Hydride elimination from an iridium(III) alkoxide complex: A case in which a vacant cis coordination site is not required

Article abstract of DOI:10.1016/S0022-328X(99)00619-1

Decomposition of trans-HIr(OCH₃)(C₆H₅)(PMe₃)₃ (2) formed by oxidative addition of methanol to Ir(C₆H₅)(PMe₃)₃ (1) was studied in detail. Thermolysis of this complex yields trans-H₂Ir(C₆H₅)(PMe₃)₃ (2) and formaldehyde. Complex 2 is less stable than its two dihydrido isomers, showing that it is the kinetic product of this reaction. The elimination process follows first order kinetics and exhibits a kinetic isotope effect of k_H/k_D=3.2±0.2, the observed activation parameters are ΔH^?_obs=8.3±1.0 kcal mol^-1; ΔS^?_obs=-34±3.5 e.u. and ΔG^?_{obs (298 K)}=18.4±2.0 kcal mol^-1. Catalysis by methanol was observed. The process does not involve a vacant coordination site cis to the coordinated methoxide, as shown by labeling experiments and by the lack of exchange with P(CD₃)_3. Thus, in this case the β-hydride elimination process does not follow the usual pathway. A mechanism, in which following methoxide dissociation, C-H cleavage of free methanol takes place, is suggested.

Full text of DOI:10.1016/S0022-328X(99)00619-1

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 593–594 (2000) 479–484
Hydride elimination from an iridium(III) alkoxide complex: a case
in which a vacant cis coordination site is not required
Ofer Blum, David Milstein *
Department of Organic Chemistry, The Weizmann Institute of Science, Reho6ot 76100, Israel
Received 2 September 1999; accepted 18 October 1999
Dedicated to Professor Fausto Calderazzo on the occasion of his 70th birthday in recognition of his outstanding contributions to
organometallic chemistry.
Abstract
Decomposition of trans-HIr(OCH3)(C6H5)(PMe3)3 (2) formed by oxidative addition of methanol to Ir(C6H5)(PMe3)3 (1) was
studied in detail. Thermolysis of this complex yields trans-H2Ir(C6H5)(PMe3)3 (2) and formaldehyde. Complex 2 is less stable than
its two dihydrido isomers, showing that it is the kinetic product of this reaction. The elimination process follows first order
kinetics and exhibits a kinetic isotope effect of kH/kD=3.290.2, the observed activation parameters are DH‡obs=8.391.0 kcal
mol−1; DS‡obs= −3493.5 e.u. and DGo‡bs (298 K)=18.492.0 kcal mol−1. Catalysis by methanol was observed. The process does
not involve a vacant coordination site cis to the coordinated methoxide, as shown by labeling experiments and by the lack of
exchange with P(CD3)3. Thus, in this case the i-hydride elimination process does not follow the usual pathway. A mechanism, in
which following methoxide dissociation, CꢀH cleavage of free methanol takes place, is suggested. © 2000 Elsevier Science S.A. All
rights reserved.
Keywords: Hydride; Iridium; Elimination; Alkoxide; Mechanism; OꢀH activation
Cl(PMe3)3 [2b]. We now report [6] that decomposition
of saturated phenyl Ir(III) hydrido alkoxide complex,
obtained by oxidative addition of methanol does not
take place by the commonly accepted mechanism. Re-
cently, a binuclear mechanism involving alkoxide com-
plexes was demonstrated [7].
1. Introduction
Late transition metal alkoxides are suggested as in-
termediates in various catalytic reactions, such as car-
balkoxylation of olefins and alkyl halides, hydro-
genation of carbon monoxide, ketones and aldehydes,
dehydrogenation of alcohols, transesterification and hy-
drogen transfer from alcohols to ketones. Decomposi-
tion of late transition metal alkoxides to metal hydrides
and the corresponding carbonyl compounds is probably
the most commonly observed reactivity mode of these
complexes [1]. This process is generally believed to
involve cleavage of a b-CꢀH bond of the coordinated
alkoxide, with concomitant formation of hydride and
h2-aldehyde (or ketone) ligands, requiring an empty cis
coordination site [2–4]. We have shown previously [5]
that water and alcohols oxidatively add to electron rich
Ir(I) complexes and have studied in detail the mecha-
nism of b-H elimination from mer-cis-HIr(OCH3)-
2. Results and discussion
Oxidative addition of methanol to Ir(C6H5)(PMe3)3
(1) [8] in benzene at room temperature leads to quanti-
tative formation of the hydrido alkoxy complex trans-
HIr(OCH3)(C6H5)(PMe3)3 (2) (Eq. (1)). Its 1H-NMR
resonance at 4.12 ppm is typical of protons a to oxygen
in late transition metal alkoxides [9]. Water oxidative
addition to 1 in THF leads to the analogous, spectro-
scopically similar hydrido–hydroxo complex trans-
HIr(OH)(C6H5)(PMe3)3 (3). The hydroxide ligand gives
1
rise to a signal at −3.06 ppm in H-NMR, which is
* Corresponding author.
typical of hydroxo complexes [1]. Complex 3 is con-
0022-328X/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00619-1

480
O. Blum, D. Milstein / Journal of Organometallic Chemistry 593–594 (2000) 479–484
Trans dihydride iridium complexes have been reported
[13].
(1)
Thermolysis kinetics in C6D6 in the presence of
methanol (98.4 mM) followed by 31P{1H}-NMR show
that reaction 1 is first order in 2 from 7 to 37°C (Fig.
1). Interestingly, reaction 1 is catalyzed by methanol
(Fig. 2). A best fit to the data is obtained with an order
Scheme 1. OꢀH oxidative addition to complex 1.
of 2.7 in methanol. The activation parameters, DHo‡bs
=
8.391.0 kcal mol−1; DSo‡bs= −3493.5 e.u. and
DG‡obs (298 K)=18.492.0 kcal mol−1 (Fig. 2), differ
substantially from those obtained for b-hydride elimi-
nation from the analogous chloro complex mer-cis-
verted to 2 by methanol (Scheme 1). The hydride
chemical shifts of 2 (−25.13 ppm) and of 3 (−23.99
ppm) point to its position trans to the methoxide, which
is the only weak s-donor in 2 [10]. Complexes 2 and 3
are uncommon examples of monomeric hydrido–hy-
droxo and –alkoxo complexes obtained by direct oxi-
dative addition [1,5,11].
Complex 2 cleanly decomposes to the dihydride com-
plex trans-H2Ir(C6H5)(PMe3)3 (4) and formaldehyde
(Eq. (1)). The presence of formaldehyde and its
oligomers was verified by the chromotropic acid test
[12]. Complex 4 was unambiguously characterized spec-
troscopically. Trans-dihydride complexes are uncom-
mon due to the large trans effect of the hydride ligand.
HIr(OCH3)Cl(PMe3)3 (5) (DH‡obs=24.1 kcal mol−1
;
DS‡obs=0.6 e.u.; DGo‡bs(298 K)=23.9 kcal mol−1) [2b].
Comparing the decompositions of 2 and of mer-trans-
DIr(OCD3)(C6H5)(PMe3)3 (2a) at 22°C (in C6D6+
CD3OD), yields kobs(2)/kobs(2a)=3.290.2 at 22°C,
implying that the CꢀH bond cleavage is involved in the
rate determining step or precedes it.
Reaction 2 yields only 4a. Integration of its hydrides
(vs. the P(CH3)3 protons) amounted to 5092% of that
expected for the non-deuterated 2 (long relaxation de-
lays of 10 s were given to allow accurate integration),
suggesting that the IrꢀD bond of 2a is not involved in
its decomposition, and that the CꢀH cleavage is irre-
versible (Eq. (2)).
(2)
The mer-trans dihydride complex 4 is undoubtedly
the kinetic product of reaction 1. It is the least stable
isomer of this complex and isomerizes slowly under the
conditions of reaction 1 (or in neat C6D6) to mer-cis-
H2Ir(C6H5)(PMe3)3 (6, Eq. (3)). The third isomer, fac-
H2Ir(C6H5)(PMe3)3 (7) was prepared separately from 1
and hydrogen in benzene and was found stable under
the conditions of reaction 1, even at elevated
temperatures.
Fig. 1. First order plots for reaction 1. [MeOH]=98.4 mM; [1]0=
12.3 mM.
Fig. 2. Dependence of reaction 1 on the methanol concentration in
C6D6 at 22°C. [2]0=12.3 mM.
(3)

O. Blum, D. Milstein / Journal of Organometallic Chemistry 593–594 (2000) 479–484
481
unlikely since it should cause deviation from the ob-
served first order dependence on 2.
While our evidence indicates clearly that reaction 1
does not proceed by the traditional b-H elimination
mechanism, the nature of the operating mechanism is
unclear. However, a mechanism in which following
dissociation of the methoxide (via a contact ion pair
HIr(C6H5)(PMe3)3+CH3O−), CꢀH cleavage of free
methanol takes place (either by oxidative addition gen-
erating a formal Ir(V) intermediate or by deprotonation
of the h2-bound CꢀH bond) [14], is in keeping with our
results (Fig. 3). In this very electron rich system, consid-
erable build-up of electron density on the methoxide
oxygen is expected, promoting inter-molecular hydro-
gen bonding [15] and methoxide dissociation. CꢀH
activation by cationic Ir(III) was reported [16]. The
stereochemical course predicted by this mechanism in-
deed leads to the unstable trans dihydride geometry of
4. The 2.7 order in methanol suggests that the CꢀH
bond cleaved may be of a free methanol molecule, and
that a hydrogen bonding network (based solely on
methanol) is involved for stabilization of the transition
state in our otherwise apolar medium. The negative and
very large activation entropy and the kinetic deuterium
isotope effect may reflect the cleavage of the CꢀH bond
in a multicentered transition state, in addition to contri-
butions from solvent rearrangement. The proposed
mechanism may also explain the different reactivities of
2 and cis-[HIr(OMe)(PMe3)4]PF6 (8) [5] which are both
kinetically stabilized against dissociation of any but the
methoxo ligand. Only the less electron rich 8 is stabi-
lized against thermolysis to the dihydrido species in
methanol–THF. Complex 2 is an unusually electron
rich Ir(III) complex that may be subject to oxidations
uncommon for d6 iridium species (Scheme 2).
(4)
The classic b-hydride elimination mechanism predicts
incorporation of the eliminated b-H into positions cis
to that of the methoxide. As 4 is the kinetic product of
reaction 1, the trans HꢀIrꢀH geometry of 4 provides
clear evidence that the decomposition of 2 does not
follow that pathway. Additionally, the reported b-H
eliminations require a vacant coordination site cis to
the alkoxide prior to the CꢀH cleavage [2–4], unlike
our observations for reaction 1. Thermolysis of 2 in the
presence of a tenfold excess of P(CD3)3 resulted in no
incorporation of this phosphine into the product, indi-
cating that phosphine dissociation does not take place.
The decompositions of the deuterium labeled 2a and
mer-trans-DIr(OCH3)(C6H5)(PMe3)3 (2b) showed no
incorporation of deuterium into the phenyl ring of the
product. Thus, a pentacoordinate intermediate having
an h2-coordinated benzene is not formed during reac-
tion 1.
A binuclear mechanism involving catalysis of reac-
tion 1 by iridium(III) as observed by Bergman [7] is
3. Experimental
3.1. General
Fig. 3. Eyring plot for reaction
1 in C6D6. [2]0=12.3 mM;
[CH3OH]=98.4 mM.
All syntheses and chemical manipulations were car-
ried out under nitrogen in a Vacuum Atmospheres
DC-882 dry box, equipped with an oxygen–water
scrubbing recirculation MO-40 ‘Dri-Train’ or under
argon, using high vacuum and standard Schlenk tech-
niques. The solutions were prepared using standard
dilution techniques to avoid weighing small amounts.
All solvents were refluxed on the proper drying agents,
distilled under argon and stored over activated (150°C
,
,
under vacuum for 12 h) 4 A molecular sieves (3 A for
methanol). All deuterated solvents were purchased from
,
Aldrich, degassed and dried over 3 A molecular sieves
for at least 1 week before use. Trimethylphosphine
(Aldrich), LiCl and LiBr (Merck) were used as received.
Cyclooctene (Merck) was freshly distilled under argon
Scheme 2. A possible mechanism of reaction 1. Participation of
additional methanol molecules (not drawn) in hydrogen bonding
during the reaction is most likely.

482
O. Blum, D. Milstein / Journal of Organometallic Chemistry 593–594 (2000) 479–484
before use. P(CD3)3 was prepared according to a litera-
ture method [17] using CD3I (Aldrich). IrCl3·3H2O was
supplied by Engelhardt.
resulting in rapid bleaching. The solvents were removed
after an hour yielding pure 3.
1
3: H-NMR (C6D6): 7.99 (tt (apparent), Jt=6.1 Hz,
1H-, 31P-, 13C- and 2H-NMR spectra were recorded at
400.19, 161.9, 100.6 and 61.4 MHz, respectively, using a
Bruker AMX 400 spectrometer. Chemical shifts are
reported in ppm downfield from Me4Si (1H, 13C),
(CD3)4Si (2H) and referenced to the residual solvent-h1
(1H) natural abundance-d1 (2H), and all-d-solvent (13C),
or downfield from external H3PO4 85% in D2O (31P).
Spectra were recorded in standard pulsed FT mode
using 90° (or less) pulses and at least 5T1 periods between
pulses to assure reliable quantitative results. When tip
angles smaller than 90° were employed, calculated delay
times were used.
Jt=1.3 Hz, 1Hortho), 7.70 (tt (apparent), Jt=6.4 Hz,
Jt=1.3 Hz, 1Hortho), 7.28 (tt (apparent), Jt=7.4 Hz,
Jt=1 Hz, 1H, Hpara), 7.15 (tt (apparent), Jt=7.1 Hz,
Jt=1.3 Hz, 1H, Hmeta), 7.04 (t (apparent), Jt=7.3 Hz,
2
1H, Hmeta), 1.27 (d, JdHꢀP=7.7 Hz, 9H, P(CH3)3), 1.00
(t, virtualJtHꢀP=3.4 Hz, 18H, 2P(CH3)3), −3.06 (s
(slightly broadened), 1H, IrOH), −23.99 (td, 2JHt ꢀP,cis
=
2
16.7 Hz, JdHꢀP,cis=12.9 Hz, 1H, IrꢀH). 31P{1H}-NMR:
2
2
−40.4 (d, JdPꢀP,cis=21.4 Hz, 2P), −50.2 (t, JtPꢀP,cis
=
21.5 Hz, 1P).
3.4. i-CꢀH elimination from mer-trans-HIr(OCH3)-
(C6H5)(PMe3)3 (2)
3.2. Generation of mer-trans-HIr(OCH3)(C6H5)(PMe3)3
(2)
3.4.1. Kinetic follow up of the decomposition of 2
C6D6 (400 ml) containing (C6H5)Ir(PMe3)3 [7] (1, 3 mg,
6.18×10−3 mmol) were taken from a stock solution,
placed in a 5 mm NMR tube and let to freeze at −30°C.
C6D6 (150 ml) containing methanol (2 ml, 4.94×10−2
mmol) were taken from another stock solution and
placed on top of the frozen solution. The NMR tube was
kept in liquid N2 until it was transferred to a ther-
mostated NMR spectrometer, where the two solutions
were let to mix. Within an hour at room temperature, 1
was completely consumed, 2 was formed and only small
amounts of mer-trans-H2Ir(C6H5)(PMe3)3 (4) were
present. Complex 2 was not isolated as a solid since it
decomposes upon removal of the solvent. It also under-
goes b-CꢀH cleavage in solution even at −30°C yielding
4, thus preventing its separation by low temperature
crystallization. Hence, 2 was always freshly prepared in
situ. Its characterization in solution is unequivocal.
Note: Free methanol in our solution appears at 3.14 ppm
(3H) and 1.41 ppm (1H).
As 2 was always prepared in situ from (C6H5)-
Ir(PMe3)3 (1) and methanol in benzene, and as this
oxidative addition is by an order of magnitude faster
than the subsequent decomposition to yield mer-trans-
H2Ir(C6H5)(PMe3)3 (4), we could study the kinetics of
both reactions on the same reaction mixture. A C6D6
solution of 2 was partitioned among several NMR tubes
such that each tube contained 2 (3 mg, 6.18×10−3
mmol) in 400 ml of C6D6. The NMR tubes were kept
frozen (−30°C) in the drybox. Before the measurement,
150 ml of a C6D6 solution containing 2 ml of methanol
(49.4×10−3 mmol) were added on top of the frozen
solution in the dry box. The tube was kept frozen (liquid
N2) for a few more minutes, then warmed to room
temperature (1.5 min) and placed in the thermostated
NMR probe. The oxidative addition was followed by
31P{1H}-NMR as described above until completion of
the process and subsequently the much slower b-CꢀH
cleavage from the product was studied on the same
reaction mixture.
2: 1H-NMR (400 MHz, in C6D6 with 0.36%
methanol): 8.41 (tt (apparent), Jt=6.4 Hz, Jt=1.3 Hz,
1Hortho), 7.59 (ddt (apparent), Jd=7.2 Hz, Jd=5.9 Hz,
Jt=1.3 Hz, 1Hortho), 7.40 (tt (apparent), Jt=7.4 Hz,
Jt=1.6 Hz, 1H, Hmeta or Hpara), 7.18 (tt (apparent),
Jt=7.2 Hz, Jt=1.3 Hz, 1H, Hmeta or Hpara), 7.07 (tt
(apparent), Jt=7.2 Hz, Jt=1.6 Hz, 1H, Hmeta or Hpara),
4.12 (s, 3H, OCH3), 1.25 (d, 2JHd ꢀP=7.6 Hz, 9H,
P(CH3)3), 1.06 (t, virtualJHt ꢀP=3.4 Hz, 18H, 2P(CH3)3),
This procedure was repeated at 7, 17, 22 and 37°C. In
all experiments, the compounds observed were 2, and 4
(at 37°C small amounts of the consequent isomerization
product mer-cis-H2Ir(C6H5)(PMe3)3 (6) were also seen).
All values were reproducible (twice) with a surprisingly
low inconsistency (less than 5%).
4: 1H-NMR (C6D6 with 0.36% methanol): 8.34 (appar-
ent ddt, 3JdHꢀH=6.6 Hz, 4JHd ꢀP,trans=5.5 Hz, 4JHt ꢀH=1.3
2
2
−25.13 (td, JtHꢀP,cis=17.7 Hz, JHd ꢀP,cis=13.6 Hz, 1H,
Hz (actually two doublets), 2Hortho), 7.12 (tt, 3JHt ꢀH=7.0
IrꢀH). 31P{1H}-NMR: −42.2 (d, JPdꢀP,cis=20.3 Hz,
Hz, JtHꢀH=1.4 Hz, 1Hpara), 7.08 (apparent tt, JdHꢀH
=
2
3
2P), −48.2 (t, JtPꢀP,cis=20.2 Hz, 1P).
7.0 Hz, 4JtHꢀH=1.4 Hz, 2Hmeta), 1.33 (d, 2JdHꢀP=7.5 Hz,
2
9H, P(CH3)3 trans to phenyl), 1.24 (t, virtualJHt ꢀP=3.3 Hz,
2
3.3. Preparation of mer-trans-HIr(OH)(C6H5)(PMe3)3
(3)
18H, 2P(CH3)3 mutually trans), −10.67 (td, JHt ꢀP,cis
=
17.4 Hz, JdHꢀP,cis=16.5 Hz, 2H, HꢀIrꢀH). 31P{1H}-
2
NMR: −47.3 (d, JdPꢀP,cis=22.6 Hz, 2P), −59.3 (t,
2
Water (100 ml) was added to an orange THF (3 ml)
solution of (C6H5)Ir(PMe3)3 (1, 70 mg, 0.144 mmol)
2JtPꢀP,cis=22.6 Hz, 1P). Elemental analysis: Anal. Calc.
C, 36.07%; H, 6.81%: Obs. C, 35.71%; H, 6.49%.

O. Blum, D. Milstein / Journal of Organometallic Chemistry 593–594 (2000) 479–484
483
3.5. Decomposition of mer-trans-DIr(OCD3)(C6H5)-
3.7. Synthesis of fac-H2Ir(C6H5)(PMe3)3 (7)
(PMe3)3 (2a)
Complex 1 (20 mg) was dissolved in 3 ml of benzene
and transferred to a Schlenk tube. The solution was
frozen (liquid N2) and the nitrogen atmosphere was
replaced by ca. 1 atmosphere of H2. The reaction mixture
was left to warm up to room temperature and the excess
pressure (above 1 atmosphere) of H2 was released. An
almost immediate color change (red–yellow) occurred.
The solution was stirred for 30 min after which the
hydrogen was released and the solvent was stripped off
under high vacuum, yielding complex 7 as a white solid.
3.5.1. Kinetic deuterium isotope effect in the
decomposition of 2
The decompositions of
2
and of mer-trans-
DIr(OCD3)(C6H5)(PMe3)3 (2a) were compared at 22°C.
The procedure described above was used here as well.
After repeating these experiments twice (and mathemat-
ically correcting for 1% non-deuterated methanol), the
value of kH/kD obtained was 3.1790.07. This value
includes contributions from both primary and sec-
ondary kinetic deuterium isotope effects, but it is suffi-
ciently large to render the existence of a substantial
primary effect beyond doubt.
1
3
7: H-NMR (C6D6): 8.09 (apparent ddd, JdHꢀH :7.5
4
4
Hz, JdHꢀP,trans:5.8 Hz, JHd ꢀH:1.8 Hz, 2H, 2Hortho),
7.17 (m, (mostly hidden by Hmeta and residual C6HD5
peaks), 1H, 1Hpara), 7.14 (m, (partly hidden by residual
C6HD5+Hpara peaks), 2H, 2Hmeta), 1.31 (d, JdHꢀP=7.9
2
3.5.2. No H/D scrambling
No H/D scrambling is observed during this decom-
position. The H-NMR hydride signals of mer-trans-
Hz, 9H, P(CH3)3 trans to phenyl), 1.23 (d, JdHꢀP=7.2
2
Hz, 18H, 2P(CH3)3 trans to hydride), −10.78 (dm,
2JdHꢀP,trans:140 Hz, 2H, 2IrꢀH trans to phosphine. No
phosphine ligands are mutually trans, as no virtual HꢀP
coupling is observed. It is not clear why a ddd splitting
1
H2Ir(C6H5)(PMe3)3 (4, at −10.67 ppm) and of
mer-trans-HDIr(C6H5)(PMe3)3 (4b, at −10.49) do not
overlap. It was therefore possible to determine that 4 is
not generated upon decomposition of 2a. The integra-
tion ratios of the hydride versus both P(CH3)3 signals
indicates that mer-trans-D2Ir(C6H5)(PMe3)3 (4a) was
not formed as well (as a ratio very close to one hydride
vs. each nine P(CH3)3 protons was found).
(instead of ddt) was observed for Hortho
.
31P{1H}-NMR:
2
2
−56.1 (d, JdPꢀP,cis=15.0 Hz, 2P), −58.3 (t, JtPꢀP,cis
=
14.9 Hz, 1P). Elemental analysis: Anal. Calc. C, 36.07%;
H, 6.81%: Obs. C, 35.86%; H, 6.55%.
3.8. Stability of 7
When this procedure was repeated in non deuterated
benzene with the appropriate CH3OH amounts, only a
single hydride signal was observed in the 2H-NMR
spectrum. The spectrum also shows no deuteride incor-
poration into either the phenyl ring or the P(CH3)3
groups.
Complex 7 (10 mg) was dissolved in a solution
identical to that in which the decomposition of 2 to 4
was studied i.e. 550 ml C6D6 containing 2 ml methanol.
No isomerization of 7 to either 4 or 6 was observed
neither after 1 week at room temperature, nor after 6 h
at 60°C.
3.6. Isomerization of mer-trans-H2Ir(C6H5)(PMe3)3 (4)
to mer-cis-H2Ir(C6H5)(PMe3)3 (6)
3.9. Decomposition of 2 in the presence of excess
P(CD3)3
The reaction mixture in which the decomposition of
2 to 4 was studied, was left for an additional 48 h at
room temperature. While only 4 was observed at the
beginning, 6 was the only complex observed at the end.
Methanol (15 ml, 0.370 mmol) was added to a C6D6
(530 ml) solution of 1 (6 mg, 1.24×10−2 mmol) in a
NMR tube at room temperature. When all of 1 was
consumed but only less than 5% of 2 decomposed to 4,
P(CD3)3 (12.8 ml, 0.124 mmol) was added by a syringe.
1
6: H-NMR (C6D6 with 0.36% methanol): 7.99 (dm,
3JdHꢀH:6.6 Hz, 2H, 2Hortho), 7.20 (m, 1H, 1Hpara), 7.18
1
31P{1H}- and H-NMR spectra were recorded periodi-
3
(ddd (apparent dt), JdHꢀHortho=3JdHꢀHpara=5.3 Hz,
cally. No release of P(CH3)3 into the solution was
observed by 31P{1H}-NMR (the 31P{1H} signal of
P(CH3)3 appears at −62 ppm, and that of P(CD3)3 at
2
4JdHꢀHmeta=1.6 Hz, 2H, 2Hmeta), 1.35 (dd JHd ꢀP=7.1
4
Hz, JdHꢀH,trans=0.7 Hz, 9H, P(CH3)3 trans to hydride),
1.31 (t, virtualJHt ꢀP=3.3 Hz, 18H, 2P(CH3)3 mutually
1
−64 ppm.). Integration of the H-NMR P(CH3)3 sig-
2
2
trans), −11.61 (dtdd, JdHꢀP,trans=134.7 Hz, JHt ꢀP,cis
=
nals versus the hydrides and the aromatic peaks shows
no P(CD3)3 incorporation into 4.
22.6 Hz, JdHꢀH,cis=5.0 Hz, JdHꢀH,trans=0.7 Hz, 1H,
2
4
IrꢀH trans to phosphine), −14.21 (tdd, JtHꢀP,cis=18.0
2
Hz, JdHꢀP,cis=17.7 Hz, JHd ꢀH,cis=5.0 Hz, 1H, IrꢀH
3.10. Examination of a solution of mer-trans-HIr(OH)-
(C6H5)(PMe3)3 (3) and P(CH3)3 by saturation transfer
2
2
trans to phenyl). 31P{1H}-NMR: −47.3 (d, JdPꢀP,cis
=
2
22.3 Hz, 2P), −56.2 (t, 2JPt ꢀP,cis=22.2 Hz, 1P). Elemen-
tal analysis: Anal. Calc. C, 36.07%; H, 6.81%: Obs. C,
35.77%; H, 6.52%.
A solution of 3 (26 mg, 5.16×10−2 mmol), P(CH3)3
(5.4 ml, 5.2×10−2 mmol) and water (1 ml, 5.55×10−2

484
O. Blum, D. Milstein / Journal of Organometallic Chemistry 593–594 (2000) 479–484
Blum, T. Koetzle, J. Chem. Soc. Dalton Trans. (1990) 142. (c) O.
Blum, D. Milstein, Angew. Chem. Int. Ed. Engl. 32 (1995) 229.
[6] A preliminary account of this work was presented: Proceedings
of the Israeli Chemical Society’s 60th Annual Meeting, 1995,
Abstract p. 134.
mmol) in C6D6 was prepared in an NMR tube, and
placed at 22°C in the 500 MHz NMR magnet. Irradiat-
ing each time only one of the 31P{1H} signals using a
DANTE pulse sequence revealed that there is no ex-
change between the free phosphine and any of the
coordinated phosphines. Repeating the experiment with
various parameters for the DANTE sequence and at
different temperatures (between 22 and 60°C) led to the
same conclusion.
[7] J.C.M. Ritter, R.G. Bergman, J. Am. Chem. Soc. 120 (1998)
6826.
[8] M. Aizenberg, D. Milstein, J. Am. Chem. Soc. 117 (1995) 6456.
[9] See Refs. [2b,5] and: (a) R.D. Simpson, R.G. Bergman,
Organometallics 12 (1993) 781. (b) D.J. Cole-Hamilton, T.A.
Stephenson, J. Chem. Soc. Dalton Trans. (1976) 2396. (c) D.S.
Glueck, L.J. Newman-Winslow, R.G. Bergman, Organometallics
10 (1991) 1462.
3.11. Generation of 2 from 3 and methanol
[10] J.P. Jesson, in: E.L. Muetterties (Ed.), Transition Metal Hy-
drides, Marcel Dekker, New York, 1971, pp. 75–189.
[11] (a) R.S. Paonessa, A.L. Prignano, W.C. Trogler, Organometal-
lics 4 (1985) 647. (b) N.S. Akl, H.A. Tayim, J. Organomet.
Chem. 297 (1985) 371. (c) M.B. Sponsler, B.H. Weiller, P.O.
Stoutland, R.G. Bergman, J. Am. Chem. Soc. 111 (1989) 6841.
(d) H. Werner, A. Michenfelder, M. Schulz, Angew. Chem. Int.
Ed. Engl. 30 (1991) 596. (e) R.D. Gillard, B.T. Heaton, D.H.
Vaughan, J. Chem. Soc. A (1970) 3126. (f) J. Gotzig, R. Werner,
H. Werner, J. Organomet. Chem. 290 (1985) 99. (g) M.J. Burn,
M.G. Fickes, J.F. Hartwig, F.J. Hollander, R.G. Bergman, J.
Am. Chem. Soc. 115 (1993) 5875. (h) C. Di Bugno, P. Leoni, M.
Pasquali, Gaz. Chim. Ital. 118 (1988) 861. (i) T. Yoshida, T.
Matsuda, T. Okano, T. Kitani, S. Otsuka, J. Am. Chem. Soc.
101 (1979) 2027. (j) T. Yoshida, T. Okano, Y. Ueda, S. Otsuka,
ibid. 103 (1981) 3411.
[12] F. Feigel, Spot Tests in Organic Synthesis, fifth ed., Elsevier,
New York, 1956, p. 331.
[13] (a) J.M. Brown, F.M. Dayrit, D. Lightowler, Chem. Comm.
(1983) 414. (b) B. Rybtchinski, Y. Ben-David, D. Milstein,
Organometallics 16 (1997) 3786.
[14] The thermodynamic and kinetic factors influencing the equi-
librium between (h2-CꢀH)Pt(II) and alkyl hydride Pt(IV) com-
plexes were discussed: S.S. Stahl, J.A. Labinger, J.E. Bercaw, J.
Am. Chem. Soc. 118 (1996) 5961. (b) ibid., Inorg. Chem. 37
(1998) 2422.
Complex 3 (12 mg, 2.32×10−2 mmol) was dissolved
in a solution containing toluene-d8 (540 ml) and
methanol (10 ml). The gradual generation of 2 and the
disappearance of 3 were observed by 31P{1H}-NMR.
Acknowledgements
The research was supported by the Israel Science
Foundation, Jerusalem, Israel and by the MINERVA
Foundation, Munich, Germany. We thank Y. Ben-
David for the preparation of P(CD3)3. D. Milstein is
the holder of the Israel Matz Professorial Chair of
Organic Chemistry.
References
[1] H.E. Bryndza, W. Tam, Chem. Rev. 88 (1988) 1163.
[2] (a) H.E. Bryndza, J.C. Calabrese, M. Marsi, D.C Roe, W. Tam,
J.E. Bercaw, J. Am. Chem. Soc. 108 (1986) 4805. (b) O. Blum,
D. Milstein, J. Am. Chem. Soc. 117 (1995) 4582. (c) H.E.
Bryndza, S.A. Kretchmer, T.H. Tulip, J. Chem. Soc. Chem.
Commun. (1985) 977. (d) K.A. Bernard, W.M. Rees, J.D. At-
wood, Organometallics 5 (1986) 390. (e) D.M. Hoffman, D.
Lappas, D.A. Wierda, J. Am. Chem. Soc. 111 (1989) 1531.
[3] H. Itagaki, N. Koga, K. Morokuma, Y. Saito, Organometallics
12 (1993) 1648.
[15] R.G. Bergman, Polyhedron 14 (1995) 3227 and references
therein.
[16] (a) H.F. Luecke, B.A. Arndtsen, P. Burger, R.G. Bergman, J.
Am. Chem. Soc. 118 (1996) 251. (b) B.A. Arndtsen, R.G.
Bergman, Science 270 (1995) 2517. (c) A theoretical study: D.L.
Strout, S. Zaric, S. Niu, M.B. Hall, J. Am. Chem. Soc. 118
(1996) 6086. (d) M.D. Su, S.Y. Chu, J. Am. Chem. Soc. 119
(1997) 5373.
[4] A.J. Gellman, Q. Dai, J. Am. Chem. Soc. 115 (1993) 714.
[5] (a) D. Milstein, J.C. Calabrese, I.D. Williams, J. Am. Chem.
Soc. 108 (1986) 6387. (b) R.C. Stevens, R. Bau, D. Milstein, O.
[17] M.L. Luetkens Jr., A.P. Sattelberger, H.H. Murray, J.D. Basil,
J.P. Fackler Jr., Inorg. Synth. 28 (1990) 305.
.