Two generalizable routes to terminal carbido complexes

Article abstract of DOI:10.1021/ja0453735

Desulfurization of the thiocarbonyl ligand in square pyramidal [Ru(CS)Cl₂(PCy₃)₂] (1-S) via sulfur atom abstraction using [Mo(H)(η²-Me₂CNAr)(N[i-Pr]Ar)₂] forms [Ru(C)Cl₂(PCy

Full text of DOI:10.1021/ja0453735

Published on Web 11/11/2005
Two Generalizable Routes to Terminal Carbido Complexes
Stephen R. Caskey, Michael H. Stewart, Jonathon E. Kivela, Joseph R. Sootsman,
Marc J. A. Johnson,* and Jeff W. Kampf
Department of Chemistry, UniVersity of Michigan, 930 North UniVersity AVenue, Ann Arbor, Michigan 48109-1055
Received August 1, 2004; E-mail: mjaj@umich.edu
Scheme 1. New Syntheses and Reactions of 1
Carbido ligands are of great interest due to their participation in
Fischer-Tropsch and related processes for catalytic formation of
hydrocarbons and “oxygenates” from synthesis gas.1-5 Dissociation
of CO on the surface of the heterogeneous late metal catalyst to
form surface carbide and oxide species is thought to be a critical
step in these heterogeneously catalyzed processes,6,7 which represent
a means of converting coal and/or natural gas into gasoline and
other liquid fuels. A bridging carbide complex is formed in the
decomposition of one Grubbs-type olefin metathesis catalyst.8
Nevertheless, well-characterized examples of the simplest potential
homogeneous model compounds, terminal (one-coordinate) carbido
complexes, are very rare.9-13 Herein we report two newly discov-
ered routes (Scheme 1) to recently reported [Ru(C)(PCy3)2Cl2] (1).10
Pursuant to synthetic studies aimed at the preparation of Ru
complexes containing terminal methylidyne and carbido moieties,
we observed that known [Ru(CH-p-C6H4Me)(PCy3)2Cl2] (2)14
reacted cleanly with 1 equiv of vinyl acetate over the course of 90
min in CH2Cl2, affording 1 as a yellow powder in 87% yield upon
removal of volatile materials under vacuum, rinsing with pentane,
and drying in vacuo. Also identifiable in the 1H NMR spectrum of
the reaction mixture were p-methylstyrene and acetic acid. The
“parent” complex [Ru(CHPh)(PCy3)2Cl2] (3)14 reacted analogously.
Use of excess vinyl acetate (20 equiv) resulted in completion within
30 min, but did not alter the products. At shorter reaction times,
resonances consistent with those expected for [Ru(CHOAc)(PCy3)-
dissociation enthalpy (BDE) values are known for its chalcogenido
derivatives [Mo(E)(N[t-Bu]Ar)3] (E ) O, S: 5-O, 5-S).16 The
Mo-O BDE in 5-O places 5 among the most powerful oxygen
atom acceptors reported.17 However, treatment of 1-O and 1-S with
1 or 2 equiv of 5 in THF or C6D6 resulted in no reaction, even
upon heating to 60 °C for several hours, possibly due to the severe
steric demands on a reaction between two such bulky reagents.
Accordingly, the smaller “masked” Mo(III) analogue, [Mo(H)(η2-
Me2CNAr)(N[i-Pr]Ar)2] (6), reported to be in equilibrium with
three-coordinate [Mo(N[i-Pr]Ar)3]18 was employed. In this case, 3
equiv of 6 converted 1-S to 1 smoothly over several hours at 30
°C (Scheme 1), as determined by 31P NMR spectroscopy.
1
Cl2] (4) were observable in the H and 31P NMR spectra of the
reaction mixtures. In contrast, when 2 was exposed to 20 equiv of
vinyl acetate in C6H6, 4 was formed cleanly within 15 min, without
the formation of detectable quantities of 1. Decomposition of 4
then took place slowly over the course of several hours, yielding 1
as the major product. Thus, the reaction proceeds via initial
metathesis to produce 4, followed in CH2Cl2 by clean elimination
of acetic acid in one or more subsequent steps. The good yield of
1 obtained via this method coupled with commercial availability
of vinyl acetate and numerous metathesis-active complexes analo-
gous to 2 renders this route exceptionally attractive for the
preparation of some carbide complexes. At present, a limitation of
this route arises from the formation of acetic acid. Surprisingly,
addition of soluble amine bases is counterproductive in that this
tends to result in the formation of multiple phosphine-containing
products and a reduced yield of 1 (see Supporting Information).
The second new route to 1 is modeled after the CO dissociation
step in the Fischer-Tropsch process. Accordingly, the carbonyl
and thiocarbonyl complexes, [Ru(CO)(PCy3)2Cl2] (1-O) and [Ru-
(CS)(PCy3)2Cl2] (1-S), were investigated as precursors to 1 via
chalcogen atom abstraction.
Also produced in this reaction was brown (µ-S)[Mo(N[i-Pr]Ar)3]2
(7), which was prepared independently via reaction of SPPh3 with
2 equiv of 6. Following reaction of 1-S with 6, pure 1 was isolated
as a yellow powder in 55% yield upon removal of volatile materials
under vacuum and subsequent extraction of 7 from the solid mixture
using pentane. Exclusion of N2 from the reaction mixture is
advantageous in order to avoid formation of purple (µ-N)[Mo(N[i-
Pr]Ar)3]2 (8),18 which has solubility similar to that of 1 in a variety
of solvents and therefore can complicate product separation. In
contrast to 1-S, 1-O remained unaffected under these conditions.
The utility of this atom-abstraction route is underscored in the
synthesis of the new carbido complex 9 from [Ru(CS)(PCy3)2(cat)]
(9-S) in 54% isolated yield. To date, we have been unable to prepare
9 from 1 by ligand exchange. We find that 1 is very inert to ligand
substitution, which limits its utility in preparing other terminal
carbido complexes. Accordingly, we find it necessary to incorporate
the desired ancillary ligand set prior to formation of the carbido
ligand, as in the synthetic scheme 1-S f 9-S f 9 rather than 1 f
9. Thus, desulfurization of 9-S is, at present, the only synthetic
route to 9. As thiocarbonyl complexes are not rare,19,20 desulfur-
ization of a CS ligand is, in principle, a general route to terminal
carbides. At present, its major drawback lies in the stoichiometric
consumption of 6, a moderately synthetically challenging and
sensitive material. However, formation of 9 demonstrates the utility
of this route when the others fail.
Following unsuccessful attempts with several relatively simple
potential atom-abstraction reagents, including PCl3, PCy3, P(OEt)3,
Ge(CH[SiMe3]2)2, 1,3-diisopropylimidazol-2-ylidene, and V(mes)3-
(THF) (mes ) 2,4,6-C6H2Me3), reactions of 1-O and 1-S with [Mo-
(N[t-Bu]Ar)3] (5)15 were examined. Conveniently, Mo-E bond
9
16750
J. AM. CHEM. SOC. 2005, 127, 16750-16751
10.1021/ja0453735 CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
apical site. The most important departures are observed in the Ru-C
bond lengths, which range from 1.632(6) Å in 19 to 1.705(26) Å
in 1-O,25 to 1.738(2) Å in 1-S.
Thus, we have demonstrated that terminal carbido complex 1 is
available via two new routes, from differing starting materials.
Treatment of 2 or 3 with vinyl acetate affords 1 in 87% isolated
yield following olefin metathesis and expulsion of acetic acid;
desulfurization of 1-S using 6 affords 1 in 55% yield. As many
carbonyl and thiocarbonyl complexes similar to 1-O and 1-S can
be obtained straightforwardly from commercial RuCl3‚xH2O, we
are investigating the steric and electronic demands of preparative
desulfurization reactions involving these and other reagents, as well
as the feasibility of related oxygen atom abstraction from carbonyl
complexes, to prepare more reactive terminal carbido complexes.
Studies of the scope and mechanism of the metathesis-facilitated
conversion of 2 to 1 via 4 are also underway.
Figure 1. A 50% thermal ellipsoid plot of [Ru(CS)Cl2(PCy3)2] (1-S).
Selected distances (Å) and angles (°): Ru1-C1, 1.7376(19); Ru1-P1,
2.4211(9); Ru1-P2, 2.4156(9); Ru1-Cl1, 2.3578(7); Ru1-Cl2, 2.3799-
(7); C1-S1, 1.576(2); Ru1-C1-S1, 179.42(14); P1-Ru1-P2, 165.789-
(18); Cl1-Ru1-Cl2, 166.290(19); C1-Ru1-Cl1, 97.87(7); C1-Ru1-Cl2,
95.84(7); C1-Ru1-P1, 96.93(6); C1-Ru1-P2, 97.16(6).
Acknowledgment. This work was supported by an award from
Research Corporation, by the Camille and Henry Dreyfus New
Faculty Awards Program, and by the University of Michigan.
M.H.S. is a fellow of the NSF-sponsored IGERT program for
Molecularly Designed Electronic, Photonic, and Nanostructured
Materials at the University of Michigan.
Complex 1 is rather unreactive and is stable to air and water.
Although its carbido ligand can act as a weak σ-donor to suitable
transition metal fragments, this interaction is accompanied by only
minor lengthening of the RutC bond.9 Reaction of 1 with ethereal
HBF4 yields [Ru(CHPCy3)(PCy3)Cl2]BF4,21 reported to be a rapidly
initiating olefin metathesis catalyst. In related reactions, we find
that 1 acts as a chalcogen atom acceptor in some cases. Although
1-O was not transformed into 1 under the conditions attempted,
the reverse transformation was successful. Pyridine-N-oxide was
unreactive toward 1 under a variety of conditions. Ozone and
iodosylbenzene each produced 1-O in small quantities; the major
product was OPCy3. In contrast, addition at ca. -90 °C of 2 equiv
of dimethyldioxirane in acetone to a pale yellow solution of 1 in
CH2Cl2 afforded known 1-O22,23 as a yellow-orange powder in
64% isolated yield, following its precipitation and extraction from
impurities using acetone.
Having prepared 1 from 1-S, we investigated the reverse
transformation. Accordingly, elemental sulfur converts 1 cleanly
into 1-S in CH2Cl2 within 16 h at 30 °C. The conversion appears
quantitative by 1H and 31P NMR spectroscopy, but at longer times,
free SPCy3 is produced. Analytically pure 1-S is thus obtained in
62% yield following filtration to remove excess sulfur, crystalliza-
tion from the filtered reaction mixture at -35 °C, and recrystalli-
zation from THF. Syntheses of 1-S and 1-O from 1 are depicted in
Scheme 1.
Supporting Information Available: All synthetic procedures;
spectroscopic and analytical data for new compounds; tables of crystal
data, atomic coordinates, structure solution and refinement, bond lengths
and angles, and anisotropic thermal parameters for compound 1-S
(PDF). This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Rofer-DePoorter, C. K. Chem. ReV. 1981, 81, 447.
(2) Snel, R. Catal. ReV. Sci. Eng. 1987, 29, 361.
(3) Vanderlee, G.; Ponec, V. Catal. ReV. Sci. Eng. 1987, 29, 183.
(4) Wojciechowski, B. W. Catal. ReV. Sci. Eng. 1988, 30, 629.
(5) Forzatti, P.; Tronconi, E.; Pasquon, I. Catal. ReV. Sci. Eng. 1991, 33,
109.
(6) Roberts, M. W. Chem. Soc. ReV. 1977, 6, 373.
(7) Sung, S. S.; Hoffmann, R. J. Am. Chem. Soc. 1985, 107, 578.
(8) Hong, S. H.; Day, M. W.; Grubbs, R. H. J. Am. Chem. Soc. 2004, 126,
7414.
(9) Hejl, A.; Trnka, T. M.; Day, M. W.; Grubbs, R. H. Chem. Commun. 2002,
2524.
(10) Carlson, R. G.; Gile, M. A.; Heppert, J. A.; Mason, M. H.; Powell, D. R.;
Vander Velde, D.; Vilain, J. M. J. Am. Chem. Soc. 2002, 124, 1580.
(11) Greco, J. B.; Peters, J. C.; Baker, T. A.; Davis, W. M.; Cummins, C. C.;
Wu, G. J. Am. Chem. Soc. 2001, 123, 5003.
(12) Enriquez, A. E.; White, P. S.; Templeton, J. L. J. Am. Chem. Soc. 2001,
123, 4992.
(13) Peters, J. C.; Odom, A. L.; Cummins, C. C. Chem. Commun. 1997, 1995.
(14) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996, 118,
100.
(15) Laplaza, C. E.; Johnson, M. J. A.; Peters, J. C.; Odom, A. L.; Kim, E.;
Cummins, C. C.; George, G. N.; Pickering, I. J. J. Am. Chem. Soc. 1996,
118, 8623.
(16) Johnson, A. R.; Davis, W. M.; Cummins, C. C.; Serron, S.; Nolan, S. P.;
Musaev, D. G.; Morokuma, K. J. Am. Chem. Soc. 1998, 120, 2071.
(17) Holm, R. H.; Donahue, J. P. Polyhedron 1993, 12, 571.
(18) Tsai, Y. C.; Johnson, M. J. A.; Mindiola, D. J.; Cummins, C. C.; Klooster,
W. T.; Koetzle, T. F. J. Am. Chem. Soc. 1999, 121, 10426.
(19) Butler, I. S. Pure Appl. Chem. 1988, 60, 1241.
(20) Broadhurst, P. V. Polyhedron 1985, 4, 1801.
(21) Romero, P. E.; Piers, W. E.; McDonald, R. Angew. Chem., Int. Ed. 2004,
43, 6161.
(22) Werner, H.; Grunwald, C.; Stuer, W.; Wolf, J. Organometallics 2003,
22, 1558.
(23) Moers, F. G.; Hoedt, R.; Langhout, J. P. J. Organomet. Chem. 1974, 65,
93.
Although 1-S was formed in good yield from 1, we found 1-S
most convenient to prepare in two steps from [RuCl2(PPh3)3] by
treating crude intermediate complex [Ru(CS)(OH2)(PPh3)2Cl2]24
with excess PCy3 without purification.
Dark orange blocks of 1-S form readily from CH2Cl2 at -35
°C; these rapidly revert to a fine yellow-orange powder upon
isolation by vacuum filtration. Nevertheless, an X-ray quality crystal
so obtained, coated immediately with paratone, and placed in a
stream of cold N2 remained intact and undamaged throughout data
collection. The molecule crystallizes in the space group P1h with a
single complex and two CH2Cl2 molecules in the asymmetric unit.
A thermal ellipsoid plot (50% probability level) is shown (Figure
1). Comparison of the X-ray structures of 1, 1-O, and 1-S reveals
very similar square pyramidal cores, with the unique ligand in the
(24) Brothers, P. J.; Roper, W. R. J. Organomet. Chem. 1983, 258, 73.
(25) Moers, F. G.; Beurskens, P. T.; Noordik, J. H. Cryst. Struct. Commun.
1982, 11, 1655-1658.
JA0453735
9
J. AM. CHEM. SOC. VOL. 127, NO. 48, 2005 16751