Rh(II) and Rh(I) two-legged piano-stool complexes: Structure, reactivity, and electronic properties

Article abstract of DOI:10.1021/ic0204981

The ligand 1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene, 3, was used to synthesize a mononuclear Rh(II) complex [(η¹:η⁶:η¹-1,4-bis[4- (diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene)Rh] [PF₆]₂, 6⁺, in a two-legged piano-stool geometry. The structural and electronic properties of this novel complex including a single-crystal EPR analysis are reported. The complex can be cleanly interconverted with its Rh(I) form, allowing for a comparison of the structural properties and reactivity of both oxidation states. The Rh(I) form 6 reacts with CO, tert-butyl isocyanide, and acetonitrile to form a series of 15-membered mononuclear cyclophanes [(η¹:η¹-1,4-bis[4- (diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene)Rh(CO)3][PF₆] (8), [(η¹:η¹-1,4-bis[4-(diphenylphosphino)butyl]-2, 3,5,6-tetramethylbenzene)Rh(CNC(CH₃)₃)₂] [PF₆] (10), and [η¹:η¹-1,4-bis[4-(diphenylphosphino)butyl]-2, 3,5,6tetramethylbenzene)Rh(CO)(CH₃CN)][PF₆] (11). The Rh(II) complex 6⁺ reacts with the same small molecules, but over shorter periods of time, to form the same Rh(I) products. In addition, a model two-legged piano-stool complex [(η¹:η⁶:η¹-1,4-bis[3- (diphenylphosphino)propoxy] -2,3,5,6-tetramethyl benzene) Rh][B(C₆F₅)₄], 5, has been synthesized and characterized for comparison purposes. The solid-state structures of complexes 5, 6, 6⁺, and 11 are reported. Structure data for 5: triclinic; P1^-; a = 10.1587(7) A; b = 11.5228(8) A; c = 17.2381(12) A; α = 96.4379(13)°; β = 91.1870(12)°; γ = 106.1470(13)°; Z= 2. 6: triclinic; P1^- a = 11.1934(5) A; b = 12.4807(6) A; c = 16.1771(7) A; α = 81.935(7)°; β = 89.943(1)°; γ = 78.292(1)°; γ = 78.292(1)°; Z = 2. 6⁺: monoclinic; P2(1)/n; a = 11.9371(18) A; b = 32.401(5) A; c = 12.782(2) A; β = 102.890(3)°; Z = 4. 11: triclinic; P1^-; a = 13.5476(7) A; b = 13.8306(7) A; c = 14.9948(8) A; α = 74.551(1)°; β = 73.895(1)°; γ = 66.046(1)°; Z = 2.

Full text of DOI:10.1021/ic0204981

Inorg. Chem. 2003, 42, 3245−3255
Rh(II) and Rh(I) Two-Legged Piano-Stool Complexes: Structure,
Reactivity, and Electronic Properties
Felicia M. Dixon,† Martin S. Masar III,† Peter E. Doan,† Joshua R. Farrell,† Frederick P. Arnold Jr.,†
,†
Chad A. Mirkin,* Christopher D. Incarvito,‡ Lev N. Zakharov,‡ and Arnold L. Rheingold‡
Department of Chemistry and Center for Nanofabrication and Molecular Self-Assembly,
Northwestern UniVersity, 2145 Sheridan Road, EVanston, Illinois 60208-3113, and Department of
Chemistry and Biochemistry, UniVersity of Delaware, Newark, Delaware 19716
Received August 5, 2002
The ligand 1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene, 3, was used to synthesize a mononuclear
Rh(II) complex [(η1:η6:η1-1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene)Rh][PF6]2, 6+, in a two-
legged piano-stool geometry. The structural and electronic properties of this novel complex including a single-
crystal EPR analysis are reported. The complex can be cleanly interconverted with its Rh(I) form, allowing for a
comparison of the structural properties and reactivity of both oxidation states. The Rh(I) form 6 reacts with CO,
tert-butyl isocyanide, and acetonitrile to form a series of 15-membered mononuclear cyclophanes [(η1:η1-1,4-bis-
[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene)Rh(CO)3][PF ] (8), [(η1:η1-1,4-bis[4-(diphenylphosphino)-
6
butyl]-2,3,5,6-tetramethylbenzene)Rh(CNC(CH ) ) ][PF ] (10), and [(η1:η1-1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-
3 3 2
6
tetramethylbenzene)Rh(CO)(CH CN)][PF ] (11). The Rh(II) complex 6+ reacts with the same small molecules, but
3
6
over shorter periods of time, to form the same Rh(I) products. In addition, a model two-legged piano-stool complex
[(η1:η6:η1-1,4-bis[3-(diphenylphosphino)propoxy]-2,3,5,6-tetramethylbenzene)Rh][B(C F ) ], 5, has been synthesized
6
5 4
and characterized for comparison purposes. The solid-state structures of complexes 5, 6, 6+, and 11 are reported.
Structure data for 5: triclinic; P1h; a ) 10.1587(7) Å; b ) 11.5228(8) Å; c ) 17.2381(12) Å; R ) 96.4379(13)°;
â ) 91.1870(12)°; γ ) 106.1470(13)°; Z ) 2. 6: triclinic; P1h; a ) 11.1934(5) Å; b ) 12.4807(6) Å; c ) 16.1771-
(7) Å; R ) 81.935(7)°; â ) 89.943(1)°; γ ) 78.292(1)°; Z ) 2. 6+: monoclinic; P2(1)/n; a ) 11.9371(18) Å; b
) 32.401(5) Å; c ) 12.782(2) Å; â ) 102.890(3)°; Z ) 4. 11: triclinic; P1h; a ) 13.5476(7) Å; b ) 13.8306(7)
Å; c ) 14.9948(8) Å; R ) 74.551(1)°; â ) 73.895(1)°; γ ) 66.046(1)°; Z ) 2.
Introduction
fication and characterization of a new class of mononuclear
Rh(II) complexes in two-legged piano-stool geometries
formed from symmetric bisphosphinoalkylarene ligands
(Scheme 1).3 This class of compounds was designed on the
basis of a systematic study of the factors that control the
stability of Rh(II) in this coordination environment that took
into account ligand steric and electronic factors. The specific
factors that contribute to the stabilization of rhodium(II) in
this geometry are (1) an electron rich and sterically demand-
Mononuclear Rh(II) compounds are relatively rare when
compared with the number of known Rh(I) and Rh(III)
complexes.1 Of the fifteen structurally characterized mono-
meric complexes, there are four predominant coordination
geometries: square planar, octahedral, trigonal bipyramidal,
and sandwich.1c,2 We recently reported preliminary identi-
* To whom correspondence should be addressed. Fax: (847) 467-5123.
E-mail: camirkin@chem.nwu.edu.
† Northwestern University.
‡ University of Delaware.
C.; Dunbar, K. R.; Bender, C. J. Am. Chem. Soc. 1991, 113, 9540.
(d) Paul, P.; Tyagi, B.; Bilakhiya, A. K.; Bhadbhade, M. M.; Suresh,
E. J. Chem. Soc., Dalton Trans. 1999, 2009. Collins, J. E.; Castellani,
M. P.; Rheingold, A. L.; Miller, E. J.; Geiger, W. E.; Rieger, A. L.;
Rieger, P. H. Organometallics 1995, 14, 1232.
(1) (a) Cotton, F. A.; Waton, R. A. Multiple Bonds Between Metal Atoms;
Wiley Interscience: New York, 1982. (b) Pandey, K. K. Coord. Chem.
ReV. 1992, 121, 1. (c) Dewit, D. G. Coord. Chem. ReV. 1996, 147,
209.
(2) (a) Harlow, R. L.; Thorn, D. L.; Baker, R. T.; Jones, N. L. Inorg.
Chem. 1992, 31, 993. (b) Connelly, N. G.; Emslie, D. J. H.; Geiger
W. E.; Hayward, O. D.; Linehan, E. B.; Orpen A. G.; Quayle, M. J.;
Rieger, P. H. J. Chem. Soc., Dalton Trans. 2001, 670. (c) Haefner, S.
(3) Dixon, F. M.; Farrell, J. R.; Doan, P. E.; Williamson, A.; Weinberger,
D. A.; Mirkin, C. A.; Stern, C. L.; Incarvito, C. D.; Liable-Sands, L.
M.; Zakharov, L. N.; Rheingold, A. L. Organometallics 2002, 21,
3091.
10.1021/ic0204981 CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/12/2003
Inorganic Chemistry, Vol. 42, No. 10, 2003 3245

Dixon et al.
Scheme 1
recorded using a Fisions VG 70-250 SE mass spectrometer, and
electrospray (ES) mass spectra were recorded on a Micromass
Quatro II triple quadrapole mass spectrometer. FT-IR spectra were
recorded on a ThermoNicolet 670 spectrometer. Electrochemical
measurements were carried out with a PINE AFRDE5 bipoten-
tiostat/galvanostat using a Au electrode with a Pt mesh counter
electrode and a silver wire quasi-reference electrode in a 0.1 M
n-Bu4NPF6/CH2Cl2 solution. All electrochemical data were refer-
enced using the Fc/Fc+ (Fc ) (η5-C5H5)Fe (η5-C5H5)) redox couple
as an internal standard. Elemental analyses were obtained from
Quantitative Technologies, Inc., Whitehouse, NJ.
Single-Crystal EPR Study. All EPR measurements were taken
on a modified Varian E-4 X-band spectrometer. Frozen solution
(5+, 19:1 CH2Cl2:CHCl3; 6+, CH2Cl2) EPR experiments were
performed in a quartz finger dewar at 77 K. For the single-crystal
experiments, a single crystal was mounted on a quartz fiber using
Paratone and vacuum grease along the 010 face. Once the crystal
was mounted, the faces were indexed using a Bruker Smart 1000
CCD diffractometer to verify its orientation. A simple goniometer
was used in the single-crystal EPR experiments (accuracy (5 deg),
and a nitrogen flow system was used to hold the temperature at
150 K. The field was calibrated using Mn(II)/MgO and diphenyl-
picrylhydrazyl (dpph) as standards. Field accuracy for the 2000 G
sweeps reported was determined to be (3 G. The microwave
frequency was measured with a frequency meter. Computer
simulations of the frozen solution spectra of 5+ and 6+ were
performed using the program QPOW.6
Theoretical Studies. Studies were performed on 6 and 6+ with
a substitution of methyl groups for phenyl rings on the phosphorus
atoms in order to simplify the calculations by reducing the number
of atoms present in the complexes. Undoubtedly, this change
introduces steric and electronic differences between the calculated
and the experimentally characterized structures. However, it has
been shown that the electronic parameters for PEtPh2 and PEtMe2
are very similar.7 The difference in the steric parameters between
the methyl and phenyl versions of 6 and 6+ can be estimated by
examining the cone angles of PEtPh2 (140°) and PEtMe2 (ap-
proximately 122°).7 The difference of nearly 20° is not negligible.
However, the experimental crystal structures of 6 and 6+ and the
theoretical structures of the model complexes are very similar (vide
infra), which suggests that the steric bulk of the phosphines does
not have a large role in the determination of the optimized structures
of these complexes. Geometries were optimized using Jaguar 4.2,8
using the LACVP* effective core potentials and valence basis set,
and the B3-GGA-II hybrid functional in C2 symmetry. LACVP*
employs ECP’s only for the rhodium atoms, and the 6-31G* all-
electron basis set for all other atoms. Structures were verified to
be local minima on the potential surface by frequency analysis of
the optimized structures. NBO analysis was performed using NBO
4.09 contained within Jaguar. The optimized geometries agree with
the X-ray crystal structure for all bond distances (Rh(I), Rh-P )
2.29 Å (∆ ) 0.04 Å), Rh-Cav 2.56 Å (∆ ) 0.19 Å); Rh(II), Rh-P
) 2.39 Å (∆ ) 0.05 Å), Rh-Cav ) 2.50 Å (∆ ) 0.17 Å)).
ing arene ring, (2) a symmetrically substituted arene ring,
and (3) a flexible tether arm with phosphine connectivity
that allows the complex to accommodate structural changes
upon oxidation.4
Since there are very few examples of Rh(II) two-legged
piano-stool complexes, little is known about their electronic
structures, the ligand influences on such structures, and the
reaction chemistry of complexes that make up this class of
compounds. Herein, we report the synthesis and full char-
acterization of a Rh(II) species in this geometry, a detailed
solution and single-crystal EPR study, and theoretical studies
aimed at understanding the metal contribution to the ground-
state electronic configuration of the complex. In addition,
we report the reactivity of both the Rh(I) and Rh(II) forms
of this complex with respect to a variety of small molecules,
including CO, tert-butyl isocyanide, and acetonitrile. Finally,
we compare the properties of this complex with those of
other Rh(II) two-legged piano-stool complexes that have been
characterized in solution.
Experimental Section
General Procedures. Unless otherwise noted, all reactions were
carried out under nitrogen using standard Schlenk techniques or in
an inert atmosphere glovebox. Diethyl ether and tetrahydrofuran
were dried and distilled over sodium/benzophenone. Methylene
chloride and pentane were dried and distilled over calcium hydride.
[RhCl(COE)2]2 (COE ) cyclooctene) was prepared according to
literature procedures.5 CP grade carbon monoxide was purchased
from Matheson Gas and passed through a column of Drierite prior
to use. All other chemicals were obtained from commercial sources
and used as received. NMR spectra were recorded on either a Varian
1
Mercury 300 MHz or an Inova 500 MHz spectrometer. H NMR
signals are reported relative to residual proton resonances in
deuterated solvents, and all signals are reported in parts per million
with coupling constants in hertz. 31P{1H} NMR chemical shifts were
measured relative to an external 85% H3PO4 standard. Fast-atom
bombardment (FAB) and electron impact (EI) mass spectra were
X-ray Crystal Structure Determinations. Diffraction intensity
data for complex 5 were collected using a Bruker Smart 1000 CCD
diffractometer, and data for complexes 6, 6+, and 11 were collected
(4) (a) Singewald, E. T.; Slone, C. S.; Stern, C. L.; Mirkin, C. A.; Yap,
G. P. A.; Liable-Sands, L. M.; Rheingold, A. L. J. Am. Chem. Soc.
1997, 119, 3048. (b) Harlow, R. L.; McKinney, R. J.; Whitney, J. F.
Organometallics 1983, 2, 1839.
(5) Ent, A.; Onderdelinden, A. L. In Reagents for Transition Metal
Complex and Organometallic Syntheses; Angelici, R. J., Ed.; John
Wiley & Sons: New York, 1990; Vol. 28, pp 90-91.
(6) QPOW: EPR analysis software was furnished by the IL EPR research
center, NIH Division of research resources Grant RR01811.
(7) Tolman, C. A. Chem. ReV. 1977, 77, 313.
(8) Schrodinger, Inc., 2002.
(9) NBO 4.0M: Glendeing, E. D.; Badenhoop, J. K.; Reed, A. E.;
Carpenter, J. E.; Weinhold, F. Theoretical Correction Program; Bruker
AXS: Madison, WI, 1996.
3246 Inorganic Chemistry, Vol. 42, No. 10, 2003

Rh(II) and Rh(I) Two-Legged Piano-Stool Complexes
Table 1. Crystallographic Data for 5, 6, 6+ and 11
5
6
6+
11
formula
fw
space group
cryst syst
a, Å
C43H47BF4O2P2Rh
847.5
P1h
triclinic
10.1587(7)
11.5228(8)
17.2381(12)
96.4379(13)
91.1870(12)
106.1470(13)
1923.17(21)
2
C44H52Cl4F6P3Rh
1032.48
P1h
C44H52Cl4F12P4Rh
1177.45
P2(1)/n
monoclinic
11.9371(18)
32.401(5)
12.782(2)
90
102.890(3)
90
4819.2(13)
4
C47H55Cl4F6NOP3Rh
1101.54
P1h
triclinic
triclinic
11.1934(5)
12.4807(6)
16.1771(7)
81.935(1)
89.943(1)
78.292(1)
2190.18(17)
2
13.5476(7)
13.8306(7)
14.9948(8)
74.551(1)
73.895(1)
66.046(1)
2428.7(2)
2
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
D(calc), g cm-3
µ(Mo KR), cm-1
temp, K
radiation
R(F), %a
R(ωF2), %b
1.463
0.583
153
1.566
0.802
120(2)
1.623
0.788
173(2)
1.506
0.730
128(2)
Mo KR (λ ) 0.71073 Å)
5.9
6.7
5.19
14.62
4.45
13.70
6.62
18.45
a R ) ∑||Fo| - |Fc||/∑|Fo|. b R(ωF2) ) {∑[ω(Fo - Fc )2]/∑[ω(Fo )2]}1/2; ω ) 1/σ2(Fo ) + (aP)2 + bP], P ) [2Fc + max(Fo,0)]/3.
2
2
2
2
2
with a Bruker Smart Apex CCD diffractometer. Crystal, data
collection, and refinement parameters are given in Table 1. The
space groups for complexes 5, 6, and 11 were chosen on the basis
of intensity statistics, and the space group for 6+ was determined
from systematic absences in the diffraction data. The structures were
solved using direct methods (5, 6, 6+) or the Patterson function
(11), completed by subsequent difference Fourier syntheses, and
refined by full matrix least-squares procedures on reflection
intensities (F2). SADABS10 absorption corrections were applied to
all data; in complexes 5 (Tmin/Tmax ) 0.87), 6 (Tmin/Tmax ) 0.86),
6+ (Tmin/Tmax ) 0.91), and 11 (Tmin/Tmax ) 0.85). Besides the main
molecules in the crystal structure of 5 there is one benzene molecule,
and in the crystal structures of complexes 6 and 11 there are two
CH2Cl2 molecules. In the crystal structure of 6, the solvate
molecules are disordered, and the program SQUEEZE11 was used
to treat them. Corrections of the X-ray data for 6 by SQUEEZE
(170 electrons/cell) were close to the required values (168 electrons/
cell). Non-hydrogen atoms in all structures were refined with
anisotropic displacement coefficients, and hydrogen atoms were
treated as idealized contributions, with thermal parameters defined
as 1.2 that of the parent atom. Data were collected and processed
using the Bruker Smart-NT and SAINT-NT programs. For complex
5, all calculations were performed using the teXsan crystallographic
software package of Molecular Structure Corporation. Software and
source scattering factors for 6, 6+, and 11 are contained in the
SHELXTL12 (5.10) program package (G.Sheldrick, Bruker XRD,
Madison, WI). Crystallographic data (excluding structure factors)
for the structures in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplementary publica-
tions CCDC-173487 (5), CCDC-175054 (6), CCDC-173489 (6+),
and CCDC-189166 (11). Copies of the data can be obtained, free
of charge, on application to CCDC, 12 Union Road, Cambridge
CB2 1EZ, U.K. (Fax, +44 1223 336033; or e-mail, deposit@
ccdc.cam.ac.uk).
refluxed at 80 °C for 12 h under a nitrogen bubbler. Aqueous
ammonium chloride (20 mL) was added to the resulting brown
solution to quench any remaining zinc. The resulting crude product
was extracted using diethyl ether to yield an orange solution. The
solution was dried with CaCl2, filtered, and then washed with diethyl
ether. Pure product (0.26 g, 31%) was obtained via column
chromatography (1 in. diameter, 2 in. of neutral alumina) by
collecting the second band with 1:4 ethyl acetate:hexanes as the
eluent. 1H NMR (CDCl3): 1.63 (m, CH2CH2(C6(CH3)4), 4H), 1.93
(m, CH2CH2Cl, 4H), 2.25 (s, C6(CH3)4, 12H), 2.69 (m, CH2(C6-
(CH3)4), 4H), 3.62 (t, CH2Cl, JH-H ) 6.6 Hz, 4H). 13C{1H} NMR
(CDCl3): δ 16.6 (s, C6(CH3)4), 27.1 (s, CH2CH2(C6(CH3)4), 30.0
(s, CH2(C6(CH3)4), 33.1 (s, CH2CH2Cl), 45.0 (s, CH2Cl), 132.3 (s,
C6o(CH3)4), 136.4 (s, C6i(CH3)4). HRMS (EI): calcd ) 314.15676,
found ) 314.15663 m/z. Elemental anal. for C18H28Cl2: calcd, %
C ) 68.56, % H ) 8.95; found, % C ) 68.52, % H ) 8.86. Mp
) 92-93 °C.
Synthesis of 1,4-Bis[4-(diphenylphosphino)butyl]-2,3,5,6-tet-
ramethylbenzene (3). A THF solution of KPPh2 (0.5M, 1.3 mL,
0.653 mmol) was added dropwise via a syringe to a flask containing
a THF solution (20 mL) of the chloro precursor described above
(100 mg, 0.318 mmol). The resulting mixture was stirred vigorously
under nitrogen for 2 h, at which time the solvent was removed in
vacuo. An extraction using CH2Cl2 and H2O yielded a white solid,
which was flashed through Celite with CH2Cl2 to remove any
residual salt. Pure 3 was obtained as a white solid after washing of
the solid with a minimal amount of EtOH (30 mL) followed by
1
cannula filtration, (0.167 g, 85%). H NMR (CD2Cl2): 1.65 (m,
PCH2CH2 and PCH2CH2CH2, 8H), 2.17 (m, PCH2, 4H), 2.25 (s,
C6(CH3)4, 12H), 2.69 (m, CH2(C6(CH3)4), 4H), 7.39-7.49 (m,
P(C6H5)2, 20H). 13C{1H} NMR (CD2Cl2): 16.6 (s, C6(CH3)4), 26.6
(d, JC-P ) 12.5, PCH2CH2), 27.9 (d, JC-P ) 9.1, PCH2CH2CH2),
30.6 (s, CH2(C6(CH3)4), 31.4 (d, JC-P ) 10.2, PCH2), 128.5 (s,
P(C6pH5)2), 128.6 (d, JC-P ) 3.67, P(C6mH5)2), 132.1 (s, C6o(CH3)4,
132.7 (d, JC-P ) 14.6, P(C6oH5)2), 136.7 (s, C6i(CH3)4), 138.9 (d,
JC-P ) 10.2, P(C6iH5)2). 31P {1H} NMR (CD2Cl2): -15.3 (s).
HRMS (EI): calcd ) 614.32312, found ) 614.32304 m/z.
Elemental anal. for C42H48P2: calcd, % C ) 82.05, % H ) 7.87;
found, % C ) 82.32, % H ) 7.76. Mp ) 114-116 °C.
Synthesis of 1,4-Bis(4-chlorobutyl)-2,3,5,6-tetramethylben-
zene. A flask equipped with a condenser was charged with
diiododurene (1 g, 2.6 mmol) and Pd(PPh3)4 (0.45 g, 15 mol %).
An excess of a 0.5 M THF solution of 4-chlorobutylzinc bromide
(∼20 mL) was cannula transferred to the reaction mixture and
Synthesis of [(η1:η6:η1-1,4-Bis[3-(diphenylphosphino)propoxy]-
2,3,5,6-tetramethylbenzene)Rh][B(C6F5)4] (5). In a glovebox,
[RhCl(COE)2]2 (40.0 mg, 0.111 mmol) and AgBF4 (22 mg, 0.113
mmol) were stirred in 3 mL of CH2Cl2 for 1 h. The resulting
(10) Sheldrick, G. M. SADABS (2.01), Bruker/Siemens Area Detector
Absorption Correction Program; Bruker AXS: Madison, WI, 1998.
(11) Van der Sluis, P.; Spek, A. L. Acta Crystallogr., Sect. A 1990, A46,
194.
(12) G. Sheldrick, Bruker XRD, Madison, WI.
Inorganic Chemistry, Vol. 42, No. 10, 2003 3247

Dixon et al.
reaction mixture was filtered through Celite to remove a light gray
precipitate, which was presumably AgCl. The filtrate was then
reacted with LiB(C6F5)4 (105 mg, 0.138 mmol), filtered through
Celite, and diluted with 125 mL of THF. A solution of 1,4-bis[3-
(diphenylphosphino)propoxy]-2,3,5,6-tetramethylbenzene (68.6 mg,
0.111 mmol) in 125 mL of THF was added dropwise at -78 °C
over 2 h to the reaction mixture. The solution was warmed to room
temperature over 1 h and then refluxed for 3 days. At this elevated
temperature, the solvent was removed in vacuo to yield an orange
powder. Redissolving the powder in CH2Cl2 and layering with Et2O
removed the remaining COE and afforded 5 (yield ) 58%, 0.090
g, 0.0064 mmol). Recrystallization from CH2Cl2/C6H6 afforded thin
red blades of 5, which were characterized by X-ray crystallography.
1H NMR (CD2Cl2): 1.38-1.50 (bm, 4H, CH2CH2CH2), 2.35-2.42
(bm, 4H, CH2P), 2.55 (s, 12H, C6(CH3)4), 4.00 (t, 4H, CH2O, JH-H
) 5.7), 7.05-7.20 (m, 20H, P(C6H5)2). 31P{1H} NMR (CD2Cl2):
δ 18.1 (d, JRh-P ) 204). MS (FAB+): [M]+ calcd ) 721.1872,
exptl ) 721.1879 m/z. Elemental anal. for C64H44P2O2RhBF20‚CCH2-
Cl2: calcd, % C ) 52.55, % H ) 3.12; found, % C ) 52.22, % H
) 3.79.
Synthesis of [(η1:η6:η1-1,4-Bis[3-(diphenylphosphino)propoxy]-
2,3,5,6-tetramethylbenzene)Rh][B(C6F5)4][BF4] (5+). Compound
5 (10.0 mg, 7.14 × 10-6 mol) and AgBF4 (1.4 mg, 7.14 × 10-6
mol) were reacted in 3 mL of CH2Cl2 for 12 h in a glovebox. The
resulting reaction mixture was filtered through Celite, and the
solvent was removed in vacuo yielding 5+ as a red-brown solid
(yield ) 94%, 0.010 g, 6.72 × 10-6 mol). A saturated CH2Cl2
solution (10 mL) of 5+ was mixed with Et2O (25 mL) to precipitate
a red-brown solid. The solid was collected on a Celite filter and
then dissolved in CH2Cl2 (20 mL). The solvent was removed in
vacuo to give 5+ as a red-brown solid, which allowed for elemental
analysis. 1H NMR (CD2Cl2): 9.4 (b, unassignable), 5.9 (b,
unassignable). 31P{1H} NMR (CD2Cl2): no signal. MS(ES): [M
- BF4]+ calcd ) 1400.2, exptl ) 1400.3 m/z, [M]2+ calcd ) 360.6,
exptl ) 360.2 m/z. Elemental anal. for C64H44P2O2RhB2F24: calcd
% C ) 51.68, % H ) 2.98; found: % C ) 51.60, % H ) 3.24.
Synthesis of [(η1:η6:η1-1,4-Bis[4-(diphenylphosphino)butyl]-
2,3,5,6-tetramethylbenzene)Rh][PF6] (6). In a glovebox, [RhCl-
(COE)2]2 (40.0 mg, 0.111 mmol) and AgPF6 (28 mg, 0.111 mmol)
were reacted in 3 mL of CH2Cl2 for 30 min. The resulting reaction
mixture was filtered through Celite, which removed a light gray
precipitate, and then diluted with 30 mL of THF. A solution of 3
(68 mg, 0.111 mmol) in 30 mL of THF was added dropwise over
30 min at -78 °C to the filtrate. The solution was warmed to room
temperature over 1 h followed by heating at 50 °C for 4 h. The
solvent was removed in vacuo at 50 °C to yield an orange solid.
Compound 6 was precipitated as a deep red solid from a
concentrated CH2Cl2 solution with pentane (18 mg, 20%). X-ray
grade crystals of 6 were grown by slow diffusion of pentane into
a CH2Cl2 solution saturated with 6. 1H NMR(CD2Cl2): 1.665 (bm,
CH2CH2(C6(CH3)4), 4H), 2.112 (bm, PCH2CH2CH2, 8H), 2.202 (s,
C6(CH3)4, 12H), 2.700 (bm, CH2(C6(CH3)4), 4H), 7.090-7.215 (m,
P(C6H5)2, 20H). 31P {1H} NMR (CD2Cl2): 28.0 (d, JRh-P ) 204).
MS(ES): [M]+ calcd ) 717.2, exptl ) 717.2 m/z. Elemental anal.
for C42H48F6P3Rh: calcd, % C ) 58.48, % H ) 5.61; found, % C
) 58.30, % H ) 5.74.
an X-ray diffraction analysis were grown by slow vapor diffusion
of diethyl ether into a CH2Cl2 solution saturated with 6+ at 0 °C.1H
NMR(CD2Cl2): 1.78 (b, unassignable), 2.11-2.25 (b, unassign-
able), 2.709 (b, unassignable), 7.7 (bm, P(C6H5)2). 31P{1H} NMR
(CD2Cl2): no signal. Elemental anal. for C42H48F12P4Rh: calcd,
% C ) 50.06, % H ) 4.80; found, % C ) 50.00, % H ) 4.95.
Despite attempts using several different techniques, mass spectral
analysis aimed at identifying the 2+ ion was unattainable due to
the reduction of 6+ to 6 during ionization.
Synthesis of [(η1:η1-1,4-Bis[4-(diphenylphosphino)butyl]-
2,3,5,6-tetramethylbenzene)Rh(CO)3][PF6] (8). In a typical reac-
tion, a red CH2Cl2 (1 mL) solution of 6 (10 mg, 0.01159 mmol)
was charged with CO for 4 h in a Teflon valved air-free NMR
tube. The reaction vessel was kept at room temperature for 40 days,
with periodic monitoring by 31P{1H} NMR spectroscopy, before
the complete conversion to 8 was observed. The resulting yellow
solution was determined to contain 8 by the following methods.
1H NMR (CD2Cl2): 1.28 (b, Ph2PCH2CH2CH2CH2, 4H), 1.90 (b,
Ph2PCH2CH2CH2, 4H), 2.14 (s, C6(CH3)4, 12H), 2.49 (b, Ph2PCH2-
CH2, 4H), 3.00 (b, CH2(C6(CH3)4), 4H), 7.47-7.61 (m, (C6H5)2P,
20H). 31P{1H} NMR (CD2Cl2): 30.4 (d, JRh-P ) 72). FT-IR (CH2-
Cl2): νCO ) 2023 (s), 2034 (s), and 2091 (w) cm-1. MS (ES): [M
- 3CO]+ calcd ) 717.2, found ) 717.0. The lability of the CO
ligands prevented elemental combustion analysis.
Synthesis of [trans-(η1:η1-1,4-Bis[4-(diphenylphosphino)bu-
tyl]-2,3,5,6-tetramethylbenzene)Rh(CO)2][PF6] (9a) and [cis-(η1:
η1-1,4-Bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethyl-
benzene)Rh(CO)2][PF6] (9b). The synthesis of complex 9a is
identical to the synthesis for 8 except the reaction mixture is heated
at 51 °C for 3 days. Complex 9b is observed as an unstable
intermediate that transforms into 9a and cannot be isolated. 9a:
1H NMR (CD2Cl2) 1.52 (b, Ph2PCH2CH2CH2CH2, 4H), 1.68 (b,
Ph2PCH2CH2CH2, 4H), 2.18 (s, C6(CH3)4, 12H), 2.58 (b, Ph2PCH2-
CH2 and CH2(C6(CH3)4), 8H), 7.44-7.67 (m, (C6H5)2P, 20H); 31P-
{1H} NMR (CD2Cl2) 23.9 (d, JRh-P ) 121). 9b: 31P{1H} NMR
(CD2Cl2) 25.3 (JRh-P ) 105 Hz). The FT-IR spectrum was
unassignable owing to the presence of 8 in the samples. MS and
EA were not possible due to the lability of the CO and the presence
of 8.
Synthesis of [(η1:η1-1,4-Bis[4-(diphenylphosphino)butyl]-
2,3,5,6-tetramethylbenzene)Rh(CNC(CH3)3)2][PF6] (10). To a
CH2Cl2 (1 mL) solution of 6 (10 mg, 0.01159 mmol) was added 2
equiv of tert-butyl isocyanide (2.6 µL, 0.02318 mmol) followed
by heating at 45 °C for 3 days, resulting in the quantitative formation
of the diisocyanide adduct, 10, as determined by NMR spectroscopy.
1H NMR (CD2Cl2): 0.86 (s, CNC(CH3)3, 18H), 1.28 (b, Ph2PCH2-
CH2CH2CH2, 4H), 1.90 (b, Ph2PCH2CH2CH2, 4H), 2.14 (s, C6-
(CH3)4, 12H), 2.49 (b, Ph2PCH2CH2, 4H), 3.00 (b, CH2(C6(CH3)4),
4H), 7.47-7.61 (m, (C6H5)2P, 20H). 31P{1H} NMR (CD2Cl2): 23.6
(d, JRh-P ) 123). FT-IR (CH2Cl2): νNC ) 2135 cm-1. MS (ES):
[M]+ calcd ) 883.4, found) 883.1. Elemental combustion analysis
was not possible due to the lability of the isocyanide ligands upon
evacuation.
Synthesis of [(η1:η1-1,4-Bis[4-(diphenylphosphino)butyl]-
2,3,5,6-tetramethylbenzene)Rh(CO)(CH3CN)][PF6] (11). Excess
CH3CN was added to a CD2Cl2 solution of 6 (10 mg, 0.01159
mmol) in a Teflon sealed air-free NMR tube followed by charging
with CO (1 atm) for 30 min. The reaction mixture was heated at
40 °C for 3 days, which resulted in the quantitative formation of
11, as determined by NMR spectroscopy. Compound 11 was
characterized in solution and by mass spectrometry. Single crystals
of 11 suitable for an X-ray diffraction analysis were grown by slow
diffusion of pentane into a CH2Cl2 solution saturated with 11 at
Synthesis of [(η1:η6:η1-1,4-Bis[4-(diphenylphosphino)butyl]-
2,3,5,6-tetramethylbenzene)Rh][PF6]2 (6+). As with 5, complex
6 (10 mg, 0.0116 mmol) was combined with excess AgPF6 in CH2-
Cl2 (3 mL) and stirred vigorously for 30 min followed by filtration
to remove a Ag0 precipitate, yielding a yellow-brown solution.
Removal of solvent yields 6+ as a yellow-brown solid (yield )
90%, 0.0105 g, 0.0104 mmol). Single crystals of 6+ suitable for
3248 Inorganic Chemistry, Vol. 42, No. 10, 2003

Rh(II) and Rh(I) Two-Legged Piano-Stool Complexes
room temperature. 1H NMR (CD2Cl2): 1.28 (b, Ph2PCH2CH2CH2-
CH2, 4H), 1.80 (b, Ph2PCH2CH2CH2, 4H), 2.14 (s, C6(CH3)4, 12H),
2.40 (b, Ph2PCH2CH2, 4H), 2.93 (b, CH2(C6(CH3)4), 4H), 7.52 (b,
(C6H5)2P, 20H). 31P{1H} NMR (CD2Cl2): 27.2 (d, JRh-P ) 97).
FT-IR (CH2Cl2): νCO ) 1976 (s), νCH3CN ) 2218 (s) cm-1. MS
(ES): [M+] calcd ) 786.3, found ) 786.1. Elemental combustion
analysis was not possible for this compound owing to the lability
of the ligands upon evacuation.
Scheme 2
Reactivity of 6+ with Small Molecules. In a typical reaction, a
CH2Cl2 solution of 6+ was charged with CO or a stoichiometric
amount of the ancillary ligand. In the case of CO and tert-butyl
isocyanide, the brown solution turns yellow upon reaction at 51
°C and results in the formation of the reduction/substitution products
8 (12 h) and 10 (24 h), respectively. See syntheses of complexes 8
and 10 for characterization. Reaction of 6+ with CO and acetonitrile
resulted in decomposition with no identifiable products.
formation of the oligomers, but without the ethers acting as
intramolecular templating agents, the formation of oligomers
was unavoidable under the conditions explored. Complexes
Results and Discussion
Synthesis and Characterization of Mononuclear Rh(I)
Complexes 5 and 6. Previous studies in our group with
mononuclear two-legged piano-stool complexes suggest that
an ideal ligand framework for the stabilization and isolation
of mononuclear Rh(II) with this geometry consists of a
symmetrically substituted arene with two tethered phosphine
moieties, where the arene is sterically protected with electron-
donating alkyl moieties and the tethers are long enough to
accommodate structural changes upon complex oxidation.
As a result, bis(phosphinoalkyl)aryl ligands 1-3 were
designed and synthesized. Ligands 1 (1,4-bis[4-(diphe-
nylphosphino)ethoxy]-2,3,5,6-tetramethylbenzene) and 2 (1,4-
bis[4-(diphenylphosphino)propoxy]-2,3,5,6-tetramethylben-
zene) have been previously reported,13 while ligand 3,
1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenz-
ene, was synthesized in two steps via a standard zinc-
mediated coupling reaction14 followed by nucleophilic
substitution. Ligand 3 is a moderately air-sensitive white solid
that has been characterized by 1H and 31P{1H} NMR
spectroscopies, mass spectrometry, combustion analysis, and
melting point determination. Complex 4 was prepared via
literature procedures, and 5 and 6 were prepared from ligands
2 and 3 and the appropriate Rh(I) precursor.13 Compound 5
is obtained in almost 60% yield after refluxing the solution
for 3 days, while 6 is formed in only 20% yield after 4 h of
heating at 50 °C. The difference in reactivity is attributed to
the ether groups in ligand 2, which can template the
formation of 5 through a binuclear intermediate (Scheme 2).
Complex 5 is the thermodynamic product of a high-yielding
reaction pathway that initially results in the formation of a
binuclear intermediate 7, which cleanly condenses into
mononuclear compound 5.13 The reaction to form complex
6, which involves a phosphinoalkylarene ligand without ether
moieties, is prone to the formation of oligomers and
polymers, which substantially lowers the yield of the target
complex. Changes in dilution (3-fold) and temperature (room
temperature to refluxing THF) were attempted to bypass the
1
5 and 6 have been characterized by H and 31P{1H} NMR
spectroscopy, mass spectrometry, elemental analysis, and
X-ray diffraction analyses. The 31P{1H} NMR spectra of 5
and 6 each exhibit single resonances (5, δ 18.1, d, JRh-P
)
204 Hz; 6, δ 28.0, d, JRh-P ) 204 Hz) that are characteristic
of complexes with two-legged piano-stool geometries about
the rhodium(I) center.
The relative kinetic and thermodynamic stabilities of the
Rh(II) complexes can be determined by assessing the
reversibility of the oxidation/reduction wave as a function
of scan rate and the corresponding E1/2 values, respectively.
Therefore, cyclic voltammetry studies were conducted on
4-6 in CH2Cl2 at room temperature. All three complexes
exhibit reversible one electron oxidation waves at all
measured scan rates (1 mV/s to 1 V/s); a cyclic voltammo-
gram of 6 at a sweep rate of 400 mV/s is provided as a
representative example of the electrochemical stability of
these complexes (Figure 1). The E1/2 values for 4-6 are 560,
528, and 410 mV vs Fc/Fc+ [Fc ) (η5-C5H5)-Fe-(η5-
C5H5)], respectively. Comparing the E1/2 value for 5 with
the E1/2 for complex 4, which has one less methylene unit,
reveals that lengthening the chelation arm results in increased
thermodynamic stability of the resulting Rh(II) complex. This
phenomenon is presumably due to the increased ability of 5
to accommodate structural changes (lengthening of the Rh-P
Figure 1. Cyclic voltammogram of 6 in CH2Cl2/0.1 M n-Bu4NPF6 using
a Au working electrode, a Pt mesh counter electrode, and a Ag wire quasi-
reference electrode at a scan rate of 400 mV/s. Ferrocene was used as an
internal reference in a subsequent scan.
(13) Farrell, J. R.; Eisenberg, A. H.; Mirkin, C. A.; Guzei, I. A.; Liable-
Sands, L. M.; Incarvito, C. D.; Rheingold, A. L.; Stern, C. L.
Organometallics 1999, 18, 4856.
(14) Reike, R. D. Aldrichimica Acta 2000, 33, 52.
Inorganic Chemistry, Vol. 42, No. 10, 2003 3249

Dixon et al.
as red-brown and brown solids, respectively. Compounds 5+
1
and 6+ have been characterized by H and 31P{1H} NMR
spectroscopy, mass spectrometry, elemental analysis, and
EPR spectroscopy. Consistent with the formation of para-
1
magnetic Rh(II) compounds, H and 31P{1H} NMR spec-
troscopy of both complexes in CD2Cl2 exhibit broad, nearly
featureless spectra. Although complex 5+ is moderately stable
in solution under an inert atmosphere, we were unable to
grow suitable crystals for characterization by single-crystal
X-ray diffraction analysis; disproportionation occurred during
crystal growth (∼1 week) resulting in the formation of 5
and unidentified products. Complex 6+ is indefinitely stable
in the solid state in air and substantially more stable than 5+
in solution, which allowed us to grow crystals suitable for
single-crystal X-ray diffraction analysis.
Solid-State Characterization of 6+. Single crystals of 6+
suitable for an X-ray diffraction analysis were grown by slow
vapor diffusion of diethyl ether into a CH2Cl2 solution
saturated with 6+ at 0 °C (Figure 3B; Tables 1 and 2).
Comparison of the structures of 6 and 6+ shows that they
are remarkably similar with only a few notable differences.
It was postulated that the ligand framework should be flexible
enough to undergo two types of structural changes, namely,
the lengthening of the Rh-P bonds and widening of the
P-Rh-P angle, to stabilize the Rh(II) metal center upon
oxidation of the Rh(I) complex 6.4a Indeed, the Rh-P bonds
Figure 2. ORTEP diagram of 5. Thermal ellipsoids are drawn at 50%
probability, and counterions, hydrogen atoms, and lattice solvent are omitted
for clarity. See Table 2 for selected bond distances and angles.
bonds and widening of the P-Rh-P angle) upon oxidation.
The need for this type of flexibility has been demonstrated
with related complexes.4 The 90 mV decrease in E1/2 in going
from 5 to 6 is attributed to the increase in electron richness
of the arene in 6. Consistent with this conclusion, it has been
shown that replacement of a single ether group with a
methylene unit in an analogous two-legged piano-stool
complex results in a 58 mV decrease in E1/2.4a
in 6+ are longer than those in 6 by about 0.1 Å (Rh-Pav
)
2.3422 Å), which is consistent with a decrease in π-back-
bonding between the phosphines and the metal that is
expected upon oxidation. The P-Rh-P angle is slightly
affected upon going from 6 to 6+ (widens by ∼0.5°), which
is attributed to the ligand having adopted a near ideal
geometry prior to oxidation. In other words, ligands that
constrain the P-Rh-P angle (e.g., 1,2-bis(diphenylphos-
phino)ethane) have led to less stable Rh(II) complexes.4a
In addition to the changes in the Rh(PPh2)2 unit upon
oxidation, the Rh-arene interaction is perturbed as evidenced
by the shortening of the Rh-centroid distance (0.03 Å) and
three of the Rh-C bonds. Like its Rh(I) analogue, the arene
ring in complex 6+ adopts a boat conformation with the bow
and stern of the boat being made up of C(19) and C(22).
The average deviation from planarity is smaller in 6+ (0.022
Å) than in 6 (0.033 Å), which is consistent with the tendency
of arene rings in 17e- two-legged piano-stool complexes to
be more planar than analogous 18e- complexes, which often
adopt boat structures.16 In contrast to complex 6, the bow
and stern of the boat-shaped arene in 6+ consist of two longer
bonds (Rh-C(19) and Rh-C(22)) making the boat point
away from the rhodium center. The arene ring flip upon
oxidation of 6 to 6+ is accompanied by a rotation of the
Rh(PPh2)2 unit about the metal-arene axis. The phosphine
groups in the Rh(I) complex 6 are rotated clockwise (θ )
21°) out of the plane defined by Rh and the carbon atoms
on the arene ring that attach the tethers, C(1) and C(4) (Figure
Solid-State Characterization of Complexes 5 and 6. The
structures of 5 and 6 were determined by single-crystal X-ray
diffraction methods and can be compared with the known
structure of 4 (Figures 2 and 3A; Tables 1 and 2).13 Crystals
of 5 suitable for X-ray diffraction analysis were grown by
slow diffusion of benzene into a CH2Cl2 solution saturated
with 5. Similarly, crystals of 6 were grown by slow diffusion
of pentane into a CH2Cl2 solution saturated with 6. As with
4, the Rh atoms in 5 and 6 are centered on the arene ring of
the two-legged piano stool. The P-Rh-P bond angles for
complexes 5 and 6 are smaller than that for 4 (4, 96.48(2)°;
5, 93.28(5)°; 6, 94.89(4)°), and the average Rh-P and Rh-
arene bond distances in complex 4 are shorter than those in
5 and 6 (4, Rh-Pav ) 2.2388 Å, Rh-arene ) 1.83 Å; 5,
Rh-Pav ) 2.2585 Å, Rh-arene ) 1.88 Å; 6, Rh-P )
2.2508 Å, Rh-arene ) 1.87 Å). These differences are
attributed to the lengthening of the tether arms in complexes
5 and 6 by one methylene unit as compared with 4. Both
durenyl moieties in 5 and 6 are nonplanar with average
deviations of planarity of 0.0468 and 0.0387 Å, respectively.
The arene rings in all three complexes adopt boat confirma-
tions with two short and four long Rh-C bonds. The bow
and stern of the boats are pointed toward the rhodium metal
center and are formed by the carbon atoms connecting the
tether arm to the arene ring (5, C(3) and C(6); 6, C(1) and
C(4)).
Synthesis and Characterization of Rh(II) complexes 5+
and 6+. Compounds 5 and 6 were chemically oxidized to
their Rh(II) forms, 5+ and 6+, with AgX (X ) BF4 or PF6)
salts (E1/2 = 650 mV/s vs Fc/Fc+ in CH2Cl2)15 and isolated
(15) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877.
(16) (a) Muetterties, E. L.; Bleeke, J. R.; Wucherer, E. J. Chem. ReV. 1982,
82, 499. (b) Radonovich, L. J.; Klabunde, K. J.; Behrens, C. B.;
McCollor, D. P.; Anderson, B. B. Inorg. Chem. 1980, 19, 1221.
3250 Inorganic Chemistry, Vol. 42, No. 10, 2003

Rh(II) and Rh(I) Two-Legged Piano-Stool Complexes
Figure 3. (A) ORTEP diagram of 6. Thermal ellipsoids are drawn at 50% probability, and counterions, hydrogen atoms, and solvent molecules are omitted
for clarity. See Table 2 for selected bond distances and angles. (B) ORTEP diagram of 6+. Thermal ellipsoids are drawn at 50% probability, and counterions,
hydrogen atoms, and solvent molecules are omitted for clarity. See Table 2 for selected bond distances and angles.
Table 2. Selected Bond Distances (Å) and Angles (deg) for 5, 6, and 6+
complex 5
complex 6
complex 6+
Rh-P(1)
Rh-P(2)
Rh-C(6)
Rh-C(1)
Rh-C(2)
Rh-C(3)
Rh-C(4)
Rh-C(5)
Rh-arene
2.2532(13)
2.2639(12)
2.269(4)
2.376(4)
2.357(2)
2.285(4)
2.356(4)
2.409(4)
1.88
Rh-P(1)
Rh-P(2)
Rh-C(1)
Rh-C(6)
Rh-C(5)
Rh-C(4)
Rh-C(3)
Rh-C(2)
Rh-arene
2.2514(9)
2.2501(10)
2.314(4)
2.413(4)
2.344(4)
2.314(4)
2.379(4)
2.330(4)
1.87
Rh-P(1)
2.3381(7)
2.3463(8)
2.316(2)
2.319(2)
2.343(3)
2.306(3)
2.315(3)
2.355(3)
1.84
Rh-P(2)
Rh-C(17)
Rh-C(18)
Rh-C(19)
Rh-C(20)
Rh-C(21)
Rh-C(22)
Rh-arene
P(1)-Rh-P(2)
93.28(5)
P(1)-Rh-P(2)
94.89(4)
P(1)-Rh-P(2)
95.41(2)
difference in orientation of the arene ring between 4-6 and
6+ cannot be due to electronic effects only. The methylene
tethers in complexes 4-6 and 6+ restrict the rotation of the
Rh(PPh2)2 fragment about the metal-arene interaction. The
ability for ML2 fragments to adopt either a through-bond or
through-atom configuration with the arene ring is the origin
of the two different orientations (boat toward or away from
metal) of the boat structure. In addition to the methylene
tethers restricting rotation, the phenyl groups of the phosphine
units sterically hinder rotation about the metal-arene axis.
For these reasons, we believe the ring flip that accompanies
the Rh(I) to Rh(II) conversion can be attributed to both steric
and electronic factors.
Figure 4. Representation of out of plane rotation of the phosphine groups
(P) in complexes 6 and 6+ where θ ) 0 is the plane defined by the carbon
atoms of the arene ring that attach the methylene tethers (CT: 6, C(1) and
C(4); 6+, C(17) and C(22)) and the Rh center. θ represents the angle of
rotation.
4). Upon oxidation of 6 to form the Rh(II) analogue, 6+, the
phosphine moieties shift closer to the tethered carbon atoms,
C(17) and C(20), where θ ) 15°.
EPR and Theoretical Studies. The paramagnetic d7 Rh-
(II) metal centers were examined by X-band EPR spectros-
copy. The frozen solution EPR spectrum of complex 5+
displays a rhombic pattern with no resolved hyperfine
splittings (Figure 5A, solid line). The best-fit computer
simulation (Figure 5A, dashed line) gives the g tensor values
and line widths of g1 ) 2.390 (85 MHz), g2 ) 2.038 (140
MHz), and g3 ) 1.997 (90 MHz). The frozen solution EPR
spectrum of 6+ (Figure 5B, solid line) appears to exhibit an
axial pattern with no resolved hyperfine splittings; however,
computer simulation of the experimental spectrum of 6+
(Figure 5B, dashed line) requires a slight rhombic splitting
of the gz feature. The parameters that best simulate the
experimental spectrum are g1 ) 2.363 (82 MHz), g2 ) 2.0245
(77 MHz), g3 ) 2.0045 (77 MHz).
The deviations from planarity of the arene rings in
complexes 4-6 and 6+ are consistent with what has been
observed with analogous η6-arene-ML2 complexes (e.g., (π-
C6H5CH3)Ni(C6F5)2).16,17 Molecular orbital diagrams for these
reported complexes, which were constructed from extended
Hu¨ckel molecular orbital calculations, reveal that the devia-
tions from planarity are due to electronic effects. Specifically,
to decrease the antibonding interaction between the metal-
arene MO and the ligand orbitals either two or four bonds
are lengthened to form the boat conformation. Although DFT
calculations performed on PMe2-substituted analogues of
complexes 6 and 6+ produce MO diagrams that are quali-
tatively consistent with those reported in the literature for
analogous two-legged piano-stool complexes, we believe the
(17) (a) Albright, T. A.; Hoffmann, R.; Tse, Y.-C.; D’Ottavio, T. J. Am.
Chem. Soc. 1979, 101, 3812. (b) Radonovich, L. J.; Koch, F. J.;
Albright, T. A. Inorg. Chem. 1980, 19, 3373.
The g| value of 2.36 for complex 6+ clearly indicates that
the odd electron is in a substantially metal-centered orbital.
Inorganic Chemistry, Vol. 42, No. 10, 2003 3251

Dixon et al.
Figure 7. Angular variation of theoretical and experimental g2 values
versus goniometer angle where 0° corresponds to the 101h face.
DFT calculations. There are four magnetically equivalent
sites in the unit cell that give rise to only one EPR signal,
which simplifies the analysis. Ignoring the slight rhombic
splitting, we treat the g tensor as axial with g| ) 2.36 and
g ) 2.02. A single rotation with the magnetic field normal
to the unique crystallographic (b) axis is sufficient to identify
the molecular direction of g|. A single crystal was mounted
and indexed on a quartz fiber using Paratone and vacuum
grease with the rotation axis coincident to the molecular b
axis. EPR measurements were recorded every 15° over a
360° range while the crystal was rotated in the ac plane.
The experimental g2 value is compared to predicted g2 values
in which the g| direction is assumed to be either x, y, or z
(Figure 7). The data best correspond to the assignment g| )
gx, which is consistent with the simple ligand-field prediction
for a dyz ground state. As complex 5+ has an EPR spectrum
similar to that of 6+, we take by analogy that the electronic
structure of 5+ is the same as that of 6+.18
Figure 5. Frozen solution (solid lines) and simulated (dashed lines) EPR
spectra of 5+ (A) and 6+ (B). Microwave frequency: 5+, 9.105 GHz; 6+,
9.090 GHz. Microwave power: 5+ and 6+, 20 mW. Modulation ampli-
tude: 5+ and 6+, 5 G. Sweep time: 5+ and 6+, 4 min. Time constant: 5+,
250 ms; 6+, 125 ms.
Chart 1
Reactivity of the Rh(I) Complex 6 and Rh(II) Complex
6+ with Small Molecules. In addition to structurally and
electronically characterizing the mononuclear Rh(I) and Rh-
(II) complexes 6 and 6+, we have investigated their reac-
tivities with small molecules. The η6-arene moieties of other
mononuclear and dinuclear two-legged piano-stool Rh(I)
complexes have been shown to be substitutionally labile in
the presence of small molecules such as CO and acetoni-
trile.13,19 These small molecules bind more strongly to the
Rh(I) metal centers, resulting in the displacement of the arene
group and subsequent formation of square planar or trigonal
bipyramidal bisphosphine complexes. Therefore, an impor-
tant issue pertains to the effect of the chelating environment
in 6 and 6+ on the lability of the arene moieties in these
complexes.
In agreement with this, using the axis system as defined in
Chart 1, the DFT calculated structure of an analogue of
complex 6+ (phenyl rings are substituted with methyl groups)
revealed a large metal contribution to the SOMO from the
dyz orbital (Figure 6); the calculation further gives the LUMO
to be the dxz orbital, which is consistent with reported
rhodium(I) two-legged piano-stool complexes.4a
To address this issue, the reactivity of the Rh(I) piano-
stool complex 6 with CO, tert-butyl isocyanide, and aceto-
nitrile was initially studied. When a CD2Cl2 solution of 6 is
charged with CO (1 atm), the Rh-arene interaction is broken,
and the tricarbonyl trans-phosphine complex, [(η1:η1-1,4-
bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene)-
We undertook a single-crystal EPR experiment of complex
6+ to substantiate the electronic structure determined by the
(18) The observed g3 > ge value for 6+ and the lack of resolved hyperfine
splittings both in 5+ and 6+ remain open questions that are being
pursued by both advanced EPR spectroscopy (ENDOR and ESEEM)
and DFT calculations.
(19) (a) Singewald, E. T.; Mirkin, C. A.; Levy, A. D.; Stern, C. L. Angew.
Chem., Int. Ed. Engl. 1994, 33, 2473. (b) Singewald, E. T.; Shi, X.;
Mirkin, C. A.; Schofer, S. J.; Stern, C. L. Organometallics 1996, 15,
3062.
Figure 6. 0.05 e/au3 isosurface of the SOMO of 6+.
3252 Inorganic Chemistry, Vol. 42, No. 10, 2003

Rh(II) and Rh(I) Two-Legged Piano-Stool Complexes
Scheme 3
Scheme 4
Rh(CO)3][PF6], 8, forms as the sole product (Scheme 3). The
conversion of 6 to 8 was monitored by 31P{1H} NMR
spectroscopy, which exhibited a downfield shift in the initial
resonance associated with 6 (δ 28.0, JRh-P ) 204 Hz) to a
new resonance at δ 30.4 (JRh-P ) 72 Hz). The latter
resonance and coupling constant are diagnostic of a trans-
bisphosphine five-coordinate complex with equatorial CO
ligands and are consistent with our structural assignment for
8.13,19 The FT-IR spectrum of 8 exhibits three metal carbonyl
bands at 2023 (s), 2034 (s), and 2091 (w) cm-1, which is
consistent with a trigonal bipyramidal metal complex with
an asymmetric ligand environment. The asymmetry of the
ligand splits the E′ mode yielding three νCO bands rather
than the two typically observed for complexes with local
D3h symmetry. Splitting of the carbonyl bands due to
asymmetric ligand environments has been observed in similar
mononuclear rhodium(I) and cobalt(I) complexes.20
Although qualitatively similar reactivity has been well-
documented for isostructural Rh(I) complexes (in particular,
CO-induced arene displacement), the reaction of 6 with CO
is significantly slower than for many of the related complexes
studied thus far under comparable conditions.13,19b For
example, the conversion of 6 to 8, under an atmosphere of
CO, takes over 20 days at room temperature whereas an
analogous binuclear complex with a two-legged piano-stool
geometry at each metal center quantitatively converts to the
tris-CO adduct per Rh(I) center under an atmosphere of CO
within 2 h.13 The decrease in reactivity of 6 toward CO
compared to similar mononuclear and dinuclear two-legged
piano-stool type complexes is a direct result of the symmetric
chelation of the η6-arene group to the Rh center and the
methyl substitution of the arene. In addition, there is another
important electronic effect that stabilizes 6. The HOMO in
6 is the dyz, which bisects the P-Rh-P plane (xz plane)
(Chart 1). The space-filling representation of the theoretical
model for 6 indicates that the metal center is exposed along
the filled HOMO (dyz) whereas the LUMO (dxz, along the
Rh-P bonds) is protected by the ligand framework, which
inhibits the substitution reaction.
analogous Rh(I) complex 6. It is well-known that 17-electron
organometallic complexes often undergo substitution reac-
tions at faster rates than their 18-electron counterparts.21
Saturating a CH2Cl2 solution of 6+ with CO and heating in
a sealed NMR tube results in a yellow solution that is EPR
silent. The 31P{1H} NMR spectrum, on the other hand,
exhibits a doublet at δ 30 (JRh-P ) 72 Hz), which is identical
to the doublet in the spectrum for CO-substituted Rh(I)
complex 8 (Scheme 4). FT-IR spectroscopy confirmed the
formation of 8 via the reduction of the Rh(II) complex 6+.
13CO-labeling studies show a new resonance at δ 125 in the
13C{1H} NMR spectrum upon reaction of 13CO with 6+,
which is consistent with the formation of 13CO2. The 13C-
{1H} NMR and the 31P{1H} NMR spectra of the 13CO
analogue of 8 taken at -70 °C exhibited a doublet of triplets
at δ 184.6 (JC-Rh ) 64 Hz, JC-P ) 12 Hz) and a doublet of
quartets at δ 30 (JRh-P ) 204 Hz; JP-P ) 13 Hz),
respectively, verifying the binding of three CO moieties per
rhodium center. A similar reduction of Rh(II) has been
observed with [PBzPh3]2[Rh(C6Cl5)4] after bubbling of CO
through a CH2Cl2 suspension to yield a Rh(I) adduct, but
the oxidation product was not identified.22 Also, Dunbar and
co-workers have reported the formation of Rh(I) and Rh-
(III) disproportionation products when CO is reacted with
[Rh(tmpp)2]2+ (tmpp ) tris(2,4,6-trimethoxyphenyl)phos-
phine).23 Since only one compound is observed in the 31P-
{1H} NMR spectrum of the product mixture formed from
the reaction between 6+ and CO, a simple two-component
disproportionation reaction has been ruled out.
To increase the rate of formation of 8, a CD2Cl2 solution
of complex 6 in a sealed NMR tube was heated at 51 °C
under an atmosphere of CO for 3 days. Instead of observing
only resonances due to 8, the 31P{1H} NMR spectrum
exhibited a new doublet at δ 23.9 (JRh-P ) 121 Hz). On the
basis of the similarity in coupling constant and chemical shift
with data for analogous trans-phosphine complexes,19b this
new product was determined to be [trans-(η1: η1-1,4-bis[4-
(diphenylphosphino)butyl]-2,3,5,6-tetramethylbenzene)Rh-
Interestingly, the substitution reaction of the Rh(II)
(21) Trogler, W. C. In Organometallic Radical Processes; Trogler, W. C.,
Ed.; Elsevier: Amsterdam, 1990; Vol. 22, pp 306-337.
(22) Garc´ıa, M. P.; Jimenez, M. V.; Cuesta, A.; Siurana, C.; Oro, L. A.;
Lahoz, F. J.; Lopez, J. A.; Catalan, M. P. Organometallics 1997, 16,
1026.
(23) Haefner, S. C.; Dunbar, K. R.; Bender, C. J. Am. Chem. Soc. 1991,
113, 9540.
complex 6+ with CO occurs dramatically faster than for the
(20) (a) Alvarez, M.; Lugan, N.; Donnadieu, B.; Mathieu, R. Organome-
tallics 1995, 14, 365. (b) Alvarez, M.; Lugan, N.; Mathieu, R. Inorg.
Chem. 1993, 32, 5652. (c) Yagupsky, G.; Brown, C. K.; Wilkinson,
G. J. Chem. Soc. A 1970, 1392.
Inorganic Chemistry, Vol. 42, No. 10, 2003 3253

Dixon et al.
1
(CO)2][PF6], 9a, the bis-CO analogue of 8. The H NMR
spectrum of the solution indicated that a very small amount
of 8 was also present in solution. Complete conversion to
9a was never observed, even after prolonged heating (14
days). Cooling of the reaction mixture containing predomi-
nantly 9a (8 is undetectable by 31P{1H} NMR) to room
temperature resulted in a decrease in 9a and increase in 8
(after 5 days, 8:9a ratio ) 2:1). The product ratio of this
mixture can be repeatedly affected by changes in temperature
in a reversible manner. While prolonged heating of the
mixture of 8 and 9a results in the formation of the tris-CO
product and one bis-CO product, monitoring of the reaction
by 31P{1H} NMR spectroscopy at an early stage shows the
appearance of a second upfield resonance δ 25.3 (JRh-P
)
105 Hz) assigned to the cis-bis-CO adduct, 9b. Comparison
with literature values of analogous complexes supports this
assignment.20a,b Due to overlapping resonances, it is difficult
Figure 8. ORTEP diagram of 11. Thermal ellipsoids are drawn at 50%
probability, and counterions, hydrogen atoms, and solvent molecules are
omitted for clarity. Selected bond distances (Å) and angles (deg): Rh-
arene(centroid) ) 5.636, Rh(1)-P(1) ) 2.3373(12), Rh(1)-P(2) ) 2.3332-
(12), Rh(1)-C(1) ) 1.819(5), Rh(1)-N(1) ) 2.054(4), P(1)-Rh(1)-P(2)
) 175.45(4).
1
to get reliable quantitative integration data in the H NMR
spectrum and to discern the assignment of the FT-IR
spectrum of the reaction mixture. Qualitatively, the reaction
involving 8 and 9a can be described as a mildly endothermic
process that favors product 9a at higher temperatures.
Complex 6 also reacts with tert-butyl isocyanide (51 °C,
10 days) through an arene displacement reaction to yield [(η1:
η1-1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetra-
methylbenzene)Rh(CNC(CH3)3)2][PF6], 10. This complex
exhibits a doublet in its 31P{1H} NMR spectrum at δ 23.6
(JRh-P ) 123 Hz), which is consistent with the formation of
a square planar Rh(I) complex with trans-phosphine atoms
as observed in similar trans-nitrile trans-phosphine macro-
cyclic complexes. The FT-IR spectrum exhibits a charac-
teristic νNC stretch at 2135 cm-1. The reaction between two
isocyanide molecules and complex 6 to form 10 was also
confirmed by mass spectrometry (M+ ) 883.1). As with CO,
the 17-electron Rh(II) complex 6+ reacts faster in the
presence of 2 equiv of tert-butyl isocyanide but still forms
the same product as that formed with the reduced complex
6. Indeed, the reaction is complete within 24 h with the only
spectroscopically observable (31P{1H} NMR and FT-IR)
metal-containing complex being 10 (Scheme 4). The reducing
agent was not identifiable for this reaction. A similar
reduction of mononuclear Rh(II) to Rh(I) in the presence of
tert-butyl isocyanide has been observed by Wilkinson and
co-workers with Rh(2,4,6-Pri3C6H2)2(tht)2 (tht ) tetrahy-
drothiophene), which occurs via hydrogen abstraction from
the solvent.24 In addition, a Rh(II) porphyrin, Rh(II) TMP
(TMP ) tetrakis(2,4,6-trimethylphenyl)porphyrinato), reacts
with CNCH3 to form the Rh(I) adducts (TMP)RhCN and
CH3Rh(TMP).25 If hydrogen abstraction was occurring to
form a Rh(I) adduct, one would not expect to see the clean
formation of product 10.
zene)-Rh(CO)(CH3CN)][PF6], 11. This reaction occurs after
18 days at room temperature. Complex 11 was characterized
by 1H and 31P{1H} NMR spectroscopy, mass spectrometry,
and an X-ray diffraction analysis. The 31P{1H} NMR
spectrum exhibits a doublet at δ 27.2 (JRh-P ) 97 Hz)
characteristic of a trans-phosphine square planar complex
with a CO ligand trans to an acetonitrile ligand. The FT-IR
spectrum exhibits characteristic stretches at 1976 and 2218
cm-1 (νCO and νMeCN, respectively). The reaction between
6+ and acetonitrile resulted in decomposition and no identifi-
able products.
The observed rates of reactions involving 6 and 6+ with
CO and tert-butyl isocyanide, respectively, suggest that, in
the case of the 17-electron complex 6+, ligand substitution
takes place prior to electron transfer. Otherwise, in the case
of 6+, one would expect to observe formation of some
reduced 6, yet this has not been observed under the conditions
explored thus far. The facilitated conversion of 6+ to 8 as
compared with 6 to 8 is likely due to the ability of the 17-
electron complex to support the coordination of an additional
ligand through an associative pathway to form a 19-electron
complex. Subsequent reduction and then arene displacement
through the addition of two more CO ligands leads to product
8. In the case of 6, ligand displacement through ring slippage
or phosphine dissociation must occur prior to CO uptake.
Solid-State Characterization of 11. Single crystals of 11
suitable for an X-ray diffraction analysis were grown by slow
diffusion of pentane into a CH2Cl2 solution saturated with
11 at room temperature (Figure 8; Table 1). Upon displace-
ment of the arene ring, the Rh-arene centroid distance
becomes 5.636 Å. The Rh-P average bond length of 2.3353
Å is slightly longer than the Rh-P average bond lengths
for the piano-stool complex 6, 2.250 Å, and compares well
with bond length data for similar square planar binuclear
complexes with trans CO/CH3CN groups at each metal center
(2.3265 and 2.335 Å).13 The geometry about the Rh(I) metal
center is slightly distorted square planar with a P(1)-Rh-
Finally, complex 6 reacts with CO in the presence of CH3-
CN in CD2Cl2 to yield the cationic Rh(I) complex [(η1: η1-
1,4-bis[4-(diphenylphosphino)butyl]-2,3,5,6-tetramethylben-
(24) Hay-Motherwell, R. S.; Koschmieder, S. U.; Wilkinson, G.; Hussain-
Bates, B.; Hursthouse, M. B. J. Chem. Soc., Dalton Trans. 1991, 2821.
(25) Wayland, B. B.; Sherry, A. E.; Bunn, A. G. J. Am. Chem. Soc. 1993,
115, 7675.
3254 Inorganic Chemistry, Vol. 42, No. 10, 2003

Rh(II) and Rh(I) Two-Legged Piano-Stool Complexes
(1)-P(2) angle of 175.45(4)°. The CO moiety is pointed
toward the arene between carbon atoms C(16) and C(17)
where the CO deviates from the plane made up of C(8)-
Rh-C(11) by 41.9°. Major differences between the arene-
bound complex 6 and the unbound complex 11 are the
orientation of the phosphine tethers with respect to the arene
ring and the conformation of the arene ring. The P(1)-Rh-
(1)-P(2) triad lies in the same plane as C(8)-Rh(1)-C(11),
which are the carbon atoms attached to the tether arms,
instead of twisting about the arene ring. Due to the lack of
metal binding, the arene ring is planar rather than distorted
in a boat orientation as in 6.16a
small molecules. In general, for the reactivity studied herein,
the chemistry of the Rh(II) forms of these two-legged piano-
stool complexes parallels that for the Rh(I) forms, but occurs
at accelerated rates. This is believed to be due to the ability
of the Rh(II) complexes to support associative reaction
pathways prior to reaction, which facilitates ligand substitu-
tion.
Acknowledgment. C.A.M. acknowledges AFOSR and
NSF for support of this research. F.M.D. acknowledges the
Illinois Minority Graduate Incentive Program for a predoc-
toral fellowship. C.A.M. and F.M.D. acknowledge Charlotte
Stern for X-ray data collection and analysis of complex 5.
P.E.D. acknowledges Professor Brian M. Hoffman for
funding (NSF MCB-9904018 and NIH HL-13531) and
helpful discussions, and Mr. Clark E. Davoust for technical
support. The University of Delaware acknowledges the NSF
for their support of the purchase of the CCD-based diffrac-
tometer (CHE00091968).
Conclusions
This manuscript reports the synthesis, characterization, and
reactivity of Rh(I) and Rh(II) two-legged piano-stool mono-
mers. The stabilizing effects of the ligands are demonstrated
and reveal three factors that increase the stability of mono-
nuclear Rh(II) in this geometry: (1) tether arms containing
four spacer units connecting the phosphine groups to the
arene ring; (2) electron richness of the arene, which is
increased by replacing ether moieties attached to the arene
ring with methylene groups; and (3) symmetrically substi-
tuted arene rings to maximize chelation. The ligand frame-
work kinetically stabilizes the unusual oxidation state due
to the increased steric bulk about the metal center by
inhibiting dimerization and reaction with solvent and other
Supporting Information Available: Tables of crystallographic
data for complexes 5, 6, 6+, and 11, and atomic coordinates for
the DFT calculated structures analogous to complexes 6 and 6+.
X-ray data available as the CIF files for complex 11. The CIF files
for 5, 6, and 6+ can be obtained from manuscript OM020248S.3
This material is available free of charge via the Internet at
http://pubs.acs.org.
IC0204981
Inorganic Chemistry, Vol. 42, No. 10, 2003 3255