Characterization and oxidative addition reactions of rhodium(I) carbonyl cupferrate diphenyl-2-pyridylphoshine complexes

Article abstract of DOI:10.1016/j.jorganchem.2013.08.037

The oxidative addition of CH₃I to the [Rh(cupf)(CO)(DPP)] complex (DPP = diphenyl-2-pyridylphoshine and cupf = N-nitroso-N- phenylhydroxylaminen) was kinetically investigated using UV/vis and infrared spectroscopy. The kinetics followed in chloroform, acetonitrile and acetone as solvents, indicated three different consecutive reactions. Firstly, a very fast reaction (intermediate formation, IM), secondly, a slower reaction with the formation of the Rh(III) alkyl complex (5.0(1) × 10^-3 M ^-1 s^-1 for acetonitrile at 20 C) and thirdly a very slow formation of the Rh(III) acyl complex as final product with a rate constant of 2.9(6) × 10^-4 s^-1. The same oxidative addition reaction in ethyl acetate as solvent exhibited only two reactions. Firstly the Rh(III) alkyl formation (1.03(3) × 10^-3 M^-1 s ^-1), which was five times slower than in the other solvents. Secondly, the Rh(III) acyl formation, which was masked by solvent IR stretching frequencies in the detection area. Rh(III) acyl, however, was isolated from ethyl acetate. There was no indication of intermediate formation with ethyl acetate. This apparent discrepancy between the rate and the mechanism for the same reaction prompted a DFT study to gain more insight into the reactants and products of the reaction, as well as to try and determine the geometry of the transition state. The DFT study predicted the formation of a linear transition state (TS), followed by the formation of the cationic five-coordinate [Rh(cupf)(CO)(DPP)(CH₃)]⁺ intermediate with the CH ₃ group in the apical position and with the iodide ion drifting away into the solvent sphere. This was in agreement with the experimental very fast first reaction. The experimentally observed difference in the rates and mechanisms of the [Rh(cupf)(CO)(DPP)] + CH₃I reaction in ethyl acetate relative to the other solvents can be attributed to the rate of the formation and/or the build-up and conversion of the I.M.

Full text of DOI:10.1016/j.jorganchem.2013.08.037

Journal of Organometallic Chemistry 745-746 (2013) 439e453
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Characterization and oxidative addition reactions of rhodium(I)
carbonyl cupferrate diphenyl-2-pyridylphoshine complexes
*
*
Walter Purcell , Jeanet Conradie , Trevor T. Chiweshe, Johan A. Venter, Linette Twigge,
Michael P. Coetzee
Department of Chemistry, University of the Free State, 205 Nelson Mandela Drive, Bloemfontein 9300, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 29 July 2013
Accepted 20 August 2013
The oxidative addition of CH3I to the [Rh(cupf)(CO)(DPP)] complex (DPP ¼ diphenyl-2-pyridylphoshine
and cupf ¼ N-nitroso-N-phenylhydroxylaminen) was kinetically investigated using UV/vis and infrared
spectroscopy. The kinetics followed in chloroform, acetonitrile and acetone as solvents, indicated three
different consecutive reactions. Firstly, a very fast reaction (intermediate formation, IM), secondly, a
slower reaction with the formation of the Rh(III) alkyl complex (5.0(1) ꢀ 10ꢁ3 Mꢁ1 sꢁ1 for acetonitrile at
20 ꢂC) and thirdly a very slow formation of the Rh(III) acyl complex as final product with a rate constant
Keywords:
Rhodium(I)
Diphenyl-2-pyridylphoshine
Carbonyl
Oxidative addition
Ionic intermediate
of 2.9(6) ꢀ 10ꢁ4
s
ꢁ1. The same oxidative addition reaction in ethyl acetate as solvent exhibited only two
reactions. Firstly the Rh(III) alkyl formation (1.03(3) ꢀ 10ꢁ3 Mꢁ1 sꢁ1), which was five times slower than in
the other solvents. Secondly, the Rh(III) acyl formation, which was masked by solvent IR stretching
frequencies in the detection area. Rh(III) acyl, however, was isolated from ethyl acetate. There was no
indication of intermediate formation with ethyl acetate. This apparent discrepancy between the rate and
the mechanism for the same reaction prompted a DFT study to gain more insight into the reactants and
products of the reaction, as well as to try and determine the geometry of the transition state. The DFT
study predicted the formation of a linear transition state (TS), followed by the formation of the cationic
five-coordinate [Rh(cupf)(CO)(DPP)(CH3)]þ intermediate with the CH3 group in the apical position and
with the iodide ion drifting away into the solvent sphere. This was in agreement with the experimental
very fast first reaction. The experimentally observed difference in the rates and mechanisms of the
[Rh(cupf)(CO)(DPP)] þ CH3I reaction in ethyl acetate relative to the other solvents can be attributed to
the rate of the formation and/or the build-up and conversion of the I.M.
Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction
catalytical processes encompass four major market sectors, namely
fuel refining (22%), polymerization (21%), chemical production
Rhodium catalysts exhibit scope and versatility that are prob-
ably unmatched by any other metal in industry today. The impor-
tance of catalysis to the chemical industry (V12 billion worth) is
illustrated by a 2006-estimate that more than 95% of chemical
products by volume are based on catalytic technologies [1]. More
than 70% of all consumer products are based on catalytic processes
and 20% of the world economy is directly or indirectly related to
catalysis. Economic and environmental considerations are mainly
the compelling reasons for the widespread use of catalysis. These
processes normally introduce lower operating costs, higher purity
and yield products with fewer side-products [2]. Industrial
(27%), and environmental remediation (30%).
The use of different rhodium(II) complexes as chiral catalysts in
the insertion of carbenes and nitrenes in drug synthesis results in
fewer reaction steps and exceptional enantioselectivity [3]. An
example is the synthesis of Ritalin which normally requires eight to
nine synthetic steps which are reduced to two steps using the
rhodium catalyst. Another article reports the use of a rhodium
catalyst capable of producing hydrogen from a renewable fuel
source in a commercially viable manner [4]. CORDIS [5] reports the
success of a rhodium catalyst to produce 50 times more hydrogen
from a feed gas containing 0.5% phenol over a broad temperature
range when compared to its commercial nickel counterpart.
Rhodium (Monsanto) [6,7] and iridium (Cativa) [8] are the cat-
alysts of choice in the production of acetic acid from methanol and
accounts for over 60% of the worldwide production. Research has
shown that the actual catalytic cycles of these two processes consist
* Corresponding authors. Fax: þ27 (0) 865255382.
E-mail addresses: purcellw@ufs.ac.za (W. Purcell), conradj@ufs.ac.za
(J. Conradie).
0022-328X/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jorganchem.2013.08.037

440
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
of six different steps, including oxidative addition, 1,1-insertion or
CO insertion, CO association and reductive elimination. Important
prerequisites for these metal complexes to undergo oxidative
addition include an unsaturated coordination sphere, the ability to
undergo a two electron oxidation process (from þ1 to þ3) as well as
non-bonding electron density on the metal center. The key to
fundamental research on these kinds of catalytic cycles is the
greater understanding of the relationships between activity and
catalyst structure, as well as ways to better predict, understand and
control catalyst molecular architecture [9].
Another group of rhodium(I) and iridium(I) complexes, namely
[M(LL0)(CO)(LX3)] (where M ¼ Rh(I) or Ir(I), LL0 ¼ mono ionic
bidentate ligands and LX3 ¼ different phosphines, phosphites, ar-
sines and stibines) were identified as possible candidates for
catalysis or model complexes to investigate all the possible pa-
rameters that influence oxidative addition reactions due to their
adherence to the above-mentioned criteria for oxidative addition
reactions. Numerous structural and kinetic studies [10aep] were
undertaken to investigate factors that influence the nucleophilicity
of the metal centers, solvent interactions and steric bulk on the rate
of oxidative addition, CO insertion and mechanistic elucidation.
Additionally, diphenyl-2-pyridylphoshine (DPP) has the ability to
not only coordinate to the metal center with the phosphorous atom
[11aed], but in some cases also with the pyridinium nitrogen,
making the phosphine ligand a potential bidentate ligand [12aec].
The increase in electron density due to the presence of a nitrogen
atom in the phosphine also has the potential to alter the Lewis
basicity of the phosphine and ultimately influence the oxidative
addition reactions for the metal complexes. With this in mind the
Double distilled water was used for dilution and Schott Duran grade
(A) type glassware was used during the metal analysis.
All the reactions were followed under pseudo-first-order con-
ditions with the typical complex concentration of 2.5 ꢀ 10ꢁ4 M for
the UV/visible measurements and 0.02 M for IR. The methyl iodide
concentrations varied between 0.1 and 1.0 M. The observed first-
order rate constants where calculated using a non-linear least-
square program according to At ¼ AN þ (A0 e AN)eꢁkobs.t with
At, AN and A0 the absorbance of the indicated species at time t, N
and 0 respectively [14]. Experimental values are presented by
points and lines represent calculated values in the subsequent
figures.
2.2. Synthesis
2.2.1. [Rh(cupf)(CO)2] [15]
RhCl3$3H2O (0.5 g, 1.9 mmol) was dissolved in a few drops of
water, added to 10 ml N,N-dimethylformamide (DMF), and refluxed
to a yellow-orange color. The solution was cooled to room tem-
perature before addition of N-nitroso-N-phenylhydroxylamine
ammonium salt (cupf, 0.3 g, 1.9 mmol). Subsequently 100 ml cold
water was added to the solution resulting in the suspension of a
yellow product, which was removed by centrifuge and dried in a
fume cupboard. The product was purified by dissolving it in 20 ml
acetone, filtering using a micro-filtration technique, and drying at
room temperature to obtain the final yellow powdered product.
Yield: 70%. IR data: n(CO) ¼ 2087, 2013 cmꢁ1. Elemental analysis of
RhO4N2C8H5: (calculated values in brackets): C, 32.81 (32.45), H,
1.98 (1.71), N, 9.09 (9.46), Rh, 34.69 (34.76) %. 1H NMR (300 MHz,
[Rh(cupf)(CO)(DPP)] complex (Hcupf
¼
N-nitroso-N-phenyl-
CDCl3, 20 ꢂC):
d 7.89 (m, 3- & 5-H, cupf), 7.46e7.41 (m, 4H, 2- & 6H,
1
hydroxylamine) was prepared and the mode of bonding as well the
phosphine’s effect on the oxidative addition of methyl iodide to
these complexes was investigated.
cupf). 13C{1H} NMR (151 MHz, CDCl3, 20 ꢂC):
d 183.16 (d, JRh-
¼ 74.9 Hz, CO), 183.11 (d, 1JRh-C ¼ 72.8 Hz, CO), 137.28 (s, 1-C, cupf),
C
129.90 (s, 4-C, cupf), 128.41 (s, 2- & 6-C, cupf), 119.63 (s, 3- & 5-C,
cupf).
2. Experimental
2.2.2. [Rh(cupf)(CO)(DPP)]
[Rh(cupf)(CO)2] (0.2 g, 0.64 mmol) was dissolved in 10 ml
methanol and the solution was slightly heated to 30 ꢂC for 5 min to
ensure homogeneity. Diphenyl-2-pyridylphosphine (0.2 g,
0.76 mmol) was added gently whilst stirring. The solution changed
from yellow to red, and a yellow product precipitated immediately
from the solution. The precipitate was removed by filtration,
washed with methanol and dried in a fume cupboard. Yield: 60%. IR
2.1. General considerations
All chemicals were of reagent grade and were used without
further purification. The RhCl3.3H2O and diphenyl-2-pyridyl
phosphine were purchased from SigmaeAldrich. Solvents were
purified and dried according to standard procedures prior to use. IR
spectra were recorded with a Digilab FTS 2000 spectrometer while
NMR data were obtained at 293 K with Bruker 300 and 600 MHz
spectrometers. For characterization of the Rh(III) alkyl and acyl
species, [Rh(cupf)(CO)(DPP)] (5 mg, 0.0101 mmol) was dissolved in
0.7 ml CDCl3 and transferred to a 5 mm diameter NMR tube. Methyl
data: n(CO) ¼ 1988 cmꢁ1. Elemental analysis of RhPN3O3C24H19
:
(calculated values in brackets): C, 54.0 (54.25), H, 3.73 (3.61), N,
8.20 (7.91), Rh, 19.16 (19.37) %. Two isomers were observed in the
31P NMR spectra, with the minor isomer (isomer A) being 37.3% and
the major isomer (isomer B) 62.7%, according to the 31P resonance
iodide (10.0
m
l) was added and a series of 1H and 31P NMR spectra
integrals. 1H NMR (300 MHz, CDCl3, 20 ꢂC):
d 8.75 (m, 3-H pyridyl-
were recorded for 30 h. A Varian Cary 50 spectrophotometer,
equipped with a temperature controlled cell changer (accuracy ꢃ
0.1 ꢂC) was use for UV/visible measurements. A cell with NaCl
windows was used to follow the IR kinetics. Elemental analyses
were performed on a LECO Truspec Micro analyzer.
A Shimadzu ICPS-7510 ICP-OES with a radial-sequential plasma
spectrometer was used for the wet chemical analysis of all the
rhodium samples in the current study [13] using cobalt
(228.616 nm) as internal standard. The vertically oriented ICP-OES
with the ‘radial viewing’ plasma was found to be suitable due to its
better detection limits compared to the axial viewing plasma. The
wavelength at 343.489 nm was the most suitable since it was free
from the spectral interference of the elements present in the
sample. The rhodium standard (1000 ppm) was purchased from
Aldrich Chemicals while analytical grade HCl (32%), HNO3 (65%), as
well as Co(NO3)2.6H2O were obtained from Merck Chemicals.
ring, isomer A & B), 8.00 (m, 6-H pyridyl-ring, isomer A & B), 7.93
(m, 3- & 5-H cupf, isomer B), 7.75 (m, 2- & 6-H phenyl, isomer A &
B), 7.68 (m, 5-H pyridyl-ring, isomer A & B), 7.47 (m, 3- & 5H cupf,
isomer A), 7.44e7.20 (m, 4-H cupf, 4-H phenyl, 3- & 5-H phenyl, 4-H
pyridyl-ring, 2- & 6-H cupf, isomer A & B). 13C{1H} NMR (151 MHz,
1
CDCl3, 20 ꢂC) (Only major isomer B reported):
d
189.52 (dd, JRh-
2
1
¼ 75.9 Hz, JP-C ¼ 25.3 Hz, CO), 156.20 (d, JP-C ¼ 74.4 Hz, 1-C
C
3
pyridyl-ring), 149.22 (d, JP-C ¼ 14.7 Hz, 3-C pyridyl-ring) 138.06
(s, 1-C, cupf), 134.87 (d, 3JP-C ¼ 9.5 Hz, 5-C pyridyl-ring), 133.55 (d,
2JP-C ¼ 11.7 Hz, 2- & 6-C phenyl), 131.11 (d, JP-C ¼ 53.1 Hz, 1-C
1
2
phenyl), 129.95 (d, JP-C ¼ 28.8 Hz, 6-C pyridyl-ring), 129.63 (d,
4JP-C ¼ 2.0 Hz, 4-C phenyl), 128.09 (s, 4-C, cupf), 127.85 (s, 2- & 6-C,
3
4
cupf), 127.17 (d, JP-C ¼ 10.6 Hz, 3- & 5-C phenyl), 123.10 (d, JP-
¼ 2.1 Hz, 4-C pyridyl-ring), 118.75 (s, 3- & 5-C, cupf). 31P{1H} NMR
C
1
(121 MHz, CDCl3, 20 ꢂC):
d
49.53 (d, JRh-P ¼ 176.2 Hz, isomer A),
49.51 (d, 1JRh-P ¼ 171.1 Hz, isomer B).

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
441
Fig. 1. IR spectra for the reaction of CH3I (0.2 M) with [Rh(cupf)(CO)(DPP)] (2.0 ꢀ 10ꢁ2 M) in acetonitrile (2 min intervals, 25.0 ꢂC).
2.2.3. [Rh(cupf)(CO)(DPP)(CH3)I]
formalism and full geometry optimization with tight convergence
criteria, was used for minimum energy and transition state (TS)
searches. Approximate structures of the TS have been determined
with linear transit (LT) scans, with a constrained optimization along
a chosen reaction coordinate, to sketch an approximate path over
the TS between reactants and products. Numerical frequency ana-
lyses [20,21] where the frequencies are computed numerically by
differentiation of energy gradients in slightly displaced geometries,
have been performed to verify the TS geometries. A TS has one
imaginary frequency. Finally, the TS was allowed to relax after
displacing the atoms according to the reaction coordinate as
determined by the eigenvectors. This gave the reactants (relax from
minimum stretch of frequency gave rhodium(I) complex and CH3I)
and the products (relax from maximum stretch of frequency gave
cationic five coordinated complex and Iꢁ) respectively. Throughout,
all calculations have been performed with no symmetry constraint
(C1) and all structures have been calculated as singlet states.
Zero point energy and thermal corrections (vibrational, rota-
tional and translational) were made in the calculations of the
thermal parameters. The enthalpy (H) and Gibbs energy (G) were
calculated from Ref. [22]
Methyl iodide (0.64 g, 4.55 mmol) was added to 5 ml chloroform
solution containing [Rh(cupf)(CO)(DPP)] (0.25 g, 0.46 mmol). The
reaction was allowed to proceed for 20 min at room temperature
during which the color changed from orange to red brown. The
solution was removed by evaporation to yield the dark brown
powdered final product. Yield: 35%. IR data: n(CO) ¼ 2061 cmꢁ1
.
Elemental analysis of RhIPN3O3C25H22
: (calculated values in
brackets): C, 45.10 (44.60), H, 3.56 (3.30), N, 6.42 (6.24), Rh, 15.51
(15.28) %. Four isomers were observed. A 1He31P HMBC experiment
was used to assign the relevant 1H and 31P resonances. 1H NMR
(300 MHz, CDCl3, 20 ꢂC):
d 9.02e7.27 (cupf, pyridyl-ring, phenyl, 4
isomers), 2.06 (dd, 2JRh-H ¼ 2.0 Hz, 3JP-H ¼ 3.7 Hz, RheCH3, alkyl 2 e
isomer A), 1.89 (dd, 2JRh-H ¼ 2.0 Hz, 3JP-H ¼ 3.7 Hz, RheCH3, alkyl 2 e
isomer B), 1.32 (dd, 2JRh-H ¼ 2.0 Hz, 3JP-H ¼ 2.0 Hz, RheCH3, alkyl 1 e
isomer B), 1.31 (dd, 2JRh-H ¼ 2.0 Hz, 3JP-H ¼ 2.0 Hz, RheCH3, alkyl 1 e
isomer A). 31P{1H} NMR (121 MHz, CDCl3, 20 ꢂC):
d
41.67 (d, 1JRh-
¼ 122.3 Hz, alkyl 1 e isomer B), 41.17 (d, 1JRh-P ¼ 126.5 Hz, alkyl 1 e
P
isomer A), 32.88 (d, 1JRh-P ¼ 120.5 Hz, alkyl 2 e isomer B),
d 30.04 (d,
1JRh-P ¼ 123.5 Hz, alkyl 2 e isomer A).
2.2.4. [Rh(cupf)(COCH3)(DPP)I]
Methyl iodide (0.76 g, 4.55 mmol) was added to 5 ml ethyl ac-
etate solution containing [Rh(cupf)(CO)(DPP)] (0.19 g, 0.44 mmol).
The reaction was allowed to proceed for 500 min at room tem-
perature during which the color changed from orange to red brown.
The solution was removed by evaporation to yield a brown
U ¼ ETBE þ EZPE þ EIE
H ¼ U þ RTðgas phaseÞ or H ¼ UðsolutionÞ
G ¼ H ꢁ TS
powdered final product. Yield: 20%. IR data: n(COCH3) ¼ 1688 cmꢁ1
.
Elemental analysis of RhIPN3O3C25H22: (calculated values in
brackets): C, 45.65 (44.60), H, 3.37 (3.30), N, 6.38 (6.24), Rh, 16.35
(15,28) %. Two isomers were observed. 1H NMR (300 MHz, CDCl3,
where U is the total energy, ETBE is the total bonding energy, EZPE is
the zero point energy, EIE is the internal energy (sum of vibrational,
rotational and translational energies), R is the gas constant, T is the
temperature and S is the entropy. The entropy (S) was calculated
from the temperature dependent partition function in ADF at
298.15 K. Solvent effects were taken into account for all calculations
reported here. The COSMO (Conductor like Screening Model) model
of solvation [23e25] was used as implemented [26] in ADF. The
COSMO model is a dielectric model in which the solute molecule is
embedded in a molecule-shaped cavity surrounded by a dielectric
20 ꢂC):
d
9.02e7.27 (cupf, pyridyl-ring, phenyl, 2 isomers), 3.05
(broad singlet, RheCOCH3, acyl a), 2.93 (broad singlet, RheCOCH3,
acyl b). 31P{1H} NMR (121 MHz, CDCl3, 20 ꢂC):
d 32.75 (broad
doublet, 1JRh-P ¼ 151.2 Hz, acyl b). Acyl a was not observed in 31P
NMR.
3. Computational chemistry
Density functional theory (DFT) and calculations were carried
out using the ADF (Amsterdam Density Functional) 2012 pro-
gramme [16e18] with the GGA (Generalized Gradient Approxi-
medium with a given dielectric constant (
used is Esurf [27] and The solvents used in the Conductor like
Screening Model (COSMO) of solvation was methanol (
¼ 32.6),
3
0). The type of cavity
3
0
mation) functional PW91 [19]. The TZP (Triple
z
polarized) basis set
chloroform (
3
0
¼ 4.8), ethyl acetate (
3
0
¼ 6.02), acetonitrile
with a fine mesh for numerical integration, a spin-restricted
(
3
0
¼ 37.5) and acetone (
3
0
¼ 20.7).

442
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
diketonato)(CO)(PR3)] (PR3 ¼ tertiary phosphine of phosphite and if
-diketonato ¼ unsymmetrical ligand) complexes exist in equilib-
b
rium and that the population ratio depends on the solvent and
temperature [31e36]. Isomers isolated for structurally similar
complexes, are controlled by the ligands bonded to the metal center
as well as crystallization conditions [37]. Thus, isomer A crystallized
in the solid state for [Rh(cupf)(CO)(PPh3)], while isomer B was
isolated for the [Rh(cupf)(CO)(P(OCH2)3CCH3)] complex. The novel
isolation of both isomers in same crystal structure was reported for
[Rh(BA)(CO)(PPh3)] (BA ¼ benzoylacetylacetonato anion) [38] and
[Rh(PhCOCHCO(CH2)CH3)(CO)(PPh3)] [35]. For complex [Rh(PhCOC
HCO(CH2)3CH3)(CO)(PPh3)] [39], both geometrical isomers crys-
tallized in the same space in the unit cell at an 89.7:10.3 ratio. The
two structural isomers of [Rh(cupf)(CO)(DPP)] in the current study
are also experimentally observed in CDCl3 with NMR spectroscopy
in a 37.3:62.7% ratio, in excellent agreement with the calculated
population of 37.0:63.0% in chloroform. The calculated geometry of
isomer A and B for [Rh(cupf)(CO)(DPP)] is square planar, with bond
lengths and angles similar to that found for the crystal structure of
Fig. 2. The two possible [Rh(cupf)(CO)(DPP)] isomers.
4. Results
4.1. Synthesis and characterization of complexes
4.1.1. [Rh(cupf)(CO)2] and [Rh(cupf)(CO)(DPP)]
A comparison between the IR spectra of the rhodium dicarbonyl
complex and that of the substituted carbonyl complex shows the
disappearance of one of the carbonyl stretching frequencies (2087,
2013 cmꢁ1). Thus, the displacement of one of the carbonyl ligands
by the phosphine ligand with a shift to a lower wavelength of
1988 cmꢁ1 for the mono substituted monocarbonyl complex is
confirmed. This is in accordance with other complexes indicating
an increase in electron density on the metal center, which can be
attributed to the electronegativity of the phosphine and the
absence of competition by the other carbonyl ligand. The decrease
the related [Rh(b-diketonato)(CO)(PR3)] complexes (Fig. 3).
Six coordinated Rh(III) alkyl complexes: The oxidative addi-
tion of iodomethane to [Rh(cupf)(CO)(DPP)] leads to a rhodium(III)
alkyl product. The 12 possible rhodium(III) oxidative addition
product isomers of the formula [Rh(cupf)(CO)(DPP)(CH3)(I)], and
calculated relative electronic energies of the optimized geometries
(in acetone solvent) are displayed in Fig. 4. Two isomers result from
trans addition of iodomethane and ten isomers result from cis
addition or from the subsequent re-arrangement of a trans addition
product. The energies indicate that four Rh(III)-alkyl isomers are
significantly more stable than the others. These are the two trans
alkyl product isomers (alkyl 1a and 2a) and the two alkyl isomers
with DPP and I above and below the square plane through the two
cupf-oxygens, CO and CH3 (alkyl 2a and 2b). Alkyl 2a and 2b can be
the products of cis addition or from re-arrangement of a trans
addition product. The DFT results highlight that the barrier to cis
addition is too high to be considered and that trans addition of
iodomethane to [Rh(cupf)(CO)(DPP)] is energetically favored. The
in n(CO) indicates an increase in the metal-carbon p-back-bonding
and a decrease in carboneoxygen triple bond character.
4.1.2. [Rh(cupf)(CO)(DPP)(CH3)I] and [Rh(cupf)(DPP)(COCH3)I]
This dynamic process is also illustrated by the solvent IR in Fig. 1
with the disappearance of the monocarbonyl peak of Rh(I) at
1980 cmꢁ1 (solid IR ¼ 1988 cmꢁ1) with time and the formation of
the Rh(III) alkyl peak at 2065 cmꢁ1 (solid IR ¼ 2061 cmꢁ1) and the
Rh(III) acyl peak at 1688 cmꢁ1 (solid IR ¼ 1688 cmꢁ1). Isolation of
the acyl complex is achieved by using ethyl acetate as solvent. The
subsequent kinetic study indicates that the acyl formation reaction
is much slower in this solvent compared to its formation in acetone,
chloroform and acetonitrile and the alkyl was isolated with the
quenching of the reaction just as the acyl began to form.
A comparison of the carbonyl stretching frequencies obtained in
this study with those from other cupf complexes reveals interesting
results. The carbonyl stretching frequency of [Rh(cupf)(CO)(DPP)]
(this study and [28]) correlates well with that of the triphenyl-
phoshine complex (1988 vs 1982 cmꢁ1) [10,29,30], but carbonyl
stretching frequency of the oxidative addition product in the cur-
rent study and [28] shifts slightly more (2061 vs 2052 cmꢁ1) [10],
DFT calculations on the structurally similar rhodium-
complexes, [Rh( -diketonato)(CO)(PPh3)] with
-diketone ¼ 4,4,4-
trifluoroꢁ1-(2-thenoyl)ꢁ1,3-propanedione,
1-phenyl-3-(2-then
b-diketonato
b
b
oyl)ꢁ1,3-propanedione, 1,3-di(2-thenoyl)ꢁ1,3-propanedione [40]
1-ferrocenyl-4,4,4-trifluorobutaneꢁ1,3-dione [33] or acetylacetone
[41], is in agreement with experimental observations [10n,32], and
also indicate trans addition of CH3I to the square planar rhodium(I)
complexes. The trans addition of iodomethane to the bisphosphite
complex [Rh(CH3COCHCOCH3)(P(OPh)3)2], in agreement with 1H
NMR data [10f], is confirmed by a DFT computational study of the
transition state and product of the oxidative addition reaction [42].
Although the crystal structure study of [Rh(cupf)(CO)(PPh3)(-
CH3)(I)] [10c] is similar to the alkyl 2a in this study, the computa-
tional results of the oxidative addition reaction of this study
indicate that this structure is most probably due to the re-
arrangement of the trans addition product as was experimentally
found for the structurally similar [Rh(fctfa)(CO)(PPh3)(CH3)(I)]
[10m] and [Rh(acac)(CO)(PPh3)(CH3)(I)] [40] complexes (Hfctfa
¼ 1-ferrocenyl-4,4,4-trifluorobutaneꢁ1,3-dione and Hacac ¼ 2,4-
pentanedione (acetylacetone)). The optimized geometry of the
trans [Rh(cupf)(CO)(DPP)(CH3)(I)] alkyl product isomers is pre-
sented in Fig. 5.
suggesting surprisingly less p-back-bonding in the current complex
with its pyridyl-ring, compared to the triphenylphosphine
complex.
4.1.3. Density functional theory (DFT) study of Rh(I) and Rh(III)-
cupf complexes
The geometry of the most stable Rh(I) and Rh(III)-cupf com-
plexes is presented using DFT results since no crystal structures are
available for the complexes.
Square planar Rh(I) complexes: Two different isomers of the
[Rh(cupf)(CO)(DPP)] reactant are possible, which are labeled Rh(I)-
isomer A and Rh(I)-isomer B. For Rh(I)-isomer A, the DPP group is
cis to the O nearest to the cupf-Ph group and for Rh(I)-isomer B, the
DPP group is trans to the O nearest to the cupf-Ph group (Fig. 2). The
DFT optimized geometries of the isomers in acetone as solvent are
shown in Fig. 3. The population of the two isomers in selected
solvents are shown in Table 1. Results indicate that the population
of isomer A varies between 30 and 42%, depending on the solvent. It
When comparing the spinespin coupling of the methyl protons of
alkyl 1 and alkyl 2 with phosphorous (spin ¼ 1/2; 3J; 1He13Ce103Rhe
31P), the values are 2.0 and 3.7 Hz respectively. This difference is due
to the relative positions of the methyl and the phosphine ligands of
Rh(III) alkyl 1 and Rh(III) alkyl 2 [41]. A 3J value of 2 Hz and 4 Hz was
previously found for the trans and cis alkyl products respectively for
is well known that the two structural isomers of related [Rh(b-

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
443
Fig. 3. The ADF/PW91/TZP/acetone optimized geometries of the two structural isomers of [Rh(cupf)(CO)(DPP)]. The experimental geometries of two Rh(cupf)(CO)(PR3)] complexes
that have been characterized by X-ray structures (CSD reference code shown) are shown for comparative purposes.
both [Rh(fctfa)(CO)(PPh3)(CH3)(I)] [10m] and [Rh(acac)(CO)(PPh3)
(CH3)(I)] [41], consistent with the proposed geometry of the DFT
calculated most stable alkyl isomers of this study (Fig. 4).
thermodynamic stable reaction product to be square-pyramidal with
the COCH3 group in the apical position. However, it is not possible to
say if the observed acyl product is the final reaction product or only a
reaction intermediate leading to the most stable acyl product.
Five-coordinated Rh(III) acyl complexes: The CO insertion in a
rhodium(III) alkyl moiety leads to a rhodium(III) acyl product. Fig. 4
gives the geometry and DFT calculated relative energies of the
possible structural Rh(III) acyl isomers of the formula
[Rh(cupf)(DPP)(COCH3)(I)]. The square pyramidal [Rh(cupf)(DPP)
(COCH3)(I)] acyl products with the COCH3 moiety in the apical posi-
tion is the most stable (Fig. 6) for the optimized geometry. The en-
ergies of trigonalebipyramidal geometries are too high or optimized
to the square pyramidal geometry of acyl 2. The structurally similar
[Rh(cupf)(P(OCH2)3CCH3)(COCH3)(I)] acyl, as well as all crystal
structures of five-coordinated rhodium(III) acyl complexes [43]
adopts a square pyramidal geometry with the COCH3 moiety in the
apical position [44]. Therefore, both computational calculations and
crystallographic data are consistent with the geometry of the
4.2. Kinetic study
The reaction between [Rh(LL0)(CO)(DPP)] and CH3I can be pre-
sented as follows:
½RhðcupfÞðCOÞðDPPÞꢄ þ CH3I/Products
(1)
4.2.1. UV/vis spectra
The rate and type of reaction of oxidative addition between
[Rh(cupf)(CO)(DPP)] and CH3I was studied in four different
Table 1
ADF/PW91 calculated population of the optimized geometries of the two structural isomers of [Rh(cupf)(CO)(DPP)] in selected solvents. The energies of the transition state
complex, the ionic intermediate and the trans Rh(III) alkyl product is also tabulated. The energy of the reactants is taken as 0.
Solvent
Rh(I)
Boltzmann population
/%
Eact/kJmolꢁ1
Bent TS
cis TS
EIM/kJmolꢁ1
Ealkyl/kJmolꢁ1
Linear TS(trans addition)
Trans addition
Trans addition
Ethyl Acetate
Acetone
Isomer-A
Isomer-B
Isomer-A
Isomer-B
Isomer-A
Isomer-B
Isomer-A
Isomer-B
Isomer-A
Isomer-B
28.5
71.5
28.4
71.6
37.0
63.0
28.3
71.7
42.3
57.7
29
28
28
27
30
28
27
28
29
28
e
113
e
24
25
15
22
26
27
12
20
20
21
ꢁ34
ꢁ41
ꢁ33
ꢁ43
ꢁ33
ꢁ42
ꢁ32
ꢁ36
ꢁ31
ꢁ36
e
149
149
e
113
e
Chloroform
Acetonitrile
Methanol
113
e
e
e
114
e
e
e
116
e
e

444
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
Fig. 4. Geometries and ADF/PW91/TZP/acetone calculated relative energies (
D
ETBE) of the possible rhodium(III) alkyl and acyl products of the [Rh(cupf)(CO)(DPP)] þ CH3I reaction.
solvents. The addition of CH3I to an acetonitrile solution containing
[Rh(cupf)(CO)(DPP)] results in a change in the UV/visible spectrum.
A relatively fast absorption increase and then a slow absorption
decrease with the formation of an isosbestic point at 420 nm for the
second reaction takes place (Fig. 7). This initial result indicates the
existence of two possible reactions, with the second resulting in the
formation of only one product. A time scan of the same reaction
indicates the presence of two distinguishable reactions, with the
first reaction (absorption increase) having t ¼ 110 s and the second
½
slower reaction (absorption decrease) with t ¼ 1400 s (for
½

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
445
Fig. 5. ADF/PW91/TZP/acetone optimized geometry of the [Rh(cupf)(CO)(DPP)(CH3)(I)] alkyl 1 product isomers that result from trans addition of CH3I to [Rh(cupf)(CO)(DPP)].
[MeI] ¼ 0.5 M). A time scan separation of the two reactions results
in two distinguishable pseudo-first-order reactions. Similar results
are obtained for acetone and chloroform with isosbestic points at
monitoring of the acyl product with IR spectra. However, product
isolation from ethyl acetate solution after 500 min clearly indicated
the n(COCH3) at 1688 cmꢁ1, confirming the formation of the acyl
product in all the solvents that were investigated.
350 nm for both. Only one absorption change with t ¼ 3000 s (for
½
[MeI] ¼ 0.12 M) was observed for ethyl acetate using UV spectra
before solvent evaporation made it impossible to further follow the
reaction.
4.2.3. NMR study
The oxidative addition reaction between CH3I and [Rh(cupf)(-
CO)(DPP)] in CDCl3 at 21 ꢂC was kinetically followed on the NMR to
get more insight in the character of the reaction products formed
(Fig. 9). The 1H NMR spectra shows a fast decrease in the signals of
the Rh(I) starting complex with the immediate formation of the
two Rh(III) alkyl 1 isomers, identified by the overlapping signals at
1.32 and 1.31 ppm (see the Experimental section for the reso-
nances). The rate of formation of the Rh(III) alkyl 1 isomers are in
agreement with the experimental rates measured on the UV/vis in
chloroform. The signals of the Rh(III) alkyl 1 isomers soon decrease
at a rate of ca 0.0006 sꢁ1. Two Rh(III) acyl isomers, identified by the
broad signals at 3.05 and 2.93 ppm, form at the same slower, [MeI]
independent rate, 0.0006 sꢁ1, as the decrease in the Rh(III) alkyl 1
isomers signals. This rate is also the same as measured at 19.8 ꢂC on
UV/vis in chloroform.
4.2.2. IR spectra
An IR study of the oxidative addition reaction between CH3I and
[Rh(cupf)(CO)(DPP)] in acetonitrile shows a fairly rapid disap-
pearance of the Rh(I)eCO peak at 1980 cmꢁ1 with the simultaneous
appearance of the Rh(III)eCO stretching frequency at 2065 cmꢁ1
(see Fig. 1). The same spectrum shows the slow appearance of the
Rh(III)eCOCH3 peak at 1688 cmꢁ1. IR spectra for the same reaction
in acetone and chloroform also shows the same disappearance of
the Rh(I) peak and the formation of the Rh(III) alkyl peak. A com-
parison of the IR spectra, however, showed a substantially slower
Rh(III) alkyl formation reaction in ethyl acetate compared to the
other three solvents. Another important aspect of the IR spectra is
the complete disappearance of the Rh(I)eCO peak.
Regarding reaction mechanisms, the IR spectra (Fig.1) show that
an absorption change in Rh(I) occurs immediately (first order
change) with the addition of CH3I, but the increase in Rh(III) only
follows first order behavior between scans 1 and 2. The position of
the peak maximum for the Rh(III) alkyl product changes from the
initial 2070 cmꢁ1 to 2065 cmꢁ1. Exactly the same behavior is
evident for acetone and chloroform solvents. The absorption
change observed during the formation of the Rh(III) alkyl product
in ethyl acetate is much slower and shows first order behavior from
the first scan with no detectable peak shift (Fig. 8). The large ab-
sorption maximum at 1680 cmꢁ1 for ethyl acetate prevented the
4.3. Mechanism and rate constant
An important result from an initial IR kinetic study in acetoni-
trile at 25 ꢂC (Fig. 1) is that the observed rate for the disappearance
of the Rh(I) complex (t w 232 s) and the appearance of the Rh(III)
½
alkyl complex (t w 224 s) are within experimental error the same,
½
while the appearance of the Rh(III) acyl is much slower with
t
½
w 1136 s. The first reaction between CH3I and [Rh(cupf)(-
CO)(DPP)] was also studied under similar conditions using UV/
visible and IR in acetonitrile to try and identify or allocate the two
Fig. 6. ADF/PW91/TZP/acetone optimized geometry of the most stable [Rh(cupf)(DPP)(COCH3)(I)] acyl product isomers.

446
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
Fig. 9. Concentration-time data of the oxidative addition of CH3I to [Rh(cupf)(-
CO)(DPP)] in CDCl3 at 21 ꢂC as obtained from 1H NMR measurements. The rate
decrease of the Rh(III) alkyl 1 isomers are the same as the rate of increase of the Rh(III)
acyl isomers. IU ¼ integration units of the methyl peak of the Rh(III)-complexes.
between CH3I and [Rh(cupf)(CO)(DPP)] follows the same reaction
mechanism for acetonitrile, acetone and chloroform, while a
different mechanism is followed in ethyl acetate.
Fig. 7. Uv/vis spectra of the oxidative addition reaction of CH3I (0.5 M) with
[Rh(cupf)(CO)(DPP)] (2.5 ꢀ 10ꢁ4 M) in acetonitrile (2 min intervals, 25.0 ꢂC).
A kinetic study of the Rh(III)-acyl formation with [CH3I] varia-
tion at different temperatures in acetonitrile, acetone and chloro-
form (the second slower reaction observed on UV/vis) indicates a
reaction which is independent of [CH3I] (Fig. 12). The same reaction
in ethyl acetate is, however, very slow (and not followed on UV/vis
due to solvent evaporation) while the large IR absorption
maximum of ethyl acetate at 1680 cmꢁ1 prevents the kinetic IR
monitoring of the acyl formation in ethyl acetate.
The following reaction mechanism (Scheme 1) can be con-
structed for three of the solvents (acetonitrile, acetone and chlo-
roform). The mechanism postulates the rapid formation of an
Rh(III)* intermediate in an equilibrium reaction followed by the
formation of the Rh(III) alkyl product. In the next step, the Rh(III)
alkyl product continues to react (intra molecular CO insertion or
methyl migration) to form the Rh(III) acyl product.
UV/visible reactions to different species formed in solution. The
disappearance of the Rh(I) complex was followed by means of IR
and compared to the fast reaction followed by means of UV/visible.
The observed rate for the two reactions in acetonitrile (for
[MeI] ¼ 0.2 M) are 2.98 ꢀ 10ꢁ3 sꢁ1 and 3.45 ꢀ 10ꢁ3 sꢁ1 respectively,
indicating that the fast reaction observed by the UV/visible spectra
is indeed the disappearance of the Rh(I) complex or the formation
of Rh(III)-alkyl complex. The [CH3I] variation at different temper-
atures (Fig. 10) followed by UV/visible spectroscopy for acetonitrile
indicates non-linear kinetics.
A complete [CH3I] variation under similar conditions for IR and
UV was also studied in acetone and the results also show non-linear
kinetics for the first reaction. The pseudo-first-order rate constants
in Table 2 confirms the excellent correlation between the fast re-
action in UV/visible and the disappearance of the Rh(I) complex or
the Rh(III)-alkyl formation in IR.
4.3.1. Rate law for first step
Non-linear kinetics results are obtained for acetonitrile, acetone
and chloroform while the first reaction in ethyl acetate shows a
linear correlation between kobs and [CH3I] at different temperatures
(Fig. 11). These results suggest that the oxidative addition reaction
The following rate law, expressed in terms of the formation of
the Rh(III) intermediate, can be deducted for the oxidative addition
of CH3I to [Rh(cupf)(CO)(DPP)] in acetonitrile, acetone and chlo-
roform as solvents
Fig. 8. IR spectra for the reaction of CH3I (0.2 M) with [Rh(cupf)(CO)(DPP)] (2.0 ꢀ 10ꢁ2 M) in ethyl acetate (2 min intervals, 25.0 ꢂC).

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
447
Fig. 10. kobs against [CH3I] for oxidative addition (alkyl formation) of CH3I to [Rh(cupf)(CO)(DPP)] in acetonitrile at different temperatures as measured on the UV/vis
spectrophotometer.
d½RhðIÞꢄ
*
h
i
d½RhðIIIÞꢄ
¼ k1½RhðIÞꢄ½CH3Iꢄ þ k ½RhðIIIÞalkylꢄ
(6)
*
ꢁ1
¼ k1½RhðIÞꢄ½CH3Iꢄ ꢁ k
RhðIIIÞ ꢁ k2½RhðIIIÞalkylꢄ
dt
ꢁ1
dt
Under pseudo-first-order conditions ([CH3I] >> [Rh(I)]):
(2)
The pseudo-first-order rate constant ([CH3I] >> [Rh(I)]) for the
alkyl formation can be presented by the following:
kobs ¼ k1½CH3Iꢄ þ k
(7)
ꢁ1
The experimental data for ethyl acetate was fitted to Equation
(7) and the calculated constants for the reaction at different tem-
peratures are given in Table 4. The absence of any notable UV/
visible spectrum change for the acyl formation reaction, the acyl/
ethyl acetate IR spectral overlap as well as the slow reaction, which
allows for substantial solvent loss during the reaction, prevents a
thorough investigation for this reaction in ethyl acetate. It is,
however, assumed that CO insertion in ethyl acetate occurs, since
the acyl product is isolated from an ethyl acetate reaction mixture.
The kinetic results in Tables 3 and 4 indicate compatible alkyl
formation rate constants for acetonitrile, acetone and chloroform,
but a substantially slower reaction for ethyl acetate at 25 ꢂC. These
results also show little or no evidence of the influence of polarity or
donocity of the different solvents on the rate of the reaction. The
kinetic results show that the establishment of the equilibrium in
the initial reaction (formation of Rh(I)* in Scheme 1 for chloroform,
acetonitrile and chloroform) is much faster than the rate of alkyl
formation as illustrated by the almost identical rate constants for
the Rh(I) disappearance and the Rh(III) alkyl formation. The K-
values of approximately one for all the solvents studied are also
large enough to ensure a constant (pre-equilibrium approximation)
supply of intermediate to ensure the complete disappearance of the
Rh(I) monocarbonyl complex. The nature of the intermediate, as
suggested by the IR and the reaction Scheme 1, is also very inter-
esting. The direct formation of the Rh(III) peak (first at 2070 cmꢁ1
and the shift to 2065 cmꢁ1) from Rh(I) as suggested by the IR, with
no indication of a Rh(I) carbonyl stretching frequency shift, sug-
gests a more loosely interaction/association between the Rh(I)
center and the CH3I molecule rather than a formal bond formation
reaction. The absence of any shoulder or new peaks with the for-
mation of the Rh(III) alkyl product in the IR study also suggests the
formation of only one type of Rh(III) alkyl isomer during the first
step of the reaction. The formation of this intermediate also seems
to play an important role in the rate of alkyl and subsequently in
acyl formation as illustrated by the much slower reactions for ethyl
acetate where no evidence of any intermediate was observed. Large
negative entropy of activation in all the solvents indicates an
associative mechanism for the acyl formation. A slight shift in the
metal (III) carbonyl stretching frequency (2004e2028 cmꢁ1) is also
k2K½CH3Iꢄ
kobs
¼
(3)
1 þ K½CH3Iꢄ
where ðK ¼ k1=kꢁ1Þpredicts non-linear kinetics and is observed
for this reaction in three of the four different solvents.
4.3.2. Rate law for second step
The rate law for the acyl product formation from the reaction
scheme can be presented in the following way:
d½RhðIIIÞacylꢄ
¼ k3½RhðIIIÞalkylꢄ
(4)
dt
Under the prevailing experimental conditions, Equation (4)
simplifies to Equation (5) (reaction is independent of [CH3I])
kobs ¼ k3
(5)
The experimental data of the first reaction in acetonitrile,
acetone and chloroform are fitted to Equation (3) and the calculated
constants for the reactions at different temperatures and in
different solvents are given in Fig. 13 and Table 3.
The fact that the oxidative addition reaction between CH3I and
[Rh(cupf)(CO)(DPP)] in ethyl acetate shows different spectral and
much slower kinetic behavior (i.e. k2 in Scheme 1 is in the same
order as k1), a more simplified mechanism is postulated in Scheme
2.
The rate law for the alkyl product formation from this reaction
scheme can be presented in the following way:
Table 2
Comparative data for the UV and IR study for the oxidative addition reaction of CH3I
to [Rh(cupf)(CO)(DPP)] in acetone at 25 ꢂC.
[CH3I] (M)
UV 103 kobs (sꢁ1
(Rh(I) disappearance or
Rh(III)-alkyl formation)
)
IR 103 kobs (sꢁ1
)
(Rh(I) disappearance)
0.1012
0.3021
0.5011
0.7013
1.016
2.5(5)
3.6(8)
4.7(5)
5.9(5)
6.3(4)
1.57(6)
3.17(6)
4.2(5)
6.0(5)
6.7(5)
observed for the reaction between [Ir2(
CH3I [45], but this is also accompanied by a shift in Ir(I) carbonyl
m
-Dcbp)(CO)2(PCy3)2]ꢁ and

448
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
Fig. 11. kobs against [CH3I] for oxidative addition (alkyl formation) of CH3I to [Rh(cupf)(CO)(DPP)] in ethyl acetate different temperatures as measured on the UV/vis
spectrophotometer.
stretching frequency, attributed to the presence of two different Ir
species in the complex.
4.4. DFT study of the oxidative addition reaction
A comparison of the rate constants of the current study with
that of other [Rh(cupf)(CO)(PX3)] complexes is summarized in
Table 5. This comparative study allows for the isolation of the steric
or the electronic influence of the different phosphine ligands on the
rate of alkyl and acyl formation. For this comparison, the cone angle
for the DPP was estimated to be the same as that for PPh3 (145ꢂ) as
the only difference between the two ligands is the pyridyl-ring for
DPP compared to the normal phenyl ring for PPh3. The results in
Table 5 clearly indicate that the cone angle for the first four phos-
phine groups are the same, however an increase of a factor of 50
(from P(p-ClC6H4)3 to DPP) was observed for the rate of Rh(III) alkyl
formation and a factor 8 increase in Rh(III) acyl formation, with the
faster reaction for the electron rich DPP. The increase in oxidative
addition rate as well as acyl formation can be attributed to the
electronic parameter or electron donating property of the DPP
ligand. It is therefore anticipated that the additional electron den-
sity of this phosphine increases the nucleophilicity of the metal
center which subsequently increase the reaction rate with the CH3I
molecule.
It is widely accepted that the SN2 oxidative addition of simple
substrates such as MeI, to square planar complexes occurs via a
stepwise mechanism (Scheme 3). The first step is nucleophilic
substitution of iodide by the metal complex, presumed to proceed
with inversion of configuration at the carbon atom. The substitution
of iodide by the metal complex and the simultaneous expulsion of
an iodide ion lead to the formation of a five-coordinated species.
Subsequent coordination of iodide to the free coordination site of
the complex completes the addition to give a six-coordinate alkyl
complex. A wealth of kinetic, stereochemical, and spectroscopic
data support this mechanism. Two SN2 processes have been re-
ported for the trans addition, the “linear/back” and the “bent” TS,
see Scheme 4 [49]. The linear/back transition state structure cor-
responds to an SN2 mechanism, characterized by a linear Rh-
Cmethyl-I arrangement and by an Rh-Cmethyl-H angle close to 90ꢂ.
The methyl hydrogen atoms are located in the equatorial plane of
the five-coordinated carbon atom, resulting in a trigonal bipyra-
midal arrangement. The bent transition state structure corresponds
to a side-on approach of the Cmethyl-I bond to the rhodium atom
(Scheme 4). The bent transition state structure leads to the same
intermediate product as the linear/back transition state structure,
namely a cationic five-coordinated rhodium complex and a free
iodide ion. Another pathway, which has been considered, is
concerted cis or front addition and is expected to lead to the
The apparent discrepancies in the rates of oxidative addition
and mechanisms for the different solvents for the oxidative addi-
tion of CH3I to [Rh(cupf)(CO)(DPP)] prompted a DFT study to gain
more insight into this reaction, as well as to try and determine the
geometry of the transition state as predicted in Scheme 1.
Fig. 12. kobs against [CH3I] for insertion reaction (acyl formation) in acetonitrile at different temperatures as measured on the UV/vis spectrophotometer.

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
449
Scheme 1. Proposed mechanism for the oxidative addition of CH3I and [Rh(cupf)(CO)(DPP)] in acetone, acetonitrile and chloroform with k1 step fast, k2 step slow and, k3 very slow.
retention of configuration at carbon [49]. In the front transition
state structure the RheI and Rh-Cmethyl bonds form simultaneously
as I-Cmethyl bond breaks, resulting in the cis addition of the methyl
iodide (Scheme 4).
imaginary frequency (ꢁ313 and ꢁ303 for addition to isomers A and
B, respectively) of the TS corresponds to a CH3I stretching vibra-
tional mode in which the IeCCH3 bond breaks and the RheCCH3
bond forms.
The activation energy (Eact) for the linear, bent and front (cis) TS
calculated in this study for the title compound is given in Table 1
and from these results it is clear that the linear TS is energetically
favored. The geometry of the “linear” TS of the [Rh(cupf)(DPP)
(CO)] þ CH3I reaction for both isomers in acetone solvent is illus-
trated in Fig. 14. The TS involves the classical backside attack by the
rhodium nucleophile, leading to inversion of stereochemistry at
carbon. Both TS structures correspond to a SN2-like approach
characterized by a near linear IeCH3eRh arrangement and by an Ie
CH3eH angle close to 90ꢂ. The methyl hydrogen atoms are located
in the equatorial plane of the five-coordinated carbon atom,
showing the expected trigonal bipyramidal geometry. This kind of
structure has been characterized in a number of related cases
[50,51] and was first proposed by Griffin et al. [49]. In the TS
structures the RheC and CeI distances agree with a concerted
formation of the RheC bond and a rupture of the CeI bond. The
The formation of the TS is followed by the formation of the
cationic five-coordinate [Rh(cupf)(CO)(DPP)(CH3)]þ intermediate
(IM) with the CH3 group in the apical position and with the iodide
ion drifting away into the solvent sphere (Fig. 14). The relative
energy of the intermediate is 0.14 eV (13 kJ molꢁ1) less than that of
the transition state. The free iodide then coordinates to the
rhodium center to form the trans rhodium(III) alkyl oxidative
ꢀ
addition product with a final RheI bond length of 2.94 A and energy
of 0.5 eV (48 kJ molꢁ1) lower than the intermediate in acetone as
solvent (Fig. 4). In the next step, an inversion of the configuration of
hydrogen at the methyl carbon occurred and the methyl group is
fully bonded to the rhodium atom.
The calculated activation energy (Ea in Table 1) or the calculated
activation Gibbs energy (DG* in Table 6) of the front transition state
leading to alkyl 2a of Fig. 15, is ca. four and two times higher than
the calculated values leading to alkyl 1a and 2a, respectively. This
Fig. 13. kobs against [CH3I] for oxidative addition (alkyl formation) of CH3I to [Rh(cupf)(CO)(DPP)] in different solvents at 25 ꢂC as measured on the UV/vis spectrophotometer.

450
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
Table 3
Summary for the oxidative addition of CH3I to [Rh(cupf)(CO)(DPP)] at different temperatures in acetonitrile, acetone and chloroform as obtained from UV/vis measurements.
3
3
Solvent
Temp
10
10
K
ΔH*
ΔS*
o
( C)
-1
(M )
-1
(kJ.mol )
-1 -1
(J.K .mol )
k2 (alkyl)
k3(acyl)
-1 -1
(M s )
-1
(s )
Acetonitrile
10.1
19.8
25.0
5.0(1)
7.0(2)
17(6)
0.043(5)
0.29(6)
0.49(6)
1.99(1)
0.2(1)
0.9(1)
48(9)
-214(7)
Acetone
10.1
19.8
25.0
3.0(2)
8.0(2)
9.0(1)
0.044(5)
0.039(6)
1.1(5)
1.7(4)
1.4(1)
0.5(1)
34(8)
12(9)
-209(6)
-173(7)
Chloroform
10.1
19.8
25.0
1.4(4)
13.4(2)
20(3)
0.11(7)
0.59(6)
1.1(5)
0.39(2)
4.6(9)
2.6(4)
Scheme 2. Proposed mechanism for the oxidative addition of CH3I and [Rh(cupf)(CO)(DPP)] in ethyl acetate.
Table 4
Summary for the oxidative addition of CH3I to [Rh(cupf)(CO)(DPP)] at different temperatures in ethyl acetate as obtained from UV/vis measurements.
3
10 k1
4
10 k-1
Temperature
K
ΔH*
ΔS*
o
( C)
-1 -1
(M s )
-1
(s )
-1
(M )
-1
(kJ.mol )
-1 -1
(J.K .mol )
10.1
19.8
25.0
0.12(2)
1.03(3)
1.59(7)
0.60(1)
0.39(2)
0.98(4)
2.0(2)
2.64(1)
1.62(2)
19(9)
-238(8)

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
451
Table 5
Rate constants for the oxidative addition of CH3I to different [Rh(cupf)(CO)(PX3)] complexes in acetone at 25 ꢂC.
PX3
Cone angle/
q
Tolman electronic
pKa
103k (alkyl)
103k (acyl)(sꢁ1
)
Reference
parameter/
n
(cmꢁ1
)
(Mꢁ1 sꢁ1
)
P(p-ClC6H4)3
PPh3
P(p-MeOC6H4)3
DPP
145
145
2072.8
2068.9
2066.1
2066.7a
2056.4
1.03
2.73
4.57
4.07
9.65
0.193(8)
1.22(2)
4.20(8)
9.0(1)
0.14(6)
0.31(1)
0.20(5)
1.1(5)
[46,47]
[46,47]
[46,47]
This study, [45,48]
[46,47]
145
w145#
170
PCy3
1.94(3)
0.24(2)
P
a
3
Calcultated value according to Tolman eq..yCOðA1Þ ¼ 2056:1 þ i ¼ 1 cicmꢁ1
.
values for the front transition state do not match the experimental
values. The good agreement between the experimental and calcu-
lated values obtained for acetone and ethyl acetate as solvent,
highlight the fact that the oxidative addition reaction in both sol-
vents follows the same reaction path. The properties and the ge-
ometry of the calculated TS in the different solvents were found to
be very similar. The activation energy and relative energies of the
ionic intermediate and alkyl 1 product, are also very similar
(Table 1). The calculated results do not show a significant difference
in the mechanism of the first oxidative addition step of the
[Rh(cupf)(CO)(DPP)] þ CH3I reaction in different solvents. The
experimentally observed difference in the mechanisms of the
[Rh(cupf)(CO)(DPP)] þ CH3I reaction in ethyl acetate relative to the
other solvents studies, can thus not be explained by the computa-
tional chemistry results of this study.
Scheme 3. The widely accepted two step SN2 mechanism for the oxidative addition of
iodomethane to square planar rhodium complexes.
The experimentally observed difference in the mechanisms of
the [Rh(cupf)(CO)(DPP)] þ CH3I reaction in ethyl acetate relative to
the other solvents studies, is thus most probably due to the rate of
formation of the TS (k1 step in Scheme 1) and/or the build-up and
conversion of the IM in the different solvents which led to the
substantial difference in the rate of alkyl formation in the two sets
of solvents.
Scheme 4. Schematic illustration of a “linear/back”, “bent” and “front/concerted cis”
transition states of the nucleophilic attack of square planar rhodium(I) on methyl io-
dide. The Monsanto catalyst [Rh(CO)2I2]ꢁ is used as example for the square planar
rhodium(I) complex. The solid arrows indicate bond formation, the dotted arrows
bond breaking and the curved arrow the movement of CO.
5. Conclusion
result is in agreement with trans addition of iodomethane to
[Rh(cupf)(CO)(DPP)] leading to the formation of two trans alkyl
product isomers (alkyl 1a and 2a).
The activation parameters of the linear and selected front
transition states were calculated for the solvents used in this study
and compared to the experimentally obtained values in Table 6. The
best agreement between experimental and calculated values is
obtained for the linear transition state, whereas the calculated
This comparative study highlighted the complexity of factors
that ultimately influence or control the rate of oxidative addition
and finally acyl formation. The substitution of a phoshine’s phenyl
ring with a pyridyl group results in an increase in oxidative addition
rate for [Rh(cupf)(CO)(DPP)] in excess of an order of magnitude
relative to other [Rh(cupf)(CO)(PR3)] complexes, PR3 ¼ tertiary
phosphine. A change in solvent does not only play an important
Fig. 14. ADF/PW91/TZP/acetone optimized geometry of the linear TS of the [Rh(cupf)(DPP)(CO)] þ CH3I reaction. Left is the TS of the oxidative addition to isomer A and right to
isomer B.

452
W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
Table 6
Experimental and ADF/PW91 calculated activation parameters of the [Rh(cupf)(CO)(DPP)] þ CH3I reaction in different solvents.
Trans addition (linear TS) Cis addition (front TS)
H*/(kJ molꢁ1 S*/(J Kꢁ1 molꢁ1 H*/(kJ molꢁ1 S*/(J Kꢁ1 molꢁ1
D
)
D
)
D
G*/(kJ molꢁ1
)
D
)
D
)
D )
G*/(kJ molꢁ1
Ethyl acetate
Experimental
19(9)
25
30
34(8)
30
31
12(9)
29
22
48(9)
26
24
ꢁ238(8)
ꢁ230
90
94
83
96
83
76
64
90
97
112
86
91
91
77
Calc. isomer A þ CH3I
Calc. isomer B þ CH3I
Experimental
Calc. isomer A þ CH3I
Calc. isomer B þ CH3I
Experimental
Calc. isomer A þ CH3I
Calc. isomer B þ CH3I
Experimental
Calc. isomer A þ CH3I
Calc. isomer B þ CH3I
Calc. isomer A þ CH3I
Calc. isomer B þ CH3I
102
110
104
ꢁ221
ꢁ145
ꢁ194
167
153
162
ꢁ175
Acetone
ꢁ209(6)
ꢁ177
ꢁ150
Chloroform
Acetonitrile
Methanol
ꢁ173(7)
ꢁ204
ꢁ249
ꢁ214(7)
ꢁ200
108
113
ꢁ170
ꢁ137
159
154
ꢁ225
26
32
ꢁ221
ꢁ151
Fig. 15. ADF/PW91/TZP/acetone optimized geometry of the cationic five-coordinate [Rh(cupf)(CO)(DPP)(CH3)]þ intermediate (IM). Left is the IM of the oxidative addition to isomer
A and right to isomer B.
[6] A. Haynes, P.M. Maitlis, G.E. Morris, G.J. Sunley, H. Adams, P.W. Badger,
role in the rate of the reaction, but also on the type of reactive
species that can experimentally be observed during the reaction
C.M. Bowers, D.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)
sequence. Results from a DFT study do not show any difference in
the mechanism of the first oxidative addition step of the
[Rh(cupf)(CO)(DPP)] þ CH3I reaction in different solvents. The
experimentally observed difference in the rates and mechanisms of
the [Rh(cupf)(CO)(DPP)] þ CH3I reaction in ethyl acetate relative to
the other solvents studied, can most probably be attributed to the
rate of the formation and/or the build-up of the IM as well as the
rate of the alkyl 1 formation.
2847e2861, http://dx.doi.org/10.1021/ja039464y.
[7] T.W. Dekleva, D. Foster, Adv. Catal. 34 (1986) 81e130, http://dx.doi.org/
10.1016/S0360-0564(08)60591-X.
[8] (a) P.M. Maitlis, A. Haynes, G.J. Sunley, M.J. Howard, J. Chem. Soc. Dalton Trans.
(1996) 2187e2196, http://dx.doi.org/10.1039/DT9960002187;
(b) A. Hayes, B.E. Mann, D.J. Gulliver, G.E. Morris, P.M. Maitlis, Am. Chem. Soc.
113 (1991) 8567e8569, http://dx.doi.org/10.1021/ja00022a079.
[9] NIST Advanced Technology Programme on Catalysis and Biocatalysis, White
Paper: Leveraging Resources and Targeting Performance, 1998.
[10] (a) S.S. Basson, J.G. Leipoldt, J.T. Nel, Inorg. Chim. Acta 86 (1984) 167e172;
(b) S.S. Basson, J.G. Leipoldt, A. Roodt, J.A. Venter, T.J. van der Walt, Inorg.
Chim. Acta 119 (1986) 35e38, http://dx.doi.org/10.1016/S0020-1693(00)
81327-4;
Acknowledgments
(c) S.S. Basson, J.G. Leipoldt, A. Roodt, J.A. Venter, Inorg. Chim. Acta 128 (1987)
31e37, http://dx.doi.org/10.1016/S0020-1693(00)84691-5;
The authors thank the Research Fund of this University and the
National Research Foundation of the Republic of South Africa for
financial support.
(d) J.G. Leipoldt, S.S. Basson, L.J. Botha, Inorg. Chim. Acta 168 (1990) 215e220,
http://dx.doi.org/10.1016/S0020-1693(00)80945-7;
(e) J.G. Leipoldt, E.C. Steynberg, R. van Eldik, Inorg. Chem. 26 (1987) 3068e
3070, http://dx.doi.org/10.1021/ic00265a029;
(f) G.J. van Zyl, G.J. Lamprecht, J.G. Leipoldt, T.W. Swaddle, Inorg. Chim. Acta
143 (1988) 223e227, http://dx.doi.org/10.1016/S0020-1693(00)83693-2;
(g) J.G. Leipoldt, G.J. Lamprecht, G.J. van Zyl, Inorg. Chim. Acta 96 (1985) L31e
L34, http://dx.doi.org/10.1016/S0020-1693(00)93728-9;
References
(h) D.M.C. Smit, S.S. Basson, A. Roodt, E.C. Steynberg, Rhodium Express 7 and 8
(1994) 12e16;
[1] J. Hagen, Industrial Catalysis, a Practical Approach, second ed., Wiley VCH,
Weinheim, 2006, pp. 425e427.
(i) Y.M. Terblans, S.S. Basson, W. Purcell, G.J. Lamprecht, Acta Crystallogr. C51
(1995) 1748e1750, http://dx.doi.org/10.1107/S010827019500254X;
(j) P. Ebenebe, S.S. Basson, W. Purcell, Rhodium Express 16 (1996) 11e16;
(k) M. Theron, E. Grobbelaar, W. Purcell, S.S. Basson, Inorg. Chim. Acta 358
(2005) 2457e2463, http://dx.doi.org/10.1016/j.ica.2004.06.070;
(l) E. Grobbelaar, W. Purcell, S.S. Basson, Inorg. Chim. Acta 359 (2006) 3800e
[2] S. Scott, Canadian Chemical News, 1999. www.allbusiness.com/north-
america/Canada/ (accessed 18.03.10).
[3] H. Kidwell, www.in-pharmatechnologist.com/content/, 29 January 2008
(accessed 16.03.10).
[4] Platinum Today, Johnson Matthey, 2004 (accessed 16.03.10).
[5] CORDIS.Europe.eu., 2006 (accessed 18.03.10).

W. Purcell et al. / Journal of Organometallic Chemistry 745-746 (2013) 439e453
453
3806, http://dx.doi.org/10.1016/j.ica.2006.03.041;
(m) J. Conradie, G.J. Lamprecht, A. Roodt, J.C. Swarts, Polyhedron. 26 (2007)
5075e5087, http://dx.doi.org/10.1016/j.bbr.2011.03.031;
(n) M.M. Conradie, J. Conradie, Inorg. Chim. Acta 361 (2008) 2285e2295,
http://dx.doi.org/10.1016/j.ica.2007.10.052;
(o) J. Conradie, J.C. Swarts, Organometallics 28 (2009) 1018e1026, http://
dx.doi.org/10.1021/om800655j;
[26] C.C. Pye, T. Ziegler, Theor. Chem. Acc. 101 (1999) 396e408, http://dx.doi.org/
10.1007/s002140050457.
[27] J.L. Pascual-Ahuir, E. Silla, I. Tuñon, J. Comput. Chem. 15 (1994) 1127e1138,
http://dx.doi.org/10.1002/jcc.540151009.
[28] http://www.chemicalize.org/structure/#!mol¼diphenyl-2-pyridylphosphine
(accessed 27.10.11).
[29] S.S. Basson, J.G. Leipoldt, A. Roodt, J.A. Venter, Inorg. Chim. Acta 118 (1986)
45e46.
[30] W. Purcell, J. Conradie, T.T. Chiweshe, J.A. Venter, H.G. Visser, M.P. Coetzee,
J. Mol. Struct. 1038 (2013) 220e229, http://dx.doi.org/10.1016/
j.molstruc.2013.01.061.
(p) A.J. Muller, J. Conradie, W. Purcell, S.S. Basson, J.A. Venter, S. Afr. J. Chem.
63 (2010) 11e19.
[11] (a) N.W. Alcock, P. Moore, P.A. Lampe, K.F. Mok, J. Chem. Soc. Dalton Trans.
(1982) 207e210, http://dx.doi.org/10.1039/DT9820000207;
(b) M.M. Olmstead, A. Maisonhat, J.P. Farr, A.L. Balch, Inorg. Chem. 20 (1981)
4060e4064, http://dx.doi.org/10.1021/ic50226a005;
[31] J.J.C. Erasmus, M.M. Conradie, J. Conradie, React. Kinet. Mech. Catal. 105 (2012)
233e247, http://dx.doi.org/10.1007/s11144-011-0413-1.
(c) Y. Inoguchi, B. Milewski-Mahrla, H. Schmidbauer, Chem. Ber. 115 (1982)
3085e3095, http://dx.doi.org/10.1002/cber.19821150914;
[32] M.M. Conradie, J. Conradie, Inorg. Chim. Acta 361 (2008) 208e218, http://
dx.doi.org/10.1016/j.ica.2007.07.010.
(d) H.J. Wasserman, D.C. Moody, R.T. Paine, R.R. Ryan, K.V. Salazar, J. Chem.
Soc. Chem. Commun. (1984) 533e534.
[33] M.M. Conradie, J. Conradie, S. Afr. J. Chem. 61 (2008) 102e111.
[34] J. Conradie, G.J. Lamprecht, S. Otto, J.C. Swarts, Inorg. Chim. Acta 328 (2002)
191e203, http://dx.doi.org/10.1016/S0020-1693(01)00731-9.
[35] K.H. Hopmann, N.F. Stuurman, A. Muller, J. Conradie, Organometallics 29
(2010) 2446e2458, http://dx.doi.org/10.1021/om1000138.
[36] N.F. Stuurman, J. Conradie, J. Organomet. Chem. 694 (2009) 259e268, http://
dx.doi.org/10.1016/j.jorganchem.2008.10.040.
[37] A.M.M. Trzeciak, J.J. Ziólkowsk, Inorg. Chim. Acta 96 (1985) 15e20, http://
dx.doi.org/10.1016/S0020-1693(00)93731-9.
[38] W. Purcell, S.S. Basson, J.G. Leipoldt, A. Roodt, H. Preston, Inorg. Chim. Acta 234
(1995) 153e156, http://dx.doi.org/10.1016/0020-1693(95)04469-P.
[39] N.F. Stuurman, A. Muller, J. Conradie, Inorg. Chim. Acta 395 (2013) 237e244,
http://dx.doi.org/10.1016/j.ica.2012.11.001.
[12] (a) U. Abram, R. Alberto, J.R. Dilworth, Y. Zheng, K. Ortner, Polyhedron. 18
(1999) 2995e3003, http://dx.doi.org/10.1016/S0277-5387(99)00216-8;
(b) P. Govindaswamy, P.J. Carroll, Y.A. Mozharivskyj, M.R. Kollipara, J. Chem.
Sci. 118 (2006) 319e326, http://dx.doi.org/10.1007/BF02708525;
(c) J.P. Farr, M.M. Olmstead, F.E. Wood, A.L. Balch, J. Am. Chem. Soc. 105 (1983)
792e798, http://dx.doi.org/10.1021/ja00342a024.
[13] T.T. Chiweshe, W. Purcell, J.A. Venter, S. Afr. J. Chem. 66 (2012) 7e16.
[14] Scientist, Version 2, MicroMath Scientific Software, Saint Louis, Mo, 2004.
ꢁ
[15] A. Rusina, A.A. Vlcek, Nature 206 (1965) 295e296, http://dx.doi.org/10.1038/
206295a0.
[16] G. te Velde, F.M. Bickelhaupt, E.J. Baerends, C. Fonseca Guerra, S.J.A. van Gis-
bergen, J.G. Snijders, T. Ziegler, J. Comput. Chem. 22 (2001) 931e967, http://
dx.doi.org/10.1002/jcc.1056.
[40] M.M. Conradie, J. Conradie, Inorg. Chim. Acta 362 (2009) 519e530, http://
dx.doi.org/10.1016/j.ica.2008.04.046.
[17] C. Fonseca Guerra, J.G. Snijders, G. te Velde, E.J. Baerends, Theor. Chem. Acc. 99
(1998) 391e403, http://dx.doi.org/10.1007/s002140050353.
[18] ADF2009.01, SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam, The
Netherlands.
[41] M.M. Conradie, J. Conradie, Dalton Trans. 40 (2011) 8226e8237, http://
dx.doi.org/10.1039/C1DT10271K.
[42] M.M. Conradie, J. Conradie, J. Organomet. Chem. 695 (2010) 2126e2133,
http://dx.doi.org/10.1016/j.jorganchem.2010.05.021.
[19] (a) J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh,
C. Fiolhais, Phys. Rev. B 46 (1992) 6671e6687, http://dx.doi.org/10.1103/
PhysRevB.46.6671;
[43] Cambridge Structural Database (CSD). Version 5, 2010, p. 32.
[44] J.A. Venter, W. Purcell, H.G. Visser, Acta Crystallogr. E65 (2009) m1528e
m1529, http://dx.doi.org/10.1107/S1600536809043050.
(b) J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Perderson,
D.J. Singh, C. Fiolhais, Phys. Rev. B 48 (1993) 4978, http://dx.doi.org/10.1103/
PhysRevB.48.4978.2.
[45] E. Grobbelaar, J. Conradie, W. Purcell, J.A. Venter, W. Purcell, T.T. Chiweshe,
J. Organomet. Chem. 696 (2011) 1990e2002, http://dx.doi.org/10.1016/
j.jorganchem.2010.10.049.
[20] L. Fan, T. Ziegler, J. Chem. Phys. 96 (1992) 9005e9012, http://dx.doi.org/
10.1063/1.462258.
[46] D.E. Graham, G.J. Lamprecht, I.M. Potgieter, A. Roodt, J.G. Leipoldt, Transit.
Met. Chem. 16 (1991) 193e195.
[21] L. Fan, T. Ziegler, J. Amer. Chem. Soc. 114 (1992) 10890e10897, http://
dx.doi.org/10.1021/ja00053a027.
[47] C.A. Tolman, Chem. Rev. 77 (1977) 313e348, http://dx.doi.org/10.1021/
cr60307a002.
[22] J.H. Espenson, Chemical Kinetics and Reaction Mechanisms, second ed., 15,,
McGraw-Hill, New York, 1995, 49, 70e75, 156.
[48] C.A. Tolman, J. Am. Chem. Soc. 92 (1970) 2953e2956, http://dx.doi.org/
10.1021/ja00713a007.
[23] A. Klamt, G. Schüürmann, J. Chem. Soc. Perkin Trans. 2 (1993) 799e805,
http://dx.doi.org/10.1039/P29930000799.
[24] A. Klamt, J. Phys. Chem. 99 (1995) 2224e2235, http://dx.doi.org/10.1021/
j100007a062.
[49] T.R. Griffin, D.B. Cook, A. Haynes, J.M. Pearson, D. Monti, G.E. Morris, J. Am.
Chem. Soc. 118 (1996) 3029e3030, http://dx.doi.org/10.1021/ja952952o.
[50] B. Bogdanov, T.B. McMahon, Int. J. Mass Spectrom. 241 (2005) 205e223,
http://dx.doi.org/10.1016/j.ijms.2004.12.018.
[25] A. Klamt, V. Jones, J. Chem. Phys. 105 (1996) 9972e9981, http://dx.doi.org/
10.1063/1.472829.
[51] M. Feliz, Z. Freixa, P.W.N.M. van Leeuwen, C. Bo, Organometallics 24 (2005)
5718e5723, http://dx.doi.org/10.1021/om050554g.