Theoretical and Synthetic Studies on Dihaptoacyl and β-Agostic Acyl Complexes of Molybdenum

Article abstract of DOI:10.1021/om9807248

Molybdenum acyl complexes of formula Mo(C(O)CH₂SiMe₂R)(S₂CX)(CO)(PMe ₃)₂ (R = Me, Ph; X = NMe₂ (2-Me, 2-Ph), N-i-Pr₂ (3-Me), NC₄H₄ (4-Me), O-i-Pr (5-Me, 5-Ph), O-t-Bu (6-Me)) containing several S-donor ligands have been prepared and characterized. Compounds 2-Me and 2-Ph exist in solution as equilibrium mixtures of the agostic and the dihaptoacyl species and crystallize, respectively, as the isomeric mixture (2-Me) and as the agostic compound (2-Ph). Complexes 3-6 show dihapto coordination both in solution and in the solid state. Confirmation of the agostic coordination of the Mo-acyl moiety has been provided by an X-ray diffraction study of Mo(C(O)CH₂SiMe₃)(S₂CNMe₂)CO(PMe ₃)₂ and by ab initio calculations performed with the representative model species Mo(C(O)CH₂R)(S₂-CNH₂)(CO)(PH₃) ₂ (R = H (8), SiH₃ (9)), which show good agreement with the structural data of the parent compounds. For 8 a value of 12.7 kcal mol^-1 has been obtained for the agostic stabilization. Calculation of the energy profile for the CO deinsertion in 8 gives an energy barrier of 4.0 kcal mol^-1 for the C_α-C_β breaking process, in good accordance with the available experimental data.

Full text of DOI:10.1021/om9807248

3294
Organometallics 1999, 18, 3294-3305
Th eor etica l a n d Syn th etic Stu d ies on Dih a p toa cyl a n d
â-Agostic Acyl Com p lexes of Molybd en u m
Gregori Ujaque, Feliu Maseras, and Agust´ı Lledo´s*
Departament de Quı´mica, Universitat Auto`noma de Barcelona,
08193 Bellaterra, Barcelona, Spain
Leopoldo Contreras, Antonio Pizzano, Dieter Rodewald, Luis Sa´nchez,* and
Ernesto Carmona*
Instituto de Investigaciones Quı´micas-Departamento de Quı´mica Inorga´nica,
CSIC-Universidad de Sevilla, 41092 Sevilla, Spain
Angeles Monge and Caridad Ruiz
Instituto de Ciencia de Materiales de Madrid, CSIC, Cantoblanco, E-28049 Madrid, Spain
Received August 26, 1998
Molybdenum acyl complexes of formula Mo(C(O)CH2SiMe2R)(S2CX)(CO)(PMe3)2 (R ) Me,
Ph; X ) NMe2 (2-Me, 2-P h ), N-i-Pr2 (3-Me), NC4H4 (4-Me), O-i-Pr (5-Me, 5-P h ), O-t-Bu
(6-Me)) containing several S-donor ligands have been prepared and characterized. Com-
pounds 2-Me and 2-P h exist in solution as equilibrium mixtures of the agostic and the
dihaptoacyl species and crystallize, respectively, as the isomeric mixture (2-Me) and as the
agostic compound (2-P h ). Complexes 3-6 show dihapto coordination both in solution and
in the solid state. Confirmation of the agostic coordination of the Mo-acyl moiety has been
provided by an X-ray diffraction study of Mo(C(O)CH2SiMe3)(S2CNMe2)CO(PMe3)2 and by
ab initio calculations performed with the representative model species Mo(C(O)CH2R)(S2-
CNH2)(CO)(PH3)2 (R ) H (8), SiH3 (9)), which show good agreement with the structural
data of the parent compounds. For 8 a value of 12.7 kcal mol-1 has been obtained for the
agostic stabilization. Calculation of the energy profile for the CO deinsertion in 8 gives an
energy barrier of 4.0 kcal mol-1 for the CR-Câ breaking process, in good accordance with the
available experimental data.
In tr od u ction
The insertion of carbon monoxide into a metal-alkyl
bond is a fundamental organometallic reaction.1 The
seemingly unrelated nonclassical M-C-H interaction,
i.e., agostic interaction,2 plays a key role in some
important organometallic transformations (e.g. C-H
activation3 and olefin polymerization4) but also in the
stabilization of unsaturated intermediates.5 Some years
ago, we reported the formation of the first agostic acyl
F igu r e 1.
CO(PMe3)2,6a and suggested that the severely distorted
Mo-CR-Câ angle it presented (structure a , Figure 1)
could be considered as a model for the transition state,
or intermediate of the CO insertion reaction into M-C
bonds. This apparent connection between migratory CO
complex of a transition metal, Mo(C(O)CH3)(S2CNMe2)-
(1) (a) Parshall, G. W.; Ittle, S. D. Homogeneous Catalysis, 2nd ed.;
Wiley-Interscience: New York, 1992. (b) Collman, J . P.; Hegedus,
L. S.; Norton, J . R.; Finke, R. G. Principles and Applications of
Organotransition Metal Chemistry; University Science Books: Mill
Valley, CA, 1987.
(2) (a) Brookhart, M.; Green, M. L. H. J . Organomet. Chem. 1983,
250, 395. (b) Brookhart, M.; Green, M. L. H.; Wong, L.-L. Prog. Inorg.
Chem. 1988, 36, 1.
(3) (a) Gleiter, R.; Hyla-Kryspin, I.; Niu, S.; Erker, G. Organome-
tallics 1993, 12, 3828. (b) Crabtree R. H. Angew. Chem., Int. Ed. Engl.
1993, 32, 789.
(4) (a) Grubbs, R. H.; Coates, G. W. Acc. Chem. Res. 1996, 29, 85.
(b) Brintzinger, H. H.; Fischer, D. F.; Mu¨lhaupt, R.; Rieger, B.;
Waymouth, R. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143. (c) Tanner,
M. J .; Brookhart, M.; DeSimone, J . M. J . Am. Chem. Soc. 1997, 119,
7617. (d) Schmidt, G. F.; Brookhart, M. J . Am. Chem. Soc. 1985, 107,
1443. (e) Han, Y.; Deng, L.; Ziegler, L. J . Am. Chem. Soc. 1997, 119,
5939.
(5) (a) J ordan, R. F.; Bradley, P. K.; Baenzinger, N. C.; Lapointe, R.
E. J . Am. Chem. Soc. 1990, 112, 1289. (b) Brookhart, M.; Lincoln, D.
M.; Bennett, M. A.; Pelling, S. J . Am. Chem. Soc. 1990, 112, 2691. (c)
Conroy-Lewis, F. M.; Mole, L.; Redhouse, A. D.; Litster, S. A.; Spencer,
J . L. J . Chem. Soc., Chem. Commun. 1991, 1601. (d) Cracknell, R. B.;
Orpen, A. G.; Spencer, J . L. J . Chem. Soc., Chem. Commun. 1984, 326.
(e) Turner, H. W.; Schrock, R. R. J . Am. Chem. Soc. 1983, 105, 4942.
(f) Feng, S. G.; White, P. S.; Templeton, J . L. J . Am. Chem. Soc. 1990,
112, 8192.
(6) (a) Carmona, E.; Sa´nchez, L.; Mar´ın, J . M.; Poveda, M. L.;
Atwood, J . L.; Priester, R. D.; Rogers, R. D. J . Am. Chem. Soc. 1984,
106, 3214. (b) Carmona, E.; Contreras, L.; Poveda, M. L.; Sa´nchez, L.
J . Am. Chem. Soc. 1991, 113, 4322. (c) Contreras, L.; Monge, A.;
Pizzano, A.; Ruiz, C.; Sa´nchez, L.; Carmona, E. Organometallics 1992,
11, 3971.
10.1021/om9807248 CCC: $18.00 © 1999 American Chemical Society
Publication on Web 07/20/1999

Dihaptoacyl and â-Agostic Acyl Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3295
Sch em e 1
insertion and agostic M-C-H interactions has been
noted by others. Thus, time-resolved infrared and opti-
cal spectroscopy studies have considered agostic acyls
as plausible alternatives to solvent stabilized- or dihap-
toacyl intermediates,7 whereas a recent theoretical
investigation of the carbonylation of CH4 by the model
catalyst RhCl(CO)(PH3)2 has shown that the transition
state for the migratory insertion step is stabilized by
the formation of a â-agostic RhC(O)CH3 bond.8
of the groups PMe3 and formation of the acyl derivatives
2-6 (Scheme 1). As shown in Scheme 1, the new acyls
are designated by numbers from 2 to 6, according to the
nature of the dithio acid ligand. Since two different R
groups have been employed, they are further denoted
with the appropriate Me or P h symbol. In some
instances two isomeric possibilities may be observed,
namely the dihaptoacyl structure of type d (see above)
or the agostic formulation a , hence these letters will
additionally be used when appropriate.
For reasons that remain unknown to us, the experi-
mental observation of agostic acyls is still restricted to
With the exception of compound 5-P h , obtained as an
oily, albeit spectroscopically pure material, complexes
3-6 are isolated as orange-red crystalline solids of the
dihapto isomer. In contrast, acyl 2-Me crystallizes from
its solutions in Et2O-petroleum ether mixtures as a
mixture of the agostic isomer 2a -Me (yellow-orange) and
the dihapto compound 2d -Me (orange-red). Careful
crystallization, or even manual separation of the crys-
talline mixture, allows the separation of the two iso-
mers. The phenyl derivative 2-P h seems to crystallize
exclusively as the agostic form, even though the dihapto
complex predominates in solution (vide infra). They are
all moderately stable to air in the solid state, but very
sensitive in solution. In particular they demonstrate
easy C-Si heterolysis9 in the presence of adventitious
water, to generate the compounds Mo(C(O)CH3)(S2CX)-
CO(PMe3)2, previously described by our group.6
The coordination mode of the acyl fragment in the
above complexes can be readily identified from their IR
spectra (Nujol mull). Thus derivatives 3d -6d show two
bands in the carbonyl region: a strong absorption in
the interval 1800-1750 cm-1 due to υ(CO) of a terminal
carbonyl ligand; and a medium intensity band in the
1490-1465 cm-1 range, characteristic of an acyl frag-
ment coordinated in the dihapto fashion.10 Contrary to
the acyl fragment spectra, the agostic isomers 2a -Me
and 2a -P h , in which back-donation to the CdO frag-
ment is absent, present the corresponding band as a
fairly strong absorption at 1615 and 1598 cm-1 respec-
tively.
Mo complexes of composition Mo(C(O)CH3)(S2CX)CO-
(PMe3)2, prepared and characterized in our laborato-
ries.6 This series of complexes has been demonstrated
to be particularly suitable for studying the preference
for the agostic coordination (a ) over the dihaptoacyl (d )
or the alkyl-carbonyl formulations (c), since the relative
stabilities of these isomeric structures appear to depend
in solution upon the nature of the bidentate, mono-
anionic S2CX ligand. For instance, compounds that
incorporate bulky and strongly electron-releasing di-
alkyldithiocarbamate groups exist in solution as equi-
librium mixtures between the a and c type compounds
while the alkylxanthate derivatives show in addition the
presence of the dihapto compound of type d . Contrarily,
the analogous complex of the less-donating 1-pyrrolyl
dithiocarbamate group exists in solution as the d isomer
only.
With the aim of improving our knowledge on the
factors that govern the adoption of structure a we
present here an extension to our previous work with
molybdenum acyls. The synthesis and characterization
of silyl derivatives of composition Mo(C(O)CH2SiMe2-
R)(S2CX)CO(PMe3)2 (R ) Me, Ph; S2CX ) alkyl xanthate
or dialkyldithiocarbamate) are described, as well as
spectroscopic studies of their behavior in solution.
Additional structural evidence for the agostic coordina-
tion is given by a new X-ray crystallographic study and
valuable complementary information is provided by the
results of a theoretical analysis of the metal-acyl
interaction by ab initio methods.
Further insight into the agostic nature of the interac-
tion between the acyl fragment and the metal center in
these compounds has been gained from an X-ray study
carried out with the 2a -Me derivative. Suitable crystals
of this compound have been obtained by careful crystal-
lization of its solutions in Et2O. Figure 2 shows an
ORTEP diagram of its molecules, while Tables 1 and 2
display crystallographic data and some selected bond
distances and angles, respectively. The geometry of 2a -
Me clearly resembles the molecular structure of other
agostic acyls studied previously in our laboratory.6 If
the acyl group is considered to occupy a single coordina-
Resu lts a n d Discu ssion
Syn t h esis of t h e Acyls Mo(C(O)CH2SiMe2R )-
(S2CX)(CO)(P Me3)2 (R ) Me, P h ). X-r a y Str u ctu r e
of th e Agostic Com p lex Mo(C(O)CH2SiMe3)(S2-
CNMe2)CO(P Me3)2, 2a -Me. Reaction of the alkaline
salts (Na+ or K+) of several dialkyldithiocarbamate and
alkylxanthate ligands, in diethyl ether (Et2O) or tet-
rahydrofuran (THF), with the chloro(acyl) complexes
Mo(η2-C(O)CH2SiMe2R)Cl(CO)(PMe3)3 (R ) Me, 1-Me;
Ph, 1-P h ) produces incorporation of the dithio acid
ligand with concomitant release of the chloride and one
(9) For other C-Si heterolyses, see for example: (a) Carmona, E.;
Contreras, L.; Gutie´rrez-Puebla, E.; Monge, A.; Sa´nchez, L. J . Orga-
nometallics 1991, 10, 71. (b) Allen, S. R.; Green, M.; Orpen, A. G.;
Williams, I. D. J . Chem. Soc., Chem. Commun. 1982, 826.
(10) Although the value of υ(CO) cannot be taken as diagnostic of
dihapto coordination, confident assignment can be made along a series
of similar compounds: Durfee, L. D.; Rothwell, I. P. Chem. Rev. 1988,
88, 1059.
(7) (a) Boese, W. T.; Ford, P. C. J . Am. Chem. Soc. 1995, 117, 8381.
(b) Boese, W. T.; Lee, B.; Ryba, D. W.; Belt, S. T.; Ford, P. C.
Organometallics 1993, 12, 4739.
(8) (a) Margl, P.; Ziegler, T.; Blo¨chl J . Am. Chem. Soc. 1996, 118,
5412.

3296 Organometallics, Vol. 18, No. 17, 1999
Ujaque et al.
Ta ble 2. Selected Bon d Dista n ces a n d An gles
for 2a -Me
Mo-P1
Mo-P2
Mo-S1
Mo-S2
Mo-C1
Mo-C2
P1-C11
P1-C12
P1-C13
P2-C21
P2-C22
P2-C23
2.432(3)
2.413(3)
2.526(3)
2.528(3)
1.864(11)
2.045(11)
1.822(14)
1.785(15)
1.839(14)
1.831(13)
1.815(12)
1.822(13)
Si-C3
Si-C4
Si-C5
Si-C6
S1-C7
S2-C7
O1-C1
O2-C2
N-C7
N-C8
N-C9
C2-C3
1.873(11)
1.859(15)
1.868(14)
1.845(15)
1.710(11)
1.710(11)
1.218(14)
1.214(13)
1.324(13)
1.466(19)
1.460(16)
1.587(15)
C1-Mo-C2
S2-Mo-C2
S2-Mo-C1
S1-Mo-C2
S1-Mo-C1
S1-Mo-S2
P2-Mo-C2
P2-Mo-C1
P2-Mo-S2
P2-Mo-S1
P1-Mo-C2
P1-Mo-C1
P1-Mo-S2
P1-Mo-S1
P1-Mo-P2
Mo-P1-C13
Mo-P1-C12
Mo-P1-C11
C12-P1-C13
C11-P1-C13
C11-P1-C12
Mo-P2-C23
Mo-P2-C22
120.1(4)
121.3(3)
108.0(3)
118.2(3)
108.6(3)
70.0(1)
74.9(3)
76.0(3)
152.7(1)
83.0(1)
75.3(3)
Mo-P2-C21
C22-P2-C23
C21-P2-C23
C21-P2-C22
C5-Si-C6
C4-Si-C6
C4-Si-C5
C3-Si-C6
C3-Si-C5
C3-Si-C4
Mo-S1-C7
Mo-S2-C7
C8-N-C9
C7-N-C9
C7-N-C8
Mo-C1-O1
Mo-C2-O2
O2-C2-C3
Mo-C2-C3
Si-C3-C2
S2-C7-N
116.7(4)
102.3(6)
100.8(6)
102.3(6)
107.2(6)
110.5(6)
109.5(6)
105.5(6)
112.8(6)
111.1(6)
86.6(4)
F igu r e 2. ORTEP diagram of Mo(C(O)CH2SiMe3)(S2-
CNMe2)CO(PMe3)2 (2a -Me).
75.7(3)
86.9(1)
86.5(4)
116.0(9)
121(1)
Ta ble 1. Cr ysta l a n d Refin em en t Da ta for 2a -Me
formula
Mr
MoS2P2SiO2NC15H35
511.5
156.9(1)
119.7(1)
114.5(5)
116.5(5)
117.5(5)
102.5(6)
100.8(6)
102.7(7)
114.0(4)
118.2(5)
122.0(9)
179.2(9)
148.1(8)
117.2(9)
94.7(6)
114.3(7)
123.1(8)
120.7(8)
116.0(6)
cryst syst
space group
a, Å
orthorhombic
P212121
9.338(4)
b, Å
12.480(3)
c, Å
21.103(9)
2459(2)
V, Å3
Z
4
S1-C7-N
F(000)
1064
1.38
22
8.7
S1-C7-S2
F(calcd), g cm-3
temp, °C
µ, cm-1
approach of one of the â-C(3)-H bonds can be envisioned
as a pivotating of the acyl ligand from the monohapto
coordination, to bring the carbon atom C(3) closer to the
Mo atom. This originates the agostic interaction (or
agostic distortion, vide infra) characterized by Mo-C(3)
and Mo-H(32) distances of 2.69(11) and 2.65(9) Å,
respectively. These values lie well in the range observed
for other agostic acyls and do not require further
comment. Also very characteristic of the agostic acyl
interaction are the values of the Mo-C(2)-O(2) and
Mo-C(2)-C(3) angles (148.1(8) and 94.7(6)° respec-
tively). They are in the intervals expected for this type
of coordination but, as already discussed at length,6
differ considerably from those corresponding to the η1-
and η2-acyl structures. Finally, it should be noted that
the formal substitution of a H atom of the C(O)CH3
fragment by a SiMe3 unit does not introduce any
significant variation in the structural parameters of
these complexes.
cryst dimens, mm
diffractometer
radiation
0.1 × 0.1 × 0.2
Enraf-Nonius CAD4
graphite-monochromated Mo KR
(λ ) 0.710 69 Å)
Ω/2θ
scan technique
data collected
no. of rflcns collcd
no. of unique data
(-11, 0, 0) to (11, 15, 26)
5418
2748
no. of unique data, (I) g 2σ(I) 2290
R(int), %
no. of std rflcns
R(F), %
Rw(F), %
average shift/error
4.1
3
3.7
5.5
0.08
tion site, the geometry of 2a -Me can be regarded as
derived from a distorted octahedron, with the chelating
dithio acid ligand and the two trimethylphosphine
fragments occupying the equatorial plane, and the
carbonyl functionalities, i.e., the terminal carbonyl
ligand and the agostic acyl moiety, positioned in the
axial sites. The geometry of the coordinated acyl ligand
deserves specific comment. The interaction between the
metal and the terminal CH2R group of the acyl ligand
produces a severe distortion with respect to the geom-
etry characteristic of the monohaptoacyl fragment,
M-C(O)R, clearly denoted by the value of the angle
Mo-C(2)-C(3) of 94.7(6)°. This is significantly shorter
than the value of 120.9° observed in CpMo(η1-C(O)CH3)-
(CO)2PPh3.11 Since the angle formed by the O(2), C(2),
and C(3) atoms (117.2(9)°) remains close to the ideal
value expected for an sp2 hybridation at C(3), the
Str u ctu r a l a n d Electr on ic F ea tu r es of Mo(C(O)-
CH2R)(S2CX)(CO)(P Me3)2 Agostic Com plexes. Theo-
retical calculations were carried out on a number of
model systems of experimentally reported acyl com-
plexes. The modeling always consisted of the replace-
ment by hydrogen atoms of substituents remote from
the metal-acyl interaction. In this way, PMe3 ligands
were modeled as PH3, and the SiMe3 substituent as
SiH3. Calculations were carried out on the parent
complex Mo(C(O)CH3)(S2CNMe2)CO(PMe3)2 (7)6a and on
(11) Churchill, M. R.; Fennessey, J . P. Inorg. Chem. 1968, 7, 953.
the species reported here Mo(C(O)CH2SiMe3)(S2CNMe2)-

Dihaptoacyl and â-Agostic Acyl Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3297
Ta ble 3. Releva n t Str u ctu r a l P a r a m eter s (in Å a n d d eg) of Mo(COCH2R)(S-S)(CO)(P R3)2 Agostic
Com p lexesa
7b
8
2a -Me
9
10c
11
X-ray
opt.
X-ray
opt.
X-ray
opt.
Mo-C(1)
Mo-C(2)
Mo-S(1)
Mo-S(2)
Mo-P(1)
Mo-P(2)
1.83(1)
2.05(1)
2.529(2)
2.547(2)
2.435(2)
2.430(3)
1.898
2.048
2.607
2.607
2.506
2.506
1.864(11)
2.045(1)
2.526(3)
2.528(3)
2.432(3)
2.413(3)
1.898
2.054
2.603
2.602
2.502
2.498
1.879(7)
2.057(7)
2.504(1)
2.497(2)
2.419(2)
2.411(1)
1.888
2.065
2.571
2.571
2.510
2.510
C(1)-Mo-C(2)
S(1)-Mo-S(2)
P(1)-Mo-P(2)
S(1)-Mo-P(1)
S(2)-Mo-P(2)
122.7(4)
69.9(8)
120.55(9)
154.80(9)
153.50(9)
125.9
120.1(4)
126.7
118.0(3)
124.4
71.1
116.9
156.4
156.4
70.0(1)
119.7(1)
156.9(1)
152.7(1)
71.5
116.5
156.8
157.0
71.9(1)
120.2(1)
155.4(1)
157.7(1)
74.0
117.4
157.8
157.8
a
b
For the atom numbering, see Scheme 2. Reference 6a. c Reference 6c.
Ta ble 4. Releva n t Bon d Dista n ces (Å) a n d An gles (d eg) for th e Mo(COCH2R) F r a gm en t in Agostic Acyl
Com p lexesa
7b
8
2a -Me
9
10c
11
X-ray
opt.
X-ray
opt.
X-ray
opt.
Mo-C(2)
Mo-C(3)
Mo-Ha
C(3)-Ha
C(2)-C(3)
2.05(1)
2.60(1)
2.06(9)
1.00(9)
1.57(2)
2.048
2.613
2.351
1.103
1.595
2.045(11)
2.69(11)
2.65(9)
1.01(9)
1.587(15)
2.054
2.600
2.247
1.108
1.588
2.057(7)
2.762(4)
2.56(6)
1.00(9)
1.561(9)
2.065
2.671
2.398
1.100
1.585
Mo-C(2)-C(3)
Mo-C(2)-O(2)
C(3)-C(2)-O(2)
C(2)-C(3)-Ha
90.9(8)
149.2(8)
120(1)
99(5)
90.8
147.5
121.7
115.7
94.7(6)
148.1(8)
117.2(9)
110(5)
90.2
147.1
122.7
114.1
98.6(4)
144.4(5)
116.9(6)
116(5)
93.1
145.7
121.2
114.4
a
b
For the atom numbering, see Scheme 2. Reference 6a. c Reference 6c.
Sch em e 2
CO(PMe3)2 (2a -Me), modeled as Mo(C(O)CH3)(S2CNH2)
(CO)(PH3)2 (8) and Mo(C(O)CH2SiH3)(S2CNH2)(CO)-
(PH3)2 (9), respectively. Additional calculations were
also carried out on the phosphonium xanthate complex
are reproduced well by the calculations, with differences
not larger than one tenth of an A° ngstrom and six
degrees, respectively, with respect the experimental
values. The largest differences can be found for complex
11, in which the most severe simplification of the
experimental system 10 has been introduced, replacing
the electron-drawing -CH2CF3 group by H. The model-
ing and computation process is therefore suitable for
this system.
In the complexes under study, six ligands are coor-
dinated to the metal. We will attempt to classify the
geometry in terms of the most frequent polyhedron
geometry found in transition metal: six-coordination.12
Given that the three species adopt very similar geo-
Mo(C(O)CH3)(S2C(PMe3)OCH2CF3)(CO)(PMe3)2 (10)6c
modeled as Mo(C(O)CH3)(S2C(PH3)OH) (CO)(PH3)2 (11).
Computed and experimental parameters concerning the
coordination sphere of the metal are collected in Table
3, and those involved in the molybdenum-acyl interac-
tion in Table 4. Atom numbering is shown in Scheme
2.
The data in Table 3 indicates a good agreement
between the experimental and calculated geometrical
parameters. Despite the simplifications introduced in
the theoretical study, bond distances and bond angles
(12) Kubacek, P.; Hoffmann, R. J . Am. Chem. Soc. 1981, 103, 4320.

3298 Organometallics, Vol. 18, No. 17, 1999
Ujaque et al.
Sch em e 3
well-known orbital splitting for an octahedral ligand
field. The three lower energy orbitals (dz , dxy and dyz)
2
are those oriented away from the ligands, while the two
2
2
higher energy orbitals (dx -y and dxz) point toward the
ligands. When going toward the bicapped tetrahedron,
the ligands initially in the y-axis are bent away toward
2
2
negative values of the Z-coordinate. The dx -y orbital
is stabilized because its two lobes directed along the
y-axis are no longer pointing toward the two ligands.
The energy of the dyz orbital is increased because the
two ligands that have moved are now pointing toward
two lobes of this orbital. The dxy and dxz orbitals do not
have contributions from the axial ligands, so they will
2
not be affected by the distortion. The dz orbital will be
only slightly affected.
The resulting orbital energy scheme for the d set of
orbitals of a bicapped tetrahedron complex consists of
two low energy orbitals of different symmetry (a′ and
a′′ in the Cs group representation), two antibonding
orbitals of the same symmetry (a′) and a highest a′′
antibonding orbital. In a low-spin d4 complex the two
nonbonding orbitals are filled so this is a favorable
electron-counting for the bicapped tetrahedral structure.
The most noteworthy feature of complexes 2a -Me, 7,
and 10 is the strong distortion of the metal-acyl frag-
ment with respect to the geometry expected for a
monohaptoacyl coordination. This distortion, mainly
reflected in the values of the Mo-C(2)-C(3) angles,
causes a pivoting of the CH2R group of the acyl ligand
toward the metal. As can be seen in Table 4, the agostic
distortion is also found in the optimized geometries of
the model complexes 8, 9, and 11. Optimized geometric
parameters of the Mo-acyl fragment agree well with the
X-ray determined values.6 The largest differences are
found for the Mo-H distance, but this could be ex-
plained by the uncertainty in the X-ray determination
of H atoms. Although the calculations resolve â-agostic
structures as the most stable of the three systems
considered, they are not able to reproduce the experi-
mental differences between 7, 2a -Me, and 10. The Mo-
C(2)-C(3) angle of 8 (90.8°) is in excellent accord with
the experimental value of 7 (90.9°).6a However, the
increase of ca. 5° of this angle, due to the substitution
of a H atom of the acyl CH2R group by SiMe3, is not
found in model system 9. There is also a difference for
the Mo-C(1)-C(3) angle of 5° between 10 and 11. The
results indicate that the very subtle electronic differ-
ences introduced in the experimental systems by chang-
ing groups in the S-S donor ligand or the R of the acyl
ligand are not accounted for by the model systems.
metrical arrangements, we will discuss only the (2a -
Me) structure. In this complex the two S atoms and two
P atoms define a plane, and the terminal carbonyl
ligand and the agostic acyl ligand are in a plane
perpendicular to that plane. The chelating dithiocar-
bamate ligand imposes a rigid angle (70.0° exp, 71.5°
calcd). The S(1)-Mo-P(2) and S(2)-Mo-P(1) angles
(83.0° exp, 85.7° calcd and 86.9° exp, 85.8° calcd
respectively) are close to the ideal octahedral value (90°).
However, the P(1)-Mo-P(2) (119.7° exp, 116.5° calcd)
and more so the C(1)-Mo-C(2) (120.1° exp, 126.7° calcd)
angles deviate substantially from those of the octahe-
dron (90° and 180° respectively). We must recall our
compounds are d4-six-coordinate species and substantial
deformations from octahedral geometry toward bicapped
tetrahedron or trigonal prism geometry have been
reported for these complexes.12
Actually 2a -Me can be described as a distorted
bicapped tetrahedron in which the two sulfur atoms of
the xanthate and the carbon atoms of the acyl and
carbonyl ligands occupy the tetrahedron vertexes whereas
the two trimethyl phosphine ligands are the capping
ligands that point toward two faces of the tetrahedron.
The angles between the “tetrahedral” ligands are S(1)-
Mo-C(1) ) 108.0° exp, 101.1° calcd, S(1)-Mo-C(2) )
118.2° exp, 121.1° calcd, C(1)-Mo-C(2) ) 120.1° exp,
126.7° calcd, and S(1)-Mo-S(2) ) 70.0° exp, 71.5° calcd
A detailed theoretical analysis of the deformations
from octahedral geometry in d4 transition-metal com-
plexes has been performed by Kubacek and Hoffmann.12
The electronic consequences of the deformation from the
octahedral geometry toward a bicapped tetrahedron can
be understood with very simple orbital arguments.13 In
Scheme 3 we have traced a qualitative picture of the
energetic evolution of the block-d orbitals along the
geometric deformation. On the left side, there is the
The origin of the agostic distortion observed in the
Mo-acyl complexes can be understood with the help of
the qualitative orbital picture presented in Scheme 3.
These species are d4-Mo(II) compounds with a formal
valence electron count for Mo of 16. The two occupied d
2
orbitals are dz and dxy. The bonding of the monohapto
C
acyl to the metal is mainly due to the interaction of the
Cacyl lone pair with a lobe of the dyz orbital. The agostic
2
2
distortion allows the involvement of the dx -y orbital
2
2
in the M-Cacyl bond. Both dyz and dx -y are a′ orbitals
and mix to reinforce the M-Cacyl bond. The agostic
distortion brings the C(3) and H(2) atoms of the acyl
fragment close to the molybdenum atom, where they can
(13) Albright, T. A.; Burdett, J . K.; Whangbo, M. H. Orbital
Interactions in Chemistry; Wiley: New York, 1985; pp 289-294.

Dihaptoacyl and â-Agostic Acyl Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3299
Ta ble 5. Op tim ized Geom etr ica l P a r a m eter s
(Å a n d d eg) for th e Meta l-Eth yl F r a gm en t in
Mo(CH2CH3)(S2CNH2)(CO)(P H3)2 (12)
Sch em e 4
12e
12s
12ef
12sf
Mo-C(2)
Mo-C(3)
Mo-Haa
2.243
2.605
2.168
1.127
1.098
1.546
1.094
2.249
2.654
2.683
1.104
1.102
1.548
1.095
2.246
3.130
2.950
1.108
1.098
1.558
1.097
2.260
3.136
3.284
1.101
1.102
1.550
1.098
C(3)-Haa
C(3)-Hb
C(2)-C(3)
C(2)-H(2)
Mo-C(2)-C(3)
Mo-C(2)-H(2)
H(2)-C(2)-C(3)
C(2)-C(3)-Ha
C(2)-C(3)-Hb
84.8
118.1
112.7
114.1
112.0
86.6
118.4
111.6
111.9
111.2
109.4b
109.4b
112.8
108.6
113.0
110.9
113.1
107.9
111.9
109.5
rel energyc
0.0
2.4
11.3
9.3
a
b
Ha refers to the hydrogen atom closest to the metal. Fixed
2
2
interact with the 4dx -y , 5s and 5py metal orbitals, all
of them empty. Thus, an agostic interaction between the
metal and the -CH2R group of the acyl ligand is at work
in these systems. In what follows we will analyze in
detail this interaction, focusing on the Mo(COCH3)(S2-
CNH2)(CO)(PH3)2 system 8.
value. c In kcal mol-1
.
Th eor etica l An a lysis of th e â-Agostic In ter a c-
tion . To our knowledge, the only agostic acyl compounds
experimentally characterized are the series of molyb-
denum complexes Mo(COCH2R)(S2CX)(CO)(PMe3)2. Sev-
eral â-agostic ethyl complexes have been reported,2 and
they are experimental and theoretical benchmarks for
ethyl â-agostic interactions.14 To go further into the
analysis of the specificity of the agostic interaction in
acyl complexes we have also analyzed theoretically the
interaction of an Mo(S2CX)(CO)(PR3)2 metal fragment
with an ethyl ligand. Given the electronic structure of
this d4-metal fragment, it may be hoped that it will also
develop a â-agostic interaction with the ethyl. In this
way, we will be able to compare acyl and ethyl â-agostic
interactions.
To better analyze this interaction, we performed for
complex 12 a series of geometry optimizations with
different constraints. These geometries are shown in
Scheme 4. The first of them, corresponding to the
absolute minimum, has no constraints, and has the
Câ-H bond eclipsed with respect to the Mo-CR bond.
It is labeled 12e. A second geometry optimization, 12s,
was performed with a forced staggered conformation
around the CR-Câ bond. Both 12e and 12s present an
important agostic distortion, with values of the Mo-
CR-Câ angle of 84.8° and 86.6°, respectively. Two other
geometry optimizations, 12ef and 12sf, were carried out
with an anagostic arrangement. This was accomplished
through freezing the Mo-CR-Câ angle to a tetrahedral
value of 109.4°. The relative energies of these four
structures, together with the geometric parameters that
describe the ethyl coordination, are collected in Table
5. The optimized geometry of the most stable structure
12e is presented in Figure 3.
F igu r e 3. Optimized geometry of the most stable structure
of Mo(CH2CH3)(S2CNH2)(CO)(PH3)2 (12e).
establish a â-agostic interaction with an ethyl ligand.
Besides the distorted Mo-CR-Câ angles, the Câ-H bond
that points to the metal is significantly lengthened
(1.127 Å vs 1.10 Å for a normal C-H bond). Moreover,
geometric parameters for the metal-ethyl fragment in
12e are very similar to those of the well characterized
â-agostic EtTiCl3(dmpe) complex.14 In particular, the
Mo-CR-Câ angle is very close to the Ti-CR-Câ angle
(86.3(6)° X-ray, 293 K,14a 84.57(9)o X-ray, 105 K14d).
Optimization of 12 with a staggered ethyl group (species
12s) gives also an agostic conformer, characterized by
a Mo-CR-Câ angle of 92.3°. This conformer is calculated
to lie only 2.4 kcal mol-1 above the minimum structure
12e. As it has been shown very recently in a thorough
study of â-agostic interactions, these interactions do not
necessarily require an ethyl group in which a Câ-H
bond points toward the metal atom.14d In 12s all the
Câ-H bond distances have the same value, pointing to
the absence of a direct M-H interaction. In agreement
with the theoretical study of EtTiCl3(dmpe),14d methyl
rotation occurs in 12 with only a small change of the
Mo-C-C angle. Early EH calculations in the agostic
[Co(C5Me5)(Et)(PMe2Ph)]+ complex indicated a different
mechanism for the methyl rotation, in which the rota-
tion about the CR-Câ bond takes place after cleavage
of the agostic interaction.15
The anagostic structures 12ef and 12sf are not
minima in the potential energy surface and have
considerably higher energies than 12e: 11.3 kcal mol-1
for the eclipsed 12ef and 9.3 kcal mol-1 the staggered
12sf. The energy of the agostic interaction in 12 is in
the range of the interaction energies determined in
experimental16 and computational17 studies of agostic
12e is the minimum energy structure for 12, indicat-
ing that a Mo(S2CX)(CO)(PR3)2 metal fragment should
(14) (a) Dawoodi, Z.; Green, M. L. H.; Mtetwa, V. S. B.; Prout, K.;
Schult, A. J .; Williams, J . M.; Koetzle, T. F. J . Chem. Soc., Dalton
Trans. 1986, 1629. (b) McGrady, G. S.; Downs, A. J .; Haaland, A.;
Scherer, W.; Mckean, D. C. Chem. Commun. 1997, 1547. (c) Popelier,
P. L. A.; Logothetis, G. J . Organomet. Chem. 1998, 555, 101. (d)
Haaland, A.; Scherer, W.; Rund, K.; McGrady, G. S.; Downs, A. J .;
Swang, O. J . Am. Chem. Soc. 1998, 120, 3762.
(15) Cracknell, R. B.; Orpen, A. G.; Spencer, J . L. J . Chem. Soc,
Chem. Commun. 1986, 1005.

3300 Organometallics, Vol. 18, No. 17, 1999
Ujaque et al.
systems (10-15 kcal mol-1). The pivoting of the methyl
group toward the metal causes a notable stabilization
of the system. The agostic interaction has been often
thought as the result of the electronic donation from a
C-H bond to a metal low-lying vacant orbital, to fill a
vacant coordination site in an electron deficient com-
pound.2,18 There are however a growing number of
experimental and theoretical studies pointing to the
need for a more elaborate analysis of agostic interac-
tions. On one hand, it has been recently proved that in
some systems steric effects of bulky phosphines can
assist in stabilizing agostic interactions.19 On the other
hand, a theoretical study of â-agostic interactions in
titanium ethyl derivatives has shown that the agostic
distortion takes place regardless of the orientation of
the C-H bond. Following that analysis, one can divide
the agostic interaction into two terms, corresponding to
the M-H and M-Câ interaction. From our calculations
it is possible to roughly separate the two contributions
in complex 12. The energy difference between the two
eclipsed conformations (11.3 kcal mol-1) gives the total
energy of the â-agostic interaction, whereas the energy
difference between the two staggered conformations (6.9
kcal mol-1) approximates the Mo-Câ contribution.
Thus, the direct M-H component of the agostic interac-
tion in 12 can be estimated as 4.4 kcal mol-1. This
component is enough to invert the relative stabilities
of the eclipsed and staggered conformations with respect
to the anagostic situation. All these energy values agree
with the presence of a strong â-agostic interaction
between the methyl group of the ethyl ligand and the
metal fragment in 12, even stronger than in TiEtCl3-
(dmpe). For instance, Becke3LYP calculations in the
titanium complex have produced staggered agostic and
anagostic structures at 0.2 and 1.8 kcal mol-1 respec-
tively, above the eclipsed agostic minimum.14c A value
of 8.4 kcal mol-1 has been calculated for the â-agostic
stabilization energy of the model [EtTiCl2]+ cation.14c
The well-known sensitivity of MP2 results to the basis
set could cast some doubt on the validity of our
quantitative analysis, in particular because we are
dealing with weak interactions. To evaluate the basis
set dependence of the results, the energy of some
selected points was recomputed in the same frozen
geometry with a much larger basis set (basis set II)
which included polarization functions in all atoms. The
energies recomputed in this way were those of the
agostic structures 12e and 12s, and those of the ana-
gostic structures 12ef and 12sf. Changes were minimal.
Ta ble 6. Geom etr ica l P a r a m eter s (Å a n d d eg) for
th e Meta l-Acyl fr a gm en t in
Mo(CH3CO)(S2CNH2)(CO)(P H3)2 (8)
8e
8s
8ef
8sf
Mo-C(2)
Mo-C(3)
Mo-Haa
C(3)-Ha
C(3)-Hb
C(2)-C(3)
C(2)-O(2)
2.048
2.613
2.351
1.103
1.097
1.595
1.248
2.044
2.661
2.800
1.098
1.097
1.623
1.248
2.106
3.180
3.158
1.100
1.098
1.552
1.276
2.128
3.198
3.406
1.101
1.097
1.549
1.275
Mo-C(2)-C(3)
Mo-C(2)-O(2)
O(2)-C(2)-C(3)
C(2)-C(3)-Ha
C(2)-C(3)-Hb
90.8
92.3
120.0b
120.0b
147.5
121.7
115.7
107.4
146.7
121.0
113.4
101.7
126.7
113.3
113.9
107.8
125.6
114.4
111.1
107.1
rel energyc
0.0
4.0
12.7
13.8
a
b
Ha refers to the hydrogen atom closest to the metal. Fixed
value. c In kcal mol-1
.
Sch em e 5
The relative energies, which were 0.0, 2.4, 11.3, and 9.3
kcal/mol with basis set I, became 0.0, 2.4, 13.1, and 11.0
kcal/mol with the more extended basis set II. The agostic
structures 12e and 12s have exactly the same energy
difference, and the anagostic species are destabilized by
ca. 2 kcal/mol with respect to the agostic ones. These
values would give an estimation of 8.6 kcal/mol for the
Mo-Câ component of the agostic stabilization and 4.5
kcal/mol for the Mo-H component. In summary, these
frozen geometry calculations prove that basis set I
provides a sufficiently accurate picture.
We have repeated the same calculations presented in
the previous paragraph for the ethyl complex 12 in the
case of the acyl complex 8. In this way, in addition to
the most stable conformation for the Mo(C(O)CH3)(S2-
CNH2)(CO)(PH3)2 system, in which the Câ-H bond
eclipses the M-CR bond (from now on 8e), we have also
optimized 8 with staggered acyl (8s), and avoiding the
â-agostic interaction by fixing the Mo-CR-Câ angle at
120° with an eclipsed (8ef) and staggered (8sf) acyl
conformations (see Scheme 5). Geometric parameters
describing the metal-acyl fragment in the four struc-
tures, together with their relatives energies, are col-
lected in Table 6. The optimized geometries of the most
stable structure are pictured in Figure 4.
(16) (a) Crabtree, R. H. Chem. Rev. 1985, 85, 245. (b) Gonza´lez, A.
A.; Zhang, K.; Nolan, S. P.; Lo´pez de la Vega, R.; Mukerjee, S. L.; Hoff,
C. D.; Kubas, G. L. Organometallics 1988, 7, 2429. (c) Burger, B. J .;
Thompson, M. E.; Cotter, W. D.; Bercaw, J . E. J . Am. Chem. Soc. 1990,
112, 1566. (d) Zhang, K.; Gonza´lez, A. A.; Mukerjee, S. L.; Chou, S.-J .;
Hoff, C. D.; Kubat-Martin, K. A.; Barnhart, D.; Kubas, G. J . J . Am.
Chem. Soc. 1991, 113, 9170.
(17) (a) Kawamura-Kuribayashi, H.; Koga, N.; Morokuma, K. J . Am.
Chem. Soc. 1992, 114, 2359. (b) Lohrenz, J . C. W.; Woo, T. K.; Ziegler,
T. J . Am. Chem. Soc. 1995, 117, 12793. (c) Margl, P.; Lohrenz, J . C.
W.; Ziegler, T.; Blo¨chl, P. E. J . Am. Chem. Soc. 1996, 118, 4434. (d)
Thomas, J . L. C.; Hall, M. B. Organometallics 1997, 16, 2318. (e)
Musaev, D. G.; Froese, R. D. J .; Svensson, M.; Morokuma, K. J . Am.
Chem. Soc. 1997, 119, 367. (f) Han, Y.; Deng, L.; Ziegler, T. J . Am.
Chem. Soc. 1997, 119, 5939. (g) Deng, L.; Woo, T. K.; Cavallo, L.; Margl,
R. M.; Ziegler, T. J . Am. Chem. Soc. 1997, 119, 6177.
(18) (a) Crabtree, R. H. Angew. Chem., Int. Ed. Engl. 1993, 32, 789.
(b) Shilov, A. E.; Shul’pin, G. B. Chem. Rev. 1997, 97, 2879.
(19) Ujaque, G.; Cooper, A. C.; Maseras, F.; Eisenstein, O.; Caulton,
K. G. J . Am. Chem. Soc. 1998, 120, 361.
At first glance, results for 8 and 12 are very similar.
In the most stable conformation of both a very acute
Mo-CR-Câ angle is found and a Câ-H bond of the

Dihaptoacyl and â-Agostic Acyl Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3301
has been interpreted in terms of C-C agostic interac-
tion. Interestingly, the variation of the CR-Câ distance
between 12e and 12ef, although slight, is in the opposite
direction. A small reduction of the C-C bond distance
is obtained in going from the anagostic to the agostic
structures (from 1.558 to 1.546 Å).
It is also worth mention that the Mo-CR bond
distance is shortened in forming the agostic interaction.
Although this trend is observed in both the ethyl and
acetyl derivatives, it becomes much more pronounced
in the agostic acyl. This fact has also been proved in
agostic alkylidene transition metal complexes,22 and has
been interpreted from extended Hu¨ckel by saying that
the strengthening of the Mo-CR bond is the driving
force for the R-agostic distortion.23 It must be remarked
that in 8 the strengthening of the Mo-CR bond and the
weakening of the CR-Câ in the agostic structures are
also coupled with a remarkable shortening of the acyl
CO distance (from 1.270 Å in 8ef to 1.248 Å in 8e).
Our theoretical studies on the acetyl complex indicate
an important activation of the CR-Câ in the agostic
structures. To determine if the activation is a conse-
quence of the presence of the agostic methyl group in
the acyl ligand or whether it is a more general phenom-
enon, we have also considered the same dithiocarbamate
metal fragment but now with a formyl ligand. The
optimized geometry of the Mo(HCO)(S2CNH2)(CO)-
(PH3)2 model complex is similar to that found for the
acetyl compound. A strong agostic distortion is taking
place in the formyl system, reflected in the very acute
Mo-CR-HR angle (65.6°). Other remarkable geometrical
parameters are the CR-H, Mo-CR, and acyl C-O bond
distances of 1.336, 1.937, and 1.334 Å, respectively. All
these values, together with the Mo-HR distance of 1.844
Å suggest that the structure we have obtained is along
the path of the migratory CO insertion reaction.
Comparison between ethyl and acyl â-agostic molyb-
denum complexes has shown that, despite the similar
structural features they present, there is an important
difference between them. Whereas the agostic ethyl
complex can be considered as an arrested structure on
an initial step of the reaction path corresponding to
â-elimination, the agostic acetyl complex can be seen
as an arrested point in the CO insertion reaction into
the Mo-C bond. In the next section we will further
analyze this aspect.
F igu r e 4. Optimized geometry of the most stable structure
of Mo(COCH3)(S2CNH2)(CO)(PH3)2 (8e).
terminal methyl eclipses the M-CR bond. Thus, an
agostic interaction is taking place. In both the ethyl and
acetyl derivatives the agostic distortion is still present
in the staggered conformation (Mo-CR-Câ angles of
86.6° and 92.3° in 12s and 8s, respectively). Clearly,
there is no need of a direct M-H interaction for the
agostic distortion to take place. In 8 and 12, methyl
rotation occurs with a low energy barrier and with only
a small variation of the Mo-CR-Câ angle. The energetic
gain associated with the distortion is also similar
(around 12 kcal mol-1). However, a more detailed
inspection of the data shows important differences
between 8 and 12. The separation of the M-H and
Mo - Câ contributions to the agostic stabilization of 8e
(12.7 kcal mol-1) gives an estimation of 9.8 kcal mol-1
for the Mo - Câ component and only 2.9 kcal mol-1 for
the M-H component. The Mo‚‚‚Câ bonding plays a more
important role in stabilizing the â-agostic acyl than it
does in the ethyl derivative. The stability of the results
on complex 8 with respect to extension of the basis set
was again tested with single point calculations on basis
set II. Changes were again very small. The relative
energies of 8e, 8s, 8ef, and 8sf with basis set II were
0.0, 4.3, 15.9, and 17.0 kcal mol-1, while 0.0, 4.0, 12.7,
and 13.8 kcal mol-1 values were obtained with basis set
I. The agostic stabilization obtained from these frozen
geometry calculations would be of 15.9 kcal mol-1 for
8e, with contributions of 12.7 and 3.2 kcal mol-1 from
the Mo-Câ and Mo-H components, respectively.
Geometric parameters of the agostic structures of 8
are in accord with this behavior. In contrast to what is
found in 12, in 8e there is not a significant increase of
the in-plane Câ-H bond distance. On the contrary, a
remarkable lengthening of the CR-Câ bond distance
occurs in 8e and 8s. This fact can be seen by comparing
the CR-Câ distance in agostic and anagostic eclipsed
structures (1.595 and 1.552 Å, respectively). The value
in 8ef is very similar to that determined by X-ray
diffraction in the monohaptoacyl complex CpMo(COCH3)-
(CO)2(PPh3) (1.555 Å), in which a normal Mo-CR-Câ
angle of 120.9 o was found.11 To ensure that the interac-
tion Mo-Câ is responsible for the increase of the C-C
distance, we have also optimized the species CH3CONa,
with a fixed Na-C-O angle of 120°. A value of 1.563 Å
was obtained for the C-C distance, confirming that our
methodology furnishes acyl C-C distances not longer
than 1.56 Å when a metal - Câ interaction is not
present. A similar stretching of C-C bonds has been
found in several theoretical studies17a,17b,20 and in a very
recent X-ray diffraction study.21 In all cases stretching
Migr a tor y CO Dein ser tion Rea ction . Our experi-
mental results6,9a,28 indicate that molybdenum and
tungsten acyl complexes M(COCH2R)(S2CX)CO(PMe3)2
can easily undergo CO deinsertion to produce the seven
coordinate alkyl-carbonyl isomers M(CH2R)(S2CX)(CO)2-
(PMe3)2. In addition, we have shown in the preceding
(22) Schultz, A. J .; Brown, R. K.; Williams, J . W.; Schrock, R. R. J .
Am. Chem. Soc. 1981, 103, 169.
(23) (a) Goddard, R. J .; Hoffmann, R.; J emmis, E. D. J . Am. Chem.
Soc. 1980, 102, 7667. (b) Eisenstein, O.; J ean, Y. J . Am. Chem. Soc.
1985, 107, 1177. (c) Demolliens, A.; J ean, Y.; Eisenstein, O. Organo-
metallics 1986, 5, 1457.
(24) Maseras, F.; Eisenstein, O. New J . Chem. 1997, 21, 961.
(25) Versluis, L.; Ziegler, T. Organometallics 1990, 9, 2985.
(26) Curtis, M. D.; Shiu, K. B.; Butler, W. M. J . Am. Chem. Soc.
1986, 108, 1550.
(27) (a) Herrick, R. S.; Templeton, J . L. Inorg. Chem. 1986, 25, 1270.
(b) Kellner, R.; Prokopwoski, P.; Malissa, H. Anal. Chim. Acta 1974,
68, 401.
(20) Koga, N.; Morokuma, K. J . Am. Chem. Soc. 1988, 110, 108.
(21) Tomaszewski, R.; Hyla-Kryspin, I.; Mayre, C. L.; Arif, A. M.;
Gleiter, R.; Ernst, R. D. J . Am. Chem. Soc. 1998, 120, 2959.
(28) Strong influence of steric effects has been found to favor six-
coordinated dihaptoacyl coordination over the seven-coordinated alkyl-
carbonyl formulation in related tungsten compounds: see ref 30.

3302 Organometallics, Vol. 18, No. 17, 1999
Ujaque et al.
the C-C bond breaking process in the molybdenum acyl
complex 8. Moreover, the alkyl-carbonyl structure 13
lies only 2.8 kcal mol-1 above the acyl isomer. Hence,
the reverse reaction, the migratory CO insertion, takes
place with an energy barrier of only 1.2 kcal mol-1. This
value is much lower than those previously calculated
for the migratory CO insertion in Rh(PH3)2Cl(H)(CH3)-
(CO) (27.4 kcal mol-1)8 and in CH3Mn(CO)5 (18 kcal
mol-1).25 The energy barriers for the C-C bond forming
and breaking processes in 8 are of the same order as
those of the methyl rotation. Interestingly, this result
is fully consistent with our previous experimental
observation of a fluxional behavior of Mo(C(O)CH3)(S2-
CNMe2)CO(PMe3)2 leading to the equivalence of the two
carbonyl groups. This process is also likely to go through
a 7-coordinate methyl species, and its free energy of
activation was estimated to be 9.4 kcal mol-1. The
difference between this value and the 4.0 kcal mol-1
computed here is likely associated with the fact that 8e
does not seem the most appropriate isomer for the
fluxionality process, which probably requires more
severe rearrangements and the involvement of higher
energy 7-coordinate isomers. For comparative purposes,
we have also computed the energy profile for the
â-elimination in 8. In this case the Câ-H bond is broken.
Optimization of several points at fixed values of the
Câ-H distances gives a monotonic energy rise: 9.0 kcal
mol-1 at Câ-H ) 1.30 Å, 25.7 kcal mol-1 at Câ-H ) 1.50
Å, 42.5 kcal mol-1 at Câ-H ) 1.75 Å, and 46.3 kcal mol-1
at Câ-H ) 1.90 Å. There is no minimum for the system
in this direction.
F igu r e 5. Energy profile for the lengthening of the C-C
distance in the complex Mo(COCH3)(S2CNH2)(CO)(PH3)2
(8e).
F igu r e 6. Optimized geometry of the Mo(CH3)(S2CNH2)-
(CO)2(PH3)2 (13) complex.
section that an important lengthening of the acyl C-C
bond is associated to the agostic distortion. To establish
the relationship between such a distortion and the C-C
bond breaking process we have studied theoretically
the reaction path for the interconversion between the
â-agostic acyl and the alkyl-carbonyl isomers of 8.
Starting from the minimum 8e, we have calculated
the energy profile for the CO deinsertion by optimizing
the geometry of the system at several fixed values of
the C-C distance. The potential energy curve obtained
is presented in Figure 5. A maximum is reached at C-C
) 2.0 Å. Full geometric optimization starting at the
geometry with C-C ) 2.1 Å leads to a stable seven-
coordinate species 13 with a methyl carbonyl nature
(C‚‚‚C ) 2.357 Å, see Figure 6).
It is clear that the â-agostic interaction between the
terminal methyl group of the acyl ligand and the metal
is able to stabilize a very advanced structure in the
reaction path that leads to the alkyl-carbonyl species.
Furthermore, this interaction assists the C-C bond
breaking process. The small energy difference between
the acyl and alkyl-carbonyl isomers of 8 justifies the
explanation that subtle changes in the basicity of the
metal, or in the donating ability or the steric require-
ments of the ancillary ligands, could significantly shift
the acyl/alkyl-carbonyl equilibrium.
Inspection of the geometric changes along the reaction
coordinate shows that, as expected, the CR-Câ length-
ening is accompanied by CRO shortening, Mo-Câ short-
ening, and closure of the Mo-CR-Câ angle. Only the
geometric parameters of the metal-acyl fragment change
appreciably; the rearrangement of the other ligands in
the coordination sphere of the metal is only minor. In
the methyl-carbonyl structure the Mo-CR and Mo-Câ
distances have values of 1.932 and 2.390 Å, respectively,
and the Mo-CR-Câ angle has closed to 66.9°. The seven-
coordinate species is better described24 as a capped
octahedron, with the carbonyl CRO group as capping
ligand, although quite distorted because of the presence
of the chelating xanthate ligand (Figure 6). We note that
other heptacoordinate species may be attainable with
a low energy barrier for the interconversion. We have
limited the study of the methyl-carbonyl derivatives to
the one directly reached by the breaking of the acyl C-C
bond in the conformation it has in the minimum energy
structure.
Solu tion Beh a vior of Com p ou n d s 2-6. Multi-
nuclear (1H, 13C{1H} and 31P{1H}) NMR studies of
solutions prepared from crystalline samples of 2a -Me
or 2a -P h in C6D5CD3, C6D6, or CD2Cl2 clearly show the
existence of two sets of resonances. By comparison with
the data already available for dihapto and agostic acyls
of composition Mo(C(O)CH3)(S2CX)CO(PMe3)2,6 the spec-
tra can unequivocally be assigned to a mixture of the
two complexes, that is, the agostic Mo(C(O)CH2SiMe2-
R)(S2CNMe2)CO(PMe3)2 (R ) Me, 2a -Me; Ph, 2a -P h )
and the dihapto Mo(η2-C(O)CH2SiMe2R)(S2CNMe2)-
CO(PMe3)2 (R ) Me, 2d -Me; Ph, 2d -P h ) isomers, the
latter being predominant. In the previously mentioned
solvents the 2d /2a ratio does not change appreciably,
either with time or temperature, once the values of 6.6
(R ) Me) and 3.4 (R ) Ph) are attained. The agostic-
to-dihapto acyl isomerization is favorable even at low
temperatures. Thus when a crystalline sample of 2a -
Me is dissolved at -40 °C and the NMR spectra are
recorded at that temperature, a 2d /2a -Me ratio of ca. 1
is observed and this value quickly increases to ca. 6
The most remarkable result of this calculation is the
extremely low energy barrier (4.0 kcal mol-1) found for

Dihaptoacyl and â-Agostic Acyl Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3303
Sch em e 6
Sch em e 8
Sch em e 7
between the isomeric agostic acyl (A), dihaptoacyl (D),
and alkyl carbonyl (C) structures (Scheme 7). The
molybdenum acyl complex Mo(C(O)CH2CMe3)(S2CNMe2)-
CO(PMe3)2 derived from the neopentyl group, i.e., with
R ) CMe3, exists only as the dihapto isomer, while the
corresponding acetyl derivative (R ) H) occurs as the
agostic complex. In this case, however, variable tem-
perature NMR studies have shown9b,c that the two
degenerate ground-state structures possible for this
compound (and its analogues) interconvert rapidly at
room temperature through the undetected alkyl(carbon-
yl) isomer C. When R ) SiMe3, an intermediate situa-
tion is found because, as described above, the solutions
of these complexes contain mixtures of the agostic and
the dihapto acyl isomers. These differences appear to
be mostly steric in origin.28
Some final comments will be devoted to the intercon-
version of structures of type A and D. Obviously, the
proposed mechanism should take into account the
different stereochemistry of the two isomers. Thus,
simple releasing of the â-CH interaction and re-
coordination of the acyl through the oxygen atom would
not be satisfactory. As mentioned both throughout this
article and in previous publications6,9a,29 molybdenum
and tungsten acyl complexes of composition M(C(O)-
CH2R)(S2CX)CO(PMe3)2 can readily undergo deinsertion
to produce the corresponding seven-coordinate alkyl-
carbonyl isomers M(CH2R)(L-L)(CO)2(PMe3)2. Since
seven-coordinate species can easily undergo ligand
rearrangements, a plausible mechanism for the isomer-
ization process could involve CO deinsertion from the
agostic complexes to give the corresponding alkyl-
carbonyls, followed by ligand redistribution and subse-
quent CO insertion to generate the dihaptoacyl com-
pounds with their preferred stereochemistry (Scheme
8). Similar seven-coordinate intermediates have been
proposed for the irreversible transformation of [Mo](η2-
C(O)R)(CNR′) complexes into their isomeric dihapto-
iminoacyl-carbonyl compounds [Mo](η2-C(NR′)R)(CO).29
In addition we have demonstrated that tungsten alkyls
of composition W(CH2R)(S2CNMe2)(CO)2PMe3 show flux-
ional behavior at room temperature due to ligand mo-
bility around the coordination sphere of the metal.30
upon warming to room temperature. Further cooling of
the mixture does not cause any observable variation.
Interestingly, an identical isomeric composition is ob-
tained when a crystalline sample of 2d -Me is dissolved.
It seems therefore clear that the agostic acyls 2a readily
evolve in solution to their dihaptoacyl isomers 2d until
equilibrium is achieved (Scheme 6). This behavior
parallels prior observations for the acetyl compounds
Mo(C(O)CH3)(S2COR′)CO(PMe3)2 (R′ ) Me, Et, i-Pr)
which crystallize as the agostic form but exist in solution
mostly as mixtures of the A and D structures (Scheme
7). As opposed to the behavior of complexes 2, compound
3 exists in solution predominantly as the dihapto
isomer, with only minor signals due to the â agostic
species.
In contrast with their acetyl-dithiocarbamate coun-
terparts, the xanthate compounds 5 and 6 occur in
solution in the form of the dihaptoacyl isomer only. The
same applies to the pyrrolyl dithiocarbamate derivative
4. It is worth mentioning at this point that the dihap-
toacyls 2d -6d display dynamic properties in solution.
For example, the two 31P nuclei of complex 2a -Me
2
appear at -90 °C as an AB spin system with J PP
)
115 Hz. This pattern of lines coallesces upon warming
and finally gives a singlet at room temperature. As
described previously, this fluxional behavior is due to a
libration motion of the η2-acyl moiety, which creates an
effective plane of symmetry.26
With regard to the relative stability of the isomeric
acyl structures, the silyl derivatives 2-6 follow the
trend already described for the corresponding acetyl
complexes (R ) H). For the latter compounds, strongly
donating chelating ligands favor the agostic coordina-
tion, whereas less electron-releasing ligands enhance
the dihapto structure. Thus, in our case the dimethyl
dithiocarbamate derivatives 2 exist in solution as
isomeric mixtures, while the complex 4d -Me, which
contains the less donating pyrrol dithiocarbamate,27
occurs in solution exclusively as the dihapto isomer.
The availability of the silyl derivatives 2-6 allows the
further establishment of a sequence of reactivity con-
cerning the tendency of compounds of composition Mo-
(C(O)CH2R)(S2CNMe2)CO(PMe3)2 to undergo exchange
Con clu sion s
Small energy differences have been observed between
the agostic and the dihapto coordination modes of the
(29) Pizzano, A.; Sa´nchez, L.; Altmann, M.; Monge, A.; Ruiz, C.;
Carmona, E. J . Am. Chem. Soc. 1995, 117, 1759.
(30) Carmona, E.; Contreras, L.; Poveda, M. L.; Sa´nchez, L. J .;
Atwood, J . L.; Rogers, R. D. Organometallics 1991, 10, 61.

3304 Organometallics, Vol. 18, No. 17, 1999
Ujaque et al.
NMR (20 °C, CD3COCD3): δ -1.7 (s, SiMe2), 15.5 (t, J CP ) 10
Hz, PMe3), 20.3 (d, J CP ) 25 Hz, PMe3), 39.4 (s, CH2), 128.5,
129.9, 134.3, 138.7 (s, Ph), 251.4 (dt, J CP ) 31, 10 Hz, CO),
271.4 (q, J CP ) 12 Hz, COR).
P r ep a r a tion of Mo(C(O)CH2SiMe2R)(L-L)CO(P Me3)2
Com p lexes [L-L ) S2CNMe2; R ) Me (2a -Me, 2d -Me), P h
(2a -P h ); S2CN(i-P r )2 R ) Me (3d -Me); S2CNC4H4, R ) Me
(4d -Me); S2CO-i-P r , R ) Me (5d -Me), P h (5d -P h ); S2CO-
t-Bu , R) Me (6d -Me)]. All these compounds were prepared
in a similar way by treating the chloroacyl Mo(η2-C(O)CH2-
SiMe2R)Cl(CO)(PMe3)3 with 1.2 equivalents of the sodium or
potassium salts of the bidentate ligands in diethyl ether or
tetrahydrofuran at room temperature. The synthesis of 2-P h
is presented as representative example.
A crystalline sample of 1-P h (0.56 g, 1.0 mmol) was
dissolved in 40 mL of THF and cooled in an ice bath.
Anhydrous NaS2CNMe2 (0.17 g, 1.2 mmol) was added to the
reaction mixture producing a color change from deep red to
orange. The resulting suspension was stirred at room tem-
perature for 2h at 0-10 °C and then the volatiles were
removed and the resulting residue was extracted with a 1:1
petroleum ether:diethyl ether mixture. Centrifugation followed
by concentration of the liquor and cooling at -10 °C afforded
2a -P h as orange crystals in 80% yield. From the corresponding
chloro-acyl complexes and the appropriate dithiocarbamate or
xanthate salts the following compounds were obtained in
similar yields. These compounds were isolated as orange-red
crystalline solids, except 4d -Me which is purple, by crystal-
lization of their solutions in petroleum ether-diethyl ether
mixtures.
acyl ligand in Mo(C(O)CH2SiMe2R)(S2CX)(CO)(PMe3)2
complexes. As previously observed in the analogous
acetyl derivatives, strongly electron-releasing ligands
appear to favor the agostic interaction, but the prefer-
ence is not very pronounced; small changes in the steric
effects around the metal, or in the nature of the
co-ligands, can unbalance the situation in favor of any
of the two isomers.
Ab initio calculations show that the agostic interaction
between the acyl ligand and the Mo atom observed in
the Mo(C(O)CH2R)(S2CNH2)(CO(PH3)2 models is sig-
nificant and lie well in the range observed for other
M-H-Câ nonclassical interactions that exist in metal
alkyls. Separation between the M-H and M-Câ con-
tributions indicates a stronger importance of the latter
(9.8 vs 2.9 kcal mol-1), thus allowing the rotation around
the CR-Câ bond without cleavage of the agostic interac-
tion. By comparison with the monohapto situation, the
interaction between the C(O)CH3 fragment and the
metal center brings the Mo and Câ atoms closer,
shortens the C-O and Mo-CR bonds, and lengthens the
CR-Câ bond. This scenario can be viewed as an ad-
vanced intermediate along the reaction path of the CO
deinsertion process in a monohapto acyl. An energy
profile of the agostic acyl to the alkyl-carbonyl carried
out by optimization of the molecular geometries at
different CR-Câ bond distances indicates a low energy
barrier for the deinsertion process, in excellent accord
Although some spectroscopic data for the complex Mo(C-
(O)CH2SiMe3)(S2CNMe2)(CO)(PMe3)2 (2a -Me and 2d -Me) have
already been reported, for the sake of completeness, their most
relevant spectroscopic data are indicated below.
with the behavior observed in solution for the Mo(C(O)-
CH3)(S2CX)CO(PMe3)2 complexes.
Mo(C(O)CH2SiMe3)(S2CNMe2)CO(P Me3)2 (2a -Me). IR
(Nujol mull cm-1): 1760 (s), 1615 (m) (υCO), 1515 (m) (υCN). 1H
NMR (20 °C, C6D6): δ 0.28 (s, SiMe3), 1.42 (d, J HP ) 8.8 Hz,
PMe3), 2.56 (brs, CH2), 2.60 (s, NMe2). 31P{1H} NMR (20 °C,
C6D6): δ 26.7 (brs). 13C{1H} NMR (20 °C, C6D6): δ 0.5 (s,
SiMe3), 15.8 (d, J CP ) 28 Hz, PMe3), 38.6 (s, NMe2). Other
carbon resonances were not observed due probably to the low
concentration of this compound in solution.
Mo(η2-C(O)CH2SiMe3)(S2CNMe2)(CO)(P Me3)2 (2d -Me).
IR (Nujol mull cm-1): 1750 (s), 1490 (m) (υCO), 1515 (m) (υCN).
1H NMR (20 °C, C6D6): δ 0.14 (s, SiMe3), 1.56 (t, J HP ) 3.6
Hz, PMe3), 3.08 (s, CH2), 2.77, 2.83 (s, NMe2). 31P{1H} NMR
(20 °C, C6D6): δ 2.5 (s). 13C{1H} NMR (20 °C, C6D6): δ -0.7
(s, SiMe3), 15.7 (t, J CP ) 11 Hz, PMe3), 37.7 (s, CH2), 39.2, 39.8
(s, NMe2), 210.3 (s, S2C), 238.2 (t, J CP ) 16 Hz, CO), 275.5 (t,
J CP ) 16 Hz, COR).
Exp er im en ta l Section
Microanalyses were carried out either by the Analytical
Service of the University of Seville or by Pascher Micro-
analytical Laboratory, Remagen (Germany). IR spectra were
recorded both as Nujol mulls or in an appropriate solvent on
a Perkin-Elmer 684 instrument. 1H, 13C, and 31P NMR spectra
were acquired on a Varian XL-200 spectrometer. 31P{1H} NMR
shifts were referenced to external 85% H3PO4, while 13C{1H}
1
and H shifts were referenced to the residual protio signals of
deuterated solvents. All data are reported in ppm downfield
from MeSi4. All manipulations were performed under oxygen-
free nitrogen or argon following conventional Schlenk tech-
niques or by using a Vacuum Atmospheres drybox. Solvents
were dried under an appropriate desiccant, deoxygenated, and
freshly distilled prior to use. Mo(η2-C(O)CH2SiMe3)Cl(CO)-
6a
6c
Mo(C(O)CH2SiMe2P h )(S2CNMe2)(CO)(P Me3)2 (2a -P h ).
Anal. Calcd for C20H37NO2P2S2SiMo: C, 41.9; H, 6.4; N 2.4.
Found: C, 42.2; H, 6.8; N, 1.6. IR (Nujol mull cm-1): 1785 (s),
(PMe3)3 and ZnCl2(PMe3)2 were prepared as described
previously. Dithiocarbamate and xanthate salts as well as
PMe3 were synthesized according to literature procedures.
Sodium dimethyldithiocarbamate was dried by heating at 90
°C at 10-2 Torr for 3 days. All other reagents were purchased
from commercial suppliers.
1
1615 (w) (υCO). H NMR (20 °C, C6D6): δ 0.58 (s, SiMe2), 1.37
(d, J HP ) 8.9 Hz, PMe3), 2.58 (s, N(CH3)2), 2.69 (brs, CH2), 7.25,
7.70 (m, Ph). 31P{1H} NMR (20 °C, C6D6): δ 27.9 (brs). 13C{1H}
NMR (20 °C, C6D6): δ -0.8 (s, SiMe2), 15.5 (d, J CP ) 28 Hz,
PMe3), 38.8 (s, N(CH3)2), 128.3, 129.0, 134.5, 140.2 (s, Ph).
Other carbon resonances were not observed, probably due to
the low concentration of this compound in solution.
Mo(η2-C(O)CH2SiMe2P h )Cl(CO)(P Me3)3 (1-P h ). Over a
yellow suspension of MoCl2(CO)2(PMe3)3 (0.45 g, 1.0 mmol) and
ZnCl2(PMe3)2 in diethyl ether, 1.2 mmol of MgClCH2SiMe2Ph
(1.4 mL, 0.86 M solution) was slowly added. On addition color
progresively turns orange-red. The mixture is stirred for 5 h
and volatiles removed under vacuum. Extraction in a 1:1
petroleum ether:diethyl ether mixture followed by centrifuga-
tion and concentration of the resulting red liquor afforded 0.37
g of 1-P h as red crystals (65% yield). Anal. Calcd for C20H40O2P3-
SiClMo: C, 42.5; H, 7.1. Found: C, 42.8; H, 7.4. IR (Nujol mull
cm-1): 1794 (s), 1507 (m) (υCO). 1H NMR (20 °C, C6D6): δ 0.52
(s, SiMe2), 1.04 (t, J HP ) 3.0 Hz, PMe3), 1.20 (d, J HP ) 8.0 Hz,
2 PMe3), 3.20 (s, CH2), 7.20, 7.58 (m, Ph). 31P{1H} NMR (20
°C, C6D6): δ 23.3 (d, J PP ) 23 Hz), 37.5 (t, J PP ) 23 Hz). 13C{1H}
Mo(η2-C(O)CH 2SiMe2P h )(S2CNMe2)(CO)(P Me3)2 (2d -
1
P h ). IR (THF cm-1): 1775 (s) (υCO). H NMR (20 °C, C6D6): δ
0.46 (s, SiMe3), 1.52 (t, J HP ) 3.6 Hz, PMe3), 2.77, 2.82 (s,
N(CH3)2), 3.28 (s, CH2), 7.19, 7.50 (m, Ph). 31P{1H} NMR (20
°C, C6D6): δ 3.6 (s). 13C{1H} NMR (20 °C, C6D6): δ -1.8 (s,
SiMe2), 15.9 (t, J CP ) 11 Hz, PMe3), 37.1 (s, CH2), 39.5, 40.1
(s, N(CH3)2), 128.2, 129.6, 134.1, 138.1 (s, Ph), 210.5 (s, S2C),
238.7 (t, J CP ) 14 Hz, CO), 275.9 (t, J CP ) 16 Hz, COR).
Mo(η2-C(O)CH2SiMe3)(S2CN-i-P r 2)(CO)(P Me3)2 (3d -Me).
Anal. Calcd for C19H43NO2P2S2SiMo: C, 40.2; H, 7.6, N, 2.5.

Dihaptoacyl and â-Agostic Acyl Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3305
Found: C, 40.5; H, 7.6, N, 2.7. IR (Nujol mull cm-1): 1760 (s),
1475 (m) (υCO), 1490 (m) (υCN). H NMR (20 °C, C6D6): δ 0.12
(s, SiMe3), 1.19 (brs, CH(CH3)2), 1.56 (t, J HP ) 3.5 Hz, PMe3),
3.05 (s, CH2), 4.60 (brs, CH). 31P{1H} NMR (20 °C, C6D6): δ
4.8 (s). 13C{1H} NMR (20 °C, C6D6): δ -0.6 (s, SiMe3), 15.7 (t,
J CP ) 11 Hz, PMe3), 19.3, 19.7 (s, CH(CH3)2), 38.0 (s, CH2),
MP2 geometries in agostic complexes are in excellent agree-
ment with experimental neutron diffraction data.34 Two dif-
ferent basis sets were used. Basis set I was used in most
calculations, including all geometry optimizations. In basis set
I, the basis for molybdenum, phosphorus, sulfur, and silicon
was that associated with the pseudopotential32 with standard
LANL2DZ contraction schemes.31 For the other atoms the
6-31G basis set was used.35
Basis set II, much more extended than basis set I, was used
to check the basis set dependence of the computed energies
through single-point calculations on structures frozen at the
geometry obtained with basis set I. Basis set II was obtained
from basis set I by adding a shell of polarization functions to
all atoms, i.e., a p shell on hydrogen atoms35a,b a d shell on
carbon, nitrogen, oxygen, phosphorus, and sulfur atoms,35a-d
and an f shell on the molybdenum atom.35e
All the geometric parameters were optimized to find the
most stable structure for each compound. To evaluate the
energy differences between agostic and anagostic geometries,
optimizations were performed for the least stable structures
keeping fixed the value of the Mo-CR-Câ at 109.4° in 12 and
120° in 8. Symmetry restrictions (Cs) were introduced in the
optimizations when possible.
X-r a y Str u ctu r e Deter m in a tion of 2a -Me. A summary
of the fundamental crystal data is given in Table 1. A crystal
of 2a -Me was coated with an epoxy resin and mounted in a
Kappa diffractometer. The cell dimensions were refined by
least-squares fitting of the values of 25 reflections. The
intensities were corrected for Lorentz and polarization effects.
Scattering factors for neutral atoms and anomalous dispersion
corrections for Mo, P, and S were taken from ref 36. The
structures were solved by Patterson and Fourier methods. An
empirical absorption correction37 was applied at the end of the
isotropic refinement. Final mixed refinement with unit weights
and fixed isotropic factors and coordinates for H atoms, except
for H(31) and H(32) for which the corresponding coordinates
were refined, led to final values of R ) 0.037 and Rw ) 0.055.
The final values of the positional parameters are given in Table
2. Most of the calculations were carried out with the X-ray 80
system.38
1
50.0, 50.8 (s, CH), 209.7 (t, J CP ) 7 Hz, S2C), 237.9 (t, J CP
14 Hz, CO), 276.0 (t, J CP ) 16 Hz, COR).
)
Mo(η2-C(O)CH2SiMe3)(S2CNC4H4)(CO)(P Me3)2 (4d -Me).
Anal. Calcd for C17H33NO2P2S2SiMo: C, 38.3; H, 6.2, N, 2.6.
Found: C, 37.8; H, 5.9, N, 2.6. IR (Nujol mull cm-1): 1780 (s),
1
1465 (m) (υCO), 1485 (m) (υCN). H NMR (20 °C, C6D6): δ 0.08
(s, SiMe3), 1.39 (t, J HP ) 3.7 Hz, PMe3), 3.00 (s, CH2), 6.10,
7.80 (t, CH). 31P{1H} NMR (20 °C, C6D6): δ 5.8 (s). 13C{1H}
NMR (20 °C, C6D6): δ -0.6 (s, SiMe3), 15.8 (t, J CP ) 12 Hz,
PMe3), 37.8 (s, CH2), 112.9, 117.9 (s, CH), 208.7 (t, J CP ) 8
Hz, S2C), 237.1 (t, J CP ) 15 Hz, CO), 276.0 (t, J CP ) 16 Hz,
COR).
Mo(η2-C(O)CH2SiMe3)(S2CO-i-P r )(CO)(P Me3)2 (5d -Me).
Anal. Calcd for C16H36NO2P2S2SiMo: C, 35.5; H, 6.8. Found:
C, 35.3; H, 6.8. IR (Nujol mull cm-1): 1795 (s), 1485 (m) (υCO).
1H NMR (20 °C, C6D6): δ 0.21 (s, SiMe3), 1.07 (d, J HH ) 6.2
Hz, CH(CH3)2), 1.44 (t, J HP ) 3.6 Hz, PMe3), 3.02 (s, CH2), 5.34
(h, J HH ) 6.2 Hz, CH). 31P{1H} NMR (20 °C, C6D6): δ 5.2 (s).
13C{1H} NMR (20 °C, C6D6): δ 1.0 (s, SiMe3), 15.7 (t, J CP ) 11
Hz, PMe3), 21.4 (s, CH(CH3)2), 37.8 (s, CH2), 74.5 (s, CH), 222.3
(t, J CP ) 6 Hz, S2C), 236.7 (t, J CP ) 15 Hz, CO), 276.7 (t, J CP
) 16 Hz, COR).
Mo(η2-C(O)CH 2SiMe2P h )(S2CO-i-P r )(CO)(P Me3)2(5d -
P h ). Satisfactory elemental analysis was precluded by its
thermal instability. IR (Nujol mull cm-1): 1800 (s), 1465 (m)
1
(υCO). H NMR (20 °C, C6D6): δ 0.40 (s, SiMe3), 1.10 (d, J HH
)
6.2 Hz, CH(CH3)2), 1.44 (t, J HP ) 3.7 Hz, PMe3), 3.23 (s, CH2),
5.40 (h, J HH ) 6.2 Hz, CH), 7.20, 7.46 (m, Ph). 31P{1H} NMR
(20 °C, THF/CD3COCD3): δ 9.7 (s). 13C{1H} NMR (20 °C,
C6D6): δ -2.3 (s, SiMe2), 15.5 (t, J CP ) 11 Hz, PMe3), 21.3 (s,
CH(CH3)2), 36.9 (s, CH2), 74.5 (s, CH), 127.9, 129.4, 133.7,
137.4 (s, Ph), 222.3 (t, J CP ) 7.8 Hz, S2C), 237.1 (t, J CP ) 13
Hz, CO), 276.6 (t, J CP ) 15 Hz, COR).
Mo(η2-C(O)CH2SiMe3)(S2CO-t-Bu )(CO)(P Me3)2 (6d -Me).
Anal. Calcd for C17H38NO2P2S2SiMo: C, 37.8; H, 7.0. Found:
C, 38.0; H, 7.1. IR (Nujol mull cm-1): 1760 (s), 1485 (m) (υCO).
1H NMR (20 °C, C6D6): δ 0.12 (s, SiMe3), 1.56 (t, J HP ) 3.7
Hz, PMe3), 1.63 (s, CMe3), 3.20 (s, CH2). 31P{1H} NMR (20 °C,
C6D6): δ 5.4 (s). 13C{1H} NMR (20 °C, C6D6): δ -0.7 (s, SiMe3),
15.7 (t, J CP ) 11 Hz, PMe3), 21.4 (s, C(CH3)3), 37.8 (s, CH2),
86.9 (s, C(CH3)3), 222.1 (t, J CP ) 6 Hz, S2C), 236.9 (t, J CP ) 14
Hz, CO), 276.5 (t, J CP ) 15 Hz, COR).
Com p u ta tion a l Deta ils. All calculations were performed
with the GAUSSIAN 94 series of programs.31 A molecular
orbital ab initio method with introduction of correlation energy
through the Moller-Plesset (MP) perturbation approach,32
excluding excitations concerning the lowest energy electrons
(frozen core approach), was applied. Effective core potentials
(ECP) were used to represent the 28 innermost electrons of
the metal atom33a as well as the 10 electron core of the
phosphorus, sulfur, and silicon atoms.33b Geometry optimiza-
tions were carried out at the second level of the Møller-Plesset
theory (MP2) with a basis set of valence double-ú quality for
all the atoms. It has already been shown that the computed
Ack n ow led gm en t. We acknowledge financial sup-
port from the DGES (Project No PB95-0639-CO2-01)
and the DGICYT (grant No. PB94-1436). We also
acknowledge J unta de Andaluc´ıa for the award of
research fellowships. The use of computational facilities
of the Centre de Supercomputacio´ i Comunicacions de
Catalunya (C4) is gratefully appreciated. Prof. O. Eisen-
stein (Montpellier) is acknowledged for fruitful discus-
sions.
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
and thermal parameters for 2a -Me. This material is available
free of charge via the Internet at http://pubs.acs.org.
OM9807248
(34) Weiss, H.; Haase, F.; Ahlrichs, R. Chem. Phys. Lett. 1992, 194,
492.
(35) (a) Hehre, W. J .; Ditchfield, R.; Pople, J . A. J . Chem. Phys. 1972,
56, 2257. (b) Hariharan, P. C.; Pople, J . A. Theor. Chim. Acta 1973,
28, 213. (c) Francl, M. M.; Pietro, W. J .; Hehre, W. J .; Binkley, J . S.;
Gordon, M. S.; Defrees, D. J .; Pople, J . A. J . Chem. Phys. 1982, 77,
3654. (d) Hullwarth, A.; Buhhme, M.; Dapprich, S.; Ehlers, A. W.;
Gobbi, A.; J onas, V.; Kuhler, K. F.; Stegmann, R.; Veldkamp, A.;
Frenking, G. Chem. Phys. Lett. 1993, 208, 237. (e) Ehlers, A. W.;
Buhme, M.; Dapprich, S.; Gobbi, A.; Hullwarth, A.; J onas, V.; Kuhler,
K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys. Lett.
1993, 208, 111.
(36) International Tables for X-Ray Crystallography; Kynoch
Press: Birmingham, UK, 1974; Vol. IV, pp 72-98.
(37) Walker, N.; Stuart, D. Acta Crystallogr. 1983, A39, 158.
(38) Stewart, J . M. The X-ray 80 System; Computer Science Center,
University of Maryland: College Park, MD, 1985.
(31) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T. A.; Petersson,
G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Laham, M. A.;
Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Cioslowski, J .;
Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andrs, J . L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.; De Frees, D. J .; Baker, J .;
Stewart, J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. Gaussian
94; Gaussian Inc.: Pittsburgh, PA, 1995.
(32) Moller, C.; Plesset, M. S. Phys. Rev. 1934, 46, 618.
(33) (a) Hay, P. J .; Wadt, W. R. J . Chem. Phys. 1985, 82, 299. (b)
Wadt, W. R.; Hay, P. J . J . Chem. Phys. 1985, 82, 284.