Synthesis and cleavage reactions of metal - Metal-bonded [MO2(S2CNR2)6](OTF)2, a source of the tris(dithiocarbamato)molybdenum(IV) fragments

Article abstract of DOI:10.1021/ic010673y

Halide abstraction from the chlorotris(dialkyldithiocarbamato)molybdenum(IV) complexes MoCl(S₂CNR₂)₃ (R = Et, Me) with silver triflate produces the diamagnetic dimeric complexes [Mo₂(S₂CNR₂)₆](OTf)₂ in good yield. The crystallographically determined structure of the diethyldithiocarbamato complex indicates that the dimer consists of two pentagonal bipyramids sharing an axial edge, with a Mo-Mo separation (2.8462(8) A) indicative of a metal-metal bond. A qualitative analysis of the bonding indicates that this bond is of order 2 and consists of one normal σ bond and one relatively weak skewed π interaction. The dimers [Mo₂(S₂CNR₂)₆](OTF)₂ react with a variety of reagents to give monomeric seven-coordinate complexes, including the new cationic molybdenum(IV) complex [Mo(PMe₂Ph)(S₂CNEt₂)₃](OTf), which has been structurally characterized. Kinetic studies of the reaction of [Mo₂(S₂CNEt₂)₆](OTf)₂ with halides indicate the presence of competing dissociative and associative substitution pathways, although neutral donors may react by different mechanisms.

Full text of DOI:10.1021/ic010673y

6676
Inorg. Chem. 2001, 40, 6676-6683
Synthesis and Cleavage Reactions of Metal-Metal-Bonded [Mo2(S2CNR2)6](OTf)2, a Source
of the Tris(dithiocarbamato)molybdenum(IV) Fragment
Sean B. Seymore† and Seth N. Brown*
Department of Chemistry and Biochemistry, 251 Nieuwland Science Hall, University of Notre Dame,
Notre Dame, Indiana 46556-5670
ReceiVed June 25, 2001
Halide abstraction from the chlorotris(dialkyldithiocarbamato)molybdenum(IV) complexes MoCl(S2CNR2)3 (R )
Et, Me) with silver triflate produces the diamagnetic dimeric complexes [Mo2(S2CNR2)6](OTf)2 in good yield.
The crystallographically determined structure of the diethyldithiocarbamato complex indicates that the dimer consists
of two pentagonal bipyramids sharing an axial edge, with a Mo-Mo separation (2.8462(8) Å) indicative of a
metal-metal bond. A qualitative analysis of the bonding indicates that this bond is of order 2 and consists of one
normal σ bond and one relatively weak “skewed π” interaction. The dimers [Mo2(S2CNR2)6](OTf)2 react with a
variety of reagents to give monomeric seven-coordinate complexes, including the new cationic molybdenum(IV)
complex [Mo(PMe2Ph)(S2CNEt2)3](OTf), which has been structurally characterized. Kinetic studies of the reaction
of [Mo2(S2CNEt2)6](OTf)2 with halides indicate the presence of competing dissociative and associative substitution
pathways, although neutral donors may react by different mechanisms.
Introduction
molybdenum(IV) fragment [Mo(S2CNR2)3]+. Here we report
that this cation appears to be unstable as a monomer, instead
dimerizing to form metal-metal-bonded dimers [Mo2(S2CNR2)6]-
(OTf)2 (4). Metal-metal bonding is a staple of molybdenum
chemistry and has been very extensively studied.7 However, the
seven-coordinate geometries of the molybdenum atoms in 4
create an unusual form of multiple bonding in this complex,
with a formally double, but very long, molybdenum-molyb-
denum bond. The strength of the metal-metal interaction helps
to hold the dimer together in solution, but 4 is cleaved by a
variety of reagents under relatively mild conditions and thus
does serve as a source of the [Mo(S2CNR2)3]+ fragment.
The coordination chemistry of molybdenum has received
considerable attention due to the occurrence of this metal in a
number of important biological systems.1,2 There has been
particular interest in sulfur-rich Mo complexes, which are used
to form reactive clusters3 and to mimic the metal-ligand
environment found in molybdoenzymes.4 Pioneering work done
by Chatt and co-workers employed the tris(dialkyldithiocar-
bamato)molybdenum unit, {Mo(S2CNR2)3}, as the core for
many model complexes, including Mo(N2COPh)(S2CNR2)3 (1),5
MoCl(S2CNR2)3 (2),5 and MoN(S2CNR2)3 (3).6
Given the utility and stability of the molybdenum tris-
(dithiocarbamate) core, we sought a convenient source of the
Experimental Section
† Current address: Department of Chemistry, Rose-Hulman Institute of
Technology, 5500 Wabash Avenue, Terre Haute, IN 47803.
(1) (a) Bolin, J. T.; Campobasso, N.; Muchmore, S. W.; Morgan, T. V.;
Mortenson, L. E. In Molybdenum Enzymes, Cofactors, and Model
Systems; Stiefel, E. I., Coucouvanis, D., Newton, W. E., Eds.; ACS
Symp. Series 535; American Chemical Society: Washington, DC,
1993; pp 186-195. (b) Rajagopalan, K. In Molybdenum and Molyb-
denum Containing Enzymes; Coughlan, M., Ed.; Pergamon: Oxford,
U.K., 1980; pp 241-272.
Unless otherwise noted, all procedures were carried out on the bench
top. Acetonitrile, chloroform, and methylene chloride were dried over
4 Å molecular sieves, followed by CaH2. Acetone was dried over 4 Å
molecular sieves. Dry ether was vacuum transferred from sodium
benzophenone ketyl. MoO2(S2CNR2)2 complexes [R ) Et (a series),
Me (b series)] were prepared using a literature procedure.8 Mo(N2-
COPh)(S2CNR2)3 (1) and MoCl(S2CNR2)3 (2) were prepared using the
1
procedures outlined in a short communication;5 details and H NMR
(2) (a) Hille, R. Chem. ReV. 1996, 96, 2757-2816. (b) Holm, R. H. Chem.
ReV. 1987, 87, 1401-1449.
spectral data for the compounds are given below. MoN(S2CNR2)3 was
prepared using a literature method.6 All other reagents were com-
mercially available and used without further purification.
(3) (a) Mayor-Lopez, M. J.; Weber, J.; Hegetschweiler, K.; Meienberger,
M. D.; Joho, F.; Leoni, S.; Nesper, R.; Reiss, G. J.; Frank, W.; Kolesov,
B. A.; Fedin, V. P.; Fedorov, V. E. Inorg. Chem. 1998, 37, 2633-
2644. (b) Dance, I.; Fisher, K. Prog. Inorg. Chem. 1994, 41, 637-
803. (c) Meienberger, M. D.; Hegetschweiler, K.; Ruegger, H.;
Gramlich, V. Inorg. Chim. Acta 1993, 213, 157-169. (d) Shibahara,
T. Coord. Chem. ReV. 1993, 123, 73-147. (e) Zimmermann, H.;
Hegetschweiler, K.; Keller, T.; Gramlich, V.; Schmalle, H. W.; Petter,
W.; Schneider, W. Inorg. Chem. 1991, 30, 4336-4341.
(4) See for example: (a) Thapper, A.; Lorber, C.; Fryxelius, J.; Behrens,
A.; Norlander, E. J. Inorg. Biochem. 2000, 79, 67-74. (b) Smith, P.
D.; Millar, A. J.; Young, C. G.; Ghosh, A.; Basu, P. J. Am. Chem.
Soc. 2000, 122, 9298-9299. (c) Smith, P. D.; Slizys, D. A.; George,
G. N.; Young, C. G. J. Am. Chem. Soc. 2000, 122, 2946-2947. (d)
Mader, M. L.; Carducci, M. D.; Enemark, J. H. Inorg. Chem. 2000,
39, 525-531.
(5) Bishop, M. W.; Chatt, J.; Dilworth, J. R. J. Organomet. Chem. 1974,
73, C59-C60.
(6) Chatt, J.; Dilworth, J. R. J. Indian Chem. Soc. 1977, 54, 13-18.
NMR spectra were measured on a General Electric GN-300 or a
Varian-300 FT-NMR spectrometer. Chemical shifts for 1H and 13C{1H}
spectra are reported in ppm referenced to TMS. Those for 19F and
31P{1H} spectra are reported in ppm referenced to external trifluoroacetic
and phosphoric acid, respectively. Infrared spectra were recorded as
evaporated films on KBr plates on a Perkin-Elmer Paragon 1000 FT-
IR spectrometer. UV-visible data were collected on a Beckman DU-
7500 diode-array spectrophotometer (300-800 nm) equipped with a
multicell transport block or a Perkin-Elmer UV/vis/NIR Lambda 19
spectrophotometer (300-3000 nm). Mass spectra were obtained on a
JEOL JMS-AX 505HA mass spectrometer using the FAB ionization
(7) Cotton, F. A.; Walton, R. A. Multpile Bonds Between Metal Atoms,
2nd ed.; Clarendon: Oxford, U.K., 1993; pp 682-719.
(8) Moore, F. W.; Larson, M. L. D. Inorg. Chem. 1967, 6, 998-1003.
10.1021/ic010673y CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/17/2001

Metal-Metal-Bonded [Mo2(S2CNR2)6](OTf)2
Inorganic Chemistry, Vol. 40, No. 26, 2001 6677
mode and 3-nitrobenzyl alcohol as a matrix. In all cases, observed
intensities were in satisfactory agreement with calculated isotopic
distributions. Elemental analyses were performed by M-H-W Labora-
tories (Phoenix, AZ) or Canadian Microanalytical Services, Ltd.
(Vancouver, BC, Canada).
established within minutes, includes the new product. 1H NMR (CD2-
Cl2): δ 1.52 (br, 36H, CH3); 17.94 (br, 24H, CH2). FABMS: m/e 1095,
(M - H)+. The complex [(Me2NCS2)3Mo(µ-N)Mo(S2CNMe2)3]OTf
(5b) was prepared analogously. 1H NMR (CD2Cl2): δ 27.88 (br, 36H,
CH3). The mixed complex [(Me2NCS2)3Mo(µ-N)Mo(S2CNEt2)3]OTf
(5c) can be obtained by mixing 3a and 4b, or 3b and 4a, in CD2Cl2.
The resulting equilibrium mixture includes 5a,b and the new product.
1H NMR (CD2Cl2): δ 1.55 (br, 18H, CH2CH3), 18.90 (br, 12H, CH2-
CH3), 26.29 (br, 18H, NCH3).
Mo(N2COPh)(S2CNEt2)3 (1a).5 To a 250-mL round-bottom flask
were added a magnetic stirbar, MoO2(S2CNEt2)2 (2.53 g, 5.96 mmol),
NaS2CNEt2‚3H2O (Aldrich, 1.60 g, 7.10 mmol), PhCONHNH2 (Aldrich,
0.88 g, 6.46 mmol), and absolute MeOH (100 mL). After stirring of
the mixture at reflux for 3 h, the orange product was collected by
filtration, washed with two 10 mL aliquots of absolute MeOH, and
Mo(N3)(S2CNEt2)3 was generated in solution. In the drybox an NMR
tube with a Teflon-lined screw cap was charged with 4a (5.9 mg, 4.3
µmol), NaN3 (Fisher, 6.0 mg, 0.09 mmol), and CD3CN (0.5 mL).
1
air-dried. Yield: 3.17 g (79%). H NMR (CDCl3): δ 1.27 (m, 18H,
1
CH3); 3.77 (m, 12H, CH2); 7.28 (t, J ) 8 Hz, 2H, meta); 7.38 (t, J )
8 Hz, 1H, para); 8.03 (d, J ) 7 Hz, 2H, ortho).
Reaction progress was monitored by H NMR spectroscopy for 24 h.
During this time period the intermediate Mo(N3)(S2CNEt2)3 was
observed. 1H NMR (CD3CN): δ 1.52 (br, 18H, CH3), 17.71 (br, 12H,
CH2CH3). Immersion of the tube in a 60 °C oil bath for 1 h led to the
quantitative formation of 3a.
Mo(N2COPh)(S2CNMe2)3 (1b).5 Following the procedure for 1a,
MoO2(S2CNMe2)2 (2.72 g, 7.38 mmol) yielded 0.70 g (16%) of orange
1
product. H NMR (CD2Cl2): δ 3.26 (s, 3H, CH3), 3.28 (s, 6H, CH3),
3.32 (s, 6H, CH3), 3.39 (s, 3H, CH3); 7.32 (t, J ) 8 Hz, 2H, meta);
7.44 (t, J ) 8 Hz, 1H, para); 8.03 (d, J ) 7 Hz, 2H, ortho).
[Mo(PMe2Ph)(S2CNEt2)3]OTf ([6]OTf). In the drybox, 4a (243 mg,
0.176 mmol), PMe2Ph (Strem, 145 mg, 1.10 mmol), and a magnetic
stirbar were added to a 50-mL round-bottom flask. The flask was
attached to the vacuum line. Dry acetone (20 mL) was added by vacuum
transfer, and the solution was stirred in vacuo for 2 d. The volume
was reduced to 15 mL, and ether (20 mL) was condensed on the reddish-
brown solution. The next day, the reaction mixture was taken into the
drybox, and the red crystals were filtered out and washed with ether.
MoCl(S2CNEt2)3 (2a).5 To a 250-mL round-bottom flask were added
a magnetic stirbar, 1a (2.05 g, 0.304 mmol), and absolute methanol
(100 mL). Gaseous HCl, generated from H2SO4 and saturated brine
and dried by passage through H2SO4, was vigorously bubbled through
the orange suspension for 15 min. Heat was generated, and the mixture
turned forest green. The flask was quickly affixed to a vacuum line,
and the volume was reduced to 30 mL. The resulting green solid was
collected on a frit under vacuum and washed with 10 mL of freshly
distilled CH3OH. The product was dried under vacuum and stored in
the drybox. Yield: 1.09 g (63%). 1H NMR (CD2Cl2): δ 1.33 (s, 18H,
CH3); 25.29 (br, 12H, CH2). UV-vis (CH2Cl2) [λ, nm (ꢀ, M-1 cm-1)]:
363 (7200), 406 (sh, 2600), 592 (80).
1
Yield: 113 mg (39%). H NMR (CDCl3): δ -18.15 (br, 6H, PCH3),
1.69 (s, 18H, CH2CH3), 7.43 (s, 1H, para), 10.74 (s, 2H, meta), 14.90
(br, 2H, ortho), 35.39 (s, 12H, CH2CH3). IR (cm-1): 1519 (s), 1460
(m), 1441 (m), 1383 (m), 1357 (m), 1274 (vs, νSO ), 1224 (m), 1205
3
(m), 1152 (m, νCF ), 1097 (w), 1076 (m), 1000 (w), 935 (m), 913 (w),
3
850 (w), 781 (w), 749 (w), 696 (w), 638 (s). UV-vis (CH2Cl2) [λ, nm
(ꢀ, M-1 cm-1)]: 418 (5400). Anal. Calcd for C24H41F3MoN3O3PS7: C,
34.82; H, 4.99; N, 5.08. Found: C, 35.00; H, 4.24; N, 5.14.
MoCl(S2CNMe2)3 (2b).5 Following the procedure above, 1b (1.00
g, 1.70 mmol) furnished the green product 2b (0.58 g, 68%). 1H NMR
(CD2Cl2): δ 38.41 (br s, 18H, CH3).
[Mo(PPh3)(S2CNEt2)3]OTf was generated in solution by addition
of PPh3 (2 equiv) to a solution of 4a in acetone-d6 or CD2Cl2. 1H NMR
(acetone-d6): δ 1.49 (br, 18H, CH3), 7.73 (br, 3H, para), 10.00 (br,
6H, meta), 11.20 (v br, 6H, ortho), 36.87 (br, 12H, CH2CH3). [Mo-
(NCCD3)(S2CNEt2)3]OTf was generated by dissolving 4a in CD3CN.
1H NMR (CD3CN): δ 1.48 (br, 18H, CH3), 30.52 (br, 12H, CH2CH3).
X-ray Structure Determinations of [Mo2(S2CNEt2)6](OTf)2‚
4CHCl3 (4a‚4CHCl3) and [Mo(PMe2Ph)(S2CNEt2)3]OTf (6). Crystals
of [Mo2(S2CNEt2)6](OTf)2‚4CHCl3 were grown by layering a solution
of 4a in chloroform with diethyl ether (1:3 v/v) and allowing the mixture
to stand at -30 °C in the drybox for 2 weeks. A small red plate (0.2
× 0.1 × 0.02 mm) was placed in inert oil and transferred to the tip of
a glass fiber in the cold N2 stream of a Bruker Apex CCD diffractometer
(T ) -100 °C). Data were reduced, correcting for absorption and decay,
using the program SADABS. The crystal was triclinic (space group
P1h). The molybdenum atoms were located on a Patterson map, and
the remaining non-hydrogen atoms were found on difference Fourier
syntheses. Hydrogens were placed in calculated positions. One of the
methyl groups (the one attached to C63) was found to be disordered
over two positions, which were given equal occupancy in the refine-
ment. Final full-matrix least-squares refinement on F2 converged at
R ) 0.0545 for 8439 reflections with Fo > 4σ(Fo) and R ) 0.0977 for
all 13 131 unique reflections (wR2 ) 0.1248, 0.1612, respectively).
All calculations used SHELXTL (Bruker Analytical X-ray Systems),
with scattering factors and anomalous dispersion terms taken from the
literature.9 Crystallographic data are found in Tables 1-3.
[Mo2(S2CNEt2)6](OTf)2 (4a). In the drybox, 2a (1.09 g, 1.89 mmol),
AgOTf (Aldrich, 540 mg, 2.10 mmol), and a magnetic stirbar were
added to a 100-mL round-bottom flask. The flask was attached to a
swivel frit, which was then affixed to a vacuum line. Dry CH2Cl2 (30
mL) was added by vacuum transfer, and the solution was stirred in the
dark for 3 h. The insoluble material was removed by filtration, and
ether (30 mL) was condensed on the resulting brown solution. The
next day, the brown solid was collected by filtration and taken into the
drybox. Yield: 1.07 g (81%). Alternatively, TlOTf can be substituted
for AgOTf in this procedure and used to prepare the product in similar
yields. 1H NMR (CD2Cl2): δ 1.34 (m, 36H, CH3); 3.79 (m, 24H, CH2).
13C{1H} NMR (CD2Cl2): δ 12.71, 12.82, 12.85, 12.89, 13.31, 14.08
(CH2CH3); 44.94, 45.34, 46.25, 46.37, 46.53, 46.77 (CH2CH3); 196.12,
196.98 (CS2). 19F{1H} NMR (CD2Cl2): δ -1.03 (s, 6F, CF3). IR (cm-1):
1523 (vs), 1460 (m), 1443 (s), 1383 (m), 1357 (m), 1273 (vs, νSO ),
3
1223 (m), 1205 (m), 1152 (s, νCF ), 1096 (w), 1076 (m), 1031 (vs,
3
ν
SO ), 970 (w), 916 (w), 850 (m), 780 (2), 754 (w), 638 (vs). FABMS:
3
m/e 1081, (M - H+)+. UV-vis (CH2Cl2) [λ, nm (ꢀ, M-1 cm-1)]: 398
(6300), 466 (4800), 517 (sh, 2500), 582 (sh, 1400), 817 (1200). Anal.
Calcd for C32H60F6Mo2N6O6S14: C, 27.86; H, 4.38; N, 6.09. Found:
C, 27.47; H, 4.41; N, 5.86.
[Mo2(S2CNMe2)6](OTf)2 (4b). Following the procedure above, 2b
(479.0 mg, 1.01 mmol) furnished 419.3 mg of [Mo2(S2CNMe2)6](OTf)2
1
(4b) as a brown solid (68%). H NMR (acetone-d6): δ 3.41 (s, 6H,
CH3), 3.42 (s, 6H, CH3), 3.49 (s, 6H, CH3), 3.51 (s, 6H, CH3), 3.53
(s, 6H, CH3), 3.74 (s, 6H, CH3). 13C{1H} NMR (CD2Cl2): δ 39.25,
39.73, 40.55, 40.90, 41.05, 42.01 (CH3); 196.92 (CS2). 19F{1H}
NMR (CD2Cl2): δ 0.34 (s, 6F, CF3). IR (cm-1): 1549 (vs), 1446 (m),
1401 (vs), 1261 (vs, νSO ), 1224 (s), 1155 (vs, νCF ), 1030 (vs, νSO ),
A 0.35 × 0.17 × 0.13 mm red block of [Mo(PMe2Ph)(S2CNEt2)3]-
OTf was deposited after slow diffusion of ether into a solution of 6 in
acetone. Data collection and reduction were done as described above.
The crystal was monoclinic, and its space group was determined to be
P21/n on the basis of the systematic absences. The molybdenum atoms
and the atoms in the first coordination shell were located using direct
methods, and remaining nonhydrogen atoms were found on difference
Fourier syntheses. Hydrogens were placed in calculated positions. There
were two crystallographically independent molecules in the unit cell,
3
3
3
991 (m), 638 (vs). FABMS: m/e 913, (M - H+)+. Anal. Calcd for
C20H36F6Mo2N6O6S14: C, 19.83; H, 3.00; N, 6.94. Found: C, 19.64;
H, 3.17; N, 6.69.
[(Et2NCS2)3Mo(µ-N)Mo(S2CNEt2)3]OTf (5a) was generated in
solution. In a typical experiment (Et2NCS2)3MoN (3a, 6.6 mg, 11.9
µmol) and 4a (10.5 mg, 7.6 µmol) were dissolved in CD2Cl2 (0.5 mL)
in the drybox. The solution was added to a screw-cap NMR tube and
monitored by 1H NMR spectroscopy. The resulting equilibrium mixture,
(9) International Tables of Crystallography; Kluwer Academic Publish-
ers: Dordrecht, The Netherlands, 1992; Vol. C.

6678 Inorganic Chemistry, Vol. 40, No. 26, 2001
Seymore and Brown
Table 3. Selected Bond Lengths (Å) and Angles (deg) for
Table 1. Crystallographic Details for
[Mo2(S2CNEt2)6](OTf)2‚4CHCl3 (4a‚4CHCl3) and
[Mo(PMe2Ph)(S2CNEt2)3]OTf (6)
[Mo(PMe2Ph)(S2CNEt2)3](OTf) (6)
molecule 1
molecule 2
2.4990(11)
[Mo2(S2CNEt2)6]-
(OTf)2‚4CHCl3
(4a‚4CHCl3)
[Mo(PMe2Ph)-
(S2CNEt2)3]-
Mo1-S11
Mo1-S12
Mo1-S13
Mo1-S14
Mo1-S15
Mo1-S16
Mo1-P1
2.4914(11)
2.4946(11)
2.4956(11)
2.4842(11)
2.5035(11)
2.4846(12)
2.5564(12)
Mo2-S21
Mo2-S22
Mo2-S23
Mo2-S24
Mo2-S25
Mo2-S26
Mo2-P2
2.4849(11)
2.4811(10)
2.4896(11)
2.5060(11)
2.4851(11)
2.5430(12)
OTf (6)
empirical formula
fw
C36H64Cl12F6Mo2N6O6S14 C24H41F3MoN3O3PS7
1857.05
827.93
temp (K)
λ (Å)
space group
tot. data collcd
no. of indep reflns
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
170
170
0.710 73 (Mo KR)
P1h
41 562
0.710 73 (Mo KR)
P21/n
63 134
S11-Mo1-S12
S11-Mo1-S13
S11-Mo1-S14
S11-Mo1-S15
S11-Mo1-S16
S11-Mo1-P1
S12-Mo1-S13
S12-Mo1-S14
S12-Mo1-S15
S12-Mo1-S16
S12-Mo1-P1
S13-Mo1-S14
S13-Mo1-S15
S13-Mo1-S16
S13-Mo1-P1
S14-Mo1-S15
S14-Mo1-S16
S14-Mo1-P1
S15-Mo1-S16
S15-Mo1-P1
S16-Mo1-P1
69.06(4)
143.10(4)
147.56(4)
74.03(3)
87.68(4)
88.81(4)
74.31(4)
141.88(4)
142.14(4)
99.17(4)
86.17(4)
69.18(3)
140.86(4)
93.96(4)
93.11(4)
75.97(4)
93.96(4)
85.22(4)
70.84(4)
101.37(4)
172.09(4)
S21-Mo2-S22
S21-Mo2-S23
S21-Mo2-S24
S21-Mo2-S25
S21-Mo2-S26
S21-Mo2-P2
S22-Mo2-S23
S22-Mo2-S24
S22-Mo2-S25
S22-Mo2-S26
S22-Mo2-P2
S23-Mo2-S24
S23-Mo2-S25
S23-Mo2-S26
S23-Mo2-P2
S24-Mo2-S25
S24-Mo2-S26
S24-Mo2-P2
S25-Mo2-S26
S25-Mo2-P2
S26-Mo2-P2
68.80(3)
141.54(4)
149.13(4)
74.84(3)
88.48(4)
85.76(4)
72.79(3)
141.26(4)
142.12(4)
97.29(4)
91.02(4)
69.14(3)
141.96(4)
94.54(4)
96.91(4)
76.39(3)
92.63(4)
86.66(4)
71.04(4)
96.65(4)
167.45(4)
13 131
17 026
13.4735(8)
15.3695(10)
19.4407(12)
96.8470(10)
92.5940(10)
110.6430(10)
3723.9(4)
2
14.544(2)
16.716(2)
29.745(3)
90
98.158(2)
90
7158.3(13)
8
1.536
γ (deg)
V (Å3)
Z
calcd F (g/cm3)
cryst size (mm)
1.656
0.02 × 0.1 × 0.2
0.13 × 0.17 × 0.35
µ (mm-1
)
1.217
0.866
R indices [I > 2σ(I)]a R1 ) 0.0545,
R1 ) 0.0511,
wR2 ) 0.1149
R1 ) 0.0838,
wR2 ) 0.1378
wR2 ) 0.1248
R1 ) 0.0977,
wR2 ) 0.1612
R indices (all data)a
a R1 ) ∑||Fo| - |Fc||/∑|Fo|; wR2 ) (∑[w(Fo2 - Fc2)2]/∑w(Fo )2)1/2
.
2
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
[Mo2(S2CNEt2)6](OTf)2‚4CHCl3 (4a‚4CHCl3)
ethylene glycol mixture circulated through the cell block and was
measured by a thermocouple inserted in the cell block. Reactions of
4a were carried out anaerobically in dichloromethane at 25.3 °C using
1 cm quartz cells fitted with septum caps and were monitored at 790
nm. Added reagents were present in >20-fold excess over 4a, whose
concentration was in the range (2-4) × 10-4 M. Three replicates were
run for each set of experimental conditions. Pseudo-first-order rate
constants were obtained for the halide substitution reactions using a
least-squares fit to the equation ln |A∞ - A| ) -kt + ln |A∞ - A0|. The
plots were linear for at least 4 half-lives.
Mo1-S11
Mo1-S12
Mo1-S21
Mo1-S22
Mo1-S31
Mo1-S32
Mo1-S61
Mo1-Mo2
2.499(2)
2.535(2)
2.439(2)
2.490(2)
2.465(2)
2.523(2)
2.495(2)
2.8462(8)
Mo2-S51
Mo2-S52
Mo2-S61
Mo2-S62
Mo2-S41
Mo2-S42
Mo2-S21
2.476(2)
2.564(2)
2.429(2)
2.486(2)
2.468(2)
2.522(2)
2.491(2)
S11-Mo1-S12
S11-Mo1-S21
S11-Mo1-S22
S11-Mo1-S31
S11-Mo1-S32
S11-Mo1-S61
S12-Mo1-S21
S12-Mo1-S22
S12-Mo1-S31
S12-Mo1-S32
S12-Mo1-S61
S21-Mo1-S22
S21-Mo1-S31
S21-Mo1-S32
S21-Mo1-S61
S22-Mo1-S31
S22-Mo1-S32
S22-Mo1-S61
S31-Mo1-S32
S31-Mo1-S61
S32-Mo1-S61
68.32(5)
148.39(6)
140.87(6)
92.01(6)
75.62(5)
80.45(6)
142.65(6)
72.61(6)
84.65(6)
134.47(6)
83.09(6)
70.63(6)
86.76(6)
74.40(5)
105.02(6)
86.51(6)
138.36(6)
92.91(6)
69.48(5)
167.32(6)
117.74(6)
S51-Mo2-S52
S51-Mo2-S61
S51-Mo2-S62
S51-Mo2-S41
S51-Mo2-S42
S51-Mo2-S21
S52-Mo2-S61
S52-Mo2-S62
S52-Mo2-S41
S52-Mo2-S42
S52-Mo2-S21
S61-Mo2-S62
S61-Mo2-S41
S61-Mo2-S42
S61-Mo2-S21
S62-Mo2-S41
S62-Mo2-S42
S62-Mo2-S21
S41-Mo2-S42
S41-Mo2-S21
S42-Mo2-S21
68.21(6)
145.74(6)
143.20(6)
91.15(6)
73.38(6)
80.94(6)
145.29(6)
75.04(6)
84.46(6)
132.84(6)
82.47(6)
70.94(6)
87.22(6)
73.98(5)
105.47(6)
87.17(6)
138.64(6)
92.63(6)
69.79(6)
166.52(6)
117.59(6)
Results
Synthesis and Characterization of [Mo2(S2CNR2)6](OTf)2.
The well-known Mo(VI) compounds MoO2(S2CNR2)2 [R ) Et
(a series), Me (b series)]8 provide a convenient entry into
molybdenum(IV) tris(dithiocarbamate) complexes MoX(S2-
CNR2)3 using a two-step route described by Chatt and co-
workers (eqs 1 and 2).5 The orange, air-stable, diamagnetic
Mo1-S21-Mo2
70.52(5)
Mo1-S61-Mo2
70.61(5)
which were related by a translation of half a unit cell along the a axis.
These two molecules were essentially identical except for the confor-
mation of one ethyl group, the orientation of the triflate counterion,
and slight differences in the conformation of the PMe2Ph ligand. Final
full-matrix least-squares refinement of F2 converged at R ) 0.0511
for 11 237 reflections with Fo > 4σ(Fo) and R ) 0.0838 for all 17 026
unique reflections (wR2 ) 0.1149, 0.1378, respectively).
Kinetic Studies. UV-visible data were collected on a Beckman
DU-7500 diode-array spectrophotometer equipped with a multicell
transport block. The temperature was regulated by a thermostated water/
diazenido complexes Mo(N2COPh)(S2CNR2)3 (1) have Cs
symmetry on the NMR time scale. This is consistent with the

Metal-Metal-Bonded [Mo2(S2CNR2)6](OTf)2
Inorganic Chemistry, Vol. 40, No. 26, 2001 6679
Figure 1. Optical spectra of (Et2NCS2)3MoCl (2a, dashed line) and
[Mo2(S2CNEt2)6](OTf)2 (4a, solid line) in CH2Cl2. Extinction coef-
ficients are given per mole of Mo.
Figure 2. SHELXTL plot (30% thermal ellipsoids) of the cation of
[Mo2(S2CNEt2)6](OTf)2‚4CHCl3 (4a‚4CHCl3).
solid-state structures of the analogous complexes Mo(N2Ar)-
(S2CNMe2)3 (Ar ) Ph, 3-NO2C6H4)10 and Mo(N2CO2Et)(S2-
CNMe2)3,11 which have a pentagonal bipyramidal geometry with
apical diazenido ligands. The forest green, slightly air-sensitive
chloro complexes MoCl(S2CNR2)3, in contrast, show only a
confirmed by X-ray crystallography (Figure 2). Each [Mo(S2-
CNEt2)3]+ fragment forms a pentagonal bipyramidal structure
with one apical group missing. One of the equatorial sulfur
atoms on each [Mo(S2CNEt2)3]+ fragment (S21 on Mo1 and
S61 on Mo2) binds to the vacant axial position on the other
molybdenum. The bridging dithiocarbamates are both on the
same face of the Mo2S2 core, giving the structure overall
(noncrystallographic) C2 symmetry. This mode of bridging is
typical of edge-sharing bioctahedral complexes linked by
bridging dithiocarbamates.13 Dimeric 4 consists of two pen-
tagonal bipyramids sharing an axial edge, a motif which has
been seen before only once, in the osmium analogue [Os2(S2-
CNEt2)6](PF6)2 (4Os).14 Indeed, the molybdenum and osmium
structures are remarkably similar on their peripheries, apart from
the generally shorter bonds to osmium (Os-Savg ) 2.415 Å in
4Os vs Mo-Savg ) 2.492 Å in 4a). However, the cores are
significantly different because the molybdenum atoms in 4a, at
2.8462(8) Å, approach each other much more closely than do
the osmium atoms in 4Os [3.682(1) Å] or than the molybdenum
atoms in dithiocarbamate-bridged Mo2(CO)4(NO)2(µ-S2CNEt2)2
[3.773(1) Å].15 The short Mo-Mo distance in 4a induces very
acute angles at the bridging sulfur atoms (70.56° vs 98.4° in
4Os) and clearly indicates a significant amount of Mo-Mo
bonding. Spectroscopic and chemical evidence leave little doubt
that the dimeric structure of 4 is retained in solution. For
example, the optical spectrum of 4a is quite different from
monomeric Mo(IV) complexes, even at submillimolar concen-
trations. The NMR spectra that show six different alkyl groups
for the dithiocarbamate ligands in 4 indicate that not only is
the integrity of the dimer retained on the NMR time scale but
that the structure is also nonfluxional. Both of these observations
contrast starkly with the behavior of the osmium analogue 4Os,
which shows only a single ethyl peak and is significantly
1
single set of alkyl resonances in their H NMR spectra, with
the hydrogens on the carbons directly attached to nitrogen
appearing far downfield (δ 25.29 for the CH2 resonance of 2a,
δ 38.41 for 2b in CD2Cl2). This indicates that the seven-
coordinate, 16-electron complexes are paramagnetic and undergo
a rapid fluxional process that renders all alkyl groups equivalent
at room temperature.
The chloride complexes 2 react rapidly in dichloromethane
with silver trifluoromethanesulfonate (silver triflate, AgOTf) to
precipitate AgCl and give air-sensitive species 4 of empirical
formula (R2NCS2)3Mo(OTf) which are soluble in polar organic
solvents such as acetone, CH2Cl2, CHCl3, and CH3CN and
insoluble in ether, benzene, and hexane. The spectroscopic
properties of the triflates 4 differ dramatically from those of
the precursor chloride complexes 2, suggesting that the reaction
is not a simple metathesis of triflate for chloride. For example,
the chloride complex 2a is pale green, appearing yellow in dilute
solution; its optical spectrum (Figure 1) shows a very weak
band at 592 nm in addition to intense charge-transfer bands in
the near-UV. The dark brown 4a has a complex optical spectrum
(Figure 1) and, in particular, shows a moderately intense
(ꢀ ) 1200 M-1 cm-1) absorption at rather long wavelengths
1
(λmax ) 817 nm). The triflate complexes 4 exhibit H and 13C
NMR spectra with dithiocarbamato resonances at normal
chemical shifts and are therefore diamagnetic, in contrast to
paramagnetic 2. Furthermore, the NMR spectra of 4 show six
inequivalent alkyl groups. For example, the spectrum of the
dimethyldithiocarbamato analogue 4b in acetone-d6 shows six
singlets of equal intensity and the methylene protons of 4a are
diastereotopic. Such a low-symmetry spectrum rules out a simple
ionic formulation as [Mo(S2CNR2)3]OTf, although the IR
spectrum does show bands characteristic of ionic triflate.12
A solution to the structural conundrum is suggested by the
FABMS of 4, which show prominent peaks at masses corre-
sponding to Mo2(S2CNR2)6. The dimeric structure of 4a was
(13) (a) Raston, C. L.; White, A. H. J. Chem. Soc., Dalton Trans. 1975,
2418-2425. (b) Mattson, B. M.; Heiman, J. R.; Pignolet, L. H.
Inorg. Chem. 1976, 15, 564-571. (c) Landgrafe, C.; Sheldrick, W. S.
J. Chem. Soc., Dalton Trans. 1994, 1885-1893. (d) Abram, U.
Z. Anorg. Allg. Chem. 1999, 625, 839-841. (e) Heard, P. J.; Kite,
K.; Nielsen, J. S.; Tocher, D. A. J. Chem. Soc., Dalton Trans. 2000,
1349-1356.
(14) Wheeler, S. H.; Pignolet, L. H. Inorg. Chem. 1980, 19, 972-979.
(15) Shiu, K.-B.; Lin, S.-T.; Fung, D. W.; Chan, T.-J.; Peng, S.-M.; Cheng,
M.-C.; Chou, J. L. Inorg. Chem. 1995, 34, 854-863.
(10) Williams, G. A.; Smith, A. R. P. Aust. J. Chem. 1980, 33, 717-728.
(11) Butler, G.; Chatt, J.; Hussain, W.; Leigh, G. J.; Hughes, D. L. Inorg.
Chim. Acta 1978, 30, L287-L288.
(12) Lawrance, G. A. Chem. ReV. 1986, 86, 17-33.

6680 Inorganic Chemistry, Vol. 40, No. 26, 2001
Seymore and Brown
dissociated into paramagnetic monomers even at concentrations
greater than 10 mM (although the dissociation is slow on the
NMR time scale).14 In the molybdenum complexes, the integrity
of the dimers appears to persist at least for an hour or so in
solution. When solutions of the diethyldithiocarbamate dimer
4a and the dimethyldithiocarbamate dimer 4b are mixed in
acetone-d6, the spectrum initially shows signals only for 4a,b
and undergoes spectral changes (possibly indicating some
formation of a mixed ethyl/methyl dimer) only after several
hours at room temperature.
Reactivity of [Mo2(S2CNEt2)6](OTf)2. The dimeric complex
4a reacts with a variety of donor ligands and oxidizing agents
to form monomeric, pentagonal bipyramidal complexes of
molybdenum(IV) or -(VI). For example, [Ph3PdNdPPh3]Cl and
[Bu4N]I react readily and quantitatively over the course of about
1 h in dichloromethane to give the halide complexes MoCl(S2-
CNEt2)3 (2a) and MoI(S2CNEt2)3, respectively. Sodium azide
in acetonitrile reacts more slowly than the halides (∼60% reacted
after 24 h at room temperature), possibly because of the low
solubility of NaN3 in this solvent. The species that forms in
this case has a broad dithiocarbamato methyl hydrogen reso-
nance (δ 1.52) and paramagnetically shifted methylene hydro-
gens (δ 17.71). This species is assigned as the fluxional, 16-
electron azido complex Mo(N3)(S2CNEt2)3 analogous to the
halide complexes. Heating this complex in situ to 60 °C for 1
h completes the conversion of the azide to the nitrido complex
3a (eq 3), which is present in only trace amounts after 24 h at
room temperature. This is reminiscent of the literature prepara-
tion of 3 from the chloro complexes 2 and NaN3 in refluxing
methanol.6
Figure 3. SHELXTL plot (30% thermal ellipsoids) of one of the two
crystallographically independent cations in [Mo(PMe2Ph)(S2CNEt2)3]-
OTf (6).
orange over several hours on exposure to air and forming
diamagnetic [MoO(S2CNEt2)3]+.16
An exception to the generally slow rate of ligand substitution
reactions of 4a is its reaction with the nitrido complex 3a, which
rapidly (<5 min) cleaves the dimer in acetone-d6 or CD2Cl2
to form an equilibrium mixture which contains the starting
materials and a µ-nitrido dimer [(Et2NCS2)3Mo(µ-N)Mo(S2CN-
Et2)3]OTf (5a), as characterized by NMR and FABMS (eq 4).
1
The H NMR spectrum of 5a shows a single resonance at δ
17.94 for the dithiocarbamato methylene protons, indicating that
The molybdenum(IV) dimer [Mo2(S2CNEt2)6](OTf)2 also
reacts with a variety of neutral donors to form paramagnetic,
fluxional adducts. For example, 4a reacts quantitatively in
acetone with PMe2Ph within 2.5 h to form the phosphine adduct
[Mo(PMe2Ph)(S2CNEt2)3]OTf (6). PPh3 reacts to form a similar
adduct over the course of 1 day. The PCH3 signals of 6 appear
upfield in the 1H NMR at δ -18.15, while the phenyl resonances
are shifted downfield, as are the methylene protons of the
dithiocarbamate ligands, appearing as a single peak at δ 35.39
ppm. The phosphine adduct 6 is soluble in chlorinated organic
solvents and is insoluble in ether, hexane, and benzene.
The structure of 6, as determined by X-ray diffraction, is
shown in Figure 3. The two crystallographically independent
cations, which differ only in the conformation of one ethyl group
and the orientation of the PMe2Ph ligand, adopt a pentagonal
bipyramidal structure, as is invariably seen for seven-coordinate
molybdenum tris(dithiocarbamato) complexes. The Sax-Mo-P
angles (170° average) are essentially linear. The Mo-Sax bond
distances (2.485 Å) are normal, indicating that the phosphine
exerts a negligible trans influence.
the Mo(S2CNEt2)3 units are fluxional and that the cation is
symmetrical on the NMR time scale. The equilibrium constant
for eq 4 has been determined by NMR at several concentrations
at 25 °C (eq 5).
Harder neutral donors also react with 4a to form paramag-
netic, fluxional monomers, albeit less favorably than do phos-
phines. For example, only 75% conversion to [Mo(NCCD3)(S2-
CNEt2)3]OTf is seen even after 20 h in neat CD3CN. No
complexation is observed in neat acetone or with water in
acetone-d6. The triflate complex is sensitive to oxygen, turning
[5a]2
[3a]2[4a]
K4 )
) 26 ( 4 M-1
(5)
(16) Young, C. G.; Broomhead, J. A.; Boreham, C. J. J. Chem. Soc., Dalton
Trans. 1983, 2135-2138.

Metal-Metal-Bonded [Mo2(S2CNR2)6](OTf)2
Inorganic Chemistry, Vol. 40, No. 26, 2001 6681
Figure 5. Kinetics of ligand substitution of 4a by PMe2Ph in CH2Cl2
at 25.3 °C. Reaction conditions are the following: (a) [4a] ) 0.2 mM,
[PMe2Ph] ) 17.4 mM (solid squares); (b) [4a] ) 0.1 mM, [PMe2Ph]
) 17.4 mM (open squares); (c) [4a] ) 0.13 mM, [PMe2Ph] ) 102.8
mM (solid triangles).
Figure 4. Observed pseudo-first-order rate constants for cleavage of
[4a](OTf)2 by [Ph3PdNdPPh3]Cl or [Bu4N]I in the presence of [Bu4N]-
PF6 as a function of halide concentration. The total salt concentration
is 0.1 M for all runs (CH2Cl2, 25.3 °C).
Analogous studies with the neutral ligand PMe2Ph give
qualitatively different results. Again, NMR and UV-vis data
suggest a clean 4a f 6 transformation (e.g., an isosbestic point
is observed at 441 nm). However, reactions with a large excess
of PMe2Ph do not show first-order decay of 4a, instead
reproducibly giving absorbance vs time profiles as shown in
Figure 5. The plots show a linear decay accounting for ∼85%
of the reaction, followed by a decelerating phase. The slope of
the “pseudo-zero-order” phase is proportional to the 4a con-
centration (compare runs a and b in Figure 5). There is no
induction period; if anything, the reaction shows a very slight
deceleration before entering the linear phase. It appears unlikely
that this behavior is due to the presence of adventitious
impurities in 4a, since different batches of the dimer, prepared
by halide abstraction using AgOTf or TlOTf, give identical
results. Similar behavior is seen when acetone is used as solvent
rather than dichloromethane.
A reaction analogous to eq 4 occurs with the dimethyldithio-
carbamato complexes 3b and 4b to form a Me2NCS2-ligated
dimer (5b). Cross reactions also occur; thus 3a and 4b, or 3b
and 4a, will react to form 5a,b and the mixed dimer [(Me2-
1
NCS2)3Mo(µ-N)Mo(S2CNEt2)3]OTf (5c). The H NMR of the
mixed species shows a broad methyl resonance at δ 1.52 and
paramagnetically shifted methyl and methylene resonances at
δ 26.29 and 18.90, respectively.
Kinetics of Cleavage of 4a. The rates of ligand substitution
of 4a were examined using UV-visible spectroscopy in
dichloromethane at 25.3 °C by monitoring the decrease in
absorbance of 4a at 790 nm. In the presence of both iodide and
chloride (in excess), 4a undergoes clean first-order decay to
the corresponding monomeric halide complexes. There is no
sign of intermediates in this reaction by NMR or UV-visible
spectroscopy (for example, the 4a f 2a transformation is
accompanied by an isosbestic point at 384 nm). If the total
concentration of salts is kept constant using [Bu4N]PF6 as an
inert electrolyte, the observed rate constants increase linearly
with increasing concentration of halide for both chloride and
iodide (Figure 4). A nonzero intercept is observed, which is
the same within experimental error for both halides [(8.4 ( 1.3)
× 10-5 s-1 for Cl-; (8.9 ( 0.3) × 10-5 s-1 for I-; total salt
concentration of 0.1 M]. If the reaction is studied as a function
of concentration of halide without adding electrolyte to maintain
the salt concentration, however, there is only a slight dependence
of the observed rate constant on halide concentration. The
observed rate constants actually decrease somewhat with
increasing chloride concentration in the absence of added
electrolyte, ranging from a high of 5.22(10) × 10-4 s-1 at 0.016
M [PPN]Cl and leveling off at ∼3.7 × 10-4 s-1 [PPN]Cl
concentrations over 0.05 M (Supporting Information, Figure S1).
This pattern of ligand substitution rates that are linearly
dependent on concentration of entering ligand at constant salt
concentration but insensitive to ligand concentration in the
absence of inert salt has been observed previously in ligand
substitution reactions of cationic metal complexes with anionic
ligands.17,18 It is most consistent with substitution taking place
within contact ion pairs, as expected in nonpolar dichloro-
methane, where adding chloride cannot increase the concentra-
tion of the ion pair, but inert electrolyte can decrease the fraction
of reactive anion in the ion pairs.19
Discussion
Bonding in [Mo2(S2CNR2)6](OTf)2 (4). The monomeric,
paramagnetic molybdenum tris(dithiocarbamato) chloride com-
plexes 25 undergo facile halide abstraction with silver or thallium
triflate. The products of halide abstraction are not monomeric
cations or triflate complexes but rather the diamagnetic dimers
[Mo2(S2CNR2)6](OTf)2 (4), where the axial site vacated by
halide has been filled by a bridging sulfur atom from an
equatorial dithiocarbamate ligand of another molybdenum
center. The structures of 4, as illustrated by the solid-state
structure of 4a (Figure 2), consist of two molybdenum atoms
each surrounded by a roughly pentagonal bipyramidal array of
sulfur atoms, with the two coordination polyhedra sharing an
axial edge.
The gross structure of 4a is strikingly similar to that of the
dicationic osmium analogue [Os2(S2CNEt2)6](PF6)2 (4Os) de-
scribed by Wheeler and Pignolet.14 Only one major discrepancy
mars the remarkable agreement between the two structures,
namely the metal-metal separation, which is clearly nonbonding
in 4Os [Os-Os ) 3.682(1) Å] and clearly bonding in 4a [Mo-
Mo ) 2.8462(8)]. The chemistry of the molybdenum dimer also
shows that it has greater cohesion than its osmium analogue.
4Os is fluxional on the NMR time scale, in equilibrium with
detectable amounts of monomeric [Os(S2CNEt2)3]+, and reacts
rapidly and completely with donor solvents such as CH3CN to
(17) (a) Romeo, R.; Arena, G.; Scolaro, L. M.; Plutino, M. R. Inorg. Chim.
Acta 1995, 240, 81-92. (b) Alibrandi, G.; Romeo, R.; Scolaro, L.
M.; Tobe, M. L. Inorg. Chem. 1992, 31, 5061-5066.
(19) Loupy, A.; Tchoubar, B.; Astruc, D. Chem. ReV. 1992, 92, 1141-
(18) Song, L.; Trogler, W. C. J. Am. Chem. Soc. 1992, 114, 3355-3361.
1165.

6682 Inorganic Chemistry, Vol. 40, No. 26, 2001
Seymore and Brown
dimer (center of Figure 6). If the reasonable assumption is made
that the metal-sulfur interaction is much stronger than the
metal-metal interaction, then the energy gap between the
orbitals will be large and the mixing between them small, and
only the Mo-S nonbonding e2′′ orbitals of the PBP Mo
fragments need to be considered (to a first approximation) in
the description of the metal-metal bonding.
The molybdenum(IV) dimers 4 have only C2 symmetry, but
if the chelation of the dithiocarbamates is ignored, then the
Mo2S12 core has nearly C2h symmetry, with the σh plane defined
by the Mo2S2 core. Two of the Mo-S nonbonding orbitals
interact in a σ fashion, which creates the Mo1-Mo2 σ bonding
and antibonding orbitals of ag and bu symmetry. The other two
orbitals produce another bonding/antibonding pair, of au and bg
symmetry, respectively (Figure 6). The latter pair are nominally
of π symmetry, but because of the constraints of the PBP ligand
framework, the orbitals are roughly parallel to each other rather
than pointed at each other. This “skewed-π” interaction could
equally well be described as a sort of slipped δ bond. A δ
interaction would normally be expected to be negligible at 2.85
Å, but the skewing of the orbitals toward each other is expected
to increase their overlap, perhaps enough to resemble that seen
in a more typical δ bond in a complex with a much shorter
metal-metal separation. A very similar bonding scheme was
suggested for Mo2(OiPr)8, which consists of two trigonal
bipyramids sharing an axial edge.23 The much shorter Mo-Mo
bond in Mo2(OiPr)8 (2.523 Å) than in 4a is probably largely
due to differences in the bridging ligands (compare the non-
bonded Mo-Mo distance of 3.335 Å in Mo2(NO)2(OiPr)4(µ-
OiPr)224 with that of 3.773 Å in Mo2(CO)4(NO)2(µ-S2CNEt2)215).
The analogy with δ bonding is strongly supported by the
optical spectrum of 4a (Figure 1). In particular, the lowest
energy band, at 817 nm (12 200 cm-1), is assigned to the au f
bg transition. The modest intensity of this dipole-allowed band
(ꢀ ) 1200 M-1 cm-1) is reminiscent of the low intensities typical
of δ f δ* transitions in metal-metal quadruply bonded dimers,
where the nominally allowed transitions are quite weak by virtue
of the small overlap between the metal d orbitals.25 The complex
optical spectrum of 4a, with multiple shoulders in the 500-
800 nm range and two strong bands at higher energy, also
appears to be qualitatively consistent with this four-orbital
picture.
Figure 6. Qualitative molecular orbital diagram for [Mo2(S2CNR2)6]-
(OTf)2 (4). Nonbridging equatorial ligands in 4 are omitted for clarity.
form monomeric seven-coordinate adducts.14 The molybdenum
dimer 4a shows six inequivalent alkyl groups on the NMR time
scale, shows no tendency to form monomers in noncoordinating
solvents, and forms monomeric adducts only slowly and in-
completely in acetonitrile. Thus, the molybdenum-molybdenum
bond in 4a appears to have both structural and chemical conse-
quences. It is not obvious, however, why 4a should be dia-
magnetic, since the Mo-Mo separation in the d2-d2 dimer is
typical of a single bond20 and is almost 0.05 Å longer than the
longest double bond length we are aware of in a symmetrical21
dimolybdenum complex, 2.798 Å in (MeCp)2Mo2(CO)4(µ-
fluorenylidene).22 Although the bond length/bond order cor-
relation must be treated with considerable caution,7 the extreme
length of the bond seemed to merit more careful analysis.
The qualitative molecular orbital diagram illustrated in Figure
6 clarifies the initially puzzling aspects of the bonding of 4a. If
one first considers the metal-ligand interactions in each
pentagonal bipyramidal (PBP) Mo fragment (idealized D5h
symmetry), the orbitals split into a set of three with Mo-S σ*
In short, the bonding in dimeric 4 appears to be well-described
as a metal-metal double bond consisting of a reasonably strong
σ bond and a rather weak δ-like “skewed-π” bond. The latter
interaction splits the au and bg orbitals enough to pair the
electrons and render the complex diamagnetic but is expected
to contribute only slightly to the metal-metal attraction. Hence,
the double bond is quite long, with a distance more typical of
a molybdenum-molybdenum single bond.
2
2
2
character (dxy, dx -y , dz ) and a degenerate pair that are π with
respect to the ligands (dxz, dyz). Tipping the axes of the two
PBP Mo fragments allows their assembly into the edge-shared
Mechanism of Ligand Substitution in [Mo2(S2CNEt2)6]-
(OTf)2. Given the presence of a metal-metal bond in 4, it was
not obvious whether these dimeric complexes would serve as
useful sources of the {Mo(S2CNR2)3}+ fragment. In fact,
addition of ligands and formation of seven-coordinate monomers
from 4a takes place readily with a variety of reagents, including
halides, phosphines, azide (ultimately forming the terminal
nitride complex 3a), and oxygen (forming a Mo(VI) oxo
complex). Qualitatively, rates span a considerable range, with
reaction of 4a with MoN(S2CNEt2)3 (3a) coming to equilibrium
(20) Cotton, F. A. J. Less-Common Met. 1977, 54, 3-12.
(21) Longer Mo-Mo distances (up to 2.94 Å) have been assigned as double
bonds in unsymmetrical phosphido-bridged dimolybdenum complexes
on the basis of simple electron-counting considerations, but the
asymmetry in these complexes also allows their formulation as having
a single bond with a pair of electrons in a localized nonbonding orbital.
(a) Endrich, K.; Korswagen, R.; Zahn, T.; Ziegler, M. L. Angew.
Chem., Int. Ed. Engl. 1982, 21, 919. (b) Adatia, T.; McPartlin, M.;
Mays, M. J.; Morris, M. J.; Raithby, P. R. J. Chem. Soc., Dalton Trans.
1989, 1555-1564. (c) Conole, G.; McPartlin, M.; Mays, M. J.; Morris,
M. J. J. Chem. Soc., Dalton Trans. 1990, 2359-2366. (d) Adams,
H.; Bailey, N. A.; Hempstead, P. D.; Morris, M. J.; Riley, S.; Beddoes,
R. L.; Cook, E. S. J. Chem. Soc., Dalton Trans. 1993, 91-100. (e)
Stichbury, J. C.; Mays, M. J.; Davies, J. E.; Raithby, P. R.; Shields,
G. P.; Finch, A. G. Inorg. Chim. Acta 1997, 262, 9-20.
(23) Chisholm, M. H.; Cotton, F. A.; Extine, M. W.; Reichert, W. W. Inorg.
Chem. 1978, 17, 2944-2946.
(24) Chisholm, M. H.; Cotton, F. A.; Extine, M. W.; Kelly, R. L. J. Am.
Chem. Soc. 1978, 100, 3354-3358.
(22) Curtis, M. D.; Messerle, L.; D’Errico, J. J.; Solis, H. E.; Barcelo, I.
D.; Butler, W. M. J. Am. Chem. Soc. 1987, 109, 3603-3616.
(25) Trogler, W. C.; Gray, H. B. Acc. Chem. Res. 1978, 11, 232-239.

Metal-Metal-Bonded [Mo2(S2CNR2)6](OTf)2
Inorganic Chemistry, Vol. 40, No. 26, 2001 6683
within minutes (eq 4), while equilibration with neat CD3CN
takes about 1 day.
However, the puzzling kinetics of the reaction of 4a with
PMe2Ph (Figure 5) should inject a note of caution into any
discussion of the mechanism of reaction of 4a with neutral
donors. In these reactions, decay of 4a is dominated by a linear
phase which accounts for ∼85% of the reaction and whose rate
is proportional to the initial concentration of 4a. While many
other dimeric systems show zero-order kinetics in entering
ligand or reagent,27-30 in no case that we know of is the reaction
zero-order in the metal complex! Indeed, no combination of
rate constants in a general scheme for ligand-induced mono-
merization (such as that given by Espenson31) will give a rate
law that is independent of [4a] over the course of the reaction.
The reaction rates with PMe2Ph do increase with increasing
phosphine concentration, suggesting that this reaction is at least
in part associative, but beyond that it is not at all clear what
detailed mechanism is operating in this case.
The rates of reaction of 4a with chloride and iodide ions are
well-behaved and allow two important mechanistic conclusions
to be drawn. First, the rate of reaction does not increase signifi-
cantly with increasing overall chloride concentration (in the
absence of other ions). This indicates that attack on [Mo2(S2-
CNEt2)6]2+ by external chloride does not contribute significantly
to the rate of the reaction; in other words, the chloride that is
involved in the reaction is ion-paired with the molybdenum
complex. This behavior appears to be typical of anionic nucleo-
philes reacting with cationic metal complexes in CH2Cl2.17,18
If the total quantity of ions is fixed, the ligand substitution rate
does increase linearly with increasing fraction of chloride or
iodide. Reaction rates for both halides extrapolate to the same
positive intercept at zero halide concentration. This suggests a
two-term rate law in this reaction (eq 6) and hence two
independent mechanisms for the reaction.
Conclusions
Halide abstraction from MoCl(S2CNR2)3 provides ready
access to the new binuclear complexes [Mo2(S2CNR2)6](OTf)2
(4). An X-ray structural study of 4a (R ) Et) reveals that the
dimer possesses an unusual edge-shared pentagonal bipyramidal
structure with a Mo2S2 core. The diamagnetism and optical
spectra are most consistent with a Mo-Mo bond order of 2,
though the Mo-Mo distance of 2.85 Å is more typical of Mo-
Mo single bonds. This is rationalized by the description of the
double bond as consisting of a normal σ bond and a weak
“skewed π” bond similar to a normal δ bond. In its ligand
substitution chemistry 4a acts as a source of the [Mo(S2-
CNEt2)3]+ fragment; it reacts with azide to form MoN(S2-
CNEt2)3 (3a) and with PMe2Ph to form the new complex
[Mo(PMe2Ph)(S2CNEt2)3]OTf (6), which has been structurally
characterized. 4a can react by both associative and dissociative
pathways. Associative pathways appear to be preferred, although
the mechanisms can be complex and may vary with the nature
of the incoming ligand.
[X-]
[X-] + [PF6 ]
kdiss ) 8.7(8) × 10-5 s-1
kCl ) 2.6(3) × 10-4 s-1
kI ) 9.8(4) × 10-5 s-1
k ) kdiss + kX
(6)
-
At low fraction of halide, a halide-independent mechanism
predominates. This is most likely simple dissociation of [Mo2(S2-
CNEt2)6]2+ into monomeric [Mo(S2CNEt2)3]+, although as-
sistance by CH2Cl2 or PF6- in this dissociation cannot be ruled
out. With increasing amounts of halide, attack by halide on the
intact dimer becomes more important, with chloride reacting
about three times more efficiently by this route than iodide.
The linear dependence of rate on halide fraction indicates that
addition of only one halide ion is sufficient to lead irreversibly
to dimer scission.
The presence of an associative, as well as a dissociative,
reaction pathway is consistent with the wide variation in rates
of reaction of 4a with nucleophiles. The ligands can be arranged
qualitatively in order of decreasing rate of reaction with 4a as
follows:
Acknowledgment. We thank Dr. Alicia Beatty for her
assistance with the X-ray structures, Prof. W. Robert Scheidt
for the use of his near-IR spectrophotometer, and Prof. A.
Graham Lappin for helpful discussions regarding the dimer
cleavage kinetics. Support from the National Science Foundation
(Grant CHE-97-33321-CAREER), the Camille and Henry
Dreyfus Foundation (New Professor Award), DuPont (Young
Professor Award), the Dow Chemical Co. (Innovation Recogni-
tion Program), and the Arthur J. Schmitt Foundation (fellowship
to S.B.S.) is gratefully acknowledged.
(R2NCS2)3MoN . Cl- ≈ PMe2Ph > PPh3 . CH3CN
This ordering correlates roughly with ligand nucleophilicity, with
the glaring exception of the molybdenum nitride 3. Even though
(Et2NCS2)3MoN (3a) is commonly described as one of the most
nucleophilic nitrides known,6,26 it is difficult to see how it could
be more nucleophilic than PMe2Ph in any simple displacement
reaction. Certainly the kinetically facile reaction of 3 is not
favored by a large thermodynamic driving force, as reaction of
3 with 4a achieves a measurable equilibrium (K4 ) 24 M-1),
while the much more slowly reacting phosphines form only
adducts at equilibrium. Possibly the ability of the Mo(VI) nitride
3 to accept electrons from the Mo(IV) dimer 4 in forming the
Mo(V)-Mo(V) product 5 may help to break up 4, giving the
nitride 3 a kinetic advantage over redox-inert ligands.
Supporting Information Available: A figure showing the of
variation of rates of chloride addition to 4a as a function of total [Cl-],
tables of crystallographic parameters, atomic coordinates, bond lengths
and angles, anisotropic thermal parameters, hydrogen coordinates for
4a‚4CHCl3 and 6, and CIF data. This material is available free of charge
via the Internet at http://pubs.acs.org.
IC010673Y
(27) Orino, H.; Shimura, M.; Yamamoto, N.; Okubo, N. Inorg. Chem. 1988,
27, 172-175 and references therein.
(28) Wilson, L. M.; Cannon, R. D. Inorg. Chem. 1985, 24, 4366-4371.
(29) Fawcett, J. P.; Poe, A.; Sharma, K. R. J. Am. Chem. Soc. 1976, 98,
1401-1407.
(30) Deal, K. A.; Burstyn, J. N. Inorg. Chem. 1996, 35, 2792-2798.
(31) (a) Lente, G.; Guzei, I. A.; Espenson, J. H. Inorg. Chem. 2000, 39,
1311-1319. (b) Lente, G.; Jacob, J.; Guzei, I. A.; Espenson, J. H.
Inorg. React. Mech. 2000, 2, 169-177.
(26) Bishop, M. W.; Chatt, J.; Dilworth, J. R.; Neaves, B. D.; Dahlstrom,
P.; Hyde, J.; Zubieta, J. J. Organomet. Chem. 1981, 213, 109-124.