Chemistry of unsaturated group 6 metal complexes with bridging hydroxy- and methoxycarbyne ligands. 2. Synthesis, structure, and bonding of 32- and 34-electron complexes

Article abstract of DOI:10.1021/om700735g

The reaction of the 30-electron methoxycarbyne complex [Mo ₂Cp₂(μ-COMe)(μ-PCy₂)(μ-CO)] with carbon monoxide gives quantitatively the electron-precise derivative [Mo ₂Cp₂(μ-COMe)(μ-PCy₂)(CO)₃], in which the methoxycarbyne ligand adopts a highly asymmetric semibridging coordination mode. This tricarbonyl species dissociates one of its CO ligands under thermal conditions (333 K) to give the 32-electron complex [Mo ₂Cp₂(μ-COMe)(μ-PCy₂)(CO)₂] in high yield. The structures and bonding of both complexes, as well as that of the isostructural carboxycarbyne complex [Mo₂Cp₂{μ- C(CO₂Me)}(μ-PCy₂)-(CO)₂] were studied by means of density functional theory. These calculations allow us to describe the intermetallic interaction in the edge-sharing bioctahedral dicarbonyls by a configuration of the type δ²σ²π ²-(δ*)², with the δ* orbitais being involved in π back-bonding to the carbonyl ligands, and the δ orbitals significantly delocalized over the Mo₂C(carbyne) triangle, then becoming the π-bonding component of the metal-carbyne bond. For these complexes, an atoms-in-molecules (AIM) analysis allows us to locate the corresponding intermetallic bond critical points, these exhibiting relatively high values of the electron density (ca. 0.43 e A^-3). Both the MO and AIM analysis of these dicarbonyl complexes suggest that the binding of the methoxy- and carboxycarbyne ligands to the dimetal center is quite similar.

Full text of DOI:10.1021/om700735g

5912
Organometallics 2007, 26, 5912-5921
Chemistry of Unsaturated Group 6 Metal Complexes with Bridging
Hydroxy- and Methoxycarbyne Ligands. 2. Synthesis, Structure, and
Bonding of 32- and 34-Electron Complexes§
M. Esther Garc´ıa,† Daniel Garc´ıa-Vivo´,† Miguel A. Ruiz,*,† Santiago Alvarez,‡ and
Gabriel Aullo´n‡
Departamento de Qu´ımica Orga´nica e Inorga´nica/IUQOEM, UniVersidad de OViedo, E-33071 OViedo,
Spain, and Departamento de Qu´ımica Inorga´nica, UniVersidad de Barcelona, E-08028 Barcelona, Spain
ReceiVed July 20, 2007
The reaction of the 30-electron methoxycarbyne complex [Mo2Cp2(µ-COMe)(µ-PCy2)(µ-CO)] with
carbon monoxide gives quantitatively the electron-precise derivative [Mo2Cp2(µ-COMe)(µ-PCy2)(CO)3],
in which the methoxycarbyne ligand adopts a highly asymmetric semibridging coordination mode. This
tricarbonyl species dissociates one of its CO ligands under thermal conditions (333 K) to give the 32-
electron complex [Mo2Cp2(µ-COMe)(µ-PCy2)(CO)2] in high yield. The structures and bonding of both
complexes, as well as that of the isostructural carboxycarbyne complex [Mo2Cp2{µ-C(CO2Me)}(µ-PCy2)-
(CO)2] were studied by means of density functional theory. These calculations allow us to describe the
intermetallic interaction in the edge-sharing bioctahedral dicarbonyls by a configuration of the type δ2σ2π2-
(δ*)2, with the δ* orbitals being involved in π back-bonding to the carbonyl ligands, and the δ orbitals
significantly delocalized over the Mo2C(carbyne) triangle, then becoming the π-bonding component of
the metal-carbyne bond. For these complexes, an atoms-in-molecules (AIM) analysis allows us to locate
the corresponding intermetallic bond critical points, these exhibiting relatively high values of the electron
density (ca. 0.43 e Å-3). Both the MO and AIM analysis of these dicarbonyl complexes suggest that the
binding of the methoxy- and carboxycarbyne ligands to the dimetal center is quite similar.
which in turn retains some multiplicity in its C-O bond
(originally a double bond in the corresponding carbonyl-bridged
precursor).
Introduction
In the previous part of this series we have shown that a variety
of cationic hydroxy- and methoxycarbyne complexes of the type
[M2Cp2(µ-COR)(µ-PR′2)2]BF4 (M ) W, R ) Me, R′ ) Ph; M
) Mo, R ) H, Me, R′ ) Et) and [Mo2Cp2(µ-COR)(µ-COR′)-
(µ-PCy2)]BF4 (R ) R′ ) Me; R ) Me, R′ ) H, Et) can be
easily obtained by the reaction of the corresponding neutral
monocarbonyl precursors with either HBF4‚OEt2 or [Me3O]BF4.1
We also reported there a complete theoretical study of these
systems, proving the usefulness of density functional theory
(DFT)2 methods for the interpretation of the electronic structure
and bonding in these highly unsaturated binuclear complexes.
Indeed, the use of the hybrid functional B3LYP for these face-
sharing bioctahedral complexes not only leads to an excellent
agreement between optimized and experimentally determined
geometries, but also allow us to describe the intermetallic bond
as one of order three following from the configuration σ2δ4.
We should note, however, that in these complexes the δ-bonding
orbitals are delocalized over the Mo2C(carbyne) triangle and
hence constitute the π-bonding component of the metal-carbyne
interaction. Therefore, a reduction of the direct overlap between
the metal atoms occurs as a consequence of the formation of
the metal-carbon multiple bond of the alkoxycarbyne ligand,
We have noted previously that the study of the reactivity of
hydroxy- and alkoxycarbyne ligands at unsaturated binuclear
centers might be of interest not only because of the unusual
transformations that can be induced,3 but also in the context of
important academic and industrial reactions like the metal-
catalyzed hydrogenation of CO.4 We thus initiated a systematic
study of the reactivity of the above-mentioned 30-electron
hydroxy- and alkoxycarbyne complexes, the results of which
will be reported separately. Besides, to expand the range of
reactive substrates at hand, we were also interested in preparing
related alkoxycarbyne complexes with metal-metal bond orders
lower than three, so we could examine the effect of these weaker
intermetallic interactions upon the metal-carbyne binding and
the reactivity associated with it. In this paper we report full
details of our study of the 32- and 34-electron complexes derived
from the addition of one or two CO ligands to the 30-electron
complex [Mo2Cp2(µ-COMe)(µ-PCy2)(µ-CO)] (1).5 To better
understand the effects of the added ligands upon the metal-
metal and metal-carbyne bonding we have analyzed the
(3) (a) Alvarez, C. M.; Alvarez, M. A.; Garc´ıa, M. E.; Garc´ıa-Vivo´, D.;
Ruiz, M. A. Organometallics 2005, 24, 4122. (b) Alvarez, M. A.; Garc´ıa,
M. E.; Riera, V.; Ruiz, M. A.; Robert, F. Organometallics 2002, 21, 1177.
(c) Alvarez, M. A.; Garc´ıa, M. E.; Riera, V.; Ruiz, M. A. Organometallics
1999, 18, 634. (d) Alvarez, M. A.; Bois, C.; Garc´ıa, M. E.; Riera, V.; Ruiz,
M. A. Angew. Chem., Int. Ed. Engl. 1996, 35, 102. (e) Garc´ıa, M. E.; Riera,
V.; Rueda, M. T.; Ruiz, M. A.; Halut, S. J. Am. Chem. Soc. 1999, 121,
1960.
(4) (a) Maitlis, P. M. J. Mol. Catal. A 2003, 204-205, 55. (b) Campbell,
I. M. Catalysis at Surfaces; Chapman and Hall: New York, 1988.
(5) Garc´ıa, M. E.; Melo´n, S.; Ramos, A.; Riera, V.; Ruiz, M. A.; Belletti
D.; Graiff, C.; Tiripicchio, A. Organometallics 2003, 22, 1983.
* To whom correspondence should be addressed. E-mail: mara@uniovi.es.
§ This paper is dedicated to the memory of Dr. Lorenzo Pueyo,
distinguished professor of physical chemistry at the Universidad de Oviedo.
† Universidad de Oviedo.
‡ Universidad de Barcelona.
(1) Garc´ıa, M. E.; Garc´ıa-Vivo´, D.; Ruiz, M. A.; Alvarez, S.; Aullo´n,
G. Organometallics 2007, 26, 4930.
(2) (a) Koch, W.; Holthausen, M. C. A Chemist’s Guide to Density
Functional Theory, 2nd ed.; Wiley-VCH: Weinheim, Germany, 2002. (b)
Ziegler, T. Chem. ReV. 1991, 91, 651.
10.1021/om700735g CCC: $37.00 © 2007 American Chemical Society
Publication on Web 10/24/2007

Synthesis of 32- and 34-Electron Complexes
Organometallics, Vol. 26, No. 24, 2007 5913
Chart 1
of the C-O stretches depending on the relative arrangement of
the CpMo(CO)x fragments (cis or trans with respect to the
average Mo2PC plane; see Chart 1). As can be appreciated in
the Figure 2, the experimental pattern of the IR spectrum of
compound 2 in the C-O stretching region is in excellent
agreement with that calculated for the cisoid structure, and it is
significantly different from that calculated for trans-2, thus
excluding the presence of significant amounts of the latter isomer
in the solutions of this compound. In agreement with such an
asymmetric structure, the CO ligands of 2 give rise to three
13C NMR resonances in the terminal region, whereas the Cp
ligands exhibit two distinct chemical environments. Unexpect-
edly, the bridging carbyne atom displays a relatively low
chemical shift (300.7 ppm), when compared to those found for
our previous 30-electron alkoxy- or hydroxycarbyne complexes
(ca. 365 ppm),1 or other alkoxycarbyne-bridged complexes
described in the literature.3,5,9 This feature can be mainly
attributed to the highly asymmetric coordination of the alkoxy-
carbyne ligand in 2, as predicted by the DFT calculations (see
later), yielding Mo-C(carbyne) distances differing by ca. 0.45
Å (Table 2), with the short distance approaching the value found
in the only terminal methoxycarbyne complex reported so far,
[WTp′(COMe)(CO)2] (Tp′ ) HB(N2C3Me2H)3),10 which is also
characterized by a quite shielded carbyne resonance (δC 228.2
ppm). As for the phosphide bridge, we note that it gives rise to
a 31P NMR resonance with a chemical shift (219.7 ppm) similar
to those measured for other electron-precise dimolybdenum
phosphide-bridged complexes such as the tricarbonyl [Mo2Cp2-
(µ-PEt2)2(CO)3] (194.7 ppm) or the hydride-phosphide complex
[Mo2Cp2(µ-H)(µ-PCy2)(CO)4] (218.8 ppm).7
Decarbonylation Reaction of Compound 2. Tricarbonyl
complexes of the group 6 elements such as 2 are rare species,
placed midway between the usually more stable tetracarbonylic
or dicarbonylic structures ([M2Cp2(µ-X)(µ-Y)(CO)n], n ) 2, 4).
For instance, the related complex [Mo2Cp2(µ-PEt2)2(CO)3]
experiences a spontaneous decarbonylation at room temperature
to give the trans-dicarbonyl complex trans-[Mo2Cp2(µ-PEt2)2-
(CO)2] in high yield.7 Thus, we decided to study the decarbo-
nylation reactions of 2 as a potential route to the corresponding
unsaturated dicarbonyl complex. Photolysis of a toluene solution
of compound 2 with visible-UV light induces rapidly a double
decarbonylation process whereby compound 1 is rapidly and
cleanly regenerated, with no intermediates being detected.
However, when toluene solutions of 2 are heated at 333 K, a
single decarbonylation occurs progressively (45 min) to give
the trans-dicarbonyl complex [Mo2Cp2(µ-COMe)(µ-PCy2)-
(CO)2] (3) in high yields (Chart 1). The process is reversible,
so the tricarbonyl 2 can be regenerated in ca. 6 h at room
temperature by stirring a toluene solution of 3 under a slight
CO overpressure (pCO ca. 4 atm). Finally, we should point out
that no further decarbonylation can be induced thermally.
structure and bonding within the new alkoxycarbyne complexes
using DFT methods and subsequent MO and AIM analysis,
following the scheme used previously with the 30-electron
alkoxycarbyne complexes.1 A parallel study of the carboxycar-
byne complex [Mo2Cp2{µ-C(CO2Me)}(µ-PCy2)(CO)2] also has
been performed, this allowing us to compare the metal-carbon
binding of the methoxycarbyne ligand with that of a non-alkoxy
carbyne ligand. With the current study we thus complete a wide
analysis of the structure and bonding of alkoxycarbyne-bridged
complexes of the group 6 elements having intermetallic bond
orders ranging formally from 3 to 1.
Results and Discussion
Reaction of Compound 1 with Carbon Monoxide. Orga-
nometallic compounds having unsaturated dimetal centers are
usually reactive toward carbon monoxide,6 to give the corre-
sponding addition products. Thus, it is not surprising that
compound 1 reacts with CO when its solutions are stirred
overnight under a slight overpressure (pCO ca. 4 atm) to give
the electron-precise tricarbonyl derivative [Mo2Cp2(µ-COMe)-
(µ-PCy2)(CO)3] (2) in quantitative yield (Chart 1), although no
intermediates are detected in this somewhat slow process. The
tricarbonylic nature of compound 2 is clearly denoted by the
presence of three strong C-O stretching bands in its IR
spectrum (Table 1). The frequencies and intensities of these
bands are very similar to those found for the isoelectronic
complexes [Mo2Cp2(µ-PEt2)2(CO)3]7 and [Mo2(µ-η5:η5-C5H4-
SiMe2C5H4)(µ-PMe2)2(CO)3],8 in which the three carbonyl
ligands are placed on the same side of the average plane defined
by the metal and phosphorus atoms, as confirmed for the latter
compound through an X-ray study. Therefore, a similar cisoid
geometry is assumed for compound 2. Moreover, we have
shown in the previous part of this series that the DFT-calculated
frequencies and intensities for the C-O stretches of the carbonyl
and methoxycarbyne ligands in our complexes are in good
agreement with the experimental values measured by IR
spectroscopy.1 Therefore we can confidently use calculated
values of intensity and frequency to distinguish between different
isomeric structures. In particular, as we will discuss later in the
context of the DFT calculations of compound 2, the interaction
between the M(CO) and M(CO)2 oscillators in this molecule is
strong enough to give rise to two substantially different patterns
(9) (a) Hersh, W. H.; Fong, R. H. Organometallics 2005, 24, 4179. (b)
Peters, J. C.; Odon, A. L.; Cummins, C. C. Chem. Commun. 1997, 1995.
(c) Bronk, B. S.; Protasiewicz, J. D.; Pence, L. E.; Lippard, S. J.
Organometallics 1995, 14, 2177. (d) Seyferth, D.; Ruschke, D. P.; Davis,
W. H. Organometallics 1994, 13, 4695. (e) Chi, Y.; Chuang, S. H.; Chen,
B.-F.; Peng, S.-M.; Lee, G.-H. J. Chem. Soc., Dalton Trans. 1990, 3033.
(f) Friedman, A. E.; Ford, P. C. J. Am. Chem. Soc. 1989, 111, 551. (g)
Keister, J. B. Polyhedron 1988, 7, 847. (h) Friedman, A. E.; Ford, P. C. J.
Am. Chem. Soc. 1986, 108, 7851. (i) Farrugia, L. J.; Miles, A. D.; Stone,
F. G. A. J. Chem. Soc., Dalton Trans. 1985, 2437. (j) Beanan, L. R.; Keister,
J. B. Organometallics 1985, 4, 1713. (k) Nuel, D.; Dahan, F.; Mathieu, R.
J. Am. Chem. Soc. 1985, 107, 1658. (l) Shapley, J. R.; Yeh, W.-Y.; Churchill,
M. R.; Li, Y.-J. Organometallics 1985, 4, 1898. (m) Green, M.; Mead, K.
A.; Mills, R. M.; Salter, I. D.; Stone, F. G. A.; Woodward, P. J. J. Chem.
Soc., Chem. Commun. 1982, 51.
(6) (a) Collman, J. P.; Boulatov, R. Angew. Chem., Int. Ed. 2002, 41,
3948. (b) Winter, M. J. AdV. Organomet. Chem. 1989, 29, 101. (c) Curtis,
M. D. Polyhedron 1987, 6, 759. (d) Cotton, F. A.; Walton, R. A. Multiple
Bonds Between Metal Atoms, 2nd ed.; Oxford University Press: Oxford,
UK, 1993.
(7) Garc´ıa, M. E.; Riera, V.; Ruiz, M. A.; Rueda, M. T.; Sa´ez, D.
Organometallics 2002, 21, 5515.
(8) (a) Heck, J. J. Organomet. Chem. 1986, 311, C5. (b) Abriel, W.;
Baum, G.; Burdof, H.; Heck, J. Z. Naturforsch. 1991, 46b, 841.
(10) Stone, K. C.; White, P. S.; Templeton, J. L. J. Organomet. Chem.
2003, 684, 13.

5914 Organometallics, Vol. 26, No. 24, 2007
Table 1. Selected IRa and NMRb Data for Compounds 1-4
ν(CO)
Garc´ıa et al.
compd
δ(P)
δ(µ-C) [JCP]
[Mo2Cp2(µ-COMe)(µ-PCy2)(µ-CO)] (1)c
[Mo2Cp2(µ-COMe)(µ-PCy2)(CO)3] (2)
[Mo2Cp2(µ-COMe)(µ-PCy2)(CO)2] (3)
[Mo2Cp2{µ-C(CO2Me)}(µ-PCy2)(CO)2] (4)e
1672 (s)
1933 (vs), 1860 (s), 1843 (m, sh)
1894 (w, sh), 1863 (s)
227.8
219.7
125.0
128.2f
352.0 [15]d
300.7 [9]
404.0 [5]
402.8f
1929 (w, sh), 1900 (vs)
a Recorded in dichloromethane solution, ν in cm-1 b Recorded in CD2Cl2 solutions at 290 K and 121.50 (31P) or 75.48 (13C) MHz, unless otherwise
.
stated; δ in ppm relative to internal TMS (13C) or external 85% aqueous H3PO4 (31P); J in Hz. c Data taken from ref 5. d COMe ligand. e Data taken from
ref 3a. f In benzene-d6 solutions.
Table 2. Selected DFT-Optimized Bond Lengths (Å) and
Angles (deg)a
cyclopentadienyl rings, which are inequivalent in the static
structure (Chart 1). This fluxional behavior can be attributed to
the rotation, fast on the NMR time scale, of the OMe group
around the O-C(carbyne) bond and it has been studied in detail
previously for other alkoxycarbyne complexes,1,3c,13 therefore
this was not further investigated.
Computational Studies. As we have shown in the previous
part of this series, the use of DFT methods for our 30-electron
methoxycarbyne-bridged complexes leads to optimized geom-
etries in excellent agreement with those experimentally deter-
mined through X-ray diffraction studies. Moreover, the analysis
of the Kohn-Sham (KS) molecular orbitals has been shown to
be a valuable tool to gain insight into the metal-metal and
metal-carbon bonding in these highly unsaturated systems.
With this in mind, we have carried out similar DFT
calculations (see the Experimental Section for further details)
for the new methoxycarbyne complexes 2 (for both the cis- and
trans-isomers) and 3, and for the carboxycarbyne complex 4.
The latter two complexes are isoelectronic and isostructural,
both of them exhibiting an edge-sharing bioctahedral geometry,
and provide an excellent opportunity to analyze the differences
in the bonding between an alkoxycarbyne ligand and the better
known alkylidyne ligands, the latter lacking the possibility of
any partial π C-O interaction. In addition, the results of these
calculations will allow us to examine the influence of the
intermetallic interaction (with formal bond orders of one and
two for the tricarbonyl and dicarbonyl species, respectively) on
the metal-alkoxycarbyne binding. Following the scheme used
in our previous work,1 we will first discuss the optimized
geometries for these molecules, this being followed by an
analysis of the electronic structure and bonding in these
complexes from two different points of view, the properties of
the relevant molecular orbitals, and the topological properties
of the electron density as managed in the AIM theory.14
Optimized Geometry of the Dicarbonyl Complexes 3 and
4. The most relevant parameters derived from the geometry
optimization of the dicarbonyls 3 and 4, and for the two isomers
of the tricarbonyl 2 can be found in Table 2, with the
corresponding views being collected in Figure 1. The experi-
mental data from the X-ray diffraction of compound 4 are also
collected in Table 2 for comparative purposes.11 As can be seen
for this latter complex, the optimized bond lengths are in quite
good agreement with the corresponding experimental data,
although those involving the metal atoms tend to be slightly
longer (less than 0.05 Å) than the corresponding values
parameter
2
trans-2
3
4b
Mo1-Mo2
Mo1-C*
Mo2-C*
Mo1-CO
3.158
2.343
1.902
1.967
1.981
1.954
1.316
3.176
2.357
1.902
1.985
1.975
1.948
1.315
2.734
2.015
2.027
1.970
2.699 {2.656(1)}
2.000 {1.993(5)}
2.006 {1.977(6)}
1.981 {1.993(8)}
Mo2-CO
C*-O
1.973
1.321
1.978 {2.015(7)}
C*-CO2Me
O-Me
1.455 {1.447(8)}
1.430 {1.448(7)}
1.220 {1.215(7)}
1.364 {1.353(7)}
2.454 {2.409(2)}
2.460 {2.394(2)}
1.447
1.449
1.440
CdO
C(O)-OMe
Mo1-P
2.613
2.414
117.3
149.3
2.621
2.421
116.7
2.449
2.463
119.3
143.8
Mo2-P
C-O-Me
Mo2-C-O
150.6
138.1c
a Mo1 refers to the atom positioned trans to the methyl group of the
COMe ligand, except for compound 4, for which the labeling is arbitrary;
C* refers to the bridgehead atom of the carbyne ligand. b Experimental
values for 4 given in braces (see ref 11). c Mo2-C-C angle.
Indeed, when toluene solutions of 2 were heated at temperatures
higher than 333 K just a generalized decomposition took place
to give a complex mixture of products, none of them being the
monocarbonyl complex 1.
The structural characterization of compound 3 can be made
easily by comparison of its spectroscopic data (Table 1) with
those previously reported for the related bis(phosphide) com-
plexes [M2Cp2(µ-PR2)2(CO)2],7 and with those of the carboxy-
carbyne complex [Mo2Cp2{µ-C(CO2Me)}(µ-PCy2)(CO)2] (4)
(Chart 1). The latter was recently prepared by us upon
demethylation of the dimethoxyacetylene complex [Mo2Cp2-
{µ-η2:η2-C2(OMe)2}(µ-PCy2)(CO)2]BF4,3a and its structure has
been confirmed through an X-ray study.11 The IR spectrum of
3 shows two C-O stretching bands with the typical pattern of
trans-dicarbonyl complexes defining angles between the CO
ligands close to 180° (weak and strong, in order of decreasing
frequencies),12 while its 31P NMR spectrum exhibits a resonance
at 125.0 ppm, in the range found for other related 32-electron
compounds of type trans-[Mo2Cp2(µ-PCy2)(µ-X)(CO)2] (X )
3e-donor ligand),3a,5,7 all of them formally displaying a Mod
Mo double bond. The bridging methoxycarbyne ligand gives
rise to a highly deshielded 13C NMR resonance at 404.0 ppm,
a chemical shift similar to that found for the complex 4 (402.8
ppm). We note that these values are some 30-50 ppm higher
than those reported for other complexes displaying bridging
alkoxycarbyne ligands,1,3,5,9 this deshielding effect probably
being related to the large magnetic anisotropy of the multiple
metal-metal bonds.6b Finally, compound 3 was found to exhibit
dynamic behavior in solution, as concluded from the presence
(13) For examples of fluxional behavior, see for example: (a) Bavaro,
L. M.; Keister, J. B. J. Organomet. Chem. 1985, 287, 357. (b) Keister, J.
B.; Payne, M. W.; Muscatella, M. J. Organometallics 1983, 2, 219. (c)
Johnson, B. F. G.; Lewis, J.; Orpen, A. G.; Raithby, P. R.; Su¨ss, G. J.
Organomet. Chem. 1979, 173, 187. (d) Keister, J. B. J. Chem. Soc., Chem.
Commun. 1979, 214. (e) Gavens, P. D.; Mays, M. J. J. Organomet. Chem.
1978, 162, 389.
1
in the H and 13C NMR spectra of just one resonance for both
(11) Garc´ıa, M. E.; Garc´ıa-Vivo´, D.; Ruiz, M. A. Unpublished results.
(12) Braterman, P. S. Metal Carbonyl Spectra; Academic Press: London,
UK, 1975.
(14) (a) Bader, R. F. W. Atoms in Molecules-A Quantum Theory; Oxford
University Press: Oxford, UK, 1990. (b) Bader, R. F. W. Chem. ReV. 1991,
91, 893.

Synthesis of 32- and 34-Electron Complexes
Organometallics, Vol. 26, No. 24, 2007 5915
Table 3. Calculated and Experimental C-O Stretching
Frequenciesa
ν(C-O)/cm-1
compd
calcdb
expc
percentd
2
2037 (100)
1981 (55)
1971 (56)
1319 (37)
2033 (77)
1976 (53)
1954 (100)
1318 (43)
1990 (20)
1976 (100)
1296 (20)
2018 (28)
1992 (100)
1750 (21)
1947 (vs)
1882 (s)
1870 (s)
4.6
5.2
5.4
trans-2
3
4
1894 (sh, w)
1862 (vs)
5.1
6.1
1929 (sh, w)
1900 (vs)
1652 (w, CdO)
4.6
4.8
5.9
a Values corresponding to the CO, CCO2Me, or COMe ligands, with
relative intensities in parentheses; experimental data for 4 taken from ref
3a. b Uncorrected values. c In petroleum ether (2) or CH2Cl2 solutions (3,
4). d (calcd - exp)/exp.
Figure 1. Optimized geometries for complexes 2, 3, and 4, with
hydrogen atoms omitted for clarity.
measured in the solid state. We also found this tendency for
our 30-electron face-sharing bioctahedral complexes bearing
bridging methoxycarbyne ligands,1 and this is in any case a
common finding with the functionals currently used in the DFT
computations of transition-metal complexes.6b,2a,15
C-O-C angle close to 120°), and is also in agreement with
the MO and AIM analysis to be discussed later on.
A frequency analysis for the dicarbonyls 3 and 4 ensured
that the optimized geometries correspond to real minima of the
potential energy surface (PES), since no imaginary frequencies
were found. The calculated and experimental values of the C-O
stretching frequencies for both the carbonyl and the methoxy-
carbyne ligands for the dicarbonyls 3 and 4 are shown in Table
3. The calculated values overestimate the experimental figures
by ca. 5-6%, which seems to be a normal bias for DFT-derived
IR stretching frequencies.1,17 For the neutral complex 1, the
C-O stretching frequency of the methoxycarbyne ligand was
calculated previously to be 1315 cm-1, and this relatively high
value was taken as another piece of evidence for the persistence
of some πC-O interaction within the methoxycarbyne ligand,
also in agreement with the subsequent MO and AIM analysis.1
In the case of 3, the calculated C-O stretching frequency (1296
cm-1) is comparable to the above figure, and therefore similar
conclusions can be extracted.
The electronic charges computed for these dicarbonyl mol-
ecules have been collected in Table S1 (see the Supporting
Information), which includes Mulliken charges18 as well as those
derived from the NBO analysis (NPA charges).19 For both
complexes, the highest negative atomic charges are found at
the oxygen atoms of the carbonyl ligands, in agreement with
their great π-acidity and the large electronegativity of the oxygen
atoms. As expected, the atomic charge at the C(carbyne) atoms
is highly dependent on the group directly attached to it. The
carboxycarbyne complex 4 displays a high negative charge of
-0.39 (Mulliken) or -0.24 e (NPA) at this site, in good
agreement with the figures previously calculated for model
compounds with terminal carbyne ligands such as [CrCl(CH)-
(CO)4] (-0.42 e) and [Cr(CH)(CO)5]+ (-0.32 e).20 However,
The intermetallic distances calculated for the dicarbonyls 3
and 4 (ca. 2.70 Å) are in good agreement with the double metal-
metal bond that is to be proposed for these 32-electron
complexes according to EAN formalism. As expected, these
values are significantly longer (ca. 0.2 Å) than those previously
calculated for our triply bonded face-sharing bioctahedral
complexes,1 doubtless as a consequence both of the decrease
in the metal-metal bond order (from 3 to 2) as well as in the
number of bridging ligands (also from 3 to 2). The geometrical
parameters are quite similar for both complexes, although the
interatomic distances within the Mo2C triangle are longer for
3, by ca. 0.03 (Mo-Mo) and 0.02 Å (Mo-C bonds). There are
two factors that could explain these observations. First, there
might be a genuine steric effect derived from the presence of
the methoxyl group in the same plane containing the metal and
carbyne atoms in the first compound. Second, a slight lengthen-
ing of the Mo-C bonds might also be expected if some πC-O
interaction in the methoxycarbyne ligand were to be present,
which is detrimental to the πM-C bonding interaction, as found
previously for our 30-electron methoxycarbyne complexes.1 A
related competing interaction is not possible for the carboxy-
carbyne complex 4, hence its shorter Mo-C lengths. In line
with this, the calculated distance for the C-C(carbyne) bond
in 4 is 1.46 Å, a value typical for single bonds between sp2-
hybridized C atoms,16 while the value of 1.32 Å for the
corresponding C-O bond in 3 is somewhat shorter than those
found for comparable single bonds, such as the C(sp2)-O bonds
in organic molecules (ca. 1.35 Å),16b or that found for the ester
moiety of the carboxycarbyne ligand of 4 (1.36 Å). This can
be taken as indicative of the presence of some multiplicity in
this bond, which is consistent with the conformation of the
COMe ligand (arrangement in the Mo2C plane, with the
(17) Yu, L.; Srinivas, G. N.; Schwartz, M. J. Mol. Struct. (THEOCHEM)
2003, 625, 215.
(18) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833.
(19) Mulliken population analyses fail to give a useful and reliable
characterization of the charge distribution in many cases, especially when
highly ionic compounds and diffuse basis functions are involved. Charges
calculated according to the Natural Population Analysis (NPA) do not show
these deficiencies and are more independent of the basis set: (a) Reed, A.
E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735. (b) Reed,
A. E.; Curtis, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899.
(20) Ushio, J.; Nakatsuji, H.; Yonezawa, T. J. Am. Chem. Soc. 1984,
106, 5892.
(15) Cramer, C. J. Essentials of Computational Chemistry, 2nd ed.;
Wiley: Chichester, UK, 2004.
(16) (a) Huheey, J. E.; Keiter, E. A.; Keiter, R. L. Inorganic Chemistry:
Principles of Structure and ReactiVity, 4th ed.; HarperCollins College
Publishers: New York, 1993. (b) Allen, F. H.; Kennard, O.; Watson, D.
G.; Brammer, L.; Orpen, G.; Taylor, R. J. Chem. Soc., Perkin Trans. 2
1987, S1.

5916 Organometallics, Vol. 26, No. 24, 2007
Garc´ıa et al.
for complex 3, with a methoxyl substituent at the carbyne ligand,
we found a significantly lower electron density at the C(carbyne)
atom, which turns out to be almost negligible (-0.03 e,
Mulliken) or slightly positive (+0.21, NPA), with values quite
similar to those previously calculated for our face-sharing
bioctahedral methoxycarbyne complexes,1 or that derived from
the semiempirical MO calculations carried out on the cluster
[Co3(µ3-COMe)(CO)9].21 This decrease in electron density at
the carbyne atom of the methoxycarbyne ligand (when compared
to the carboxycarbyne ligand) cannot be attributed to the
presence of the already mentioned πC-O interaction in the
former, since such an interaction is expected to increase (rather
than decrease) the electron density at the C(carbyne) atom.
Therefore the low negative charge at the bridgehead carbon atom
of the methoxycarbyne ligand must be attributed essentially to
the high electronegativity of the oxygen atom bound to it.
Optimized Geometries of the Tricarbonyl Complex 2. The
optimized geometries of the tricarbonyl isomers cis-2 and
trans-2 are fully consistent with the decrease in the intermetallic
bond order and the high asymmetry expected for these mol-
ecules. The intermetallic distances found for both isomers (ca.
3.16 Å) are consistent with the electron-precise nature of these
34-electron complexes, for which a single metal-metal bond
has to be proposed according to the EAN formalism, and they
are significantly longer (by ca. 0.4 Å) than the corresponding
lengths in the dicarbonyls 3 or 4, as expected. In both isomers
the metal atoms and the bridging phosphorus and carbon atoms
define a puckered Mo2CP rhombus, with dihedral angles of ca.
160° (cis-2) and 150° (trans-2). The most relevant feature in
these central frameworks is the highly asymmetric coordination
of the methoxycarbyne ligand in both cases, which is placed
much closer to the monocarbonylic metal center. This asym-
metry is so pronounced that the longer distance (ca. 2.35 Å) is
remarkably greater than the value expected for a comparable
Mo-C single bond, i.e., the distances found for symmetrically
bridging carbonyl ligands (ca. 2.10 Å).1,5 Conversely, the second
Mo-C bond length is much shorter, with a value of only 1.90
Å, a figure approaching that for the terminal methoxycarbyne
complex ([WTp′(COMe)(CO)2]; Tp′ ) HB(N2C3Me2H)3; Wt
C ) 1.86(1) Å).10 Thus we can describe the coordination of
the methoxycarbyne ligand in these molecules as semibridging
or nearly terminal, with an Mo-C bond close to triple (Chart
1), in agreement with the 13C NMR data of compound 2
discussed above. The dicyclohexylphosphide bridge is also more
strongly bound to the monocarbonylic metal fragment, but the
asymmetry here is less pronounced, with the Mo-P lengths
differing by ca. 0.2 Å.
Figure 2. Experimental (left, petroleum ether solution) and DFT-
calculated (middle and right) IR spectra for compound 2 in the
C-O stretching region.
Molecular Orbitals of the 32-Electron Complexes 3 and
4. A detailed study of the molecular orbital diagrams for edge-
sharing bioctahedral complexes was reported early by Hoffmann
and co-workers,22 and was revisited and widened recently by
some of us.23 The metal-metal bonding in these systems follows
from the occupation of one σ, one π, and one δ bonding MO’s,
with their corresponding antibonding combinations being the
next ones in energy. Using this scheme as a first approach, we
might therefore expect a configuration of the type σ2π2δ2(δ*)2
for the 32-electron complexes 3 and 4, and thus a formally
double metal-metal bond, in agreement with the EAN formal-
ism. This situation, however, could be significantly modified
in our complexes as a result of orbital mixing with the
π-acceptor ligands present there, as turns out to be the case.
The most relevant molecular orbitals of compounds 3 and 4
are depicted in Figure 3, along with their associated energy and
prevalent bonding character. The occupied frontier orbitals
exhibit some differences. In both compounds the highest
occupied molecular orbital (HOMO) corresponds to the metal-
metal δ* orbital, with an important contribution of the π*C-O
orbitals of the terminal carbonyls, thus implying the presence
of significant metal to carbon π back-bonding, which necessarily
decreases the antibonding character of this orbital with respect
to the metal-metal binding. Below this, there is a group of
three orbitals with similar energies and then the corresponding
δ-component of the Mo-Mo bond for these compounds [MO
120 (3) and MO 127 (4)]. The latter orbitals are so low in energy
because of their pronounced delocalization over the Mo2C
triangle (by mixing with the relevant p orbital of the C atom),
and hence represent effectively the π component of the metal-
carbyne bond. A similar effect was previously found for our
face-sharing bioctahedral methoxycarbyne complexes, in which
one of the δ components of the triple metal-metal bond is
involved in the π Mo-C bonding of the carbyne ligand.1 The
three orbitals below the HOMO in compounds 3 and 4
correspond to the σ and π components of the double metal-
metal bond and to the upper-energy component of the metal-
ligand σ-bonding orbitals holding the central Mo2PC framework,
but there are some differences in the metal-metal overlaps
implied. For the carboxycarbyne complex 4, the intermetallic
A frequency analysis for both isomers (cis-2 and trans-2)
ensured that the optimized geometries correspond to real minima
of the PES, and also allowed us to distinguish between both
isomers on the basis of their IR patterns (Figure 2 and Table
3). The calculated patterns of the C-O stretching bands for
both isomers are different enough to safely exclude the presence
of significant amounts of the trans-isomer in the solutions of
compound 2, which exhibit an IR spectrum with the pattern
calculated for the cisoid structure, and NMR resonances
indicative of the presence of a single species in solution.
Furthermore, we note that the cisoid structure is calculated to
be slightly more stable than that of the isomer trans-2 (by ∆G
) 2 kJ mol-1 at 298 K and 1 atm). This small difference itself
would not justify the absence of trans-2 in solution, which
possibly is less favored by the solute-solvent interactions,
although we have not studied this effect.
(22) Shaik, S.; Hoffmann, R.; Fiesel, C. R.; Summerville, R. H. J. Am.
Chem. Soc. 1980, 102, 4555.
(23) Palacios, A. A.; Aullo´n, G.; Alemany, P.; Alvarez, S. Inorg. Chem.
2000, 39, 3166.
(21) Aitchison, A.; Farrugia, L. J. Organometallics 1987, 6, 819.

Synthesis of 32- and 34-Electron Complexes
Organometallics, Vol. 26, No. 24, 2007 5917
Figure 3. Selected molecular orbitals of complexes 2, 3, and 4, with their energies and main bonding character indicated below.
bond consists of the two classical σ and π components, while
for the methoxycarbyne complex this bond is made up of two
nearly equivalent bent (or banana) bonds. At the moment we
can give no satisfactory explanation to this result.
As for the bonding within the methoxycarbyne moiety, there
is a molecular orbital (MO 103, see Table S3 in the Supporting
Information) that reveals the presence of significant π-bonding
interaction in the O-C(carbyne) bond of this ligand, therefore
competing with the Mo-C π bonding (MO 120), as found for
our 30-electron methoxycarbyne complexes,1 and thus this needs
not be further discussed. Finally we note that the lowest
unoccupied molecular orbitals (LUMO) of these dicarbonyl
complexes are largely metal-based molecular orbitals (MO 125
(3) and MO 132 (4), see the Supporting Information), and
exhibit π*M-M antibonding character. As could be expected for
such unsaturated molecules, the incorporation of electron-donor
ligands to the dinuclear center under orbital-controlled conditions
is to be accompanied by a reduction of the intermetallic bond
order. In fact, for the complex 2, which is the CO-addition

5918 Organometallics, Vol. 26, No. 24, 2007
Garc´ıa et al.
Table 4. Topological Properties of the Electron Density at
the Bond Critical Pointsa
2
3
4
2
2
2
bond
F
∇ F
F
∇ F
F
∇ F
Mo1-Mo2
Mo1-C*
Mo1-C*
Mo2-C*
Mo2-C*
C*-OMe
O-Me
0.419b
0.918
0.929b
0.866
0.874b
2.029
1.633
1.66b 0.444b
1.98b
6.65
0.478
1.085
2.084
3.23
8.30
0.82
6.23
0.918
5.29b 0.925b
5.81b
6.96
6.73
0.938
5.89b 0.902b
2.47
6.14b
1.582 -9.95
0.885 10.68
0.865 10.08
0.901 11.27
2.984 24.75
2.975 24.22
2.938 22.70
-11.08
1.651
-9.97
Figure 4. Electron density maps in the Mo2C(carbyne) plane of
compounds 2 (A), 3 (B), and 4 (C). Nuclear positions (b), bond
critical points (2), ring critical points (+), and bond critical paths
(bold lines) are also indicated.
Mo1-CO
Mo2-CO
C-O
0.885
0.878
2.956
2.952
10.20
10.24
23.51
23.40
0.862
0.877
2.990
2.969
10.09
9.87
25.02
24.10
Mo1-P
0.499
0.560
2.14
3.22
0.528
0.523
3.27
3.00
0.526
0.524
1.927
2.736
1.956
0.374
3.17
3.12
plexes with chemically predictable M-M bonds,24 such as [Co2-
(CO)8]25 and [Fe2(CO)9],26 or for the phosphide-bridged com-
plexes [Fe2(µ-H)x(µ-PH2)(CO)4(PH3)2](x-1)+ (x ) 1, 2).27 Since
the quality of the DFT-derived electron density is quite
dependent on the basis sets used, we then tested different
combinations of basis sets and effective core potentials (ECP).
The best results were obtained when using a combination of
the Stuttgart-Dresden relativistic effective core potentials
(SDD-RECP) for molybdenum atoms and 6-31G* basis for light
atoms. These computations were carried out on the previously
optimized structures for these complexes, since we did not
observe significant differences between the optimized structure
of the complex 3 by using either the LANL2DZ or the SDD
pseudopotentials. Keeping in mind the relativistic derivation of
the SDD ECP, we can assume that it will yield a more accurate
description of the electron density especially in the intermetallic
region, then being more reliable to analyze the metal-metal
bonding. The values of F and its laplacian so computed at the
Mo-Mo and Mo-C(carbyne) bcp’s for compounds 3 and 4
have been added to Table 4 (the values for other bonds were
similar to those obtained with the usual basis set and are not
included).
Mo2-P
C*-CO2Me
CdO(OMe)
C(O)-OMe
-18.29
3.97
-11.52
1.62
C*MoPMoc 0.203
MoC*Moc
1.150 0.362
0.419b
1.62
2.03b 0.444b
2.13b
a Values of the electron density at the bond critical points (F) are given
in e Å-3; values of the laplacian of F at these points (∇ F) are given in e
2
Å-5; Mo1 refers to the atom positioned trans to the methyl group of the
COMe ligand, except for compound 4, for which the labeling is arbitrary;
C* refers to the bridgehead atom of the carbyne ligand; all the data were
computed with the LANL2DZ ECP. b Data computed with the SDD ECP.
c Values at the ring critical point.
product of 3, we find a longer metal-metal distance, in
agreement with this bond-order reduction, as discussed below.
Molecular Orbitals of cis-2. The MO’s of this highly
asymmetric molecule exhibit a strong mixing of atomic orbitals
and are difficult to interpret. The HOMO of this compound is
a π* combination of metal AO’s with a significant contribution
of the π*C-O orbital of the terminal carbonyls, and correlates
in shape with the LUMO of its precursor 3, as expected. Mixing
with the π*C-O orbital of the terminal carbonyls is also observed
in the orbitals following the HOMO, which represent the Mo-
Mo σ bond (MO 129) and the Mo-C(carbyne) π bonding. In
contrast to the tricentric π bonding interaction between both
metal centers and the C(carbyne) atom that is found for the
symmetrically bridged carbyne complexes 3 or 4, the corre-
sponding π interaction in 2 is now more localized between the
carbon atom and the monocarbonyl metal center, and it can be
recognized through the orbitals MO 130 (Mo-C π bonding)
and MO 128 (distorted tricentric Mo-C-Mo π bonding). This
is in agreement with the geometric parameters already discussed,
revealing a very asymmetric binding of the carbyne ligand,
strongly bound to the monocarbonylic center.
Topological Analysis of the Electron Density. In the
previous part of this series we have shown that the analysis of
the electron density under the AIM theory provides very
valuable information (complementary to that of the MO analysis)
for the characterization, interpretation and even quantification
of the different bonds present in our highly unsaturated 30-
electron methoxycarbyne complexes.1 We have therefore per-
formed a similar analysis for the complexes 2 to 4, with special
attention to the data relative to the metal-metal and metal-
carbyne bonds. The values of the electron density (F) and its
First, we note that the use of the SDD/6-31G* basis for the
dicarbonyl complexes 3 and 4 leads to the location of the
corresponding metal-metal bcp’s (Figure 5). As expected, the
values of F at these points (ca. 0.43 e Å-3) are intermediate
between those calculated for triply bonded carbonyl complexes
(ca. 0.60 e Å-3) and that computed for the singly bonded [Mo2-
Cp2(CO)6] (ca. 0.17 e Å-3),1 which is in agreement with the
double metal-metal bond proposed for these 32-electron
complexes, also sustained by the MO analysis. In contrast, no
intermetallic bcp could be located for the tricarbonyl 2, probably
due to the much longer M-M separation, this leading to a quite
flat plateau of almost equal F along the intermetallic vector
(Figure 5), where the mathematical conditions of the bcp cannot
be satisfied. Note, however, that the electron density at the
intermetallic region of 2 does not vanish, but keeps a minimum
value of ca. 0.22 e Å-3, similar to the density of the singly
bonded [Mo2Cp2(CO)6] at the intermetallic bcp (ca. 0.17 e Å-3).1
2
Finally, we note that the values of ∇ F at the intermetallic bcp
for compounds 3 and 4 are moderately positive (1.65-2.76 e
Å-5). This feature was also found for our 30-electron meth-
oxycarbyne complexes,1 and seems to be a characteristic of
2
laplacian (∇ F) at the most relevant bond critical points (bcp)
(24) Macchi, P.; Sironi, A. Coord. Chem. ReV. 2003, 238-239, 383.
(25) Low, A. A.; Kunze, K. L.; MacDougall, P. J.; Hall, M. B. Inorg.
Chem. 1991, 30, 1079.
(26) Bo, C.; Sarasa, J. P.; Poblet, J. M. J. Phys. Chem. 1993, 97, 6362.
(27) Phillips, A.; Ienco, A.; Reinhold, J.; Bo¨ttcher, H.-C.; Mealli, C.
Chem. Eur. J. 2006, 12, 4691.
of these complexes are collected in Table 4.
The first significant observation is that in none of the cases
could a Mo-Mo bcp be located with the usual basis set (Figure
4), a circumstance previously found for other carbonyl com-

Synthesis of 32- and 34-Electron Complexes
Organometallics, Vol. 26, No. 24, 2007 5919
ester like the methyl acetate, as far as the values of F and its
laplacian are concerned (see Table 4 and the Supporting
Information), except for the value of F at the bcp of the C-CO2-
Me bond in 4 (1.93 e Å-3), somewhat higher than the
corresponding figure for the Me-C bond in methyl acetate (1.75
e Å-3), as expected when comparing C(sp2)-C(sp2) and
C(sp3)-C(sp2) bonds (shorter interatomic separation in the
former bonds). It is important to notice that all these bonds
2
exhibit large negative values of ∇ F, in agreement with their
covalent nature.14 From this general behavior, however, we have
to exclude the CdO bond, which exhibits a positive value for
this function at the corresponding bcp (3.97 and 4.45 e Å-5 for
4 and methyl acetate, respectively). Actually this is the usual
feature found for multiple C-O bonds, and can be mainly
attributed to the displacement of the bcp toward the carbon
atom.24
As for the C-OMe bonds in the methoxycarbyne complexes
cis-2 and 3 we find that the electron density at the corresponding
bcp’s (ca. 2.05 e Å-3) has values comparable to those found
for our 30-electron methoxycarbyne complexes (2.03-2.13 e
Å-3),1 and all these are in turn slightly higher than the values
found for the C(O)-OMe bonds in 4 or methyl acetate. We
take this alone as an indication of the presence of some
multiplicity in the C-ÃMe bonds of our methoxycarbyne
complexes. Besides this, there is an independent feature pointing
to the same conclusion: the values for the lapacian of F at the
C-ÃMe bcp’s in all our methoxycarbyne complexes are
positive (ca. 0.8-2.5 e Å-5, see Table 4 and ref 1), a common
feature of multiple C-O bonds as stated above, while the
corresponding figures for the C(O)-OMe bonds in 4 or methyl
acetate keep the high negative values (ca. -11.5 e Å-5) expected
for single covalent C-O bonds.
Figure 5. Electron density map in the Mo2C(carbyne) plane of
complex 3, with nuclear positions (b), bond critical points (2),
ring critical points (+), and bond critical paths (bold lines) also
indicated (top). Below are represented the electron density profiles
along the metal-metal vector for the complexes 1, 2, and 3.
metal-metal interactions in transition metal complexes,28 at-
tributed to the diffuse character of the electrons involved in the
bonding.24
As for the M-C(carbyne) bonds, the values of F at the
corresponding bcp’s in the dicarbonyls 3 and 4 are quite similar,
with figures close to the values calculated for our 30-electron
methoxycarbyne complexes (in the range 0.83-0.93 e Å-3).1
The average value for the carboxycarbyne complex 4 (0.93 e
Å-3) is only slightly higher than the corresponding average value
for the methoxycarbyne complex 3 (0.90 e Å-3). From these
data we can conclude that the strength of the interaction between
the dimetal center and a COR ligand is not significantly different
from that of a CR ligand having an electron-withdrawing
hydrocarbon group. The tricarbonyl 2 displays a strongly
asymmetric coordination of the carbyne ligand, and this causes
the electron density at the bcp of the shorter bond to be
considerably higher (1.09 e Å-3) than the values calculated for
symmetrical methoxycarbyne bridges,1 indeed approaching the
electron density calculated for the terminal carbyne CrtCH
(1.17 e Å-3).29 Conversely, the longer Mo-C bond in cis-2 is
characterized by a much lower value of F at the corresponding
bcp (0.48 e Å-3), which is substantially lower than the values
found for comparable M-C single bonds, such as the electron
density calculated for CrCH3 (0.77 e Å-3),29 or the values of
ca. 0.71 e Å-3 found for the Mo-C bonds of the carbonyl
bridges in our 30-electron complexes.1 This justifies our
description of the coordination of the carbyne ligand in 2 as
strongly semibridging, with Mo-C bonds of orders higher than
two and lower than one, respectively (Chart 1).
Concluding Remarks
The 30-electron methoxycarbyne complex 1 does behave as
a substrate having a triple intermetallic bond, then being able
to incorporate two CO molecules to give the corresponding
electron-precise tricarbonyl complex 2, which in turn renders
the 32-electron dicarbonyl 3 through a reversible thermal
decarbonylation. The DFT methods used to study these com-
plexes lead to optimized geometries which in the case of its
unsaturated dicarbonyl 3 are in excellent agreement with that
experimentally determined for the related carboxycarbyne
complex 4, and the frequency calculation for the two possible
isomers of the tricarbonyl 2 allows the unambiguous identifica-
tion of its dominant structure in solution as the cisoid one. The
analysis of the molecular orbitals for the 32-electron edge-
sharing bioctahedral complexes 3 and 4 gives support to the
presence of a double metal-metal bond derived from a
configuration of the type δ2σ2π2(δ*),2 in which the δ* orbitals
are involved in the π back-bonding to the terminal carbonyls,
while the δ orbitals are delocalized over the Mo2C(carbyne)
triangle and then constitute the π-bonding component of the
metal-carbyne bond, which is very similar in both the meth-
oxycarbyne and the carboxycarbyne ligands. Yet, some residual
π-bonding interaction at the C-O bond of the methoxycarbyne
ligand can be identified, which weakens the Mo-C bond of
this ligand to a small extent. The tricarbonyl 2 exhibits a
methoxycarbyne ligand with a strongly asymmetric, nearly
terminal coordination to the metal atoms, with the latter kept at
a short distance by a σ-type bonding orbital.
The bonding within the CO2Me substituent of the carbyne
ligand of 4 is comparable to that of this fragment in an organic
(28) Gervasio, G.; Bianchi, R.; Marabello, D. Chem. Phys. Lett. 2004,
387, 481 and references cited therein.
(29) Vidal, I.; Melchor, S.; Dobado, J. A. J. Phys. Chem. 2005, 109,
7500.
The topological analysis of the electron density within all
these molecules is fully consistent with the MO description of
the bonding just given. In particular, we find that the electron

5920 Organometallics, Vol. 26, No. 24, 2007
Garc´ıa et al.
under vacuum and the residue was dissolved in dichloromethane-
petroleum ether (1:4) and chromatographed on an alumina column
(activity IV, 20 × 2 cm) at 253 K. Elution with dichloromethane-
petroleum ether (1:2) gave a pale pink fraction that yielded, after
removal of solvents, compound 3 as a red microcrystalline solid
(0.52 g, 68%). Anal. Calcd for C26H35Mo2O3P: C, 50.50; H, 5.70.
Found: C, 50.67; H, 5.93. νCO (CH2Cl2) 1894 (w, sh), 1863 (s)
density in the intermetallic region of these molecules reaches a
relatively flat plateau. As a result, only for the double-bonded
dicarbonyls 3 and 4, and using a high quality basis, can the
intermetallic bcp’s be located, with values of F at these points
of ca. 0.4 e Å-3, which are figures intermediate between those
of the triply bonded [Mo2Cp2(CO)4] and the singly bonded [Mo2-
Cp2(CO)6]. The electron densities at the Mo-C bonds of the
dicarbonyl complexes 3 and 4 are similar, and then support the
idea that the strengths of binding to the metal of the methoxy-
carbyne and carboxycarbyne ligands are comparable, with the
former being slightly weakened because of the presence of some
π-interaction in the C-OMe bond, which in turn is supported
by the relatively high value of F and positive (instead of
negative) value of its laplacian at the corresponding bond critical
point.
1
cm-1. H NMR δ 5.45 (s, 10H, Cp), 4.78 (s, 3H, OMe), 2.4-1.1
(m, 22H, Cy). 31P{1H} NMR δ 125.0 (s). 13C{1H} NMR δ 404.0
(d, JCP ) 5, µ-COR), 227.8 (d, JCP ) 11, CO), 90.2 (s, Cp), 70.5
(s, OMe), 44.9 [d, JCP ) 19, C1(Cy)], 36.0 [s, C2,6(Cy)], 34.2 [s,
C6,2(Cy)], 28.5 [d, JCP ) 8, C3,5(Cy)], 28.4 [d, JCP ) 10, C5,3(Cy)],
26.6 [s, C4(Cy)].
Computational Details. All computations described in this work
were carried out with the GAUSSIAN03 package,31 in which the
hybrid method B3LYP was applied with the Becke three parameters
exchange functional32 and the Lee-Yang-Parr correlation func-
tional.33 Effective core potentials (ECP) and their associated
double-ú LANL2DZ basis set were used for the molybdenum and
phosphorus atoms,34 supplemented by an extra d-polarization
function in the case of P.35 The light elements (O, C, and H) were
described with 6-31G* basis.36 Geometry optimizations were
performed under no symmetry restrictions, using initial coordinates
derived from X-ray data of the same or comparable complexes,
and frequency analyses were performed to ensure that a minimum
structure with no imaginary frequencies was achieved in each case.
For interpretation purposes, natural population analysis (NPA)
charges19b were derived from the natural bond order (NBO) analysis
of the data.19b Molecular orbitals and vibrational modes were
visualized by using the Molekel program.37 For the AIM analysis
of F we also carried out single-point calculations on the previously
optimized geometries but using a combination of the relativistic
effective core potentials (RECP) from the Stuttgart-Dresden group
(SDD) to represent the innermost electrons of the Mo atom together
with their associated valence basis set of double-ú quality,38 and
the 6-31G* basis for light elements. The topological analysis of F
was carried out with the Xaim routine.39
Experimental Section
General Procedures and Starting Materials. All manipulations
and reactions were carried out under a nitrogen (99.995%)
atmosphere with standard Schlenk techniques. Solvents were
purified according to literature procedures and distilled prior to
use.30 Petroleum ether refers to that fraction distilling in the range
338-343 K. Compound 1 was prepared as described previously.5
Chromatographic separations were carried out with use of jacketed
columns cooled by a closed 2-propanol circuit, kept at the desired
temperature with a cryostat. Commercial aluminum oxide (Aldrich,
activity I, 150 mesh) was degassed under vacuum prior to use. The
latter was mixed under nitrogen with the appropriate amount of
water to reach the activity desired. All other reagents were obtained
from the usual commercial suppliers and used as received. IR
stretching frequencies were measured in solution and are referred
to as ν (solvent). Nuclear magnetic resonance (NMR) spectra were
routinely recorded at 300.13 (1H), 121.50 (31P{1H}), or 75.47 (13C-
{1H}) at 290 K in CD2Cl2 solutions unless otherwise stated.
Chemical shifts (δ) are given in ppm, relative to internal tetra-
methylsilane (1H, 13C) or external 85% aqueous H3PO4 (31P).
Coupling constants (J) are given in Hz.
Preparation of [Mo2Cp2(µ-COMe)(µ-PCy2)(CO)3] (2). A
toluene solution (10 mL) of compound 1 (0.100 g, 0.169 mmol)
was placed in a bulb equipped with a Young’s valve. The bulb
was cooled at 77 K, evacuated under vacuum, and then refilled
with CO. The valve was then closed, and the solution was allowed
to reach room temperature and further stirred for 14 h to give an
orange solution containing compound 2 as the major product. After
removal of solvents under vacuum, dichloromethane (10 mL) and
then petroleum ether (15 mL) were added. Removal of the solvents
from the latter mixture under vacuum gave compound 2 as an
orange powder (0.099 g, 91%). Anal. Calcd for C27H35Mo2O4P:
C, 50.17; H, 5.46. Found: C, 50.21; H, 5.39. νCO (CH2Cl2) 1933
(vs), 1860 (s), 1843 (m, sh) cm-1. 1H NMR δ 5.34, 5.00 (2 × s, 2
× 5H, 2 × Cp), 4.44 (s, 3H, OMe), 2.5-0.4 (m, 22H, Cy). 31P-
{1H} NMR δ 219.7 (s). 13C{1H} NMR δ 300.7 (d, JCP ) 9,
µ-COMe), 242.8 (d, JCP ) 27, CO), 238.3 (d, JCP ) 10, CO), 236.5
(31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03, Revision B.02; Gaussian, Inc.: Wallingford, CT, 2004.
(32) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(33) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(34) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(s, CO), 92.9, 90.9 (2 × s, 2 × Cp), 70.5 (s, OMe), 54.8 [d, JCP
)
12, C1(Cy)], 39.5 [d, JCP ) 16, C1(Cy)], 36.0 [d, JCP ) 3, C2,6
(Cy)], 35.1 [s, C2,6(Cy)], 34.6 [d, JCP ) 3, C2,6(Cy)], 29.2 [s, C2,6
-
-
(35) Ho¨llwarth, A.; Bo¨hme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi,
A.; Jonas, V.; Ko¨hler, K. F.; Stegman, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 237.
(36) (a) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
(b) Petersson, G. A.; Al-Laham, M. A. J. Chem. Phys. 1991, 94, 6081. (c)
Petersson, G. A.; Bennett, A.; Tensfeldt, T. G.; Al-Laham, M. A.; Shirley,
W. A.; Mantzaris, J. J. Chem. Phys. 1988, 89, 2193.
(37) Portmann, S.; Lu¨thi, H. P. MOLEKEL: An Interactive Molecular
Graphics Tool. CHIMIA 2000, 54, 766.
(38) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor.
Chim. Acta 1990, 77, 123.
(Cy)], 28.9 [d, JCP ) 9, C3,5(Cy)], 28.6 [d, JCP ) 10, C3,5(Cy)],
27.9 [d, JCP ) 13, C3,5(Cy)], 27.7 [d, JCP ) 14, C3,5(Cy)], 26.8 [s,
C4(Cy)], 26.5 [s, C4(Cy)].
Preparation of [Mo2Cp2(µ-COMe)(µ-PCy2)(CO)2] (3). A
toluene solution (20 mL) of compound 2 (0.080 g, 0.124 mmol)
was stirred at 333 K for 45 min to give a red solution containing
compound 3 as the major product. Solvents were then removed
(30) Amarego, W. L. F.; Chai, C. Purification of Laboratory Chemicals,
5th ed.; Butterworth-Heinemann: Oxford, UK, 2003.
(39) Ortiz, J. C.; Bo, C. Xaim; Departamento de Qu´ımica F´ısica e
Inorga´nica, Universidad Rovira i Virgili: Tarragona, Spain, 1998.

Synthesis of 32- and 34-Electron Complexes
Organometallics, Vol. 26, No. 24, 2007 5921
Supporting Information Available: Mulliken and NPA charges,
molecular orbitals, and data from the AIM topological analysis for
complexes 2, 3, and 4. This material is available free of charge via
the Internet at http://pubs.acs.org.
Acknowledgment. We thank the MEC of Spain for a grant
(to D.G.) and financial support (Projects BQU2003-05471 and
CTQ2005-08123-C02-02BQU). The computing resources at the
Centre de Supercomputacio´ de Catalunya (CESCA) were made
available to us through a grant from Fundacio´ Catalana per a la
Recerca (FCR) and Universitat de Barcelona.
OM700735G