Kinetic study of the oxidative addition reaction between methyl iodide and [Rh(FcCOCHCOCF3)(CO)(PPh3)]: Structure of [Rh(FcCOCHCOCF3)(CO)(PPh3)(CH3)(I)]

Article abstract of DOI:10.1016/j.poly.2007.07.004

The kinetics of oxidative addition of CH₃I to [Rh(FcCOCHCOCF₃)(CO)(PPh₃)], where Fc = ferrocenyl and (FcCOCHCOCF₃)^- = fctfa = ferrocenoylacetonato, have been studied utilizing UV/Vis, IR, ¹H and ³¹P NMR techniques. Three definite sets of reactions involving isomers of at least two distinctly different classes of Rh^III-alkyl and two different classes of Rh^III-acyl species were observed. Rate constants for this reaction in CHCl₃ at 25 °C, applicable to the reaction sequence below, were determined as k₁ = 0.00611(1) dm³ mol^-1 s^-1, k_-1 = 0.0005(1) s^-1, k₃ = 0.00017(2) s^-1 and k₄ = 0.0000044(1) s^-1 while k_-3 ? k₃ and k_-4 ? k₄ but both ≠0. The indeterminable equilibrium K₂ was fast enough to be maintained during Rh^I depletion in the first set of reactions and during the Rh^IIIalkyl2 formation in the second set of reactions. From a ¹H and ³¹P NMR study in CDCl₃, K_c1 was found to be 0.68, K_c2 = 2.57, K_c3 = 1.00, K_c4 = 4.56 and K_c5 = 1.65. {A figure is presented}. The above reaction sequence represents a completely general reaction sequence for the oxidative addition of iodomethane to any β-diketonato complex of the type [Rh(β-diketonato)(CO)(PPh₃)], although the equilibrium K₂ may not necessarily always be fast. A temperature dependence study in chloroform led to the activation enthalpies, ΔH^#, and activation entropies, ΔS^#, for both the first and second sets of reactions. A solvent dependence study utilizing five different solvents showed the rate of the first set of reactions are directly proportional to the dielectric constant of the solvent. The molecular formula of all the Rh^III-alkyl and Rh^III-acyl species are [Rh(FcCOCHCOCF₃)(CH₃)(CO)(PPh₃)(I)] and [Rh(FcCOCHCOCF₃)(COCH₃)(PPh₃)(I)], respectively (Fc = ferrocenyl), but the geometries of the two Rh^III isomers and the initial Rh^I species all differ due to different coordination spheres. A crystal structure determination (Z = 4, monoclinic, P2₁/c) of [Rh(FcCOCHCOCF₃)(CO)(PPh₃)(CH₃)(I)], an isomer of the Rh^III-alkyl2 species, is also reported.

Full text of DOI:10.1016/j.poly.2007.07.004

Polyhedron 26 (2007) 5075–5087
www.elsevier.com/locate/poly
Kinetic study of the oxidative addition reaction between methyl
iodide and [Rh(FcCOCHCOCF3)(CO)(PPh3)]: Structure
of [Rh(FcCOCHCOCF3)(CO)(PPh3)(CH3)(I)]
*
Jeanet Conradie, Gert J. Lamprecht, Andreas Roodt, Jannie C. Swarts
Department of Chemistry, University of the Free State, Bloemfontein 9300, South Africa
Received 16 May 2007; accepted 10 July 2007
Available online 28 August 2007
Abstract
The kinetics of oxidative addition of CH3I to [Rh(FcCOCHCOCF3)(CO)(PPh3)], where Fc = ferrocenyl and (FcCOCH-
1
COCF3)ꢀ = fctfa = ferrocenoylacetonato, have been studied utilizing UV/Vis, IR, H and 31P NMR techniques. Three definite sets of
reactions involving isomers of at least two distinctly different classes of RhIII-alkyl and two different classes of RhIII-acyl species were
observed. Rate constants for this reaction in CHCl3 at 25 ꢀC, applicable to the reaction sequence below, were determined as
k1 = 0.00611(1) dm3 molꢀ1 sꢀ1, kꢀ1 = 0.0005(1) sꢀ1, k3 = 0.00017(2) sꢀ1 and k4 = 0.0000044(1) sꢀ1 while kꢀ3 ꢁ k3 and kꢀ4 ꢁ k4 but
both ¼0. The indeterminable equilibrium K2 was fast enough to be maintained during RhI depletion in the first set of reactions and dur-
1
ing the RhIIIalkyl2 formation in the second set of reactions. From a H and 31P NMR study in CDCl3, Kc1 was found to be 0.68,
Kc2 = 2.57, Kc3 = 1.00, Kc4 = 4.56 and Kc5 = 1.65.
First set of reactions
Second set of reactions
Third set of reactions
The above reaction sequence represents a completely general reaction sequence for the oxidative addition of iodomethane to any b-
diketonato complex of the type [Rh(b-diketonato)(CO)(PPh3)], although the equilibrium K2 may not necessarily always be fast. A tem-
perature dependence study in chloroform led to the activation enthalpies, DH#, and activation entropies, DS#, for both the first and
second sets of reactions. A solvent dependence study utilizing five different solvents showed the rate of the first set of reactions are directly
proportional to the dielectric constant of the solvent. The molecular formula of all the RhIII-alkyl and RhIII-acyl species are [Rh(FcC-
OCHCOCF3)(CH3)(CO)(PPh3)(I)] and [Rh(FcCOCHCOCF3)(COCH3)(PPh3)(I)], respectively (Fc = ferrocenyl), but the geometries of
the two RhIII isomers and the initial RhI species all differ due to different coordination spheres. A crystal structure determination
(Z = 4, monoclinic, P21/c) of [Rh(FcCOCHCOCF3)(CO)(PPh3)(CH3)(I)], an isomer of the RhIII-alkyl2 species, is also reported.
ꢁ 2007 Elsevier Ltd. All rights reserved.
Keywords: Rhodium complexes; Ferrocene; b-Diketonates; Kinetics; Oxidative addition
1. Introduction
Platinum group transition metal complexes are well
known for their medical applications [1] and are often
*
Corresponding author. Tel.: +27 0 51 4012781; fax: +27 0 51 4446384.
E-mail address: swartsjc.sci@ufs.ac.za (J.C. Swarts).
0277-5387/$ - see front matter ꢁ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2007.07.004

5076
J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
employed as either homogeneous [2] or heterogeneous [3]
catalysts in a variety of chemical reactions. Rhodium-
based homogeneous catalysts in particular may be used
inter alia in carbon–carbon bond formation [4], cycloprop-
anation [5], decarbonylation of aldehydes [6], hydrobora-
tion of olefins [7], hydroformylation of olefins [8],
hydrogenation [9], hydrosilation [10] and carbon monoxide
insertion reactions [11–13]. Oxidative addition of iodome-
thane to square planer rhodium(I) or iridium(I) complexes
and subsequent CO insertion reactions is important in the
Monsanto and Cativa industrial processes [12,14] for the
production of acetic acid from methanol. The rate-deter-
mining step in the rhodium-based Monsanto process is
the oxidative addition of methyl iodide to [Rh(CO)2I2]ꢀ.
In contrast, during cis-[Ir(CO)2I2]ꢀ-catalyzed methyl car-
bonylation in the Cativa process, CO insertion in
[CH3Ir(CO)2I3]ꢀ to give [(CH3CO)Ir(CO)2I3]ꢀ is rate-
determining. It is known [15–17] that rhodium(I) com-
plexes of the type [Rh(L,L0)(CO)(PPh3)], L,L0 = mono
charged bidentate ligands, can also facilitate the formation
of an acetyl species following oxidative addition of methyl
iodide. Depending on what L,L0 is, a variety of reaction
mechanisms for this reaction has been reported. These
mechanisms are summarized in Eq. (1) (L,L0) = methyl(2-
methyl-amino-1-cyclopentene-1-dithiocarboxylato, [17]),
(2) (L,L0 ꢀthioacetylacetonato [15d]), (3) (L,L0 = inter alia
representative. Structurally it differs from the alkyl species
in that the CO ligand inserted between the Rh center and
the alkyl, here a methyl ligand, to generate an acyl ligand.
We were interested to establish whether a ferrocenyl (Fc)
fragment as part of the b-diketonato ligand in complexes of
the type [Rh(b-diketonato)(CO)(PPh3)] may enhance acyl
formation during oxidative addition. Utilizing L,L0 =
(FcCOCHCOCF3)ꢀ = fctfa, we show with this communi-
cation that incorporation of the ferrocenyl group in the
b-diketonato ligand significantly enhances the rate of acyl
formation as well as the ensuing carbonyl insertion and
deinsertion to form new RhIII-alkyl and RhIII-acyl species
during the oxidative addition of CH3I to [Rh(fctfa)-
(CO)(PPh3)]. We especially demonstrate how a variety of
different spectroscopic techniques (UV/Vis, IR, 1H and
31P NMR) may be used to obtain different information
packages and a better insight into the overall process that
takes place during this general type of reaction. Especially
satisfying was the realization that each of the previous pro-
posed mechanisms (Eqs. (1)–(4)) are only special cases of
the more general mechanism given in the abstract for the
reaction between [Rh(fctfa)(CO)(PPh3)] and CH3I.
2. Experimental
2.1. Materials and apparatus
1,1,1-trifluoro-2,4-pentanedionato
[15a])
and
(4)
(L,L0 = inter alia 2,4-pentanedionato [15b]) below. The
notation RhIII-alkyl1 and RhIII-acyl1 in the four equations
below indicates the first formed alkylated, [RhIII(R)], or
acylated rhodium(III) species, [RhIII(COR)]. When the last
NMR measurements, at 298 K unless otherwise stated,
were recorded on a Bruker Advance DPX 300 NMR spec-
trometer [1H (300.130 MHz), and 31P (121.495 MHz)].
Chemical shifts are reported as d values relative to SiMe4
1
number in the notation changes to ‘‘2’’, such as in RhIII
-
at 0 ppm for the H spectra or relative to 85% H3PO4 at
alkyl2 in Eqs. (3) and (4), it shows that after the first alkyl-
ated or acylated species has formed, it converted to a
second, different but more stable, alkylated or acylated
geometrical isomer.
0 ppm for the 31P spectra. Infrared spectra were recorded
on a Hitachi 270-50 infrared spectrometer either in a KBr
matrix or in chloroform or acetone solutions. UV measure-
ments were recorded on a GBC-916 UV/Vis spectrometer.
Liquid reactants and solvents were distilled prior to use,
water was double distilled. DMF was dried according to
published methods [18]. Solutions of [Rh(fctfa)(CO)(PPh3)]
in acetone, methanol, ethyl acetate, benzene and chloro-
k1;K
k2
RhI þ CH3I ꢀ RhIII-alkyl1 ! RhIII-acyl1
ð1Þ
ð2Þ
k
ꢀ1
k1
k2
RhI þ CH3I ! RhIII-alkyl1 ! RhIII-acyl1
1
RhI þ CH3I kꢀ1;K RhIII-alkyl1
form [19] were tested for stability by means of H NMR
and overlay IR and UV spectra for at least 24 h.
k
ꢀ1
Synthesis of [Rh(fctfa)(CO)(PPh3)(CH3(I)]: Methyl
iodide (1.9 g, 13.4 mmol) was added to [Rh(fctfa)-
(CO)(PPh3)] [20] (75 mg, 0.1 mmol) dissolved in hexane
(70 ml) at 30 ꢀC. N2 was purged through the reaction vessel
before it was sealed and left at room temperature in the
dark. After 5 days crystals suitable for X-ray structure
determination were obtained from the reaction mixture,
90% yield, m.p.: 235 ꢀC; C, 47.5; H 3.2, C34H28F3FeO3-
k2
k3
! RhIII-acyl1 ! RhIII-alkyl2
ð3Þ
RhI þ CH3I kꢀ1;K RhIII-alkyl1
k
ꢀ1
k2
k3
ꢀ RhIII-acyl1 ꢀ RhIII-alkyl2
ð4Þ
k
k
ꢀ3
ꢀ2
In, for example, Eq. (4), [Rh(H3CCOCHCOCH3)(CO)-
(PPh3)(CH3)(I)] would be a typical formula describing the
two different RhIII-alkyl1 or RhIII-alkyl2 geometrical iso-
mers. The first example of a single crystal structure deter-
mination of a RhIII-alkyl species of this type is described
later in this publication and Fig. 8 shows clearly the coor-
dination modes. For the RhIII-acyl1 isomer of Eq. (4), the
formula [Rh(H3CCOCHCOCH3)(COCH3)(PPh3)(I)] is
RhPI requires C, 47.6; H, 3.3%). IR (KBr), mmax/cmꢀ1
:
1
2056 (C@O); H NMR(CDCl3): isomer A: 1.78 (3H, d d,
CH3), 4.30 (5H, s, C5H5 and 1H, m, C5H4), 4.40 (1H, m,
C5H4), 4.48 (1H, m, C5H4), 4.77 (1H, m, C5H4), 5.31
(1H, s, CH) and 7.25–7.50 (15H, m, aromatic); isomer B:
1.78 (3H, d d, CH3), 4.24 (5H, s, C5H5 and 1H, m,
C5H4), 4.38 (1H, m, C5H4), 4.60 (1H, m, C5H4), 4.80

J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
5077
(1H, m, C5H4), 5.45 (1H, s, CH) and 7.25 – 7.50 (15H, m,
aromatic).
surements. The activation parameters DH#, DS# and DG#
for the reaction in chloroform and acetone were obtained
from kinetic runs between 15.0 and 35.0 ꢀC. At least three
temperatures, held constant within 0.1 ꢀC, were employed
to establish the temperature dependence for each substitu-
tion reaction.
2.2. Crystallography
Crystal density of [Rh(fctfa)(CO)(PPh3)CH3(I)] was
determined by flotation in sodium iodide solution. A crys-
tal of dimensions 0.075 · 0.150 · 0.650 mm was used for
data collection on a Syntex P-1 diffractometer with Nb fil-
tered utilizing Mo Ka radiation. The x/2h-scan technique
was used with variable scan width Dx = (0.43 + 0.34
tanh)ꢀ, scan speed 5.49ꢀ minꢀ1 in x and a maximum scan
time of 60 s per reflection. Unit cell parameters were deter-
mined from least-squares refinement of 25 reflections with
15 < h < 20ꢀ (total measuring range 3 < h < 25ꢀ). Empirical
absorption corrections [21] were applied with minimum
and maximum correction factors of 0.4529 and 0.7413,
respectively. The mean intensity of three standard reflec-
tions, measured over 7200 s of X-ray exposure time, varied
2.4. Calculations
Pseudo first-order rate constants, k1obs, for reaction set 1
were calculated by fitting kinetic data to the first-order
equation [26]
½Aꢂt ¼ ½Aꢂ0 expðꢀk1obstÞ
ð5Þ
with [A]t and [A]0 the absorbance of the indicated species at
time t and 0 (UV/Vis or IR experiments), or integral values
for the specified peaks on NMR spectra, and were con-
verted to second order rate constants, k1, by determining
the slope of the linear plots of k1obs against the concentra-
tion of the incoming CH3I ligand with
from the initial value by less tꢀh1an ꢀ1%. All possible reflec-
˚
tions with (sinh)/k < 0.59 A
in the range ꢀ12 < h <
kobs ¼ kꢀ1 þ k1½CH3Iꢂ
ð6Þ
12.0 < k < 18.0 < l < 24 were measured, giving 4420 unique
reflections of which 4414 were considered observed
[I > 2.0r(I)]. The structure was solved by the heavy-atom
method using SHELXS86 [22] and was subjected to aniso-
tropic full-matrix least-squares refinement on F (a total of
380 parameters refined) using SHELXS93 [23]. The H-atom
positions were calculated riding on the adjacent C atoms
Here k is the first-order rate constant for the reverse
ꢀ1
step [26] in the general mechanism
k1
A þ CH3I ꢀ B
k
ꢀ1
with A = [Rh(fctfa)(CO)(PPh3)] and B = {[Rh(fctfa)(CH3)-
(CO)(PPh3)(I)] ꢀ [Rh(fctfa)(COCH3)(PPh3)(I)]}, or more
conveniently expressed as [RhIII-alkyl1 ꢀ RhIII-acyl1].
First-order rate constants for reaction set 2, k3, and reac-
tion set 3, k4, were also determined from Eq. (5) by replac-
ing k1obs with k3 or k4. Kinetic data for the consecutive
reaction treatment [26] of IR data obtained in acetone
˚
assuming C–H = 0.930 A, and were refined with an overall
temperature factor. Neutral-atom scattering factors were
taken from Cromer and Mann [24] and anomalous-disper-
sion corrections for rhodium from International Tables for
X-ray Crystallography [25]. Final R = 0.0588 and wR =
0.1692 (unit weights), (Dq)max = 1.841, (Dq)min = ꢀ1.087
ꢀ3
˚
k1
k3
e A and (D/r)max = ꢀ0.745. The high value of the maxi-
mum residual electron density (Dq)max of 1.841 is due to
irrelevant peaks near the heavy Rh-atom.
and applicable to the general reaction A ! B ! C were fit-
ted to Eq. (5) when monitoring the disappearance of A, Eq.
(7) for monitoring the appearance and disappearance of B
and Eq. (8) for monitoring the appearance of C, here RhIII
alkyl2. A and B were defined above.
-
2.3. Kinetic measurements
The kinetic rate constants were determined utilizing UV/
Vis (by monitoring the change in absorbance at 375, 400,
530 and 580 nm), IR (by monitoring formation and disap-
pearance of the carbonyl peaks at 2082, 2064, 1990, 1729
k1obs½Aꢂ0
½Bꢂt ¼
½expðꢀk1obstÞ ꢀ expðꢀk3tÞꢂ
ð7Þ
k3 ꢀ k1obs
ꢀ
ꢁ
1
½Cꢂt ¼ ½Aꢂ0 1 ꢀ
½k3 expðꢀk1obstÞ ꢀ k1obs expðꢀk3tÞꢂ
k3 ꢀ k1obs
1
and 1714 cmꢀ1), H and 31P NMR spectroscopy (by mon-
ð8Þ
itoring the change in integration units of the various signals
with time as specified in Table 2 and Fig. 6). A linear rela-
tionship between UV absorbance, A, and concentration, C,
confirmed the validity of the Beer Lambert law (A = eCl
with l = path length = 1 cm) for the [Rh(fctfa)(CO)(PPh3)]
complex. All kinetic measurements were monitored under
pseudo first-order conditions with CH3I concentrations
7–10000 times the concentration of the [Rh(fctfa)-
(CO)(PPh3)] complex. At least five different concentrations
within this range were utilized. The concentration of
[Rh(fctfa)(CO)(PPh3)] was 60.0004 mol dmꢀ3 for UV/Vis
measurements and 60.01 mol dmꢀ3 for IR and NMR mea-
Activation parameters [26] were obtained from a fit on
Eq. (9) with ki (i = 1, 3) the kinetic rate constants for reac-
tion steps 1 and 3.
ki
T
DH# DS#
RT
kB
h
ln ¼ ꢀ
þ
þ ln
R
ð9Þ
All kinetic mathematical fits were done utilizing the fit-
ting program MINSQ [27]. The error of all the data is pre-
sented according to crystallographic conventions. For
example k1obs = 0.0446(1) sꢀ1 implies k1obs = 0.0446
0.0001 sꢀ1
.

5078
J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
3. Results and discussion
6
35 °C
-13
5
3.1. The UV/Vis study, solvent and temperature dependence
-15
-17
4
3
2
1
As illustrated in Fig. 1, three general reaction steps for
the oxidative addition reaction between CH3I and
[Rh(fctfa)(CO)(PPh3)] could be identified by UV/Vis spec-
trophotometry. Due to the larger change in absorbance for
the second and third reaction step at 375 nm, more accu-
rate k3 and k4 rate constants were obtainable at this wave-
length. For reaction step 1, data obtained at 530 nm gave
3.2
3.3
3.4
3.5
103T-1/K-1
25 °C
15 °C
-10.4
-10.9
-11.4
35 °C
more accurate k1 and k rate constants.
ꢀ1
20
15
10
5
The first reaction showed a first-order dependence on
CH3I, while the second and third (not illustrated) reactions
are independent (i.e. zero order) of CH3I concentration,
Fig. 2. It is clear that the relationship between k1obs and
[CH3I] in chloroform does not pass through the origin.
This implies [26] that reaction step one is an equilibrium
process of the form
25 °C
15 °C
3.2
3.3
3.4
3.5
103T-1/K-1
k1
0
A þ B ꢀ C
0
0.5
1
1.5
2
k
ꢀ1
[CH3I]/mol dm-3
However, when the reaction was studied in methanol,
Fig. 2. CH3I concentration dependence for the UV/Vis monitored
oxidative addition of CH3I to [Rh(fctfa)(CO)(PPh3)] in chloroform at
530 nm for the first reaction step, bottom, and for the second reaction
step, top. Insets: The linear dependence between ln(k/T) and 1/T allowed
the determination of activation parameters.
acetone, ethyl acetate and benzene, k was found to be
ꢀ1
so small as to be undetected. The temperature dependence
in chloroform for the first and second stages of the oxida-
tive addition reaction is also demonstrated in Fig. 2.
The increase in the rate of the first step of oxidative
addition depends linearly on the dielectric constant, DE
[28], of each solvent, Fig. 3. Chloroform does not fit the
trend suggested by the other solvents because it follows a
chloroform
methanol
0.006
different mechanism (k is measurable).
ꢀ1
Rate constants, results of the solvent dependence study
and activation parameters which were obtained from the
temperature dependence study utilizing Eq. (9) are summa-
rized in Table 1.
0.004
acetone
0.002
2
375 nm
ethylacetate
1 st
375 nm
1
benzene
4
3
2
1
0
0
1st
2 nd
0
10
20
30
dielectric constant
530 nm
580 nm
400 nm
1
0
Fig. 3. Relationship between the second-order rate constant, k1, in units
of dm3 molꢀ1 sꢀ1, and the dielectric constant of the solvent during the UV
monitoring of the first reaction set of the oxidative addition of CH3I to
[Rh(fctfa)(CO)(PPh3)] at 25 ꢀC.
0
2 nd
reaction
0
0.2
t/hours
0.4
530 nm
580 nm
0
2
t/hours
4
Of special interest is the observation that DS# values for
3
rd reaction
all the reaction steps that are independent of [CH3I] (the
375 nm
k
and k3 steps), approaches zero within experimental
ꢀ1
400 nm
error, whereas DS# for the second order reaction step
related to k1 are large negative. This is consistent with an
interpretation that the transition state for the forward step
1 reaction occurs via an associative mechanism where the
coordination number changes from 4 to 6. For the other
reaction steps, the coordination number changes from 6
to 5 implying the transition state undergoes much less dra-
matic coordination changes. This results in DS# approach-
ing zero.
530 nm
60
0
20
40
t/hours
Fig. 1. Relative absorbance vs. time data at 25 ꢀC for the UV/Vis monitored
oxidative addition of CH3I to [Rh(fctfa)(CO)(PPh3)] in chloroform at the
indicated wavelengths for the first reaction step (insert top left), the second
reaction step (insert top right) and the third reaction step (main graph,
starting at approximately 5 h), [Rh(fctfa)(CO)(PPh3)] = 0.00015 mol dmꢀ3
[CH3I] = 0.1416 mol dmꢀ3
,
.

Table 1
Kinetic rate constants for the oxidative addition reaction of CH3I to [Rh(fctfa)(CO)(PPh3)] as obtained by various spectroscopic methods at 25 ꢀC by considering each reaction step in isolation during
data work-up with Eq. (5), as well as by a consecutive reaction treatment of data with Eqs. (7) and (8)
Method
UV/Vis
Solvent (DE),a
(k/nm)
Reaction sequence 1bDH# = 29(3),
Reaction sequence 1bDH# = 81(6),
Reaction sequence 2b,c DH# = 93(1),
RS 3b,c106 k4 (sꢀ1
)
DS# = ꢀ188(9) k1 (dm3 molꢀ1 sꢀ1
)
DS# = ꢀ40(20) k (sꢀ1
)
DS# = ꢀ5(3) k3 (sꢀ1
)
ꢀ1
CHCl3 (4.8) k = 580
k = 530
k = 400
0.0060(1)
0.0061(1)
0.0005(1)
0.0005(1)
0.00016(2)
0.00017(2)
0.00021(2)
0.00020(1)
4.4(1)
4.3(1)
3.7(1)
k = 375
C6H6 (2.27)
CH3COOEt (6.02)
(H3C)2CO (20.7)
CH3OH (32)
0.000141(2)
0.000716(4)
0.00370(4)d
0.00658(3)
ꢃ0
ꢃ0
ꢃ0
ꢃ0
IRc
CHCl3
0.0061(2)
0.0041(1)
0.0046(4)
0.008(2)
0.0006(2)
ꢃ0
0.00017(1)
0.00020(4)
0.00010(1)
0.00018(1)
0.00016(1)
4.7(4)
2.6(5)
(H3C)2CO, Eq. (5)e
(H3C)2CO, Eq. (7)e
(H3C)2CO, Eq. (8)e
(H3C)2CO, Eq. (7)e
ꢃ0
ꢃ0
5.5(3)
5(2)
1H NMR
31P NMR
CHCl3 (4.8)
CHCl3 (4.8)
0.006(1)
0.005(1)
0.00011(1)
0.00012(2)
Activation parameters in CHCl3 are summarized in the table heading. RS is the reaction sequence and Et is the ethyl.
a
DE is the dielectric constant at 25 ꢀC, except for CHCl3 where DE is at 20 ꢀC. The sited wave lengths were used to monitor the reaction by UV/Vis in CHCl3.
b
c
Units of DH# is kJ molꢀ1, that of DS# is J molꢀ1 Kꢀ1. For the extremely slow reaction sequence 3 (t1/2 < 35 h), no activation parameters were determined. Also, k and k ¼ 0 but very small.
ꢀ3
ꢀ4
Cited rate constants are the average from several reactions with different CH3I concentrations.
d
e
In acetone, DH# = 50(4) kJ molꢀ1, DS# = ꢀ124(4) J molꢀ1 Kꢀ1
.
IR data for each reaction set fitted to the indicated kinetic model Eqs. (5), (7) or (8), conditions for the consecutive reaction treatment of IR data were [CH3I] = 0.2604 mol dmꢀ3
,
[Rh(fctfa)(CO)(PPh3)] = 0.002 mol dmꢀ3, T = 25 ꢀC.

5080
J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
Results of the UV/Vis section only of this study are inter-
The half-life of the first reaction step under pseudo first-
order conditions with [CH3I] = 1.673 mol dmꢀ3 and at
25 ꢀC, Fig. 4, is 63 s. Since it is highly unlikely that the
RhIII-acyl1 species can be formed by fast isomerization of
an undetected RhIII-alkyl species with total different ligand
coordination pattern than the detected RhIII-alkyl1 species
at exactly the same rate as the detected RhIII-alkyl1 species,
this result is considered to imply that an equilibrium exist
between the RhIII-alkyl1 and RhIII-acyl1 species that is fast
enough to be maintained on the time scale of [Rh(fctfa)-
(CO)(PPh3)] disappearance. The same conclusion could
preted to imply that the first reaction step represents an oxi-
dative addition process whereby the RhI nucleus converts to
a RhIII nucleus. The product of this oxidative addition step
may be either an alkyl or acyl species, [Rh(fctfa)(CH3)-
(CO)(PPh3)(I)] or [Rh(fctfa)(COCH3)(PPh3)(I)] (or both),
but the UV/Vis technique cannot distinguish between these
possibilities. Since the second and third reaction steps are
independent of [CH3I], one may deduce that these two reac-
tions either involve slow solvent exchange with any of the
ligands of the newly formed RhIII complex or it may be
isomerization from an alkyl to an acyl product (or vice
1
be drawn using H NMR data and will be discussed under
versa). To obtain more insight on the nature of the RhIII
-
the NMR heading. This first set of reactions may, there-
fore, be presented by reaction sequence (10) where ki repre-
sents forward rate constants and kꢀi reverse rate constants,
i = 1 or 2.
products, an infrared study of the reaction was conducted
as this technique is ideal to distinguish between metal–CO
bonds that vibrates at ꢃ1900–2100 cmꢀ1, and metal–
COCH3 bonds that vibrates at ꢃ1600–1800 cmꢀ1
3.2. Infrared study
.
ð10Þ
3.2.1. Chloroform as solvent
Three reaction steps could also be identified by IR spec-
troscopy, Fig. 4. In the first reaction step, the rate of disap-
pearance of the RhI species [RhI(fctfa)(CO)(PPh3)] in
chloroform at 25 ꢀC, as monitored by the disappearance
of the CO peak at 1990 cmꢀ1 corresponds exactly with
The k step in Eq. (10) was only detected in CHCl3
ꢀ1
solutions. The equilibrium constant between the RhIII
-
.
alkyl1 and RhIII-acyl1 species is defined as K2 = k2/k
ꢀ2
Pseudo first-order rate constants k1obs for reaction 10 uti-
the rate of formation of both a RhIII-alkyl1 species, [RhIII
-
lizing nine different CH3I concentrations between 0.12
and 1.70 mol dmꢀ3 were determined. A graph (not shown)
of k1obs versus [CH3I] in chloroform solution, was linear
with y-intercept 0.0006(2) sꢀ1. This satisfies Eq. (6) with
(fctfa)(CH3)(CO)(PPh3)(I)], monitored by the appearance
of the CO-peak at 2082 cmꢀ1, and a RhIII-acyl1 species,
[RhIII(fctfa)(COCH3)(PPh3)(I)], monitored at 1729 cmꢀ1
.
3rd reaction
Rh(III)-
alkyl2
2nd reaction
1
Rh(III)-
acyl2
Rh(III)-
alkyl2
1st reaction
1.0
Rh(I)
Rh(III)-
alkyl1
Rh(III)-
alkyl1
0
2064
1714
Rh(III)-
acyl1
0.4
0
2082 2064
1729
Rh(III)-
acyl1
CH3I
0
2082
1990
1729
wave number/cm-1
Fig. 4. The reaction between CH3I (1.673 mol dmꢀ3) and [Rh(fctfa)(CO)(PPh3)] (0.0100 mol dmꢀ3) in chloroform at 25 ꢀC as monitored by IR
spectroscopy. The first set of reactions (bottom spectrum, spectrum times are t = 0 (. . .. . ., not shown for RhI due to high amplitude), 74, 149, 224 and
299 s) shows the disappearance of Rh(I) at 1990 cmꢀ1 (k1obs = 0.011 sꢀ1) and the simultaneous appearance at the same rate of a RhIII-alkyl1, at
2082 cmꢀ1, and a RhIII-acyl1 species at 1729 cmꢀ1. The middle insert illustrates the simultaneous disappearance of RhIII-alkyl1 (t = 554, 734, 1094, 1919,
2819, 3719, 4905, 7605, 10305 and 20205 s) and RhIII-acyl1 (t = 554, 1919, 3719, 7605 and 20205 s) and the formation of a RhIII-alkyl2 species at
2064 cmꢀ1 (k3 = 0.00017(3) sꢀ1 for all three cases) during the second set of reactions. The third set of reactions (top right inset) is illustrated by the
disappearance of the RhIII-alkyl2 species (spectra were recorded at t = 40408, 121558, 248726 and 335308 s) and the simultaneous formation of a new
RhIII-acyl2 species at 1714 cmꢀ1 (k4 = 4.3(2) · 10ꢀ6 sꢀ1).

J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
5081
k1 the second order forward rate constant and kꢀ1 the first-
where the superscripts ‘‘1’’ and ‘‘2’’ refer to RhIII-alkyl1,
RhIII-acyl1, and RhIII-alkyl2 or RhIII-acyl2, respectively.
As was the case with the first set of reactions, k3 and k4 ob-
tained by the IR method are consistent with the results
from the UV/Vis method.
order reverse rate constant for reaction (10). Individual val-
ues for k1 and k are tabulated in Table 1. The mutual
ꢀ1
consistency of k1 and k values as determined by the
ꢀ1
UV/Vis and IR methods are obvious.
A second, much slower (t1/2 ꢄ 1.1 h ꢄ 4000 s) set of
reactions that are CH3I independent is illustrated in the
middle insert in Fig. 4, where the RhIII-acyl1 species at
1729 cmꢀ1 and the RhIII-alkyl1 species at 2082 cmꢀ1 both
disappear at the same rate (k2 = 0.00017(1) sꢀ1) as the for-
3.2.2. Consecutive reaction treatment in acetone as solvent
In chloroform, the kobs versus [CH3I] graph for the first
step did not pass through the origin for both the UV/Vis
and IR detection methods, and this is indicative of a notice-
able equilibrium in the first reaction step of the total reac-
tion. However, IR results in acetone indicate, in agreement
with the UV/Vis results in acetone (Table 1), that this equi-
mation of
(k3 = 0.00017(3) sꢀ1) at 2064 cmꢀ1. Again, this is regarded
as additional proof that the equilibrium between the RhIII
a
new RhIII-alkyl2 geometrical isomer
-
acyl1 species and the RhIII-alkyl1 species is fast. The
[CH3I]-independent second reaction step is 0.011/
0.00017 ꢄ 65 times slower (from rate constants from
Fig. 4) than the [CH3I]-dependent first reaction at
[CH3I] = 1.673 mol dmꢀ3. Kinetic data of the second reac-
tion is consistent with the equilibrium
librium was shifted so far to the right that the k step is
ꢀ1
not observable. The apparent absence of a kꢀ1 step in ace-
tone allowed us to treat results from an IR study in acetone
utilizing consecutive reaction kinetic models.
The IR observed reaction sequence of this reaction in
acetone in the range 1960–2100 cmꢀ1 was basically the
same as observed in chloroform. However, the formation
of the RhIII-acyl complexes could not be measured due to
the strong IR absorbance of the solvent, acetone, in the
1650–1800 cmꢀ1 region. The RhI peak was monitored at
1987 cmꢀ1 and the RhIII-alkyl1 and RhIII-alkyl2 species
at 2082 and 2059 cmꢀ1, respectively. By treating data for
every reaction sequence in isolation utilizing Eq. (5), as
was done in chloroform, rate constants was determined
for all the reaction steps and corresponded to those of
the UV/Vis study. Rate constants are listed in Table 1.
Because the equilibrium K2 was found to be fast enough
to be maintained during Rh(I) depletion, and since CH3I
was in large excess over Rh(I), it was possible to determine
the rate constants k1, k3 and k4 for the studied reaction in
acetone by using the consecutive reaction treatment on
k3
A ꢀ B
ð11Þ
k
ꢀ3
where A = [RhIII-alkyl1ꢀRhIII-acyl1] and B = RhIII
-
alkyl2. As discussed later, kꢀ3 n k3. The structure of the
RhIII-alkyl2 geometrical isomer having the molecular for-
mula [RhIII(fctfa)(CH3)(CO)(PPh3)(I)] has also been solved
and is reported later in this publication.
The second set of reactions was followed by a very slow
(t1/2 ꢄ 40 h) [CH3I]-independent third reaction and corre-
sponds to the slow first-order disappearance of the RhIII
-
alkyl2 species (k4 = 4.9(5) · 10ꢀ6 sꢀ1) at 2064 cmꢀ1 and
the appearance, at the same rate, of a new RhIII-acyl2 geo-
metrical isomer at 1714 cmꢀ1 (k4 = 4.4(2) · 10ꢀ6 sꢀ1
,
Fig. 4, top right inset). The long half-life of the third reac-
tion implied that it could not be followed with great accu-
racy, as solvent evaporation became difficult to control.
The kinetic data of the third reaction is consistent with
the general reaction sequence
k1
k3
k4
the general reaction sequence, A ! B ! C ! D (A, B and
C was defined in Sections 3.1 and 3.2.1, D = RhIII-acyl2).
By ignoring the very slow k4 step (i.e. D) in the reaction
sequence and upon performing a least-squares fit of the
available IR absorbance vs. time data of the RhIII-alkyl1
peak for the reaction in acetone to Eq. (7), see graph shown
in Fig. 5, top, the rate constants were determined as
k1obs = 0.0012(1) sꢀ1 (k1 = 0.0046 dm3 molꢀ1 sꢀ1), and
k3 = 0.00010(1) sꢀ1. The value of k1obs corresponds very
close with those obtained from data treatment utilizing
Eq. (5) (values summarized in Table 1). The value of k3
was slightly (37%) smaller than the value of 0.00016(2) sꢀ1
which was determined by treating reaction 2 in isolation.
In addition, fitting of the available absorbance/time
data of the RhIII-alkyl2 peak to Eq. (8), see Fig. 5, bot-
tom, resulted in k1obs = 0.0022(6) sꢀ1 (k1 ꢄ 0.0084 dm3
molꢀ1 sꢀ1), while k3 = 0.00018(1) sꢀ1. This value of k3
corresponds well to the value obtained by treating data
sets as separate, isolated quantities for each reaction
according to Eq. (5). Utilizing Eq. (8), k1 was 83% (1.83
times) larger than the value obtained from Eq. (5) but it
is still kinetically equivalent. This drift in k1 and k3 was
accentuated because of the difficulty of obtaining large
k4
RhIII-acyl2 ! RhIII-alkyl2
ð12Þ
Utilizing the combined information obtained by the
UV/Vis and IR study only, the overall reaction sequence
for this oxidative addition reaction may therefore be repre-
sented as
ð13Þ

5082
J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
2.0
1.5
1.0
0.5
0.0
All things considered, the calculated rate constants
obtained by the consecutive reaction treatment (Eqs. (7)
and (8)) and as individual data sets utilizing Eq. (5) are
regarded as mutually consistent and confirm the proposed
mechanism. It is, however, important to note that consis-
tency between the consecutive reaction treatment of data
and treatment of kinetic data of each reaction step in isola-
tion would not have been observed if reaction step 1 was not
much faster than reaction step 2 and if reaction step 3 was
not much slower than reaction step 2. If the rates did not
differ as much as demonstrated by the half lives cited in Sec-
tion 3.2.1, reaction steps 1, 2 and 3 would be overlapping so
much that accurate rate constants would not be obtainable
by regarding each reaction step in isolation.
When the consecutive reaction treatment of data was
applied to reactions performed in chloroform (data for
Section 3.2.1), the mathematical models (Eqs. (7) and (8))
were not found to be appropriate as data could not be sat-
isfactorily fitted to it. The breakdown of the model is
0.4
0.2
0.0
attributed to the detected k step in the mechanism
ꢀ1
k1
k3
A ꢀ B ! C
k
ꢀ1
The mathematical treatment of such monophasic revers-
ible reactions is quite complex [29].
0
20
40
time/min
1
3.3. H and 31P NMR study
The reaction between CH3I and [Rh(fctfa)(CO)(PPh3)]
was also monitored by 1H and 31P NMR in order to obtain
additional insight into this reaction. By carefully comparing
the positions and integrals of the different signals, the spec-
tral parameters of different isomers could be identified,
Table 2, see also Ref. [19]. The same reaction sequence as
100
200
300
400
500
0
time/min
Fig. 5. Relative IR absorbance vs. time data of the reaction between CH3I
(0.2604 mol dmꢀ3) and [Rh(fctfa)(CO)(PPh3)] (0.002 mol dmꢀ3) in ace-
tone at 25 ꢀC. Top: Kinetic data were fitted to the consecutive reaction
model Eq. (7), relative absorbancies of the RhIII-alkyl1 species are
illustrated. Bottom: A data fit of RhIII-alkyl2 IR absorbances to Eq. (8) is
shown. Inset: An enlargement of the initial stages of RhIII-alkyl2
formation. The slow initial formation of the RhIII-alkyl2 species in the
region t < 10 min is apparent.
1
observed by IR and UV/Vis, was observed utilizing H
NMR, Fig. 6 and 31P NMR, Fig. 7. The new feature intro-
duced by the NMR study is the existence of more than one
geometrical isomer for each intermediate. The two main iso-
mers of each intermediate will be referred to as A and B, e.g.
RhIII-alkyl1A and RhIII-alkyl1B, the choice of the labels are
arbitrary and have no significance. The A and B isomers of
each species exist in a fast equilibrium with each other,
because the observed rate constant for the disappearance
or formation of an A and a B isomer of the same species
was found to be within experimental error the same and also
because the ratio Rh(I)A/Rh(I)B = 1.5 (obtained by inte-
amounts of data points for relatively fast reactions
quickly enough with overlay spectra utilizing an infrared
spectrophotometer. It led to too little available data to
fit complicated equations.
By ignoring the fast k1 step (i.e. A) in A ! B ! C ! D
and by applying the consecutive reaction kinetic model,
Eq. (7), to intermediate C for the compound series B,
k1
k3
k4
gral evaluation) is not the same as the ratio RhIII
-
C
and D, values of k3 and k4 were found to
1.6(1) · 10ꢀ4 and 5.5(3) · 10ꢀ6 sꢀ1. The value for k3 cor-
responded well to the previously determined values. The
value for k4 is larger by a factor 2 than the value
obtained from the treatment of data in isolation, which
alkyl1A/RhIII-alkyl1B = 0.39 or the ratio RhIII-acyl1A/
RhIII-acyl1B = 1.0 (Table 2). If the reaction was (RhIA +
CH3I ꢀ {RhIII-alkyl1A ꢀ RhIII-acyl1A}) for the A iso-
mers and separately (RhIB + CH3I ꢀ {RhIII-alkyl1B ꢀ
RhIII-acyl1B}) for the B isomers, and if no equilibrium
between isomers A and B existed, the ratio A:B of these iso-
mers would have been the same throughout the entire reac-
tion sequence involving steps 1, 2 and 3.
gave k4 = 2.6(5) · 10ꢀ6 sꢀ1
.
This twofold difference
between the k4 values are again not regarded as signifi-
cant. Eq. (8) could not be fitted to formation data of
D, from the reaction in acetone as solvent, due to inter-
ference of the acetone carbonyl peak in the region of
RhIII-acyl2 detection.
Rate constants obtained from 1H NMR are summarized
in Table 1. It can be seen good agreement exist between the

J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
5083
kinetic rate constants obtained by utilizing UV/Vis, IR, 1H
NMR and 31P NMR spectroscopic methods.
The structure of one of the isomers of the [RhI(fctfa)-
(CO)(PPh3)] is reported elsewhere [20], while the structure
of the [RhIII(fctfa)(CH3)(CO)(PPh3)(I)]2B species is
reported in this communication. The structure of the
other RhIII acyl compounds is as yet unknown and the
indicated formulas relate to molecular stoichiometry only
and not to a specific geometry. The 1H and 31P NMR
spectrum of the crystalline isolated RhIII-alkyl2B isomer
were identical to the spectrum of isomer RhIII-alkyl2B
as observed during the monitoring of the second set of
reactions during the oxidative addition reaction of
[Rh(fctfa)(CO)(PPh3)] with CH3I in CDCl3 solutions.
An important observation was that upon dissolving the
solid crystalline RhIII-alkyl2B isomer in CDCl3, isomeri-
zation to the RhIII-alkyl2A isomer sets in. After 4 days,
traces of Rh(I), RhIII-alkyl1 and RhIII-acyl1 were also
observed on 1H and 31P NMR. This means that Eq.
(13) also has a k and k step, but these rate constants
ꢀ3
ꢀ4
are so small that the k and k steps are simply not
ꢀ3
ꢀ4
kinetically determinable. Taking into account there exist
two main isomers of each reactant and reaction product,
the complete reaction sequence for the oxidative addition
of CH3I to [Rh(fctfa)(CO)(PPh3)] is therefore
First set
of reactions
ð14Þ
Second set
of reactions
Third set of reactions
The above mechanism is a completely general mecha-
nism. As such it caters for all previously encountered spe-
cial cases sighted in Eqs. (1)–(4). Each one of these four
equations can be generated by careful manipulation of
the size of the individual rate constants. For example,
Eq. (14) will simplify to Eq. (3) provided kꢀ2 = k
=
ꢀ3
k4 = kꢀ4 = 0 sꢀ1. The equilibrium constants for the equilib-
rium between the isomers, defined as Kci = [RhB-species]/
1
[RhA-species], i = 1,2,. . .5, were determined from the H
NMR integration data of suitable identifiable and well
resolved peaks at 25 ꢀC in CDCl3 and are documented in
Table 2.

5084
J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
First set of reactions
Third set of reactions
RhIA
RhIII-alkyl1A
and 1B
RhIB
RhIII-acyl1A
and 1B
PPM
PPM
PPM
6.05
PPM
PPM
6.10
3.05
3.00
1.50
1.45
3.05
3.00
2.95
2.90 1.80
1.75
Second set of reactions
PPM
PPM
PPM
PPM
PPM
6.05
5.45
5.40
5.35
5.30
3.05
3.00
1.80
1.75
1.45
Fig. 6. Fragments of the 1H NMR spectra in CDCl3 illustrating reaction sequences during the oxidative addition and the ensuing carbonyl insertion and
deinsertion reactions of 0.3674 mol dmꢀ3 CH3I to 0.0183 mol dmꢀ3 [Rh(fctfa)(CO)(PPh3)] in CDCl3 at T = 25 ꢀC. Top left: Selected spectra for the first set
of reactions illustrating changes in the signal of the methine proton of the b-diketonato ligand of the indicated isomers in the region 6.04–6.08 ppm, the
growing signals of the CH3-group of indicated RhIII-acyl isomers at 3.05 and 2.98 ppm and of RhIII-alkyl isomers at 1.47 and 1.42 ppm. Bottom: Signal
intensity changes of the second reaction sequence showing the methine proton of the b-diketonato ligand of the indicated isomers at 5.30–6.08 ppm, the
CH3 signal disappearance for the RhIII-acyl isomers at 3.05 and 2.98 ppm and RhIII-alkyl isomers at 1.47 and 1.42 ppm, and the increase in signal intensity
of the CH3-group of the overlapping alkyl 2 isomers at 1.78 ppm. Top right: The third set of reactions is illustrated by changes in signal intensity of the
CH3-group of the RhIII-alkyl2 isomers at 1.78 ppm, and of the RhIII-acyl2A and 2B species at 3.01 and 2.92 ppm. The multiplicity of the signal of the CH3
group of the RhIII-alkyl1 and 2 isomers is due to coupling with Rh (spin 1/2) and P (spin 1/2). Supplementary data shows further spectra highlighting
NMR changes.
Rh -alkyl1B
Rh -alkyl1A
with other bidentate ligands [15,30] is that the iodo ligand,
I, is bonded in the axial position of the octahedron, above
or below the equatorial plane of the octahedron formed by
Rh -alkyl2B
O(1), O(2), C(5)O and C(6)H3. The Rh–I bond axis devi-
Rh A and Rh B
Overlapping
ated 3.6(1)ꢀ towards O(2) from the expected 90ꢀ and the
Rh -alkyl2B
Rh–P bond deviates by more than 2ꢀ away from O(1)
and O(2).
Rh -alkyl2A
This resulted in a 2.26(5)ꢀ deviation from linearity for
the I–Rh–P bond axis. A similar deviation from linearity
PPM
50
40
30
20
(3.4ꢀ) for the I–Rh–P bond axis was found for [Rh(cupf)-
(CO)(PPh3)(CH3)(I)] [15]. The RhIII–I bond length of the
˚
˚
PPM
50
40
30
20
present study, 2.716(1) A, is shorter (ꢃ0.1 A) than Rh–I
bonds in complexes having the CH3 group trans to the
iodine atom [30] and this is the result of the smaller trans
influence of PPh3 compared to that of an alkyl group.
The various ligand–rhodium–ligand bond angles in the
equatorial plane formed by O(1), O(2), C(6) and C(5),
viz. C(5)–Rh–C(6) = 88.4(4)ꢀ, C(5)–Rh–O(2) = 97.3(3)ꢀ,
O(2)–Rh–O(1) = 88.0(2)ꢀ and O(1)–Rh–C(6) = 86.1(3)ꢀ,
deviate 2–7ꢀ from 90ꢀ. The Rh–C(5)–O(3) bond angle,
175.7(8)ꢀ, is not completely linear, with the C(5)–O(3) bond
50
40
30
20
Fig. 7. Selected 31P NMR spectra illustrating doublet 31P peaks of the
indicated reactants and products during the three sets of reactions for the
oxidative addition of 0.500 mol dmꢀ3 CH3I to 0.021 mol dmꢀ3
[Rh(fctfa)(PPh3)(CO)] in CDCl3 at 25 ꢀC. Bottom left: at time t = 0
(reagent). Middle top: t = 1070 s (towards the end of the first set of
reactions). Bottom right: t = 69520 s (near the end of the second set of
reactions).
3.4. The crystal structure data of
[Rh(fctfa)(CO)(PPh3)(CH3)(I)]
˚
length 1.12(1) A, significantly shorter than the C(5)–O(3)
bond distance of 1.147(5) A in [Rh (fctfa)(CO)(PPh3)] [20].
I
˚
The numbering system of the atoms in [Rh(fctfa)-
(CO)(PPh3)(CH3)(I)] is shown in Fig. 8 while the crystal
data are summarized in Table 3. Selected bond lengths
and bond angles are given in Table 4. An obvious feature
of [Rh(fctfa)(CO)(PPh3)(CH3)(I)] and similar known
RhIII-alkyl complexes in which the fctfa ligand is replaced
In accordance with the p bonding concept [31] the
˚
Rh–C(5) bond length of 1.830(8) A is larger than the
I
˚
corresponding bond length, 1.801(5) A, in [Rh (fctfa)-
(CO)(PPh3)] because decreasing electron density on the
rhodium center in changing the oxidation state from I to
III, imply less electron donating capability from the metal

J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
5085
Table 4
˚
Selected bond lengths (A) and angles (ꢀ) for [Rh(fctfa)(CO)(PPh3)-
(CH3)(I)]
Atoms
Bond
Atoms
Bond
Rh–C(5)
Rh–O(1)
Rh–C(6)
Rh–O(2)
Rh–P
1.830(8)
2.075(5)
2.076(8)
2.160(5)
2.320(2)
2.716(1)
1.822(7)
1.12(1)
F(1)–C(4)
C(1)–C(2)
C(2)–C(3)
C(3)–C(4)
C(1)–C(11)
C(11)–C(12)
C(11)–C(15)
C(12)–C(13)
C(21)–C(22)
C(41)–C(42)
1.39(3)
1.40(1)
1.38(1)
1.51(1)
1.47(1)
1.41(1)
1.45(1)
1.40(1)
1.39(2)
1.40(1)
Rh–I
P–C(51)
O(3)–C(5)
O(2)–C(3)
F(2)–C(4)
1.271(9)
1.40(3)
Fig. 8. A perspective view of [Rh(fctfa)(CO)(PPh3)(CH3)(I)] showing
atom labelling. Thermal ellipsoids were drawn at a 30% probability level.
The view from above down the ferrocenyl axis (top left) shows the
cyclopentadienyl ring deviates 12.7ꢀ from a fully eclipsed conformation.
Right: The cyclopentadienyl rings are planar and almost parallel, they
deviate 3.1(5)ꢀ from co-planarity with the plane through the Rh chelate.
Angles
Atoms
Angles
C(5)–Rh–O(1)
O(1)–Rh–O(2)
O(2)–Rh–P
173.8(3)
88.0(2)
92.9(2)
89.7(1)
114.2(7)
111.9(7)
119.6(9)
O(1)–C(1)–C(2)
O(2)–C(3)–C(2)
C(2)–C(3)–C(4)
C(3)–C(2)–C(1)
Rh–C(5)–O(3)
P–Rh–I
126.5(6)
129.3(7)
118.9(7)
125.3(7)
175.7(8)
177.74(5)
O(1)–Rh–I
O(1)–C(1)–C(11)
O(2)–C(3)–C(4)
C(34)–C(33)–C(32)
d-orbital into the p* orbital of the carbonyl ligand. This
leads to a longer (weaker) Rh–C bond length and a stron-
˚
The standard deviation of the last decimal is given in brackets.
ger C–O bond. The Rh–O(1) bond is 0.085 A shorter than
the Rh–O(2) bond which indicates the structure is not sym-
metric around the b-diketonato ligand. However, both Rh–
O(1) and Rh–O(2) bonds in [RhIII(fctfa)(CO)(PPh3)-
atoms of the b-diketonato skeleton were found, Table 4.
The largest deviations were at the C atoms bound to
O(1) and O(2). In the case of the free Hfctfa [32] the corre-
sponding bond angles were all within experimental error
120ꢀ.
(CH3)(I)] are longer than the corresponding bonds in
I
˚
˚
[Rh (fctfa)(CO)(PPh3)] [20], 2.048(3) A and 2.070(3) A.
The Rh–O(1) bond in both the Rh(I) and Rh(III) com-
plexes is trans to the CO group, but Rh–O(2) is trans to
the PPh3 group in [RhI(fctfa)(CO)(PPh3)] and trans to the
CH3 group in [RhIII(fctfa)(CO)(PPh3)(CH3)(I)]. The
C(3)–O(2) and C(1)–O(1) bonds in the b-diketonato skele-
ton of the [Rh(b-diketonato)(CO)(PPh3)(CH3)(I)] complex
The cyclopentadienyl rings are within experimental
error, planar and almost parallel [dihedral angle 1.4(6)ꢀ]
while the average angle of deviation from a fully eclipsed
configuration is 12.7ꢀ, Fig. 8. The cyclopentadienyl ring
which is bonded to the b-diketonato chelate ring at C(1)
deviates 3.1(5)ꢀ from co-planarity with the plane defined
by atoms O(1), O(2), C(1), C(2), C(3) and Rh, i.e. the plane
through the chelate ring.
˚
˚
are similar, 1.271(9) A and 1.265(8) A, respectively, and
both bonds are smaller than the corresponding bonds in
free Hfctfa [32]. The change in oxidation state from
Rh(I) to Rh(III) did not influence the C(3)–O(2) and
C(1)–O(1) bond lengths. The C(1)–C(2) and C(2)–C(3)
bonds of the present Rh(III) complex are similar, contrary
to what was found for both [RhI(fctfa)(CO)(PPh3)] (corre-
˚
The average C–C bond distance, 1.38(1) A, in the phenyl
rings is within experimental error near the normal value
˚
(1.394 A) for the aromatic C–C bond [33]. The bond angles
around P differ no more than 7ꢀ from 109ꢀ280, the angle for
˚
sponding bonds were 1.413(5) and 1.368(6) A), or the free
˚
a regular tetrahedron. The mean P–C distance is 1.82(1) A,
Hfctfa ligand where asymmetric enolization was observed
˚
˚
and almost the same [20] as in the parent rhodium(I) com-
(C(1)–C(2) = 1.432(6) A, C(2)–C(3) = 1.345(6) A). As was
the case for [Rh(fctfa)(CO)(PPh3)], deviations up to 10ꢀ
from the expected 120ꢀ for sp2 hybridization for the C
I
˚
plex [Rh (fctfa)(CO)(PPh3)], where it was 1.824 A. The
RhIII–P bond distance for the present complex of
Table 3
Crystal data and structure refinement for [Rh(fctfa)(CO)(PPh3)(CH3)(I)]
Empirical formula
Formula weight
Crystal system
Space group
C
34H28F3FeO3RhPI
Z, density (calculated) (g/cm3)
4, 1.716
293(2)
0.71073
1.961
858.2219
Monoclinic
P21/c
Temperature (K)
˚
Wavelength (A)
Absorption coefficient (mmꢀ1
)
Unit cell dimensions
˚
a (A)
10.769(2)
15.153(3)
20.413(5)
94.31(2)
3221(1)
Reflections collected
Unique reflections
Refinement method
Data [I > 2r(I)]; parameters
Final R indices [I > 2r(I)]
4717
˚
b (A)
4420 [Rint = 0.0588]
Full-matrix least-squares on F2
4414; 429
˚
c (A)
b (ꢀ)
3
˚
Volume (A )
R = 0.0588, Rw = 0.1692

5086
J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
˚
2.320(2) A is, however, much longer than the Rh–P bond
posed mechanisms for a variety of such reactions are
special cases of the general mechanism that emanated from
this study. The structure of the intermediate RhIII-alkyl2B
geometric isomer, [Rh(fctfa)(CO)(PPh3)(CH3)(I)], was
solved. It was asymmetric in nature and the phosphine
and iodo ligands were found to be the axial ligands.
I
˚
distance for [Rh (fctfa)(CO)(PPh3)], 2.232(1) A, indicating
a change in oxidation state from Rh(I) to Rh(III) results
in a weaker Rh–P bond. The RhIII–P bond length for the
[Rh(fctfa)(CO)(PPh3)(CH3)(I)] complex is, however, in
the same order as was found for a variety of other RhIII
-
alkyl complexes including [Rh(cupf)(CO)(PPh3)(CH3)(I)]
˚
[15c], 2.327(4) A, and [Rh(ox)(CO)(PPh3)(CH3)I] [30b],
Acknowledgements
˚
2.321(2) A, (average of two molecules found in one unit
cell, Hcupf = N-hydroxy-N-nitroso-benzeneamine).
Financial assistance by the South African NRF under
Grant 2054243 and the Central Research Fund of the Uni-
versity of the Free State is gratefully acknowledged.
The trifluoromethyl group in [Rh(fctfa)(CO)(PPh3)-
(CH3)(I)] is disordered, similar to what was found for the
free b-diketone Hfctfa [32]. An intermolecular interaction
was observed between F and either a hydrogen on a ferr-
ocenyl group, or a hydrogen on a phenyl ring from another
Appendix A. Supplementary material
0
molecule (indicated by H0): F(3)–H (53) = 2.661 A,
CCDC 646438 contains the supplementary crystallo-
graphic data for [Rh(fctfa)(CO)(PPh3)(CH3)(I)]. These
data can be obtained free of charge via http://www.ccdc.
cam.ac.uk/conts/retrieving.html, or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge
CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail: depos-
it@ccdc.cam.ac.uk. Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.poly.2007.07.004.
˚
0
0
˚
˚
F(3)–H (12) = 2.669 A, F(12)–H (53) = 2.652 A, F(12)–
0
0
˚
˚
H (13) = 2.667 A, F(13)–H (36) = 2.606 A. These intermo-
lecular interactions link one molecule to another in the unit
cell and it is partially the cause of the CF3 disorder. How-
ever, the CF3 disorder is also the result of intramolecular
interaction that prevents rotation of the CF3 group. Two
preferred CF3 orientations are observed. The first involves
a strong intramolecular interaction between F(2) and H(2)
(the hydrogen bonded to C(2)) and a weak interaction
between both F(1) and F(3) with O(2). The contact dis-
References
˚
˚
tances are F(2)–H(2) = 2.362 A, F(1)–O(2) = 2.702 A and
F(3)–O(2) = 2.718 A, respectively. The other CF3 orienta-
˚
[1] S.E. Sherman, S.J. Lippard, Chem. Rev. 87 (1987) 1153.
[2] W.A. Herrmann, Kontakte 1 (1991) 22.
[3] W.A. Herrmann, Kontakte 3 (1991) 29.
[4] C. Mercier, P. Chabardes, Pure Appl. Chem. 66 (1994) 1509.
[5] (a) J. Adams, D.M. Spero, Tetrahedron 47 (1991) 1765;
(b) A.J. Anciaux, A. Demonceau, A.F. Noels, R. Warin, A.J. Hubert,
tion, involving F(12), F(11) and F(13), is the result of a
strong interaction between F(12) and O(2) and a weak
interaction of both F(11) and F(13) with H(2). The intra-
molecular
contact
distances
are
F(12)–O(2) =
˚
˚
˚
´
P. Teyssie, Tetrahedron 24 (1983) 2169.
2.485 A, F(11)–H(2) = 2.727 A and F(13)–H(2) = 2.624 A,
respectively.
[6] J.M. O’ Connor, J. Ma, Tetrahedron 57 (1992) 5075.
[7] (a) K. Burgess, M.J. Ohlmeyer, Chem. Rev. 91 (1991) 1179;
(b) D.A. Evans, G.C. Fu, A.H. Hoveyda, J. Am. Chem. Soc. 114
(1992) 6671.
4. Conclusion
[8] (a) N. Sakai, S. Mano, K. Nozaki, H. Takaya, J. Am. Chem. Soc. 115
(1993) 7033;
(b) G.D. Cuny, S.L. Buchwald, J. Am. Chem. Soc. 115 (1993)
2066.
[9] (a) M.J. Burk, Y.M. Wang, J.R. Lee, J. Am. Chem. Soc. 118 (1996)
5142;
A three-step reaction sequence was found for the oxida-
tive addition of iodomethane to [Rh(fctfa)(CO)(PPh3)], the
final product of the reaction was found to be a RhIII-acyl
species. UV/Vis studies alone could only provide reaction
rates and the amount of reaction steps. Utilizing IR spec-
troscopic techniques, alkyl and acyl RhIII species could
be identified, but only NMR spectroscopy revealed that
each reaction product exists in solution as two main iso-
mers in a fast equilibrium with each other. This study is
the first to observe a fast equilibrium between RhIII-alkyl
and RhIII-acyl intermediates (equilibrium K2) and it was
the first to determine the actual rate of conversion from a
RhIII-alkyl2 species to a RhIII-acyl2 species. The incorpora-
tion of a ferrocenyl group in the b-diketonato ligand
slightly accelerated the rate of CH3I oxidative addition
when compared with the rate of oxidative addition to either
the corrꢀesponding CH3COCHCOCF3ꢀ [15a] or PhCOCH-
COCF3 complex [34]. A completely general mechanism
for the oxidative addition of CH3I to [Rh(b-diketona-
to)(CO)(PPh3)] could be derived and all previously pro-
(b) D.A. Evans, G.C. Fu, Chem. Rev. 93 (1993) 1307;
(c) W.R. Cullen, S.J. Rettig, F.B. Wickenheiser, J. Mol. Catal. 66
(1991) 251.
[10] W.R. Cullen, F.B. Wickenheiser, J. Organomet. Chem. 370 (1989)
141.
[11] J.H. Jones, Platinum Metals Rev. 44 (2000) 94.
[12] P.M. Maitlis, A. Haynes, G. J Sunley, M.J. Howard, J. Chem. Soc.,
Dalton Trans. (1996) 2187.
[13] J.D. Atwood, Coord. Chem. Rev. 83 (1988) 93.
[14] (a) A. Haynes, B.E. Mann, G.E. Morris, P.M. Maitlis, J. Am. Chem.
Soc. 115 (1993) 4093;
(b) A. Haynes, P.M. Maitlis, G.E. Morris, G.J. Sunley, H. Adams,
P.W. Badger, C.M. Bowers, D.B. Cook, P.I.P. Elliot, T. Ghaffer, 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) 2847.
[15] (a) S.S. Basson, J.G. Leipoldt, J.T. Nel, Inorg. Chim. Acta 84 (1984)
167;
(b) S.S. Basson, J.G. Leipoldt, A. Roodt, J.A. Venter, T.J. van der
Walt, Inorg. Chim. Acta 119 (1986) 35;

J. Conradie et al. / Polyhedron 26 (2007) 5075–5087
5087
(c) S.S. Basson, J.G. Leipoldt, A. Roodt, J.A. Venter, Inorg. Chim.
Acta 128 (1987) 31;
(d) J.G. Leipoldt, S.S. Basson, L.J. Botha, Inorg. Chim. Acta 168
(1990) 215;
(e) J.A. Venter, J.G. Leipoldt, R. van Eldik, Inorg. Chem. 30 (1991)
2207.
[24] D.T. Cromer, J. Mann, Acta. Crystallogr., Sect. A 24 (1968) 321.
[25] A.J.C. Wilson (Ed.), International Tables for X-ray Crystallography,
vol. C, Kluver Academic Publishers, Dortrecht, 1992.
[26] J.H. Espenson, Chemical Kinetics and Reaction Mechanisms, 2nd
ed., McGraw-Hill, New York, 1995, pp. 15, 49, 70–75, 156.
[27] MINSQ, Least Squares Parameter Estimation, Version 3.12, Micro-
Math, 1990.
[28] R.C. Weast, Handbook of Chemistry and Physics, 49th ed., The
Chemical Rubber Co., Ohio, 1969, pp. E58–E59.
[29] R.G. Wilkins, Kinetics and Mechanism of Reactions of Transition
Metal Complexes, 2nd thoroughly revised ed., VCH, Weinheim, 1991,
p. 15.
[30] (a) J.P. Collman, P.A. Christian, S. Current, P. Denisevich, T.R.
Halbert, E.R. Schmittou, K.O. Hodgson, Inorg. Chem. 15 (1976) 223;
(b) K.G. van Aswegen, J.G. Leipoldt, I.M. Potgieter, G.J. Lamprecht,
A. Roodt, G.J. van Zyl, Trans. Met. Chem. 16 (1991) 369.
[31] F.A. Cotton, G. Wilkinson, Basic Inorganic Chemistry, John Wiley &
Sons, New York, 1976, pp. 473–480.
[16] D.M.C. Smit, Synthesis and kinetic study of rhodium(I) complexes
containing organic tripod ligands (in Afrikaans), Ph.D. thesis,
University of the Orange Free State, R.S.A., 1995.
[17] (a) G.J.J. Steyn, A. Roodt, J.G. Leipoldt, Inorg. Chem. 31 (1992)
3477;
(b) G.J.J. Steyn, A. Roodt, J.G. Leipoldt, Rhodium Ex. 1 (1993) 25.
[18] B.S. Furniss, A.J. Hannaford, P.W.G. Smith, A.R. Tatchell, Vogel’s
Textbook of Practical Organic Chemistry, 5th ed., John Wiley &
Sons, New York, 1994, p. 409.
[19] J. Conradie, G.J. Lamprecht, S. Otto, J.C. Swarts, Inorg. Chim. Acta
328 (2002) 191.
[20] G.J. Lamprecht, J.C. Swarts, J. Conradie, J.G. Leipoldt, Acta
Crystallogr., Sect. C 49 (1993) 82.
[21] L.G. Akselrud, Yu. N. Grin, P. Yu. Zavalii, V.K. Pacharsky, V.S.
Fundamensky, CSD-universal program package for single crystal
and/or powder structure data refinement, XII European Crystallo-
graphic Meeting, Moscow, August 1989, Collected Abstracts, v.3, p.
155.
[22] G.M. Sheldrick, Acta Crystallogr., Sect. A 46 (1990) 467.
[23] G.M. Sheldrick, SHELX93, Program for Crystal Structure Determina-
tion, University of Go¨ttingen, Germany, 1993.
[32] (a) W.C. du Plessis, T.G. Vosloo, J.C. Swarts, J. Chem. Soc., Dalton
Trans. (1998) 2507;
(b) W.C. du Plessis, J.C. Erasmus, G.J. Lamprecht, J. Conradie, T.S.
Cameron, M.A.S. Aquino, J.C. Swarts, Can. J. Chem. 77 (1999)
378.
[33] L.E. Sutin, Tables of Interatomic Distances and Configuration in
Molecules and Ions, Supplement 1956–1959, The Chemical Society,
London, p. S16s.
[34] D. Lamprecht, G.J. Lamprecht, unpublished results.