Formation of rhodium carbonyl thiolate dimers via elimination of dihydrogen; Crystal and molecular structure of [Rh2(CO)4(μ-SC6H4CH3)2]

Article abstract of DOI:10.1016/S0022-328X(00)00657-4

Reaction of [Rh₆(CO)₁₅(NCMe)] with p-thiocresol [(4-Me)C₆H₄SH] leads to the formation of [Rh₂(CO)₄(μ-SC₆H₄CH₃)₂] as the main product along with a small amount of [Rh₆(CO)₁₆]. An approximately 30-fold excess of the thiol is required in order to obtain a good yield of the thiolate-bridged dimer while reaction of [Rh₄(CO)₁₂] with an excess of p-thiocresol leads to an apparently clean conversion to the dimeric Rh(I) complex. Mass spectrometric measurements show that the latter reaction involves evolution of H₂, and CO evolution is indicated by the retardation of the reaction in CO saturated solution; these results suggest the following reaction stoichiometry: [Rh₄(CO)₁₂]+4RSH→2[Rh₂(CO)₄(μ-SR)₂]+2H₂+4CO. Kinetic measurements show that the reaction proceeds in three stages which are proposed to involve two rapid pre-equilibria and a final irreversible and relatively slow conversion to the products. The crystal and molecular structure of [Rh₂(CO)₄(μ₂-SC₆H₄CH₃)₂] is reported.

Full text of DOI:10.1016/S0022-328X(00)00657-4

Journal of Organometallic Chemistry 623 (2001) 65–73
www.elsevier.nl/locate/jorganchem
Formation of rhodium carbonyl thiolate dimers via elimination of
dihydrogen; crystal and molecular structure of
[Rh2(CO)4(m-SC6H4CH3)2]
Nitsa K. Kiriakidou-Kazemifar a, Matti Haukka b, Tapani A. Pakkanen b,
Sergey P. Tunik* c,1, Ebbe Nordlander* a,2
a Inorganic Chemistry 1, Chemical Center, Lund Uni6ersity, Box 124, SE-221 00 Lund, Sweden
b Department of Chemistry, Uni6ersity of Joensuu, PO Box 111, FIN-80101 Joensuu, Finland
c Department of Chemistry, St. Petersburg Uni6ersity, Uni6ersitetskii pr., 2, St. Petersburg 198904, Russian Federation
Received 7 June 2000; accepted 5 September 2000
Abstract
Reaction of [Rh6(CO)15(NCMe)] with p-thiocresol [(4-Me)C6H4SH] leads to the formation of [Rh2(CO)4(m-SC6H4CH3)2] as the
main product along with a small amount of [Rh6(CO)16]. An approximately 30-fold excess of the thiol is required in order to
obtain a good yield of the thiolate-bridged dimer while reaction of [Rh4(CO)12] with an excess of p-thiocresol leads to an
apparently clean conversion to the dimeric Rh(I) complex. Mass spectrometric measurements show that the latter reaction
involves evolution of H2, and CO evolution is indicated by the retardation of the reaction in CO saturated solution; these results
suggest the following reaction stoichiometry: [Rh4(CO)12]+4RSH“2[Rh2(CO)4(m-SR)2]+2H2+4CO. Kinetic measurements
show that the reaction proceeds in three stages which are proposed to involve two rapid pre-equilibria and a final irreversible and
relatively slow conversion to the products. The crystal and molecular structure of [Rh2(CO)4(m2-SC6H4CH3)2] is reported. © 2001
Elsevier Science B.V. All rights reserved.
Keywords: Rhodium; Thiol; Cluster; Hydrodesulfurization; Mechanism
1. Introduction
of sulfur-containing ligands to transition metal car-
bonyl clusters and the reactivity of metal carbonyl
clusters containing such ligands. Our aims are to relate
observed coordination modes to possible intermediates
in HDS catalysis and to investigate the viability of
using clusters as stoichiometric reagents (or catalysts) in
homogeneous desulfurization reactions. Our efforts
have thus far been concentrated on ruthenium and
osmium clusters [6–8] as it has been shown that the
sulfides of these two metals exhibit high HDS activity
[2]. The activities of rhodium and iridium sulfides are
comparable to those of ruthenium and osmium, and we
have therefore begun to investigate the reactivities of
rhodium clusters towards sulfur-containing ligands.
Here, we wish to report the redox reactions of two
rhodium carbonyl clusters with p-thiocresol, which in
both cases lead to the formation of [Rh2(CO)4(m2-
SC6H4CH3)2]. Reaction of the thiol with [Rh4(CO)12]
leads to the formation of the dimeric product in high
The commercial and environmental importance of
hydrodesulfurization (HDS) catalysis serves as an im-
petus for research on the mechanism(s) of desulfuriza-
tion of thiols, thioethers and thiophenes in
heterogeneous and homogeneous systems in which
transition metal sulfides or coordination complexes
serve as catalysts or stoichiometric reagents [1–4]. Be-
cause of the experimental difficulties encountered dur-
ing in situ studies of commercial catalysts (MoꢀCoꢀS or
MoꢀNiꢀS phases) [2], the use of metal complexes and
metal clusters has become an important alternative
means to study HDS mechanisms [3–5]. We are cur-
rently involved in a broad study of coordination modes
1 *Corresponding author.
2 *Corresponding author. Tel.: +46-46-108118; fax: +46-46-
2224439; e-mail: ebbe.norlander@inorg.lu.se
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00657-4

66
N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
yield with concomitant evolution of dihydrogen. A
preliminary investigation of the kinetics of the latter
redox reaction and a proposal for its mechanism is also
described.
2.2. Reaction of [Rh4(CO)12] with p-thiocresol
[Rh4(CO)12] (40 mg, 0.05 mmol) was dissolved in
hexane (20 ml) and solid p-thiocresol (53 mg, 0.43
mmol) was added to the solution. The mixture was
stirred under nitrogen until no wCꢀO resonances of the
starting material could be detected by IR spectroscopy.
The solution was concentrated under reduced pressure
and subjected to preparative TLC (dichloromethane–
hexane, 2:3, 6:6) to give [Rh2(CO)4(m-SC6H4CH3)2] (1)
(56 mg, 93%) as the main product. Trace amounts of
two other yellow compounds with lower Rf than 1 were
detected but were not characterized.
2. Experimental
The clusters [Rh4(CO)12] [9] and [Rh6(CO)15(NCMe)]
[10] were prepared by literature methods. para-Thiocre-
sol (Lancaster Chemicals), tetrabutylammonium bro-
mide (Acros) and carbon monoxide (AGA) were used
as received. All reactions were carried out under an
atmosphere of dry, oxygen free nitrogen (unless other-
wise stated), using solvents which were freshly distilled
from appropriate drying agents. Infrared spectra were
recorded in n-hexane or dichloromethane solution in
0.5 mm NaCl cells, using a Nicolet Avatar FTIR
spectrometer. 1H-NMR spectra were recorded on a
Varian Unity 300 MHz spectrometer using the solvent
resonance as an internal standard. Thin layer chro-
matography was performed on commercial plates pre-
coated with Merck Kieselgel 60–0.5 mm thickness.
Products are given in order of decreasing Rf values.
Fast atom bombardment (FAB+) and electron impact
(EI+) ionization mass spectra were obtained on a JEOL
SX-102 instrument. The matrix for the FAB+ spectra
consisted of 3-nitrobenzyl alcohol and CsI was used as
calibrant. The EI+ spectra were obtained using a PLOT
2.3. Detection of e6olution of dihydrogen during the
reaction of [Rh4(CO)12] with p-thiocresol
Solid [Rh4(CO)12] (50 mg, 0.07 mmol) and
[(4-Me) C6H4SH] (66 mg, 0.54 mmol) was put into one
compartment of
a
‘double-horn’ Schlenk vessel,
and 20 ml of hexane into the other one. The hexane was
degassed using freeze–pump–thaw techniques and the
vessel was filled with CO (the use of argon or nitrogen
as an inert component of the system would make it
impossible to carry out the chromatography of the gas
phase as described below). The reaction was started by
transferring hexane into the compartment of the
Schlenk vessel which contained the solid reagents. The
mixture was stirred for 1 h, and heated to +50°C to
transfer gaseous products from solution into gas phase.
A sample of the gas-phase was removed with a syringe
and injected into a mass spectrometer at 50°C (EI+).
The composition of the gas-phase was analyzed by
GC–MS which unambiguously displayed the peak of
H2. A control experiment was performed under exactly
the same conditions, but in the absence of
[Rh4(CO)12] — no peak due to H2 could be detected in
this case.
,
5 A molecular sieve column (30 m×0.32 mm×30 mm)
and CO as the carrying gas. All kinetic runs were
monitored with an Applied Photophysics SX.18MV
stopped flow spectrophotometer interfaced with a PC
driven by PC Pro-K software. The temperature was
maintained constant at 303 K with a Lauda RM6
thermostat bath.
2.1. Reaction of [Rh6(CO)15(NCMe)] with p-thiocresol
[(4-Me)C6H4SH]
2.4. Monitoring of the reaction of [Rh4(CO)12] with
p-thiocresol
A total of 170 mg of p-thiocresol (1.35 mM) were
dissolved in 30 ml of dichloromethane and 35 mg of
solid [Rh6(CO)15(NCMe)] (0.032 mM) were added to
the solution. The reaction mixture was left stirring for 4
h at room temperature (r.t.) after which the solution
was concentrated to a minimal amount and separated
by TLC using a hexane–dichloromethane mixture (5:1,
6:6) as eluant. Two products were detected and iso-
lated: dark reddish-brown [Rh2(CO)4(m-SC6H4CH3)2]
(1) (32 mg, 0.057 mmol) and [Rh6(CO)16] (8 mg, 7.5
mmol). Rh2(CO)4(m-SC6H4CH3)2 (1) IR [wCO (cm−1),
CH2Cl2]: 2081w, 2063s, 2015s, 1963sh; 1H-NMR [l
(ppm), CDCl3]: 7.62 [d, (SC6H4CH3), 2H], 7.07 [d,
(SC6H4CH3), 2H], 2.33 [s, (SC6H4CH3), 3H], MS
[FAB+, m/z, amu] 565 (M+), loss of 4 CO observed.
2.4.1. IR
A 20-fold excess of p-thiocresol was added to a 0.13
mM solution of [Rh4(CO)12] in hexane which was
prethermostated to 30°C. The resultant solution was
put into an IR cell and the reaction course was moni-
tored by IR in the 2200–1500 cm−1 range. A clean
conversion
SC6H4CH3)2] (1) was observed.
of
[Rh4(CO)12]
into
[Rh2(CO)4(m-
2.4.2. Stopped-flow spectrometry
Kinetic runs were carried out in hexane under
pseudo-first-order conditions using at least 20-fold ex-
cess of the thiol. The [Rh4(CO)12] concentration used in
all experiments was 1.2×10−4 M. The reaction course

N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
67
Table 1
for each concentration of the thiol ligand. The fre-
quency of the data acquisition was varied in the ranges
0–2, 2–20 s, and 0–20, 20–500 s, respectively, in order
to acquire maximum amount of points for each reac-
tion stage. The first two stages, which were complete in
2 and 60 s, respectively, were accompanied by relatively
small changes in the absorbance at 400 nm, while the
final stage was accompanied by substantial UV ab-
sorbance changes and was assigned to final product
accumulation. The observed rate constants for the three
reaction steps were obtained by treatment of the data
from the time intervals 0–2, 0–20 and 0–500 s using
double exponential analysis to take into account over-
lapping successive kinetics. In order to investigate the
effect of carbon monoxide on the rate of the reaction,
one kinetic run was carried out in a CO saturated
solution; CO was bubbled through solutions of both
reagents for 15 min prior to the start of the reaction.
Crystal data and structure refinement for [Rh2(CO)4(m-SC6H4CH3)2]
(1)
Chemical formula
Molecular weight (g mol−1
Crystal size (mm)
Color
Crystal system
Space group
C18H14O4Rh2S2
564.23
0.3×0.2×0.05
Orange–red
Triclinic
)
(
P1
Unit cell dimensions
,
a (A)
8.912(18)
9.975(2)
12.623(3)
68.37(3)
78.62(3)
71.25(3)
983.8(3)
2
,
b (A)
,
c (A)
h (°)
i (°)
k (°)
3
,
V (A )
Z
Dcalc (Mg m−3
)
1.905
1.909
552
Absorption coefficient v (mm−1
F(000)
)
q range (°)
h range
k range
l range
3.91–27.46
−11–11
−12–12
−16–16
8057
4090
3678
291
2.5. X-ray data collection and structure solution for
[Rh2(CO)4(v-SC6H4CH3)2] (1)
No. of collected reflections
No. of unique reflections
No. of observed reflections a
No. of parameters
Final R indices a
Single crystals of 1 suitable for X-ray analysis were
obtained by slow evaporation of a pentane solution of
the complex at r.t. The X-ray data was collected at 120
K on a Nonius Kappa CCD diffractometer using f-
0.0238
0.0593
1.080
a
wR2
,
Goodness-of-fit
scans [11] and Mo Ka radiation (u=0.71073 A). The
T (K)
120(2)
cell refinement and data reduction were carried out
using the Denzo and Scalepack programs [12]. The
structure was solved by direct methods using SHELXS97
[13] and structure refinement was carried out with the
SHELXL97 program [14]. All non-hydrogen atoms were
refined anisotropically. Hydrogens were located from
the difference Fourier map and refined isotropically.
a I\2|(I), w=1/[|2(Fo2)+(xP)2+yP], P=(Fo2+2F2c)/3.
was followed by monitoring absorbance changes at 400
nm by stopped flow spectrometry. The reaction kinetics
were studied in two time intervals, 0–20 and 0–500 s,
Fig. 1. An ORTEP plot of the molecular structure of [Rh2(CO)4(m-SC6H4CH3)2] (1) showing the atom labeling scheme. Thermal ellipsoids are
drawn at the 50% probability level.

68
N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
Table 2
for both clusters, leaving a coordinatively unsaturated
and two-electron deficient intermediate which readily
reacts with nucleophiles [15,16]. In an attempt to detect
,
Selected bond lengths (A) and angles (°) for [Rh2(CO)4(m-
SC6H4CH3)2] (1)
a
putative agostic intermediate of the form
Bond lengths
[Rh6(CO)15(RSH)], [Rh6(CO)15(NCMe)] was reacted
with p-thiocresol in dichloromethane solution. In the
presence of considerable excess of the thiol the main
product of this reaction was identified as [Rh2(CO)4(m2-
SC6H4CH3)2] (1) on the basis of its IR, NMR and mass
spectra, which were similar to previously prepared
[Rh2(CO)4(m2-SR)2] compounds [17–23]. In addition, it
was possible to grow crystals of 1 suitable for X-ray
analysis and the solid state structure of 1 was deter-
mined by a single-crystal X-ray diffraction study (vide
infra).
Rh(1)ꢀC(1)
Rh(1)ꢀS(1)
Rh(1)ꢀS(2)
Rh(1)ꢀRh(2)
Rh(2)ꢀC(3)
Rh(2)ꢀC(4)
Rh(2)ꢀS(1)
Rh(2)ꢀS(2)
C(1)ꢀO(1)
C(2)ꢀO(2)
C(3)ꢀO(3)
C(4)ꢀO(4)
1.880(3)
2.3504(8)
2.3571(11)
3.1041(7)
1.869(3)
1.880(3)
2.3662(12)
2.3701(10)
1.132(3)
1.133(4)
1.131(3)
1.134(3)
Bond angles
The outcome of the reaction of [Rh6(CO)15(NCMe)]
is dependent on the relative amount of thiol. The
parent cluster, [Rh6(CO)16] is always formed as a side
product, even when a 30-fold excess of the thiol is used.
The use of only a slight (1.5-fold) excess of the ligand
results in a very low yield (B10%) of [Rh2(CO)4(m2-
SC6H4CH3)2] while the main product is [Rh6(CO)16]. In
the latter reaction, conversion of the starting material
to [Rh6(CO)16] occurs in 40 min at r.t. In the absence of
thiol, the relatively labile [Rh6(CO)15(NCMe)] is trans-
formed into [Rh6(CO)16] in an intermolecular CO ‘re-
distribution’ reaction which leads to approximately 50%
conversion in approximately 24 h at r.t. [24]. It is
plausible that the reaction of [Rh6(CO)15(NCMe)] with
thiol is of a bimolecular nature where an initial labile
[Rh6(CO)15(RSH)] species undergoes a redistribution/
disproportionation reaction to form [Rh6(CO)16] in the
presence of relatively small concentrations of free thiol,
while in the presence of a relatively large excess of the
thiol, further reaction of [Rh6(CO)15(RSH)] with thiol
to form 1 occurs.
C(2)ꢀRh(1)ꢀC(1)
C(2)ꢀRh(1)ꢀS(1)
C(1)ꢀRh(1)ꢀS(1)
C(2)ꢀRh(1)ꢀS(2)
C(1)ꢀRh(1)ꢀS(2)
S(1)ꢀRh(1)ꢀS(2)
C(2)ꢀRh(1)ꢀRh(2)
C(1)ꢀRh(1)ꢀRh(2)
S(1)ꢀRh(1)ꢀRh(2)
S(2)ꢀRh(1)ꢀRh(2)
C(3)ꢀRh(2)ꢀC(4)
C(3)ꢀRh(2)ꢀS(1)
C(4)ꢀRh(2)ꢀS(1)
C(3)ꢀRh(2)ꢀS(2)
C(4)ꢀRh(2)ꢀS(2)
S(1)ꢀRh(2)ꢀS(2)
C(3)ꢀRh(2)ꢀRh(1)
C(4)ꢀRh(2)ꢀRh(1)
S(1)ꢀRh(2)ꢀRh(1)
S(2)ꢀRh(2)ꢀRh(1)
C(5)ꢀS(1)ꢀRh(1)
C(5)ꢀS(1)ꢀRh(2)
Rh(1)ꢀS(1)ꢀRh(2)
C(11)ꢀS(2)ꢀRh(1)
C(11)ꢀS(2)ꢀRh(2)
Rh(1)ꢀS(2)ꢀRh(2)
91.38(12)
95.68(8)
172.39(9)
173.92(8)
93.93(9)
78.87(3)
125.05(8)
123.98(9)
49.06(3)
49.14(3)
90.75(12)
170.62(9)
96.37(9)
94.84(8)
174.04(9)
78.30(3)
130.37(8)
125.51(9)
48.62(2)
48.77(3)
111.56(9)
116.13(9)
82.31(3)
109.15(9)
114.53(9)
82.09(3)
3.2. Crystal and molecular structure of
[Rh2(CO)4(v2-SC6H4CH3)2] (1)
Relevant crystallographic data are summarized in Table
1.
The
molecular
structure
of
[Rh2(CO)4(m2-
SC6H4CH3)2] (1) is shown in Fig. 1; crystallographic
data and selected bond lengths and angles are given in
Table 2. The structure of 1 closely resembles those of
[Rh2(CO)4(m-SR)2] (R=Ph, Et [22], SC6H4F [23]). Two
thiolate bridges join the two rhodium atoms to form a
central wingtip Rh2S2 fragment; the conformation of
the molecule may be described as a bent syn, endo
structure, the latter notation referring to the relative
orientation of the thiolate substituents with respect to
each other and the metal ligands, respectively [25]. The
sulfur atoms are coordinated to the Rh2 center in a
slightly asymmetric manner, the differences between the
long and short metalꢀsulfur bonds for two sulfur atoms
3. Results and discussion
3.1. Reaction of [Rh6(CO)15(NCMe)] with p-thiocresol
We have earlier reported on the reaction of the
labilized cluster [Os3(CO)11(NCMe)] with thiols to form
products of the general formula [(m-H)Os3(CO)10(m-
SR)]; this reaction proceeds via the formation of an
intermediate which contains an agostic Os–H–SR in-
teraction [7]. It is known that the reactivity of
[Rh6(CO)15(NCMe)]
is
similar
to
that
of
,
[Os3(CO)11(NCMe)] in that substitution of the labile
acetonitrile ligand proceeds via a dissociative pathway
is 0.015 and 0.013 A. The Rh2S wings of the central
core form a dihedral angle of 114.4°, which is very close

N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
69
to the values found for the two crystallographically
independent molecules in [Rh2(CO)4(m-SC6H5)2] (115.2
and 113.7°, respectively [22]). The RhꢀRh intramolecu-
sistent with a metalꢀmetal bond. The angular parame-
ters of the ligand environment of each rhodium atom
range from 90.8 to 96.4° for the CꢀRhꢀC and CꢀRhꢀS
bond angles and are equal to 78.9 and 78.3° for the
SꢀRhꢀS angles. Deviations of the ligands from ideal-
ized planar arrangement are rather small and do not
,
lar distance, 3.104(1) A, is slightly longer than those in
,
[Rh2(CO)4(m2-SR)2] (R=Ph: RhꢀRhav=3.092 A;
,
C6H4F: RhꢀRhav=3.073 A). These distances are longer
,
than usual RhꢀRh bonding contacts in cluster com-
exceed 0.05 A. Thus, the rhodium atoms are in slightly
,
,
pounds, e.g. 2.73 A in [Rh4(CO)12] [26] and 2.78 A in
[Rh6(CO)16] [27] and imply a very weak Rh–Rh inter-
action, if any. A recent calculational study indicates
that the molecular energy is stabilized by a weak Rh–
Rh interaction in the model [Rh2(CO)4(m-SH)2] [25]. On
the other hand, it should be noted that the intermetal
distorted square planar coordination geometries, in ac-
cordance with their formal oxidation states of +1 (d8).
3.3. Reaction of [Rh4(CO)12] with p-thiocresol
Previous syntheses of [Rh2(CO)4(m2-SR)] complexes
[17–20] involve reaction of [Rh2(CO)4Cl2] with thiols
,
distance in [Rh2(CO)4(m-SEt)2] [22], 2.820(2) A, is con-
Fig. 2. (a) GC trace of blank experiment for detection of hydrogen gas (cf. Section 2.3). (b) GC trace of gaseous products from the reaction of
[Rh4(CO)12] with p-thiocresol. (c) Mass spectrum showing the presence of H2 (m/z=2 amu), H2O (18 amu, artifact) and CO (28 amu, carrier gas)
in the above-mentioned reaction.

70
N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
Fig. 3. Time-resolved IR spectra for the reaction of [Rh4(CO)12] with p-thiocresol. An expansion of the 2060–2000 cm−1 region (inset) shows two
of the isosbestic points at approximately 2055 and 2026 cm−1. The spectra have been collected with a resolution of 2 cm−1 and smoothed using
a Savitsky–Golay algorithm [34].
resulting in substitution of the bridging chlorides by
thiolate moieties. The concomitant HCl evolution ap-
pears to be a thermodynamic driving force for this
reaction; the thiolate complexes that are formed are
relatively unstable in comparison with the parent chlo-
ride derivative [17,19]. Reaction of [Rh4(CO)12] with
p-thiocresol in hexane or dichloromethane at r.t. also
leads to an apparently clean and quantitative conver-
sion to [Rh2(CO)4(m2-SR)] (1). The presence of atmo-
spheric oxygen does not display any effect on the
composition of the reaction products and the reaction
rate (vide infra). It has been proposed that reactions of
mono-, di- and trinuclear iron carbonyl compounds
[28–30] and [M3(CO)12] (M=Ru, Os) [31] to form
metal carbonyl thiolate clusters proceeds via evolution
of H2; however, direct detection of dihydrogen has
hitherto not been observed in such reactions. In order
to confirm that H2 is indeed formed in the reaction of
[Rh4(CO)12] with p-thiocresol, the reaction system was
monitored by GC–MS (cf. Section 2.3). The results
unambiguously show that H2 is formed during the
reaction while a blank experiment (cf. Section 2.3) did
not display any detectable evolution of dihydrogen
(Fig. 2). It thus appears that the reaction stoichiometry
may be summarized by Eq. (1).
[Rh2(CO)4(m-SR)2] with isosbestic points at 2026, 2055
and 2078 cm−1 remaining throughout the reaction
(Fig. 3). The reaction was monitored at 2069 and 2063
cm−1, respectively ([L]=3.2×10−3 M), and a kinetic
curve was obtained. The observed rate constants were
obtained using single exponential analysis and an aver-
age value was calculated (kobs=0.004 s−1). No IR
resonances other than those attributable to the starting
material and the product could be detected. Thus, no
appreciable amounts of any intermediate species could
be detected by IR spectroscopy under the time scale
applied (0–900 s). Neither could any hydride or other
intermediate species be detected by 1H-NMR
spectroscopy.
However, kinetic results obtained using stopped flow
techniques showed that Reaction (1) is a multistage
process, which includes two sequential preequilibria
and a final irreversible process to form the binuclear
product. These three successive stages of the reaction
have been detected in various time intervals; experimen-
tal kinetic curves along with the corresponding fits are
shown in Fig. 4. At 303 K and [Rh4(CO)12]=1.2×
10−4 M, the first two stages are complete in 2 and 60 s,
respectively, while the third, relatively slow stage occurs
in the 60–900 s time interval, which is in agreement
with the above-mentioned IR experiments. The kinetic
parameters for these three stages were obtained using
double exponential analysis to avoid ‘mixing effects’ of
the sequential reaction kinetics; the rate constants ob-
tained at various ligand concentrations are given in
Table 3. The rate constants for the first two stages
appear to be independent of ligand concentration while
the rate constant for the third increases slightly with
increasing ligand concentration. Further studies cover-
ing a wider thiol concentration range are currently
[Rh4(CO)12]+4RSH
“2[Rh2(CO)4(m-SR)2]+2H2+4CO
3.4. Kinetic/mechanistic studies
(1)
In order to gain insight into the mechanistic features
of Reaction (1), a preliminary kinetic study has been
undertaken. Monitoring of the reaction by IR spec-
troscopy indicates clean conversion of [Rh4(CO)12] into

N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
71
Table 3
being undertaken; these studies confirm the dependence
a
Rate constants for the reaction of [Rh4(CO)12
]
and p-thiocresol in
of the rate of the third stage on the thiol concentra-
tion [33]. The absence of a dependence of the first
two observed rate constants on ligand concentra-
tion implies unimolecular rate determining steps
for both reactions. It is likely that these reactions
involve activation of the cluster by, for example, CO
dissociation or metalꢀmetal bond scission, to generate
vacant coordination sites. The observed rate constant
for the third stage, which is derived from stopped-flow
hexane b
[HSC6H4CH3]
kobs1 (s−1
)
kobs2 (s−1
)
kobs3 (s−1
)
(mol dm−3
)
0.003 (IR
0.00490.001
monitoring) c
0.0025
0.005
0.01
0.01 (CO
saturated
solutions)
0.02
3.0890.17 0.05690.004 0.003290.0001
2.4490.18 0.06790.005 0.003790.0002
2.5790.12 0.06590.003 0.003790.0001
0.002390.002
measurements (kobs=3.2×10−3 −1, [L]=2.5×10−3
s
M), is close to that obtained via IR spectroscopy (vide
2.3390.24 0.05190.002 0.005090.0002
a [Rh4(CO)12]=1.2×10−4 M.
b Temperature=303 K, u=400 nm.
c [Rh4(CO)12]=1.3×10−4 M.
supra) within the limits of experimental uncertainty.
The reaction scheme depicted in Scheme 1 may explain
the experimental observations.
For the three detected reaction stages, each subse-
quent step is slower than the preceding one (cf. Table
3). This should normally result in accumulation of
reaction products from each stage of the reaction and it
should thus be possible to observe reaction intermedi-
ates but, as previously mentioned, no intermediates
could be detected by IR spectroscopy. This fact along
with the isosbesticity of the obtained spectra implies a
very low concentration of the products formed in stages
A and B of Scheme 1. The proposed mechanism can
match the experimental data if both the A and B stages
are completely reversible and the equilibria are shifted
−2
A,B
to the left, i.e. kA1 ,B, k ꢀk2A,B, k−A,B2; this also explains
the low concentration of the intermediates formed in
these stages. The shift of the equilibria is in agreement
with the expected lability of the thiol ligand in the
coordination sphere of the Rh4 cluster. The reversibility
of the A and B stages does not prevent single exponen-
tial treatment of the kinetics of each individual stage;
however, the observed rate constants are in this case a
combination of the constants for the forward and back
reactions. The above-mentioned assumptions imply that
the kobs values for the A and B stages should be
independent of ligand concentration, as is observed
experimentally. It should be noted that the present data
does not permit the distinction between intermediates
of
a
dissociative nature, e.g. ‘[Rh4(CO)11]’,
‘[Rh4(CO)10(HSR)]’, and intermediates that are formed
by opening of the cluster framework/metalꢀmetal bond
scission, e.g. ‘[Rh4(CO)12]*’, ‘[Rh4(CO)11(HSR)]*’ (‘but-
terfly’ clusters).
In order to probe whether carbon monoxide retards
the reaction between [Rh4(CO)12] and p-thiocresol, the
kinetics of the reaction in CO saturated hexane solution
were studied by stopped-flow spectrometry. Under
Fig. 4. Kinetic traces for the three stages detected in the reaction of
[Rh4(CO)12] with p-thiocresol at 303 K; [Rh4(CO)12]=1.2×10−4 M,
[SC6H4CH3]=1×10−2 M. Observed data (×, ꢀ, ꢁ) and theoreti-
cal fits (solid lines) for (a) all three stages (0–600 s) (b) the first and
second stage (0–24 s) (c) the first stage (0–2 s).

72
N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
Scheme 1. Proposed mechanism for the reaction of [Rh4(CO)12] with p-thiocresol.
these conditions, only a slow reaction that could be
fitted to a single exponential function could be detected
tion, and CO evolution is indicated by the retardation
of the reaction rate(s) in CO-saturated solution. These
observations are in accordance with the stoichiometry
in Eq. (1) (vide supra). Although elimination of dihy-
drogen in reactions involving transition metal carbonyl
clusters have been proposed earlier, we provide here, to
our knowledge, the first experimental evidence for such
a reaction.
Kinetic studies of the latter reaction by IR spec-
troscopy and stopped-flow spectrometry show that it
proceeds in three stages, involving two preequilibria
and a final relatively slow conversion to the products
(Scheme 1). A more thorough kinetic study which aims
to establish rate laws for each reaction stage as well as
activation parameters is currently in progress.
(kobs=2.3×10−3
s−1
for
[RSH]=0.01
M,
[Rh4(CO)12]=1.2×10−4 M); this observation is in
agreement with Scheme 1. The presence of excess CO
shifts the A and B equilibria further to the left, making
the corresponding UV absorbances negligible and pre-
venting detection of the two initial stages of the reac-
tion. The observed rate constant, which thus
corresponds to the third stage of the reaction, is consid-
erably smaller than that measured in the absence of
CO.
We propose that the intermediate that is formed as a
product in equilibrium B is [Rh4(CO)10(RSH)2]. A pos-
sible mechanism for subsequent formation of dihydro-
gen involves oxidative addition of an SꢀH bond at a Rh
center or across a RhꢀRh bond to generate a thiolate
ligand and a hydride. The very low affinity of
[Rh4(CO)12] and [Rh6(CO)16] towards hydrides is well
known and a hydride species is therefore expected to be
labile [32]; the hydride may combine with the proton of
the second (coordinated) thiol to form H2. The dihy-
drogen evolution is accompanied by the formation of
two thiolate bridges and cluster core cleavage to afford
the binuclear Rh(I) product and a coordinatively unsat-
urated [Rh2(CO)y] species. The latter species subse-
quently adds two further thiol ligands and undergoes a
reaction sequence similar to that described to complete
the reaction stoichiometry (cf. Scheme 1).
5. Supplementary material
Observed and calculated mass spectra for
[Rh2(CO)4(m-SC6H4CH3)2] (four pages) are available on
request. Crystallographic data for the structural analy-
sis have been deposited with the Cambridge Crystallo-
graphic Data Centre, CCDC no. 143964. Copies of
this information may be obtained free of charge
from The Director, CCDC, 12 Union Road, Cam-
bridge CB2 1EZ, UK (Fax: +44-1223-336033; e-mail:
deposit@ccdc.cam.ac.uk
or
www:
http://www.ccdc.cam.ac.uk).
4. Summary and conclusions
Acknowledgements
Reaction of [Rh6(CO)15(NCMe)] with a large excess
of p-thiocresol leads to the formation of [Rh2(CO)4(m-
SC6H4-p-Me)]. On the other hand, the presence of a
relatively low concentration of thiol appears to catalyze
the conversion of [Rh6(CO)15(NCMe)] into [Rh6(CO)16]
in a CO redistribution reaction. Reaction of the thiol
with [Rh4(CO)12] leads to quantitati6e conversion to
[Rh2(CO)4(m-SC6H4-p-Me)] with concomitant H2 evolu-
This research has been sponsored by the Swedish
Natural Science Research Council (NFR), the Swedish
Engineering Research Council (TFR), The Nordic
Council of Ministers, the Royal Swedish Academy of
Sciences and the European Union (TMR Network
Metal Clusters in Catalysis and Organic Synthesis). We
thank the Wenner-Gren Center Foundation for a visit-
ing professorship for S.P.T., Professor Lars Ivar Elding

N.K. Kiriakidou-Kazemifar et al. / Journal of Organometallic Chemistry 623 (2001) 65–73
73
for the generous loan of a stopped-flow spectrometer
and helpful advice, Professor Anthony J. Poe¨ for help-
ful discussions and Einar Nilsson for experimental as-
sistance in the detection of dihydrogen.
[13] G.M. Sheldrick, SHELXS-97, Program for Crystal Structure De-
termination, University of Go¨ttingen, Germany, 1997.
[14] G.M. Sheldrick, SHELXL-97, Program for Crystal Structure
Refinement, University of Go¨ttingen, Germany, 1997.
[15] K. Dahlinger, A.J. Poe¨, P.K. Sayal, V.C. Sekhar, J. Chem. Soc.
Dalton Trans. (1986) 2145.
[16] A.J. Poe¨, S.P. Tunik, Inorg. Chim. Acta 268 (1998) 189.
[17] E.S. Bolton, R. Havlin, G.R. Knox, J. Organomet. Chem. 18
(1969) 153.
[18] J.V. Kingstone, G.R. Scollary, J. Inorg. Nucl. Chem. 33 (1971)
4373.
[19] W. Hieber, K. Heinicke, Z. Naturforsch. Teil. B 16 (1961) 554.
[20] J. Cooke, M. Green, F.G.A. Stone, J. Chem. Soc. (A) (1968)
170.
[21] P. Kalck, R. Poilblanc, Inorg. Chem. 14 (1975) 2779.
[22] J. Pursiainen, T. Teppana, S. Rossi, T.A. Pakkanen, Acta Chem.
Scand. 47 (1993) 416.
[23] C. Claver, A.M. Masdeu, N. Ruiz, C. Foces-Foces, F.H. Cano,
M. Carmen Apreda, L.A. Oro, J. Garcia-Alejandre, N. Torres, J.
Organomet. Chem. 398 (1996) 177.
[24] S.P. Tunik, unpublished results.
[25] G. Aullo´n, G. Ujaque, A. Lledo´s, S. Alvarez, Chem. Eur. J. 5
(1999) 1391.
[26] C.N. Wei, G.R. Wilkes, L.F. Dahl, J. Am. Chem. Soc. 89 (1967)
4792.
[27] E.R Corey, L.F. Dahl, W.J. Beck, J. Am. Chem. Soc. 85 (1963)
1202.
[28] K. Farnery, M. Kilner, J. Chem. Soc. (A) (1970) 634.
[29] N.S. Nametkin, V.D. Tyurin, M.A. Kukina, J. Organomet.
Chem. 149 (1978) 355.
References
[1] R.J. Angelici, Polyhedron (1997) 3073.
[2] H. Topsøe, B.S. Clausen, F.E. Massoth, Hydrotreating catalysis,
in: J.R. Anderson, M. Boudart (Eds.), Catalysis-Science and
Technology, vol. 11, Springer–Verlag, Berlin, 1996.
[3] (a) C. Bianchini, A. Meli, Acc. Chem. Res. 31 (1988) 109. (b) C.
Bianchini, J.A. Casares, A. Meli, V. Sernan, F. Vizza, R.A.
Sa´nchez-Delgado, Polyhedron (1997) 3099.
[4] M. Brorson, K. Kiriakidou, J.D. King, F. Prestopino, E. Nord-
lander, Metal clusters as models for hydrodesulfurization cata-
lysts, in: P. Braunstein, L. Oro, P.R. Raithby (Eds.), Metal
Clusters in Chemistry, vol. 2, Wiley–VCH, Weinheim, 1999, pp.
741–781.
[5] T.B. Rauchfuss, Prog. Inorg. Chem. 39 (1991) 259.
[6] M. Monari, R. Pfeiffer, U. Rudsander, E. Nordlander, Inorg.
Chim. Acta 247 (1996) 131.
[7] K. Kiriakidou, M.R. Plutino, F. Prestopino, M. Monari, L.I.
Elding, E. Valls, R. Gobetto, S. Aime, E. Nordlander, J. Chem.
Soc. Chem. Commun. (1998) 2721.
[8] J.D. King, M. Monari, E. Nordlander, J. Organomet. Chem. 573
(1999) 272.
[9] S. Martinengo, P. Chini, G. Giordano, J. Organomet. Chem. 27
(1971) 389.
[30] J.A. de Beer, R.J. Haines, J. Organomet. Chem. 24 (1970) 757.
[31] B.F.G. Johnson, R.D. Johnston, P.L. Josty, J. Lewis, I.G.
Williams, Nature 213 (1967) 901.
[32] B.T. Heaton, L. Strona, S. Martinengo, R.J. Goodfellow, I.H.
Sadler, J. Chem. Soc., Dalton Trans. (1982) 1499.
[33] I.O. Kochevoi, S.P. Tunik, A.J. Poe¨, unpublished results.
[34] OMNIC spectroscopic software v. 4.1, Nicolet Instrument Corpo-
ration, Madison, Wisconsin, USA, 1997
[10] S.P. Tunik, A.V. Vlasov, V.V. Kryvikh, Inorg. Synth. 31 (1996)
239.
[11] COLLECT, Data collection software, Enraf–Nonius, Delft, The
Netherlands, 1998.
[12] Z. Otwinowski, W. Minor, Processing of X-ray diffraction data
collected in oscillation mode, in: C.W. Carter, Jr., R.M. Sweet
(Eds.), Macromolecular Crystallography, part A, Methods in
Enzymology, vol. 276, Academic Press, 1997, pp. 307–326.
.
.