A kinetic investigation of the oxidative addition reactions of the dimeric Bu4N[Ir2(μ-Dcbp)(CO)2(PCy3) 2] complex with iodomethane

Article abstract of DOI:10.1016/j.jorganchem.2010.10.049

The IR and UV/visible kinetic results of the oxidative addition of iodomethane to Bu₄N[Ir₂(μ-Dcbp)(CO) ₂(PCy₃)₂] (Dcbp = 3,5-dicarboxylatepyrazolate anion) showed three time separable reactions. The first, very fast reaction corresponds to the a Ir(I)-Ir(III) alkyl species formation within 10 ^-3 s. The second, relative fast reaction corresponds to Ir(III)-Ir(III) alkyl formation with a rate constant of 3.25(4) × 10 ^-2 M^-1 s^-1 while the third and slowest reaction corresponds to Ir(III)-Ir(III) acyl formation with a rate constant of 1.42 × 10^-5 s^-1. The IR data clearly show the existence of a number of equilibria with the formation of an Ir(I)-Ir(III) alkyl product which then react to form the Ir(III)-Ir(III) which then slowly react to form the Ir(III)-Ir(III) acyl product. A solvent study indicated increased oxidative addition activity in the presence of polar solvents, which is indicative of a polar transition state. The large negative entropy of activation for the Ir(III)-Ir(III) alkyl formation step (k₂) of -178(23) JK^-1 mol^-1 is indicative of an associative process. DFT calculations successfully identified the stereochemistry of the starting complex, [Ir ₂(μ-Dcbp)(CO)₂(PCy₃)₂] ^- as well as that of the Ir-alkyl and acyl isomers. A reaction pathway, using the IR data and DFT calculations, is proposed for the reaction.

Full text of DOI:10.1016/j.jorganchem.2010.10.049

Journal of Organometallic Chemistry 696 (2011) 1990e2002
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
A kinetic investigation of the oxidative addition reactions of the dimeric Bu4N
[Ir2(m-Dcbp)(CO)2(PCy3)2] complex with iodomethane
,
b
,
a
a
,
a
Ebeth Grobbelaar a, Jeanet Conradie a , Johan A. Venter , Walter Purcell
*
**
, T.T. Chiweshe
a Department of Chemistry, University of the Free State, Bloemfontein 9300, South Africa
b Center for Theoretical and Computational Chemistry and Department of Chemistry, University of Tromsø, Tromsø N-9037, Norway
a r t i c l e i n f o
a b s t r a c t
Article history:
The IR and UV/visible kinetic results of the oxidative addition of iodomethane to Bu4N[Ir2(m-Dcbp)
Received 12 May 2010
Received in revised form
15 October 2010
(CO)2(PCy3)2] (Dcbp ¼ 3,5-dicarboxylatepyrazolate anion) showed three time separable reactions. The first,
very fast reaction corresponds to the a Ir(I)eIr(III) alkyl species formation within 10ꢀ3 s. The second, relative
fast reaction corresponds to Ir(III)eIr(III) alkyl formation with a rate constant of 3.25(4) ꢁ 10ꢀ2 Mꢀ1 sꢀ1
while the third and slowest reaction corresponds to Ir(III)eIr(III) acyl formation with a rate constant
of 1.42 ꢁ 10ꢀ5 sꢀ1. The IR data clearly show the existence of a number of equilibria with the formation of an Ir
(I)eIr(III) alkyl product which then react to form the Ir(III)eIr(III) which then slowly react to form the
Ir(III)eIr(III) acyl product. A solvent study indicated increased oxidative addition activity in the presence of
polar solvents, which is indicative of a polar transition state. The large negative entropy of activation for the
Ir(III)eIr(III) alkyl formation step (k2) of ꢀ178(23) JKꢀ1 molꢀ1 is indicative of an associative process. DFT
Accepted 21 October 2010
Keywords:
Oxidative addition
Iridium
3,5-Dicarboxylatepyrazolate
Iodomethane
Equilibrium
calculations successfully identified the stereochemistry of the starting complex, [Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ
Acyl formation
as well as that of the Ir-alkyl and acyl isomers. A reaction pathway, using the IR data and DFTcalculations, is
proposed for the reaction.
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
World consumption of acetic acid was about 6 million metric
tons in 2000, which was mainly used as raw material in the
The importance of iridium complexes as catalysts was under-
lined in 2000 when the [Ir(CO)2(I)2]- complex (Cativa) replaced the
rhodium analogue (Monsanto) for the first time as catalyst in a BP
plant in Malaysia for the carbonylation of methanol to produce
acetic acid [1]. In addition to cheaper iridium prices, benefits such
as increased throughput due to faster reaction rates, milder reac-
tion conditions needed for acetic acid production and the robust-
ness of the catalyst under a wide variety of experimental conditions
were achieved from this change in catalyst. Iridium complexes are
also used in a number of other industrial processes. [Ir(cod)(PCy3)
py]þ is used in the reduction of steroids and the hydrogenation of
alkenes and alkynes while the [Ir(CO)Cl(PPh3)2] complex is used in
the conversion of higher alcohols and diols to carboxylic acids as
well as the polymerisation of styrene [2].
production of key petrochemical products and intermediates such
as vinyl acetate monomer, acetate esters, cellulose acetate, mono-
chloroacetic acid, etc. The majority of the acetic acid production
(65%) was synthesized by methanol carbonylation technology. One
of the key steps in the carbonylation of methanol to produce acetic
acid involves the oxidative addition of methyl iodide to the metal
complex, which was approximately 150 times faster for the Ir
complex compared to the rhodium complex [3]. With the change in
metal centres the rate determining step changed to the insertion of
CO into the IreCH3 bond to from the subsequent acyl intermediate.
Reductive elimination of the CH3COI group with the subsequent
reaction with water completed the acetic acid formation [1].
From these results it should be clear that the metal centre plays an
important role in the rate of the catalytic process and possiblyas well as
to the selectivity of the products that are formed in this process.
Research done by Garcia et al. [4e7] indicated an increase in catalytic
activity when the monomeric complex is replaced by dimeric metal
complexes. It was found for instance that the catalytic hydrogenation of
* Corresponding author. Department of Chemistry, University of the Free State,
Bloemfontein 9300, South Africa. Tel.: þ27 514012200; fax: þ865255382.
** Corresponding author.
E-mail addresses: conradj@ufs.ac.za (J. Conradie), purcellw@ufs.ac.za (W.
Purcell).
cyclohexene by [H(CO)(PPh3)2Ru(
m
-bim)M(cod)] (bim ¼ 2,20 bi-imi-
dazolate, M ¼ Rh, Ir) is higher than the corresponding monomeric
complex. In the case of the production of acetic acid from methanol the
0022-328X/$ e see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2010.10.049

E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
1991
2. Experimental
2.1. General considerations
All the preparations were performed under a dry nitrogen
atmosphere and unless otherwise stated all the chemicals were
reagent grade and used without further purification. IR spectra
were recorded with a Hitachi 270-50 spectrophotometer while the
NMR spectra were obtained at 293 K on a Bruker 300 and 600 MHz
spectrometer. The chemical analyses were performed at Mikroa-
nalytisches Labor Pascher in Remagen-Bandorf, Germany.
Fig. 1. General chemical presentation of the di-nuclear [Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ
complex.
2.2. Synthesis of tetrabutylammonium
pyrazolatocarbonyl- tricyclohexylphosphine iridate(I)], (Bu4N)
[Ir2( -Dcbp)(CO)2(PCy3)2]
m-3,5-dicarboxylato-
presence of two metal ions open the possibility of different oxidative
addition patterns such as reversible and irreversible oxidative addition
and addition on only one or both metal ions [8]. Another new exciting
possibility in these type of reactions is the possibility of co-operative
effects that may take place between the different metal ions which can
lead to metalemetal bond formation and/or the migration of the
added fragments between the two metal ions [9e12]. Oxidative
addition on one metal ion in some cases can lead to the deactivation of
the second metal ion to the same reaction. This type of deactivation is
attributed to the electron donation from the metal ion in the lower
oxidation state tothe metal ion in the higheroxidation statewhich may
result in the activation of the one metal ion towards oxidative addition
while the other metal centre may then be activated for reductive
elimination which can be very useful in the design of catalysts [13].
Two recent studies [14,15] on di-nuclear rhodium complexes con-
taining flexible trans-chelating diphosphines clearly demonstrated the
increased reactivity of di-nuclear complexes towards oxidative addi-
tion reactions. The current study is an extension of the kinetic inves-
tigation that was recently completed on the Ir-Dcbp cod dimeric
complex [16], Dcbp ¼ 3,5-dicarboxylatepyrazolate anion.
m
(Bu4N)[Ir2(m-Dcbp)(cod)2] [13] (0.1 g, 0.14 mmol) was dissolved in
35 ml acetone. Carbon monoxide was bubbled for 10 min through the
solution, while stirring and maintaining a nitrogen atmosphere over
the solution. The bright yellow solution changed to light yellow
during this process. The formation of the (Bu4N)[Ir2(m-Dcbp)(CO)4]
complex was confirmed with liquid IR spectra (
n
(CO) ¼ 2077 and
1987 cm-1 [17]). Tricyclohexylphosphine (PCy3) (0.22 g, 0.8 mmol)
was added slowly to the first solution while stirring and maintaining
a nitrogen atmosphere. A light yellow precipitate formed after
45 min. The solution was stirred for another 2 h after which the
volume of the solution was reduced to half of its original volume
using a flow of nitrogen gas. The precipitate was centrifuged, washed
with ether and dried in a vacuum desiccator over P2O5 for 8 h. Yield:
63%, IR data (KBr): n(OCO) ¼ 1617 mꢀ1
,
n(CO) ¼ 1932 cmꢀ1, Elemental
analysis: (calculated values in brackets): C, 49.6 (50.95); H, 7.10
(7.03); N, 2.35 (3.01), O, 8.25 (6.87), P, 3.75 (4.43), Ir, 27.01 (27.59)%.1H
NMR (CDCl3), Dcbp,
0.95 (12H, t), PCy3, NBu4, 2.20e1.10, (38H, m),13C NMR (CDCl3,
77.01 MHz) 206.6 (CO), 172.4 (Dcbp), 65.8 (NBu4), 33.6 (PCy), 30.9
(NBu4), 30.0 (Dcbp), 27.7 (PCy3), 27.6 (PCy3), 26.3 (PCy3),19.7 (NBu4),
15.2 (NBu4), 31P NMR (CDCl3, 36.44 MHz)
30.51.
(s ¼ singlet, t ¼ triplet and m ¼ multiplet).
d6.75 ppm (1H, s), NBu4, 3.25 (8H, s), 2.58 (8H, s),
d
Table 1
DFT molecular energies (eV) of the possible stereo isomers of [Ir2(m-Dcbp)
(CO)2(PX3)2]ꢀ (X ¼ Cy or H) relative to the lowest energy isomer as zero.
d
Geometry
Energy/eV X ¼ Cy Energy/eV X ¼ H
2.3. Kinetics
The solvents used during the kinetic study were purified and
dried prior to use by standard procedures [18]. The kinetic
measurements were performed at atmospheric pressure on a GBC
(Model 916) and a Varian (Cary 50) UV/vis spectrophotometer
equipped with a thermostatted 6 cell changer (0.1 ꢂC) while the IR
spectra and kinetic runs were performed on a Hitachi 270-50
0.97
0.38
0.00
0.13
0.10
0.00
spectrophotometer using acetone as solvent (2200e1850 cmꢀ1
)
and KBr -disks (4000e251 cmꢀ1).
The oxidative addition reactions with CH3I were followed on
UV/vis at 425.0 nm with the exception of acetonitrile and dichloro-
methane, which were followed at 420.0 nm. Typical experimental
conditions were [Ir2-complex] ¼ 0.5e5.5 ꢁ10ꢀ4 M and [CH3I] varied
between 1.0 ꢁ 10ꢀ2 and 0.2 M thus ensuring good pseudo first order
kinetics. The observed first order rate constants where calculated
obstÞ
according to At ¼ AN þ ðA0 ꢀ ANÞeðꢀk
with At , AN and A0 the
absorbance of the indicated species at time t, N and 0 respectively,
using a non-linear least-square program [19]. The experimentally
determined pseudo first order rate constants were converted to
second orderrate constants, k2, by determining the slope of thelinear
plots of kobs against the concentration of the incoming iodomethane
ligand. The metal complex solutions were freshly prepared before
each experiment to minimise the extent of the solvolysis reaction.

1992
E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
Fig. 2. The OLYP/TZP optimized geometry of [Ir2(
d(IreIr) ¼ 4.44 Å. At the top is a front and side view of the molecule to illustrate the spatial orientation of the ligands attached to Ir(I). H atoms have been omitted for clarity.
m .
-Dcbp)(CO)2(PCy3)2]ꢀ Selected bond lengths (Å) and angles (ꢂ) are shown. The IreNeNeIr dihedral angle is 41.8ꢂ.
Fig. 3. The first (black arrows) and second (blue arrows) steps in the oxidative addition
step between CH3I (0.0113 M) and (Bu4N)[Ir2(
m
-Dcbp)(CO)2(PCy3)2] (4.9 ꢁ 10ꢀ4 M)
Fig. 4. Ir(III)-acyl formation (red arrow) as the third step in the oxidative addition reac-
with the formation of the Ir(III)-alkyl complex in dichloromethane, T ¼ 25.0 ꢂC, with
2 min intervals. Inset: The absorbance-time data measured at the indicated wave-
lengths for the second step. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article).
tion between CH3I (0.176 M) and (Bu4N)[Ir2(
m
-Dcbp)(CO)2(PCy3)2] (0.024 M), ¼ 25.0 ꢂC
and time intervals of 60 min. Inset: The absorbance-time data measured at 1714 cmꢀ1 for
the third step. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article).

E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
1993
Table 2
Kinetic results and activation parameters (forward and reverse) for the oxidative
addition reaction between CH3I and (Bu4N)[Ir2(
temperatures in dichloromethane.
m
-Dcbp)(CO)2(PCy3)2] at different
1
0.5
0
(a)
(b)
Temperature (ꢂ C)
(102) k2 (Mꢀ1 sꢀ1
)
(103) k (sꢀ1
)
K2 (Mꢀ1
a
)
ꢀ2
15.1
25.0
34.9
2.46(3)
3.25(4)
5.61(5)
1.40(3)
2.30(5)
2.50(5)
17.6(4)
14.1(3)
22.4(5)
399 nm
414 nm
D
D
Hs (kJ molꢀ1
Ss (J K-1 mol-1
)
28(7)
ꢀ178(23)
17(9)
ꢀ238(30)
)
(c)
s
¼ Activation.
a
K2 ¼ k2/k
.
ꢀ2
350
400
450
500
of the starting complex indicate that a fair presentation of the
expected formula has been made. 1H, spectra also verified the
presence of the Dcbp anion through its weak methine signal
wavelength / nm
(d
6.75 ppm) whilst the multiplet signals of the phosphine and NBu4þ
Fig. 5. Spectral changes during the reaction between Bu4N[Ir2(
m
-Dcbp)(CO)2(PCy3)2]
(4.9 ꢁ 10ꢀ4 M) and CH3I (0.390 M) in dichloromethane (2 min intervals, 25.0 ꢂC).
species dominated the NMR spectrum. 13C and 31P NMR are also
consistent with the expected formulation of the starting complex.
In orderto continuewith theidentification of thestereochemistry
2.4. Computational study
of the di-nuclear [Ir2(
tional study by means of density functional theory has been per-
formed on the three possible isomers of [Ir2(
-Dcbp)(CO)2(PCy3)2]ꢀ.
m
-Dcbp)(CO)2(PCy3)2]ꢀ complex, a computa-
Density Functional Theory (DFT) calculations were carried out
using the Amsterdam Density Functional 2007 (ADF) program
system [20] with the OLYP [21,22] generalized gradient approxi-
m
The isomer with the PCy3 groups trans to N, as expected from trans
influence considerations, is the most stable, see Table 1. The most
unstable isomer is the isomer with the PCy3 groups adjacent to each
other and trans to O, while the most stable isomer has the PCy3
groups in the sterically unhindered position on the “outside” of
the molecule. To further investigation of the sterical effect of the
mation (GGA). The ZORA/TZP [23] (Triple
z polarized) all electron
basis set, a fine mesh for numerical integration (5.5 for geometry
optimizations and 6.0 for frequency calculations), a spin-restricted
(gas-phase) formalism and full geometry optimization with tight
convergence criteria as implemented in the ADF 2007 program,
were used. Optimized structures were verified as a minimum
through frequency calculations. No symmetry limitations were
imposed on the calculations.
PCy3 groups, calculations on a simplified version, [Ir2(m-Dcbp)
(CO)2(PH3)2]ꢀ were performed which led to the same results, sug-
gesting that the trans influence also plays an important role (as
opposed to the sterical effect of PCy3) in the substitution of the
specific CO by a phosphine group in [Ir2(
highlights of [Ir2(
-Dcbp)(CO)2(PX3)2]ꢀ are presented in Fig. 2. The
most significant feature of the structure of the complex is the
distortion from planar geometry. The -Dcbp backbone is still flat as
in the uncoordinated -Dcbp ligand [25], while the iridium atoms are
m
-Dcbp)(CO)4]ꢀ. Structural
m
3. Results and discussion
m
3.1. Synthesis of complexes
m
located slightly above and below the plane of the ligand with an
The novel dimeric (Bu4N)[Ir2(
by the substitution reaction of CO for PCy3 in (Bu4N)[Ir2(
m
-Dcbp)(CO)2(PCy3)2] was obtained
IreNeNeIr dihedral angle of 41.8ꢂ. Similar distortion from planar
m-Dcbp)
geometry is observed in the crystal structure of [Rh2(m
-Dcbp)(CO)4]ꢀ
(CO)4]. Since the nitrogen atom of the heterocyclic ring has a larger
trans influence than the oxygen atom [24], a structure with the PCy3
group trans to N is expected as illustrated in Fig. 1. Symmetric
substitution of one CO on each iridium(I) centre is compatible with
the single absorption observed at 1932 cmꢀ1 in the CO region of its
infrared spectrum. The results obtained from the chemical analysis
[17], with a dihedral angle of 30.3ꢂ. The geometry around the
two iridium atoms is square-planar, with iridium located within
0.09 Å from the plane defined by PeCCOeNDcbpeODcbp, similar as the
distance between Rh and the CCOeCCOeNDcbpeODcbp plane in
10
-8
dichlormethane
34.9°C
16
-9
8
12
-10
6
25.0°C
0.0032 0.0034 0.0036
-1
1/T / K
tetrahydrofuran
8
4
chloroform
15.1°C
4
1,2-dichlorethane
2
benzene
0
0
0.00
0.10
0.20
0.00
0.10
0.20
-3
-3
[CH3I] / mol dm
[CH3I] / mol dm
Fig. 6. kobs vs [CH3I] for the oxidative addition reaction (alkyl formation) between CH3I
and (Bu4N)[Ir2( -Dcbp)(CO)2(PCy3)2] in dichloromethane (380.0 nm).
Fig. 7. kobs vs [CH3I] for the oxidative addition reactions between CH3I and (Bu4N)
[Ir2(
-Dcbp)(CO)2(PCy3)2] (alkyl formation) in different solvents (25 ꢂC).
m
m

1994
E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
1658 cmꢀ1 for the duration of the reaction. These studies indicated
that the OCO peak remains fairly stable for the duration of the study
which confirmed the stability of the complex.
Table 3
Kinetic results for the oxidative addition reactions between CH3I and (Bu4N)[Ir2(
Dcbp)(CO)2(PCy3)2] (alkyl formation) in different solvents at 25.0 ꢂC.
m-
(102) k2 (Mꢀ1 sꢀ1) (103) k-2 (sꢀ1) K2 (Mꢀ1
)
D
Gob (kJ molꢀ1
)
a
Solvent
Dichloromethane 3.25(4)
2.30(5)
0.29(2)
1.12(5)
0.33(1)
0.292(8)
14.1(3)
58(4)
13.3(7)
38(1)
ꢀ6.6(1)
ꢀ10.0(7)
ꢀ6.4(3)
ꢀ9.0(2)
ꢀ3.6 (1)
3.2. Kinetics
Tetrahydrofuran
Chloroform
1.68(2)
1.49(4)
1,2-Dichlorethane 1.26(1)
3.2.1. Infrared study
Benzene
0.128(6)
4.38(1)
The oxidative addition reaction of CH3I to [Ir2(m-Dcbp)
(CO)2(PCy3)2]ꢀ was first explored on IR since this technique allow to
a
K ¼ k2/k
.
Go ¼ ꢀꢀR2TlnK.
b
D
distinguish between metal-CO bonds in M(I)-carbonyl (resonates at
w1900e2000 cmꢀ1), M(III)-alkyl (resonates at w2000e2100 cmꢀ1
)
and M(III)-acyl (resonates at w1700e1750 cmꢀ1) species, M ¼ metal
such as Rh [29] or Ir [30]. Three reactions were observed on the IR.
The first and second reactions are illustrated in Fig. 3 and the third
reaction in Fig. 4.
[Rh2(m
-Dcbp)(CO)4]ꢀ. The dihedral angles formed by the pyrazole
ring and the two coordination meanplanes are 26.3ꢂ (Ir1) and ꢀ24.9ꢂ
(Ir2) (ca 20ꢂ (Rh1) and ꢀ10ꢂ (Rh2) for [Rh2(
m
-Dcbp)(CO)4]ꢀ). A
The addition of CH3I to [Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ resulted in
consequence of this distortion is an Ir1-Ir2 distance equal to 4.44 Å
(4.53 Å for [Rh2(m
-Dcbp)(CO)4]ꢀ, which is somewhat longer than ca
a very fast first reaction which is observed by an initial fast disap-
pearance of the iridium(I) peak at 1932 cmꢀ1 with the simulta-
neous formation of a peak at 2004 cmꢀ1. This first reaction was too
fast to measure on the timescale of the IR.
3.9 Å that was found in related structures with the flat undistorted
anions coordinated to two metals (M ¼ Pd, Cu or Zn) [26].
The UV/visible spectra of the complex that was dissolved in the
different solvents used in this study, showed no appreciable
absorption change within 120 min, indicating the relative stability
of the starting complex in these solvents. Upon addition of CH3I to
A second, slower reaction, resulted in the CO peak formation at
2028 cmꢀ1 (Ir(III) formation) (Fig. 3) with the simultaneous
disappearance of the CO peak at 1932 cmꢀ1 (Ir(I) disappearance)
which appears to be shifted to 1942 cmꢀ1. Kinetic results performed
on the second reaction showed that the reaction rate measures at
1942 cmꢀ1 (disappearance of CO) of 2.4(1) ꢁ 10ꢀ3 sꢀ1 is within
experimental error the same as that at 2028 cmꢀ1 (formation of
alkyl species), 2.5(1) ꢁ 10ꢀ3 sꢀ1 (dichloromethane at 25 ꢂC with
[complex] ¼ 4.9 ꢁ 10ꢀ4 M and [CH3I] ¼ 0.0113 M). A third, very slow
reaction (Fig. 4) resulted in the decrease of the peak at 2028 cmꢀ1
[Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ, different product species, in equilib-
rium with the starting complex, were observed on infrared. This
equilibrium between the different products and reactant rendered
the isolation of any of the products not possible. However, due to
the distinct infrared stretching frequency of CO groups [27,28], it
was possible to clearly distinguish between alkyl and acyl products
as described in the next section. In addition, the different products
is further characterised by means of a computational chemistry
study. The possibility of side reactions, as well as the stability of the
complex was monitored at the solvent OCO stretching frequency of
and the simultaneous n
(CO) peak formation at 1714 cmꢀ1 with an
observed rate constant of 1.1 ꢁ 10ꢀ5 sꢀ1 which is attributed to Ir
(III)-acyl formation.
Fig. 8. Possible products of CH3I oxidative addition to [Ir2(
m
-Dcbp)(CO)2(PCy3)2]ꢀ (top) and a simplified model (bottom).

E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
1995
Table 4
DFT molecular energies (eV) of selected stereo isomers of [(Ir(II)-Ir(II))(
the lowest energy isomer as zero.
m
-Dcbp)(CO)2(PX3)2(CH3)(I)]ꢀ and [(Ir(I)-Ir(III))(
m
-Dcbp)(CO)2(PX3)2(CH3)(I)]ꢀ (X ¼ Cy or H) relative to
Ir(II)eIr(II)
no
Geometry
Energy/eV
X ¼ Cy
X ¼ H
-
O
O
CH3
O
N
N
O
Ir
1 Trans
e
a
Ir
X3P
OC
I
PX3
CO
-
O
O
O
O
CO
O
CO
Ir
O
N
N
3 Cis
e
3.79
Ir
X3P
CH3
I
PX3
-
O
I
CO
Ir
O
N
N
O
4 Cis
e
2.84
Ir
X3P
CH3
CO
PX3
-
O
CO
Ir
O
N
N
O
Ir
3 Trans
e
b
X3P
CH3
I
PX3
CO
-
O
O
CO
Ir
O
N
N
O
Ir
4 Trans
e
2.71
X3P
CH3
CO
PX3
I
-
O
O
CO
O
CH3
Ir
O
N
N
1 Cis
e
2.36
Ir
X3P
OC
I
PX3
(continued on next page)

1996
E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
Table 4 (continued )
Ir(II)eIr(II)
no
Geometry
Energy/eV
X ¼ Cy
X ¼ H
-
-
O
O
I
CH3
2 Cis
e
1.95
O
N
N
O
Ir
Ir
X3P
OC
CO
PX3
O
O
CH3
Ir
O
N
N
O
2 Trans
Ir(I)eIr(III)
1
1.05
1.48
Ir
X3P
OC
CO
PX3
I
-
-
-
O
O
CO
Ir
O
N
N
O
0.60
0.46
Ir
X3P
I
CO
PX3
H3C
O
O
I
O
N
N
O
2
0.33
0.39
Ir
Ir
X3P
CH3
CO
PX3
OC
O
O
CH3
Ir
O
N
N
O
3
0.00
0.00
Ir
X3P
OC
CO
PX3
I
a ¼ optimized to 2 cis.
b ¼ optimized to 4 cis.
Bold ¼ Isomer with lowest relative energy.
1. Alkyl 1 convert into Alkyl 2 through an unobserved acyl inter-
mediate. This is possible, but highly unlikely.
2. Alkyl 1 is actually an Ir(II)eIr(II) species with the CH3 bonded to
one iridium metal and I to the other. It is not expected that both
Ir(I) centres get oxidized to Ir(II) simultaneously during the
An interesting aspect of the final IR spectra is the stable n(CO)
peaks at 2028 and 1942 cmꢀ1 for the duration of the reaction
suggesting equilibria during the reaction sequence.
Results obtained from this IR study can be summarized as
follows:

E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
1997
Fig. 9. The OLYP/TZP optimized geometry of [Ir(I)-Ir(III)(m
-Dcbp)(CO)2(PCy3)2(CH3)(I)]ꢀ-alkyl 1. Selected bond lengths (Å) and angles (ꢂ) are shown. The IreNeNeIr dihedral angle is
33.2ꢂ. d(IreIr) ¼ 4.47 Å. H atoms have been omitted for clarity, except for the methyl H.
first, (fast) reaction, since as soon as the first Ir(I) is oxidized to
Ir(II), the second Ir(I) will be more difficult to be oxidized.
3. The two alkyl isomers formed are both in equilibrium with the
starting complex, possibly a cis and trans isomer, and that only
one of these isomers undergo CO insertion or methyl migration
while an equilibrium is maintained with the starting complex.
the oxidative addition reaction, implies that only one product is
formed during this reaction. A comparison of the kinetic results
obtained from the UV/visible and IR indicated that the second
reaction observed in the UV/visible region corresponds to the Ir
(III)-alkyl formation (n
(CO) at 2028 cmꢀ1) with observed rate
constants of 2.6 ꢁ 10ꢀ3 (UV) and 2.5 ꢁ 10ꢀ3 sꢀ1 (IR) respectively
([CH3I] ¼ 0.01 M).
This does not explain the n .
(CO) peak at 1942 cmꢀ1
A third, much slower reaction (not shown in Fig. 5) was also
observed on UV/visible with an observed rate constant of
1.42 ꢁ 10ꢀ5 sꢀ1 (t1/2 ¼ 13.6 h), which is of the same order as the Ir
(III)-acyl formation reaction observed on IR (n(CO) stretching
frequency growth at 1714 cmꢀ1) with observed rate constant of
1.1 ꢁ 10ꢀ5 sꢀ1 ([CH3I] ¼ 0.35 M). The long half-life of this reaction
implied that it could not be followed with great accuracy due to, for
example, evaporation problems of the solvent. The first (fast)
reaction was investigated with a stopped-flow method, but the
results were inconclusive.
The second slower reaction, corresponding to the Ir(III)-alkyl 2
formation (2028 cmꢀ1) reaction on IR, was followed at three different
temperatures on the UV/visible (380 nm) to ascertain the activation
parameters for this reaction. Since the second reaction was much
slower than the first reaction, and much faster that the third reaction,
the kinetic data of the second reaction could be treated in isolation
from the other reactions. From the results obtained under pseudo
first order conditions, shown in Fig. 6, it is clear that the rate law
kobs ¼ k2[CH3I] þ kꢀ2 (see Equation(2)) gives a fair presentationof the
reaction (straight line) and that the substantial intercept is inter-
preted as the reverse reaction i.e. the dissociation of the CH3I from
the Ir centre. Results are tabulated in Table 2.
4. Alkyl 1 is a mixed Ir(III)ealkyleIr(I) species. This gives a logical
explanation for the
n
(CO) peaks at 2004 and 1942 cmꢀ1; the
n
n
(CO) peak at 2004 cmꢀ1 due to the CO bonded to Ir(III) and the
(CO) peak at 1942 cmꢀ1 due to the CO bonded to Ir(I).
3.2.2. UV/visible study
The UV/visible spectra (Fig. 5) of the reaction between (Bu4N)
[Ir2( -Dcbp)(CO)2(PCy3)2] and methyl iodide indicated two clearly
m
3.2.3. Solvent study
distinguishable reactions within the first 30 min. A fairly quick
spectrum change in Fig. 5 (a to b) resulted after the addition of
CH3I to the metal complex. A subsequent slower spectrum change
(b to c) was observed which also resulted in the formation of
isosbestic points at 399 and 414 nm in the UV/visible spectra. The
isosbestic point, which was detected for the second slower part of
A solvent study was also performed on the second reaction, the
alkyl 2 formation reaction, to try and find valuable information on
the transition state of the complex. Five different solvents with
a wide variety of dielectric constants and donocity properties were
selected and the oxidative addition reaction between Bu4N

1998
E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
Table 5
DFT molecular energies (eV) of the selected alkyl stereo isomers of [(Ir(III)-Ir(III))(
m
-Dcbp)(CO)2(PX3)2(CH3)2(I)2]ꢀ (X ¼ Cy or H) relative to the lowest energy isomer as zero.
Geometry
Energy/eV
Geometry
Energy/eV
X ¼ Cy
X ¼ H
X ¼ Cy
X ¼ H
-
-
-
-
-
-
-
-
-
-
O
O
O
O
O
O
O
O
O
CO
I
CO
Ir
I
e
e
e
e
1.08
0.53
0.46
0.45
0.42
0.41
O
N
N
O
O
N
N
O
Ir
Ir
Ir
X3P
I
I
PX3
X3P
H3C
CO
H3C
PX3
CH3
CO
H3C
O
O
O
O
CO
Ir
CH3
CO
Ir
I
0.99
0.96
0.90
0.35
0.47
0.42
O
N
N
O
O
N
N
N
N
O
Ir
Ir
X3P
CH3
H3C
PX3
X3P
I
OC
PX3
I
CH3
I
OC
O
CH3
Ir
CO
Ir
CO
Ir
I
O
N
N
N
N
O
O
N
O
Ir
X3P
I
H3C
PX3
X3P
I
OC
PX3
CH3
I
OC
H3C
O
O
CO
Ir
I
I
CO
Ir
O
N
O
O
N
O
Ir
Ir
X3P
I
H3C
PX3
X3P
CO
H3C
PX3
CO
I
H3C
H3C
O
O
O
O
CO
Ir
CO
Ir
CH3
Ir
CH3
Ir
O
N
O
e
0.89
0.19
0.24
O
N
N
O
X3P
CH3
I
PX3
CH3
X3P
CO
OC
PX3
I
I
I
-
-
-
O
O
O
O
CO
Ir
CH3
Ir
I
I
0.83
0.84
0.00
0.00
O
N
N
O
O
N
N
O
Ir
Ir
X3P
CH3
H3C
PX3
X3P
CO
OC
PX3
CO
CH3
I
I
O
O
CH3
Ir
CO
Ir
e
0.59
O
N
N
O
X3P
I
I
PX3
CH3
OC
Bold ¼ Isomer with lowest relative energy.

E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
1999
[Ir2(
m
-Dcbp)(CO)2(PCy3)2] and CH3I were studied at 25 ꢂC. The
position of the complex were performed in order to identify the
more stable isomers. Calculations on the most stable isomers were
repeated for the full experimental system. Results for selected alkyl
addition products resulting from one CH3I molecule addition are
presented in Table 4. From the relative energies given in Table 4 it is
clear that all Ir(I)eIr(III) alkyl products are more stable than the Ir
(II)eIr(II) products. This finding is in agreement with the IR results
presented above that Alkyl 1 is a mixed Ir(III)ealkyleIr(I) species.
results are presented in Fig. 7 while the calculated rate constants
and activation parameters are given in Table 3.
Results obtained from the solvent variation study (Fig. 7 and
Table 3) show the influence of both the polarity and donocity of the
different solvents on the oxidative addition reaction. The results
clearly show a 25x increase in reaction rate when the reaction is
performed in halogenated solvents (polar solvents) compared to
that of benzene (non-polar) which is substantially smaller than that
The first product of the oxidative addition reaction to [Ir2(m-Dcbp)
observed for the same reaction in the case of [Ir(
m-pz)2(CO)4],
(CO)2(PCy3)2]ꢀ, the Alkyl 1 species, is thus a mixed Ir(III)ealkyleIr
(I) complex. The computational results further indicate that trans
addition to one of the iridium metal centres occurs. The optimized
where an increase of up to a factor of 103 was observed when more
polar solvents were used during the oxidative addition reaction.
These relative moderate increases in oxidative addition reaction
rate are normally an indication that a concerted three centre
transition state was formed during the reaction.
geometry of the [Ir(I)-Ir(III)(
is presented in Fig. 9.
m
-Dcbp)(CO)2(PCy3)2(CH3)(I)]ꢀ-alkyl 1
Addition of another CH3I molecule to the mixed Ir(III)-alkyleIr
(I) (Alkyl 1) lead to a Ir(III)eIr(III)-alkyl product, or the mixed Ir(III)-
alkyleIr(I) could first isomerizes to form an Ir(III)-acyleIr(I) species.
The possibility of the formation of an Ir(III)-acyleIr(I) species is
however discarded, since the energy of the optimized Ir(III)e
acyleIr(I) þ CH3I species is much higher than that of the Ir(III)eIr
(III)-alkyl product, see Fig. 11. In Table 5 selected alkyl addition
3.2.4. Computational study
Oxidative addition of CH3I to [Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ may
lead to different alkyl species. Ignoring the stereochemistry, three
possible alkyl products can be envisioned, see Fig. 8. While only 3
isomeric forms of the Ir(I)eIr(I) reactant complex is possible, 36
isomers of the Ir(II)eIr(II) complex, 48 isomers of the Ir(I)eIr(III)
complex and 576 isomers (enantiomers included) of the Ir(III)eIr
(III) complex are possible! To calculate the optimized geometry of
each of these isomers (the Ir(III)eIr(III) complex has 132 atoms!) by
means of computational chemistry, such as DFT, is thus a senseless
project when taking the computational hours needed into account.
Therefore we first calculate the relative energies of simplified
versions where the PCy3 group is substituted for PH3 in the full
model (Fig. 8 bottom).
Test calculations on both the full as well as the simplified system
indicated that complexes with both the PCy3 (full model) or PH3
(simplified model) groups on the “outside” position of the complex,
as was found for the Ir(I)eIr(I) starting complex (Fig. 2) are ca 1 eV
more stable than isomers having the PCy3 (full model) or PH3
(simplified model) groups in other positions. Calculations on all
isomers with the PH3 (simplified model) groups on the “outside”
products of two CH3I molecules to [Ir2(
as well as their relative energies, are given. The energy of the
most stable [Ir(III)-Ir(III)(
-Dcbp)(CO)2(PCy3)2(CH3)2(I)2]ꢀ isomer,
m
-Dcbp)(CO)2(PCy3)2]ꢀ,
m
presumable the Alkyl 2 product observed on IR, is slightly higher
(0.13 eV) than that of the mixed Ir(III)-alkyleIr(I) (Alkyl 1) þ CH3I
(Fig. 11), in agreement with a fast equilibrium between Ir(I)-Ir(I)
and the mixed Ir(III)-alkyleIr(I) (Alkyl 1) species, leading to the Ir
(III)eIr(III)-alkyl (Alkyl 2) product of oxidative addition. The
computational results further indicate that trans addition of CH3I to
both the iridium metal centres occurs, with the two iodide ions
trans to each other. The optimized geometry of [Ir(III)-Ir(III)(m-
Dcbp)(CO)2(PCy3)2(CH3)2(I)2]ꢀ-alkyl 2 is presented in Fig. 10.
CO insertion in [Ir(III)-Ir(III)(m
-Dcbp)(CO)2(PCy3)2(CH3)2(I)2]ꢀ-
alkyl 2 leads to an acyl product, as observed on IR. To further
identification of the stereochemistry of the acyl product, density
functional theory calculations have also been performed on
Fig. 10. The OLYP/TZP optimized geometry of [Ir(III)-Ir(III)(m
-Dcbp)(CO)2(PCy3)2(CH3)2(I)2]ꢀ-alkyl 2. Selected bond lengths (Å) and angles (ꢂ) are shown. The IreNeNeIr dihedral
angle is 27.9ꢂ. d(IreIr) ¼ 4.53 Å. H atoms have been omitted for clarity, except for the methyl H.

2000
E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
Table 6
DFT molecular energies (eV) of the selected acyl stereo isomers of [(Ir(III)-Ir(III))(m-
Dcbp)(PX3)2(COCH3)2(I)2]ꢀ (X ¼ Cy or H) relative to the lowest energy isomer as
zero.
No
Geometry
Energy/eV
X ¼ Cy
X ¼ H
5 acyl
0.75
0.00
4 acyl
0.70
a
Fig. 11. The calculated reaction pathway for the CH3I oxidative addition to [Ir2(m-Dcbp)
(CO)2(PCy3)2]ꢀ
.
selected Ir(III)eIr(III) acyl isomers (Table 6). Again it was found that
acyl species with the PCy3 groups on the “outside” position of the
complex are more stable than isomers having the PCy3 groups in
other positions by at least 1 eV. The optimized geometry of the
lowest energy [(Ir(III)-Ir(III))(
isomer is presented in Fig. 12.
3 acyl
0.40
0.02
m
-Dcbp)(PX3)2(COCH3)2(I)2]ꢀ-acyl
Both the acyl moieties are above the ligand plane with one of the
iodide ions in and the other below the ligand plane. The rearrange-
ment of the ligands attached to iridium led to a geometry with the
IreNeNeIr dihedral angle reduced to 4.0ꢂ. The geometry around the
two iridium atoms is square pyramidal. The base of the one pyramid
2 acyl
0.34
0.13
is in the
m-Dcbp ligand plane, but the other is perpendicular to it with
CCOCH3 and Om
in the apical positions of the two pyramids
-Dcbp
respectively. Such an arrangement has the least sterical strain in the
-Dcbp backbone and the least sterical interaction between the
groups attached to iridium. Results obtained for the most stable [(Ir
(III)-Ir(III))(
-Dcbp)(PH3)2(COCH3)2(I)2]ꢀ-acyl isomer for the simpli-
m
m
fied system with much less sterical strain, were not in agreement
with the results obtained for the full experimental system.
1 acyl
0.00
b
3.2.5. Proposed mechanistic pathway
The most probable mechanistic pathway for the overall oxida-
tive addition of CH3I to [Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ, taking into
account all the information that was obtained from the IR spectra
and DFT calculations and which are presented in Scheme 1, is used
to interpret the observed rate constant for the reactions. Using the
following observations
a ¼ Optimized to 5 acyl.
b ¼ Optimized to 2 acyl.
Bold ¼ Isomer with lowest relative energy.
a) fast oxidative addition step (k1/k
path, the n(CO) at
ꢀ1
1942 cmꢀ1 and shoulder formation at 2004 cmꢀ1
b) a slower oxidative addition step (k2/k path, (CO) formation
)
The second reaction can be treated in isolation due to the very
fast first reaction and the very slow third reaction. Under pseudo-
first order conditions, [IreIr] < < < [CH3I] and integration, Equation
(2) can be obtained
n
ꢀ2
at 2028 cmꢀ1
c) a very slow carbonyl insertion/methyl migration reaction (k3)
to form the Ir(III)eIr(III) acyl product ( (CO) formation at
1714 cmꢀ1
)
n
Ã
kobs ¼ k2½CH3I þ k
(2)
ꢀ2
)
A graph of kobs vs [CH3I] will yield a straight line with slope ¼ k2 and
intercept kꢀ2 as indicated in Fig. 6 and the results reported in Table 2.
It was decided to concentrate on the k2/kꢀ2 pathway in this study.
The forward rate constant for Ir(III) alkyl 2 formation of 3.25
(4) ꢁ 10ꢀ2 Mꢀ1 sꢀ1 (25.0 ꢂC in dichloromethane) is smaller than the
the rate of disappearance of the [Ir(I)eIr(III)] complex can be
expressed as follows:
Rate ¼ K2½IrðIÞ ꢀ IrðIIIÞꢃ½CH3Iꢃ þ K ½IrðIÞ ꢀ IrðIIIÞꢃ½CH3Iꢃ
ꢀ1
ꢀK1½IrðIÞ ꢀ IrðIÞꢃ½CH3Iꢃ ꢀ K ½IrðIIIÞ ꢀ IrðIIIÞꢃ
(1)
ꢀ2
corresponding reaction for the (Bu4N)[Ir2(m-Dcbp)(cod)2] complex

E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
2001
Fig. 12. The OLYP/TZP optimized geometry of [Ir(III)-Ir(III)(m
-Dcbp)(PCy3)2(COCH3)2(I)2]ꢀ-acyl. Selected bond lengths (Å) and angles (ꢂ) are shown. The IreNeNeIr dihedral angle is
4.0ꢂ. d(IreIr) ¼ 4.18 Å. H atoms have been omitted for clarity, except for the methyl H.
(7.0(1) ꢁ 10ꢀ2 Mꢀ1
s
ꢀ1) [16], while the reverse reactions are
by Stobart et al. [31] that was obtained for the oxidative addition
comparable in magnitude. This decrease in Ir(III)-alkyl 2 formation
rate for the title complex are an indication of deactivation by the
PCy3/CO ligand combination towards the Ir(I) centre compared to
reactions between CH3I and [Ir(m
-pz)2(CO)4] (32(2) kJ molꢀ1 and
-152(8) JKꢀ1 molꢀ1 respectively). The negative activation entropy
obtained for the reductive elimination step (kꢀ2 path) is contrary to
what is expected due to the fact that both the IreCH3 and IreI bonds
need to break to from the CH3I molecule. A possible explanation for
this negative activation entropy is the probable coordination of
solvent molecules during the transition state formation.
the cod ligands in the (Bu4N)[Ir2(m-Dcbp)(cod)2] complex. The
positive activation enthalpy for the k2 path (see Table 2) of 28
(7) kJ molꢀ1 is indicative of an associative mechanism while the
negative activation entropy(-178(23) JKꢀ1 molꢀ1) corresponds to
CH3eI bond dissociation and which is associated with a polar tran-
sition state. These results corresponds favourably with that obtained
The rate of acyl formation (1.42 ꢁ 10ꢀ5 sꢀ1) is clearly indicative
of a complex which is deactivated towards CO insertion or methyl
Scheme 1. Reaction pathway for the CH3I oxidative addition to [Ir2(
m
-Dcbp)(CO)2(PCy3)2]ꢀ as obtained by an IR and computational study.

2002
E. Grobbelaar et al. / Journal of Organometallic Chemistry 696 (2011) 1990e2002
[11] D.O.K. Fjeldsted, S.R. Stobart, M.J. Zaworotko, J. Am. Chem. Soc. 107 (1985)
8258e8259.
[12] J.A. Bailey, S.L. Grundy, S.R. Stobart, Organometallics 9 (1990) 536e337.
[13] T.G. Schenk, C.R.C. Milne, J.F. Sawyer, B. Bosnich, Inorg. Chem. 24 (1985)
2338e2344.
[14] Z. Freixa, P.C.J. Kamer, M. Lutz, A.L. Spek, P.W.N.M. van Leeuwen, Angew.
Chem. Int. Ed. 44 (2005) 4385e4388.
[15] C.M. Thomas, R. Mafua, B. Therrien, E. Rusanov, H. Stoeckli-Evans, G. Süss-Fink,
Chem. Eur. J. 8 (2002) 3343e3352.
[16] E. Grobbelaar, W. Purcell, S.S. Basson, Inorg. Chim. Acta 359 (2006)
3800e3806.
[17] J.C. Bayon, G. Net, P. Esteban, P.G. Rasmussen, D.F. Bergstrom, Inorg. Chem. 30
(1991) 4771e4777.
migration if compared to reaction rates of complexes such as
[Rh(acac)(CO)(PPh3)] (1.6
ꢁ
10ꢀ3
s
ꢀ1), [Rh(hpt)(CO)(PPh3)]
(1.0 ꢁ 10ꢀ2
s
ꢀ1) [32] (hpt ¼ thiohydroxamate anion) [18] and
[Rh(tta)(CO)(PPh3)] (1.8
ꢁ
10ꢀ3 sꢀ1
)
(tta
¼
2-thenoyltri-
fluoroacetonato) [29]. The relative large K2 values (Table 3) points
to a situation where the equilibrium lies to the right and that
kobs ¼ k2.
4. Conclusions
[18] D.D. Perrin, W.L.F. Armarego, Purification of Laboratory Chemicals, third ed.
Pergamon Press, Oxford, 1988.
[19] Scientist, Version 2, MicroMath Scientific Software, Saint Louis, Mo, 2004.
[20] The ADF program system was obtained from Scientific Computing and
Combining the results obtained from an IR study for the CH3I
oxidative addition to [Ir2(m
-Dcbp)(CO)2(PCy3)2]ꢀ with the results
obtained from a computational chemistry study, led to the reaction
scheme which is proposed in Scheme 1.
Modeling, Amsterdam (http://www.scm.com/). For
a description of the
methods used in ADF, see: G.T. Velde, F.M. Bickelhaupt, E.J. Baerends,
C.F. Guerra, S.J.A. van Gisbergen, J.G. Snijders, T.J. Ziegler, J. Comput. Chem. 22
(2001) 931e967.
Acknowledgements
[21] N.C. Handy, A. Cohen, Mol. Phys. 99 (2001) 403e412.
[22] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B. 37 (1988) 785e789.
[23] E. van Lenthe, A.E. Ehlers, E.J. Baerends, J. Chem. Phys. 110 (1999) 8943e8953
and references therein.
[24] J.G. Leipoldt, S.S. Basson, E.C. Grobler, A. Roodt, Inorg. Chim. Acta 99 (1985) 13e17.
[25] J.-P. Xiao, Q.-X. Zhou, J.-H. Tu, Acta Crystallogr. Sect. E Struct. Rep. Online 63
(2007) o2785.
The authors thank the Research Fund of this Universityof the Free
State, the Research Council of Norway (JC) and the National Research
Foundation of the Republic of South Africa (JC) for financial support.
[26] P. King, R. Clerac, C.E. Anson, A.K. Powell, Dalton Trans. (2004) 852e861;
J.C. Bayon, P. Esteban, G. Net, P.G. Rasmussen, K.N. Baker, C.W. Hahn,
M.M. Gumz, Inorg. Chem. 30 (1991) 2572e2574;
Appendix A. Supplementary material
Optimized Cartesian coordinates for selected systems can be
foundintheonline version, at doi:10.1016/j.jorganchem.2010.10.049.
X.-H. Li, Acta Crystallogr. Sect. E Struct. Rep. Online 61 (2005) m2405em2407;
Q.-Q. Dou, Y.-K. He, L.-T. Zhang, Z.-B. Han, Acta Crystallogr. Sect. E Struct. Rep.
Online 63 (2007) m2908e2909;
J.-I. Tian, S.-P. Yan, D.-Z. Liao, Z.-H. Jiang, P. Cheng, Inorg. Chem. Commun. 6
(2003) 1025e1033;
J.-F. Zhang, X.-H. Li, Acta Crystallogr. Sect. E Struct. Rep. Online 62 (2006)
m125em127.
References
[1] J.H. Jones, Platinum Met. Rev. 44 (2000) 94e105.
[2] R.S. Dickson, Homogeneous Catalysis with Compounds of Rhodium and
Iridium. D. Reidel Publishing Co., Dordrecht, 1985.
[27] L.J. Damoense, W. Purcell, A. Roodt, Y.S. Varshavsky, Rhodium Express 14
(1995) 4e9.
[28] L.J. Damoense, W. Purcell, A. Roodt, J.G. Leipoldt, RhodiumExpress 5 (1994) 10e13.
[29] F.N. Stuurman, J. Conradie, J. Organomet. Chem. 694 (2009) 259e268;
J. Conradie, J.C. Swarts, Organometallics 28 (2009) 1018e1026;
M. Conradie, J. Conradie, Inorg. Chim. Acta 361 (2008) 208e218;
M. Conradie, J. Conradie, Inorg. Chim. Acta 361 (2008) 2285e2295;
J. Conradie, G.J. Lamprecht, S. Otto, J.C. Swarts, Inorg. Chim. Acta 328 (2002)
191e203.
[30] L. Vaska, J.W. Di Luzio, J. Am. Chem. Soc. 33 (1961) 2784e2785;
L. Vaska, D.L. Catone, J. Am. Chem. Soc. 88 (1966) 5324e5325;
J.G. Cordaro, R.G. Bergman, J. Am. Chem. Soc. 126 (51) (2004) 16912e16929;
A. Haynes, A.J.H.M. Meijer, J.R. Lyons, H. Adams, Inorg. Chem. 48 (1) (2009)
28e35;
[3] A. Haynes, P.M. Maitlis, G.E. Morris, G.J. Sunley, H. Adams, P.W. Badger,
C.M. Bowers, B. Cook, P.I.P. Elliot, T. Ghaffar, H. Green, T.R. Griffin, M. Payne,
J.-M. Pearson, M.J. Taylor, P.W. Vickers, R.J. Watt, J. Am. Chem. Soc. 126 (2004)
2847e2861.
[4] M.P. García, M.A. Esteruelas, M. Martin, L.A. Oro, J. Organomet. Chem. 467
(1994) 151e159.
[5] M.A. Esteruelas, M.P. García, A.M. López, L.A. Oro, Organometallics 10 (1991)
127e133.
[6] M.P. Garcia, A.M. López, M.A. Esteruelas, F.J. Lahoz, L.A. Oro, J. Chem. Soc.
Chem. Commun. (1988) 793e795.
[7] M.A. Esteruelas, M.P. Garcia, A.M. López, L.A. Oro, Organometallics 11 (1992)
702e705.
[8] J.L. Atwood, K.A. Beveridge, G.W. Bushnell, K.R. Dixon, D.T. Eadie, S.R. Stobart,
M.J. Zaworotko, Inorg. Chem. 23 (1984) 4050e4057.
[9] D.G. McCollum, B. Bosnich, Inorg. Chim. Acta 270 (1) (1998) 13e19.
[10] K.A. Beveridge, G.W. Bushnell, S.R. Stobart,Organometallics2(1983)1447e1451.
E. Grobbelaar, S. Lötter, D. Visser, J. Conradie, W. Purcell, Inorg. Chim. Acta 362
(2009) 3949e3954.
[31] R.D. Brost, S.R. Stobart, J. Chem. Soc. Chem. Commun. (1989) 498e500.
[32] G.J.J. Steyn, A. Roodt, J.G. Leipoldt, Rhodium Express 1 (1993) 24e29.