallics, Vol. 18, No. 20, 1999 4163
[
P tCl
2 2 2
(SEt ) ] (m ixtu r e of cis a n d tr a n s isom er ). Since
this complex is a widely used precursor for [Me
2
2 2
Pt(µ-SEt ) -
PtMe
effectively directly from PtCl
general, PtCl
2
], we found that it can be prepared conveniently and cost-
instead of using44 [PtCl
]
2-
. In
2
4
II
2
is not suitable for synthesizing Pt complexes
directly because of its strong tendency to disproportionate. This
problem, however, was not encountered in the following
procedure: To PtCl
2
(4 g, 15 mmol) was added diethyl sulfide
(5 mL, 46 mmol) in 120 mL of water. The suspension was
heated under reflux and vigorously stirred for 2 h. The
resulting yellow solution was extracted 6 times with 50 mL of
dichloromethane. Removing the solvent in vacuo left the
complex in the form of yellow crystals. Yield: 6.2 g, 93%.
[
3 3 3 3
P t(BP MA)(F CSO )][F CSO ] (4). A 10 mg sample of 1
was dissolved in 1 mL of dichloromethane. Addition of 1.1
equiv of triflic acid led to the formation of 2a . When an
additional 1 equiv of triflic acid was added, the formation of
gas bubbles (methane) was immediately observed. The reaction
was complete within 10 s. The solvent was reduced to one-
third (in vacuo) and the reaction mixture cooled to 0 °C.
Microcrystalline 4 was obtained in essentially quantitative
F igu r e 6. Qualitative reaction coordinate diagram for the
unimolecular reactions of 2a and its deuterated analogue
2
a -P t-D. Following the formation of the five-coordinate
intermediate SS1, two reaction paths are possible: forma-
2
tion of η -bound methane cis to amine (SS3) leads to
yield. Anal. Calcd for C14
H
13
N
3
PtF
6 2 6
S O : C, 24.3; H, 1.9; N,
2
elimination; formation of η -bound methane trans to amine
1
6
.1; S, 9.3. Found: C, 23.9; H, 2.0; N, 5.9; S, 9.1. H NMR (CD -
2
(SS2) leads to deuterium/proton scrambling.
3
Cl
2
): δ 8.47 (dd, J H-H ) 1.5 Hz, 6.3 Hz, J Pt-H ) 30 Hz, 1 H,
H ortho to N), 8.11 (dt, J H-H ) 1.4 Hz, 8 Hz, 1 H, H para),
.56 (m, 2 H, H meta), 4.7 (m, 2 H, CH ). IR (KBr): ν[SO ]:
1255(s), 1288(s), 1337(m) cm , characteristic for triflate,
activation: The substitution of a ligand by methane
strongly depends on the leaving group, which must
preferably be a weak ligand. The interconversion be-
7
2
3
-
1
consistent with the presence of both coordinated and ionic
2
45,46
tween the η -methane complex and the five-coordinate
triflate.
1
In Situ Ch a r a cter iza tion . The H NMR data of [Pt(CH
3
)
2
-
methyl(hydrido)platinum(IV) complex can be controlled
utilizing the trans labilizing effect of strong donors. This
is suggested by the present study, and clearly much
more can be done in this area. The stabilization of the
Pt(IV) hydrido species depends on the donor strength
and the geometrical requirements of pendant arms or,
alternatively, on the donor strength and concentration
of additional ligands. If a catalytic process is desired,
this stabilization must be balanced and not too strongly,
otherwise a “dead-end” situation will arise.
(BPMA)] (1), cis-[Pt(H)(CH
3
)
2
(BPMA)](BF
4
) (2a ), trans-[Pt(H)-
(
CH (BPMA)] (BF ) (2b), and [Pt(CH
3
)
2
4
3
4
)(BPMA)](BF ) (3) were
1
3
identical with those reported in the literature. Similar
spectra were obtained for the triflate salts.
Gen er a l P r oced u r es a n d High -P r essu r e NMR. The
reactions were performed under inert conditions (dry argon)
using standard glovebag and high-vacuum line techniques. The
glassware was flame dried in vacuo prior to use. CD Cl was
2
2
2
dried using CaH and freshly distilled. For a typical run, a
sufficient amount of 1, typically 2-5 mg, was weighed and
transferred to a WILMAD screw cap NMR tube equipped with
a PTFE-septum, and 1 mL of solvent was added. The ap-
In our opinion the construction of reaction coordinate
diagrams for these important reactions is much aided
by kinetics which include not only the use of isotope
labeling but also the application of high pressure as an
additional kinetic parameter. Currently, we are working
on systems that should reveal the activation volume for
propriate amount of acid (HBF
4
‚OEt
2 2 3
, 85% in OEt , or F -
CSO H, both obtained from Aldrich in the highest purity
3
available) was added using a microliter syringe, and the NMR
spectra were recorded after 10 min (after which the final
temperature was reached and the shimming was completed).
For the high-pressure studies, the DRX 400WB spectrom-
eter was used together with a homemade high-pressure
probe.47 In addition to the published design, a H lock channel
was implemented to stabilize the magnetic field during the
experiment. The magnetic field inhomogenity over the sample
2
processes where elimination from the η -methane com-
plex is rate determining. The ultimate goal would be to
gain insight into the volume changes associated with
alkane activation itself.
2
-9
volume was reduced to 3 × 10 . The high-pressure NMR tube
was sealed with a MACOR plug and an O-ring made of VITON.
Plug and O-ring were shown to be stable in the employed
solvent and acid. For the high-pressure studies on 2b, the
procedure described in the paper was used, with the exception
that the equilibration and the “freezing” of the equilibrium
Exp er im en ta l Section
The NMR spectra were acquired on a BRUKER Avance DRX
4
00WB or DPX 300 spectrometer. Chemical shifts are reported
in ppm relative to TMS. Elemental analyses were performed
on a Carlo Erba elemental analyzer, type 1106 (C, H, N) or
(by excess acid) were done in a dry Schlenk tube. The solution
1
108 (C, H, N, S).
was then transferred to the high-pressure NMR tube. The
spectra were evaluated (Fourier transformed and integrated)
using the WIN-NMR program package. Kinetic data of the
observed first-order reactions were obtained by fitting the data
(standard nonlinear least-squares procedure) to the integrated
3 2
Syn th eses. [P t(CH ) (BP MA)] (1). The title complex was
prepared in water-free form using a modified procedure from
1
3
the literature: Instead of [Me
2
Pt(µ-SMe
2 2 2 2
) PtMe ], [Me Pt(µ-
SEt PtMe ] (which was synthesized in a procedure similar
2
)
2
2
4
3
to that given by Scott and Puddephatt ) was used. When the
reaction was complete (within 1 h), the reaction mixture was
cooled to 0 °C. The complex 1 precipitated as a yellow
microcrystalline powder, which was filtered off under argon
rate law: C ) C
0
exp(-kt).
(44) van Asselt, R.; Rijnberg, E.; Elsevier, C. J . Organometallics
1994, 13, 706.
45) Diver, C.; Lawrance, G. A. J . Chem. Soc., Dalton Trans. 1988,
31.
(
and dried in vacuo. Yield: 75%. Anal. Calcd for C14
19 3
H N Pt:
9
C, 39.6; H, 4.5; N, 9.9. Found: C, 39.7; H, 4.6; N, 9.8.
(
(
46) Lawrance, G. A. Chem. Rev. 1986, 86, 17.
47) Zahl, A.; Neubrand, A.; Aygen, S.; van Eldik, R. Rev. Sci.
(43) Scott, J . D.; Puddephatt, R. J . Organometallics 1983, 2, 1643.
Instrum. 1994, 65, 882.

4
164 Organometallics, Vol. 18, No. 20, 1999
Fekl et al.
Deu ter iu m Scr a m blin g Stu d ies. N-deuterated [Pt(CH
3
)
2
-
Ack n ow led gm en t. The authors gratefully acknowl-
edge financial support from the Deutsche Forschungs-
gemeinschaft, Fonds der Chemischen Industrie, and the
Max-Buchner-Forschungsstiftung. We thank Dr. S.
Schindler (Erlangen) for a gift of the BPMA ligand and
Dr. K. Goldberg (Seattle) for valuable discussions.
(
BPMA)] (1) was prepared using a procedure similar to that
given for 1 above. The BPMA ligand used for this purpose was
deuterated at the nitrogen position by adding 2 mL of CD OD
3
to 0.1 mL of the ligand. Exchange of the amine proton was
complete after 24 h (followed by NMR). After removal of the
solvent, the ligand was used as described above. Acetone-d
Aldrich, highest quality available) and F CSO D (Aldrich, 98
atom % D) were used for the protonation to generate 2a -P t-
D. The NMR tubes were pretreated for several days using CD
6
(
3
3
3
-
Su p p or tin g In for m a tion Ava ila ble: Kinetic time traces
and Eyring plots for kobscis and kobstrans. This material is
available free of charge via the Internet at http://pubs.acs.org.
OD and dried in vacuo. The sum of 2a -P t-D and 2a -C-D
concentration was obtained from the integration of the aro-
matic signal at 7.8 ppm, corresponding to the two para protons
of 2a . Since integration of the hydride signal allowed direct
determination of 2a -C-D, the concentration of 2a -P t-D was
determined from the difference in concentration.
OM9901426