Activation of the sulfhydryl group by Mo centers: Kinetics of reaction of benzyl radical with a binuclear Mo(μ-SH)Mo complex and with arene and alkane thiols

Article abstract of DOI:10.1021/ja049321r

This paper provides evidence from kinetic experiments and electronic structure calculations of a significantly reduced S-H bond strength in the Mo(μ-SH)Mo function in the homogeneous catalyst model, CpMo(μ-S) ₂(μ-SH)₂MoCp (1, Cp = η ⁵-cyclopentadienyl). The reactivity of 1 was explored by determination of a rate expression for hydrogen atom abstraction by benzyl radical from 1 (log(k_abs/M^-1 s^-1) = (9.07 ± 0.38) - (3.62 ± 0.58)/θ) for comparison with expressions for CH₃(CH₂)₇SH, log(k_abs/M ^-1 s^-1) = (7.88 ± 0.35) - (4.64 ± 0.54)/θ, and for 2-mercaptonaphthalene, log(k_abs/M ^-1 s^-1) = (8.21 ± 0.17) - (4.24 ± 0.26)/θ (θ = 2.303RT kcal/mol, 2σ error). The rate constant for hydrogen atom abstraction at 298 K by benzyl radical from 1 is 2 orders of magnitude greater than that from 1-octanethiol, resulting from the predicted (DFT) S-H bond strength of 1 of 73 kcal/mol. The radical CpMo(μ-S) ₃(μ-SH)MoCp, 2, is revealed, from the properties of slow self-reaction, and exclusive cross-combination with reactive benzyl radical, to be a persistent free radical.

Full text of DOI:10.1021/ja049321r

Published on Web 05/08/2004
Activation of the Sulfhydryl Group by Mo Centers: Kinetics of
Reaction of Benzyl Radical with a Binuclear Mo(µ-SH)Mo
Complex and with Arene and Alkane Thiols
James A. Franz,* Jerome C. Birnbaum,* Douglas S. Kolwaite, John C. Linehan,
Donald M. Camaioni, and Michel Dupuis
Contribution from The Pacific Northwest National Laboratory, P.O. Box 999,
Richland, Washington 99352
Received February 6, 2004; E-mail: james.franz@pnl.gov; jerome.birnbaum@pnl.gov
Abstract: This paper provides evidence from kinetic experiments and electronic structure calculations of
a significantly reduced S-H bond strength in the Mo(µ-SH)Mo function in the homogeneous catalyst model,
CpMo(µ-S)2(µ-SH)2MoCp (1, Cp ) η5-cyclopentadienyl). The reactivity of 1 was explored by determination
of a rate expression for hydrogen atom abstraction by benzyl radical from 1 (log(kabs/M-1 s-1) ) (9.07 (
0.38) - (3.62 ( 0.58)/θ) for comparison with expressions for CH3(CH2)7SH, log(kabs/M-1 s-1) ) (7.88 (
0.35) - (4.64 ( 0.54)/θ, and for 2-mercaptonaphthalene, log(kabs/M-1 s-1) ) (8.21 ( 0.17) - (4.24 (
0.26)/θ (θ ) 2.303RT kcal/mol, 2σ error). The rate constant for hydrogen atom abstraction at 298 K by
benzyl radical from 1 is 2 orders of magnitude greater than that from 1-octanethiol, resulting from the
predicted (DFT) S-H bond strength of 1 of 73 kcal/mol. The radical CpMo(µ-S)3(µ-SH)MoCp, 2, is revealed,
from the properties of slow self-reaction, and exclusive cross-combination with reactive benzyl radical, to
be a persistent free radical.
Scheme 1. Dihydrogen Activation and Catalytic Homolytic
Pathways in Hydrogenation of Olefins and Dehydrogenation of
Saturated Hydrocarbons Involving MonS2n Clusters
Introduction
Promoted heterogeneous molybdenum sulfide catalysts are
among the most important industrial catalysts, because of their
role in hydrodesulfurization (HDS) of petroleum feedstock.1
MoS clusters are found in oxidase, reductase, dehydrogenase,
and nitrogenase enzymes.2 The chemistry of soluble metal
sulfides is recognized to be central to the understanding of the
function of bio-organometallic reaction centers in broad cat-
egories of enzymatic transformations.3 Molybdenum-complexed
sulfhydryl species play an important role in heterogeneous
catalysis in hydrocarbon conversion processes. Much interest
continues to surround the development of detailed understanding
of molecular-level mechanistic chemistry of promoted Mo2S4
systems in hydrogenolysis and hydrodesulfurization (HDS) and
in C-S and C-H bond activation. The properties of the S-H
and S-C bonds in the Mo-(µ-SR)-Mo group (R ) H,C) and
in mixed metal (Mo, Co, Ni) µ2-SR and in µ3-SR sulfide systems
control the activation of S-C bonds in HDS and control
catalytic pathways of hydrogenation. The µ-S and µ-SH
structures may participate in dihydrogen activation, as depicted
in Scheme 1. Recent theoretical studies4 have reviewed current
structural understanding of HDS catalysis and have defined the
thermodynamics of hydrogenation of various functional groups
of Mo2S4. Studies of homogeneous models of CoMoS catalysts
have identified mixed metal systems that, most remarkably, may
lead to the reduction of S-H and S-aryl bond strengths to as
low as 20 kcal/mol.5 The reduction of S-H and S-C bond
strengths in complexes of organosulfur compounds at CoMoS
clusters provides efficient pathways both for C-S bond scission
(HDS) and for efficient reverse-radical-disproportionation-based
hydrogenation (kRRD, Scheme 1) and C-C bond hydrogenolysis.
Radical-forming reactions involving hydrogen atom transfer
from metal-complexed sulfhydryl groups to unsaturated hydro-
carbons may involve both µ- and σ-bonded sulfhydryl groups.
(1) (a) Whitehurst, AdV. Catal. 1998, 42, 345 and references therein. (b)
Transition Metal-Sulfur Chemistry; Stiefel, E. I., Matsumoto, K., Eds.; ACS
Symposium Series 653; American Chemical Society: Washington, DC,
1996. (c) Topsøe, H.; Clausen, B. S.; Massoth, F. E. In Catalysis, Science
and Technology; Anderson, J. R., Boudart, M., Ed.; Hydrotreating Catalysis,
Science and Technology, Vol. 11; Sprinter Verlag: Berlin, 1996.
(2) Molybdenum Enzymes, Cofactors, and Model System; Stiefel, E. I.,
Coucouvanis, D., Newton, W. E., Eds.; ACS Symposium Series 535;
American Chemical Society: Washington, DC, 1993.
(4) Travert, A.; Nakamura, H.; van Santen, R. A.; Cristol, S.; Paul. J.-F.; Payen,
E. J. Am. Chem. Soc. 2002, 124, 7084-7095.
(5) Curtis, M. D. Hydrodesulfurization Catalysis and Catalyst Models Based
on Mo-Co-S Clusters and Exfoliated MoS2. In Transition Metal-Sulfur
Chemistry; Stiefel, E. I., Matsumoto, K., Eds.; ACS Symposium Series
653; American Chemical Society: Washington, DC, 1996; pp 154-175.
(3) Rauchfuss, T. B. Inorg Chem. 2004, 43, 14-26.
9
6680
J. AM. CHEM. SOC. 2004, 126, 6680-6691
10.1021/ja049321r CCC: $27.50 © 2004 American Chemical Society

Activation of the Sulfhydryl Group by Mo Centers
A R T I C L E S
Figure 1. Evolution of products during photolysis of a 0.02 M solution of phenylbenzyl ketone (PBK) in benzene at 315 K in the presence of [CpMo(µ-
S)(µ-SH)]2, 1, 1.4 × 10-4 M. Product evolution is depicted for benzaldehyde (b), toluene (9), and bibenzyl (2). The rate constant for abstraction of a
hydrogen atom from 1 by benzyl radical, kabs ) 3.4 × 106 M-1 s-1, was determined using data at 2.0 s (eq 10a). The solid lines are predicted concentrations
of benzaldehyde, toluene, and bibenzyl found by numerical integration of the equations of Scheme 2, employing a rate of photolysis of PBK, 5.7 × 10-6
M/s, and k4 ) 0.5(k5 + k6) ) 0.5(k7 + k8) ) 2.2 × 109 and k9 ) 4.4, 4.4, and 2.2 × 109 M-1 s-1, respectively. Fits of depicted data are sensitive to the choice
of k4 but insensitive to the choice of k5 and k6, and equivalent fits are obtained with low termination rates for 2, e.g., <104 M-1 s-1. The rate constant for
abstraction of hydrogen by benzoyl radical (PhCO•) from 1, 3 × 107 M-1 s-1, was found by variation of rate constants to optimize the fit of the above data
to numerical integration of eqs 1-9 (Scheme 2) subject to the constraints listed above. Convenient programs are available for numerical integration and
optimization.16
The role of reverse-radical-disproportionation (kRRD and kRRD2
depicted in Scheme 1) is now well-established in radical
initiation6 and hydrogenation pathways, as well as in controlled
polymerization.7 The rate of radical initiation via RRD is directly
dependent on ∆H°RRD or ∆H°init (cf. Scheme 1) and in turn on
the M-H or S-H (M ) Co, Mo, Ni) bond strengths of the
hydrogen atom donors. The underlying thermochemistry of S-H
bonds of µ-SH bridges between Mo atoms or Mo and Co or Ni
atoms remains largely unexplored. No direct measure of
homolytic reactivity of the µ2-SH or µ3-SH structures has been
reported, nor have S-H bond strengths been measured for MoS
or MoCoS clusters that are models for activation of S-H and
C-S bonds important to catalysis. A more detailed knowledge
of homolytic reactivity and thermochemistry of the µ-S and
µ-SH groups is desirable for understanding stepwise catalytic
transformations involving C-S and S-H bond activation.
The novel reactivity of bridging µ-S groups in hydrogen
activation, catalytic hydrogenation, and bond formation involv-
ing olefins, acetylenes, and alkyl functionality was early
recognized by M. Rakowski-DuBois and co-workers. The
soluble binuclear species [CpMo(µ-S)(µ-SH)]2, 1, and related
species have been the subject of considerable interest, because
of the relationship of the patterns of reactivity of the embedded
Mo2S4 cluster to key heterogeneous catalytic reaction steps of
MoS and promoted catalysts.8 The complex 1 contains two
bridging hydrosulfido ligands and two bridging sulfur atoms.
Complex 1 catalyzes HD formation from mixtures of H2/D2 or
H2/D2O, demonstrating that the complexed hydrosulfido group
has the ability to activate dihydrogen as well as engaging in
acid-base equilibria with Brønsted acids. Extended Hu¨ckel
calculations revealed the bridging sulfur atoms to exhibit HOMO
and LUMO orbitals with proper symmetry to confer donor and
acceptor properties to the sulfur bridge atoms similar to those
associated with a transition metal.9 The complex reacts directly
with olefins and acetylenes, displacing dihydrogen from the
sulfhydryl groups by olefins and acetylenes to form bridging
bis-alkyl- or bis-alkenyl-dithiolate complexes, respectively.8 The
characteristic of the Mo2S4 cluster of 1 to activate H2 and to
react directly with olefins reflects the novel reactivity available
to the µ2-S functional groups of the cluster. In previous work,
we examined hydrogen transfer from thiols to carbon-centered
radicals and to thiyl radicals, supporting quantitative mechanistic
studies using hydrogen atom transfer from thiols as basis rate
standards.10 The present work extends those studies to systems
of catalytic interest. Although wide interest persists in molyb-
denum hydride and molybdenum-sulfhydryl systems,11 only a
(6) Ru¨chardt, C.; Gerst, M.; Ebenhoch, J. Angew. Chem., Int. Ed. Engl. 1997,
36, 1406.
(7) Tang, L.; Papish, E. T.; Abramo, G. P.; Norton, J. R.; Baik, M.-H.; Friesner,
R. A.; Rappe, A. J. Am. Chem. Soc. 2003, 125, 10093-10102 and
references therein.
(8) Rakowski DuBois, M.; VanDerveer, M. C.; DuBois, D. L.; Haltiwanger,
R. C.; Miller, W. K. J. Am. Chem. Soc., 1980, 102, 7456.
(9) DuBois, D. L.; Miller, W. K.; Rakowski DuBois, M. J. Am. Chem. Soc.
1981, 103, 3429.
(10) (a) Franz, J. A.; Bushaw, B. A.; Alnajjar, M. S. J. Am. Chem. Soc. 1989,
111, 268. (b) Alnajjar, M. S.; Garrossian, M. S.; Autrey, S. T.; Ferris, K.
F.; Franz, J. A. J. Phys. Chem. 1992, 96, 7037.
(11) Examples include: (a) Rakowski DuBois, M. Catalysis by Sulfido Bridged
Dimolybdenum Complexes. In Catalysis by Di-and Polynuclear Metal
Cluster Complexes; Adams, R. D., Cotton, F. A., Eds.; Wiley-VCH Inc.:
New York, 1998; pp 127-143 and references therein. (b) Reference 4. (c)
Sarker, N.; Bruno, J. W. J. Am. Chem. Soc. 1999, 121, 2174. (d) Fettinger,
J. C.; Kraatz, H. B.; Poli, R.; Quadrelli, E. A.; Torralba, R. C. Organo-
metallics 1998, 17, 5767. (d) Bayse, C. A.; Hall, M. B.; Pleune, B.; Poli,
R. Organometallics 1998, 17, 4309. (e) Hascall, T.; Murphy, V. J.; Parkin,
G. Organometallics 1996, 15, 3910. (f) Tsai, Y. C.; Johnson, M. J. A.;
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6681

A R T I C L E S
Franz et al.
Scheme 2. Kinetic Steps in the Steady-State Photolysis of
Phenylbenzyl Ketone (PBK) in the Presence of
[CpMo(µ-S)(µ-SH)]2 (1)
undergo self-reaction to form a dimer but predominately
undergoes cross-termination with a benzyl radical to give the
isomeric cross-termination product CpMo(µ-SCH2Ph)(µ-S)2(µ-
SH)MoCp (3a,b). Evidence suggests that radical 2 does not form
a dimer and undergoes self-reaction exclusively, and very
slowly, by disproportionation, possibly forming transient CpMo-
(µ-S)4MoCp (4), which forms the thermodynamically favored
rearrangement product 5 (CpMo(dS)(µ-S)2Mo(dS)Cp) (Scheme
2). The slow self-disproportionation of 2, taken with the selective
cross-combination of 2 with transient radicals (benzyl), dem-
onstrates that 2 is a persistent free radical.
Results
Kinetics of Reaction Benzyl Radical with [CpMo(µ-S)(µ-
SH)]2, 1. Rate constants for hydrogen atom abstraction by the
benzyl radical from 1 were determined by a competition of
hydrogen atom abstraction by benzyl radical to form toluene
with benzyl radical self-termination to produce bibenzyl during
photolysis of phenylbenzyl ketone (PBK) or dibenzyl ketone
(DBK) in benzene solutions. The benzyl radical was produced
by broadband visible UV irradiation (>330 nm) of optically
dilute solutions of phenylbenzyl ketone (PBK) and 1.
Under conditions of constant rate of photolysis of phenyl-
benzyl ketone and a short extent of conversion of the hydrogen
donor, typically less than 5-7%, the relationship among toluene,
bibenzyl (BB), and the average hydrogen atom donor concentra-
tion is given by eq 10a, where [DH]av is the average concentra-
tion of the thiol donor over a period of photolysis time, ∆t:
[Toluene]kt1/2
[BB]1/2[DH]av∆t1/2
kabs
)
(10a)
(10b)
abs[BB]1/2 1/2/kt1/2
∆
t
[Toluene] ) [DH]0(1 - e-k
)
In practice, conversion of donor (1) up to 50% leads to less
than 20% error in rate constant kabs using eq 10a. An integrated
rate equation, eq 10b, may be applied for the case of extensive
conversion of starting donor at constant rate of photolysis of
PBK or DBK. For the PBK experiments, the yields of
benzaldehyde and toluene were used to calculate the average
concentration of hydrogen donor consumed. Combining experi-
mental values of bibenzyl and toluene at each photolysis time
(0.5-8.0 s) with the rate constant for self-termination of
bibenzyl, kt (see below), provides values of kabs at each
temperature. The rate of photolysis of PBK employed in the
few kinetic studies have been carried out involving hydrogen
atom transfer from molybdenum hydrides,12,13 along with several
measurements of MosH bond strengths.14 We present a kinetic
characterization of the activation of the Mo-(µ-SH)-Mo group
in the Mo2S4H2 cluster, 1, along with estimates of the Mo(µ-
SH)Mo SsH bond strength based on a correlation of abstraction
rates with rate constants for reactions of carbon-centered radicals
with thiols and DFT bond dissociation enthalpy calculations
comparing the S-H bond strength of 1 with arene and alkane
thiols and H2S. New rate expressions for reaction with benzyl
radical with alkane and aryl thiols are presented for comparison
to the reactivity of the MoSH cluster.13,15 Evidence is presented
showing that the radical CpMo(µ-S)2(µ-SH)(µ-S•)MoCp, 2,
formed from hydrogen atom abstraction from 1, does not
kinetic measurements was approximately (3-7) × 10-7 M s-1
,
corresponding to a steady-state concentration of benzyl radical
of about 2 × 10-8 M. The rate of formation of toluene and
bibenzyl was found to be linear during the first several seconds
of photolysis, justifying the use of eq 10a for calculation of
rate constants. The resulting temperature-dependent rate con-
stants for the reaction of benzyl radical with [CpMo(µ-S)(µ-
SH)]2 are shown in Figure 2, and the rate expression is presented
in Table 1.
Mindiola, D. J.; Cummins, C. C. J. Am. Chem. Soc. 1999, 121, 10426. (g)
Nolan, S. P.; Lopez de la Vega, R.; Hoff, C. D. J. Organomet. Chem. 1986,
315, 187. (h) Hoff, C. D. J. Organomet. Chem. 1985, 282, 201. (i) Landrum,
J. T.; Hoff, C. D. J. Organomet. Chem. 1985, 282, 215.
(12) (a) Eisenberg, D. C.; Norton, J. R. Isr. J. Chem. 1991, 31, 55. (b) Eisenberg,
D. C.; Lawrie, C. J. C.; Moody, A. E.; Norton, J. R. J. Am. Chem. Soc.
1991, 113, 4888.
Hydrogen Abstraction by Benzyl Radical from 1-Oc-
tanethiol in Benzene and Cyclohexane. Rate constants for
(13) Franz, J. A.; Linehan, J. C.; Birnbaum, J. C.; Hicks, K. W.; Alnajjar, M.
S. J. Am. Chem. Soc. 1999, 121, 9824.
(14) (a) Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1989, 111, 6711. (See also
Additions and Corrections. J. Am. Chem. Soc. 1990, 112, 2843.) (b)
Reference 11g.
(15) Franz, J. A.; Naushadali, S. K.; Alnajjar, M. S. J. Org. Chem., 1986, 51,
19.
(16) (a) Macey, R. I.; Oster, G. F. Berkeley Madonna, version 8.0.1; 2000. See:
http://www.berkeleymadonna.com/. (b) Hinsberg, W.; Houle, F.; Allen, F.;
Yoon, E. Chemical Kinetics Simulator, v. 101; IBM Almaden Research
Center. See: https://www.almaden.ibm.com/st/computational_science/ck/
msim/.
9
6682 J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004

Activation of the Sulfhydryl Group by Mo Centers
A R T I C L E S
tolysis of dibenzyl ketone (DBK) in benzene solutions and in
cyclohexane solutions. Benzyl radical was produced by broad-
band visible UV irradiation (>330 nm) of optically dilute
solutions of DBK and 1-octanethiol, as shown in Scheme 3.
Under the conditions employed, the relationship among toluene,
bibenzyl, and the average donor concentration, [DHav], is given
by eq 10a. In typical experiments, less than 6% of the donor
was consumed. The rate of photolysis of DBK employed in the
kinetic measurements was about 2 × 10-6 M s-1, corresponding
to a steady-state concentration of benzyl radical of about 1 ×
10-7 M. The yield of toluene was used to calculate an average
thiol concentration during the photolysis. The temperature-
dependent plot of data for the reaction of benzyl radical with
1-octanethiol is shown in Figure 2, covering a range of 295-
374 K, and the rate expression is presented in Table 1. Kinetics
of reaction of benzyl radical were also carried out in cyclohexane
for comparison with benzene, to rule out formation of a portion
of the product toluene by reaction of benzyl radical with
octanethiylcyclohexadienyl radical, the adduct of 1-octanethiyl
radical to the solvent. This would potentially react with benzyl
radical to give toluene and octylphenylsulfide. In experiments
with benzene employing extended photolysis times (>60 s),
octylphenylsulfide was detectable by GC/MS. However, at
photolysis times (<5 s) used in kinetic measurements, the sulfide
was not detectable and kinetics carried out in cyclohexane
yielded Arrhenius rate parameters close to those obtained in
benzene, [log(kabs/M-1 s-1) ) (7.75 ( 0.21) - (4.79 ( 0.31)/
θ, (θ ) 2.303RT kcal/mol, errors are 2σ)]. The participation of
Figure 2. Temperature-dependent rate data for abstraction of hydrogen
atom by benzyl radical from [CpMo(µ-S)(µ-SH)]2 (1), 2-mercaptonaph-
thalene, and octanethiol in benzene.
Scheme 3
hydrogen abstraction by benzyl radical from 1-octanethiol were
determined in competition experiments by steady-state pho-
Table 1. Rate Parameters for the Reaction of Benzyl Radical with a µ-SH Complexed Molybdenum Thiol 1, Arene and Alkane Thiols, and a
Molybdenum Hydride (Errors Are 2σ)
q
q
hydrogen
donor
X−H BDE
(kcal/mol)
log A
∆S (eu)
E h
(kcal/mol)
∆H
k (M-1 s-1
298 K
)
mean temp
(Tm, K)
temp range
(K)
a
(M-1 ‘s-1
)
(Tm, K)
(kcal/mol)
b
[CpMo(µ-S)(µ-SH)]2
C10H7-2-SHb,d
73(calcd)e
77.9g
9.07 ( 0.38
8.21 ( 0.17
7.89 ( 0.35
7.75 ( 0.21
8.89 ( 0.22
-19.2 ( 1.8
-23.2 ( 0.8
-24.7 ( 1.6
-25.3 ( 1.0
-20.0 ( 1.0
3.62 ( 0.58
4.24 ( 0.26
4.64 ( 0.54
4.79 ( 0.31
2.31 ( 0.33
2.96 ( 0.58
3.57 ( 0.26
3.96 ( 0.54
4.12 ( 0.31
1.66 ( 0.33
2.6 × 106
1.3 × 105
3.1 × 104
1.7 × 104
1.6 × 107
331
337
341
333
325
295-372
298-376
295-374
295-372
297-373
CH3(CH2)7SHb
87.3g
CH3(CH2)7SHc
87.3g
Cp*Mo(CO)3Ha,b,f
70a
a Nolan, S. P.; Vega, R. L.; Hoff, C. D. Organometallics 1986, 5, 2529-2537. Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1989, 111, 6711-6717. See
Additions and Corrections, J. Am. Chem. Soc. 1990, 112, 2843. b Solvent, benzene. c Solvent, cyclohexane. d 2-Mercaptonaphthalene. e Estimated from DFT
bond dissociation isodesmic calculations, this work; see text. f Cp* ) pentamethyl-η5-cyclopentadienyl. g Reference 34. Bordwell, F. G.; Zhang, X.-M.;
a
Satish, A. V.; Cheng, J. P. J. Am. Chem. Soc. 1994, 116, 6605. h A ) Arrhenius preexponential factor, Ea ) Arrhenius exponential constant in k ) Ae-E /RT
.
1
1
Figure 3. (a) H NMR spectra of photolysis of DBK and 1. (b) H NMR spectrum of photolysis of DBK and 1, 3000 s.
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6683

A R T I C L E S
Franz et al.
Scheme 4. Photolysis of DBK in the Presence of 1 Yields Predominately 3a,b and Trace Amounts of 5; See Texta
a Compounds 5 and 6 are not detected. Benzyl radical undergoes nearly exclusive combination with 2 to form 3.
the addition/cross-termination pathway in production of toluene
appears negligible.
CH2 resonances (Figure 3b) at 2.70 and 2.78 ppm exhibiting
twice the integrated intensity of the hydride resonances. The
resonances at 2.7 and 2.8 ppm are assigned to the cross-
termination product isomers 3a and 3b. (Note phenylacetalde-
hyde formed by trapping by 1 of the intermediate phenylacetyl
radical.) Under conditions of extensive conversion, 3a and 3b
are further converted to the dibenzyl substituted cluster 8. To
aid in the identification of isomers of 3 and 8, hydride 1 was
subject to exchange of µ2-SH groups with benzylthiol (eq 23):
Hydrogen Abstraction by Benzyl Radical from 2-Mer-
captonaphthalene. Kinetics of reaction of benzyl radical with
2-mercaptonaphthalene were determined in benzene under
conditions employed for octanethiol. Temperature-dependent
rate data are depicted in Figure 2 and presented in Table 1.
Slow Self-Reaction of 2 and Dominant Cross-Termination
of 2 with Benzyl Radical. The photolysis of dibenzyl ketone
was carried out in the presence of a higher concentration of 1,
0.0005 M, than employed in the kinetic experiments, to facilitate
the observation of photoproducts derived from 2 and benzyl
1
radical by H NMR. The photoconversion steps are described
in eqs 16-22. Figure 3a shows the consumption of the two
isomers of 1 and the appearance of a new pair of hydride
resonances at -1.64 and -1.7 ppm in CDCl3. The new hydride
peaks are accompanied by the appearance of two new benzylic
Exchange of benzylthiol with 1 provided two products, 3a
and 3b, with 13C and 1H NMR spectra identical to those
produced in the photolysis experiments, followed by further
conversion of isomers 3 to 8. The exchange reaction, eq 23,
9
6684 J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004

Activation of the Sulfhydryl Group by Mo Centers
A R T I C L E S
Table 2. 13C and 1H Data for Isomers of 1, 3, 5, and 8
compound
solvent
1H NMR data
13C NMR data
1a
1b
1a
1b
3a
CDCl3
CDCl3
benzene-d6
benzene-d6
CDCl3
6.37(10H,s), -1.41 (1H,s)
6.37(10H,s), -1.48 (1H,s)
5.89 (10H,s), -1.32 (1H, s)
5.89 (10H,s), -1.40 (1H,s)
6.22(10H,s), 2.80(2H,s), -1.78(1H,s)
98.07
98.15
141.35,128.89(2C),128.23(2C), 126.85,
98.07(10C), 44.52
3b
CDCl3
6.23(10H,s), 2.74(2H,s), -1.86(1H,s)
140.56, 129.05(2C), 128.45(2C), 126.83,
98.05(10C), 43.49
3a
3b
5
benzene-d6
benzene-d6
benzene-d6
CDCl3
5.87(10H,s), 2.78(2H,s), -1.65(1H,s)
5.87(10H,s), 2.71(2H,s), -1.70(1H,s)
5.78 (10H, s)
5
6.34 (10H, s)
8a
CDCl3
7.23(6H,m), 6.86(4H,m), 6.08(10H,s),
2.764(4H,s)
141.2(2C), 129.05(4C), 128.41(4C), 126.74(2C),
98.16(10C), 42.09
8b
8a
CDCl3
7.21(6H,m), 6.84(4H,m), 6.13(10H,s),
2.712(4H,s)
7.2(6H,m), 6.8(4H,m), 5.85(10H,s),
2.73(4H,s)
140.71(2C), 129.23(4C), 128.37(4C), 126.62(2C),
98.11(10C), 38.10
benzene-d6
was found to proceed much more rapidly in CDCl3 than in
benzene-d6, and the second step (eq 23b) was found to be
appreciably faster than the initial step (eq 23a). Preparative
column chromatography of the exchange mixture containing 1,
3, and 8 on alumina (eluent, CH2Cl2) led to nearly pure isomers
of 8 that were found to be conformationally stable. Similar to
1, the isomers of 3 are always found in the same ratio, reflecting
a facile acid/base-catalyzed pathway available to the SH protons
that retains the equilibrium mixture of 3a and 3b. (Note that
the assignment of isomeric structures remains (3a vs 3b and
8a vs 8b) ambiguous.)
assigned as 5, and isomers of 3 and 8.
hν8 ButO•
ButOOBut
9
(24)
(25)
(26)
kabs′′
ButO• + 1
8 ButOH + 2
kdisp
2 2
8 1 + Cp2Mo2(S)4 (4 and/or 5)
kcomb
kdec
2 2
8 6 8 1 + (4 and/or 5)
(27)
(28)
Cp2Mo2(S)4 (4) f 5
Cp2Mo2(S)4 (4 and/or 5) f precipitate
Traces of the disproportionation product assigned as 5 were
identified by prompt (within 10 min) 1H NMR spectroscopy of
the product mixture from photolysis of DBK in the presence of
1 in benzene-d6. The product 5 was present at less than 5% of
(29)
Formation of Disproportionation Product 5. The lack of
formation of 6 via dimerization of 2 in the photochemical
experiments above involving reactions of benzyl radical and 2,
together with the predominance of products 3a,b, raises the
question of whether 2 participates in slow self-disproportion-
ation to form 4 and/or 5 in the absence of reactiVe radicals
such as benzyl. To observe self-reaction of radical 2 in the
absence of reactive alkyl radicals, 2 was produced by reaction
of a tert-butoxyl radical with 1. Thus, freeze-thaw degassed
solutions of 1 and di-tert-butylperoxide in benzene-d6 were
photolyzed with the diffuse beam of a 1-kW high pressure xenon
arc lamp and promptly examined by 1H NMR spectroscopy over
the time period 5-90 min. Approximately 25% of 1 was cleanly
converted to a new Cp resonance at 5.78 ppm, which dis-
appeared (by precipitation) with a half-life of 120 min at 298
K. No new S-H resonances in the region -1 to -3 ppm,
corresponding to cross-termination or self-termination products
from 2, were observed. Because of the very high expected rate
of reaction of tert-butoxyl radical with 1 (5-10 × 108 M-1
s-1), tert-butoxyl radical is entirely consumed in abstraction (eq
25), to the exclusion of cross-termination reactions with 2 (and
unlike experiments in which benzyl radical undergoes ap-
preciable cross-termination with 2 during the steady-state
photolysis of DBK). Thus, 2 reacts slowly by self-reaction (eqs
26, 27). The absence of new SH resonances attributable to 6
suggests that 5 can form by a slow, apparent disproportionation
of 2 (eq 26). An alternative pathway that cannot be formally
1
the concentration of 3a,b. No H NMR transitions attributable
to 6 were detectable among the photoproducts. (Isomers of 6
are expected to give at least two unique SH protons in the range
-1 to -3 ppm.) Although radical 2 is predicted in kinetic
modeling to be present at significant steady-state population and
would be expected to undergo self-reaction to form 4 (or 5) or
6 (see e.g., Scheme 4), no new hydride resonances attributable
to 6 in the 1H NMR spectrum were detectable, and 5 was shown
to be transient in solution due to extremely low solubility in
CDCl3 and benzene-d6 (see below). Although self-termination
product 6 was not detected in the present experiments, formation
of a disulfide linkage between Mo2S4 binuclear units is well-
known.17 The present experiments do not rule out the formation
of 6 followed by rapid decomposition of 6 to give 1 and 5,
e.g., equivalent to the self-disproportionation products of radical
2.
However, in promptly examined experiments in which DBK
was photolyzed in the presence of 1, no more than trace amounts
of 4 or 5 were formed. Thus, 4 and/or 5 are not formed by
cross-disproportionation of benzyl radical and 2 or by self-
disproportion of 2, competitive with cross-combination of benzyl
radical with 2. Table 2 presents NMR data for 1, the structure
(17) (a) Birnbaum, J. C.; Godziela, G.; Maciejewski, M.; Tonker, T. L.;
Haltiwanger, R. C.; Rakowski DuBois, M. Organometallics 1990, 9, 394.
(b) Birnbaum, J. C.; Laurie, J. C. V.; Rakowski DuBois, M. Organometallics
1990, 9, 156.
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6685

A R T I C L E S
Franz et al.
ruled out involves the formation of 6 by dimerization of 2,
followed by rapid decomposition of 6 to 1 and insoluble 5 (eq
27). The transient 1H NMR peak observed at 5.89 ppm is
assigned to 5, the disproportionation product of 2.
[Cp′Mo(µ-SH)(µ-S)]2 (13) and [Cp*Mo(µ-SH)(µ-S)]2 (14), by
oxidation with azo-benzene, which is reduced to 1,2-diphenyl-
hydrazine. Both 11 and 12 are rapidly reconverted to 13 and
14 by exposure to H2.22 In this work, compound 5 was
independently prepared by oxidation of 1 with azo-benzene in
CDCl3 and benzene-d6 and gaVe an identical Cp chemical shift
and solubility characteristics as those of 5 formed in reactions
of 1 in radical reactions of benzyl or tert-butoxyl radicals. When
5 was formed by oxidation of 1 with azo-benzene, it was also
found to be highly insoluble in benzene and CDCl3. Finally,
insoluble 5 formed in the azo-benzene oxidation reaction was
rapidly reconverted to 1, under 1 atm of H2 at 60-80 °C. Initial
formation of 4 followed by photolytic or thermal conversion to
5 is not ruled out by these results.
Bruce and Tyler23 reported the photolytic interconversion of
12, 16, and 15, and Brunner et al.18 reported thermal conversion
of 16 to 15. Based on these reports, conversion of 4 to 5 is
expected.24 Based on our theoretical results, it may be possible
that the structure proposed to be 15 is actually the syn-isomer
of 12.
A. Assignment of Product of Oxidation of 1 to Structure
5. Assignment of the observed transient species to 5 is based
on the observation of a strong MdS stretch observed in the
FTIR spectrum of the solid product isolated from the reaction
of azo-benzene and 1: the dark solid exhibited absorptions (KBr
pellet) at 523 (w), 481 (s), 432 (m), and 406 (m) cm-1
,
respectively, assigned to νModS, ν′ModS, νMosSsMo and ν′MosSs
Mo frequencies, respectively. Candidate structures for oxidation
product(s) of 1 include 4, 5, 9 (the syn-isomer of 5), and 10,
with relative energies depicted for 4, 5, and 9 (0 K B3LYP
DFT energies and geometries calculated with a hybrid basis
set consisting of LANL2DZ ECP at Mo and 6-31G(d) at C,H,
and S). Brunner et al.18 proposed a structure for an isomer of
B. Further Evidence for Slow Self-Reaction of 2. An
experiment to semiquantitatively estimate the rate of self-
termination of 2 was carried out. In an NMR scale photolysis
experiment, an oxygen-free solution of DBK (0.015 M) and 1
(0.003 M) in benzene-d6 was photolyzed with a 1-kW high Xe
lamp for 180 s to give toluene (0.004 M) and 3 but no detectable
5, under conditions of a lower limit of detection of 5 of ∼10-5
M. From the yield of toluene and the known rate constant, kabs
) 2.6 × 106 M-1 s-1 (Table 1), the steady-state benzyl radical
the formula Cp*2Mo2S4 (Cp* ) η5-C5 (CH3)5) to be (Cp*MoS)2-
(µ-S2), analogous to 10. An X-ray crystal structure was not
available to verify the structure. While the structures 4, 5, and
9 were readily located in DFT calculations in this work, all
attempts to locate a stable structure corresponding to 10 failed,
with the initial structures of 10 opening during geometry
optimization to form 9 (relative energies shown).
The syn-isomer 9 is predicted to lie 1.6 kcal/mol higher in
energy than anti-isomer 5. A strong frequency observed at 481
cm-1 (ModS) suggests 5 or 9 to be likely, with 10 ruled out as
a candidate. From computational results suggesting 5 to be most
stable, we have selected 5 as the most probable candidate. This
assignment does not rule out the possibility of initial formation
of 4, either in azo-benzene oxidation of 1 or in the photochemi-
cal free radical oxidation of 1 by benzyl or tert-butoxyl radicals,
followed by thermal or photoinduced conversion of 4 to 5 in
the present experiments.
•
concentration is estimated to be [PhCH2 ] = [toluene]/(∆t)(kabs)-
(2*[1]) ) [toluene]/(180 s)(2.6 × 106)(0.006) )1.4 × 10-9 M.
Assuming a conventional cross-termination rate constant for the
reaction of benzyl radical and 2, 2 × 109 M-1 s-1, together
with the yield of 3 and the benzyl radical concentration from
above, the steady-state concentration of 2 can be estimated from
•
the relation [2] = [3]/(180 s)([PhCH2 ])(2 × 109 M-1 s-1) ) 3
× 10-6 M. Finally, from the relation [5]/∆t = kdisp[2]2 and an
upper limit of [5] < 10-5 M, a value of kdisp < 5 × 103 M-1
s-1is estimated. The rate constant for self-reaction of 2 is thus
at least 105-fold slower than normal bimolecular radical
termination rates. From the estimates of radical concentrations
above, the buildup of a high concentration of the persistent
radical (2) causes the transient radical (benzyl) to be steered to
follow the cross-termination path. The properties of 2 closely
reflect the behavior expected for persistent free radicals.25,26 This
is dramatically shown in Figure 3a, which illustrates the clean,
exclusive formation of cross-termination products 3a,b.
While an early report of the preparation of 519 was found to
be problematic,20 both [Cp′MoS(µ-S)]2 (11, Cp′ ) MeCp) and
[Cp*MoS(µ-S)]2 (12, Cp* ) pentamethyl-η5-cyclopentadienyl)
have been prepared21 in high yields from the parent compounds,
(18) Brunner, H.; Meier, W.; Wachter, J.; Guggolz, E.; Zahn, T.; Ziegler, M.
Organometallics 1982, 1, 1107-1113.
(22) Rakowski DuBois, M.; DuBois, D. L.; VanDerVeer, M. C.; Haltiwanger,
R. C. Inorg. Chem. 1981, 20, 3064.
(23) Bruce, A. E.; Tyler, D. R. Inorg. Chem. 1984, 23, 3433-3434.
(24) The authors thank the reviewers for pointing out the possible photolytic
interconversion of 4 to 5.
(25) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925.
(26) Karatekin, E.; O’Shaughnessy, B.; Turro, N. J. J. Chem. Phys. 1998, 108,
9577.
(19) Beck, W.; Danzer, W.; Thiel, G. Angew. Chem., Int. Ed. Engl. 1973, 12,
582.
(20) Rakowski DuBois, M.; Haltiwanger, R. C.; Miller, D. J.; Glatzmeier, G. J.
Am. Chem. Soc. 1979, 101, 5245.
(21) Casewit, C. J.; Coons, D. E.; Wright, L. L.; Miller, W. K.; DuBois, M. R.
Organometallics 1986, 5, 951-955.
9
6686 J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004

Activation of the Sulfhydryl Group by Mo Centers
A R T I C L E S
Table 3. Electronic Energies and Total Enthalpy Corrections, 298 K (hartree/particle)
enthalpy
correctiong
a
a
a
f
species
LANL2DZ
LANL2DZdpa
ELANL2DZ
ECEP-121G
LANL2DZ (6-31G*)e
LANL2DZ(6-311++G(2d,dp)
H•b
-0.498 91b
-10.657 75
-11.289 78
-49.973 49
-50.599 89
-241.687 58
-242.301 55
-395.309 02
-395.921 60
-563.272 48
-0.501 94b
-10.679 60
-11.326 47
-49.999 76
-50.640 56
-241.763 49
-242.391 48
-395.421 56
-396.047 90
-563.452 86
-0.502 26b
-398.774 34
-399.425 04
-438.102 12
-438.747 09
-629.903 56
-630.534 64
-783.588 13
-784.217 42
-2115.922 73
-1259.863 22
-876.289 803
-2116.544 66
-0.502 26b
-398.774 34
-399.425 04
-438.102 12
-438.747 09
-629.903 56
-630.534 64
-783.58 813
-784.217 42
-2116.886 94
0.00236
0.008 80
0.017 76
0.040 23
0.049 04
0.095 94
0.104 29
0.144 97
0.153 33
0.200 55
0.194 68
0.083 87
0.209 39
HS•
-398.740 027 8
-399.385 435 5
-438.059 669 1
-438.698 348 4
-629.809 997 9
-630.434 123
-398.774 737 2
-399.426 311 3
-438.103 971 3
-438.748 834 8
-629.906 204 3
-630.537 837 1
H2S
CH3S•
CH3SH
PhS•
PhSH
Nap-2-S•c
Nap-2-SHd
2
-2115.688 803
-1259.678 176
-876.207 532 9
-2115.950 427
-1259.876 768
-876.300 824 4
PhSSPh
CH3SSCH3
1b
1a
6b
6c
5
4
3b
3a
-563.879 90
-563.879 71
-1126.571 42
-1126.5453
-562.703 13
-562.673 09
-834.202 60
-834.203 18
-564.071 90
-2117.490 92
-4231.883 01
-4231.416 025
-4231.939 390
0.403 54
-2115.354 31
-2115.325 61
0.192 54
0.192 07
a Single-point electronic energies calculated using the basis set indicated with geometries optimized at the B3LYP/LANL2DZ ECP level of theory (UB3LYP
for open shell systems). ELANL2DZ uses the LANL2DZ ECP (Mo) and the 6-311++G(2d,2p) basis set (CHS). ECEP-121G uses the CEP-121G ECP basis
on Mo and the 6-311++G(2d,2p) functions on C,H,S. b The electronic energy of hydrogen atom was set to -0.5 hartree for BDE calculations. c Naphthalene-
2-thiyl radical. d 2-Mercaptonaphthalene. e Geometry optimized using B3LYP/LANL2DZ ECP for Mo and 6-31G* for CHS. f Single-point energy at B3LYP/
LANL2DZ (Mo) and 6-311++G(2d,2p) (CHS) on geometry of footnote e. g B3LYP/LANL2DZ ECP frequencies used for thermal corrections.
Table 4. UB3LYP S-H Bond Dissociation Enthalpies (298 K), kcal/mol, from Semidirect, Equation 31 (D) and Isodesmic, Equation 32 (I)
Calculations (B3LYP/LANL2DZ Geometries Employed)
H S (expt 91.2)
2
1
PhSH (expt 79.1)b
Nap-2-SH (expt 77.9)b
CH SH (expt 87.3)b
3
basis set
D
D
I
D
I
D
I
D
I
LANL2DZ
78.7
88.0
90.4
90.4
63.3
70.6
72.5
72.2
76.6
74.7
73.2
73.0
67.8
76.6
78.5
78.5
81.1
80.6
79.3
79.3
66.9
75.5
77.4
77.4
80.3
79.6
78.1
78.1
75.3
84.3
86.9
86.9
88.7
88.4
87.7
87.7
LANL2DZdp
ELANL2DZa
ECEP-121Ga
a See footnote a, Table 3. b Reference 34 and references therein.
Discussion
for accurate prediction by semidirect (eq 31) methods or by
computing heats of atomization of molecules by CBS and related
methods28 not practicable for systems of the size of 1 and 2.
These trends are evident in Table 4. For H2S, thiophenol,
2-mercaptonaphthalene, and CH3SH, the errors in BDE average
-0.6 kcal/mol (eq 25), and errors in the isodesmic BDEs
average +0.3 kcal/mol at the highest level of theory (Table 4).
At the UB3LYP/ELANL2DZ//LANL2DZ level of theory, the
bond dissociation energies have converged to within 1 kcal/
mol for all of the model thiol systems. From the pattern of
convergence of the two computational approaches, the bond
strength of 1 is predicted to be 73 kcal/mol. Inclusion of a larger,
triple-ú quality basis set at Mo (CEP-121G) yielded only slight
change in the predicted bond strength.
Benzyl Radical Selectivities in Hydrogen Atom Abstrac-
tions of Aryl, Alkyl, and Molybdenum Thiols and a Molyb-
denum Hydride. The rate constant for hydrogen atom abstrac-
tion by benzyl radical from 1 at 298 K (2.6 × 106 M-1 s-1) is
20-fold faster than hydrogen abstraction from 2-mercaptonaph-
thalene and 100-fold faster than abstraction from octanethiol.
The enhanced rates of hydrogen atom transfer from 1 to benzyl
radical suggest a significant reduction in the S-H bond strength
of 1. Electronic structure calculations were carried out to
estimate the SH bond strength, DH°, of 1 by direct dissociation
(eq 31) and by the use of isodesmic calculations to calculate
∆H°298 for eq 32 followed by application of the bond strength
of H2S, 91.2 kcal/mol, as a reference (eq 32):
Structural Assignment of Isomers of 1, 3, and 6. LANL2DZ
ECP calculation of relative energies of the five isomers of 1
reveals the favored isomers to be those substituted on opposing
sulfur bridges (1a,b):
RSH f RS• + H•
RSH + HS• f RS• + H2S
(31)
(32)
Calculated values of bond strengths by direct dissociation, e.g.,
eq 31, typically result in bond strengths lower than experiment
that converge toward experimental values with improved basis
set (Table 3). Bond strengths derived from isodesmic (and
isogyric)27 reactions (eq 32) tend to converge toward experi-
mental values at a somewhat more modest basis set than required
(27) Cramer, C. J. Essentials of Computational Chemistry, Theories and Models;
John Wiley & Sons Ltd.: Chichester, West Sussex, U.K., 2003.
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6687

A R T I C L E S
Franz et al.
In agreement with M. Rakowski-Dubois and co-workers, who
assigned the more stable of the isomers of 1 to the two trans
isomers (1a,b),8 DFT calculations predict 1b and 1a to be more
stable than the three cis isomers, 1c-e, by 10-20 kcal/mol.
The electronic energies of 1a,b and 3a,b are too close to reliably
assign a and b isomers.
Dimers 6 are also favored that have substituents on the sulfur
bridge opposing the disulfide link. Thus, 6b, with hydrogen
atoms at remote bridge sulfur atoms in a transoid relationship
to the disulfide bridge, is about 16 kcal/mol more stable than
6d, which has a hydrogen atom attached to the sulfur bridge
adjacent to the disulfide bond.
The above observations suggest that both self-combination
and self-disproportionation of 2 are quite slow processes,
whereas reaction of 2 with benzyl radical is a fast, diffusion-
controlled process. The preference of 2 to form 5 rather than 6
may be related to steric congestion in forming the S-S bond
of 6. DFT calculations (B3LYP/ELANL2DZ//B3LYP/LANL2DZ)
predict a 2.43 Å S-S bond for 6b, longer than typical dialkyl
disulfide bonds, 2.05-2.07 Å, and longer than a reported S-S
bond length, 2.147 Å, of the related dicationic complex
[(MeCpMo(µ-S))2S2CH2]2 (SO3CF3)2.17 Thus, the modest basis
set LANL2DZ ECP, lacking d-type functions on S, was
inadequate for predicting the S-S bond length of 6b. The
geometry of 6b was thus optimized with a hybrid basis set
consisting of LANL2DZ ECP at Mo and the all-electron 6-31G*
basis (CHS) giving an improved bond length of 2.08 Å for 6b.
Optimization of the geometry of [(MeCpMo(µ-S))2S2CH2]2
(SO3CF3)2 at the latter basis yielded an S-S bond length of
2.168 Å, in good agreement with the experimental data. Errors
in calculated bond lengths are less than 0.03 Å, and errors in
bond angles are less than 2°. (See the Supporting Information
for a detailed comparison of bond lengths and bond angles for
the dimer [(MeCpMo(µ-S))2S2CH2]2.) Although the structural
predictions are reliable, the error in bond strengths from direct
dissociation calculations remains consistently low by about 10
kcal/mol. However, the magnitude of error in describing sulfur-
centered radicals is consistent for arylthiyl and alkanethiyl
radicals, allowing useful, if semiquantitative, estimates of bond
strengths in isodesmic calculations. For example, the difference
between S-S bond strengths in PhSSPh and CH3SSCH3 is
predicted to be 17.5 kcal/mol (experiment 16 kcal/mol (Table
5)). The S-S bond strength of 6b is thus estimated using
isodesmic reactions, using thermochemical data from Table 5
with single-point electronic energies calculated at B3LYP/
LANL2DZ ECP (Mo) and 6-311++G(2d,2p) (CHS) on the
B3LYP/LANL2DZ ECP (Mo) and 6-31G* (CHS) geometries,
together with thermal enthalpy corrections to 298 K:
Thus, isomers of 1, 3, and 6 favor alkyl or hydrogen atoms
disposed on nonadjacent bridge sulfur atoms. Because of the
close proximity in energy found for the favored isomer pairs
(1a,b; 3a,b; and 6), the assignments (a,b) for the isomers may
be reversed in the discussions above. It is interesting to note
that the electron density of the HOMO of 2 in Figure 4 is
suggestive of favored bond formation at the sulfur opposed to
the S-H group.
Figure 4. HOMO of radical 2 formed by hydrogen atom abstraction from
1. Unpaired spin density is localized primarily on the µ-S bridge oppsite
the S-H group.
Combination and Disproportionation Reactions of 2. The
photolysis experiments of DBK and 1 revealed that benzyl
radical undergoes cross-combination reactions with radical 2
to the near exclusion of cross-disproportionation (kcomb/kdisp
3/5 > 20:1):
>
DH°(6b) ) (PhSSPh(expt) - (PhSSPh(calcd) -
6b (calcd)) ) 49.5 - (38.6 - 22.7) ) 34 kcal/mol
DH°(6b) ) CH3SSCH3(expt) - (CH3SSCH3 (calc) -
6b (calc)) ) 65.4 - (56.1 - 22.7) ) 32 kcal/mol
Self-reaction of 2 in the absence of reactive radicals produces
no detectable dimer(s) 6, but only 5:
The exercise yields an estimate of 33 kcal/mol for the S-S
bond of 6b. This result, even allowing for substantial composite
error (perhaps as much as 10 kcal/mol) in the calculations and
in the experimental thermochemistry, predicts that 6b should
(28) Henry, D. J.; Sullivan, M. B.; Radom, L. J. J. Chem. Phys. 2003, 118 (1)
4849-4860.
9
6688 J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004

Activation of the Sulfhydryl Group by Mo Centers
A R T I C L E S
Table 5. Thermochemical Values, kcal/mol, for Estimation of
Isodesmic S-S Bond Strengths
theoretically predicted value of 73 kcal/mol for the S-H bond
of 1, compared to ∆H° ) -13 ( 4 kcal/mol predicted from
the relationship of Figure 5.
DH°, S−S
DH°, S−S (calcd)g
species
PhSH
∆H°f,298
DH°, S−H
(expt)
direct (isodesmic)
A pattern of reactivity of the donating species shows an
increase in rate constants by about an order of magnitude from
1-octanethiol to 2-mercaptonaphthalene to [CpMo(µ-S)(µ-SH)]2.
(Cp*Mo(CO)3H is included in Table 1 to illustrate the reactivity
of a directly Mo-bonded hydrogen atom compared to the
reactivity of the MoS-H group. The reactivity increases by a
factor of 6 on going from 1 to Cp*Mo(CO)3H, DH° ) 69 kcal/
mol.31 Steric crowding about the molybdenum hydride and
interaction of the bulky substituents about the R-carbon of the
abstracting radical with the inaccessible Mo-H exert significant
control over the rates of hydrogen atom donation for Cp*Mo-
(CO)3H.13) For [CpMo(µ-S)(µ-SH)]2, 1, the radially disposed
S-H groups appear readily accessible to abstracting radicals,
with substantial separation from the Cp groups. The radical 2,
while apparently disfavoring self-combination, readily undergoes
cross-combination with transient, reactive radicals on the
periphery of the S4 ring. The dimethyl analogue of the
molybdenum dithiol complex, [CpMo(µ-S)(µ-SCH3)]2, reveals
a pseudo square planar array of sulfur atoms approximately the
size of a cyclopentadienyl ring.8 This suggests that the hydro-
sulfido ligands on [CpMo(µ-S)(µ-SH)]2 would be fully exposed
and unhindered by the distant Cp ligands. This assumption is
confirmed by the observed reactivity order shown in Figure 5,
on assignment of the calculated estimate of DH° ) 73 kcal/
mol for the S-H bond of 1.
26.9a
-5.5b
58.3c
79.1e
87.3f
CH3SH
PhSSPh
CH3SSCH3
6b
49.5
65.4
38.6(47)
56.1(68)
22.7(33)
-5.8d
PhS•
53.9
29.8
CH3S•
a Scott, D. W.; McCullough, J. P.; Hubbard, W. N.; Messerly, J. F.;
Hossenlopp, J. A.; Frow, F. R.; Waddington, G. J. Am. Chem. Soc. 1956,
78, 5463. b Good, W. D.; Lucina, J. L.; McCullough, J. P. J. Phys. Chem.
1961, 65, 2229. c Mackle, H.; Mayrick, R. G. Trans. Faraday Soc. 1962,
50, 238. d Hubbard, W. N.; Douslin, D. R.; McCullough, J. P.; Scott, D.
W.; Todd, S. S.; Messerly, J. F.; Hossenlopp, J. A.; George, A.; Waddington,
G. J. Am. Chem. Soc. 1958, 80, 3547. e Bordwell, F. G.; Zhang, X.-M.;
Satish, A. V.; Cheng, J.-P. J. Am. Chem. Soc. 1994, 116, 6605. f CRC
Handbook of Chemistry and Physics, 76th ed.; Lide, D. R., Ed.; CRC Press:
Boca Raton, FL, 1995. g Single-point energies at B3LYP/LANL2DZ ECP
(Mo) and 6-311++G(2d,2p) (CHS) calculated using the B3LYP/LANL2DZ
ECP (Mo) and 6-31G* (CHS) geometry data and total thermal corrections
of Table 3 to obtain direct dissocation values and isodesmic values in
parentheses.
Conclusions
The first kinetics of reaction of a carbon-centered radical with
the µ2-S-H group in a Mo2S4 cluster are reported. New
temperature-dependent rate data for abstraction of hydrogen
atom from a µ2-S-H group of the transition metal dithiol, 1,
have been measured, together with new absolute rate expressions
for abstraction of arene- and alkane-thiol S-H hydrogen atoms
by benzyl radical. The observed trend in reactivity is semi-
quantitatively correlated in a Polanyi relationship and found to
be consistent with the 73 kcal/mol S-H bond strength predicted
for 1 by DFT calculations. The reduced S-H bond strength of
1, in comparison to organic thiols, suggests that the Mo(µ2-S-
H)Mo group will be more reactive in reverse-radical-dispro-
portionation (RRD) pathways than conventional organic thiols
and in reactions involved in the catalysis of hydrogenation/
dehydrogenation pathways involving MoS catalyst substructure.
Accordingly, Mo(µ2-S-C)Mo bond strengths are lower than
conventional thiols, enhancing homolytic S-C bond cleavage
in HDS for nonpromoted catalysts. For promoted catalysts,
addition of a third metal center, CoMo2(µ3-S-H) and CoMo2-
(µ3-S-C) may further dramatically lower S-H and S-C
homolytic bond strengths.4 Finally, radical 2 exhibits properties
of favoring cross-termination with reactive alkyl radicals but
does not undergo combination to the dimer and only undergoes
disproportionation very slowly in the absence of reactive
transient radicals, fitting the definition of a persistent free radical.
The build-in of strain on formation of 6 results in a weakened
S-S bond and may contribute to the strongly disfavored
dimerization of 2.
Figure 5. Plot of natural logarithm of abstraction rate constant, 298 K
(carbon-centered radical, thiol) vs enthalpy of reaction: (1) n-Bu, PhSH;
(2) i-Pr, PhSH; (3) t-Bu, PhSH; (4) benzyl, 1; (5) benzyl, naphthyl-2-SH;
(6) benzyl, PhSH; (7) PhCH(•)Ph, PhSH; (8) benzyl, n-octylSH; (9) PhC-
(•)OHCH3, PhSH. Experimental values are taken from this work and from
the following: Franz, J. A.; Bushaw, B. A.; Alnajjar, M. S. J. Am. Chem.
Soc. 1989, 111, 268-275. Alnajjar, M. S.; Franz, J. A. J. Am. Chem. Soc.
1992, 114, 1052-1058. Kandanarachchi, P. H.; Autrey, T.; Franz, J. A. J.
Org. Chem. 2002, 67, 7937-7945.
not decompose homolytically to regenerate 2 at an appreciable
rate at ambient temperatures and should have been readily
detectable. The S-S bond strength of 6b is appreciably less
than that of PhSSPh and may reflect significant strain built into
the crowded Cp4Mo4S8H2 cluster. This finding may be related
to the failure of radical 2 to combine to form 6.
Evans-Bell-Polanyi Correlation of Hydrogen Atom
Transfer Reactivity. The reactivity for hydrogen atom transfer
to carbon-centered radicals from thiols in nonpolar solvents is
broadly correlated with ∆H° for the S-H atom transfer
reaction.29 As depicted in Figure 5, rate constants for hydrogen
atom transfer range from ∼108 M-1 s-1 for reaction of primary
radicals with thiophenol (∆H° ) -22 kcal/mol) to 3 × 103
M-1 s-1 for reaction of a stabilized benzylic radical, PhCH(•)-
(OH)CH3, with thiophenol (∆H° ) 2 kcal/mol) to 6 M-1 s-1
for reaction of triphenylmethyl radical with thiophenol.30 The
value of ln(kabs/M-1 s-1) for reaction of benzyl radical with 1
is plotted against the value of ∆H° ) -15.5 kcal/mol from the
(29) Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1938, 34, 11.
(30) Colle, T. H.; Glaspie, P. S.; Lewis, E. S. J. Org. Chem. 1978, 43, 2722-
2725.
(31) (a) Nolan, S. P.; Lopez de la Vega, R.; Hoff, C. D. Organometallics 1986,
5, 2529. (b) Reference 6a.
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6689

A R T I C L E S
Franz et al.
1
Electronic Structure Calculations. Geometry optimizations
were carried out using the B3LYP32 density functional with the
LANL2DZ ECP basis set which includes the D9533 basis set
on first row elements and the Los Alamos ECP plus D95 basis
set for second row and higher elements,34 using the Gaussian
9835 or NWChem 4.536 programs. Single-point calculations were
carried out with the LANL2DZ geometries of the Mo species
and organosulfur species using three basis sets, (1) the LANL2DZ
ECP basis set augmented with diffuse and polarization functions,
LANL2DZdp,37 (2) a basis set (denoted ELANL2DZ) consisting
of the LANL2DZ ECP basis set for Mo enhanced with the
6-311++G(2d,2p) basis set for H, C, and S, and (3) ECEP-
121G, corresponding to the CEP-121G ECP38 basis set for Mo
and the 6-311++G(2d,2p)39 basis set for C, H, and S. For
examination of S-S bonding in structures 6, geometries were
optimized at B3LYP/LANL2DZ ECP (Mo) and 6-31G* (CHS)
with single-point energies calculated at B3LYP/LANL2DZ ECP
(Mo) and 6-311++G(2d,2p) (CHS). Analytical gradient cal-
culations were employed to confirm ground states and to correct
electronic energies to ambient temperature enthalpy values. Total
enthalpy corrections to the electronic energy at 298 K were
computed according to eq 30:40
cm-1 were replaced by /2RT. The electronic energy for the
hydrogen atom was taken as -0.5 hartree. The present approach
resembles the methodology of DiLabio et al.42 with exception
of the use of UB3LYP treatment of open shell systems.
Electronic energies and total thermal corrections are presented
in Table 3, and calculated bond dissociation energies are
presented in Table 4. Cartesian coordinates for B3LYP/
LANL2DZ geometries are available in the Supporting Informa-
tion.
Experimental Section
Reagents. [CpMo(µ-S)(µ-SH)]2 was synthesized using a literature
procedure.8 Phenylbenzyl ketone (PBK) and dibenzyl ketone (DBK)
were purchased from Aldrich and recrystallized from methanol or
purified by radial chromatography on silica gel using methylene
chloride/hexane eluent. Benzene was purchased from Aldrich and triply
fractionally distilled to reduce trace toluene to ∼10-7 M. 1-Octanethiol
was purchased from Aldrich and distilled. Cyclohexane, tert-butylben-
zene, and dichloromethane were purchased from Aldrich and used
without further purification.
Rate Constants for Reaction of Benzyl Radical with [CpMo(µ-
S)(µ-SH)]2, 1-Octanethiol, and 2-Mercaptonaphthalene. Pyrex reac-
tion tubes (5 mm) containing 200 µL of solution (0.02 M DBK or
0.02 M PBK, 10-4 M [CpMo(µ-S)(µ-SH)]2 (or 0.01 M CH3(CH2)7SH,
or 2 × 10-3 M 2-mercaptonaphthalene), and 10-3 M tert-butylbenzene
in benzene) were freeze-thaw degassed 3 times and flame sealed with
an approximate headspace of 600 ( 50 µL. The tubes were placed
inside a thermostated oven equipped with a quartz window and allowed
to achieve temperature equilibrium for 1 min. The sample was
photolyzed with a water-filtered 1-kW high-pressure xenon arc lamp
for controlled periods of time (0.50-20.00 ( 0.005 s) using a Uniblitz
model 225L0AOT522952 computer-controlled electronic shutter. Re-
agent concentrations were corrected for solvent vapor pressure and
density.43 Rate constants for hydrogen abstraction were calculated using
eq 10c at 0.5-8.0 s photolysis times, using the measured product
concentrations, the calculated average molybdenum-thiol donor con-
centration, and the benzyl self-radical termination rate, 2kt, calculated
from the expression ln(2kt/M-1 s-1) ) 27.23-2952.5/RT. This expres-
sion was calculated using the von Smoluchowski equation, eq 33, using
the Spernol-Wirtz modification of the Debye-Einstein equation, eq
34, where f in eq 34 is the SW microfriction factor.44
H298° ) 3/2RT + nRT +
k,b
R
(Θ (1/ + 1/(eΘ /T - 1))) + RT (30)
∑
k,a 2
k
Θk,x ) (c1)hc/kbλk
x ) a, c1 ) 0.9806
x ) b, c1 ) 0.9989
n ) 1, linear
n ) 3/2, nonlinear
Equation 30 includes a translational, a rotational, an internal
(ZPE and vibrational), and a PV term. Each vibration, λk, was
scaled by 0.9806 for the ZPE term (x ) a) and by 0.9989 for
the vibrational term (x ) b).41 Each vibration was examined to
identify group rotations, and scaled internal rotations under 260
(32) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (b) Becke,
A. D. J. Chem. Phys. 1993, 98, 5648.
(33) Dunning, T. H., Jr.; Hay, P. J. In Modern Theoretical Chemistry; Schaefer,
H. F., III, Ed.; Plenum: New York, 1976; Vol. 3, p 1.
(34) (a) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. (b) Wadt, W.
R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284. (c) Hay, P. J.; Wadt, W. R.
J. Chem. Phys. 1985, 82, 299.
2kt ) (8π/1000)σFDAB
DAB ) kT/6πrAηf
N
(33)
(34)
(35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.11; Gaussian,
Inc.: Pittsburgh, PA, 2001.
Fischer and co-workers45 have measured self-termination rate
expressions for benzyl radical in cyclohexane, ln(2kt/M-1 s-1) ) 26.2-
2497.6/RT, and in toluene, ln(2kt/M-1 s-1) ) 27.05-2797.4/RT. For
self-termination of benzyl in hexane, we previously developed the
expression ln(2kt/M-1 s-1) ) 26.0-1803.6/RT.46 Parameters for estima-
(38) (a) Stevens, W.; Basch, H.; Krauss, J. Chem. Phys. 1984, 81, 6026. (b)
Stevens, W. J.; Krauss, M.; Basch, H.; Jasien, P. G. Can. J. Chem. 1992,
70, 612. (c) Cundari, T. R.; Stevens, W. J. J. Chem. Phys. 1993, 98, 5555.
(39) (a) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1980, 72, 650. (b) Blaudeau, J.-P.; McGrath, M. P.; Curtiss, L. A.; Radom,
L. J. Chem. Phys. 1997, 107, 5016. (c) Clark, T.; Chandrasekhar, J.;
Schleyer, P.v. R. J. Comput. Chem. 1983, 4, 294.
(36) Straatsma, T. P.; Apra, E.; Windus, T. L.; Dupuis, M.; Bylaska, E. J.; de
Jong, W.; Hirata, S.; Smith, D. M. A.; Hackler, M. T.; Pollack, L.; Harrison,
R. J.; Nieplocha, J.; Tipparaju, V.; Krishnan, M.; Brown, E.; Cisneros, G.;
Fann, G. I.; Fruchtl, H.; Garza, J.; Hirao, K.; Kendall, R.; Nichols, J. A.;
Tsemekhman, K.; Valiev, M.; Wolinski, K.; Anchell, J.; Bernholdt, D.;
Borowski, P.; Clark, T.; Clerc, D.; Dachsel, H.; Deegan, M.; Dyall, K.;
Elwood, D.; Glendening, E.; Gutowski, M.; Hess, A.; Jaffe, J.; Johnson,
B.; Ju, J.; Kobayashi, R.; Kutteh, R.; Lin, Z.; Littlefield, R.; Long, X.;
Meng, B.; Nakajima, T.; Niu, S.; Rosing, M.; Sandrone, G.; Stave, M.;
Taylor, H.; Thomas, G.; van Lenthe, J.; Wong, A.; Zhang, Z. NWChem, A
Computational Chemistry Package for Parallel Computers, version 4.5;
Pacific Northwest National Laboratory: Richland, WA 99352-0999, U.S.A.,
High Performance Computational Chemistry Group, Pacific Northwest
National Laboratory, Richland, WA 99352, U.S.A., 2003.
(40) McQuarrie, D. A. Statistical Mechanics; Harper and Row: New York, 1976.
(41) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502.
(42) DiLabio, G. A.; Pratt, D. A.; LoFaro, A. D.; Wright, J. S. J. Phys. Chem.
A 1999, 103, 1653-1661.
(43) (a) Reid, R. C.; Prausnitz, J. M.; Sherwood, T. K. The Properties of Liquids
and Gases; McGraw-Hill: New York, 1977. (b) Weast, R. C.; Astle, M. J.
Handbook of Chemistry and Physics; CRC Press: Boca Raton, FL, 1982-
83.
(44) Fischer, H.; Paul, H. Acc. Chem. Res. 1987, 20, 200.
(45) (a) Lehni, M.; Schuh, H.; Fischer, H. Int. J. Chem. Kinet. 1979, 11, 705.
(b) Claridge, R. F. C.; Fischer, H. J. Phys. Chem. 1983, 87, 1960. (c)
Huggenberger, C.; Fischer, H. HelV. Chim. Acta 1981, 64, 338.
(37) Check, C. E.; Faust, T. O.; Bailey, J. M.; Wright, B. J.; Gilbert, T. M.;
Sunderlin, L. S. J. Phys. Chem. A 2001, 105, 8111.
9
6690 J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004

Activation of the Sulfhydryl Group by Mo Centers
A R T I C L E S
tion of the rate expression for self-termination of benzyl radical in
benzene employed in these kinetics, ln(2kt/M-1 s-1) ) 27.23-2952.5/
RT, are listed.47 The maximum error in estimation of self-termination
rate constants by this method is estimated to be less than ca. 25%.44
Photolysis of 1,3-Dibenzyl Ketone (DBK) and 1 in Benzene-d6
for Product Analysis. A solution of 1 (0.005 M) and DBK (0.06 M)
in degassed benzene-d6 was photolyzed at room temperature with the
diffused, water-filtered output of a 1-kW high-pressure Xe lamp for
times varying from 0 to 3000 s. The photolysis solution was analyzed
freeze-thaw degassed solution of 1 (12 mg) and benzylthiol (3.5 uL)
in CDCl3 (1 mL) was heated at 65 °C for 18 h under 2 atm of H2 in a
sealed thick-walled NMR tube. 1H NMR revealed complete conversion
of 1 to 3a (5.6%), 3b (9.4%), 8a (25%), and 8b (60%). NMR data are
presented in Table 2. (The identical reaction carried out in benzene-d6
at room temperature for 8 days gave no exchange. When heated at
80°C for 5 h in benzene-d6, 1 was converted to 3 in ca. 5% yield.)
Fast Atom Bombardment Mass Spectrometric Analysis. The
above exchange experiment in CDCl3 was repeated but with heating
at 50 °C for 15 h and terminated at ca. 50% conversion of 1. The major
products were 3a, 3b, and minor products were 8a, 8b. Solvent was
removed in a vacuum, and the product mixture was analyzed by fast
atom bombardment/secondary ion mass spectrometry (FAB/LSIMS).
Multiple ion peaks corresponding to matrix-induced peaks (CpMo)2S3
(417.6), (CpMo)2S4 (449.6), and (CpMo)3S4 (610.5) were observed
together with major peak groups corresponding to (CpMo)2S4(H)(CH2-
Ph) (3) (540.6) and (CpMo)2S4(CH2Ph)2 (8) (632.5).
1
by H NMR to obtain the spectra of Figure 3. Products phenylacetal-
dehyde, toluene, 3a,b, and 8 were observed.
Photolysis of DBK and 1 in Benzene-d6 with Prompt H NMR
1
Analysis. In the experiment above, the transient product 4 was not
observed, possibly due to its decay during extended data acquisition
times. Thus, a freeze-thaw degassed and sealed solution of DBK (0.05
M) and 1 (0.005 M) in benzene-d6 in a Pyrex NMR tube was photolyzed
with the unfiltered light of a 1-kW high-pressure Xe lamp. 50%
conversion of 1 was obtained within several minutes, followed by
prompt (within 5 min) analysis of the product mixture. At approximately
50% conversion of 1, the observed products were toluene, phenyl-
acetaldehyde, and 3a and 3b, the latter identified by the characteristic
pairs of Cp, benzylic CH2, and sulfhydryl groups (see Table 2). The
Cp resonance assigned to 4 was present at low concentration, with a
ratio 4/(3a + 3b) of 0.04.
Fast Atom Bombardment Mass Spectrometric Analysis. The
above photolysis experiment in benzene-d6 was repeated but terminated
at ca. 50% conversion of 1. The major products were 3a, 3b, and minor
products were 8a, 8b. Solvent was removed in a vacuum, and the
product mixture was analyzed by fast atom bombardment/electrospray
ionization mass spectrometry (FAB/ESI). Multiple ion peaks corre-
sponding to matrix-induced peaks (CpMo)2S3 (417.6), (CpMo)2S4 (449),
and (CpMo)3S4 (613) were observed together with major peak groups
corresponding to (CpMo)2S4(H)(CH2Ph) (3) (541) and (CpMo)2S4(CH2-
Ph)2 (8) (633).
Attempted Chromatographic Fractionation of Exchange Prod-
ucts of PhCH2SH and 1 in CHCl3. In a 200-mL vacuum flask, 420
mg of 1 (0.93 mmol), and 210 µL of benzylmercaptan (1.8 mmol) were
dissolved in 100 mL of CHCl3 and heated at 65 °C for 23 h. The
reaction was evacuated to dryness and transferred in CH2Cl2 to an
alumina column in a nitrogen glovebox. Three 30-mL fractions were
collected with CH2Cl2 as eluent and taken to dryness, redissolved in
CDCl3 and benzene-d6, and analyzed by 1H and 13C NMR. Fraction 1
contained 12% 1, 7% 3a, 12% 3b, 1% 8a, 68% 8b. Fraction 2 contained
39% 1, 11% 3a, 18% 3b, 27% 8a, 6% 8b. Fraction 3 contained 27%
1
1, 3% 3a, 5% 3b, 64% 8a, 1% 8b. Careful spectroscopic (13C and H
NMR) analysis of the fractions in CDCl3 and benzene-d6 allowed
complete 1H and 13C spectral assignment of all isomers (Table 2). Note
that the absolute a,b structural assignments of isomers of 3 and the
a,b assignments of 8 are not established.
Oxidation of 1 to 5 in CDCl3 and Benzene-d6. A sample of 1 (15
mg) and azo-benzene (15 µL) in 1 mL of oxygen-free CDCl3 was sealed
1
Photolysis of ButOOBut in Benzene-d6. Observation and Decay
of Transient 5. A solution of 1 (0.002 M) and ButOOBut (0.17 M) in
benzene-d6 in a freeze-thaw degassed and sealed Pyrex NMR tube
was photolyzed with a 1-kW unfiltered high-pressure xenon lamp. After
10 min of photolysis at room temperature, a single Cp peak at 5.78
ppm, assigned to 5, was present at 2.7 × 10-4 M at 5 min of elapsed
time following the end of photolysis. At 15, 25, 35, 55, 75, and 95
min, the concentration of 5 was found to be 2.64, 2.62, 2.58, 2.09,
1.86, and 1.63 × 10-4 M, corresponding to a half-life of 120 min, due
to precipitation of highly insoluble 5. No new sulfhydryl peaks were
observed corresponding to 6 (e.g., in the region -1 to -3 ppm).
Exchange Reaction of 1 with PhCH2SH to Form 3 and 8. Similar
in an NMR tube and then examined by H NMR spectroscopy as a
function of time over a period of 12 h. 1 (Cp in CDCl3 at 6.38 ppm)
was cleanly oxidized within 1 h to give 1,2-diphenylhydrazine and a
single new Cp peak at 6.34 ppm, 5. Over a period of 12 h, the peak
assigned to 5 decreased in intensity to about 20% of peak value and
less than ca. 1% by 20 h, accompanied by the deposition of solid
material (reflecting very low solubility of 5). The sample was opened
on a vacuum line, charge with H2 (1 atm) and resealed. Warming at 80
°C for 10 min led to restoration of 1 along with complete conversion
of the remaining azo-benzene to 1,2-diphenylhydrazine. A second
sample of 1, 15 mg, and azo-benzene in 1 mL of benzene-d6 was sealed
in an NMR tube. 1 (Cp at 5.89 ppm in benzene-d6) was cleanly and
to a procedure employed by Rakowski-DuBois and co-workers,8
a
1
quantitatively converted to a single product with a Cp peak in the H
NMR spectrum at 5.78 ppm, 5. The Cp resonance of 5 decayed to
about 20% of its initial intensity over a period of about 10 h.
(46) Franz, J. A.; Suleman, N. K.; Alnajjar, M. S. J. Org. Chem. 1986, 51, 19.
(47) Diffusion coefficients DAB, eq 7, were calculated for estimation of the self-
termination rate of benzyl radical in benzene using the following data. For
the radical model, toluene: mp 178 K, bp 383.8 K, MW 92.14, F ) 6.02
× 10-8 cm (average of van der Waals (Edward, J. T. J. Chem. Educ. 1970,
47 (4), 261), LeBas (Ghai, R. L.; Dullien, F. A. L. J. Phys. Chem. 1974,
78, 2283), and other (Spernol-Wirtz Spernol, A.; Wirtz, K. Z. Naturforsch.
Acknowledgment. This work supported by the Office of
Science, Office of Basic Energy Sciences, of the U.S. Depart-
ment of Energy under Contract DE-ACO6-76RLO 1830. The
authors thank Dr. Robert Barkley of the Central Analytical
Laboratory, University of Colorado, Boulder, CO, for perform-
ing FAB analysis of reaction products of 1.
1
1953, 8a, 522) diameters, σ ) /4, density of toluene, d(T, K) ) 1.1787-
1.048 × 10-3 (T, K). For the solvent benzene, mp 278.7, bp 353.3, MW
78.1, Andrade viscosity, ln(η, cP) ) -4.117 + 2.0973 × 103/RT, d(T, K)
) 1.192-1.059 × 10-3(T, K). The microfriction factor ×a6 of Spernol
r
and Wirtz is given by ×a6 ) (0.16 + 0.4rA/rB)(0.9 + 0.4TA - 0.25TBr),
and reduced temperatures are given by TXr ) (T - TXf)/(TXb - TXf), where
f
b
TX and TX are freezing and boiling points of species X ) benzyl (A) or
Supporting Information Available: Tables of kinetic data,
computational summaries from electronic structure calculations,
and supporting spectral data. This material is available free of
charge via the Internet at http://pubs.acs.org.
benzene (B). The radii in the microfriction factor term are given by rX
)
(3VX(ø)/4πN)1/3, where ø ) 0.74, the volume fraction for cubic closest
packed spheres, and VX are density-based molecular volumes, and N is
Avogadro’s number. The resulting Smoluchowski expression for self-
termination of the benzyl radical in benzene is ln(2kt/M-1 s-1) ) 27.23-
2952.47/RT. A similar treatment for benzyl self-termination in toluene is
in near-perfect agreement with experimental values (ref 42).
JA049321R
9
J. AM. CHEM. SOC. VOL. 126, NO. 21, 2004 6691