Quantifying the electronic cis effect of phosphine, arsine and stibine ligands by use of rhodium(I) Vaska-type complexes

Article abstract of DOI:10.1016/S0020-1693(03)00436-5

The cis effects of phosphine, arsine and stibine ligands have been evaluated by measuring the IR stretching frequency in dichloromethane of the carbonyl ligand in a series of Rh(I) Vaska-type complexes, trans-[RhCl(CO)(L) ₂]. These data were correlated with those obtained by Tolman for the electronic trans influences in the [Ni(L)(CO)₃] complexes. The electronic contribution, χ_Fc, of ferrocenyl was determined as 0. 8 from these plots by evaluating PPh₂Fc as ligand. In order to accommodate arsine and stibine ligands an additional correction term, to compensate for differences in the donor atom, was added to Tolman's equation for calculation of the Tolman electronic parameter of phosphine ligands. In the resulting equation: ν(CO_Ni)=2056.1+∑_i=1 ³χ_i+C_L values for C_L of C _P=0, C_As=-1.5 and C_Sb=-3.1 are suggested for phosphine, arsine and stibine ligands, respectively. The crystal and molecular structures of trans-[RhCl(CO)(PPh₂Fc)₂]·2C ₆H₆, trans-[RhCl(CO){P(NMe₂)₃} ₂] and trans-[RhCl(CO)(AsPh₃)₂] are reported. The Tolman cone angles for PPh₂Fc and P(NMe₂)₃ were determined as 169° and 166°, while the effective cone angles for PPh₂Fc, P(NMe₂)₃ and AsPh₃ were determined as 171°, 168° and 147°, respectively.

Full text of DOI:10.1016/S0020-1693(03)00436-5

Inorganica Chimica Acta 357 (2004) 1–10
www.elsevier.com/locate/ica
Quantifying the electronic cis effect of phosphine, arsine and
stibine ligands by use of rhodium(I) Vaska-type complexes
b
Stefanus Otto a,*,1, Andreas Roodt
a
University of the Free State, P.O. Box 339, Bloemfontein 9300, South Africa
b
Department of Chemistry and Biochemistry, Rand Afrikaans University, P.O. Box 524, Johannesburg 2006, South Africa
Received 4 December 2002; accepted 10 March 2003
Abstract
The cis effects of phosphine, arsine and stibine ligands have been evaluated by measuring the IR stretching frequency in di-
chloromethane of the carbonyl ligand in a series of Rh(I) Vaska-type complexes, trans-[RhCl(CO)(L)2]. These data were correlated
with those obtained by Tolman for the electronic trans influences in the [Ni(L)(CO)3] complexes. The electronic contribution, vFc, of
ferrocenyl was determined as 0.8 from these plots by evaluating PPh2Fc as ligand. In order to accommodate arsine and stibine
ligands an additional correction term, to compensate for differences in the donor atom, was added to TolmanÕs equation for cal-
P
3
culation of the Tolman electronic parameter of phosphine ligands. In the resulting equation: mðCONiÞ ¼ 2056:1 þ i¼1 vi þ CL
values for CL of CP ¼ 0, CAs ¼ )1.5 and CSb ¼ )3.1 are suggested for phosphine, arsine and stibine ligands, respectively. The crystal
and molecular structures of trans-[RhCl(CO)(PPh2Fc)2] ꢀ 2C6H6, trans-[RhCl(CO){P(NMe2)3}2] and trans-[RhCl(CO)(AsPh3)2] are
reported. The Tolman cone angles for PPh2Fc and P(NMe2)3 were determined as 169° and 166°, while the effective cone angles for
PPh2Fc, P(NMe2)3 and AsPh3 were determined as 171°, 168° and 147°, respectively.
Ó 2003 Elsevier B.V. All rights reserved.
Keywords: Rhodium complexes; Electronic parameters; Crystal structures
1. Introduction
from the metal by back donation into the empty 3d
orbitals and rꢁ orbitals of the P atom [1]. The second
aspect influencing the coordination of a ligand is the
steric factor associated with the specific ligand.
In order to rationalize certain parameters, like steric
bulk and electron donating capability, for ligands with
no or very little data available it is important to do so in
terms of other well-known ligand systems. It is also of
further importance to rationalize the information ob-
tained in such a way that all ligands employed are
compared on an identical scale.
Two main aspects are involved in the coordination of
tertiary phosphine ligands to transition metal atoms.
Firstly, the electronic character of the M–P bond which
is a combined effect of the r bond formed by donation
of the lone pair electron from the P to the M atom and,
by the ability of the P ligand to accept electron density
The diversity of tertiary phosphines in terms of their
Lewis basicity and bulkiness renders them excellent
candidates to tune the reactivity of square-planar com-
plexes towards a variety of chemical processes, such as
oxidative addition and substitution reactions [2]. The
most widely quoted parameter to indicate Lewis basicity
of tertiary phosphines is the pKa(H2O) value thereof,
which is a measure of the Brønsted basicity. In general,
substituents with better electron donating capabilities
will increase the Lewis basicity of the phosphine, while
electron-withdrawing substituents increase the p accep-
tor ability.
In order to evaluate the electronic characteristics of a
specific ligand one needs a probe, related to electron
density in some way, that can be measured conveniently
and is sensitive to changes induced in the system by the
* Corresponding author. Tel.: +27-16960-4456; fax: +27-11-5223218.
E-mail address: fanie.otto@sasol.com (S. Otto).
1
Present address: Sasol Technology R&D, P.O. Box 1, Sasolburg
1948, South Africa.
0020-1693/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0020-1693(03)00436-5

2
S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
specific ligand. The effect of tertiary phosphine ligands
with different electron donor and acceptor capabilities
was illustrated by Tolman [3] by measuring, in CH2Cl2
medium, the CO stretching frequency of the trans car-
bonyl group in Ni(0) complexes of the general formula
[Ni(L)(CO)3]. These measurements gave a good indica-
tion of the relative ‘‘electronic trans influences’’ of the
various phosphine ligands. Tolman also found that in
this specific system the electronic measurement made
was independent of the steric size of the phosphine li-
gands, a very important factor to keep in mind. Several
additional ways of evaluating the properties of phos-
phine ligands are described in the literature – these in-
clude NMR measurements of first order Pt–P, P–Se or
P–BH3 coupling constants [4–6], or by measuring the
CO stretching frequency of the rhodium analogues of
the Vaska-type complexes in solution [7].
VaskaÕs compound, trans-chlorocarbonylbis(triphe-
nylphosphine)iridium(I), was first prepared by Angoletta
[8] in 1959 and was later correctly formulated by Vaska
[9] in 1961. The analogous Rh(I) complexes were already
known at that time and investigated to some extent [10].
These complexes were soon recognized as important
model compounds for studies in homogeneous catalysis
[11] and are widely known and used today.
2. Experimental
2.1. Preparation of ligands and complexes
The PPh2Fc and PPhFc2 ligands were prepared ac-
cording to the literature procedures [26] while all other
ligands were obtained from commercially available
sources. The Vaska-type complexes were prepared in
acetone medium by reacting [Rh(CO)2(l-Cl)]2 with 4.4
equivalents of the respective ligands resulting in near
quantitative yields. The SbPh3 complexes were prepared
as reported previously [23].
2.2. Crystallography
Crystals of
a
trans-[RhCl(CO)(PPh2Fc)2] ꢀ 2C6H6
were obtained after recrystallization from benzene,
trans-[RhCl(CO){P(NMe2)3}2] were obtained directly
from the acetone reaction medium and trans-[RhCl(CO)
(AsPh3)2] were recrystallized from dichloromethane.
The densities of the crystals were determined experi-
mentally by flotation in aqueous NaI. All data collec-
tions were done on
a
Bruker SMART CCD
ꢀ
diffractometer using Mo Ka (0.71073 A) and x-scans at
293(2) K. After completion of the data collection the
first 50 frames were repeated to check for decay of
which none was observed. All reflections were merged
and integrated using SAINT [27] and were corrected for
Lorentz, polarization and absorption effects using
While organometallic complexes containing tertiary
phosphine ligands were extensively studied, the analo-
gous arsine and stibine complexes received very little at-
tention. As a consequence basic parameters that are well
established for phosphine ligands have not been extended
to include these ligand systems. As part of our continued
interest in the PPh2Fc ligand [12–15] and also in arsine
[16–21] and stibine ligands [22–24] it was of importance to
quantify the electronic properties of these ligands.
S
ADABS [28]. The structures were solved by the heavy
atom method and refined through full-matrix least
squares cycles using the SHELXL97 [29] software pack-
P
2
age with
atoms were refined with anisotropic displacement pa-
rameters while the H atoms were constrained to parent
ðjFoj ꢂ jFcjÞ being minimized. All non-H
The rhodium(I) Vaska-type complexes render them-
selves as excellent candidates for such a study and sev-
eral investigations of this nature have been reported
before [7]. Advantages of using this system include their
ease and safety of preparation and high degree of sta-
bility. Furthermore, the CO stretching frequency is
easily identifiable and gives a sensitive measure of the
‘‘electronic cis influence’’ of the ligands employed. An
extensive range of complexes were thus prepared in or-
der to construct a database from which a correlation
could be made to gain some insight into the character-
istics of the PPh2Fc, As and Sb ligands [25]. The ligands
were investigated in order to determine the electronic
effect, v, of the Fc substituent, as well as the relative
contributions of As and Sb compared to P as donor
atoms. The crystal structures of trans-[RhCl(CO)
(PPh2Fc)2] ꢀ 2C6H6 (1), trans-[RhCl(CO){P(NMe2)3}2]
(2), and trans-[RhCl(CO)(AsPh3)2] (3), are reported to
gain more information about the coordination mode of
these ligands. In addition the steric demand of the li-
gands was calculated based on the Tolman cone angle
model [3].
ꢀ
sites using a riding model (aromatic C–H ¼ 0.93 A; al-
ꢀ
iphatic C–H ¼ 0.96 A). The graphics were done with the
D
IAMOND [30] Visual Crystal Structure Information
System software.
2.3. Cone angle calculations
The cone angles for the ligands were calculated ac-
cording to TolmanÕs model [3]. Per definition a Rh–P
ꢀ
bond distance of 2.28 A was used to calculate the Tol-
man cone angles (hT) for the phosphine ligands, while
the observed Rh–P/As bond distances were used to
calculate the effective cone angles (hE) of the phosphine
and arsine ligands. A van der Waals radius of hydrogen
ꢀ
of 1.2 A was used in all calculations. In the tables con-
taining the geometrical parameters of the different
structures the individual half angles (h=2) are reported
for the cone angle calculations of each ligand. These half
angles represent the largest angle towards each of the
three substituents on the ligand as measured to the
hydrogen atom listed in brackets.

S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
3
Table 1
Crystallographic data for complexes 1, 2 and 3
1a
2b
3c
Empirical formula
Formula weight
Crystal system
Space group
C57H50ClOFe2P2Rh
1062.97
triclinic
C13H36ClN6OP2Rh
492.78
C37H30ClOAs2Rh
778.81
triclinic
monoclinic
P21=n
7.963(1)
12.479(1)
11.143(1)
90
P1
P1
ꢀ
a (A)
9.4583(19)
12.954(2)
10.466(3)
97.73(3)
86.67(3)
107.61(3)
1211.0(4)
1
9.9134(4)
10.1532(5)
18.1213(8)
75.393(10)
75.625(10)
72.398(10)
1652.86(13)
2
ꢀ
b (A)
ꢀ
c (A)
a (°)
b (°)
c (°)
98.92(1)
90
ꢀ3
V (A )
1093.89(19)
2
1.496
Z
Dc (g cmꢂ3
)
1.458
1.089
1.565
2.612
l (mmꢂ1
)
1.062
0.901/0.857
512
Tmax/Tmin
F ð000Þ
0.908/0.743
544
0.637/0.479
776
Crystal size (mm)
h limit (°)
Index ranges
0.29 ꢃ 0.21 ꢃ 0.09
1.66–32.02
ꢂ13 6 h 6 12
ꢂ15 6 k 6 19
ꢂ15 6 l 6 15
12 906
0.15 ꢃ 0.13 ꢃ 0.10
2.47–31.86
ꢂ11 6 h 6 11
ꢂ18 6 k 6 18
ꢂ12 6 l 6 16
11 862
0.33 ꢃ 0.20 ꢃ 0.19
5.86–28.33
ꢂ13 6 h 6 13
ꢂ13 6 k 6 13
ꢂ24 6 l 6 22
18 287
Collected reflections
Independent reflections
Observed reflections [I > 2rI]
Rint
7349
5294
3420
2229
8045
6097
0.0259
0.0434
0.0229
Data/restraints/parameters
S
7349/0/305
0.984
0.0352
3420/0/131
0.901
0.0303
8045/0/380
1.029
0.0319
d
RðI > 2rIÞ
wRe
R (all data)
0.0765
0.0605
0.0611
0.0639
0.0661
0.0513
wR
0.0835
0.383; )0.458
0.0667
0.348; )0.663
0.0739
0.747; )0.810
ꢂ3
ꢀ
Dqmax; Dqmin (e A
)
a trans-[RhCl(CO)(PPh2Fc)2] ꢀ 2C6H6
b trans-[RhCl(CO){P(NMe2)3}2]
c trans-[RhCl(CO)(AsPh3)2]
P
P
d R ¼ ½ð DF Þ=ð FoÞꢄ
P
2.4. Spectroscopic measurements and data fitting
P
2
2 1=2
e wR ¼ ½wðFo2 ꢂ Fc2Þ ꢄ= ½wðFo2Þ ꢄ
parameters are summarized in Tables 2–4, respec-
tively. The molecular structures with thermal ellipsoids
and numbering schemes for 1, 2 and 3 are shown in
Figs. 1, 2 and 3, respectively.
All IR spectroscopic measurements were made on a
Hitachi 270-50 spectrometer as dichloromethane solu-
tions using a NaCl cell with the background spectrum
of dichloromethane being corrected for, the measure-
ments being accurate to within 1 cmꢂ1. It was immi-
nent to determine the values of m(CORh) in solution
since packing effects may have a significant influence
on the measured value as shown before [31]. The data
were fitted to a simple quadratic function ðy ¼ ax2 þ
bx þ cÞ using the SCIENTIST least-squares program [32].
3.1.1. Crystal structure of trans-[RhCl(CO) (PPh2 Fc)2] ꢀ
2C6H6 (1)
As part of a preliminary study [25] crystals of trans-
[RhCl(CO)(PPh2Fc)2] ꢀ CH2Cl2 were isolated and the
2
structure was determined. Since severe disorders were
2
1
2
trans-[RhCl(CO)(PPh2Fc)2] ꢀ CH2Cl2; Enraf Nonius CAD 4; Mo
ꢀ
Ka ¼ 0:71073 A; empirical formula, C45: 5H39Cl2OP2Fe2Rh; FW,
ꢀ
944.20; crystal system, triclinic; space group, P1; a, 9.4680(10) A; b,
ꢀ
ꢀ
12.995(2) A; c, 18.085(3) A; a, 107.93(2)°; b, 96.31(2)°; c, 95.19(2)°; V,
3
ꢀ
3. Results and discussion
2086.0(5) A ; Z, 2; Dc, 1.500 g cmꢂ3; Dm, 1.454 g cmꢂ3; l, 1.315 mmꢂ1
;
T
max=Tmin, 0.737/ 0.668; F ð000Þ, 954; crystal size, 0.45 ꢃ 0.30 ꢃ 0.28
mm; 1:71 6 h 6 24:97°; index ranges, 0 6 h 6 11, ꢂ15 6 k 6 15,
3.1. Crystallography
ꢂ21l 6 21; collected reflections, 5411; independent reflections, 5071;
R
int, 0.0154; observed reflections, I > 2rI 5396; data/restraints/param-
Details of the data collections and the refinement
parameters are given in Table 1 and selected geometrical
eters, 5396/72/545; S, 1.156; R=wR(I > 2rI), 0.0382/ ꢂ03.1223; R=wR (all
ꢀ
data), 0.0382/0.1223; Dqmax=Dqmin, 1.142/)0.796 e A
.

4
S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
Table 2
ꢀ
Bond lengths (A), bond-, torsion- and steric angles (°) for trans-[RhCl(CO)(PPh2Fc)2] ꢀ 2C6H6
Rh–P
2.3346(9)
1.787(6)
0.581(5)
1.831(2)
1.807(2)
2.036(3)
Rh–Cl
O–C(0)
O–Cl0
2.3670(18)
1.138(6)
0.559(3)
1.822(2)
2.038(2)
Rh–C(0)
C(0)–Cl0
P–C(11)
P–C(31)
Fe–Cp(4)avg
P–C(21)
Fe–Cp(3)avg
C(0)–Rh–P
C(0)–Rh–Cl
P–Rh–Cl
87.8(2)
C(0)0–Rh–P
C(0)0–Rh–Cl
P0–Rh–Cl
92.2(2)
0.8(3)
179.2(3)
92.51(5)
103.10(10)
102.92(9)
112.26(7)
179.1(8)
87.49(5)
102.25(10)
118.50(7)
115.87(7)
C(11)–P–C(21)
C(21)–P–C(31)
C(21)–P–Rh
O–C(0)–Rh
C(11)–P–C(31)
C(11)–P–Rh
C(31)–P–Rh
Cl–Rh–P–C(11)
Cl–Rh–P–C(31)
)12.39(9)
109.72(9)
Cl–Rh–P–C(21)
Rh–P–C(31)–C(41)
)132.44(9)
)46.72(8)
ðh=2Þ (H12)
ðh=2ÞE (H26)
67.3
81.9
105.0
169
ðh=2Þ (H12)
ðh=2ÞT (H26)
68.2
82.9
105.8
171
ðh=2ÞE (H42)
ðh=2ÞT (H41)
E
T
hE
hT
Table 3
Selected bond lengths (A), bond, torsion and steric angles (°) for trans-[RhCl(CO){P(NMe2)3}2]
ꢀ
Rh–P
2.3416(6)
1.723(6)
0.724(6)
1.6932(18)
1.6691(19)
1.465(3)
1.449(3)
1.459(3)
Rh–Cl
C(0)–O
Cl–O0
2.447(4)
1.145(9)
Rh–C(0)
Cl–C(0)0
P–N(1)
0.425(10)
1.6637(17)
P–N(3)
P–N(2)
N(1)–C(11)
N(2)–C(21)
N(2)–C(22)
N(1)–C(12)
N(3)–C(31)
N(3)–C(32)
1.464(3)
1.446(3)
1.459(3)
P–Rh–P0
180.0
89.1(2)
C(0)–Rh–Cl
P–Rh–Cl
179.5(2)
90.9(2)
C(0)–Rh–P
C(0)–Rh–Cl0
N(2)–P–N(1)
N(3)–P–N(1)
N(3)–P–N(2)
C(11)–N(1)–P
C(21)–N(2)–P
C(31)–N(3)–P
C(12)–N(1)–C(11)
C(31)–N(3)–C(32)
0.5(2)
101.60(9)
99.12(9)
O–C(0)–Rh
N(1)–P–Rh
N(2)–P–Rh
177.1(11)
120.94(7)
110.38(7)
112.78(7)
119.71(16)
122.90(16)
124.24(16)
113.07(19)
111.07(10)
116.03(15)
122.50(16)
122.10(16)
112.96(18)
112.78(18)
N(3)–P–Rh
C(12)–N(1)–P
C(22)–N(2)–P
C(32)–N(3)–P
C(21)–N(2)–C(22)
Cl–Rh–P–N(1)
Cl–Rh–P–N(3)
–4.86(14)
112.03(15)
Cl–Rh–P–N(2)
)123.08(14)
ðh=2Þ (H11C)
ðh=2ÞE (H21A)
70.5
90.2
88.2
166
ðh=2Þ (H11C)
ðh=2ÞT (H21A)
71.4
91.6
89.6
168
ðh=2ÞE (H31B)
ðh=2ÞT (H31B)
E
T
hE
hT
encountered in both the rhodium co-ordination com-
pound and the solvent of crystallization the current
structure determination was selected for a detailed report
in this paper. Compound 1, trans-[RhCl(CO) (PPh2Fc)2] ꢀ
2C6H6, crystallize isomorphous to the previously re-
ported trans-[PtCl2(PPh2Fc)2] ꢀ 2C6H6 [12] and trans-
[PdClMe(PPh2Fc)2] ꢀ 2C6H6 [14] complexes in P1 with the
metal atom situated on an inversion centre resulting in a
50% statistical disorder in the CO and Cl positions. Each
Rh moiety co-crystallized with two benzene solvent
molecules in the unit cell. It is clearly the bulky nature of
the PPh2Fc ligand that predominantly determines the

S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
5
Table 4
Selected bond lengths (A), bond, torsion and steric angles (°) for trans-[RhCl(CO)(AsPh3)2]
ꢀ
Rh–As(1)
Rh–C(1)
2.4226(4)
2.015(7)
1.944(3)
1.943(3)
1.945(3)
0.707(8)
Rh–As(2)
Rh–Cl
2.4217(4)
2.3599(12)
1.951(3)
1.946(3)
1.943(3)
As(1)–C(111)
As(1)–C(121)
As(1)–C(131)
C(0)–O
As(2)–C(211)
As(2)–C(221)
As(2)–C(231)
As(1)–Rh–As(2)
C(0)–Rh–As(1)
C(0)–Rh–As(2)
O–C(0)–Rh
176.020(15)
88.61(16)
91.48(16)
175.7(11)
112.43(8)
114.15(8)
121.50(8)
102.07(12)
C(0)–Rh–Cl
Cl–Rh–As(1)
Cl–Rh–As(2)
179.01(17)
92.28(3)
87.59(3)
C(111)–As(1)–Rh
C(121)–As(1)–Rh
C(131)–As(1)–Rh
[C–As(1)–C]avg
C(211)–As(2)–Rh
C(221)–As(2)–Rh
C(231)–As(2)–Rh
[C–As(2)–C]avg
120.79(9)
114.67(8)
110.31(9)
103.03(13)
Cl–Rh–As(1)–C(111)
Cl–Rh–As(1)–C(121)
Cl–Rh–As(1)–C(131)
)147.50(10)
97.78(10)
Cl–Rh–As(2)–C(111)
Cl–Rh–As(2)–C(121)
Cl–Rh–As(2)–C(131)
)156.09(12)
)31.21(10)
84.92(10)
)26.87(11)
ðh=2Þ (H116)
ðh=2ÞE (H126)
77.2
74.6
66.4
145
ðh=2Þ (H216)
ðh=2ÞE (H226)
65.7
71.6
85.4
148
ðh=2ÞE (H132)
ðh=2ÞE (H236)
E
E
hEð1Þ
hEð2Þ
orientation of the ferrocenyl moiety [14]. In this regard
the Rh–P–C(31)–C(41) solid state torsion angle of
)46.72(8)° results in effective- and Tolman cone angles
of hE ¼ 169° and hT ¼ 171°.
3.1.2. Crystal structure of trans-[RhCl(CO){P
(NMe2)3}2] (2)
The molecular structure of 2, trans-[RhCl(CO)
{P(NMe2)3}2], shows the coordination complex to ex-
hibit a square planar geometry with angles deviating only
slightly from the expected 90°. As in 1 the Rh atom is
situated on a special position in the monoclinic space
group P21=n resulting in a 50% statistical disorder be-
tween the CO and Cl positions. Some of the most note-
worthy observations made from the structure include an
Fig. 1. Molecular diagram showing the numbering scheme and dis-
placement ellipsoids (30% probability) in trans-[RhCl(CO)
(PPh2Fc)2] ꢀ 2C6H6. Hydrogen atoms and the benzene solvent mole-
cules were omitted for clarity. In the numbering scheme of the Ph and
Cp rings the first digit refers to the number of the ring, 1 and 2 for Ph
and 3 and 4 for Cp, the second digit refers to the number of the C atom
in the rings.
ꢀ
exceptionally long Rh–Cl bond of 2.443(7) A as well as a
ꢀ
relatively long Rh–P bond distance of 2.3426(7) A.
As encountered with other phosphine ligands in the
case of the P–C bond distances one of the P–N bond
distances is considerably longer than the other two with
P–N(1) ¼ 1.695(2) compared to P–N(2) ¼ 1.668(2) and
P–N(3) ¼ 1.661(2), respectively. The O–C–Rh angle is
175.8(14)° and deviates significantly from linearity.
The effective (hE) and Tolman (hT) cone angles of the
P(NMe2)3 ligand were determined as hE ¼ 166° and
hT ¼ 168°, respectively.
mode of packing when a relatively small metal core is
present. Except for a fairly short Rh–Cl bond distance of
ꢀ
2.3670(18) A all the other bonds were within expected
limits. The O–C–Rh moiety is close to linear with an angle
ꢀ
of 179.1(8)°. The P–C bond of 1.807(2) A towards the
ferrocenyl is significantly shorter than those to the phenyl
3.1.3. Crystal structure of trans-[RhCl(CO)(AsPh3)2]
(3)
ꢀ
substituents that averages 1.826(2) A, and is comparable
with that found in previously reported crystallographic
studies containing the PPh2Fc ligand [12–15].
From previous studies it was found that the steric
demand of the PPh2Fc ligand is very sensitive to the
Although several structures of rhodium complexes
containing bidentate arsine ligands are reported in the
literature [33], the monodentate tertiary arsine com-
plexes are far less studied. Several polymorphs of the

6
S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
C (31)
C (32)
N(3)
O
0)
C (12)
P'
P
R h
N(1)
C l
N(2)
C (22)
C (21)
C (11)
Fig. 2. Molecular diagram showing the numbering scheme and displacement ellipsoids (30% probability) in trans-[RhCl(CO){P(NMe2)3}2]. Hy-
drogen atoms (thermal ellipsoids of arbitrary size) were included to give an indication of the size of the P(NMe2)3 ligand.
bonds in similar complexes. Contrary to the P–C
bonds in phosphine ligands, it was found that all the
As–C bonds are comparable in length, an effect
probably induced by the longer As–C bonds compared
to P–C bonds resulting in less steric strain in the As
ligand system. It was found, however, that one of the
C–As–Rh tetrahedral angles on each ligand deviates
quite significantly, 120.50(10)° and 120.74(10)° for
C(131)–As(1)–Rh and C(211)–As(2)–Rh, respectively,
from the other two, as well as from the normal 109°.
An O–C–Rh angle of 175.0(12)°, deviating from line-
arity, was also observed while the C–O bond refined
ꢀ
to a slightly obscure value of 0.717(7) A, an effect
probably caused by the long Rh–C bond of 2.017(7)
ꢀ
A. The structure was checked for a random disorder
in the CO and Cl positions, but no exclusive evidence
Fig. 3. Molecular diagram showing the numbering scheme and dis-
placement ellipsoids (30% probability) in trans-[RhCl(CO)(AsPh3)2].
for such an effect was observed. Similar effects have
been observed before and are well documented in the
literature [34,35].
The effective (hE) cone angles of the AsPh3 ligands
were determined as hEð1Þ ¼ 145° and hEð2Þ ¼ 148° for
As(1) and As(2), respectively.
The first digit in the phenyl rings refers to the number of the As ligand
(1 or 2), the second digit to the number of the ring (1–3) and the third
digit to the number of the C atom in the ring (1–6). Hydrogen atoms
were omitted for clarity.
analogous PPh3 complex have been reported to date
[31], but surprisingly this structure is not isomorphous
to any of these. The structure of 3, trans-
[RhCl(CO)(AsPh3)2], shows the coordination complex
to exhibit a slightly distorted square planar geometry,
with the bulky arsine ligands in a trans orientation as
expected. The As(1)–Rh(1)–As(2) angle of 175.97(15)°
deviates slightly from linearity and is a probable reason
why the molecule crystallizes on a general position in P1
rather than on an inversion centre.
3.1.4. Solid state comparisons
The Rh–Cl bond distances for the complexes listed in
ꢀ
Table 5 average 2.371(7) A and range from 2.315(3) to
ꢀ
2.443(7) A for the bis-SbPh3 and the P(NMe2)3 com-
plexes, respectively. It was also found that the Rh–Cl
bond distance decreased from 2.382(1) to 2.3538(14) to
ꢀ
2.315(3) A for the PPh3 to AsPh3 to SbPh3 complexes;
indicative of a decrease in the cis effect when descending
in group 15. This can most probably be attributed to a
combination of the formation of more electrophilic
metal centres combined with a decrease in the steric
crowding experienced in the molecules in that order
[22,23,36–41].
ꢀ
The Rh–Cl bond distance of 2.3538(14) A is
slightly shorter than that generally encountered while
ꢀ
the average Rh–As bond length of 2.4225(4) A is al-
ꢀ
most 0.1 A longer than that normally found for Rh–P

S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
7
Table 5
Comparison of trans-[RhCl(CO)(L)2] complexes
Rh–La (A)
L–Pt–L (°)
Reference
ꢀ
ꢀ
Rh–Cl (A)
ꢀ
Rh–C (A)
ꢀ
C–O (A)
L
PPh3
2.322(1)
2.317(2)
2.314(2)
2.382(1)
2.363(2)
2.357(2)
2.3654(15)
2.415(7)
2.443(7)
2.3581(12)
2.381(2)
2.347(2)
2.3599(12)
2.315(3)
1.77(1)
1.140(2)
1.142(8)
1.154(8)
1.162(6)
1.056(14)
1.15(2)
180.0(1)
178.2(1)
174.11(6)
177.67(6)
180
[36]
[37]
[38]
[39]
TW
TW
[40]
[41]
[22]
TW
[23]
PPh2Me
PPhMe2
PBz3
1.795(7)
1.800(6)
1.783(6)
1.814(14)
1.731(9)
1.798(5)
1.800(2)
1.788(10)
2.015(7)
1.797(13)
2.3160(16)
2.3344(14)
2.3426(7)
2.3325(11)
2.322(2)
PPh2Fc
P(NMe2)3
P(p-Tol)3
P(p-F-Ph)3
As(p-Tol)3
AsPh3
180
1.139(6)
1.150(7)
1.139(10)
0.707(8)
1.175(13)
175.67(4)
178.7(2)
175.69(4)
176.020(15)
180
2.4120(10)
2.4222(4)
2.5655(2)
SbPh3
a Average of unequivalent bonds where applicable.
Unfortunately, only two Rh–Cl bond distances are
available for complexes containing arsine ligands, but
they do seem to be slightly shorter than in the analogous
phosphine complexes. The Rh–C and C–O bond lengths
values, but is purely intended as representative of the
experimental points within the region described by the
data. Even though the equation could in principle be
used to calculate the m(CONi) value with fair accuracy
for any ligand if the corresponding m(CORh) value is
known, no physical meaning is attached to any of the
constants. Tolman rationalized the measurements on the
[NiL(CO)3] complexes in terms of Eq. (1) that can now
ꢀ
average 1.82(2) and 1.09(3) A, respectively, but there are
usually such large s.u.Õs associated with these values that
it may be a risk to make conclusions based purely
thereupon.
An increase in the Rh–L bond lengths is observed
when descending periodically from PPh3 to AsPh3 to
SbPh3 as ligands as illustrated by the values of 2.322(1),
Table 6
IR (CH2Cl2), data for the [Ni(L)(CO)3] and trans-[RhCl(CO)(L)2]
complexes
ꢀ
2.4226(4) and 2.5655(2) A, respectively. The general L–
C bond distances for the P, As and Sb ligands increase
ꢀ
from ca. 1.8 to 1.9 to 2.1 A while the C–L–C tetrahedral
a
L
m(CONi
)
(cmꢂ1
)
m(CORh) (cmꢂ1
)
angles for the triphenyl derivatives of group 15 elements
were reported [42] to decrease from N to P to As to Sb
to Bi in the sequence 116°, 109°, 102°, 97°, and 94°.
These changes in bond length and tetrahedral angles are
in accordance with the increase in the covalent radii of
the ligands going from N(0.70) to P(1.10) to As(1.21) to
1
2
PCy3
PiPr3
AsEt3
2056.4
2059.2
1943
1950
1954
1955
1957c
1958
1964
1964
1964
1966
1968
1970
1970
1971
1973
1973
1974
1975
1975
1976
1977
1979
1983
1983
1983
1993
1996
2005
3
2059.6b
2060.3
2060.5b
2061.7
2061.9
2063.7
2060.6
2064.8
2067.0
2066.4
2065.5b
2065.8b
2066.7
2066.8b
2066.6
2067.9b
2066.1
2066.7
2068.2
2068.9
2075.9
2071.3
2072.8
2080.7
2082.8
2090.9
4
PnBu3
5
PPhFc2
PEt3
6
7
P(NMe2)3
PPhEt2
ꢀ
Sb(1.41) to Bi(1.47) A.
8
In the PPh2Fc ligand all bond distances and angles of
the coordination polyhedron were within normal limits
indicating its behaviour to be similar to most other
phosphine ligands investigated. Investigating the Vaska
complex thereof did, however, make it possible to get an
additional value for the cone angle, as well as an esti-
mation of its electron donating capability.
9
PPhCy2
PPh2Cy
PPh2Me
PBz3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
PPh2Fc
SbPh3
PPh2Et
As(p-Tol)3
P(o-Tol)3
AsPh3
3.2. Spectroscopic measurements
P(p-OMe-Ph)3
P(p-Tol)3
PPh2(p-Tol)
PPh3
The synthesized complexes and the corresponding IR
)
data collected are listed in Table 6, and these m(CORh
values were plotted against the m(CONi) data reported by
Tolman. A good fit of the data, within the region of
interest, was obtained by fitting a simple quadratic
function, y ¼ ax2 þ bx þ c, through the points with
PPh2(C6F5)
P(p-F-Ph)3
P(p-Cl-Ph)3
PPh2Cl
PPh(C6F5)2
P(C6F53
values of a ¼ ð7:2 ꢅ 1:1Þ ꢃ 10ꢂ3
,
b ¼ ꢂ28 ꢅ 4 and
)
c ¼ ð29 ꢅ 4Þ ꢃ 103 being obtained. This mathematical
equation is in no means implicated to give a funda-
mental quantum mechanical relationship between the
a Ref. [3].
b Obtained from Fig. 4.
c Measurement in KBr.

8
S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
be used to calculate the electronic parameter m(CONi) for
phosphine ligands containing common substituents
for vFc the Vaska complex of PPhFc2 was also prepared.
The solubility of this complex was, however, so limited
in all common organic solvents evaluated (even aceto-
nitrile and DMF) that no solution measurements could
be made. A solid state stretching frequency of 1955 cmꢂ1
was obtained in KBr discs resulting in a m(CONi) value
of 2060.5 and a vFc value of only 0.05. This value is
unfortunately not consistent with what one would ex-
pect, most probable due to the solid state-solution state
extrapolation, and will hence not be used in any dis-
cussions further on.
3
X
mðCONiÞ for PX1X2X3 : mðCONiÞ ¼ 2056:1 þ
vi:
i¼1
ð1Þ
In Eq. (1) each substituent has a value for v (t-Bu ¼ 0;
Cy ¼ 0.1; Et ¼ 1.8; Ph ¼ 4.3, etc.) that has to be known
to calculate the value of m(CONi). This equation implies
that P(tBu)3 has the highest electronic trans influence
t
(strongest electron donor) and on replacing the Bu
In order to determine the contributions of As and Sb
relative to P an additional term correcting for the donor
atom (CL; L ¼ P, As or Sb) was added to Eq. (1) thus
giving Eq. (2):
groups with other groups, a correction must be applied
since a decrease in the trans influence results in an in-
crease in the CO stretching frequency. This observation
is best explained in terms of a decrease in electron
density on the metal centre giving rise to poorer back
donation from the metal into the anti-bonding orbital
on the C atom. The decrease in the back donation re-
sults in a weaker Rh–C bond and thus a stronger CO
bond that in turn results in an increase in the CO
stretching frequency.
The m(CONi) versus m(CORh) data from Table 6 were
fitted to the curve shown in Fig. 4. It is clear that the
m(CORh) values cover almost twice the range than the
m(CONi) values with Dm(CORh) ¼ 62 cmꢂ1 and
Dm(CONi) ¼ 34.5 cmꢂ1. From this graph the corre-
sponding m(CONi) values for the PPh2Fc, As and Sb li-
gands were obtained and are reported in Table 6.
Substituting the m(CONi) value for PPh2Fc (2065.5
cmꢂ1) obtained from Fig. 4 into Eq. (1) and solving
(vPh ¼ 4.3) gives: 2065.5 ¼ 2056.1 + 2(4.3) + vFc; thus
vFc ¼ 0.8. This value of 0.8 for vFc indicates it to be a
good electron-donating group, comparable to o-OMe–
Ph (v ¼ 0:9). In an attempt to obtain an additional value
3
X
mðCONiÞ ¼ 2056:1 þ
vi þ CL:
ð2Þ
i¼1
According to definition the phosphine ligands have
CP ¼ 0. Substituting the m(CONi) values for AsEt3
(2059.6), AsPh3 (2067.9), As(p-Tol)3 (2066.8) and SbPh3
(2065.8) obtained from Fig. 4 into Eq. (2) and solving
gave the following values for CAs ¼ )1.9 (AsEt3), )1.1
(AsPh3) and +0.2 (As(p-Tol)3) and CSb ¼ )3.1(SbPh3).
From these results the CAs value of +0.2 obtained for
As(p-Tol)3 seems a bit strange compared to the negative
values obtained for the other As ligands. Direct sub-
traction of the m(CORh) values of As(p-Tol)3 and P(p-
Tol)3 gives 1973–1976 ¼ )3 cmꢂ1 and dividing by 2 to
convert to the smaller range for m(CONi) suggests a value
of )1.0 to )1.5 cmꢂ1. Based on these arguments an
average value of 1.5 is proposed for CAs and the values
of )3.1 is retained for CSb.
Fig. 4. Non-linear fit of m(CONi) vs. m(CORh) for all known complexes ðj) in Table 6, fitted values for new complexes (*) were obtained from the
least-squares calculations.

S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
9
At first though it might seem like the m(CORh) values
Acknowledgements
for both the As and the Sb complexes suggest that the
Rh metal centre should be more electron rich (lower
m(CORh) values). However, it is clear that the Rh–Cl
bond distance for these complexes significantly shortens,
Financial assistance from the South African NRF
and the research funds of the University of the Free
State, the Rand Afrikaans University, SIDA (Swedish
International Cooperation Development Agency) is
gratefully acknowledged. Lund University and the
University of the Witwatersrand are acknowledged for
the use of their diffractometers.
ꢀ
see Table 5 (2.382(1), 2.354(2) and 2.315(3) A) for the
PPh3, AsPh3 and SbPh3 complexes, respectively. Thus it
seems as if the electron density on the Rh metal centre
decreases significantly from PPh3 to AsPh3 to SbPh3.
However, it cannot be excluded that this observation
might be due to steric effects.
It is thus suggested that the decrease in electron
density on the metal centres is a combined result of the
longer Rh–L bond distances and the corresponding
smaller steric demands of the SbPh3 versus AsPh3 versus
PPh3 ligands, which is in fact manifested by stronger M–
Cl bonds. The increased apparent higher electron den-
sity of the complexes, based on the m(CORh) values, is
thus in the order PPh3 > AsPh3 > SbPh3. The impor-
tance of the smaller steric demand is further manifested
in the fact that five-coordinate complexes for the Sb
system form with relative ease [23]. As in all experi-
mental measurements it has to be remembered that, as
References
[1] F.A. Cotton, G. Wilkinson, Advanced Inorganic Chemistry,
Wiley, New York, 1988.
[2] G.O. Spessard, G.L. Miessler, Organometallic Chemistry, Pre-
ntice-Hall, New Jersey, 1997.
[3] C.A. Tolman, Chem. Rev. 77 (1997) 313.
[4] D.W. Allen, B.F. Taylor, J. Chem. Soc., Dalton Trans. (1982)
51.
[5] A.H. Cowley, M.C. Damasco, J. Am. Chem. Soc. 93 (1971) 6815.
[6] R.W. Rudolph, C.W. Schultz, J. Am. Chem. Soc. 93 (1971) 6821.
[7] K.G. Moloy, J.L. Petersen, J. Am. Chem. Soc. 117 (1995) 7696.
[8] M. Angoletta, Gazz. Chim. Ital. 89 (1959) 2359.
[9] L. Vaska, J.W. DiLuzio, J. Am. Chem. Soc. 83 (1961) 2784.
[10] J. Chatt, B.L. Shaw, Chem. Ind. (1961) 290.
[11] A. Roodt, S. Otto, G.J. Steyl, Coord. Chem. Rev. (2003) (in
press).
illustrated, many factors may influence the m(CORh
)
values and appropriate caution needs to be exercised
when conclusions are made.
[12] S. Otto, A. Roodt, Acta Cryst. C53 (1997) 1414.
[13] S. Otto, A. Roodt, J.J.C. Erasmus, J.C. Swarts, Polyhedron 17
(1998) 2447.
4. Conclusions
[14] S. Otto, A. Roodt, J. Smith, Inorg. Chim. Acta 303 (2000) 295.
[15] G. Steyl, S. Otto, A. Roodt, Acta Cryst. E57 (2001) m352.
[16] A. Roodt, S. Otto, Acta Cryst. C51 (1995) 1105.
[17] S. Otto, A. Roodt, Acta Cryst. C52 (1996) 1636.
[18] S. Otto, A. Roodt, Acta Cryst. C53 (1997) 280.
ꢀ
[19] M.H. Johansson, S. Otto, A. Roodt, A. Oskarsson, Acta Cryst.
B56 (2000) 226.
A clear trend was illustrated between the stretching
frequencies of the Rh and Ni complexes. The rhodium
Vaska-type complexes were then used in a similar fashion
as the [NiL(CO)3] complexes to establish the electronic
parameters of various ligands, as an example the elec-
tronic parameters were determined for the ferrocenyl
moiety. A correction term was added to TolmanÕs original
equation to account for ligands containing different do-
nor atoms of group 15 in the periodic table, i.e., phos-
phorous, arsenic and antimony. The rhodium Vaska-type
system has the added benefits that the complexes are easily
accessible and safe to work with and it covers almost twice
the range than the [NiL(CO)3] complexes making more
significant measurements feasible.
In addition to the Tolman cone angle model the im-
plementation of the effective cone angle principle was
extended which takes the real metal ligand bond distance
into account when the cone angle is calculated. This
principle is especially applicable in calculating the cone
angles for the arsine and stibine ligands since no default
bond distance has been established for these ligands.
[20] S. Otto, A.J. Muller, Acta Cryst. C57 (2001) 1405.
[21] S. Otto, M.H. Johansson, Inorg. Chim. Acta 329 (2002) 135.
[22] S. Otto, S.N. Mzamane, A. Roodt, in: G.J. Leigh, N. Winterton
(Eds.), Modern Coordination Chemistry, The Legacy of Joseph
Chatt, Royal Society of Chemistry, Oxford, UK, 2002, p. 328.
[23] S. Otto, A. Roodt, Inorg. Chim. Acta 331 (2002) 199.
[24] S. Otto, A. Roodt, Acta Cryst. C58 (2002) m565.
[25] S. Otto, Ph.D. Dissertation, University of the Free State, South
Africa, 1999.
[26] G.P. Sollot, H.E. Mertway, S. Portnoy, J.L. Snead, J. Org. Chem.
28 (1963) 1090.
[27] SAINT, Siemens Analytical X-ray Instruments Inc., Madison,
Wisconsin, USA, 1995.
€
[28] G.M. Sheldrick, SADABS, University of Gottingen, Germany.
[29] G.M. Sheldrick, SHELXL97, University of Gottingen, Germany.
€
[30] K. Brandenburg, M. Brendt, DIAMOND, Release 2.1c, Crystal
Impact GbR, Postfach 1251, D-53002, Bonn, Germany, 2000.
[31] G. Kemp, A. Roodt, W. Purcell, Rhodium Expr. 12 (1995) 21.
[32] SCIENTIST for Windows: Program for Least-squares Parameter
Estimation, Release 4.00, Micromath Inc., Utah, 1990.
[33] M. Cowie, S.K. Dwight, Inorg. Chem. 20 (1981) 1534.
[34] R.L. Harlow, S.A. Westcott, D.L. Thorn, R.T. Baker, Inorg.
Chem. 31 (1992) 323.
5. Supplementary material
[35] A.L. Del Pra, G. Zanotti, P. Segala, Cryst. Struct. Commun. 8
(1979) 959.
[36] K.R. Dunbar, S.C. Haefner, Inorg. Chem. 31 (1992) 3676.
Complete crystallographic data in CIF format.

10
S. Otto, A. Roodt / Inorganica Chimica Acta 357 (2004) 1–10
[37] F. Dahan, R. Choukroun, Acta Cryst. C41 (1985) 704.
[38] H. Schumann, S. Jurgis, M. Eisen, J. Blum, Inorg. Chim. Acta 172
(1990) 191.
[40] S. Otto, S.N. Mzamane, A. Roodt, Acta Cryst. C55 (1999) 67.
ꢁ
[41] A. Monge, E. Gutierrez-Puebla, J.V. Heras, E. Pinilla, Acta Cryst.
C39 (1983) 446.
ꢀ
[39] A.J. Muller, A. Roodt, S. Otto, A. Oskarsson, S. Yong, Acta
Cryst. E58 (2002) m715.
[42] A.N. Sobolev, I.P. Romm, V.K. Belsky, E.N. Guryanova, J.
Organomet. Chem. 179 (1979) 153.