Dehydrohalogenation as a source of OsHnCl(PPh3)3 (n = 1, 3)

Full text of DOI:10.1021/ic981402z

4168
Inorg. Chem. 1999, 38, 4168-4170
Experimental Section
Dehydrohalogenation as a Source of
OsHnCl(PPh3)3 (n ) 1, 3)
General. All reactions and manipulations were conducted using
standard Schlenk and glovebox techniques. Solvents were dried and
distilled under nitrogen or argon, and stored in airtight solvent bulbs
with Teflon closures. OsHCl(PPh3)3 does not react with N2 (1 atm, 25
°C). All NMR solvents were dried, vacuum-transferred, and stored in
a glovebox under an argon atmosphere. Styrene (Aldrich) and NEt3
(Aldrich) were degassed with freeze-pump-thaw cycles. Styrene was
stored in an airtight bulb at -3 °C. (NH4)2[OsCl6] (Aesar) was used as
received with no further purification. OsCl2(PPh3)3 was synthesized
according to published procedures.12
German Ferrando and Kenneth G. Caulton*
Department of Chemistry, Indiana University,
Bloomington, Indiana 47405-4001
ReceiVed December 8, 1998
Synthesis of [OsHnCl(PPh3)3] (n ) 1, 3). A 0.100 g amount of
OsCl2(PPh3)3 (0.095 mmol) was dissolved in 20 mL of benzene, and
NEt3 (0.143 mmol, 20 µL) was added to the solution. The solution
was frozen, and the flask was evacuated and then filled with H2 (600
mmHg, 0.969 mmol). The reaction mixture was allowed to warm to
room temperature and stirred vigorously for 1.5 h, giving a pale green
solution. At this point, the only Os-containing species (as determined
by NMR-scale studies in C6D6) is [OsH3Cl(PPh3)3], characterized by a
Introduction
The compound RuHCl(PPh3)3 is one of increasing and
versatile utility in metal-mediated organic synthesis; initial
screening for some metal-promoted organic transformation often
includes RuHCl(PPh3)3.1-6 It has the necessary (i.e., reactive)
hydride ligand, it is unsaturated, and it has the bulky PPh3 ligand,
in the event that a leaving group is required during catalysis. It
would therefore be desirable to have access to the osmium
analogue, OsHCl(PPh3)3, to obtain different selectivity or to
detect intermediates which are too transient for the ruthenium
analogue. This compound was reported7 in 1974, together with
its catalytic activity for olefin hydrogenation at 50 °C. However,
this work dates from before the extensive study of polyhydride
complexes8 and the knowledge of H2 as a ligand,9 and thus its
characterization is incomplete by contemporary standards. In
particular, we were attracted by the fact that the hydride
chemical shift for the reported OsHCl(PPh3)3, -9.3 ppm,
differed unexpectedly from that10 of RuHCl(PPh3)3, -17.7 ppm.
Moreover, there is an empirical correlation that the chemical
shift of a hydride trans to an empty site in a five-coordinate
species has a very high field chemical shift, certainly more
negative than -15 ppm and sometimes as high as -35 ppm.
For comparison,11 see the chemical shifts in I. We report here
a study of these points, together with a reliable synthesis of
authentic OsHCl(PPh3)3 and a correct identification of the
material reported earlier.7
1
broad signal in H NMR at -9.4 ppm and a 31P NMR signal at 3.4
ppm in C6D6. The reported 31P{1H} NMR signal was located at +8.0
ppm referenced to free PPh3 in C6H6, which is at 3.4 ppm when
referenced to external H3PO4. Conversion to [OsHCl(PPh3)3] can be
accomplished by three methods.
Method A. After 1.5 h of stirring under H2, the light green/brown
suspension is evacuated to dryness, followed by addition of fresh solvent
and evacuation, a total of four successive cycles, leading to a brown
product which is [OsHCl(PPh3)3].
Method B. After 1.5 h of stirring, the H2 is removed in a vacuum
(briefly) and the suspension is treated with styrene (0.096 mmol, 11
µL). This dehydrogenation can be performed at room temperature in
24 h or at 60 °C in 1 h. The dark brown solution is evacuated to dryness
to give a brown solid.
Method C. The resulting OsH3Cl(PPh3)3 solution formed in C6D6
was frozen, the excess H2 was evacuated, and excess C2H4 was added.
Within 10 min, OsHCl(C2H4)2(PPh3)2 formed. The 1H NMR spectrum
showed a triplet at -1.9 ppm (2JH-P ) 22 Hz) and 31P{1H} NMR gives
a sharp peak at 0.63 ppm. When this brown solution is evacuated to
dryness, the coordinated C2H4 is released and PPh3 existing in solution
binds again, leading to OsHCl(PPh3)3. The following NMR data are
diagnostic; aromatic 1H NMR resonances were also seen (as follows).
OsHCl(C2H4)2(PPh3)2: 1H NMR (20 °C, C7D8) δ -1.9 ppm (Os-
H, t, JHP ) 22 Hz), δ 3.4 ppm (C2H4). 31P{1H} NMR (20 °C, C7D8) δ
0.63 ppm.
1
OsH3Cl(PPh3)3: H NMR (-80 °C, toluene-d8): -11.7 ppm (br)
and -4.4 ppm (br) in a 2:1 ratio. 31P{1H} NMR (-80 °C, toluene-d8):
2
δ -0.243 ppm (br), 7.59 ppm (d, JP-P ) 14.6 Hz) in a 1:2 intensity
ratio.
1
OsHCl(PPh3)3: H NMR (-70 °C, toluene-d8) δ -24.26 ppm (dt,
2JH-P ) 30 Hz, JH-P′ ) 18 Hz). 31P{1H} NMR (-50 °C, toluene-d8)
2
2
* Corresponding author. E-mail: caulton@indiana.edu.
(1) Mizushima, E.; Yamaguchi, M.; Yamagishi, T. Chem. Lett. 1997, 3,
237.
δ 26.03 ppm (d, JP-P ) 11 Hz), δ 23.85 ppm (br) in a 2:1 intensity
ratio.
(2) Alibrandi, G.; Mann, B. E. J. Chem. Soc., Dalton Trans. 1994, 7,
951.
Results
(3) Grushin, W.; Alper, H. J. Organomet. Chem. 1991, 56, 5159.
(4) Kando, T.; Kotachi, S.; Tsuji, Y.; Watanabe, Y.; Mitsudo, T.
Organometallics 1997, 16, 2562.
Synthesis and Reactivity. Repetition of the literature syn-
thesis (eq 1) produced the previously reported material in 1 h
(5) Tsuji, Y.; Kotashi, S.; Huh, K. T.; Watanabe, Y. J. Organomet. Chem.
1990, 55, 580.
(6) Hiraki, K.; Ochi, N.; Sasada, Y.; Hayashida, H.; Fuchita, Y.;
Yamanaka, S. J. Chem. Soc., Dalton Trans. 1985, 873.
(7) Oudeman, A.; Van Rantwijk, F.; Van Bekkum, H. J. Coord. Chem.
1974, 4, 1.
(8) Morris, R. H.; Jessop, P. G. Coord. Chem. ReV. 1992, 121, 155.
(9) Kubas, G. J. Acc. Chem. Res. 1988, 21, 120.
(10) Hallman, R. S.; McGarvey, B. R.; Wilkinson, G. J. Chem. Soc. (A)
1968, 3143.
C6H6
OsCl2(PPh3)3 + H2 + NEt3
8
(1)
at 25 °C. We have investigated the influence of the NEt3:Os
ratio and found that a larger amount of NEt3 replaces a second
chloride, leading to the formation of OsH4(PPh3)3.13 The
optimum ratio to avoid formation of the tetrahydride species
(11) Heyn, R. H.; Macgregor, S. A.; Nadasdi, T. T.; Ogasawara, M.;
Eisenstein, O.; Caulton, K. D. Inorg. Chim. Acta 1997, 259, 5.
(12) Hoffmann, P. R.; Caulton, K. G. J. Am. Chem. Soc. 1974, 97, 4221.
10.1021/ic981402z CCC: $18.00 © 1999 American Chemical Society
Published on Web 08/14/1999

Notes
Inorganic Chemistry, Vol. 38, No. 18, 1999 4169
without slowing the synthesis of the desired OsHnCl(PPh3)3 at
room temperature in benzene was 1.5 equiv of NEt3 for a period
of 1.5 h.
Characterization of OsHnCl(PPh3)3 (n ) 1, 3). The room-
temperature NMR spectra reported above are too simple for any
reasonable structures for these molecules. For example, the
structure14 of RuCl2(PPh3)3 shows a mirror-symmetric, square-
pyramidal coordination sphere with two equivalent P, but one
unique P. We therefore investigated the variable-temperature
NMR spectra of these two osmium hydrides.
The hydride 1H NMR signal of the product is broad, lacking
the reported resolved multiplet structure (i.e., JPH ) 3 Hz),7
and the 31P{1H} spectrum shows only one broad signal at 3.4
ppm, all at 20 °C in benzene.
When a benzene solution of this compound is subject to
vacuum removal of solvent, the NMR spectra of the resulting
material in C6D6 at 20 °C shows an additional hydride and 31P-
{1H} signal. Noteworthy is the fact that the new hydride signal
is significantly upfield (-23.9 ppm) of that of the original
product (-9.4 ppm). We postulated that this was evidence for
the operation of the equilibrium in eq 2, which vacuum-shifts
OsHCl(PPh3)3. The crystal structure of the ruthenium
analogue of this compound is a useful reference point to
understand the expected spectra for OsHCl(PPh3)3. The structure
of RuHCl(PPh3)3 is that of a square-based pyramid (analogous
to that of RuCl2(PPh3)3), distorted by phosphine bending toward
the small hydride ligand (II). Important distortions away from
OsH3Cl(PPh3)3 h OsHCl(PPh3)3 + H2
(2)
to the right, by H2 removal. The previously reported product
was thus OsH3Cl(PPh3)3, as a consequence of the synthesis being
carried out under excess H2. Indeed, if a benzene solution
containing both species is saturated with H2, NMR spectra
recorded within 10 min of H2 addition show complete disap-
pearance of the signals attributed to OsHCl(PPh3)3. The change
is immediate as judged by the solution color change from brown
to pale green, together with dissolution of the less soluble
OsHCl(PPh3)3, to form the more soluble OsH3Cl(PPh3)3.
Subjecting the primary synthetic product, OsH3Cl(PPh3)3, to four
benzene solvent removal cycles (i.e., strip solution to dryness,
dissolve resultant solid in fresh benzene, strip, etc.) effects
essentially complete disappearance of OsH3Cl(PPh3)3. Attempts
to remove H2 by hydrogenating styrene at 25 °C show complete
conversion to OsHCl(PPh3)3 and PhEt within 24 h. This
conversion is conveniently carried out with equimolar styrene
at 60 °C in 1 h.
3-fold symmetry of a trigonal bipyramid with H and Cl axial
are shown by (1) the angle P-Ru-P (153°), much larger than
120° (i.e., trigonal bypramid) and (2) the “apical” phosphorus,
*P, being closer to Ru by 0.14 Å.15 Below 0 °C in toluene-d8,
the 31P{1H} NMR spectrum of OsHCl(PPh3)3 begins to decoa-
lesce, and below -30 °C, two resonances are seen in a 2:1
intensity ratio at chemical shifts which average to the single
line seen at 20 °C. At -50 °C, multiplet structure can be
resolved (JPP′ ) 11 Hz), which is consistent with mirror
symmetry (AM2 31P{1H} spin system). The hydride spectrum
is consistent with these observations, since it is a doublet of
triplets below -30 °C (AM2H spin system with JHP ) 30 Hz,
JH-P′ ) 18 Hz), but then transforms, in the -20 to +20 °C
region, to a quartet (JHP ) 22 Hz). On warming to 70 °C, the
hydride resonance broadens, indicative of a second, distinct
dynamic process (see below).
When OsH3Cl(PPh3)3 is treated with excess ethylene at 25
°C in toluene-d8 or CD2Cl2, the reagent osmium complex
disappears within 10 min. This reaction is thus much faster than
that with equimolar styrene. However, the product (ethane is
also detected) is not merely OsHCl(PPh3)3, but rather it is that
shown in eq 3. This addition and phosphine displacement is
OsH3Cl(PPh3)3. This molecule shows a 2:1 31P{1H} NMR
pattern in the range -90 to -40 °C in toluene-d8. At -55 °C
and below, multiplet structure can be resolved (JPP ) 14.6 Hz).
In the range -40 to 0 °C, these coalesce to a single signal,
which sharpens more at 20 °C and still further by +70 °C. The
hydride signal in the range -30 to +70 °C is only one line,
somewhat sharper at the higher temperatures, but never sharp
enough to show resolved coupling to phosphorus. Between -55
and -50 °C, the hydride signal is too broad to detect, but
decoalescence occurs below -55 °C, showing two signals in a
2:1 ratio at -11.7 and -4.4 ppm with no resolved structure.
Is this molecule a classical trihydride, or a dihydrogen
complex (i.e., OsH(H2))? Measurement of T1 of the coalesced
signal in C7D8 at 300 MHz shows a value of 26 ms at -40 °C,
which is short enough to indicate that the complex exists as
OsHCl(H2)(PPh3)3, an octahedral d6 Os(II) complex. Poor
solubility prevents lower temperature measurements, and thus
obtaining a minimum T1 value for this molecule, or for the even
less soluble OsHCl(PPh3)3. The line width of the decoalesced
H on Os is too broad (135 Hz at -80 °C in CD2Cl2) to resolve
OsHCl(PPh3)3 + nC2H4 h
OsHCl(C2H4)n(PPh3)2 + PPh3 (3)
most evident by the hydride triplet in the product, and the
production of free PPh3, as seen in the 31P{1H} NMR spectrum.
This reaction is reversible, and vacuum removal of volatiles
also removes coordinated ethylene, to produce OsHCl(PPh3)3.
1
The H NMR spectrum at 25 °C in toluene shows separate
resonances for free (5.2 ppm) and coordinated (3.4 ppm)
ethylene, although the coordinated ethylene is broad. The 25
°C 31P{1H} NMR spectrum in toluene shows separate sharp
lines for OsHCl(C2H4)n(PPh3)2 and for added OsHCl(PPh3)3,
separated by about 23 ppm. The hydride resonance is a triplet,
but integration of this signal against that of coordinated ethylene,
to determine n, the number of coordinated ethylenes, is not
reliable since there is evidence of deuteration of free and
coordinated ethylene by solvent (toluene-d8). However, the
hydride chemical shift is so far downfield (-1.9 ppm) that it
indicates the presence of a ligand trans to hydride; the formula
cis,trans-OsHCl(C2H4)2(PPh3)2 is consistent with all observa-
tions.
1
1
any J(H-D) in the H NMR spectrum of OsHD2Cl(PPh3)3,
made from D2 and OsHCl(PPh3)3. It is noteworthy that the
chemical shift of coordinated H2 is upfield of that of the hydride
ligand.
The above spectra were recorded on a solution containing
both complexes, and thus in equilibrium, together with some
(undetectable) quantity of H2. The mole fraction of OsHCl-
(13) Levison, J. J.; Robinson, D.; Uttley, H. F. J. Chem. Soc., Dalton Trans.
1971, 93, 3068.
(14) LaPlaca, S. J.; Ibers, J. A. Inorg. Chem. 1965, 4, 778.
(15) Skapski, A. C.; Troughton, P. G. H. Chem. Commun. 1968, 1230.

4170 Inorganic Chemistry, Vol. 38, No. 18, 1999
Notes
(PPh3)3 increases with increasing temperature, consistent with
1
a positive ∆S° for eq 2. Since, even at +70 °C, separate H
and 31P NMR signals are seen for the two species, their rate of
equilibration is slower than ∼1 s-1. However, the second
dynamic process detected by 1H NMR for OsHCl(PPh3)3 at +70
°C could be attributed to (1) the rate of its collision with H2 to
form OsH3Cl(PPh3)3 or (2) dissociation of PPh3, to form the
14-electron transient OsHCl(PPh3)2. To investigate this further,
we have studied the line width of the 31P{1H} NMR signal of
free PPh3 added to OsHCl(PPh3)3. At +70 °C, we find no
detectable broadening of free PPh3, which then favors the first
suggestion, above, for the origin of the broadening of the hydride
signal.
It must be reiterated, however, that any 1H or 31P site
exchange between OsHCl(PPh3)3 and OsH3Cl(PPh3)3 is too slow
to coalesce their NMR signals. Moreover, the line width of the
hydride signal of OsHCl(PPh3)3 is increased by any traces of
added H2, which can also be provided, via eq 2, by any OsH3-
Cl(PPh3)3 present; thus, the hydride signal of OsHCl(PPh3)3 can
sometimes be broadened to the point of being very difficult to
detect at +20 °C.
whether H2 is trans to Cl or to L (PPh3). These two closely
related species are now better characterized for organic synthetic
applications, which may be further influenced by the tendency
of any tris-triphenylphosphine complex to lose one PPh3 (cf.
eq 3).
Acknowledgment. This work was supported by the National
Conclusions
Science Foundation.
All of the observations reported here are consistent with eq
4, which emphasizes that we cannot establish with certainty
IC981402Z