Bis(dithiolene)molybdenum analogues relevant to the DMSO reductase enzyme family: Synthesis, structures, and oxygen atom transfer reactions and kinetics

Article abstract of DOI:10.1021/ja003546u

A series of dithiolene complexes of the general type [Mo^IV(QR′)(S₂C₂Me₂) ₂]^1- has been prepared and structurally characterized as possible structural and reactivity analogues of reduced sites of the enzymes DMSOR and TMAOR (QR' = PhO^-, 2-AdO^-, PrⁱO^-), dissimilatory nitrate reductase (QR' = 2-ADS^-), and formate dehydrogenase (QR' = 2-AdSe^-). The complexes are square pyramidal with the molybdenum atom positioned 0.74-0.80 A above the S₄ mean plane toward axial ligand QR'. In part on the basis of a recent clarification of the active site of oxidized Rhodobacter sphaeroides DMSOR (Li, H.-K.; Temple, C.; Rajagopalan, K, V.; Schindelin, H. J. Am. Chem. Soc. 2000, 122, 7673), we have adopted the minimal reaction paradigm Mo^IV + XO ? Mo^IVO + X involving desoxo Mo(IV), monooxo Mo(VI), and substrate/product XO/X for direct oxygen atom transfer of DMSOR and TMAOR enzymes. The [Mo(OR')(S₂C₂Me₂)₂]^I- species carry dithiolene and anionic oxygen ligands intended to simulate cofactor ligand and serinate binding in DMSOR and TMAOR catalytic sites. In systems with N-oxide and S-oxide substrates, the observed overall reaction sequence is [Mo^IV(OR')(S₂C₂Me₂) ₂]^I- + XO → [Mo^IVO(OR')(S₂C₂Me₂) ₂]^I- →[Mo^VO(S₂C₂Me₂) ₂]^I-. Direct oxo transfer in the first step has been proven by isotope labeling. The reactivity of [Mo(OPh)(S₂C₂Me₂)₂]^I- (1) has been the most extensively studied. In second-order reactions, 1 reduces DMSO and (CH₂)₄SO (k₂ ≈ 10^-6, 10^-4 M^-I s^-I; ΔS^? = -36, -39 eu) and Me₃NO (k₂ = 200 M^-I s^-I; ΔS^? = -21 eu) in acetonitrile at 298 K. Activation entropies indicate an associative transition state, which from relative rates and substrate properties is inferred to be concerted with X-O bond weakening and Mo-O bond making. The Mo^VIO product in the first step, such as [Mo^VIO(OR')(S₂C₂Me₂) ₂]^I-, is an intermediate in the overall reaction sequence, in as much as it is too unstable to isolate and decays by an internal redox process to a Mo^VO product, liberating an equimolar quantity of phenol. This research affords the first analogue reaction systems of biological N-oxide and S-oxide substrates that are based on desoxo Mo(IV) complexes with biologically relevant coordination. Oxo-transfer reactions in analogue systems are substantially slower than enzyme systems based on a k_cat/K_M criterion. An interpretation of this behavior requires more information on the rate-limiting step(s) in enzyme catalytic cycles. (2-Ad = 2-adamantyl, DMSOR = dimethyl sulfoxide reductase, TMAOR = trimethylamine N-oxide reductase).

Full text of DOI:10.1021/ja003546u

1920
J. Am. Chem. Soc. 2001, 123, 1920-1930
Bis(Dithiolene)molybdenum Analogues Relevant to the DMSO
Reductase Enzyme Family: Synthesis, Structures, and Oxygen Atom
Transfer Reactions and Kinetics
Booyong Shim Lim and R. H. Holm*
Contribution from the Department of Chemistry and Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138
ReceiVed September 29, 2000
Abstract: A series of dithiolene complexes of the general type [MoIV(QR′)(S2C2Me2)2]1- has been prepared
and structurally characterized as possible structural and reactivity analogues of reduced sites of the enzymes
DMSOR and TMAOR (QR′ ) PhO-, 2-AdO-, PriO-), dissimilatory nitrate reductase (QR′ ) 2-AdS-), and
formate dehydrogenase (QR′ ) 2-AdSe-). The complexes are square pyramidal with the molybdenum atom
positioned 0.74-0.80 Å above the S4 mean plane toward axial ligand QR′. In part on the basis of a recent
clarification of the active site of oxidized Rhodobacter sphaeroides DMSOR (Li, H.-K.; Temple, C.; Rajagopalan,
K. V.; Schindelin, H. J. Am. Chem. Soc. 2000, 122, 7673), we have adopted the minimal reaction paradigm
MoIV + XO a MoVIO + X involving desoxo Mo(IV), monooxo Mo(VI), and substrate/product XO/X for
direct oxygen atom transfer of DMSOR and TMAOR enzymes. The [Mo(OR′)(S2C2Me2)2]1- species carry
dithiolene and anionic oxygen ligands intended to simulate cofactor ligand and serinate binding in DMSOR
and TMAOR catalytic sites. In systems with N-oxide and S-oxide substrates, the observed overall reaction
sequence is [MoIV(OR′)(S2C2Me2)2]1- + XO f [MoVIO(OR′)(S2C2Me2)2]1- f [MoVO(S2C2Me2)2]1-. Direct
oxo transfer in the first step has been proven by isotope labeling. The reactivity of [Mo(OPh)(S2C2Me2)2]1-
(1) has been the most extensively studied. In second-order reactions, 1 reduces DMSO and (CH2)4SO (k2 ≈
10-6, 10-4 M-1 s-1; ∆S‡ ) -36, -39 eu) and Me3NO (k2 ) 200 M-1 s-1; ∆S‡ ) -21 eu) in acetonitrile at
298 K. Activation entropies indicate an associative transition state, which from relative rates and substrate
properties is inferred to be concerted with X-O bond weakening and Mo-O bond making. The MoVIO product
in the first step, such as [MoVIO(OR′)(S2C2Me2)2]1-, is an intermediate in the overall reaction sequence, inasmuch
as it is too unstable to isolate and decays by an internal redox process to a MoVO product, liberating an
equimolar quantity of phenol. This research affords the first analogue reaction systems of biological N-oxide
and S-oxide substrates that are based on desoxo Mo(IV) complexes with biologically relevant coordination.
Oxo-transfer reactions in analogue systems are substantially slower than enzyme systems based on a kcat/KM
criterion. An interpretation of this behavior requires more information on the rate-limiting step(s) in enzyme
catalytic cycles. (2-Ad ) 2-adamantyl, DMSOR ) dimethyl sulfoxide reductase, TMAOR ) trimethylamine
N-oxide reductase)
Introduction
structural and functional synthetic analogues. The structures and
reactivity of such analogues should be of considerable value in
interpreting corresponding properties of enzyme sites. In this
investigation, the enzymes of interest are DMSOR and TMAOR.
Crystal structures and molybdenum EXAFS results for Rs5-8
and Rc9-12 DMSOR, EXAFS data for Rs biotin sulfoxide
It is now well-established that all known hydroxylase/
oxotransferase molybdenum1-3 and tungsten4 enzymes contain
mononuclear catalytic sites in which the metal atom is coordi-
nated by one or two pterin-dithiolene cofactor ligands, as
illustrated in Figure 1. The Hille classification of molybdoen-
zymes recognizes three major families.1 Among them is the
DMSO reductase (DMSOR) family, which includes DMSOR
itself, trimethylamine N-oxide reductase (TMAOR), nitrate
reductase, and formate dehydrogenase. (These and other ab-
breviations are listed in the Chart.) The increasing availability
of structural results for catalytic sites in this family provides
much of the information necessary for the formulation of
(5) Schindelin, H.; Kisker, C.; Hilton, J.; Rajagopalan, K. V.; Rees, D.
C. Science 1996, 272, 1615-1621.
(6) Li, H.-K.; Temple, C.; Rajagopalan, K. V.; Schindelin, H. J. Am.
Chem. Soc. 2000, 122, 7673-7680.
(7) George, G. N.; Hilton, J.; Rajagopalan, K. V. J. Am. Chem. Soc.
1996, 118, 1113-117.
(8) George, G. N.; Hilton, J.; Temple, C.; Prince, R. C.; Rajagopalan,
K. V. J. Am. Chem. Soc. 1999, 121, 1256-1266.
(9) Schneider, F.; Lo¨we, J.; Huber, R.; Schindelin, H.; Kisker, C.;
Kna¨blein, J. J. Mol. Biol. 1996, 263, 53-69.
(1) Hille, R. Chem. ReV. 1996, 96, 2757-2816.
(2) Roma˜o, M. J.; Kna¨blein, J.; Huber, R.; Moura, J. J. G. Prog. Biophys.
Mol. Biol. 1997, 68, 121-144.
(10) McAlpine, A. S.; McEwan, A. G.; Shaw, A. L.; Bailey, S. J. Biol.
Inorg. Chem. 1997, 2, 690-701.
(11) McAlpine, A. S.; McEwan, A. G.; Bailey, S. J. Mol. Biol. 1998,
275, 613-623.
(3) Kisker, C.; Schindelin, H.; Rees, D. C. Annu. ReV. Biochem. 1997,
66, 233-267.
(12) Baugh, P. E.; Garner, C. D.; Charnock, J. M.; Collison, D.; Davies,
E. S.; McAlpine, A. S.; Bailey, S.; Lane, I.; Hanson, G. R.; McEwan, A.
G. J. Biol. Inorg. Chem. 1997, 2, 634-643.
(4) Johnson, M. K.; Rees, D. C.; Adams, M. W. W. Chem. ReV. 1996,
96, 2817-2839.
10.1021/ja003546u CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/10/2001

Bis(Dithiolene)molybdenum Analogues
J. Am. Chem. Soc., Vol. 123, No. 9, 2001 1921
ligand at ∼1.9 Å as Ser147. Possibly the same type of disorder
exists in the highly homologous Rc DMSOR, which has been
described as having a seven-coordinate dioxo site when oxidized
and a six-coordinate monooxo site when reduced,10-12 and in
Sm TMAOR, but the matter has not been clarified.
Presented in Figure 1 are schematic minimal formulations
of the oxidized and reduced sites of three enzymes in the
DMSOR family for which structural information is available.
In doing so, we adopt the six-coordinate structure of Rs
DMSOR, assume the indicated reduced site for dissimilatory
nitrate reductase, and do not include the selenosulfide-thiolate
ligand form proposed on the basis of EXAFS analysis of two
formate dehydrogenases.15,16 We take as the minimal reaction
paradigm for direct oxygen atom transfer MoIV + XO h MoVIO
+ X,17 involving desoxo Mo(IV) and monooxo Mo(VI). Direct
oxo transfer has been demonstrated for Rs DMSOR by 18O
labeling,18 and, on the basis of results that follow, is highly
probable for TMAOR. Recent research has resulted in the
synthesis of bis(dithiolene)Mo(IV,VI) complexes of the de-
sired structural types.17,19-21 Oxo-transfer reactivity of [MoIV-
(OSiButPh2)(bdt)2]1- with sulfoxides is sluggish and incomplete,
even at 50 °C, and with Me3NO, some conversion to
[MoVIO(OSiButPh2)(bdt)2]1- occurs, but the reaction is not
clean.19 Also, this complex is not reduced by sulfides. To
provide an electronic environment more closely related to that
apparently present at DMSOR and TMAOR catalytic sites, we
have turned to complexes that are derived from the dialkyl-
dithiolene 1,2-Me2C2S22-. In this work, we describe the
synthesis and structures of [Mo(QR)(S2C2Me2)2]1- (Q ) O, S,
Se), intended as analogues of desoxo Mo(IV) sites in the
DMSOR family (Figure 1), and the oxo-transfer reactivity of
complexes with axial anionic oxygen ligands. Elsewhere, we
have reported molybdenum K-edge spectra and EXAFS of these
complexes in order to facilitate site identification by X-ray
absorption spectroscopy.22
Figure 1. Depictions of simplified active-site structures of enzymes
in the DMSOR family and the structure of the pterin-dithiolene cofactor
(R absent or a nucleotide). Examples cited are crystallographic results
(see text); the oxo ligand may be protonated in certain cases. EXAFS
analyses indicate a selenosulfide-thiolate ligand in two FDH enzymes-
(not shown).
Chart 1. Designation of Complexes and Abbreviations
In view of the existence of molybdenum and tungsten iso-
enzymes,23,24 we have endeavored to develop tungsten dithiolene
17,25-27
chemistry
in parallel with that of molybdenum dithio-
lenes. The oxo-transfer reactivity of the related tungsten com-
(13) Temple, C. A.; George, G. N.; Hilton, J. C.; George, M. G.; Prince,
R. C.; Barber, M. J.; Rajagopalan, K. V. Biochemistry 2000, 39, 4046-
4052.
(14) Czjzek, M.; Dos Santos, J.-P.; Pommier, J.; Giordano, G.; Me´jean,
V.; Haser, R. J. Mol. Biol. 1999, 284, 435-447.
(15) George, G. N.; Colangelo, C. M.; Dong, J.; Scott, R. A.; Khangulov,
S. V.; Gladyshev, V. N.; Stadtman, T. C. J. Am. Chem. Soc. 1998, 120, 0,
1267-1273.
(16) George, G. N.; Costa, C.; Moura, J. J. G.; Moura, I. J. Am. Chem.
Soc. 1999, 121, 2625-2626.
(17) Lim, B. S.; Sung, K.-M.; Holm, R. H. J.Am. Chem. Soc. 2000, 122,
7410-7411.
(18) Schultz, B. E.; Holm, R. H.; Hille, R. J. Am. Chem. Soc. 1995,
117, 827-828.
(19) Donahue, J. P.; Goldsmith, C. R.; Nadiminti, U.; Holm, R. H. J.
Am. Chem. Soc. 1998, 120, 12869-12881.
(20) Musgrave, K. B.; Donahue, J. P.; Lorber, C.; Holm, R. H.; Hedman,
B.; Hodgson, K. O. J. Am. Chem. Soc. 1999, 121, 10297-10307.
(21) Lim, B. S.; Donahue, J. P.; Holm, R. H. Inorg. Chem. 2000, 39,
263-273.
(22) Musgrave, K. B.; Lim, B. S.; Sung, K.-M.; Holm, R. H.; Hedman,
B.; Hodgson, K. O. Inorg. Chem. 2000, 39, 5238-5247.
(23) Buc, J.; Santini, C.-L.; Giordani, R.; Czjzek, M.; Wu, L.-F.;
Giordano, G. Mol. Microbiol. 1999, 32, 159-168.
(24) Stewart, L. J.; Bailey, S.; Bennett, B.; Charnock, J. M.; Garner, C.
D.; McAlpine, A. S. J. Mol. Biol. 2000, 299, 593-600.
(25) Lorber, C.; Donahue, J. P.; Goddard, C. A.; Nordlander, E.; Holm,
R. H. J. Am. Chem. Soc. 1998, 120, 8112.
(26) Goddard, C. A.; Holm, R. H. Inorg. Chem. 1999, 38, 5389-5398.
(27) (a) Sung, K.-M.; Holm, R. H. Inorg. Chem. 2000, 39, 1275-1281.
(b) Sung, K.-M.; Holm, R. H. J. Am. Chem. Soc. 2001, 123, 1931-1943.
reductase,13 and a crystal structure of Sm TMAOR14 have been
reported. The structure of oxidized Rs DMSOR at 1.3 Å recently
determined by Li et al.6 clarifies an earlier structure at lower
resolution (2.2 Å5) by showing that the enzyme as crystallized
contains a disordered mixture of two types of sites: square
pyramidal dioxo [MoVIO2 (O‚Ser)(S2pd)] and distorted trigonal
prismatic monooxo [MoVIO(O‚Ser)(S2pd)2]. The latter is the
probable catalytic site. In the fully reduced enzymes, X-ray and
EXAFS data are consistent with the minimal formulation desoxo
[MoIV(O‚Ser)(S2pd)2] with possible additional coordination by
water or hydroxide. Crystallography has identified the oxygen

1922 J. Am. Chem. Soc., Vol. 123, No. 9, 2001
Lim and Holm
plexes [W(OR)(S2C2Me2)2]1- and their comparison to molyb-
denum dithiolenes is described in a separate report.27b
MoNOS4: C, 40.92; H, 5.00; N, 2.17; S, 19.86. Found: C, 40.78; H,
5.10; N, 2.26; S, 19.91.
(Et4N)[MoO(S2C2Ph2)2]. This compound was prepared by the
oxidation of (Et4N)2[MoO(S2C2Ph2)2] with iodine in a procedure
analogous to that for (Et4N)[MoO(S2C2Me2)2].21 The product was
isolated as black crystals (50%). IR (KBr): νMoO 926 cm-1. Absorption
spectrum (acetonitrile): λmax (ꢀM) 600 (sh, 715), 867 (3770) nm. Anal.
Calcd. for C36H40MoNOS4: C, 59.48; H, 5.55; N, 1.93; S, 17.64.
Found: C, 59.39; H, 5.62; N, 1.87; S, 17.55.
(Et4N)[Mo(OPri)(S2C2Me2)2]. A solution of LiOPri (22 mg, 0.33
mmol) in 1 mL of THF was added to a suspension of [Mo(CO)2(S2C2-
Me2)2] (125 mg, 0.32 mmol) in 2 mL of acetonitrile. The reaction
mixture was stirred overnight to produce a brown solution. A solution
of Et4NCl (55 mg, 0.33 mmol) in 1 mL of acetonitrile was added. The
mixture was stirred for 10 min and filtered; the filtrate was reduced to
dryness. The black solid residue was dissolved in a minimal volume
of acetonitrile. Several volume equivalents of ether were layered onto
the solution, and the mixture was allowed to stand for 2 days. The
product was isolated as a brown microcrystalline solid (85 mg, 51%).
1H NMR (anion, CD3CN): δ 0.84 (d, 6), 2.48 (s, 12), 3.73 (m, 1).
Absorption spectrum (acetonitrile): λmax (ꢀM) 264 (703), 325 (9980),
378 (4570), 460 (1380), 630 (sh, 305) nm. Anal. Calcd. for C19H39-
MoNOS4: C, 43.74; H, 7.53; N, 2.68; S, 24.58. Found: C, 43.66; H,
7.46; N, 2.64; S, 24.65.
Experimental Section
Preparation of Compounds. All reactions and manipulations were
conducted under a pure dinitrogen atmosphere using either an inert
atmosphere box or standard Schlenk techniques. Acetonitrile and
dichloromethane were freshly distilled from CaH2, and methanol was
distilled from magnesium; CD3CN, THF-d8, and Me2SO (Aldrich) were
dried over freshly activated molecular sieves. Ether and THF were
distilled from sodium/benzophenone and stored over 4 Å molecular
sieves. Tetramethylenesulfoxide (Fluka) was distilled from CaH2 under
reduced pressure before use. In the following preparations, all volume
reduction and evaporation steps were performed in vacuo. The
compounds NaOPh, (Et4N)(OC6F5), LiOPri, Li(2-AdO), and Li(2-AdS)
were prepared from the corresponding phenol, alcohol, thiol, or selenol
using NaOMe, Et4NOH (25% in MeOH), or BunLi (1.6 M in hexane).
The compound 2-AdSH was obtained by a published method.28
Elemental selenium was obtained from Lancaster.
Bis(2-adamantyl)diselenide. To a mixture of 2-bromoadamantane
(5.00 g, 0.023 mol) and Mg (9.00 g, 0.37 mol) was added 250 mL of
ether. The reaction was initiated by adding 0.1 mL of 1,2-dibromo-
ethane. The reaction mixture was refluxed overnight (without mechan-
ical stirring to avoid formation of 2,2′-biadamantane29), filtered, and
the filtrate was introduced into a flask containing selenium powder
(1.8 g, 0.023 mol). THF (50 mL) was added, and the mixture was stirred
overnight. The pale yellow solution was opened to the air, stirred for
another 6 h, and filtered. The filtrate was reduced to dryness to give a
yellow solid, which was dissolved in a minimal volume of hexanes
and chromatographed on silica gel with ether in pentane (0-2%). The
product was obtained as a yellow powder (3.5 g, 71%) after removal
of solvents: mp 215-217 °C. 13C NMR (CDCl3): 27.3, 27.9, 32.6,
(Et4N)[Mo(2-AdO)(S2C2Me2)2]. A solution of Li(2-AdO) (26 mg,
0.16 mmol) in 1 mL of THF was added to a solution of [Mo(CO)2-
(S2C2Me2)2] (64 mg, 0.16 mmol) in 2 mL of THF. The reaction mixture
was stirred overnight. To the brown solution was added a solution of
Et4NCl (27 mg, 0.16 mmol) in 2 mL of acetonitrile. The mixture was
stirred for 10 min; solvent was removed to give a black solid. The
solid was then dissolved in a minimal volume of THF and several
volume equivalents of ether were added. Large black needlelike crystals
of the product were collected, washed with ether, and dried (65 mg,
1
1
65%). H NMR (anion, CD3CN): δ 1.18-1.21 (d, br, 2), 1.44-1.70
34.0, 37.7, 39.2, 56.7. H NMR (CDCl3): δ 1.56-2.19 (m, 28), 3.64
(s, 2). EI-MS+: 430 (M+).
(m, br, 12), 2.46 (s, 12), 3.55 (t, br, 1). Absorption spectrum
(acetonitrile): λnm (ꢀM) 264 (8340), 327 (9000), 373 (3970), 453 (sh,
1690), 528 (sh, 864), 850 (sh, 389) nm. Anal. Calcd. for C26H47-
MoNOS4: C, 50.87; H, 7.72; N, 2.28; S, 20.89. Found: C, 50.84; H,
7.79; N, 2.30; S, 20.73.
(Et4N)[Mo(OPh)(S2C2Me2)2]. A suspension of NaOPh (24 mg, 0.21
mmol) in 2 mL of acetonitrile was added to a suspension of [Mo-
(CO)2(S2C2Me2)2]21 (54 mg, 0.20 mmol) in 2 mL of acetonitrile. The
mixture was stirred for 6 h to generate a brown solution. A solution of
Et4NCl (34 mg, 0.21 mmol) in 1 mL of acetonitrile was added. The
mixture was stirred for 10 min and filtered. The filtrate was reduced
to one-third of its original volume, several volume equivalents of ether
were layered onto the solution, and the mixture was allowed to stand
for 2 days. The product was obtained as black bar-type crystals (54
mg, 48%). 1H NMR (anion, CD3CN): δ 2.54 (s, 12), 6.38 (d, 2), 6.78-
(t, 1), 7.01 (m, 2). Absorption spectrum (acetonitrile): λnm (ꢀM) 275
(sh, 10900), 345 (12300), 395 (4100), 478 (1790), 565 (sh, 750), 730
(728) nm. Anal. Calcd. for C32H37MoNOS4: C, 47.55; H, 6.71; N, 2.52;
S, 23.08. Found: C, 47.48; H, 6.72; N, 2.50; S, 23.15.
(Et4N)[Mo(2-AdS)(S2C2Me2)2]. A solution of Li(2-AdS) (38 mg,
0.22 mmol) in 1 mL of THF was added to a suspension of [Mo(CO)2-
(S2C2Me2)2] (84 mg, 0.22 mmol) in 2 mL of acetonitrile. Upon stirring,
the solution became brown-green; the reaction mixture was stirred for
3 h. To the mixture was added a solution of Et4NCl (36 mg, 0.22 mmol)
in 2 mL of acetonitrile; stirring was continued for 10 min to give a
brown-green solution. Solvents were removed, leaving a red-brown
solid, which was dissolved in a minimal volume of acetonitrile. Several
volume equivalents of ether were introduced by vapor diffusion. The
1
product separated as black blocklike crystals (70 mg, 51%). H NMR
(anion, CD3CN): δ 1.22-1.79 (m, br, 14), 2.74 (s, 12), 2.82 (s, br, 1).
Absorption spectrum (acetonitrile): λmax (ꢀM) 278 (9030), 342 (15800),
447 (3610), 528 (sh, 1900) nm. Anal. Calcd. for C26H47MoNS5: C,
49.57; H, 7.52; N, 2.22; S, 25.45. Found: C, 49.64; H, 7.63; N, 2.26;
S, 25.54.
(Et4N)[Mo(OPh)(S2C2Ph2)2]. The preceding method was followed
but with use of [Mo(CO)2(S2C2Ph2)2].21 The product was obtained as
dark brown crystals (48%). 1H NMR (anion, CD3CN): δ 6.54 (m, 2),
6.83 (m, 1), 7.05 (m, 2), 7.19-7.38 (m, 20). Absorption spectrum
(acetonitrile): λmax (ꢀM) 320 (22700), 349 (29000), 400 (sh, 6470), 470
(sh, 2300), 560 (798), 690 (428) nm. Anal. Calcd. for C42H45-
MoNOS4: C, 62.73; H, 5.64; N, 1.74; S, 15.95. Found: C, 62.65; H,
5.71; N, 1.68; S, 15.97.
(Et4N)[Mo(2-AdSe)(S2C2Me2)2]. The preceding method was fol-
lowed but with use of 2-AdSe- anion generated in situ by treating (2-
AdSe)2 with LiBHEt3 (1 M in THF). The brown-green reaction solution
was reduced to dryness, and the solid residue was redissolved in 5 mL
of acetonitrile. The solvent was removed slowly in vacuo. This proce-
dure was repeated 10 times. The final solid material was washed with
acetonitrile (2 × 2 mL) to obtain a red-brown solid. The filtrate was
subject to the same procedure to afford additional red-brown solid.
(These tedious dissolution/solvent removal steps are required to promote
the removal of carbonyl groups from the starting material.) The com-
bined solids were dissolved in a minimal volume of acetonitrile to
generate a bright red-brown solution, which was filtered through Celite.
Several volume equivalents of ether were added to the filtrate. After
several days, the product was obtained as black needlelike crystals
(30%). 1H NMR(anion, CD3CN): δ 1.34-1.81 (m, br, 14), 2.79(s, 12),
3.29(s, br, 1). Absorption spectrum (acetonitrile): λmax (ꢀM) 293 (6730),
351 (15600), 396 (sh, 5570), 463 (sh, 2900), 530 (sh, 1640), 750 (sh,
(Et4N)[Mo(OC6F5)(S2C2Me2)2]. A solution of (Et4N)(OC6F5) (80
mg, 0.26 mmol) in 1 mL of THF was filtered through Celite and added
to a solution of [Mo(CO)2(S2C2Me2)2] (100 mg, 0.26 mmol) in 2 mL
of THF. The reaction mixture was stirred for 6 h to generate a brown
precipitate. This material was collected by filtration and washed with
THF(1 × 1 mL) and Et2O (2 × 2 mL) to afford the product as a brown
1
solid (75 mg, 45%). H NMR (anion, THF-d8): δ 2.57 (s, 12). 19F
NMR (THF-d8, BF3Et2O reference): -161 (d, 2), -170 (t, 2), -175
(t, 1). Absorption spectrum (THF): λmax (ꢀM) 358 (10700), 425 (4060),
502 (sh,1610), 594 (sh, 735), 740 (262) nm. Anal. Calcd. for C22H32F5-
(28) Greidanus, J. W. Can. J. Chem. 1970, 48, 3593-3597.
(29) Molle, G.; Bauer, P.; Dubois, J. E. J. Org. Chem. 1982, 47, 4120-
4128.

Bis(Dithiolene)molybdenum Analogues
J. Am. Chem. Soc., Vol. 123, No. 9, 2001 1923
Table 1. Crystallographic Dataa for Compounds Containing 2-7
(Et4N)[2]‚0.5MeCN
(Et4N)[3]‚THF
(Et4N)[4]
(Et4N)[5]
(Et4N)[6]‚0.3Et2O
(Et4N)[7]‚0.3Et2O
formula
fw
crystal system
space group
Z
C
43H46.5MoN1.5OS4
C26H40F5MoNO2S4
717.77
triclinic
P1h
2
C19H39MoNOS4
521.69
triclinic
Ph1
2
C26H47MoNOS4
613.87
orthorhombic
Pnma
C27.2H50MoNO0.3S5
652.12
orthorhombic
I4(1)/a
C27.2H50MoNO0.3S4Se
699.02
orthorhombic
I4(1)/a
824.50
monoclinic
Cc
8
4
16
16
a, Å
b, Å
c, Å
44.188(4)
9.8658(8)
19.673(2)
9.2550(9)
13.295(1)
13.569(1)
79.465(2)
81.938(2)
78.231(2)
1597.6(3)
1.492
9.3067(9)
11.514(1)
12.674(1)
87.551(2)
76.498(2)
71.353(2)
1250.5(2)
1.386
23.9500(6)
11.4467(3)
11.1152(3)
36.233(2)
36.233(2)
9.8980(6)
36.413(2)
36.413(2)
9.7755(6)
R, deg
â, deg
γ, deg
V, Å3
99.032(2)
8470(1)
1.293
0.540
1.87-28.30
1.037
3047.2(1)
1.338
0.723
1.70-25.00
1.053
12995(1)
1.333
0.743
1.12-28.31
1.078
12961(1)
1.433
1.808
1.12-22.49
1.069
d
calc, g/cm3
µ, mm-1
0.725
1.54-22.50
1.087
0.867
1.65-28.32
1.003
θ range, deg
GOF(F2)
R1b (wR2 ), %
4.23(8.86)
3.79(9.73)
3.49(8.01)
3.61(9.47)
4.34(9.99)
2.94(7.53)
c
a Obtained with graphite monochromated Mo KR (λ ) 0.71073 Å) radiation. b R1 ) ∑|Fo| - |Fc|/∑|Fo|. c wR2 ) {∑[w(Fo2 - Fc2)2]/∑[w(Fo )2]}1/2
.
2
69) nm. Anal. Calcd. for C26H47MoNS4Se: C, 46.14; H, 7.00; N, 2.07;
S, 18.95; Se, 11.67. Found: C, 46.21; H, 7.01; N, 2.13; S, 19.08; Se,
11.56.
Kinetics Measurements. Sample preparations and reactions were
performed under strictly anaerobic conditions in acetonitrile or THF
solutions. Reactions were monitored spectrophotometrically with a Cary
3 instrument equipped with a cell compartment thermostated to (0.5
°C. S-oxide substrate reduction reactions were followed by observing
the appearance of the 830-nm band of [MoVO(S2C2Me2)2]1- except for
the [Mo(OPh)(S2C2Ph2)2]1- system, in which disappearance of the 349-
nm band of this complex was monitored. N-oxide substrate reduction
reactions were monitored by the appearance of the 610-nm band of
[MoVIO(OPh)(S2C2Me2)2]1-. Decreases in intensity of the 345- and 327-
2-Thiaindane S-oxide-[18O]. This compound was prepared in a
manner similar to described procedures.30,31 A solution of bromine (0.52
mL, 0.01 mol) was added to a solution of 2-thiaindane32 (1.36 g, 0.01
mol) in 100 mL of hexane with stirring. Immediately, yellow 2-thia-
indane S-dibromide formed and was collected by filtration. The
dibromide was added in small portions with stirring to a mixture of
freshly distilled triethylamine (1.40 mL, 0.01 mol) and H2[18O] (95%,
0.25 mL, 0.012 mol) in 100 mL of dichloromethane. The yellow color
bleached as the reaction proceeded. After stirring for 3 h, the reac-
tion mixture was filtered, and the filtrate was reduced to dryness. The
residue was twice flash-chromatographed on silica gel with ethyl acetate
in hexane (0-90%). Removal of solvent afforded the product as 0.50
g (32%) of off-white solid. IR (KBr): ν(S[18O]) 998 cm-1, ν(S[16O])
1035 cm-1. 1H NMR (CDCl3): δ 4.05(d, 2), 4.21(d, 2), 7.24-7.31(m,
4). EI-CI mass spectrometry showed the product to contain 82% 18O.
18O Transfer Reactions. (a) with Ph2Se[18O]. (Et4N)[Mo(OPh)-
(S2C2Me2)2] (7.6 mg, 13.7 mmol) and Ph2Se[18O] (62%)25 (4.55 mg,
18.1 mmol) were stirred in 1 mL of acetonitrile for 3 h. The solvent
was removed in vacuo, and the residue was washed with ether (5 × 2
nm bands of [Mo(OPh)(S2C2Me2)2]1- and [Mo(OPri)(S2C2Me2)2]1-
,
respectively, were followed for hydrolysis reactions. Sharp isosbestic
points demonstrated that these oxo-transfer reactions proceeded
cleanly: [Mo(OPh)(S2C2Me2)2]1- f [MoO(OPh)(S2C2Me2)2]1- (372
nm), [MoO(S2C2Me2)2]1- (552 nm), [MoO(S2C2Me2)2]2- (739 nm);
[Mo(OPh)(S2C2Ph2)2]1- f [MoO(S2C2Ph2)2]1- (583 nm); [Mo(OC6F5)-
(S2C2Me2)2]1- f [MoO(S2C2Me2)2]1- (548 nm); [Mo(OPri)(S2C2Me2)2]1-
f [MoO(S2C2Me2)2]1- (305, 532 nm), [MoO(S2C2Me2)2]2- (299, 711
nm); [MoO(OPh)(S2C2Me2)2]1- f [MoO(S2C2Me2)2]1- (747 nm).
Reaction systems contained the initial concentrations [[Mo(OR′)-
(S2C2R2)2]1-]0 ) 0.314-1.11 mM and excess substrate: [Me2SO]0 )
8.46-10.1 M, [(CH2)4SO]0 ) 0.218-1.53 M, [H2O]0 ) 0.0914-0.261
M, [Me3NO]0 ) 0.973-1.31 mM, [(C6H4CH2)3NO]0 ) 1.11-1.62 mM.
For each substrate, reactions were run at a minimum of four
temperatures in the range of 288-338 K. At each temperature, at least
four runs were performed under pseudo-first-order conditions for
S-oxide reduction and hydrolysis reactions or under second-order
conditions for N-oxide reduction reactions. Plots of ln[(A - A∞)/(A0 -
A∞)] vs time (S-oxide reductions and hydrolysis reactions) or ln[1 +
{([N-oxide]0 - [Mo]0)(A0 - A∞)}/[Mo]0(A - A∞)] vs time (N-oxide
reductions) were linear over 3 half-lives. Linear plots of kobs vs substrate
concentration yielded the second-order rate constants, k2, for S-oxide
reduction and hydrolysis reactions. The second-order rate constants,
k2, for N-oxide reductions were determined as mean values from at
least five independent measurements at different concentrations. These
second-order rate constants, k2, were used to determine activation
parameters by means of Eyring plots. The first-order rate constant, k1,
for the reaction [MoO(OPh)(S2C2Me2)2]1- f [MoO(S2C2Me2)2]1- was
obtained as a mean value of two independent measurements in the range
[MoVI]0 ) 0.7-2.0 mM. Errors were estimated using linear least-squares
error analysis with uniform weighting of the data points33 or standard
deviations.
mL). IR (KBr): ν(Mo[18O]) 874 cm-1, ν(Mo[16O]) 917 cm-1
.
FAB-MS: 18O product incorporation 37%. (b) with C6H4(CH2)2S[18O]:
(Et4N)[Mo(OPh)(S2C2Me2)2] (2.35 mg, 4.23 mmol) and C6H4(CH2)2S-
[18O] (82%) (55.0 mg, 0.357 mmol) were stirred in 1 mL of acetonitrile
over 2 days at 50 °C. The solvent was removed in vacuo, and the residue
was washed with ether (5 × 5 mL). IR(KBr): ν(Mo[18O]) 875 cm-1
.
18
FAB-MS:
O product incorporation 30%.
Reactions of CO with (Et4N)[Mo(2-AdQ)(S2C2Me2)2] (Q ) O, S,
Se). Carbon monoxide was bubbled through a solution of (Et4N)[Mo-
(2-AdQ)(S2C2Me2)2] (Q ) O, 3.40 mM; Q ) S, 1.80 mM; Q ) Se,
5.39 mM) for 20 min. IR spectra of the resulting solutions showed νCO
at 1955, 1944, and 1952 cm-1 for Q ) O, S, and Se, respectively.
Absorption spectra were subsequently measured.
Detection of Tetrahydrothiophene. A mixture of (Et4N)[Mo(OPh)-
(S2C2Me2)2] (800 µL, 2.30 mM), tetramethylenesulfoxide (80 µL, 1.01
M), and toluene (1 µL, internal reference) was stirred at 50 °C for 8 h.
A small protion (1 µL) of the resulting solution was injected for gas
chromatography, and the peak profiles were analyzed by comparison
with those of authentic tetramethylenesulfoxide and tetrahydrothiophene
at the same temperature gradients (50-150 °C for 25 min). A mixture
of acetonitrile (800 µL), tetramethylenesulfoxide (80 µL), and toluene
(1 µL) without (Et4N)[Mo(OPh)(S2C2Me2)2] that was treated in the
same way as the above mixture showed no peak corresponding to
tetrahydrothiophene.
In the sections that follow, molybdenum complexes are numerically
designated following the Chart, which also includes abbreviations.
X-ray Structure Determinations. The six compounds listed in
Table 1 were structurally characterized by X-ray crystallography.
Crystals of (Et4N)[3]‚THF were grown from a saturated THF solution
held at -25 °C for several days. All other crystals were obtained by
(30) Drabowicz, J.; Midura, W.; Mikolajczyk, M. Synthesis 1979, 39-
40.
(31) Kreher, R. P.; Kalischko, J. Chem. Ber. 1991, 124, 645-654.
(32) Holland, H. L.; Turner, C. D.; Andreana, P. R.; Nguyen, D. Can. J.
Chem. 1999, 77, 463-471.
(33) Clifford, A. A. MultiVariate Error Analysis; John Wiley & Sons:
New York, 1973.

1924 J. Am. Chem. Soc., Vol. 123, No. 9, 2001
Lim and Holm
Table 3. Bond Distances (Å) and Angles (deg) for Complexes
Table 2. Bond Distances (Å) and Angles (deg) for Complexes
with Axial Oxygen Ligands
with Axial Thiolate and Selenolate Ligands
1a
2b
3
4
5
6
7
Mo-O1
Mo-Sb
C-Sb
1.898(5) 1.868(8) 1.933(3) 1.843(2) 1.839(3)
Mo-S5/Se1a
2.312(1)
2.315(3)
1.76(2)
1.32(1)
101.4(4)
0.746
129.4
2.439(1)
2.317(3)
1.76(1)
2.32(5)
1.77(2)
1.330(5) 1.35(3)
2.31(1)
1.77(2)
2.309(1) 2.325(8) 2.333(6)
1.765(2) 1.77(3) 1.776(1)
Mo-Sb
C-Sb
C-Cc
1.322(3) 1.321(6) 1.330(1)
C-Cc
1.324(1)
Mo-O1-C 134.7(4) 145(1)
128.1(2) 139.0(2) 120.0(4)d
Mo-S5/Se1-C9a
102.5(1)
0.740
129.8
δe
θf
0.795
125.7
0.798
125.5
0.752
128.3
0.783
126.7
0.762
128.6
δd
θe
a Ref 10. b Mean values. c Chelate ring. d Values taken from the
a Values taken from the major occupancy. b Mean values. c Chelate
ring. d Perpendicular displacement of Mo atom from S4 least-squares
plane. e Dihedral angle between MoS2 planes.
major occupancy of the disordered 2-adamantyl group. e Perpendicular
displacement of Mo atom from S4 least-squares plane. f Dihedral angle
between MoS2 planes.
Results and Discussion
vapor diffusion where the first component specified is the parent
solvent: (Et4N)[1], (Et4N)[2]‚0.5MeCN, (Et4N)[4], (Et4N)[6,7]‚0.3Et2O
from acetonitrile/ether; (Et4N)[5], from THF/ether. Crystals were
mounted on glass capillary fibers in grease and cooled in a stream of
dinitrogen (-60 °C). Diffraction data were obtained with a Siemens
(Bruker) SMART CCD area detector system using ω scans of 0.3°/
frame, 30-s frames such that 1271 frames were collected for a
hemisphere of data. The first 50 frames were re-collected at the end of
the data collections to monitor for crystal decay; no significant decay
was observed. Cell parameters were determined using SMART
software. Data reduction was performed with SAINT software, which
corrects for Lorentz polarization and decay. Absorption corrections
were applied by using the program SADABS. Space groups were
assigned by analysis of symmetry and observed systematic absences
determined by the program XPREP and by successful refinement of
the structures. All of the structures were solved by direct methods
with SHELXS and subsequently refined against all of the data by the
standard technique of full-matrix least-squares on F2 (SHELXL-97,
SHELXL-93).
Asymmetric units contain one ((Et4N)[3]‚THF,[4],[6]‚0.3Et2O,[7]‚
0.3Et2O]) or two ((Et4N)[2]‚0.5MeCN) formula weights, except for
(Et4N)[5], which contains one-half formula weight, owing to an imposed
mirror plane. The structure of (Et4N)[2]‚0.5MeCN was refined as a
racemic twin. Disordered atoms or groups were frequently found in
the structures. The 2-adamantyl group of (Et4N)[5] is disordered over
three different positions, one of which is generated by the imposed
mirror plane, and was refined with 0.5, 0.25, and 0.25 occupancy
factors. The selenium atom in (Et4N)[7]‚0.3Et2O is disordered over two
sites and was refined with occupancy factor 0.92. The 2-adamantyl
groups of (Et4N)[6,7]‚0.3Et2O are disordered over two sites and were
refined with occupancy factors 0.83 and 0.85, respectively. Methylene
groups of the cations in (Et4N)[4-7] are disordered over two sites and
were refined with site occupancy factors of 0.72, 0.51, 0.64, and 0.63,
respectively. The acetonitrile solvate molecule in (Et4N)[2])‚0.5MeCN
and the THF solvate molecule in (Et4N)[3])‚THF were found to be
disordered over two positions and were refined accordingly. Ether
solvate molecules in (Et4N)[6,7]‚0.3Et2O are found in a chainlike
arrangement over the lattice and were treated isotropically. All non-
hydrogen atoms other than in solvate molecules were described
anisotropically. In the final stages of refinement, hydrogen atoms were
added at idealized positions and refined as riding atoms with a uniform
value of Uiso. Final refined structures were examined for any overlooked
symmetry with the checking program PLATON. Crystallographic and
final agreement factors are included in Table 1. Metric parameters for
the structures are collected in Tables 2 and 3. Because of the large
quantity of data, mean values are given where feasible instead of
individual data. (See paragraph at the end of this article for available
Supporting Information.)
Synthesis and Strutures of Compounds. As shown previ-
ously, the complexes [Mo(CO)2(S2C2R2)2] (R ) Me, Ph) readily
undergo carbonyl substitution reactions to afford [Mo(OAr)-
(S2C2R2)2]1- and [Mo(CO)(QAr)(S2C2R2)2]1- (Q ) S, Se).21
The failure to expel both carbonyl groups is probably due to
longer Mo-Q bond distances and attendant decreased steric
interaction between carbonyl and thiolate or selenolate ligands.
The synthesis of new Mo(IV) complexes on the basis of
dicarbonyl precursors is outlined in Figure 2, together with
yields. Phenolate complexes 1-3 and alkoxide complexes 4 and
5 are readily prepared in stoichiometric reactions. Monocarbonyl
intermediates were not detected in the synthesis. Reaction of 5
with carbon monoxide indicates an equilibrium with this
complex and [Mo(CO)(2-AdO)(S2C2Me2)2]1- (νCO 1955 cm-1),
which is strongly disfavored. Axial ligands have been varied
in 1-5 to examine their effects on oxo-transfer reaction rates.
Reaction of [Mo(CO)2(S2C2Me2)2] with one equivalent of
2-AdS- or 2-AdSe- results in a color change from purple to
green-brown, followed by development of a bright red-brown
solution. By examining the reaction of 6 and 7 with carbon
monoxide, it was shown that the green-brown chromophore is
the intermediate [Mo(CO)(2-AdQ)(S2C2Me2)2]1- (Q ) S, νCO
1944 cm-1; Q ) Se, νCO 1952 cm-1), which is similar to the
monocarbonyls [Mo(CO)(QPh)(S2C2Me2)2]1- (Q ) S, Se),
which have been isolated but not decarbonylated.21 For reasons
not yet clear, the complexes [Mo(CO)(2-AdQ)(S2C2Me2)2]1-
are readily decarbonylated, whereas certain other species with
large bulky ligands, such as [Mo(CO)(SeC6H2-2,4,6-Pri3)-
(S2C2R2)2]1-, are not.21 The structures of 2-7 are shown in
Figures 3 and 4; selected metric data are collected in Tables 2
and 3. Together with previous data,19,21 the structural features
of five-coordinate bis(dithiolene)Mo(IV) complexes are now
well-established: square-pyramidal with mean Mo-S bond
lengths of 2.31-2.33 Å, displacements δ of the molybdenum
atom above the S4 mean plane of 0.74-0.80 Å, dihedral angles
θ between MS2 chelate ring planes of 126-130°, and chelate
ring C-C and C-S distances consistent with an enedithiolate
ligand oxidation state. The latter distances (Tables 2 and 3) may
be compared to CdC, 1.33 Å; C-S, 1.75 Å; and CdS, 1.67
Å.34 The axial Mo-O distances in 1-5 are somewhat variable
(Table 2), being longest with the least basic ligand (C6F5O-)
and shortest with the most basic ligands (PriO-, 2-AdO-). Note
that in the series [Mo(2-AdQ)(S2C2Me2)2]1-, the Mo-Se bond
distance in 7 is 0.60 Å longer than the Mo-O bond in 5,
providing a basis for relative carbonyl labilities in monocarbonyl
intermediates. The absorption spectra of four oxygen-ligated
complexes are compared in Figure 5 and those of 5-7, in Figure
Other Physical Measurements. Absorption spectra were recorded
1
with a Varian Cary 50 Bio spectrophotometer. H NMR spectra were
obtained with Varian Mercury 400/500 spectrometers. Infrared spectra
were measured with a Nicolet Nexus 470 Fourier transform IR
instrument. EPR spectra were obtained with a Bruker ESP-300
spectrometer. Gas chromatographic analysis was performed with a
Hewlett-Packard 5890 Series II gas chromatograph equipped with an
HP-5 cross-linked 5% silicone column.
(34) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen, A.
G.; Taylor, R. In International Tables of Crystallography; Wilson, A. J.
C., Ed.; Kluwer Academic Publishers: Boston, 1995; Section 9.5.

Bis(Dithiolene)molybdenum Analogues
J. Am. Chem. Soc., Vol. 123, No. 9, 2001 1925
Figure 2. Scheme showing the synthesis of five-coordinate bis(dithiolene)Mo(IV) complexes 1-7.
Figure 3. Structures of four complexes with axial anionic oxygen ligandss[Mo(OPri)(S2C2Me2)2]1- and [Mo(2-AdO)(S2C2Me2)2]1- (upper), and
[Mo(OC6F5)(S2C2Me2)2]1-, and [Mo(OPh)(S2C2Ph2)2]1- (lower)sshowing 50% probability ellipsoids and atom-labeling schemes. The major position
of the three disordered sites of 2-adamantyl group in [Mo(2-AdO)(S2C2Me2)2]1- is shown.
6. For at least the two highest energy features, the bands shift
to the blue in the order Q ) O > S > Se, indicating charge-
transfer transitions involving the axial ligands. Brief comparisons
with enzyme spectra have been made.21
The series [Mo(2-AdQ)(S2C2Me2)2]1- (Q ) O, S, Se) is the
first in which axial ligands have been varied in the manner of
reduced sites in the DMSOR enzyme family (Figure 1).
Complexes 6 and 7 are intended as representatives of the sites
of dissimilatory nitrate reductase and the Ser147Cys mutant
of Rs DMSOR8 and of formate dehydrogenase, respectively.
Their reactivity properties will be described elsewhere. Com-
plexes 1-5 are the closest approaches currently available to
the five-coordinate site in DMSOR and possibly TMAOR. The
oxo-transfer reactivity of this set of complexes is next con-
sidered.
Oxo Transfer Reactions and Kinetics. The reactions of three
types of oxo-donor substrates XO ) Se-oxide, N-oxide, and
S-oxide were investigated. As will be seen, the final product in
all three of the systems is the monooxo Mo(V) complex 9 or
10 arising from reaction sequence 1 in which the monooxo Mo-

1926 J. Am. Chem. Soc., Vol. 123, No. 9, 2001
Lim and Holm
terms of direct (or primary36) oxo-transfer reactions of all
substrates XO examined in this work. In these and subsequent
reactions, the blue-violet MoVO product is identified by its
characteristic absorption spectrum and, on occasion, by its EPR
spectrum with g ) 1.996 and <aMo> ) 28.85 G.
(b) N-oxides. Complexes 1-5 react with Me3NO, although
at varying rates. In solutions containing a 0.8-1.5 mM complex
and 1.0 equiv of N-oxide, formation of the MoVIO complex in
reaction sequence 1 is complete within 10 minutes for 1-3 and
after ∼25 min for 4. For 5, the reaction is 34% complete after
30 min; this relatively slow reaction is attributed to the steric
effects of the 2-adamantyl group. The reaction mixtures change
from brown (1-5) to green (typified by 8), and over longer
times to blue-violet (9) or green (10). The intervention of a
MoVIO intermediate was demonstrated with 8, whose tungsten
analogue [WO(OPh)(S2C2Me2)2]1- is stable under oxo-transfer
conditions.17,27b This complex has a methyl resonance at 2.19
ppm and prominent absorption bands at 514 and 637 nm. When
reaction 4, with complex 1 and Me3NO as substrate, is
monitored by 1H NMR, signals appear at 2.11 (Me3N) and 2.26
(Me) ppm. In the absorption spectrum, bands grow in at λmax
) 610 and 795 nm, red-shifted as compared to the bands of the
tungsten complex. These spectral changes are entirely consistent
with the formation of 8 which, despite numerous attempts, has
not been isolated in pure form. Similar observations are made
for reaction 4 involving 2-5. The systems [Mo(OR′)(S2C2-
Me2)2]1-/Me3NO (R′ ) Ph, Pri, 2-Ad), when monitored spec-
trophotometrically after 16 h, show 90-94% conversion of 1,
1
4, or 5 to 9. The H NMR spectra clearly demonstrate the
presence of phenol, 2-propanol, and 2-adamantanol. In reaction
system 4, 83% of bound phenoxide appears as phenol by NMR
integration. Thus, 8 and analogous complexes apparently decay
by an internal redox process 5 in which a phenoxyl or alkoxyl
radical is developed and then quenched by capture of a hydrogen
Figure 4. Structures of the complexes [Mo(2-AdS)(S2C2Me2)2]1-
(upper) and [Mo(2-AdSe)(S2C2Me2)2]1- (lower) with axial thiolate and
selenolate ligands, showing 50% probability ellipsoids and the atom-
labeling schemes. The major positions of two disordered sites of
2-adamantyl groups are shown.
[MoIV(OR′)(S2C2R2)2]1- + R′′3NO f
(VI) complex is formed by oxo transfer and decays to Mo(V).
All of the reactions were carried out in acetonitrile solutions
unless noted otherwise.
[MoVIO(OR′)(S2C2R2)2]1- + R′′3N (fast) (4)
[MoVIO(OR′)(S2C2R2)2]1-
f
[MoIV(OR′)(S2C2R2)2]1- + XO f
[MoVO(S2C2R2)2]1- + R′O‚ (slow) (5)
[MoVIO(OR′) (S2C2R2)2]1- f [MoVO(S2C2R2)2]1- (1)
atom from solvent or trace water. Considerable difficulty was
encountered in the earlier isolation of pure [MoVIO(OSiButPh2)-
(bdt)2]1-, which is extremely sensitive hydrolytically.19 Indeed,
complex 8 cannot be observed in situ unless the acetonitrile
solvent is rigorously dry, and its decomposition to 9 is
qualitatively more rapid in the presence of traces of water.
Reactions 4 with Me3NO, N-methylmorpholine N-oxide, and
(PhCH2)3NO were selected for kinetics study. With Me3NO as
a substrate, reactions of 2 and 3 are too rapid for convenient
spectrophotometric measurements, and reactions of 4 and 5
sufficiently slow that in sequence 1, formation of 9 is com-
petitive with oxo transfer. The time course of reaction 4 with
Me3NO is illustrated in Figure 7a. The intense absorption band
of 1 at 345 decreases, and those at 610 and 795 nm of 8
grow in, as time increases. A tight isosbestic point occurs at
372 nm, indicating that clean conversion of 1 to 8 without
substantial accumulation of 9. All reactions adhere to the sec-
ond-order rate law 6. Kinetics data for N-oxide reduction
reactions are presented in Table 4. Activation parameters were
obtained from Eyring plots. The rate constants are ∼10-200
M-1 s-1, with the slowest rate for the more hindered substrate.
(a) Direct Atom Transfer. The first step in reaction sequence
1 was examined using two [18O]-labeled substrates in reactions,
resulting in the conversions 2 and 3. Reaction 2 with 1.3 equivs
of selenoxide proceeds to completion within several hours at
room temperature. Because of the slow reactivity of sulfoxides
[Mo(OPh) (S2C2Me2)2]1- + Ph2Se[18O] (62%) f
[Mo[18O](S2C2Me2)2]1- (37%) (2)
[Mo(OPh) (S2C2Me2)2]1- + C6H4(CH2)2S[18O] (82%) f
[Mo[18O](S2C2Me2)2]1- (30%) (3)
(vide infra), reaction 3 was conducted for 2 days at 50 °C with
a large excess (84 equivs) of 2-thiaindane S-oxide. Earlier, we
had demonstrated direct oxo transfer to a MoIVO center using
Ph2S[18O].35 As in that experiment, some loss of label is
observed upon workup of product because of trace water
contamination and is probably abetted in reaction 3 by the
forcing conditions that were used. We interpret these results in
(35) Schultz, B. E.; Gheller, S. F.; Muetterties, M. C.; Scott, M. J.; Holm,
R. H. J. Am. Chem. Soc. 1993, 115, 2714-2722.
(36) Holm, R. H. Chem. ReV. 1987, 87, 1401-1449.

Bis(Dithiolene)molybdenum Analogues
J. Am. Chem. Soc., Vol. 123, No. 9, 2001 1927
Table 4. Rate Constants and Activation Parameters for Reactions 4 and 8
a Not measured. b Ref 17. c Calculated value from the Eyring equation. d Measured in THF. e Measured in neat substrate.
The negative entropies of activation indicate an associative
transition state consistent with a second-order reaction. The
conversion 8 f 9 is illustrated in Figure 7b. The 610-nm feature
-d[MoIV]/dt ) k2[MoIV][XO]
-d[MoVIO]/dt ) k1[MoVIO]
(6)
(7)
of 8 decreases in intensity as the 830 nm band of 9 increases.
The final spectrum is that of 9 measured separately. The reaction
follows first-order rate law 7. The observed rate constants k1/
[MoVIO]0 ≈ 0.087-0.10 M-1 s-1 at ambient temperature are
smaller than k2, thereby allowing both steps in sequence 1 to
be examined separately and accounting for the isosbestic points
in Figure 7.
(c) S-oxides. In reaction sequence 1, the relative rates of the
two steps are reversed, as compared to reactions with N-oxides.
Oxo-transfer reaction 8 is quite slow, even in neat DMSO or
tetramethylene S-oxide (TMSO), and reaction 5 is fast. Con-
sequently, the appearance of only the MoVO product is
measurable by spectrophotometry or EPR. In reaction 8 with
complex 1 tetramethylene sulfide was detected by gas chroma-
Figure 5. UV/vis spectra of [Mo(OC6H3Pri2)(S2C2Me2)2]1- (- ‚ -),
[Mo(OPh)(S2C2Me2)2]1- (-‚‚), [Mo(OPri)(S2C2Me2)2]1- (‚‚‚‚‚), and
[Mo(2-AdO)(S2C2Me2)2]1- (s) in acetonitrile; selected band maxima
are indicated.
s-1, are small compared to the values k1/[MoVIO]0 ≈ 0.089-
[MoIV(OR′)(S2C2R2)2]1- + R′′2SO f
0.31 M-1 s-1, supporting analysis of the data in terms of rate
[MoVIO(OR′)(S2C2R2)2]1- + R′′2S (slow) (8)
law 6. The large negative entropies of activation again indicate
TMSO
an associative mechanism. The rate constant ratio k2
/
DMSO
k2
≈ 115 for reactions of 1 must arise from the less
tography, thereby supplying additional evidence for direct oxo
transfer, and phenol was formed quantitatively. Complexes 2-4
also react with DMSO and TMSO. The reactions of complex 5
were very sluggish, owing to the steric bulk of the adamantyl
group, and were not pursued. Spectrophotometric monitoring
of reaction 8 with 1 and TMSO is shown in Figure 8. Over
time, absorption bands of 1 at 345, 396, and 475 nm decrease
in intensity, but those of 9 at 579 and 830 nm increase. A sharp
isosbestic point at 552 nm indicates no appreciable build-up of
intermediate 8.
hindered nature of TMSO in binding to molybdenum in the
associative transition state. Because of its increased reaction
rate, TMSO was used as substrate in all of the other reactions.
Reduction of the basicity of the dithiolene ligands by substitution
of phenyl for methyl has little effect on reaction rates; with
TMSO as substrate, k2(2)/k2(1) ) 2.3. Variation of axial ligands
increases the rate constant in the order C6F5 > Ph . Pri. To
achieve an appreciable rate for 4, its reaction was measured in
neat TMSO. Although not strictly comparable to other values,
the rate constant is a factor of 300 smaller than the value for 1
with the same substrate.
Rate constants and activation parameters for reaction 8 are
collected in Table 4. The rate constants k2 ≈ 10-3-10-6 M-1

1928 J. Am. Chem. Soc., Vol. 123, No. 9, 2001
Lim and Holm
Figure 6. UV/vis spectra of [Mo(2-AdO)(S2C2Me2)2]1- (- ‚‚ -),
[Mo(2-AdS)(S2C2Me2)2]1- (‚‚‚‚‚), and [Mo(2-AdSe)(S2C2Me2)2]1-
(s) in acetonitrile; selected band maxima are indicated.
Figure 8. UV/vis spectra for the reaction [Mo(OPh)(S2C2Me2)2]1-
+
(CH2)4SO f [MoVIO(OPh)(S2C2Me2)2]1- + (CH2)4S (slow) followed
by the reaction [MoVIO(OPh)(S2C2Me2)2]1- f [MoVO(S2C2Me2)2]1-
(fast) at room temperature. The system contained initially 1.05 mM
[MoIV(OPh)(S2C2Me2)2]1- and 2.69 M (CH2)4SO.
Table 5. Kinetics and Activation Parameters for Reaction 9
R′
k2298 (M-1 s-1
)
∆H‡ (kcal/mol)
∆S‡(eu)
Ph
H2O
D2O
H2O
1.2(2) × 10-2
4.7(1) × 10-3
1.1(1) × 10-2
10.0(2)
a
10.6(4)
-34(1)
a
-32(1)
Pri
a Not measured.
Figure 7. Spectrophotometric monitoring of reactions: (a) the sys-
tem[MoIV(OPh)(S2C2Me2)2]1- + Me3NO f [MoVIO(OPh)(S2C2Me2)2]1-
initially containing 0.744 mM [MoIV(OPh)(S2C2Me2)2]1- and 0.818 mM
Me3NO in acetonitrile at room temperature; (b) the conversion of
[MoVIO(OPh)(S2C2Me2)2]1- to [MoVO(S2C2Me2)2]1- in a system initially
containing 1.39 mM [MoIV(OPh)(S2C2Me2)2]1- and 1.45 mM Me3NO
in acetonitrile at room temperature.
(d) Hydrolysis. We have emphasized the instability of 8 to
internal redox reaction 5 and to traces of water. The initial
hydrolysis product preceding formation of 9 has not been
identified. Complexes 1-5 are also highly sensitive to water.
Even under conditions intended to be rigorously dry, reaction
9 can proceed, resulting in the formation of MoIVO complexes.
One of these, orange-brown 11, has been isolated and structur-
ally characterized.21 To differentiate kinetically hydrolysis from
oxo transfer, the reactions of 1 and 4 with water have been
investigated kinetically under pseudo-first-order conditions in
acetonitrile solutions. The reactions, monitored spectrophoto-
Figure 9. Summary of the oxo-transfer analogue reaction systems
based on [MoIV(OR′)(S2C2R2)2]1- (XO ) N-oxide, S-oxide), in-
cluding the conversion of the MoVIO intermediate to the MoVO final
product.
) 2.6 indicates that outright O-H bond dissociation is not
involved in the rate-determining step. With complex 1, reaction
9 is ∼104 slower than oxo transfer from Me3NO. It is, how-
ever, ∼102-104 faster than oxo transfer from S-oxides, em-
phasizing the requirement for anhydrous conditions when
studying reaction 8.
[MoIV(OR′)(S2C2Me2)2]1- + H2O f
[MoIVO(S2C2Me2)2]2- + R′OH + H+ (9)
metrically (not shown), proceed with well-defined isosbestic
points and were treated as second-order within the water
concentration range employed. The final spectra are that of 11
measured separately.21 Rates constants are the same for 1 and
(e) Mechanistic Aspects. Reaction sequence 1 is summarized
in Figure 9. We assume the structure of the MoVIO interme-
diate as being the same as [WO(OPh)(S2C2Me2)2]1-, which has
been isolated and structurally characterized.17,27b Reactions of
1 and related complexes with N-oxides and S-oxides share the
D
4 and are given in Table 5. The modest isotope effect k2H/k2

Bis(Dithiolene)molybdenum Analogues
J. Am. Chem. Soc., Vol. 123, No. 9, 2001 1929
common features of second-order kinetics, associative transi-
tion states, and appreciable enthalpies of activation (8-15
kcal/mol). These features are anticipated in earlier molybde-
num oxo-transfer systems,35-40 developed prior to proof of
pterin-dithiolene ligand binding to molybdenum. All utilize the
minimal reaction MoIVO + XO f MoVIO2 + X, not yet
known to be a physiological process, involving complexes
with sulfur ligands but not dithiolenes. Of these systems, that
based on the complex [MoIVO(ButL-NS2)2] is the most thor-
oughly analyzed.38 Certain considerations raised in that work
apply here.
DMSORs reduce both DMSO and Me3NO,49 as exemplified
by Ec DMSOR,50 whereas TMAORs such as the Ec enzyme
reduce N-oxides but do not reduce S-oxides, including DMSO
and TMSO, at a detectable rate.23 As shown here, complex 1 is
intrinsically competent to reduce both substrate types, although
at widely different rates (Table 4). Second, rates of analogue
systems and enzymes are markedly different. Second-order rate
constants under single-turnover conditions are unavailable for
enzyme systems. For comparison to enzymes, recourse is taken
to kcat/KM values, which are effectively second-order rate
constants for substrate reduction under steady state (Michaelis-
Menten) conditions. The following values have been reported
for the indicated substrates: Ec DMSOR, DMSO, 1.9 × 106
Reaction rates are highly substrate-dependent, with k2 values
spanning a range of 108. For the five reactions whose activa-
tion parameters were determined, T|∆S‡| is 40-53% of ∆G‡.
Activation free energies are, therefore, not dominated by
enthalpic contributions, but two such contributions are indicated
by relative rates. The much faster reactions of N-oxides than
S-oxides correlate with two substrate properties, gas-phase bond
dissociation energies and basicities. For Me3NO, D(N-O) )
61 kcal/mol has been obtained from calorimetric data,41 and
71 kcal/mol from a combination of calorimetric and com-
putational results.42 For DMSO and TMSO, D(S-O) ) 87
kcal/mol from thermodynamic data43 and 86 kcal/mol by
computation,44 respectively. The gas-phase proton affinity of
Me3NO (235 kcal/mol) is substantially larger than that of
DMSO (211 kcal/mol).45 Also, Me3NO is a stronger base in
aqueous solution (pKBH, 4.6)46 than is DMSO (pKBH, -1.8)47
and TMSO (pKBH, -1.3),48 although these results were not
all obtained by the same procedure. Thus, the fast rates of
reaction of substrate XO are associated with weaker X-O bonds
and higher basicity. Further, the fastest rate of TMSO reduc-
tion is found with 3, which carries the most electronegative
axial ligand, a feature that should promote substrate binding.
We infer a concerted transition state involving both X-O bond-
weakening and Mo-O bond-making with concomitant struc-
tural rearrangement to a configuration approaching that of the
MoVIO intermediate. The structure of the latter (e.g., 8) may be
safely assumed to be that of [WO(OPh)(S2C2Me2)2]1-, which
has a distorted cis-octahedral stereochemistry.17,27b The mech-
anism of oxo transfer will be considered in more detail in a
future report on the reactivity of [WIV(OR′)(S2C2R2)]1- com-
plexes, which have been examined with a wider range of sub-
strates.27b The WVIO products are stable, thereby obviating the
last step in the reaction scheme for molybdenum complexes
(Figure 9).
M-1 s-1 50,52
;
Ec TMAOR, Me3NO, 2.2 × 106 M-1 s-1 23
;
Sm
TMAOR, Me3NO, 7.1 × 106 M-1 s-1 51
.
Although comparison among enzymes and between enzymes
and analogue systems is obviously inexact, the conclusion is
inescapable that, in terms of second-order rate constants, the
reduced enzymes in in vitro-reconstituted systems are orders
of magnitude more reactive than Mo(IV) complexes such as 1.
Under steady-state conditions, all steps in a catalytic cycle
proceed at the same rate but not necessarily with the same rate
constant. There is little information on which stepssubstrate
binding, oxygen atom transfer from substrate to the molybdenum
center, product release, or proton-coupled electron transfers
controls the turnover rate of an enzyme. In an Rc DMSOR
system, formation of an enzyme-substrate complex is rate-
limiting at pH 5.5, whereas at pH 8.0, the oxo-transfer step is
limiting.49 Tight binding of both pterin dithiolene ligands and
serinate to molybdenum during the catalytic cycle of Rs
DMSOR is supported by resonance Raman spectra.53 We have
no reason to believe that the corresponding metal-ligand
binding does not obtain in the analogue reaction systems.
However, it is appropriate to note several of the many real or
possible differences between analogue and enzyme systems. The
enzymes possess a crevice or “funnel”,5,9-11,14 apparently for
efficacious passage of substrate from the enzyme surface and
binding at the catalytic site,5 a probable rate-enhancing feature.
The molybdenum cofactor is held in the protein structure by a
large number of hydrogen bonds involving the pterin nucleus
and phosphate-nucleotide side chains. These interactions, ad-
jacent residues, and any protein-induced structural features
presumably condition the active site in a manner favorable to
oxo transfer. The [MoVIO(O‚Ser)(S2pd)2] site in oxidized Rs
DMSOR is described by Li et al.6 as a distorted trigonal prism,
whereas the tungsten analogue of 8 is closer to the octahedral
limit. Possibly, the trigonal prismatic structure is influenced by
the protein and is nearer to the transition state geometry for
Relation to Enzymes. This work provides the first analogue
reaction systems for reduction of biologically relevant N-oxide
and S-oxide substrates by desoxo bis(dithiolene)-Mo(IV) com-
plexes with axial ligands that simulate serinate coordinate in
Rs DMSOR (Figure 1). The steps of substrate binding and
reduction by oxo transfer doubtless occur in the enzyme, as
shown for Rc DMSOR,49 and the transition state for oxo transfer
from substrate may have characteristics in common with the
analogue systems. Thereafter, interesting differences are evident,
albeit from a limited enzyme reactivity database.23,49-51 First,
(44) Jenks, W. S.; Matsunaga, N.; Gordon, M. J. Org. Chem. 1996, 61,
1275-1283.
(45) Hunter, E. P. L.; Lias, S. G. J. Phys. Chem. Ref. Data 1998, 27,
413-656, and references therein.
(46) Nyle´n, P. Z. Anorg. Allg. Chem. 1941, 246, 227-330.
(47) Landini, D.; Modena, G.; Scorrano, G.; Taddei, F. J. Am. Chem.
Soc. 1969, 91, 6703-6707.
(48) Curci, R.; Di Furia, F.; Levi, A.; Lucchini, V.; Scorrano, G. J. Chem.
Soc., Perkin II 1975, 341-344.
(49) Adams, B.; Smith, A. T.; Bailey, S.; McEwan, A. G.; Bray, R. C.
Biochemistry 1999, 38, 8501-8511.
(37) Holm, R. H. Coord. Chem. ReV. 1990, 100, 183-221.
(38) Schultz, B. E.; Holm, R. H. Inorg. Chem. 1993, 32, 4244-4248.
(39) Enemark, J. H.; Young, C. G. AdV. Inorg. Chem. 1994, 40,
1-88.
(40) Laughlin, L. J.; Young, C. G. Inorg. Chem. 1996, 35, 1050-1058.
(41) Acree, W. E., Jr.; Tucker, S. A.; Ribeiro da Silva, M. D. M. C.;
Matos, A. R.; Gonc¸alves, J. M.; Ribeiro da Silva, M. A. V. J. Chem.
Thermodyn. 1995, 27, 391-398.
(42) Haaland, A.; Thomassen, H.; Stenstrøm, Y. J. Mol. Struct. 1991,
263, 299-310
(43) Donahue, J. P.; Holm, R. H. Polyhedron 1993, 12, 571-589.
(50) Simala-Grant, J. L.; Weiner, J. H. Eur. J. Biochem. 1998, 251, 510-
515, and references therein.
(51) Santos, J.-P. D.; Iobbi-Nivol, C.; Couillant, C.; Giordano, G.;
Me´jean, V. J. Mol. Biol. 1998, 284, 421-433.
(52) We previously misquoted the second-order rate constant for oxida-
tion of Me2S-reduced DMSOR by DMSO as 4.3 × 102 M-1 s-1 17
;
This
value is the binding constant of substrate to reduced enzyme;49 the value
of kcat/KM apparently has not been determined.
(53) Garton, S. D.; Hilton, J.; Oku, H.; Crouse, B. R.; Rajagopalan, K.
V.; Johnson, M. K. J. Am. Chem. Soc. 1997, 119, 12906-12916.

1930 J. Am. Chem. Soc., Vol. 123, No. 9, 2001
Lim and Holm
oxo transfer. In a reaction sequence such as that in Figure 9, a
relatively electron-deficient Mo(IV) center is expected to
enhance substrate binding, but a relatively electron-rich center
should promote formation of the MoVIO state by oxo transfer.
As shown previously, dithiolene ligands by virtue of Mo-S
bond covalency can act as an “electronic buffer” to maintain
charge distribution at the molybdenum center as other ligands
are varied or the oxidation state is changed.54 In one set of
oxomolybdenum complexes containing a dithiolene ligand,
MoVO/MoIVO redox potentials are linearly related to calcu-
lated charges per sulfur atom.55 The latter are dependent on
the dithiolene itself. The buffer effect can exist in both ana-
logue and enzyme, but ligand structural and environmental
effects in the latter could render the buffered charge distribu-
tion more favorable to the rate-limiting step in oxo transfer, as
compared to analogues, which are simplified to the point of
lacking structure beyond the chelate rings. Last, complexes
such as 1 are not represented here as highly accurate analogues
of the reduced sites of DMSOR and TMAOR. They are,
however, the nearmost approaches currently available and
provide the basis for further evolution of analogue com-
plexes that more closely approximate the reaction rates of
enzyme sites.
Acknowledgment. This research was supported by NSF
Grant CHE 98-76457. We thank K.-M. Sung for valuable
discussions, Prof. H. Schindelin for access to ref. 6 prior to
publication, and Prof. E. N. Jacobsen for use of a gas
chromatograph.
Supporting Information Available: Crystallographic data
for the compounds in Tables 1 and 2 and selected Eyring plots
for reactions 8 and 9 (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
(54) Westcott, B. L.; Gruhn, N. E.; Enemark, J. H. J. Am. Chem. Soc.
1998, 120, 3382-3386.
(55) Helton, M. E.; Gruhn, N. E.; McNaughton, R. L.; Kirk, M. L. Inorg.
Chem. 2000, 39, 2273-2278.
JA003546U