Removal of metal-metal bonding in a dimetallic paddlewheel complex: Molecular and electronic structure of bis(phenyl) dirhodium(III) carboxamidate compounds

Article abstract of DOI:10.1021/om800631b

The unique structural features and chemical stabilities of bis(phenyl) tetracarboxamidatodirhodium(III) are reported, and their electronic structures are mapped through XPS, electrochemical, and computational methods. Comparison with the structures of dirhodium(II,II) and dirhodium(II,III) oxidative precursors portrays the diphenyl dirhodium(III) compounds as two square-pyramidal rhodium units that have undergone conrotatory motion in order to optimize metal-ligand bonding. Axial phenyl ligands are severely distorted from their expected Rh-Rh-C linear array. XPS data for this series of dirhodium compounds are consistent with the absence of a rhodium-rhodium bond for the diphenyl dirhodium(III) compounds, and electrochemical measurement shows a single reversible Rh₂⁶⁺/Rh₂⁷⁺ redox couple. Notably, they exhibit high thermal stability, and Bronsted acid removal of carboxamidate ligands precedes the formation of benzene. The ability of a phenyl group to impart unusual stability to rhodium(III) compounds is explained by theoretical analysis of the electronic structure of bis-σ-(phenyl)-tetrakis-μ-(carboxamido)dirhodium(III), and comparison is made with previously reported dinitrosyl dirhodium(III) complexes. Formally, two phenyl radicals in combination with a rhodium-rhodium bond are transformed into two phenyl-rhodium bonds. The severe distortion of phenyl rings from linearity is suggested to result from a long-range interaction between the electron-deficient rhodium and distal oxygen atoms.

Full text of DOI:10.1021/om800631b

5836
Organometallics 2008, 27, 5836–5845
Removal of Metal-Metal Bonding in a Dimetallic Paddlewheel
Complex: Molecular and Electronic Structure of Bis(phenyl)
Dirhodium(III) Carboxamidate Compounds
Joffrey Wolf,† Rinaldo Poli,*,†,§ Jian-Hua Xie,‡ Jason Nichols,‡ Bin Xi, Peter Zavalij,‡ and
Michael P. Doyle*,‡
Department of Chemistry and Biochemistry, UniVersity of Maryland, College Park, Maryland 20742,
Laboratoire de Chimie de Coordination, UPR CNRS 8241 Lie´e par ConVention a` l’UniVersite´ Paul
Sabatier et a` l’Institut National Polytechnique de Toulouse, 205 Route de Narbonne,
31077 Toulouse Cedex 4, France, Institut UniVersitaire de France, 103, BouleVard Saint-Michel,
75005 Paris, France, and Department of Chemistry, Purdue UniVersity, West Lafayette, Indiana 47907
ReceiVed July 5, 2008
The unique structural features and chemical stabilities of bis(phenyl) tetracarboxamidatodirhodium(III)
are reported, and their electronic structures are mapped through XPS, electrochemical, and computational
methods. Comparison with the structures of dirhodium(II,II) and dirhodium(II,III) oxidative precursors
portrays the diphenyl dirhodium(III) compounds as two square-pyramidal rhodium units that have
undergone conrotatory motion in order to optimize metal-ligand bonding. Axial phenyl ligands are severely
distorted from their expected Rh-Rh-C linear array. XPS data for this series of dirhodium compounds
are consistent with the absence of a rhodium-rhodium bond for the diphenyl dirhodium(III) compounds,
and electrochemical measurement shows a single reversible Rh26+/Rh2 redox couple. Notably, they
7+
exhibit high thermal stability, and Brønsted acid removal of carboxamidate ligands precedes the formation
of benzene. The ability of a phenyl group to impart unusual stability to rhodium(III) compounds is explained
by theoretical analysis of the electronic structure of bis-σ-(phenyl)-tetrakis-µ-(carboxamido)dirhodium(III),
and comparison is made with previously reported dinitrosyl dirhodium(III) complexes. Formally, two
phenyl radicals in combination with a rhodium-rhodium bond are transformed into two phenyl-rhodium
bonds. The severe distortion of phenyl rings from linearity is suggested to result from a long-range
interaction between the electron-deficient rhodium and distal oxygen atoms.
Ir2(hpp)4Cl2 (Scheme 1, right), also from group 9, has been
reported to have a doubly bonded dimetallic core with two
unpaired electrons.7 Isoelectronic Ru24+ complexes are reported
to have the σ2π4δ2δ*2π*2 electronic configuration,1,8 implying
a Ru-Ru double bond and a spin triplet ground state, whereas
Introduction
The recent definitive treatise on multiple bonds between metal
atoms introduced oxidation states with the statement that “the
most common oxidation states for the M2n+ units in paddlewheel
complexes correspond to values of n of 4, 5, and 6”.1 Examples
of the n ) 6 systems are less common. Our report last year of
remarkably stable bis(phenyl)dirhodium(III) caprolactamate
6+
Ru2 compounds have the π4δ2π*4 electronic configuration
with a net δ single bond.1,9,10
Homodinuclear σ-alkynyl complexes of both ruthenium and
rhodium are known,11 although with rhodium only in the (II,III)
oxidation state.12 Electron donation from the axial carbon ligand
to rhodium increases the Rh-Rh bond distance, but only by a
fraction of the lengthening found with 1, and their electronic
properties imply retention of a metal-metal bond.
(Scheme 1, left)2 was only the second example of Rh2
6+
compounds, the previously reported Rh2(O2CR)4(NO)4 (R )
Me, Et, iPr) (Scheme 1, center) exhibiting the same structural
type and an essentially identical Rh-Rh separation.3 Although
no specific oxidation state assignment was mentioned in the
Lippard report, the bent geometry observed for the NO ligands
justifies the assignment of a formal oxidation state of III to the
Rh centers, and this was later supported by computational
The ability of paddlewheel complexes to hold together two
metals without a metal-metal bond is well-established,13,14 and
studies.4-6 One of the surprising features of these Rh2
6+
compounds is the apparent absence of a rhodium-rhodium
(4) Sizova, O. V. J. Mol. Struct. (THEOCHEM) 2006, 760, 183–187.
(5) Sizova, O. V.; Skripnikov, L. V.; Sokolov, A. Y.; Ivanova, N. V.
Russ. J. Coord. Chem. 2007, 33, 588–593.
bond. In contrast, a paramagnetic paddlewheel Ir26+ compound,
(6) Sizova, O. V.; Sokolov, A. Y.; Skripnikov, L. V.; Baranovski, V. I.
Polyhedron 2007, 26, 4680–4690.
(7) Cotton, F. A.; Murillo, C. A.; Timmons, D. J. Chem. Commun. 1999,
1427–1428.
(8) Cotton, F. A.; Ren, T.; Eglin, J. L. Inorg. Chem. 1991, 30, 2552–
2558.
* Corresponding author. E-mail: rinaldo.poli@lcc-toulouse.fr.
† Universite´ Paul Sabatier et Institut National Polytechnique de Toulouse.
§ Institut Universitaire de France.
‡ University of Maryland.
Purdue University.
(1) Cotton, F. A.; Murillo, C. A.; Walton, R. A. Multiple Bonds Between
Metal Atoms, 3rd ed.; Springer Science: New York, 2005.
(2) Nichols, J. M.; Wolf, J.; Zavalij, P.; Varughese, B.; Doyle, M. P.
J. Am. Chem. Soc. 2007, 129, 3504–3505.
(3) Hilderbrand, S. A.; Lim, M. H.; Lippard, S. J. J. Am. Chem. Soc.
2004, 126, 4972–4978.
(9) Ren, T. Organometallics 2002, 21, 732–738.
(10) Bear, J. L.; Li, Y. L.; Han, B. C.; Kadish, K. M. Inorg. Chem.
1996, 35, 1395–1398.
(11) Hurst, S. K.; Ren, T. J. Organomet. Chem. 2003, 670, 188–197.
(12) Yao, C. L.; Park, K. H.; Khokhar, A. R.; Jun, M. J.; Bear, J. L.
Inorg. Chem. 1990, 29, 4033–4039.
10.1021/om800631b CCC: $40.75
2008 American Chemical Society
Publication on Web 10/22/2008

Bis(phenyl) Dirhodium(III) Carboxamidate Compounds
Organometallics, Vol. 27, No. 22, 2008 5837
Figure 1. Structures of dirhodium pyrrolidinates (a) 1a, (b) 2a′, and (c) 3a. Ellipsoids are shown at the 30% probability level. Hydrogen
-
atoms, BF4 (for 2a′), and solvent molecules are omited for clarity.
Scheme 1. Comparison of Two Isoelectronic Group 9 Paddlewheel Dimers
the structural and electronic influences that result have substan-
tial theoretical and practical implications.15 In this contribution,
we report the molecular structures and chemical stabilities of
the diphenyl tetracarboxamidatodirhodium(III) system and their
study by XPS spectroscopy, electrochemistry, and computational
methods.
prepared their diamagnetic bis(phenyl)dirhodium(III) analogues
(3, eq 2). The synthesis of dirhodium(II,III) complexes was
achieved by oxidation of the dirhodium(II) carboxamidate using
copper(II) triflate, nitrosonium salts (eq 1),17,18 or phenyldia-
zonium tetrafluoroborate; oxidations with nitrosonium salts were
the most versatile and gave dirhodium(II,III) product without
byproducts in solution. As mentioned in the Introduction, the
synthesis (using sodium tetraphenylborate) and characterization
of bis(phenyl)dirhodium(III) caprolactamate (3c) has been
described in a preliminary communication.2 Since this first
report, a new methodology for convenient access to diaryl
analogues directly from dirhodium(II) carboxamidates and
arylboronic acids catalyzed by copper sulfate (eq 2) has been
communicated,19 and the spectral data of these products have
been described.
Results and Discussion
Synthesis and Characterization. In order to identify the
unique structural features of bis(phenyl)dirhodium(III) carboxa-
midates, we have prepared the dirhodium(II) pyrrolidinate (pyr),
valerolactamate (val), and caprolactamate (cap) bis-acetonitrile
complexes (1),16 oxidized these compounds to the corresponding
paramagnetic Rh(II)Rh(III) derivatives (2, eq 1), and also
The crystal structures of all nine compounds have been
obtained. Those of the pyrrolidinate systems Rh2(pyr)4(MeCN)2
(1a), [Rh2(pyr)4(PhCN)2]BF4 (2a′), and PhRh(pyr)4RhPh (3a)
are presented in Figure 1. Unpublished data for the other six
compounds are provided as Supporting Information. A selection
of bond lengths and angles for all nine compounds is reported
in Tables 1 and 2. As can be seen from these data, the diphenyl-
dirhodium(III) compounds exhibit severe structural distortions
relative to their dirhodium(II,II) and dirhodium(II,III) analogues.
They can best be evaluated by comparing a few key bond angles
[the values are averages over the three structures within each
family and listed in the order 1, 2, 3]: Laxial-Rh-Rh [175(2)°,
174(4)°, 159(4)°]; Rh-O-C [118(2)°, 118(2)°, 126(1)°];
(13) Mendiratta, A.; Cummins, C. C.; Cotton, F. A.; Ibragimov, S. A.;
Murillo, C. A.; Villagran, D. Inorg. Chem. 2006, 45, 4328–4330.
(14) Aakeroy, C. B.; Schultheiss, N.; Desper, J. Dalton Trans. 2006,
1627–1635.
(15) Liu, C. S.; Wang, J. J.; Yan, L. F.; Chang, Z.; Bu, X. H.; Sanudo,
E. C.; Ribas, J. Inorg. Chem. 2007, 46, 6299–6310.
(16) Rh2(pyr)4(Hpyr)2 and Rh2(val)4(Hval)2 have been prepared and their
crystal structures reported: Bear, J. L.; Lifsey, R. S.; Chau, K. L.; Ahsan,
M. Q.; Korp, J. D.; Chavan, M.; Kadish, K. M. J. Chem. Soc., Dalton Trans.
1989, 9, 3–100.
(17) Wang, Y.; Wolf, J.; Zavalij, P.; Doyle, M. P. Angew. Chem., Int.
Ed. 2008, 47, 1439–1442.
(18) For uses of nitrosonium salts as oxidants, see: Connelly, N. G.;
Geiger, W. E. Chem. ReV. 1996, 96, 877–910.
(19) Xie, J.-H.; Nichols, J. M.; Lubek, C.; Doyle, M. P. Chem. Commun.
2008, 2671–2673.

5838 Organometallics, Vol. 27, No. 22, 2008
Wolf et al.
Table 1. Selected Bond Lengths of Dirhodium Carboxamidates in
the (II,II), (II,III), and (III,III) Oxidation States
bond lengths of dirhodium complexes, Å
bond
1a
2a′a
3a
from Rh2(pyr)4 Rh1-Rh2
Rh2-O1
2.4540(5)
2.077(2)
2.080(2)
2.003(3)
2.011(3)
2.241(2)
2.4432(4)
2.030(6)
2.034(5)
1.973(6)
1.994(8)
2.248(8)
2.5738(12)
2.093(6)
2.095(5)
1.993(7)
2.012(6)
2.037(7)
Rh2-O2
Rh1-N1
Rh1-N2
Rh-Lax
bond
1b
2b′b
3b
from Rh2(val)4 Rh1-Rh2
2.4187(5)
2.061(2)
2.073(2)
2.023(3)
2.034(3)
2.254(3)
2.3872(3)
2.0143(16)
2.0211(16)
2.0078(19)
2.016(2)
2.3071(16)
2c
2.5162(5)
2.105(3)
2.061(2)
1.997(3)
2.027(3)
1.996(3)
Rh2-O1
Rh2-O2
Rh1-N1
Rh1-N2
Rh-Lax
bond
1c
3c
from Rh2(cap)4 Rh1-Rh2
2.4221(4)
2.050(2)
2.053(2)
2.038(3)
2.045(3)
2.336(3)
2.3840(6)
2.023(2)
2.030(2)
1.988(3)
1.996(3)
2.288(2)
2.5188(3)
2.0825(16)
2.0777(16)
2.012(2)
2.0078(19)
2.010(2)
Rh2-O1
Rh2-O2
geometrical distortions, there is a significant lengthening of the
Rh-Rh bond on going from 2 to 3, whereas the bond shortens
on going from 1 to 2.
Rh1-N1
Rh1-N2
Rh-Laxial
a The bis-benzonitrile complex was prepared after failed attempts to
prepare the bis-acetonitrile complex. b The bis-aquo complex was
formed and crystallized, even though reaction was performed in
acetonitrile.
X-ray photoelectron spectroscopy (XPS) data for these nine
compounds are given in Table 3. As evidenced by this direct
measurement of the binding energy of d-electrons that
constitute the dirhodium core, there is a significant change
in binding energy upon oxidation of dirhodium(II,II) to
dirhodium(II,III) (∼0.8 eV) but an insignificant change upon
oxidation of dirhodium(II,III) to dirhodium(III,III) (<0.2 eV).
Table 2. Selected Bond Angles for Dirhodium Carboxamidates in
the (II,II), (II,III), and (III,III) Oxidation States
bond angles of dirhodium complexes, deg
The relative electron richness of the diphenyldirhodium(III)
carboxamidate complexes is also suggested by the results of an
electrochemical investigation. Cyclic (CV) and differential pulse
(DPV) voltammograms of 3a, 3b, and 3c were measured in
CH2Cl2, and their half-wave potentials (E1/2) are reported in
Table 4, along with their electronic absorption maxima (λmax).
angle
1a
2a′a
3a
from Rh2(pyr)4 Rh1-Rh2-O1
Rh1-Rh2-O2
Rh2-Rh1-N1
Rh2-Rh1-N2
Rh2-O1-C1
90.04(6)
90.01(6)
85.30(7)
85.35(7)
115.16(19)
115.9(2)
123.40(2)
124.00(2)
89.66(16)
89.94(15)
85.9(2)
86.15(18)
115.8(5)
117.7(5)
122.4(5)
123.7(5)
178.22(16)
2b′b
89.61(5)
89.66(4)
87.11(5)
87.23(5)
118.89(14)
119.08(14)
121.10(16)
122.07(16)
172.27(5)
2c
88.58(7)
88.50(7)
88.42(8)
88.60(8)
119.8(2)
119.5(2)
120.8(2)
121.2(2)
171.28(7)
82.43(15)
83.87(14)
89.32(18)
90.61(18)
121.1(4)
122.4(5)
118.0(5)
118.5(5)
164.20(20)
3b
74.58(7)
85.05(7)
99.10(8)
88.47(8)
133.0(2)
122.1(2)
121.8(4)
119.8(2)
157.34(11)
3c
79.29(4)
78.39(5)
94.46(5)
95.32(5)
128.11(15)
128.97(15)
114.45(16)
113.51(15)
156.19(7)
Rh2-O2-C2
Figure 3 shows the CV and DPV traces for PhRh(pyr)4RhPh
Rh1-N1-C1
6+
Rh1-N2-C2
(3a). For each of the complexes, a single, reversible Rh2
/
Rh1-Rh2-Laxial 176.90(7)
Rh27+ redox couple (A) was observed. The oxidation potential
angle
from Rh2(val)4 Rh1-Rh2-O1
Rh1-Rh2-O2
1b
of 3 increases with decreasing number of methylene units in
88.03(7)
88.85(7)
87.08(8)
87.89(8)
118.8(2)
119.3(2)
121.5(2)
121.8(2)
4+/5+
the bridging cyclic amide, which follows that of the Rh2
couple of the parent complexes.20 For complex 3a, an additional
Rh2-Rh1-N1
Rh2-Rh1-N2
irreversible redox couple (B) near the reduction limit of the
Rh2- O1-C1
5+/6+
solvent was observed and is assigned to the Rh2
redox
Rh2-O2-C2
Rh1-N1-C1
couple. Attempts to isolate and characterize the oxidized
PhRh(pyr)4RhPh or any of its derivatives have thus far been
unsuccessful.
Rh1-N2-C2
Rh1-Rh2-Laxial 173.84(8)
angle
from Rh2(cap)4 Rh1-Rh2-O1
Rh1-Rh2-O2
1c
89.62(6)
89.37(6)
86.70(7)
86.81(7)
118.3(2)
118.96(19)
121.5(2)
121.7(2)
The stabilities of bis(phenyl)dirhodium(III) carboxamidates
are noteworthy. They are air stable and thermally stable to
temperatures below 130 °C. They do not react with nucleo-
philes such as cyanide or azide, but they are sensitive to
electrophilic attack. Treatment of compound bis-σ-(6-meth-
oxynaphthalen-2-yl)-tetrakis-µ-(caprolactamato)dirhodium(I-
II)19 with trifluoroacetic acid, for example, resulted in
quantitative conversion to 2-methoxynaphthalene (eq 3) and
caprolactam (capH). Product analysis as a function of time
with bis-σ-(phenyl)-tetrakis-µ-(caprolactamato)dirhodium(III)
demonstrated the stability of the phenyl-rhodium(III) bond
(Figure 4); caprolactam was released from the complex prior
to release of benzene, but attempts to isolate the anticipated
diphenyl derivative of dirhodium(III) trifluoroacetate were
unsuccessful. Although bis-σ-(phenyl)-tetrakis-µ-(caprolac-
tamato)dirhodium(III) was stable as a solid to temperatures
above 200 °C with a measured decomposition point at 256
°C, when this compound was heated in refluxing chloroben-
zene over long periods of time, biphenyl as well as mono-
Rh2-Rh1-N1
Rh2-Rh1-N2
Rh2-O1-C1
Rh2-O2-C2
Rh1-N1-C1
Rh1-N2-C2
Rh1-Rh2-Laxial 174.42(7)
a The bis-benzonitrile complex was prepared after failed attempts to
prepare the bis-acetonitrile complex. b The bis-aquo complex was
formed and crystallized, even though reaction was performed in
acetonitrile.
Rh-N-C [122(1)°, 122(1)°, 118(1)°]. These distortions take
the form of severe bending of the Rh-Rh-Ph bond axis from
180°, a major opening of the Rh-O-C angle, and a less
pronounced closing of the Rh-N-C angle from the ideal 120°.
The positioning of the rhodium units in the diphenyl-dirhodium
structures appears as if two square-pyramidal units have
undergone conrotatory motion so as to minimize repulsions and
metal-metal interactions (Figure 2). In addition to these

Bis(phenyl) Dirhodium(III) Carboxamidate Compounds
Organometallics, Vol. 27, No. 22, 2008 5839
Figure 2. ORTEP views of (a) PhRh(pyr)4RhPh (3a), (b) PhRh(val)4RhPh (3b), and (c) PhRh(cap)4RhPh (3c) showing comparable structural
distortions. Ellipsoids are shown at the 30% probability level. Hydrogen atoms and solvent molecules are omitted for clarity. The corresponding
Rh-Rh-C bond angles are 164.2° (3a), 157.3° (3b), and 156.2° (3c).
Table 3. XPS Data for Dirhodium Carboxamidates in the (II,II),
(II,III), and (III,III) Oxidation Statesa
compound
Rh-3d5/2, eV
Rh-3d3/2, eV
Rh2(pyr)4(MeCN)2 (1a)
Rh2(val)4(MeCN)2 (1b)
Rh2(cap)4(MeCN)2 (1c)
Rh2(pyr)4(PhCN)2BF4 (2a′)
Rh2(val)4(H2O)2BF4 (2b′)
Rh2(cap)4(MeCN)2BF4 (2c)
PhRh(pyr)4RhPh (3a)
308.000
308.129
308.080
308.905
308.841
309.094
309.045
309.011
309.105
312.563
312.843
312.812
313.514
313.567
313.718
313.739
313.723
313.725
PhRh(val)4RhPh (3b)
PhRh(cap)4RhPh (3c)
compounds.21,22 The B3LYP functional provides the best
agreement between experimental and calculated structures,1 and
it has therefore been adopted in this study. The study used the
standard LANL2DZ basis set in order to provide a qualitative
description of the electronic structure. A model system was
constructed with the amidato ligands being replaced with simple
formamidato ligands, [OCCHNH]-.
a Calibrated to carbon 1s at 284.6 eV.
Table 4. Electrode Potentials (V) from CV Measurements and λmax
of 3
compound
A (V)a
B (V)b
λmax (nm)c
3a
3b
3c
0.91
0.81
0.78
-1.36
442
429
430
(a) [Rh2(amidato)4], [Rh2(amidato)4(MeCN)2], and [Rh2(ami-
dato)4(MeCN)2]+ Systems. Before analyzing the more interesting
bis(aryl)dirhodium(III,III) derivatives, it is necessary to start with
the dirhodium(II,II) and dirhodium(II,III) precursors. These
families of compounds have already been extensively studied;
thus our analysis is limited to a comparison of the optimized
and experimental geometries and to a comparison of the basic
electronic structural features with those previously presented
for other members of the same families23-30 (fuller details are
available as Supporting Information). Geometry optimizations
were carried out for the simple tetrakis(amidato)dirhodium(II,II)
system (I in Scheme 2), for its bis(acetonitrile) adduct (II), and
for the one-electron oxidation product of the latter (III). All
these calculations were carried out only with the experimentally
observed head-head-tail-tail (cis-2,2)20 relative orientation of
the four bridging amidato ligands (see Scheme 2). The salient
features of the optimized geometries are shown in Table 5. There
is excellent agreement between optimized and experimental
parameters, attesting to the suitability of the computational
a Reversible anodic wave reported as E1/2, measured with platinum
wire/glassy carbon working electrodes against a Ag/AgCl reference
electrode in CH2Cl2 with tetra-n-butylammonium hexafluorophosphate as
the supporting electrolyte. b Irreversible cathodic reduction reported as
cathodic peak potential (Epc) in parentheses. c Measured in CH2Cl2.
(20) Doyle, M. P.; Ren, T. Prog. Inorg. Chem. 2001, 49, 113–168.
(21) Cotton, F. A.; Feng, X. J. Am. Chem. Soc. 1997, 119, 7514–7520.
(22) Cotton, F. A.; Feng, X. J. J. Am. Chem. Soc. 1998, 120, 3387–
3397.
(23) Norman, J. G., Jr.; Kolari, H. J. J. Am. Chem. Soc. 1978, 100, 791–
799.
Figure 3. Cyclic (CV) and differential pulse voltammetry (DPV)
measurements of 3a recorded in 0.2 M CH2Cl2 solution of Bu4NPF6
at a scan rate of 0.10 V/s.
(24) Norman, J. G., Jr.; Renzoni, G. E.; Case, D. A. J. Am. Chem. Soc.
1979, 101, 5256–67.
and dichlorobiphenyl are formed, suggestive of homolytic
cleavage of the rhodium-phenyl bond.
(25) Bursten, B. E.; Cotton, F. A. Inorg. Chem. 1981, 20, 3042–8.
(26) Cotton, F. A.; Feng, X. J. Inorg. Chem. 1989, 28, 1180–1183.
(27) Kawamura, T.; Maeda, M.; Miyamoto, M.; Usami, H.; Imaeda, K.;
Ebihara, M. J. Am. Chem. Soc. 1998, 120, 8136–8142.
(28) Mougenot, P.; Demuynck, J.; Benard, M. Chem. Phys. Lett. 1987,
136, 279–282.
(29) Nakatsuji, H.; Ushio, J.; Kanda, K.; Onishi, Y.; Kawamura, T.;
Yonezawa, T. Chem. Phys. Lett. 1981, 79, 299–304.
(30) Nakatsuji, H.; Onishi, Y.; Ushio, J.; Yonezawa, T. Inorg. Chem.
1983, 22, 1623–1630.
Computational Study. The structural novelty of the
bis(aryl)dirhodium(III) family and their peculiar properties
(metal-metal bond length, stability) in comparison with iso-
electronic systems (see Introduction) warrant a more detailed
analysis of their electronic structure. DFT methods have proven
quite successful when applied to dinuclear metal-metal bonded

5840 Organometallics, Vol. 27, No. 22, 2008
Wolf et al.
Figure 4. Reaction of PhRh(cap)4RhPh (3.7 µmol) with (a) 10 and (b) 15 molar equiv of trifluoroacetic acid in CDCl3 at room temperature.
Scheme 2. Tetraformamidato Systems Used for the DFT Calculations
Table 5. Selected DFT-Optimized Geometric Parameters (distances in Å, angles in deg) for the Model Compounds I-III and Comparison with
the Experimental Geometries
system II
system III
parameter
(a) Distances
system I
opt
exptla
opt
exptlb
Rh-Rh
Rh-N (amidato)
Rh-O
Rh-N(acetonitrile)
(b) Angles
2.454
2.047
2.085
2.497
2.039
2.112
2.286
2.43 (2)
2.03 (2)
2.066 (4)
2.28 (5)
2.481
2.011
2.064
2.285
2.40 (3)
2.00 (2)
2.025 (7)
2.28 (3)
Rh-Rh-N (amidato)
87.15
89.25
86.36
88.80
176.61
87 (1)
89.3 (7)
175 (2)
86.95
88.32
177.54
87 (1)
89.3 (2)
174 (4)
Rh-Rh-O
Rh-Rh-N (acetonitrile)
a Averages over all chemically equivalent parameters of compounds 1a, 1b, and 1c. b Averages over all chemically equivalent parameters of
compounds 2a′, 2b′, and 2c.
method and of the chosen model. The Rh-Rh distance trend is
perfectly reproduced by the calculation. The very minor
elongation (0.045 Å) computed on going from I to II agrees
with the notion that this bond is rather insensitive to the presence
ofaxialligands,anexperimentalexamplebeingRh2(O2C-C6H2iPr3-
2,4,6)4 and its bis(acetone) adduct (distances of 2.350(1) and
2.370(1) Å, respectively).31,32
The specific Rh-Rh σ, π, and δ* interactions are symmetry-
mixed with symmetry-adapted linear combinations (SALC) of
the Rh-O and Rh-N bonds. The effect of this mixing is that
each of these interactions is mainly distributed in two molecular
orbitals. Within each pair of orbitals, the one carrying the greater
Rh-Rh character is that at lower energy for σ and π and that
at higher energy for δ*. From the energetic point of view, the
major Rh-Rh σ MO is found at higher energy than the major
Rh-Rh π MO, and the major Rh-Rh δ* MO is higher than
the Rh-Rh π* MO. Upon MeCN axial ligation, the σ
competition between the Rh-NCMe and Rh-Rh interactions
has the effect of raising the Rh-Rh σ level. Weak axial ligands
(e.g., H2O) establish weak interactions with the Rh atoms and
weakly affect the Rh-Rh σ MO energy,23 whereas stronger
ligands (e.g., phosphines) have a more pronounced effect and
greatly raise the energy of the Rh-Rh σ MO to the point that
this may even become the HOMO.25 In the present case, the
interaction is relatively weak but sufficient to push the Rh-Rh
σ MO higher than the Rh-Rh δ MO but lower than the Rh-Rh
π* MOs. Upon MeCN ligation, the energy of all orbitals is
raised by ca. 0.4 eV on average, reflecting the additional electron
density provided to the Rh atoms by the axial ligands. On the
other hand, upon one-electron oxidation, all orbitals drop by
ca. 1.5 eV on average as a result of the increased effective
positive charge, but the MO energy ordering remains ap-
proximately unchanged (the σ level remains between the δ and
π* levels). For all three systems, the HOMO (doubly occupied
for I and II, singly occupied for III) is the Rh-Rh δ* MO.
(b) [Rh2(amidato)4Ph2] System. A geometry optimization
was carried out on the model system Rh2(OCHNH)4(C6H5)2 with
the same (cis-2,2) arrangement as in the I-III congeners (IV,
see Scheme 2). The input geometry was constructed directly
from the experimentally observed geometry of compound 3c.
Since this geometry is very close to an ideal C2h symmetry,
that symmetry was imposed in the initial geometry optimization.
(31) Cotton, F. A.; Hillard, E. A.; Murillo, C. A. J. Am. Chem. Soc.
2002, 124, 5658–5660.
(32) Cotton, F. A.; Hillard, E. A.; Liu, C. Y.; Murillo, C. A.; Wang,
W. N.; Wang, X. P. Inorg. Chim. Acta 2002, 337, 233–246.

Bis(phenyl) Dirhodium(III) Carboxamidate Compounds
Organometallics, Vol. 27, No. 22, 2008 5841
Figure 5. Two mutually perpendicular views of the geometries, and relative energies, of system IV obtained under three different conditions:
(i) constrained optimization in C2h symmetry; (ii) full optimization without symmetry constraints; (iii) idealized geometry with linear
Ph-Rh-Rh-Ph moiety.
Table 6. Selected DFT-Optimized Geometric Parameters (distances
in Å, angles in deg) for the Model Compound IV and Comparison
with the Experimental Geometries
system IV
(no symmetry)
system IV
(C2h symmetry)
parameter
expta
(a) Distances
Rh-Rh
2.644
2.639
2.022
2.138
1.999
2.54 (3)
2.01 (2)
2.09 (2)
2.01 (2)
Rh-N
2.025, 2.053
2.095, 2.133
1.999
Rh-O
Rh-C
(b) Angles
Rh-Rh-N
Rh-Rh-O
Rh-Rh-C
90.22, 81.83
89.75, 81.50
166.92
90.73
81.13
166.44
93 (1)
81 (1)
159 (4)
a Averages over all chemically equivalent parameters of compounds
3a, 3b, and 3c.
The converged geometry (Figure 5, i) remained very close to
the initial one, but the frequency analysis revealed one imaginary
frequency whose motion showed the tendency for the two
Rh-Ph axes to move in opposite directions away from the
symmetry plane. Therefore, a free optimization in C1 symmetry
was subsequently carried out. The resulting geometry (Figure
5, ii) was essentially Ci-symmetric, with the expected 3N-6
positive vibrational frequencies, and with a coplanar arrangement
for the two Ph groups. The energy of the global minimum is
just 0.2 kcal/mol lower than the C2h structure. The ideal
geometry where the Ph-Rh-Rh-Ph moiety was forced to be
coplanar and linear (Figure 5, iii), on the other hand, turned
out to be energetically less favored by 8.2 kcal/mol. These
combined results indicate that the origin of the experimentally
observed distortion of the Rh-Ph bonds is not in the steric
interaction between the Ph groups and the amidato ligands, but
rather in an electronic effect that forces the three-bond
Ph-Rh-Rh-Ph sequence out of colinearity. Of the two
different and nearly isoenergetic distortion modes (i) and (ii),
the first one is adopted in the crystal structure for possible
packing reasons.
Figure 6. Simplified interaction diagram for system IV, built from
the Rh2(OCHNH)4 fragment (V) and two phenyl radicals.
coordination geometry, yielding alternating long-short Rh-N
and Rh-O distances. The longer distances are associated with
the smaller Rh-Rh-X (X ) N, O) angles.
The electronic structure analysis helps rationalize the mo-
lecular constitution and geometry of the novel dinuclear RhIII
complexes. The essential features are reported in Figure 6, while
fuller details are presented in the Supporting Information. The
relevant MOs for IV derive from the full geometry optimization,
while those of the Rh2(OCHNH)4 fragment (V) and the Ph
radicals are calculated for the corresponding fixed geometries
Selected optimized geometric parameters for system IV are
shown in Table 6. Like for systems II and III, the optimized
structure of IV (C2h symmetry) agrees very well with the
observed structures. Except for the Rh-Rh distance (ca. 0.1 Å
longer in the optimized geometry), all distances are in very close
agreement with the experimentally determined structure. A
significant difference is observed in the Rh-Rh-Ph angle (ca.
10° smaller in the X-ray structures), presumably because of
crystal-packing effects. The distortion in the lower-energy Ci
structure is different. It causes a slight asymmetry in the Rh

5842 Organometallics, Vol. 27, No. 22, 2008
Wolf et al.
Rh-Rh interaction in this molecule can therefore be formally
described as π4δ2π*4δ*2, leading to a formal bond order of zero.
Using a basic, qualitative valence bond approach, we can
describe the molecule as two interpenetrating octahedra, using
the familiar d2sp3 hybrids for bonding. This would lead to the
establishment of all three σ interactions (two Rh-Ph bonds and
the Rh-Rh bond) by implication of the higher energy 5s and
5pz orbitals for the two rhodium atoms. Obviously, this
implication is not extensive enough to keep the third bonding
orbital sufficiently low in energy to be occupied. We note that
the involvement of the higher energy s orbital has been
invoked33 to rationalize the observed short distance of paddle-
wheel d3-d3 W2(O2CEt)4R′2 compounds (E ) Et and R′ )
CH2Ph or CH2tBu),34 which are able to maintain their W-W
σ interaction as in the parent d4-d4 W2(O2CEt)4 compound.
The involvement of the higher energy orbitals seems a more
common feature for 5d elements, as a result of relativistic effects.
The effect of the σ interaction scheme on the electronic structure
is such that the HOMO of IV is the same orbital, the δ* orbital,
as the HOMO of I. The two are essentially identical because
the Ph groups do not significantly contribute. This agrees with
the observed reversible electrochemical behavior upon oxidation:
the electron is pulled off the essentially nonbonding δ* orbital;
thus the Rh27+ oxidation product is predicted to have essentially
no Rh-Rh bonding interaction and to display a structure closely
Figure 7. MOLDEN contour plots of relevant orbitals for system
IV.
as found in IV. The frontier orbitals of interest for the Ph radical
are the filled, familiar Huckel-type π1, π2, and π3 orbitals, plus
the σ orbital pointing out from the unsaturated carbon in a radial
direction. The latter is expected to be used for the formation of
the RhIII-Ph σ bond.
Upon interaction between V and the two Ph radicals, we
observe that those molecular orbitals that are not involved in a
significant interaction change their energy only slightly. More
specifically, there is a slight energy increase for the phenyl
orbitals 17 (π1) and 20 (π3) and a slight energy decrease for V
orbitals 51 and 52 (Rh-Rh π), 59 (Rh-Rh δ), and 63 (Rh-Rh
δ*). Inspection of the contour plots of the related orbitals in
system IV (see Supporting Information) shows that they do not
involve a significant mixing of components from the different
fragments.
The greater energy change experienced by the Ph π2 orbitals
is caused by the combination with the appropriately oriented
Rh-Rh π and π* orbitals of V, resulting in Rh-Ph π-type
interaction. However, since these are filled-filled interactions
(both the in-phase and out-of-phase combinations are filled),
there is no net Rh-Ph π bonding. The interaction is stronger
(greater mixing and energy changes) for the Rh-Rh π* orbital
because of a better energy match. The in-phase and out-of-phase
combinations are found in orbitals 94 and 103, while the
corresponding combinations with the Rh-Rh π MO are found
in orbitals 93 and 102.
The most important interaction between V and the two Ph
radicals results from the combination of the σ-type orbitals.
There are four such orbitals in the frontier region, two for
Rh2(OCHNH)4 (the filled Rh-Rh σ orbital, mixed with a
nonbonding ligand-based combination in MOs 54 and 55, Vide
supra, and the empty Rh-Rh σ* orbital, MO 64) and two for
the two Ph radicals (the half-filled frontier σ orbital, MO 21).
This gives rise to two filled MOs (86 and 91) and to two empty
MOs (105 and 108) in IV, all of which are shown in Figure 7
(a more extensive sample of MO contour diagrams for IV is
presented in the Supporting Information). The two filled orbitals
have mostly Rh-Ph σ character but also feature other contribu-
tions: the former has substantial Rh-Rh and C-C/C-H σ
contributions, whereas the latter has Rh-Rh σ* character. Thus,
very little Rh-Rh σ bonding is present in the molecule. Most
of the Rh-Rh σ interaction is found in the empty MO 105,
which also has significant Rh-Ph σ* character and is therefore
raised above the energy of the Rh-Rh δ* orbital. Finally, MO
108 has both Rh-Rh and Rh-Ph σ* character (Figure 7). The
6+
related to that of its Rh2 precursor.
One remaining point of interest is the origin of the nonlinear
arrangement of the Ph-Rh-Rh-Ph moiety. One possible cause,
related to the Ph-Rh-Rh-Ph σ bonding scheme, is a second-
order Jahn-Teller effect involving the filled Rh-Ph and the
empty Rh-Rh σ orbitals. However, this idea is contradicted
by the higher energy of the bonding orbitals 86 and 91 in the
optimized minimum relative to the idealized linear geometry
(iii in Figure 5, orbital energies available in the Supporting
Information). This suggests that the Ph-Rh-Rh-Ph σ-bonding
framework prefers the linear configuration. The relative weak-
ness of this interaction and its symmetry-mixing with other
bonding features, however, do not allow its clear identification
in the molecular orbitals. Therefore, the reason for the molecular
distortion of 3 out of linearity must be attributed to other factors.
A possible clue comes from the observation of very narrow
Rh-Rh′-O bond angles in 3 (less than 90°), suggesting the
presence of an attractive interaction between Rh and the O atom
bonded to Rh′. The through-space crystallographic distances
between rhodium and its distal oxygens are 3.00 and 3.24 Å
for 3a, 2.82 and 3.11 Å for 3b, and 2.90 and 2.97 Å for 3c.
Given the absence of direct Rh-Rh bonding, the Rh atoms are
in fact electronically unsaturated (16-electron configuration) and
therefore seek electron density from the distal O lone pairs.
However, we note that an alkylrhodium(III) compound with a
nearly identical coordination geometry, the O2N2 coordination
environment being provided by a Schiff base ligand, maintains
a mononuclear five-coordination geometry with an open coor-
dination site.35
(c) [Rh2(O2CCH3)4(NO)2] System. Compounds Rh2(O2CR)4-
(NO)2 (R ) Me, Et, iPr), featuring the same paddlewheel
structure as compounds 3 with axial NO ligation and a very
similar Rh-Rh distance, have been recently described.3 The
(33) Chisholm, M. H.; Clark, D. L.; Huffman, J. C.; Van Der Sluys,
W. G.; Kober, E. M.; Lichtenberger, D. L.; Bursten, B. E. J. Am. Chem.
Soc. 1987, 109, 6796–6816.
(34) Chisholm, M. H.; Hoffman, D. M.; Huffman, J. C.; Vandersluys,
W. G.; Russo, S. J. Am. Chem. Soc. 1984, 106, 5386–5388.
(35) Anderson, D. J.; Eisenberg, R. Inorg. Chem. 1994, 33, 5378–5379.

Bis(phenyl) Dirhodium(III) Carboxamidate Compounds
Organometallics, Vol. 27, No. 22, 2008 5843
Figure 9. Contour plots of relevant orbitals for system VI.
diphenyl system, the situation is complicated by the greater
number of participating orbitals [the Rh-Rh σ and σ* orbitals
from fragment V, plus the N lone pair and the N-O π| and π|*
orbitals from each of the two NO ligands, for a total of eight
resulting MOs]. The four lowest energy combinations (orbitals
51, 52, 61, and 65 in Figure 8) have their major contribution
from the Rh-N σ and N-O π| bonds. Contour diagrams of
the other four orbitals are shown in Figure 9 (other relevant
orbitals are shown in the Supporting Information). They are
qualitatively similar to those of system IV, including their
occupation. Thus, as with the diphenyl formamidato system,
the formation of the two Rh-NO bonds essentially destroys
the Rh-Rh interaction, since the Rh-Rh σ* component that
was empty in the dinuclear RhII fragment becomes now engaged
in the axial σ-bonding system. That there is no N-Rh-Rh-N
distortion in the nitrosyl complex is likely due to the symmetric
nature of the three-atom bridges; since oxygen atoms constitute
both sides, the molecule remains symmetric.
(d) Comparison of Electronic Structures. A comparison
between the two MO diagrams of IV in Figure 6 and VI in
Figure 8 shows that those orbitals with the largest Rh contribu-
tion, and not involved in the σ interactions (namely the π, π*,
δ, and δ* orbitals), have lower energies for system VI than IV.
This is rationalized by the more electronegative nature of the
carboxylato ligands relative to the amidato ligands. The most
important difference between the two systems, however, is in
the σ manifold. For IV, the Rh-Rh(σ*) of the Rh2(OCHNH)4
fragment interacts strongly with the Ph donor orbitals, yielding
a stabilized bonding orbital and leaving the Rh-Rh(δ*) orbital
as the HOMO. The unexpectedly low observed increase in
binding energy for the d electrons (XPS data in Table 3) upon
going from compounds 2 to compounds 3 is consistent with
the relatively small change of electronic structure in the frontier
region and with the reversible oxidative electrochemistry (Vide
supra).
Figure 8. Simplified interaction diagram for system VI, built from
Rh2(O2CCH3)4 and two nitrosyl radicals.
electronic structure of these compounds (with R ) H and
CH3) has also been reported.4-6 Here, we wish to com-
pare the electronic structures of the system Rh2(O2CCH3)4-
(NO) (VI) and IV. For this purpose, the geometry of VI has
been optimized at the same computational level as IV. The
orbital interaction diagram for VI with respect to its parent
Rh2(O2CCH3)4 and NO fragments is shown in Figure 8, while
fuller details of the electronic structure of VI are presented in
the Supporting Information.
The axial donor, in this case, mostly interacts with the dimetal
system via the N lone pair (which is mixed with the O lone
pair in MO 7). However, because of the Rh-N-O bending,
the NO π| and π|* orbitals can also contribute to the Rh-Rh
and Rh-N σ-bonding scheme. The perpendicular N-O π
orbitals (π and π *) have the correct symmetry to mix with
the Rh-Rh π and π* components of the Rh2(O2CCH3)4
fragment, much like what happens in structure IV for the phenyl
π2 orbitals (Vide supra).
Going from the fragments to the full system, as with the
formamidato system discussed above, the metal-metal π, δ,
π*, and δ* orbitals are not significantly affected. The N-O π
orbitals can potentially establish Rh-N π interactions but,
because of their relatively deep energy, do not significantly mix
with the metal orbitals and therefore are also little effected by
the axial coordination. As for the amidato system, the most
important change concerns the σ-type orbitals. Relative to the
On the other hand, the interaction between the Rh-Rh(σ*)
orbital of the Rh2(O2CCH3)4 fragment and the orbitals of the
NO axial donors is not nearly as strong and yields a MO at
relatively high energy, the HOMO of system VI. The nature of
this orbital is Rh-Rh(σ*) and Rh-N(σ); see Figure 9, thus
oxidation of this compound may be expected to lead to rupture
of the Rh-NO interactions and to the establishment of a Rh-Rh
bond. Thus, although systems IV and VI have a closely related

5844 Organometallics, Vol. 27, No. 22, 2008
Wolf et al.
MeCN/H2O or MeOH/H2O (9:1) as solvent in the presence of 1%
triflic acid. Elemental analyses were carried out by the Service
d’Analyse du Laboratoire de Chimie de Coordination in Toulouse.
Dirhodium(II) acetate, carboxamidate ligands, and reagents de-
scribed in this report were obtained commercially.
electronic structure and very similar bonding parameters, their
chemical behavior toward oxidation should be drastically
different.
It is also interesting to compare the electronic structure of
6+
these two systems with that of the isoelectronic Ir2 complex
Tetrakis-µ-(pyrrolidinato)bis(acetonitrile)dirhodium(II) (1a).16 In
a dry flask under N2 were mixed Rh2(OAc)4 (500 mg, 1.31 mmol),
2-pyrrolidinone (2.3 g, 27.0 mmol), and anhydrous chlorobenzene
(40 mL). The flask was equipped with a Soxhlet apparatus with a
thimble containing a 2:1 NaHCO3/sand mixture (5 g). The solution
was kept under N2 and stirred at reflux for 24 h, after which the
solution was cooled and solvent was removed under reduced
pressure. The residue was washed with diethyl ether (2 × 50 mL)
and pentane (50 mL). The purple solid obtained was heated at 80
°C in acetonitrile (15 mL), and some methanol was added to
dissolve the entire solid. The mixture was crystallized at 20 °C to
give a pink-red solid. Yield: 565 mg (80%). Suitable X-ray quality
crystals (pink-red prisms) were obtained by slow evaporation of
Ir2(hpp)4Cl2 (shown on the right in Scheme 1). No computational
study on this system has been presented as yet, to the best of
our knowledge. There is little doubt, however, that the general
features of the electronic structure will parallel those presented
above for systems IV and VI. The question of importance is
the strength of the σ interaction between the axial ligands and
the metal-metal framework. For the electronegative and spheri-
cally symmetrical Cl atom, the four resulting MOs (analogous
to 86, 91, 105, and 108 for system IV or to 87, 94, 96, and 100
for system VI) would certainly end up at lower energy relative
to the block of metal-metal orbitals of π and δ symmetry. The
stronger relativistic effect for the 5d Ir element should also
contribute to a stabilization of the orbitals of M-M σ symmetry
through a greater involvement of the metal s orbital. In case
the second highest of these four orbitals (having M-M σ and
M-Lax σ* contributions) is stabilized to the point of falling
below the energy of the HOMO, then the molecule will indeed
have a formal metal-metal double bond, because the two
electrons needed to fill the M-M σ orbital originate from a
previously filled antibonding orbital of π* or δ* type. If δ* is
located below π*, the molecule will display two unpaired
electrons, as is indeed observed for compound Ir2(hpp)4Cl2.
1
diethyl ether from an acetonitrile solution. H NMR (400 MHz,
CD3CN + CD3OD): δ 3.55 (br, N-CH2), 3.44 (m, N-CH2), 2.34 (t,
3J ) 7.6 Hz, 8H, CH2-C), 1.98 (q, 3J ) 7.6 Hz, 8H, N-CH2-CH2).
13C NMR (100 MHz, CD3CN + CD3OD): δ 186.89 (s, NCO),
55.41 (s, N-CH2), 33.12 (s, CH2-C), 21.91 (s, N-CH2-CH2). MS
(ESI), m/z, (%): 584.01, (100) [Rh2C18H28N5O4]+ ) M(MeCN)+;
625.04, (28) [Rh2C20H31N6O4] ) M(MeCN)2+. UV/visible spectrum
(MeCN + MeOH, one drop), λ (ꢀ, M-1 cm-1): 350 nm (224); 505
nm (235). XPS: Rh-3d a 5/2 308.000 eV; Rh-3d b 5/2 309.082
eV; Rh-3d a 3/2 312.563 eV; Rh-3d b 3/2 313.648 eV.
Tetrakis-µ-(pyrrolidinato)bis(acetonitrile)dirhodium(II,III) Tet-
rafluoroborate (2a). A solution of NOBF4 (22.3 mg, 191 µmol) in
anhydrous acetonitrile (9 mL) was slowly added (over 3 min) to a
suspension of 1a (108.5 mg, 174 µmol) in CH2Cl2 (25 mL). The
suspension rapidly turned from purple to dark red-purple. After
complete addition the mixture was stirred for an additional 30 min
at room temperature. Solvents were removed under reduced
pressure, and the residue was dissolved in CH2Cl2 (5 mL), then
filtered. The filtrate was concentrated (ca. 2-3 mL), precipitated
in diethyl ether (50 mL), filtered, and dried under vacuum. A dark
red-violet solid was obtained (117.2 mg, 94.8%), but the bis-
acetonitrile complex did not form crystals suitable for X-ray
analysis. Suitable X-ray crystals (black plate) of the bis(benzonitrile)
adduct, 2a′, were obtained after an axial ligand exchange (by heating
the product 5 min at 60 °C in a small amount of benzonitrile) and
a layer diffusion of diethyl ether in a 2:1 benzonitrile/CH2Cl2
solution. A satisfactory elemental analysis for this compound could
not be obtained, presumably because of the facile axial ligand
exchange. NMR silent; HRMS (ESI) calcd for Rh2C18H27N5O4
(M(MeCN)+) 583.01731, found 583.01516. MS (ESI), m/z, (%):
574.1, (100) [Rh2C17H28N4O5]+ ) M(MeOH)+; 542.0, (24)
[Rh2C16H24N4O4]+ ) M+; 87.0, (100) [BF4]-. UV/visible spectrum
(CH2Cl2), λ (ꢀ, M-1 cm-1): 489 nm (4150); 530 nm (-); 1030 nm
(1440). XPS: Rh-3d a 5/2 308.905 eV; Rh-3d b 5/2 310.722 eV;
Rh-3d a 3/2 313.514 eV; Rh-3d b 3/2 315.069 eV.
Conclusion
The ability of a phenyl group to impart unusual stability to
rhodium(III) compounds is explained by theoretical analysis of
the electronic structure of bis-σ-(phenyl)-tetrakis-µ-(carboxa-
mido)dirhodium(III). Formally, two phenyl radicals in combina-
tion with a rhodium-rhodium bond are transformed into two
phenyl-rhodium bonds. The absence of a rhodium-rhodium
σ-bonding interaction is suggested in the lengthening of the
rhodium-rhodium distances from those of dirhodium(II,II) and
dirhodium(II,III) complexes, as well as from XRD data and is
confirmed by DFT analysis. The stability of the phenyl-rhodium
bonds is seen in the preferential removal of equatorial carboxa-
midate ligands by trifluoroacetic acid and in the homolytic loss
of phenyl radical from 3 at high reaction temperatures. The
observed severe distortion of phenyl rings from linearity is
suggested to have its origin in long-range interactions between
the electron-deficient rhodium and the distal oxygen atoms.
n+
These structural characteristics are unique among M2 com-
pounds, and they suggest intriguing potential for future structural
revelations.
Experimental Section
Bis-σ-(phenyl)-tetrakis-µ-(pyrrolidinato)dirhodium(III) (3a). To a
MeOH/CH2Cl2 solution (5 mL, 1:1) of 2a (30.5 mg, 42.9 µmol)
were added NaBPh4 (73.4 mg, 214.5 µmol) and CuOTf (1.5 mg,
7.1 µmol). The initial purple solution rapidly turned to yellow-
green, and the mixture was stirred at room temperature for 18 h.
During this period a yellow-green solid precipitated. Solvents were
removed under reduced pressure, MeOH (10 mL) was added to
the residue, and the mixture was stirred 5 min to obtain a
homogeneous suspension. The mixture was filtered, washed with
MeOH (2 × 5 mL) and Et2O (10 mL), and dried under vacuum to
give a yellow-green product. Yield: 20.6 mg (69%). Suitable crystals
(yellow prism) for X-ray analysis were obtained by slow evapora-
tion of dichloromethane from a CH2Cl2 solution. Anal. Calcd for
C28H34N4O4Rh2 (MW ) 696.41): C 48.29, H 4.92, N 8.05. Found:
General Procedures. All oxidation reactions were performed
in air. NaBPh4 and CuOTf were purchased and used as received.
1H NMR (400 MHz) and 13C NMR (100 MHz) spectra were
obtained from solutions in CDCl3. Chemical shifts are reported in
parts per million (ppm, δ) downfield from Me4Si (TMS); coupling
constants are reported in hertz (Hz). High-resolution mass spectra
were done on an ESI instrument. UV-visible spectra were obtained
using a xenon flash lamp. Both cyclic and differential pulse
voltammograms were recorded in 0.2 M (n-Bu)4NPF6 solution
(CH2Cl2, N2-degassed) on a voltammetric analyzer with a glassy
carbon working electrode, a platinum-wire auxiliary electrode, and
a Ag/AgCl reference electrode. The ferrocenium/ferrocene couple
was observed at 0.42 V (vs Ag/AgCl). All mass spectra were
recorded on JEOL AccuTOF-CS from CH2Cl2 solution using

Bis(phenyl) Dirhodium(III) Carboxamidate Compounds
Organometallics, Vol. 27, No. 22, 2008 5845
C 48.10, H 4.93, N 7.29. 1H NMR (400 MHz, CDCl3): δ 7.33 (m,
4H, o-Ph), 7.1 (m, 6H, m,p-Ph), 3.64 (m, 4H, N-CH2), 3.01 (m,
4H, N-CH2), 2.52 (m, 4H, CH2-C), 2.41 (m, 4H, CH2-C), 1.94
(m, 4H, N-CH2-CH2), 1.82 (m, 4H, N-CH2-CH2). 13C NMR (100.6
Computational Details. All geometry optimizations were per-
formed using the B3LYP three-parameter hybrid density functional
method of Becke,37 as implemented in the Gaussian03 suite of
programs.38 The basis functions consisted of the standard LANL2DZ
for all atoms, which included the Hay and Wadt effective core
potentials (ECP)39 for Rh. Unless otherwise stated, the geometry
optimizations were carried out without any symmetry constraint
and the final geometries were characterized as local minima of the
potential energy surface (PES) by verifying that all second
derivatives of the energy were positive. The unrestricted formulation
was used for the open-shell, mixed-valence Rh25+ system. The value
of S2 at convergence was 0.7579, very close to the expected value
of 3/4, indicating minor spin contamination.
1
MHz, CDCl3): δ 183.00 (s, NCO), 141.18 (d, JRh-C ) 34 Hz),,
136.60 (s, o-Ph), 127.50 (s, m-Ph), 125.21 (s, p-Ph), 53.46 (s,
N-CH2), 34.12 (s, CH2-C), 21.82 (s, N-CH2-CH2). UV/visible
spectrum (CH2Cl2): 442 nm (ꢀ ) 4492 M-1 cm-1). HRMS (ESI)
calcd for Rh2C28H35N4O4 697.07684, found 697.07580 (MH+).
XPS: Rh-3d a 5/2: 309.045 eV; Rh-3d a 3/2: 313.739 eV.
Compounds 1b, 2b, 2b′, 3b, 1c, 2c, and 3c. These compounds
were synthesized by the same general procedures described above
for compounds 1a, 2a, 2a′, and 3a. The details are given as
Supporting Information.
Reaction of Bis-σ-(6-methoxynaphthalen-2-yl)-tetrakis-µ-(ca-
prolactamato)dirhodium(III) with Trifluoroacetic Acid. A solu-
tion of trifluoroacetic acid (114 mg, 1.0 mmol) in CH2Cl2 (1 mL)
was slowly added (over 5 min) to a solution of bis-σ-(6-
methoxynaphthalen-2-yl)-tetrakis-µ-(caprolactamato)dirhodium(I-
II) (52 mg, 53.7 µmol) in anhydrous CH2Cl2 (6 mL). The reaction
solution rapidly turned from green to orange. After the addition
was complete, the mixture was stirred at room temperature for 24 h.
Solvents were then removed under reduced pressure, and the residue
was chromatographed on silica gel using hexanes/EtOAc (10:1) as
eluent. The product, 2-methoxynaphthalene, was obtained as a white
Acknowledgment. We are grateful to the National Science
Foundation (CHE-0456911) for their generous support. We
also thank the Fond Social Europe´en for funding J.W. and
Bristol-Meyers-Squibb for supporting J.N. with an American
Chemical Society Organic Division Fellowship. R.P. thanks
the CICT (Project CALMIP) for granting free computer time.
Supporting Information Available: Synthesis and characteriza-
tion of 2 and 3. Crystallographic data for eight compounds (1-3)
in CIF format are available. Cartesian coordinates and relative
energies of all optimized structures. Energies, composition, and
contour diagrams of the relevant molecular orbitals for geometry-
optimized systems I, II, III, IV, and VI. This material is available
free of charge via the Internet at http://pubs.acs.org.
1
solid (15.5 mg, 91% yield). Mp: 71-73 °C; lit.36 71-73 °C. H
NMR (400 MHz, CDCl3): δ 7.76-7.74 (m, 3H), 7.44-7.39 (m,
1H), 7.34-7.29 (m, 1H), 7.14-7.10 (m, 2H), 3.91 (s, 3H).
Reaction of 3c with Trifluoroacetic Acid. In a dry NMR tube
were added 3.0 mg (3.7 µmol) of bis(phenyl)dirhodium(III)
caprolactamate (3c) and 0.5 mL of CDCl3. After the solid was
dissolved and a clear green solution was obtained, 13 µL (18.5
µmol, 5 equiv), 26 µL (37.0 µmol, 10 equiv), or 39 µL (55.5 µmol,
15 equiv) of trifluoroacetic acid solution in CDCl3 (1.35 mmol/
mL) was then added. The mixture was rapidly oscillated for a
minute and then submitted to 1H NMR to follow reaction progress.
Spectra were taken every 5 min for 1 h and then every hour, and
data were obtained by using CHCl3 in CDCl3 as an internal standard.
OM800631B
(36) Arsenijevic, L.; Arsenijevic, V.; Horeau, A.; Jacques, J. Org. Synth.
1973, 53, 5–8.
(37) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(38) Frisch, M. J.; et al. Gaussian 03, ReVision B.04; Gaussian, Inc.:
Wallingford, CT, 2003.
(39) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270–283.