Mono- and binuclear complexes of rhodium involving a new series of hemilabile o-phosphinoaniline ligands

Article abstract of DOI:10.1039/b822170g

Monophosphines of the type Ph_xPAr_3-x (x = 0, 1 or 2, Ar = o-N-methylanilinyl) and the diphosphine, Ar₂PCH ₂PAr₂ (mapm) have been synthesized for use as chelating and/or bridging P,N-ligands within mon

Full text of DOI:10.1039/b822170g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Mono- and binuclear complexes of rhodium involving a new series of
hemilabile o-phosphinoaniline ligands†
Lindsay J. Hounjet,a Matthias Bierenstiel,‡a Michael J. Ferguson,b Robert McDonaldb and Martin Cowie*a
Received 10th December 2008, Accepted 16th February 2009
First published as an Advance Article on the web 6th April 2009
DOI: 10.1039/b822170g
Monophosphines of the type PhxPAr3-x (x = 0, 1 or 2, Ar = o-N-methylanilinyl) and the diphosphine,
Ar2PCH2PAr2 (mapm) have been synthesized for use as chelating and/or bridging P,N-ligands within
mono- and binuclear rhodium(I) complexes, respectively. The previously prepared phosphines,
PhxPAr¢3-x (x = 0, 1 or 2, Ar¢ = o-N,N-dimethylanilinyl) and Ar¢2PCH2PAr¢2 (dmapm), have also been
used to prepare analogous mono- and binuclear complexes. Variable temperature 1H NMR
spectroscopy of the mononuclear complexes, [RhCl(CO)(L)] (L = PhPAr2, PhPAr¢2, PAr3 and PAr¢3),
and line-shape analyses of the resultant spectra indicate the substantially increased lability of the
N,N-dimethylanilinyl donors relative to the related monomethylanilinyl groups. X-Ray structural
analyses of the mononuclear complexes suggest that the enhanced Type II hemilability in the
dimethylanilinyl complexes compared to their monomethyl analogues results from increased steric
interactions involving the coordinated dimethylanilinyl substituents. In the case of the binuclear,
dmapm-bridged compound [Rh2Cl2(CO)2(m-dmapm)], there are additional transannular repulsions
between the chloro ligand on one metal and the coordinated dimethylanilinyl group on the other, which
˚
result in a Rh–Rh separation of over 4.1 A. For the analogous mapm-bridged species, the transannular
interactions between the chloro ligands and the amine hydrogens are in fact attractive, resulting in a
˚
much closer Rh–Rh separation (3.450 A). The chloride substituents of [Rh2Cl2(CO)2(m-mapm)] can be
replaced to generate the complexes, [Rh2(X)2(CO)2(m-mapm)] (X = I, CF3SO3, CH3CO2), the last of
which also exhibits pronounced transannular hydrogen-bonding interactions in the solid state.
unsaturation” has obvious applications in catalysis,20–27,35 in which
Introduction
the labile donor stabilizes the catalyst precursor prior to substrate
coordination and assists in displacing the catalyst-modified sub-
strate, regenerating the catalyst precursor, after the transformation
is complete. In this context, ligands containing “soft” phosphorus
and “hard” nitrogen donors have found many applications as
hemilabile ligands in the chemistry of low-valent, late-transition-
metal complexes,20–24,26–38 in which phosphorus binds strongly while
nitrogen is more labile.
We have sought to combine two of the above themes through
the use of diphosphine ligands with pendent amine groups,
in which the diphosphine moiety binds effectively to and
bridges a pair of late metals, holding them in close proximity,
while the chelating amines function as labile groups. In earlier
studies we43 and others44–48 used bis(di(o-N,N-dimethylanilinyl)
phosphino)methane (dmapm = Ar¢2PCH2PAr¢2; Ar¢ = o-C6H4-
NMe2) as a bridging diphosphine ligand that has chelating,
hemilabile dimethylanilinyl groups. However, in our study we
proposed that unfavorable steric repulsions involving the pairs
of methyl substituents on the anilinyl groups have appeared to
inhibit close approach of the adjacent metals, so we subsequently
set out to synthesize the somewhat less bulky monomethylanilinyl
analogue, Ar2PCH2PAr2 (mapm; Ar = o-C6H4NHMe), in order to
reduce the steric demand of the amine donors. In addition, we set
out to prepare a series of monophosphine analogues of mapm in
order to compare the reactivities of related mononuclear and bin-
uclear diphosphine-bridged species, thereby gaining information
on possible influences of adjacent metals on substrate activation
Bi- and multidentate ligands occupy an important position in
the chemistry of transition metals.1–7 Not only do such groups
find applications in mononuclear complexes, where they offer
additional stability compared to related monodentate ligands,
through the chelate effect,8 they can also be used to bridge two or
more metals in multinuclear complexes.9–18 Multidentate ligands
can also be extended to a series of “hybrid” ligands, capable of
binding to the metal(s) through different donor atoms.19–42 This not
only introduces the flexibility of ligand fine-tuning, in which the
metal(s) can be sterically and electronically “tuned” through the
use of different combinations of donor sites within these hybrids,
but also introduces the concept of hemilability,19–38 in which one
or more donor sites in the multidentate ligand bind more strongly
to the metal(s) under study while other donor site(s) bind weakly.
These labile donors are capable of stabilizing the complex in the
absence of substrate, while being readily and reversibly displaced
by the appropriate substrate. The resulting “incipient coordinative
aDepartment of Chemistry, University of Alberta, Edmonton, Alberta,
Canada T6G 2G2. E-mail: martin.cowie@ualberta.ca; Fax: +1 (1)780
4928231; Tel: +1 (1)780 4925581
bX-Ray Crystallography Laboratory, University of Alberta, Edmonton,
Alberta, Canada T6G 2G2
† CCDC reference numbers 713707–713713. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b822170g
‡ Present address: Department of Chemistry, Cape Breton University,
Sydney, Nova Scotia, Canada, B1P 6L2
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4213
©

View Article Online
and on the possibility of cooperative substrate activation by
the adjacent metals. Such cooperative substrate activation has
been elegantly demonstrated in a related dirhodium system16 that
utilized a non-labile tetraphosphine ligand in which the central
pair of phosphorus nuclei bridged the metals while the outer pair
each chelated to a different metal.
A further aspect of interest in these monomethylanilinyl phos-
phines is the possibility of deprotonating the amine groups yielding
amido functionalities. The reversible transformation of chelating
amine to amido groups has generated enormous recent interest in
the catalytic hydrogenation of polar substrates such as ketones.49–51
In this paper we report the synthesis of a series of hybrid
monomethylanilinyl phosphine ligands and the generation of a
series of mononuclear and binuclear complexes of rhodium using
these hybrid ligands. The steric influences of these monomethy-
lanilinyl derivatives with regards to their lability and their
structural influences are compared with the analogous species
containing the dimethylanilinyl groups.
these species the molar conductivities were determined as K =
23 and 12 cm2 X-1 mol-1, respectively.
Preparation of P,N-ligands
(a) Diphenyl(o-N-methylanilinyl)phosphine (Ph2PAr) 1. In a
200 mL Schlenk flask N-methylaniline (1.73 mL, 15.9 mmol) was
dissolved in 30 mL of freshly distilled, dry tetrahydrofuran (THF)
and cooled to -78 ◦C (acetone/dry-ice bath). n-BuLi (2.5 M
in hexanes, 6.3 mL, 16 mmol) was added dropwise via syringe
resulting in immediate slow gas evolution and formation of a white
precipitate. The reaction mixture was allowed to warm to ambient
temperature (approx. 45 min) after which CO2(g) was passed
through the reaction mixture via a syringe needle attachment at a
moderate rate (approx. 0.5 mL s-1) for 15 min resulting in a clear,
light yellow solution. The solution was allowed to stir for 15 min
before cooling to -78 ◦C. t-BuLi (1.7 M in THF, 11 mL, 19 mmol)
was added dropwise via syringe producing a white precipitate in a
bright yellow-orange solution. The reaction mixture was stirred at
-78 ◦C for 5 min, allowed to warm to -35 ◦C (acetonitrile/dry-ice
bath) and stirred for 1 h to generate the dilithiated intermediate.
Chlorodiphenylphosphine (2.85 mL, 15.9 mmol) in 15 mL of
dry THF was added dropwise via syringe. The cooling bath was
removed and the reaction mixture was allowed to warm to ambient
temperature. HCl (2 M, 15 mL) was added carefully to quench the
reaction, leading to release of CO2(g). After cessation of CO2(g)
effervescence, the solution was neutralized with a 30% (w/w)
KOH–H2O solution. 50 mL of water was then added and the
aqueous layer was extracted with 3 ¥ 50 mL of Et2O. The combined
organic layers were then washed with 100 mL of water, dried over
anhydrous Na2SO4, filtered and the solvent was removed in vacuo.
The o-phosphinoaniline was recrystallized from approx. 50 mL
of boiling ethanol (2.78 g, 60.1%) yielding a white, crystalline
product (found: C, 78.11; H, 6.30; N, 4.81. Calc. for C19H18NP: C,
78.33; H, 6.23; N, 4.81%); dH(400 MHz; CD2Cl2; Me4Si) 2.84 (3H,
s/br, CH3), 4.86 (1H, m/br, NH), 6.65 (2H, m, HAr), 6.79 (1H,
m, HAr), 7.33 (11H, m, HAr). dC(101 MHz; CD2Cl2; Me4Si) 30.8
(1C, s, CH3). dP(162 MHz; CD2Cl2; H3PO4) -21.8 (s). HRMS (EI,
70 eV). Found: m/z 291.11697 for [M]+. Calc. for C19H18NP: m/z
291.11768.
Experimental
General comments
All solvents were deoxygenated, dried (using appropriate drying
reagents) and distilled before use and stored under nitrogen.
Reactions were performed under an argon atmosphere using
standard Schlenk techniques. RhCl3·3H2O, Ph2PCl, PhPCl2, PCl3
and Cl2PCH2PCl2 were purchased from Strem Chemicals. n-BuLi
and t-BuLi were purchased from Sigma-Aldrich. Dry CO2(g) was
52
purchased from Supelco. The compounds [Rh(m-Cl)(COD)]2
53
(COD
=
1,5-cyclooctadiene) and [Rh(m-Cl)(CO)2]2
were
prepared by the literature routes. The monophosphine lig-
ands, bis(o-N,N-dimethylanilinyl)-phenylphosphine (PhPAr¢2),
tris(o-N,N-dimethylanilinyl)phosphine (PAr¢3),54 and the diphos-
phine ligand, bis(di(o-N,N-dimethylanilinyl)phosphino)methane
(dmapm),44 were prepared as previously reported. o-Bromo-N,N-
dimethylaniline was prepared from commercially available o-
bromoaniline by exhaustive methylation with dimethylsulfate.55
NMR spectra were recorded on Bruker AM-400, Varian Inova-
400 or Varian Unity-500 spectrometers operating at 400.0, 399.8
or 499.8 MHz, respectively, for 1H; at 161.9, 161.8 or 202.3 MHz,
respectively, for 31P; and at 100.6, 100.6 or 125.7 MHz, respectively,
for 13C nuclei. J values are given in Hz and overlapping, unresolved
aromatic 13C NMR signals, observed in the typical 80–120 ppm
range, are not reported. Spectroscopic data for all metal complexes
(5–14) are provided in Table 1. Solution-phase infrared spectra
(KBr cell) were recorded on either a FT-IR Bomem MB-100 spec-
trometer or a Nicolet Avatar 370 DTGS spectrometer. Elemental
analyses were performed by the Microanalytical Laboratory of the
University of Alberta. Electrospray ionization mass spectra were
run on a Micromass Zabspec spectrometer in the departmental
MS facility. In all cases, the distribution of isotope peaks for
the appropriate parent ion matched very closely that calculated
from the formulation given. SpinWorks version 2.5.556 was used
for line-shape analyses and NMR spectral simulations. Con-
ductivity measurements were carried out under inert conditions
on 10-3 M solutions of [Rh2(OTf)2(CO)2(m-mapm)] (12) and
[Rh2(OAc)2(CO)2(m-mapm)] (14) in dry nitromethane using a
Yellow Springs Instrument Model 31 conductivity bridge. For
(b) Di(o-N-methylanilinyl)phenylphosphine (PhPAr2) 2. The
dilithiated intermediate was prepared from N-methylaniline
(2.10 mL, 19.3 mmol) as described in part (a). Dichloro-
phenylphosphine (1.31 mL, 9.65 mmol) was added dropwise via
syringe and the mixture was allowed to warm slowly to ambient
temperature. The resultingsolutionwas thenacidified, neutralized,
extracted, dried and filtered as described in part (a). The solvent
was removed in vacuo and the o-phosphinoaniline was cleanly
precipitated from approx. 50 mL of boiling ethanol (1.40 g, 45.4%)
yielding a white powder (found: C, 74.64; H, 6.49; N, 8.66. Calc.
for C20H21N2P: C, 74.98; H, 6.61; N, 8.74%); dH(400 MHz; CD2Cl2;
Me4Si) 2.85 (6H, s/br, CH3), 4.71 (2H, m/br, NH), 6.67 (4H, m,
HAr), 6.82 (2H, m, HAr), 7.37 (7H, m, HAr). dC(101 MHz; CD2Cl2;
Me4Si) 30.9 (2C, s, CH3). dP(162 MHz; CD2Cl2; H3PO4) -38.0 (s).
HRMS (EI, 70 eV). Found: m/z 320.14380 for [M]+. Calc. for
C20H21N2P: m/z 320.14423.
(c) Tri(o-N-methylanilinyl)phosphine (PAr3) 3. The dilithi-
ated intermediate was prepared from N-methylaniline (2.10 mL,
4214 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 1 Spectroscopic data for the rhodium complexes
NMRb
1
1
Compound
IR/cm-1a d(31P{ H})/ppmc d(1H)/ppmd
d(13C{ H})/ppmd
[RhCl(CO)(Ph2PAr)] (5)
[RhCl(CO)(PhPAr2)] (6)
1992 (s)
1996 (s)
58.0 (d, 1JRhP
169 Hz, 1P)k
41.7 (d, 1JRhP
156 Hz, 1P)k
=
=
NH: 5.57 (br, 1H)k NMe: 2.87 (d, 3JHH
=
CO: 189.3 (dd, 1JRhC = 73 Hz, 2JPC = 18
CO: 188.9 (dd/br, 1JRhC = 55 Hz)k NMe:
43.9 (s/br, 1C), 30.2 (s/br)k
6.5 Hz, 3H)k
Hz)k NMe: 44.1 (s)k
NH: 5.73 (br, 2H)k NMe: 2.73 (br, 6H)k
NH: 7.20 (3H),l 6.37 (br, 1H), 6.32 (br,
1H), 5.95 (br, 3H)g NMe: 2.88 (d, 3JHH
5.9 Hz, 3H), 2.77 (d, 3JHH = 5.0 Hz, 3H),
2.62 (d, 3JHH = 5.7 Hz, 9H), 2.56 (d,
3JHH = 4.8 Hz, 9H)g
=
[RhCl(CO)(PAr3)] (7)
1994 (s)
1987 (s)
27.9 (d, 1JRhP
=
NH: 7.00 (1H),l 5.04 (br, 1H), 4.64 (br,
1H)k NMe: 2.79 (br, 9H)k
CO: 189.6 (dd, 1JRhC = 73 Hz, 2JPC = 16
149 Hz, 1P)k
Hz)k NMe: 30.3 (s/br)k
NH: 7.38 (1H),l 5.23 (br, 1H), 4.36 (br,
1H)i NMe: 2.81 (d, 3JHH = 5.0 Hz, 3H),
2.71 (d, 3JHH = 4.9 Hz, 3H), 2.63 (d,
3JHH = 6.0 Hz, 3H)i
[RhCl(CO)(PhPAr¢2)] (8)
49.7 (d, 1JRhP
=
=
NMe2: 2.75 (s, 12H)k
CO: 189.1 (dd, 1JRhC = 74 Hz, 2JPC = 17
173 Hz, 1P)k
Hz)k NMe2: 48.5 (s)k
NMe2: 3.01 (s/br, 3H), 2.94 (s/br, 3H),
2.69 (s/br, 3H), 1.89 (s/br, 3H)f
NMe2: 2.69 (s, 18H)k
[RhCl(CO)(PAr¢3)] (9)
1988 (s)
2000 (s)
1999 (s)
37.8 (d, 1JRhP
CO: 190.1 (dd, 1JRhC = 76 Hz, 2JPC = 17
186 Hz, 1P)k
Hz)k NMe2: 47.8 (s)k
NMe2: 2.83 (s/br, 9H), 2.34 (s/br, 9H)f
NH: 7.75 (2H), 6.94 (2H)k,l CH2: 3.94
(m, 2H)k NMe: 3.17 (d, 3JHH = 6.0 Hz,
6H) 2.78 (d, 3JHH = 4.8 Hz, 6H)k
[Rh2Cl2(CO)2(m-mapm)] (10)
[Rh2Cl2(CO)2(m-dmapm)] (11)
23.3 (m, 1JRhP
=
CO: 185.3 (dd, 1JRhC = 71 Hz, 2JPC = 18
Hz)k NMe: 42.9 (s, 2C), 30.2 (s, 2C)k
CH2: 33.7 (t, 1JPC = 31 Hz)k
160 Hz, 2P)e,k
41.0 (d, 1JRhP
=
CH2: 4.59 (t/br, 2JPH = 12.4 Hz, 2H)k
NMe2: 3.70 (s/br, 6H), 2.97 (s/br, 6H),
2.73 (s/br, 6H), 2.53 (s/br, 6H)k
CO: 187.6 (dd, 1JRhC = 80 Hz, 2JPC = 28
Hz)j NMe2: 52.3 (s, 2C), 51.9 (s, 2C),
47.9 (s, 2C), 47.1 (s, 2C)j CH2: 27.3 (t,
1JPC = 29 Hz)j
175 Hz, 2P)k
41.8 (m, 1JRhP
=
=
=
=
CH2: 4.59 (t, 2JPH = 11.6 Hz, 2H)f NMe2:
3.68 (s, 6H), 2.86 (s, 6H), 2.70 (s, 6H),
2.27 (s, 6H)f
179 Hz, 2P)e,h
[Rh2(OTf)2(CO)2(m-mapm)] (12) 2000 (s)
31.5 (m, 1JRhP
NH: 7.64 (2H)k,l, 6.62 (m/br, 2H)k CH2: N/A (poorly soluble)
176 Hz, 2P)e,k
3.97 (m, 2H)k NMe: 3.09 (d, 3JHH
=
6.0 Hz, 6H) 2.85 (d, 3JHH = 5.0 Hz, 6H)k
NH: 7.60 (2H), 6.62 (2H)k,l CH2: 3.96
(m, 2H)k NMe: 3.25 (d, 3JHH = 6.0 Hz,
6H) 2.76 (d, 3JHH = 5.0 Hz, 6H)k
[Rh2I2(CO)2(m-mapm)] (13)
2000 (s)
20.3 (m, 1JRhP
N/A (poorly soluble)
162 Hz, 2P)e,k
[Rh2(OAc)2(CO)2(m-mapm)] (14) 1991 (s)m 28.2 (m, 1JRhP
NH: 8.91 (m/br, 2H), 8.12 (m/br, 2H)k
N/A (slowly decomposes in CD2Cl2)
155 Hz, 2P)e,k
CH2: 3.81 (m, 2H)k NMe: 3.14 (d, 3JHH
=
6.0 Hz, 6H), 2.78 (d, 3JHH = 4.8 Hz, 6H)k
a IR abbreviations: s = strong, m = medium, w = weak. Only nCO signals given. Dichloromethane solution; in units of cm-1. b NMR abbreviations: s =
singlet, d = doublet, t = triplet, m = multiplet, br = broad, dd = doublet of doublets. NMR data in CD2Cl2. c 31P chemical shifts referenced to external
d
85% H3PO4. 1H and 13C chemical shifts r◦eferenced to tetramethylsilane. Chemical shifts for phenyl groups not given. e 2nd-order◦effects complicate
observed signal pattern. f NMR data at -80 C. g NMR data at -60 ◦C. h NMR data at -40 ◦C. i NMR data at -20 ◦C. j NMR data at 0 C. k NMR data at
27 ◦C. l Multiplicities of NH signals could not be determined (due to overlap with aromatic proton signals); chemical shifts were determined by gradient
correlation spectroscopy (GCOSY) analysis. m THF solution.
19.3 mmol) as described in part (a). Trichlorophosphine (0.56 mL,
6.4 mmol) was added dropwise via syringe and the mixture was
allowed to slowly warm to ambient temperature. The resulting
solution was then acidified, neutralized, extracted, dried and
filtered as described in part (a). The solvent was removed in vacuo
and the o-phosphinoaniline precipitated from approx. 50 mL of
boiling ethanol (0.436 g, 19.4%) yielding an off-white powder;
(d) Bis(di(o-N-methylanilinyl)phosphino)methane (mapm) 4.
The dilithiated intermediate was prepared from N-methylaniline
(1.75 mL, 16.2 mmol) as described in part (a). In a 25 mL Schlenk
flask bis(dichlorophosphino)methane (0.53 mL, 4.0 mmol) was
dissolved in 2 mL of freshly distilled, dry THF. The diphosphine
solution was added dropwise over 5 min to the reaction mixture via
cannula and the mixture was allowed to slowly warm to ambient
temperature. The resultingsolutionwas thenacidified, neutralized,
extracted, dried and filtered as described in part (a). The solvent
was removed in vacuo and the o-phosphinoaniline was cleanly
precipitated from approx. 20 mL of boiling ethanol (0.378 g,
18.9%) yielding an off-white powder (found: C, 69.19; H, 6.80;
N, 10.76; Calc. for C29H34N4P2: C, 69.59; H, 6.85; N, 11.19%);
3
dH(400 MHz; CD2Cl2; Me4Si) 2.85 (9H, d/br, JHH = 5.0 Hz,
CH3), 4.57 (3H, m/br, NH), 6.69 (6H, m, HAr), 6.82 (3H, m,
HAr), 7.34 (3H, m, HAr). dC(101 MHz; CD2Cl2; Me4Si) 30.8 (3C, s,
CH3). dP(162 MHz; CD2Cl2; H3PO4) -53.6 (s). HRMS (EI, 70
eV). Found: m/z 349.17050 for [M]+. Calc. for C21H24N3P: m/z
349.17078.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4215
©

View Article Online
dH(400 MHz; CD2Cl2; Me4Si) 2.74 (2H, m, CH2), 2.77 (12H, d,
3JHH = 4.8 Hz, CH3), 4.62 (4H, m/br, NH), 6.58 (4H, m, HAr), 6.69
(4H, m, HAr), 7.24 (8H, m, HAr). dC(100 MHz; CD2Cl2; Me4Si) 21.7
(1C, t, 1JPC = 16 Hz, CH2), 31.1 (4C, s, CH3). dP(162 MHz; CD2Cl2;
H3PO4) -60.9 (s). HRMS (ES+). Found: m/z 501.23282 for [M+ +
H]. Calc. for C29H35N4P2: m/z 501.23315.
0.406 mmol) and PhPAr¢2 (282 mg, 0.812 mmol) and isolated as
a yellow solid (366 mg, 87.5%). Single crystals suitable for X-ray
crystallographic analysis were obtained by dissolving the complex,
under Ar atmosphere, in a minimum volume of CH2Cl2 and
layering the solution with anhydrous n-pentane in an NMR tube
(found: C, 53.49; H, 4.92; N, 5.48. Calc. for [C21H20ClN2OPRh]:
C, 53.66; H, 4.89; N, 5.44%).
Preparation of metal complexes
(e) Chlorocarbonyl(diphenyl(o-N-methylanilinyl)phosphine)rho-
dium(I) [RhCl(CO)(Ph2PAr)] 5. In a 50 mL Schlenk flask under
anhydrous conditions and Ar atmosphere, [Rh(m-Cl)(COD)]2
(200 mg, 0.406 mmol) and Ph2PAr (236 mg, 0.811 mmol) were
dissolved in dichloromethane (10 mL) at ambient temperature.
CO(g) was passed through the solution for 10 min at an
approximate rate of 0.5 mL s-1 and the reaction mixture was
stirred for 18 h at ambient temperature. The solvent was
reduced to approximately 2 mL under vacuum and a yellow
solid precipitated upon addition of 20 mL of dry n-pentane.
The yellow solid was filtered, washed with 10 mL of n-pentane
and dried in vacuo (334 mg, 90.4%). Single crystals suitable for
X-ray crystallographic analysis were obtained by dissolving the
complex, under Ar atmosphere, in a minimum volume of CH2Cl2
and layering the solution with anhydrous n-pentane in an NMR
tube (found: C, 50.80; H, 3.71; N, 2.96; Cl, 10.63. Calc. for
[C20H18ClNOPRh]·0.25CH2Cl2: C, 50.78; H, 3.89; N, 2.92; Cl,
11.10%).
(i) Chlorocarbonyl(tris(o-N,N-dimethylanilinyl)phosphine)rho-
dium(I) [RhCl(CO)(PAr¢3)] 9. The compound was prepared
as described in part (e) using [Rh(m-Cl)(COD)]2 (58 mg,
0.13 mmol) and PAr¢3 (92 mg, 0.24 mmol) and isolated as
a yellow solid (101 mg, 76.6%). Single crystals suitable for
X-ray crystallographic analysis were obtained by dissolving
the complex, under Ar atmosphere, in a minimum volume of
CH2Cl2 and layering the solution with anhydrous n-pentane
in an NMR tube (found: C, 52.81; H, 5.31; N, 7.32; Cl, 7.43.
Calc. for [C25H30ClN3OPRh]·0.11CH2Cl2: C, 53.20; H, 5.37; N,
7.41; Cl, 7.57%). Although the crystal structure indicates no
dichloromethane content, a microcrystalline sample was analyzed
1
here. Chloride analysis and H NMR analysis in CDCl3 (both
obtained at approximately the same time) were used to determine
dichloromethane content.
(j) Dichlorodicarbonyl(l-P,N,P¢,N¢-bis(di(o-N -methylanili-
nyl)phosphino)methane)dirhodium(I) [Rh2Cl2(CO)2(l-mapm)] 10.
In a 50 mL Schlenk flask under anhydrous conditions and Ar
atmosphere, [Rh(m-Cl)(CO)2]2 (77 mg, 0.20 mmol) and mapm
(103 mg, 0.206 mmol) were dissolved in THF (15 mL) by stirring at
ambient temperature. Solvent was slow◦ly removed from the bright
red–orange solution by heating to 40 C under a steady flow of
Ar(g). Dichloromethane (3 mL) was added to the resultant orange
solids yielding a red solution with a bright-yellow precipitate. The
precipitate was isolated by Schlenk filtration, washed three times
with 1 mL aliquots of dichloromethane and dried in vacuo (118 mg,
71%). Single crystals suitable for X-ray crystallographic analysis
were obtained by dissolving the complex, under Ar atmosphere,
in a minimum volume of CH2Cl2 and layering the solution with
anhydrous Et2O in an NMR tube (found: C, 42.60; H, 3.96; N,
6.15; Cl, 13.67. Calc. for [C31H34Cl2N4O2P2Rh2]·0.75CH2Cl2: C,
42.51; H, 3.99; N, 6.25; Cl, 13.68%).
(f) Chlorocarbonyl(di(o-N-methylanilinyl)phenylphosphine)rho-
dium(I) [RhCl(CO)(PhPAr2)] 6. The compound was prepared
as described in part (e) using [Rh(m-Cl)(COD)]2 (187 mg,
0.383 mmol) and PhPAr2 (245 mg, 0.765 mmol) and isolated
as a yellow solid (305 mg, 81.9%). Single crystals suitable for
X-ray crystallographic analysis were obtained by dissolving
the complex, under Ar atmosphere, in a minimum volume of
CH2Cl2 and layering the solution with anhydrous n-pentane
in an NMR tube (found: C, 48.82; H, 4.08; N, 5.38. Calc. for
[C21H21ClN2OPRh]·0.5CH2Cl2: C, 48.80; H, 4.19; N, 5.29%).
(g) Chlorocarbonyl(tri(o-N-methylanilinyl)phosphine)rhodium-
(I) [RhCl(CO)(PAr3)] 7.
Method a. The compound was prepared as described in part (e)
using [Rh(m-Cl)(COD)]2 (174 mg, 0.352 mmol) and PAr3 (246 mg,
0.704 mmol) and isolated as a yellow solid (348 mg, 95.8%). Single
crystals suitable for X-ray crystallographic analysis were obtained
by dissolving the complex, under Ar atmosphere, in a minimum
volume of CH2Cl2 and layering the solution with anhydrous n-
pentane in an NMR tube.
(k) Dichlorodicarbonyl(l-P,N,P¢,N¢-bis(di(o-N,N-dimethyla-
nilinyl)phosphino)methane)dirhodium(I) [Rh2Cl2(CO)2(l-dmapm)]
11. In a 100 mL Schlenk flask under anhydrous conditions
and Ar atmosphere, [Rh(m-Cl)(CO)2]2 (149 mg, 0.383 mmol)
and dmapm (227 mg, 0.408 mmol) were dissolved in 10 mL of
dichloromethane at ambient temperature and stirred. Stirring was
stopped after 30 min and a light Ar(g) stream was left blowing
over the saturated red–orange solution for 18 h producing cube-
shaped, orange–yellow crystals. The crystals were then washed
with 1 mL of dry dichloromethane and dried in vacuo (278 mg,
81.5%). Single crystals suitable for X-ray crystallographic analysis
were obtained by dissolving the complex, under Ar atmosphere,
in a minimum volume of CH2Cl2 and layering the solution with
anhydrous n-pentane in an NMR tube (found: C, 47.35; H, 4.89;
N, 6.37. Calc. for [C35H42Cl2N4O2P2Rh2]: C, 47.27; H, 4.76; N,
6.30%).
Method b. In a 50 mL Schlenk flask under anhydrous conditions
and Ar atmosphere, [Rh(m-Cl)(CO)2]2 (27 mg, 68 mmol) and PAr3
(48 mg, 0.14 mmol) were dissolved in dry THF (5 mL) at ambient
temperature. The yellow solution was stirred for 5 min before
10 mL of dry n-pentane were added and the resulting yellow
precipitate was allowed to settle before removing the supernatant
via cannula. The compound was then dried in vacuo (61 mg, 86%)
producing a yellow solid (found: C, 51.05; H, 4.83; N, 7.83. Calc.
for [C22H24ClN3OPRh]: C, 51.23; H, 4.69; N, 8.15%).
(h) Chlorocarbonyl(bis(o-N,N -dimethylanilinyl)phenylphos-
phine)rhodium(I) [RhCl(CO)(PhPAr¢2)] 8. The compound was
prepared as described in part (e) using [Rh(m-Cl)(COD)]2 (200 mg,
4216 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
(l) Bis ( trifluoromethanesulfonato)dicarbonyl(l - P, N, P¢,N¢ -
bis(di(o-N-methylanilinyl)phosphino)methane)dirhodium(I) [Rh2-
(OTf)2(CO)2(l-mapm)] 12. In a 25 mL Schlenk tube under an-
hydrous conditions and Ar atmosphere, [Rh2Cl2(CO)2(m-mapm)]
(51 mg, 61 mmol) and AgOTf (34 mg, 13 mmol) were dissolved
in 5 mL of dichloromethane at ambient temperature and stirred
for 12 h in the dark. The resultant orange–brown slurry was then
left unstirred and the precipitate allowed to settle before filtering
the orange solution through celite into a 50 mL Schlenk flask.
The solvent was removed in vacuo and the complex was washed
with 1 mL of dichloromethane before drying in vacuo (43 mg, 66%)
producing an orange solid (found: C, 37.34; H, 3.56; N, 5.63. Calc.
for [C33H34F6N4O8P2Rh2S2]: C, 37.37; H, 3.23; N, 5.28%).
(SADABS [5, 9, 10, 11, 14] or TWINABS [6]) or through Gaussian
integration from indexing of the crystal faces (7). Structures were
solved using the direct methods programs SHELXS–9758 (5, 7,
9, 10, 11) and SIR9759 (14), or the Patterson search/structure
expansion facilities within the DIRDIF-9960 program system (6).
Refinements were completed using the program SHELXL-97.58
Hydrogen atoms were assigned positions based on the sp2 or
sp3 hybridization geometries of their attached carbon or nitrogen
atoms, and were given thermal parameters 20% greater than those
of their parent atoms. See Table 2 for a listing of crystallographic
experimental data.
(b) Special refinement conditions. (i) Compound 5: attempts
to refine peaks of residual electron density as solvent (DCM) car-
bon or chlorine atoms were unsuccessful. The data were corrected
for disordered electron density through use of the SQUEEZE pro-
(m) Diiododicarbonyl(l-P,N,P¢,N¢-bis(di(o-N -methylanili-
nyl)phosphino)methane)dirhodium(I) [Rh2I2(CO)2(l-mapm)] 13.
In a 100 mL Schlenk flask under anhydrous conditions and Ar
atmosphere, [Rh2Cl2(CO)2(m-mapm)] (106 mg, 0.127 mmol) was
dissolved in 10 mL of dichloromethane at ambient temperature
and stirred. Under similar conditions, KI (207 mg, 1.25 mmol)
was dissolved in 8 mL of methanol at ambient temperature.
The KI solution was transferred to the [Rh2Cl2(CO)2(m-mapm)]
solution via cannula and the resultant orange solution was stirred
at ambient temperature for 1 h producing an orange–brown slurry.
The solvents were removed in vacuo yielding a brown solid. Water
(40 mL) was added with stirring and the product was extracted
with 3 ¥ 5 mL of dichloromethane into a 50 mL Schlenk flask.
The solution was stirred vigorously while adding 15 mL of Et2O
followed by 10 mL of n-pentane producing a yellow precipitate
which was allowed to settle before the supernatant was decanted.
The complex was then dried under a brisk flow of Ar and dried
further in vacuo (85 mg, 66%) producing a yellow solid (found: C,
36.38; H, 3.43; N, 5.15. Calc. for [C31H34I2N4O2P2Rh2]: C, 36.64;
H, 3.37; N, 5.51%).
cedure as implemented in PLATON.61 A total solvent-accessible
3
˚
void volume of 237.7 A with a total electron count of 83
(consistent with two molecules of solvent dichloromethane, or
0.25 molecules per formula unit of the complex molecule) was
found in the unit cell.
(ii) Compound 6: the crystal used for data collection was found
to display non-merohedral twinning. Both components of the twin
were indexed with the program CELL_NOW.62 The second twin
component can be related to the first component by 180◦ rotation
about the [-1/4 1 0] axis in real space and about the [0 1 0] axis
in reciprocal space. Using all reflection data (exactly overlapped,
partially overlapped and non-overlapped), integrated intensities
for the reflections from the two components were written into
a SHELXL-97 HKLF 5 reflection file with the data integration
program SAINT (version 7.06A).63
(iii) Compound 10: the disordered dichloromethane electron
density was treated in the same manner as for 5. A total solvent-
3
˚
accessible void volume of 445.8 A with a total electron count of
125 (consistent with three molecules of solvent dichloromethane,
or 0.75 molecules per formula unit of the complex molecule) was
found in the unit cell.
(n) Diacetatodicarbonyl(l-P,N,P¢,N¢-bis(di(o-N-methylanili-
nyl)phosphino)methane)dirhodium(I) [Rh2(OAc)2(CO)2(l-mapm)]
14. In a 50 mL Schlenk flask under anhydrous conditions and
Ar atmosphere, 25 mL of dry THF was added to [Rh2Cl2(CO)2(m-
mapm)] (161 mg, 0.193 mmol) and KOAc (186 mg, 1.90 mmol).
The resulting dark-red slurry was stirred for 18 h and then filtered
through celite. The solvent volume was reduced to approx. 2 mL
in vacuo before dry n-pentane was added and the resultant yellow–
brown slurry stirred for 5 min. The precipitate was allowed to settle
before the supernatant was removed via cannula. The complex
was then dried in vacuo (145 mg, 78.8%) producing a dark,
yellow–green solid (found: C, 47.57; H, 4.76; N, 6.01. Calc. for
[C35H40N4O6P2Rh2]: C, 47.74; H, 4.58; N, 6.36%). Single crystals
suitable for X-ray crystallographic analysis were obtained from a
saturated 1 : 1 THF–n-pentane solution under Ar atmosphere.
Results and discussion
P,N-Ligands
The simple, five-step, one-pot syntheses of the targeted (o-
N-methylanilinyl)phosphine compounds, 1–4, as illustrated in
Scheme 1, were carried out using the method reported by Budze-
laar for the synthesis of diphenyl(o-N-methylanilinyl)phosphine64
(1), which was in turn based on the methodology of Katritzky
et al.65 Synthetic versatility was achieved using the commer-
cially available phosphorus synthons, chlorodiphenylphosphine
(Ph2PCl), dichlorophenylphosphine (PhPCl2), trichlorophos-
phine (PCl3) and bis(dichlorophosphino)methane (Cl2PCH2PCl2)
in conjunction with the nitrogen-containing precursor, N-
methylaniline. None of the prepared P,N-ligands was sensitive
to air or water and all were readily soluble in ether, enabling
their purification by standard ether extraction. The toxic and
odorous byproducts of the hydrolysis of chlorophosphines are
typically water-soluble and were removed during the aqueous
work-up along with any unreacted lithium reagents. In general,
these ligands are thermally stable, white solids and can be purified
by recrystallization from a minimal amount of boiling ethanol.
X-Ray structure determinations
(a) General. Data for compounds 5, 6, 7, 9 and 11 were col-
lected using a Bruker SMART 1000 CCD detector/PLATFORM
diffractometer57 using Mo Ka radiation, with the crystals cooled
to -80 ◦C. Data for compound 14 were collected using a
Bruker APEX II CCD detector/D8 diffractometer57 using Mo
Ka radiation, with the crystal cooled to -100 ◦C. The data
were corrected for absorption through use of a multi-scan model
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4217
©

View Article Online
4218 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Scheme 1 Ligand syntheses (1–4).
We suggest that increased steric congestion at phosphorus after
each subsequent ortho-arylation of the phosphine tends to hinder
production of the more heavily aminated P,N-ligands as illustrated
by the lower yields of these targets.
with [Rh(m-Cl)(COD)]2 at ambient temperature under strictly
inert conditions in dichloromethane, before passing carbon
monoxide through the reaction mixtures (Scheme 2). The
complexes were then precipitated by addition of n-pentane
and were obtained in moderate to high yields. A more di-
rect route using [Rh(m-Cl)(CO)2]2 as a starting material had
previously been exploited by Roundhill et al. to prepare the
N,N-dimethylanilinyl compound, [RhCl(CO)(Ph2PAr¢)],68 and we
have also used this methodology to prepare [RhCl(CO)(PAr3)]
(7). The monophosphines, PhPAr¢2 and PAr¢3, first prepared
by Venanzi and coworkers,54 have also been used to prepare
the N,N-dimethyl analogues of 6 and 7, [RhCl(CO)(PhPAr¢2)]
(8) and [RhCl(CO)(PAr¢3)] (9), respectively. At 27 ◦C the 1H
NMR signal for the NMe protons of [RhCl(CO)(Ph2PAr)]
The challenge of synthesizing (o-N-methylanilinyl)phosphines
can be attributed to the reactivity of the 2◦ amino group of N-
alkylanilines66 that, under basic nucleophilic conditions, leads
to unwanted side reactions, thereby necessitating its protection
(with CO2 to afford the O-lithiocarbamate) prior to ortho-
functionalization of the arene. o-Metallation to afford the dilithi-
ated intermediate is problematic and rigorous exclusion of air and
moisture is required. In this step it is necessary to use the more
basic t-BuLi as the o-metallating agent since n-BuLi failed to react
with the O-lithiocarbamate precursor. For example, in attempts
to use n-BuLi as the o-metallating agent for the preparation of
compound 4, bis(di-n-butylphosphino)methane—resulting from
reaction of the precursor, bis(dichlorophosphino)methane, with
n-BuLi that had failed to react in the o-metallation step—was
instead isolated in quantitative yield.
3
(5) appears as a fully resolved doublet with JHH = 6.5 Hz.
The complexes, [RhCl(CO)(PhPAr2)] (6), [RhCl(CO)(PAr3)] (7),
[RhCl(CO)(PhPAr¢2)] (8) and [RhCl(CO)(PAr¢3)] (9), which con-
tain coordinated and pendent amine groups (vide infra) all
1
display only a single H NMR resonance for N-methyl protons
Very recently, Lee and co-workers have reported that addition
of 1 equiv. of THF in diethyl ether significantly enhanced
product yields for syntheses involving the o-metallation of
tetrahydroquinoline67 derivatives with t-BuLi. We have not yet
used this methodology to determine the effect of adding stoichio-
metric THF on product yields of (o-N-methylanilinyl)phosphines.
1H NMR spectra of the ligands exhibit broad NH signals (due
to quadrupolar broadening by nitrogen) between dH 4.6 and 4.9
with the general trend that less shielded NH protons belong to the
at ambient temperature, indicating the rapid exchange of these
coordinated and pendent groups—a feature indicative of the (Type
II) hemilabile nature of these complexes.21 Within the series of
compounds, 5–9, the greater the number of anilinyl substituents
on the phosphine, the greater the shielding of the 31P nuclei and
the greater the 1JRhP (Table 1).
1
more heavily aminated phosphines. The H NMR signals of the
NMe protons at ca. dH 2.8 appear either as broad singlets or as
doublets at ambient temperature, the latter situation arising from
observable, vicinal coupling to the NH protons (3JHH = ca. 5 Hz).
1
31P{ H} NMR signals within the series of monophosphines show
a significant upfield shift of the singlet resonances (from dP -21.8
to -53.6) as the number of amino substituents increases (from one
to three), whereas the diphosphine mapm (4) exhibit a signal at an
even higher field, at dP -60.9.
Scheme 2 Syntheses of mononuclear rhodium complexes (5–7).
In order to determine how the degree of N-methyl substitution
affects the lability of the anilinyl groups, we carried out variable
temperature NMR experiments on the related dimethyl- and
monomethylanilinyl complexes, [RhCl(CO)(PhPAr¢2)] (8; Ar¢ =
C6H4NMe2) and [RhCl(CO)(PhPAr2)] (6; Ar = C6H4NHMe),
Mononuclear complexes
Mononuclear rhodium complexes were readily prepared by
the reaction of the above monophosphine P,N-ligands (1–3)
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4219
©

View Article Online
respectively. Upon cooling to -20 ◦C, 1H NMR analysis of
[RhCl(CO)(PhPAr¢2)] (8) reveals significant broadening of the
single resonance representing all N-methyl protons, and at -71 ◦C
three distinct signals are evident, in a 1 : 1 : 2 intensity ratio—two
for the diastereotopic methyl groups of the coordinated amine, and
one for both methyl groups of the pendent amine (Scheme 3)—
indicating that amine exchange at rhodium is slow on the NMR
timescale at that temperature. Upon further cooling to -79 ◦C, four
distinct N-methyl proton resonances of equal intensity are present
in the spectrum suggesting that lone-pair inversion of the pendent
amine has slowed to allow the resolution of the two chemically
unique environments at this nitrogen.
at the temperatures 13, -20 and -60 ◦C were used to calculate
DG for this process [eqn (1)], giving a value of 2.3 0.5 kJ mol-1
at 95% confidence. We assume that the major diastereoisomeric
pair corresponds to the pair of enantiomers B/B¢, on the basis of
steric considerations, in which the methyl group on the coordinated
anilinyl moiety avoids the larger pendent anilinyl substituent in
favor of the smaller phenyl group, and on the basis of its having a
lower dipole moment which should be favored in the low-polarity
solvent. This is also the structure found in the solid state for 6 (vide
infra).
DGdiaster = -RT ln(Kdiaster
)
(1)
◦
Cooling to 13 C results in the resolution of the two different
N-methyl signals (for the coordinated and pendent amines) of
the major enantiomeric pair into two doublets indicating the
coalescence point for this enantiomerization. At 10 ◦C the N-
methyl signal for the minor enantiomeric pair begins to split
into two more doublets indicating the coalescence point for the
enantiomerization of the minor stereoisomers (proposed to result
from A ꢀ A¢, Scheme 4).
Line-shape analyses for the methyl resonances of
[RhCl(CO)(PhPAr2)] (6) and [RhCl(CO)(PhPAr¢2)] (8) were
undertaken to compare the rates of exchange processes within
these compounds along with the corresponding values of DG‡,
calculated using eqn (2).69
Scheme 3 Enantiomerization of [RhCl(CO)(PhPAr¢2)] (8).
The variable-temperature 1H NMR spectroscopic study of
[RhCl(CO)(PhPAr2)] (6) proves to be more complicated than that
of its dimethylated counterpart (8). As is immediately apparent
from Scheme 4, any particular coordination geometry of the
complex possesses two stereogenic centers (one at phosphorus, the
other at the coordinated amine) giving rise to four stereoisomers
existing as two diastereomeric pairs of enantiomers (A/A¢ and
B/B¢). The 1H NMR spectrum of [RhCl(CO)(PhPAr2)] (6) at
27 ◦C reveals two broad, nearly coalescing signals representing two
diastereomeric pairs of enantiomers, the intensity ratio of which
is approximately 3 : 1, suggesting a thermodynamic preference for
one pair of rapidly interconverting enantiomers over the other. The
relative concentrations of the diastereomers, which also provide a
measure of the equilibrium constant for the diastereoisomerization
via amine–donor exchange, only vary from 2.80 at 13 ◦C, to 3.20 at
-60 ◦C. Subsequently, values of K for the diastereoisomerization
k = (kBT)/h exp[–DG‡/(RT)]
(2)
For compound 6, the rate of enantiomerization for the major
stereoisomers (presumably kBB¢ = kB¢B) was determined as 17 s-1
at 286 K and this value was used to calculate DG‡ (286K) =
63.2 kJ mol-1. Similarly, the rate of enantiomerization for the
minor stereoisomers (presumably kAA¢ = kA¢A) was determined as
25 s-1 at 283 K and this value was used to calculate DG‡ (283K) =
61.6 kJ mol-1. Exchange parameters for the enantiomerization of
the N,N-dimethyl analogue, 8, were also obtained: k(194 K) =
274 s-1, DG‡ (194K) = 38.4 kJ mol-1. A comparison of the
DG‡ values for compounds 6 and 8 indicates that, despite its
stronger Lewis basicity, the N,N-dimethylanilinyl group of 8
renders the complex much more labile than its monomethylated
counterpart, 6. This labilization of the dimethylanilinyl donor can
be rationalized on the basis of the more severe steric repulsion
involving its more highly substituted anilinyl groups (vide infra).
The H NMR spectrum of [RhCl(CO)(PAr3)] (7) at -20 ◦C
1
displays three well-resolved doublets representing all mutually
non-equivalent N-methyl groups, while at 27 ◦C, a rapid three-
site exchange process results only in a broad singlet. The 1H
NMR spectrum of [RhCl(CO)(PAr¢3)] (9) exhibits only one signal
for all methyl protons at room temperature but, interestingly, at
-80 ◦C only two methyl signals are observed, each integrating as 9
protons. The appearance of two equal-intensity methyl resonances
in the low-temperature spectrum of 9 can be rationalized by the
geometry of the PAr¢3 ligand of 9 in the solid state (vide infra),
in which the pendent amine groups each have methyl groups in
clearly different environments. It appears that rapid exchange of
the amine donors at rhodium by rotation about the Rh–P bond,
even at -80 ◦C, occurs in a propeller-like manner, resulting in two
chemically distinct average methyl environments in solution.
In order to compare the structural differences between the
monomethyl- and dimethylanilinyl analogues, the single-crystal
Scheme 4 Possible isomerization mechanisms of [RhCl(CO)(PhPAr2)]
(6) depicting the possibility of four kinetically independent mechanisms of
amine donor exchange at rhodium.
4220 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
X-ray structures of the three mononuclear [RhCl(CO)(L)] com-
plexes (L = Ph2PAr (5), PhPAr2 (6), PAr3 (7); Ar = o-C6H4NHMe)
have been determined and are compared to the previously
reported dimethylanilinyl complex, [RhCl(CO)(Ph2PAr¢)] (Ar¢ =
o-C6H4NMe2).70 In addition, we have determined the structure of
the compound [RhCl(CO)(PAr¢3)] (9) as a further comparison.
The ORTEP diagrams of compounds 5, 6, 7 and 9 are shown
in Fig. 1 and a comparison of their structural parameters,
along with those of [RhCl(CO)(Ph2PAr¢)], is given in Table 3.
All compounds have the expected square-planar geometry at
rhodium in which the carbonyl ligand is opposite the weaker
trans-directing amine group while the chloro ligand is opposite
the phosphine moiety which has the greater trans effect. All five
compounds also have quite comparable structural parameters in
which the bond lengths and angles are as expected. Certainly,
within the series of monomethylanilinyl complexes (5–7) all
related parameters are closely comparable, indicating that the
incorporation of additional N-methylanilinyl groups (Ph2PAr vs.
PhPAr2 vs. PAr3) has no obvious structural influence on the
metal coordination geometries, although minor differences in
the orientations of the aryl groups are observed between the
three complexes. Similarly, within the pair of dimethylanilinyl
compounds the structural parameters are closely comparable.
However, a comparison of the monomethyl- and dimethylanilinyl
compounds shows significant differences between the two classes.
A visual comparison of the two trisubstituted species 7 and 9,
shown in Fig. 1, indicates that the most significant differences
between the monomethyl- and dimethylanilinyl analogues relate to
the coordinated amine groups. In the case of the dimethylanilinyl-
containing compounds, [RhCl(CO)(Ph2PAr¢)] and 9, the Rh–N
˚
distance (2.1947(6) and 2.2019(6) A, respectively) is greater
than that for the three monomethylanilinyl-containing species
˚
(av. 2.135(5) A). This lengthening for the dimethylated compounds
is also accompanied by a slight widening of the N–Rh–Cl angle,
which is greater than 90◦ for the dimethylanilinyl compounds
and less than 90◦ for the monomethylanilinyl analogues. Both
differences appear to result from the greater steric crowding in
the dimethylamines, which weakens the Rh–N bond and gives
rise to greater repulsions involving the adjacent chloro ligand.
These structural comparisons are consistent with the significantly
greater lability of the dimethylanilinyl species as discussed above
for compounds 6 and 8.
We had initially intended to compare the above monophos-
phine complexes with the mononuclear diphosphine equivalent,
[RhCl(CO)(P,N-mapm)], for which we had assumed a phosphine
binding mode, analogous to compounds 5–9 (in which the
diphosphine ligand is bound to Rh via one phosphorus and an
adjacent amine, while the other end of the diphosphine remains
uncoordinated and pendent), would be observed. The related
complex, [RhCl(CO)(P,N-dmapm)], was previously shown to
have this structure type.43 However, all attempts to prepare this
mononuclear mapm analogue gave the binuclear diphosphine-
bridged species, [Rh2Cl2(CO)2(m-mapm)] (vide infra), as the major
Fig. 1 ORTEP diagrams showing one of two crystallographically
independent molecules of [RhCl(CO)(Ph2PAr)] (5), and one of two
crystallographically independent molecules of [RhCl(CO)(PhPAr2)] (6),
[RhCl(CO)(PAr3)] (7) and [RhCl(CO)(PAr¢3)] (9). Gaussian ellipsoids
for all non-hydrogen atoms are depicted at the 20% probability level.
Hydrogens are shown artificially small, except for aryl hydrogens which
are omitted.
Table 3 Selected structural parameters for the mononuclear complexes
[RhCl(CO)(Ph2PAr)] (5)
[RhCl(CO)(PhPAr2)] (6)
[RhCl(CO)(PAr3)] (7)
[RhCl(CO)(Ph2PAr¢)]70
[RhCl(CO)(PAr¢3)] (9)
˚
˚
˚
˚
˚
Atoms
Rh–P
Rh–N(1)
Rh–C(1)
Rh–Cl
Bond lengths/A
Bond lengths/A
Bond lengths/A
Bond lengths/A
Bond lengths/A
2.1933(7), 2.1909(7)a
2.129(2), 2.140(2)
1.819(3), 1.809(3)
2.3936(7), 2.3757(7)
2.2150(8), 2.2035(8)a
2.131(2), 2.139(2)
1.825(3), 1.819(3)
2.3786(8), 2.3797(8)
2.2199(4)
2.1368(13)
1.8160(17)
2.3787(4)
2.1947(6)b
2.1865(2)
1.807(2)
2.2019(6)
2.1883(18)
1.801(3)
2.3867(7)
2.3941(6)
Atoms
Angles/◦
Angles/◦
Angles/◦
84.05(4)
88.16(4)
93.16(5)
95.00(6)
Angles/◦
85.03(5)
91.11(5)
92.62(7)
91.24(7)
Angles/◦
84.62(5)
92.17(5)
90.08(7)
93.86(8)
P–Rh–N(1)
Cl–Rh–N(1)
Cl–Rh–C(1)
P–Rh–C(1)
83.49(6), 83.24(7)
86.69(6), 89.11(7)
98.04(9), 95.36(10)
91.83(9), 92.65(10)
83.28(6), 82.83(7)
88.13(6), 87.16(7)
92.87(9), 96.06(10)
95.71(9), 94.07(10)
a Two crystallographically independent molecules. b Correct bond lengths and angles for [RhCl(CO)(Ph2PAr¢)] obtained from Table 5 within ref. 70.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4221
©

View Article Online
product accompanied by minor amounts of uncharacterized
side products. None of these side products displayed spectra
characteristic of our targeted mononuclear species. It appears
that the greater steric accessibility of the mapm ligand favors
the formation of the bimetallic complexes over the mononuclear
pendent complexes.
analysis which shows strong correlations to N-methyl resonances)
while the methylene group of the bridging mapm ligand appears
as a multiplet at dH 3.94. The N-methyl groups appear as two
sharp doublets at dH 3.17 and 2.78. The downfield NH signal of
the (presumably) coordinated amine (dH 7.75) exhibits a strong
GCOSY correlation to the more upfield NMe signal (dH 2.78)
while the more upfield NH signal (dH 6.94) shows a similar
correlation to the more downfield NMe signal (dH 3.17) providing
a means for the assignment of coordinated and pendent NMe
Binuclear complexes
The binuclear mapm-bridged complex [Rh2Cl2(CO)2(m-mapm)]
(10; mapm = Ar2PCH2PAr2) was prepared, as alluded to above, by
adding THF to a flask containing [Rh(m-Cl)(CO)2]2 and mapm (4)
at ambient temperature; the dmapm analogue, [Rh2Cl2(CO)2(m-
dmapm)] (11; dmapm = Ar¢2PCH2PAr¢2) was prepared by a
similar procedure (Scheme 5). Compound 11 had been previously
reported71 but had not been structurally characterized. We were
interested in establishing the structural differences that would
result from substituting the amine hydrogen in 10 by a methyl
group, and also in whether such a substitution would influence
the lability of the coordinated amine groups. Both compounds
display a single carbonyl stretch in the IR spectrum at around
2000 cm-1, characteristic of Rh(I), and also show a doublet of
1
signals. In the 31P{ H} NMR spectrum a well-resolved multiplet,
resembling the low-temperature resonance for 11, appears at
1
dP 23.3. The observed and simulated56 31P{ H} NMR spectra,
assuming an AA¢XX¢ spin system, for compounds 10 and 11 are
given in Fig. 2. All derived parameters (10: 1JRhP = 157 Hz, 2JPP
=
3
2
1
53.6 Hz, JRhP = 3.2 Hz, JRhRh = -0.05 Hz; 11: JRhP = 176 Hz,
2JPP = 46.1 Hz, 3JRhP = 2.7 Hz, 2JRhRh = -0.05 Hz) are consistent
with those reported for the related diphosphine-bridged species
[Rh2(m-Cl)(COD)2(m-dppm)][BF4] (dppm = Ph2PCH2PPh2).72 As
noted for the monophosphine compounds, substitution of the
amine hydrogen in 10 by a methyl group to give 11 substantially
labilizes this coordinated amine, again probably due to steric
repulsions between this larger tertiary amine and other ligands on
Rh. Although exchange between the free and coordinated aniline
groups in 11 is facile at ambient temperature, there is no evidence
of fluxionality at this temperature for 10.
1
doublets for the pair of carbonyls in the 13C{ H} NMR spectra at
around dC 186, displaying typical one-bond coupling to Rh and
two-bond coupling to P (see Table 1). At ambient temperature
the 1H NMR spectrum of 11 shows a well-resolved broad triplet
resonance at dH 4.59 (2JPH = 12.4 Hz) for the methylene group
of the dmapm ligand, but shows only very broad, unresolved
resonances for the methyl groups, between approximately dH 2.2
and 3.7, and for the aromatic protons. Upon cooling to -80 ◦C the
methyl resonances appear as sharp singlets at dH 3.68, 2.86, 2.70
and 2.27, each integrating as six protons while the signal for the
methylene protons also sharpens significantly.
Scheme
5
Preparations of [Rh2Cl2(CO)2(m-mapm)] (10) and
1
Fig. 2 Calculated (A) and observed (B) 31P{ H} NMR spectra of
[Rh2Cl2(CO)2(m-dmapm)] (11).
◦
compound 10 at 27 C. Calculated (C) and observed (D) 31P{ H} NMR
1
spectra of compound 11 at -40 ◦C.
This temperature dependence suggests fluxionality, presumably
involving the sequential exchange of dimethylanilinyl groups at
each Rh via a transient Cs-symmetric intermediate that renders
the methylene hydrogens inequivalent. The 1H NMR temperature
dependence is paralleled by differences in the 31P NMR spectra
in which the diphosphine appears as a broad doublet at dP
41.0 (1JRhP = 173 Hz) at ambient temperature but sharpens
to a well-resolved multiplet characteristic of an AA¢XX¢ spin
system at -80 ◦C. In contrast, the resonances in the 1H and
In order to gain a better understanding of the influence of
the additional N-methyl substituent we have carried out the
X-ray structure determination of compounds 10 and 11 and
both structures are shown in Fig. 3, with a summary of metrical
parameters given in Table 4. Both structures are similar in having
a face-to-face arrangement of the two Rh square planes that are
bridged by the diphosphine unit of mapm or dmapm and both
square planes are also staggered with respect to each other by
approximately 40◦ (10) and 44◦ (11), allowing the ligands on
one metal to avoid those on the other. Furthermore, in both
cases the coordinated aniline group on one metal occupies one
side of the approximate Rh2P2 plane while that on the other
1
the 31P{ H} NMR spectra of the mapm analogue (10) are sharp
and well resolved, showing no evidence of fluxionality over the
full temperature range between 25 ◦C and -80 ◦C. In the 1H
NMR spectrum the amine hydrogens overlap two aromatic proton
resonances at dH 7.75 and 6.94 (as indicated by GCOSY NMR
4222 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 4 Selected structural parameters for [Rh2Cl2(CO)2(m-mapm)] (10)
and [Rh2Cl2(CO)2(m-dmapm)] (11)
indicate that these groups are close to their minimum sepa-
ration. An additional consequence of the above repulsions is
the slight bending of the Cl ligands away from these contacts
(P(1)–Rh(1)–Cl(1) = 174.51(6)◦, 175.79(5)◦; P(2)–Rh(2)–Cl(2) =
175.62(5)◦, 176.26(5)◦). Furthermore, the repulsions that force the
Rh planes apart in 11 give rise to significantly enlarged Rh–P–C(3)
(av. 122.5 ◦) and P(1)–C(3)–P(2) (av. 120.0 ◦) angles compared to
those in 10 (116.4 ◦ and 114.0◦, respectively).
[Rh2Cl2(CO)2(m-
mapm)] (10)
[Rh2Cl2(CO)2(m-
dmapm)] (11)
˚
˚
Atoms
Distances/A
Distances/A
Rh(1)–Rh(2)
Rh(1)–Cl(1)
Rh(2)–Cl(2)
Rh(1)–N(1)
Rh(2)–N(3)
Cl(1)–H3N
Cl(2)-H1N
3.4500(4)
2.3810(9)
2.393(1)
2.140(3)
2.162(3)
2.51
4.1211(7), 4.3185(6)a
2.398(1), 2.412(1)
2.396(1), 2.409(1)
2.181(4), 2.217(4)
2.200(4), 2.202(4)
In contrast, the contacts between the chloro ligands in 10 and
the amino hydrogens (Cl(1)–HN(3) and Cl(2)–HN(1)) associated
˚
with the adjacent metal, at 2.51 A, are much shorter than a
2.51
normal van der Waals separation73 and indicate the presence of
a reasonably strong hydrogen bond that actually appears to be
pulling the metal coordination planes together. This attraction
is further manifested in a slight bending of the chlorine ligands
towards the amine to which it is hydrogen-bonded (P(1)–Rh(1)–
Cl(1) = 171.19(4)◦; P(2)–Rh(2)–Cl(2) = 171.10(4)◦). Interestingly,
the pendent methylanilinyl groups in 10 are oriented such that
the amino hydrogens are aimed towards the adjacent metals
above and below the vacant coordination sites on the outsides of
the face-to-face dimer. However, these Rh(1)–HN(2) and Rh(2)–
Atoms
Angles/◦
116.2(1)
116.6(1)
114.0(2)
Angles/◦
124.8(2), 120.1(2)
122.3(2), 122.8(2)
118.0(2), 122.0(3)
Rh(1)–P(1)–C(3)
Rh(2)–P(2)–C(3)
P(1)–C(3)–P(2)
a Two crystallographically independent molecules.
˚
HN(4) contacts (2.59 and 2.57 A, respectively) appear to be
normal and do not suggest an attractive interaction between these
hydrogens and the metals. As a consequence, the Rh(1)–P(1)–
C(21) and Rh(2)–P(2)–C(41) angles (123.8(1)◦ and 122.8(1)◦) are
much larger than the other angles at phosphorus which range from
102.8(1)◦ to 105.8(2)◦, suggesting that the above Rh–H contacts
are repulsive, forcing the anilinyl groups away from the metals
1
slightly. Certainly, the downfield shift in the H NMR spectrum
of these amine hydrogens (dH 6.94) argues against an agostic
interaction in solution, for which we would expect an upfield
shift. Nevertheless, the Rh–H contacts observed in the solid state
cannot be too unfavorable, given the orientation of the pendent
methylanilinyl groups which project the amine hydrogens into
the vicinities of the two metals, rather than away from them as
observed for the dimethylanilinyl groups in 11.
Fig. 3 ORTEP diagrams of [Rh2Cl2(CO)2(m-mapm)] (10, left) and
[Rh2Cl2(CO)2(m-dmapm)] (11, right). Thermal ellipsoids as in Fig. 1.
metal coordinates to the opposite face such that both complexes
are C2-symmetric. In addition, a slight lengthening of the Rh–
˚
Cl bonds is observed in 11 (av. 2.404 A) compared to 10 (av.
The apparent lack of amine lability in complex 10 prompted our
attempts to prepare a species with inherently lower coordinative
saturation at the bimetallic core. Specifically, a cationic, chloro-
bridged complex similar to [Rh2(COD)2(m-Cl)(m-dppm)][BF4] 72
was targeted in which the remaining chloride was bridging and
could serve as a source of coordinative unsaturation. The targeted
complex, [Rh2(m-Cl)(CO)2(m-mapm)]+, involving mapm as the
bridging diphosphine, was selected due to the proximity of the
Rh centers of the parent complex 10 relative to the dmapm
analogue, 11. Unfortunately, reaction of 10 with a variety of silver
salts, including AgBF4, AgPF6 and AgOTf, failed to yield the
monochloride. Reactions of 10 with AgBF4 and AgPF6 under
a variety of conditions and solvent systems routinely resulted in
multiple decomposition products, while reaction of 10 with AgOTf
yielded the symmetric disubstituted species, [Rh2(OTf)2(CO)2(m-
mapm)] (12) (Scheme 6). The molar conductivity of 12 in CH3NO2
was determined to be 23 cm2 X-1 mol-1, suggesting some degree
˚
2.387 A), probably also due to repulsions involving the more
bulky dimethylaniline groups. Again, as for the monophosphine
analogues discussed above, the Rh–N distances are longer for
˚
the dimethylanilinyl complex 11 (2.181(4)–2.217(4) A) compared
˚
to the monomethylanilinyl species 10 (2.140(3), 2.162(3) A)—a
consequence of steric crowding in the former.
However, the major differences between compounds 10 and
11 become obvious on visual comparison of the two in Fig. 3
which demonstrates that the two Rh square planes in 11 are
significantly tilted, resulting in a much larger Rh ◊ ◊ ◊ Rh separation
˚
(4.1211(7), 4.3185(6) A for the two independent molecules) than
˚
in 10 (3.4500(4) A). This tilt of the Rh square planes in 11 is
evident in the dihedral angle between these planes of 27.78(5)◦
and 24.5(1)◦ in the two independent molecules, whereas the Rh
planes in 10 are close to parallel (dihedral angle = 2.93(8)◦). The
opening up of the cavity between the two Rh square planes in
11 clearly results from repulsion between the chloro ligand on one
metal and one methyl of the coordinated dimethylanilinyl group on
1
of triflate dissociation. Additionally, 19F{ H} NMR spectroscopy
of 12 at 27 ◦C reveals two broad coalescing signals suggesting
˚
the adjacent metal, leading to close contacts of ca. 2.94 A between
two chemically distinct fluorine environments by a possible ex-
1
˚
Cl(1) and the methyl hydrogens on C(37), and of ca. 2.92 A between
change of inner- and outer-sphere triflate moieties. 31P{ H} NMR
Cl(2) and the C(17) methyl group. Both separations are slightly
spectroscopy of 12 shows a downfield shift (relative to the parent
complex, 10) of the multiplet resonance (dP 31.5, 1JRhP = 176 Hz)
73
˚
less than a normal van der Waals separation of 3.00 A and
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4223
©

View Article Online
Scheme 6 Chloride-replacement reactions of [Rh2Cl2(CO)2(m-mapm)]
(10).
and the 1H NMR spectrum of 12 is similar to that of 10. In spite of
some degree of apparent triflate ion dissociation, we were unable to
isolate the presumed cationic monotriflate species. Expecting that
exchange of the chloro substituents of 10 for the larger iodide ions
could favor formation of an iodide-bridged species, we synthesized
[Rh2I2(CO)2(m-mapm)] (13, Scheme 6) through reaction of 10 with
a five-fold excess of KI in dichloromethane–methanol. However,
subsequent reactions of 13 with AgBF4, AgPF6 and AgOTf
yielded either decomposition in the first two cases or the bis-
triflato species 12, as was observed for 10. Although the targeted,
cationic, halogeno-bridged complexes could not be prepared, the
weakly coordinating sulfonate ligands of 12 appear to be labile, as
suggested by the molar conductivity and the observation of both
Fig. 4 ORTEP diagram of [Rh2(OAc)2(CO)2(m-mapm)] (14). Thermal
ellipsoids as in Fig. 1.
˚
1.84 A). In spite of the larger acetate compared to chloro ligand,
˚
the Rh(1)–Rh(2) separation in 14 (3.2227(2) A) is actually less
˚
than in 10 (3.4500(4) A).
This mutual approach of both metals results in a slight
pyramidalization of both square planes, as shown in Fig. 4, with
˚
˚
the metals being 0.08 A and 0.07 A out of the planes defined by the
four attached ligands. This distortion appears not to result from
any mutual attraction of the metals, but instead from repulsion
due to pendent amine hydrogens above and below the pair of
almost-parallel square planes. We had noted for compound 10
1
free and coordinated triflate ions in solution. The 31P{ H} NMR
spectrum of 13, although similar to 10, shows a slight upfield
1
shift of the multiplet resonance to dP 20.3 with JRhP = 162 Hz.
1
As expected, the ambient temperature H NMR spectrum of 13
˚
that these contacts (~2.58 A) were probably repulsive; in 14 the
1
is very similar to that of 10. Unfortunately, 13C{ H} spectra for
˚
Rh–HN contacts (~2.50 A) are even shorter and in this case lead
complexes 12 and 13 could not be obtained due to their poor
solubilities in a variety of solvents.
to a significant deviation of the metals from their respective planes.
Again, the very low-field chemical shifts of these protons (dH 8.91,
8.12) argue against any type of agostic interaction in solution, for
which we would expect a significant upfield shift.
The acetato complex, [Rh2(OAc)2(CO)2(m-mapm)] (14), could
also be prepared by reaction of 10 with KOAc in THF. The low
molar conductivity of 14 (12 cm2 X-1 mol-1 in CH3NO2) suggests
little acetate-ion dissociation, and may result from minor amounts
of salt impurities. Interestingly, compound 14 exhibits strong
solvatochromic tendencies, transforming from a deep-red solution
to a dark, yellow–green powdery solid upon removal of solvent
Conclusions
A number of P,N-ligated, mono- and binuclear complexes
of rhodium have been synthesized and fully characterized.
The complexes [RhCl(CO)(PhPAr2)] (6), [RhCl(CO)(PAr3)]
(7), [RhCl(CO)(PhPAr¢2)] (8), [RhCl(CO)(PAr¢3)] (9) and
[Rh2Cl2(CO)2(m-dmapm)] (11) are shown by NMR to display
fluxional behavior consistent with the hemilabile nature of
these systems. The more highly substituted dimethylanilinyl
ligands are found to be more labile than the monomethyl
analogues. X-Ray structural comparisons of related dimethyl-
and monomethylanilinyl species show greater steric repulsions
and concomitant weaker Rh–amine interactions for the former,
offering a rationalization for the greater lability of the NMe2-
substituted ligands. The N-methylamino-tethered, binuclear com-
plex, [Rh2Cl2(CO)2(m-mapm)] (10), has a greatly reduced inter-
atomic Rh ◊ ◊ ◊ Rh separation compared to its N,N-dimethylated
counterpart, [Rh2Cl2(CO)2(m-dmapm)] (11), owing to greater
steric repulsions between the two Rh coordination planes in the
latter case. While the mapm-bridged binuclear complex is expected
to have greater potential for bimetallic cooperativity owing to
significantly closer approach of the metals, it may be that the
1
in vacuo. Furthermore, the H NMR spectrum of this complex
exhibits highly deshielded NH protons at dH 8.91 and 8.12 which
were identified, via GCOSY analysis, by their strong correlations
1
to NMe protons. 31P{ H} NMR data show the expected multiplet
1
at dP 28.2 with 1JRhP = 155 Hz. 13C{ H} NMR data could not be
obtained from CD2Cl2 due to decomposition to multiple products
in solution over a 24 h period. Interestingly, one of these multiple
decomposition products has been identified as [Rh2Cl2(CO)2(m-
mapm)] via 31P NMR spectroscopy. It was also noticed that 14 is
quite hygroscopic as a solid, as the incorporation of water, which
1
can be evidenced by H NMR, led to its slow decomposition to
multiple uncharacterized products, as indicated by 31P NMR.
The structure of 14 has been determined crystallographically,
and an ORTEP diagram of this species is shown in Fig. 4. As is
obvious from a comparison of Fig. 3 and 4, compounds 10 and
14 have closely related structures. Again, the hydrogen atoms of
the coordinated amine are hydrogen bonded to the anionic ligand
(in this case acetate) on the adjacent metal, as demonstrated by
˚
the close O–H contacts (O(4)–H(3N) = 1.94 A; O(6)–H(1N) =
4224 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
lower steric bulk of the monomethylanilinyl group, which allows
this closer approach, may actually work to the detriment of the
system, owing to the lower lability of these groups. What effects
the two competing influences will have must await subsequent
reactivity studies.
12 J. Andrieu, P. Braunstein, A. Tiripicchio and F. Ugozzoli, Inorg. Chem.,
1996, 35, 5975–5985.
13 T. K. Ronson, H. Adams and M. D. Ward, Inorg. Chim. Acta, 2005,
358, 1943–1954.
14 M. A. Jalil, S. Fujinami, T. Honjo and H. Nishikawa, Polyhedron, 2001,
20, 1071–1078.
15 K. Mashima, Y. Kaneda, A. Fukumoto, M. Tanaka, K. Tani, H.
Nakano and A. Nakamura, Inorg. Chim. Acta, 1998, 270, 459–
466.
16 M. E. Broussard, B. Juma, S. G. Train, W.-J. Peng, S. A. Laneman and
G. G. Stanley, Science, 1993, 260, 1784–1788.
17 J. P. Farr, M. M. Olmstead, C. H. Hunt and A. L. Balch, Inorg. Chem.,
1981, 20, 1182–1187.
18 F. E. Wood, J. Hvoslef, H. Hope and A. L. Balch, Inorg. Chem., 1984,
23, 4309–4315.
19 J. C. Jeffrey and T. B. Rauchfuss, Inorg. Chem., 1979, 18, 2658–2666.
20 C. S. Slone, D. A. Weinberger and C. A. Mirkin, Prog. Inorg. Chem.,
1999, 48, 233–350.
Our failure to prepare cationic, halide-bridged species probably
results from the strain inherent in such a product, in which the
halide bridge would be required to lie opposite both ends of the
bridging diphosphine. In addition, the staggered arrangement of
the Rh coordination planes in the dichloro precursor (10) appears
necessary in order to minimize unfavorable contacts between
these planes. Replacement of one chloride ligand by a bridging
arrangement of the remaining chloride would force an eclipsed
conformation of the planes leading to a closer and less favorable
approach of the anilinyl and carbonyl groups on adjacent metals.
Nevertheless, it should still be possible to achieve an anion-bridged
structure through the use of bidentate groups such as acetates,
which should give rise to less strain while maintaining more
favorable contacts between the planes, although we have until now
failed to isolate such species in this chemistry.
The subsequent chemistries of the mapm-bridged species 10 and
12–14 will be investigated in order to determine whether ligand
hemilability and effects of metal–metal cooperativity will play a
role. Furthermore, the potential of using the acetate moieties in 14
as an internal base for deprotonation of one or more of the amine
groups to generate catalytically active amido-rhodium species51 is
an immediate goal of these studies.
21 P. Braunstein and F. Naud, Angew. Chem., Int. Ed., 2001, 40, 680–699.
22 P. Espinet and K. Soulantica, Coord. Chem. Rev., 1999, 193–195, 499–
556.
23 P. Braunstein, J. Organomet. Chem., 2004, 689, 3953–3967.
24 J. Andrieu, J.-M. Camus, P. Richard, R. Poli, L. Gonsalvi, F. Vizza and
M. Peruzzini, Eur. J. Inorg. Chem., 2006, 51–61.
25 A. Bader and E. Lindner, Coord. Chem. Rev., 1991, 108, 27–110.
26 H. Yang, M. Alvarez-Gressier, N. Lugan and R. Mathieu,
Organometallics, 1997, 16, 1401–1409.
27 F. Speiser, P. Braunstein and L. Saussine, Organometallics, 2004, 23,
2625–2632.
28 M. Bassetti, Eur. J. Inorg. Chem., 2006, 4473–4482.
29 H. Yang, N. Lugan and R. Mathieu, Organometallics, 1997, 16, 2089–
2095.
30 R. Ferna´ndez-Gala´n, F. A. Jalo´n, B. R. Manzano and J. Rodr´ıguez-de
la Fuente, Organometallics, 1997, 16, 3758–3768.
31 M. P. Anderson, A. L. Casalnuovo, B. J. Johnson, B. M. Mattson,
A. M. Mueting and L. H. Pignolet, Inorg. Chem., 1988, 27, 1649–
1658.
Acknowledgements
32 J. C. Jeffrey, T. B. Rauchfuss and P. A. Tucker, Inorg. Chem., 1980, 19,
3306–3315.
We thank the Natural Sciences and Engineering Research Council
of Canada (NSERC) and the University of Alberta for financial
support for this research and NSERC for funding the Bruker
PLATFORM/SMART 1000 CCD diffractometer, the Bruker
D8/APEX II CCD diffractometer and the Nicolet Avatar IR
spectrometer. We thank the Department’s Analytical and Instru-
mentation Laboratory for performing elemental analyses as well
as the NMR Spectroscopy Laboratory and Mr Rahul Samant for
assistance with variable temperature NMR spectroscopic analyses
and helpful discussions. We thank Dr Guy Bernard for carrying
out 31P NMR spectral simulations.
33 I. Bertini, P. Dapporto, G. Fallani and L. Sacconi, Inorg. Chem., 1971,
10, 1703–1707.
34 A. Del Zotto, G. Nardin and P. Rigo, J. Chem. Soc., Dalton Trans.,
1995, 3343–3351.
35 K. V. Baker, J. M. Brown, N. A. Cooley, G. D. Hughes and R. J. Taylor,
J. Organomet. Chem., 1989, 370, 397–406.
36 K. Tani, M. Yabuta, S. Nakamura and T. Yamagata, J. Chem. Soc.,
Dalton Trans., 1993, 2781–2789.
37 M. Habib, H. Trujillo, C. A. Alexander and B. N. Storhoff, Inorg.
Chem., 1985, 24, 2344–2349.
38 J. P. Farr, F. E. Wood and A. L. Balch, Inorg. Chem., 1983, 22, 3387–
3393.
39 B. R. Aluri, M. K. Kindermann, P. G. Jones, I. Dix and J. Heinicke,
Inorg. Chem., 2008, 47, 6900–6912.
40 M. T. Whited, E. Rivard and J. C. Peters, Chem. Commun., 2006, 1613–
1615.
References
41 M. D. Fryzuk and P. A. MacNeil, J. Am. Chem. Soc., 1981, 103, 3592–
3593.
42 D. Soulivong, C. Wieser, M. Marcellin, D. Matt, A. Harriman and L.
Toupet, J. Chem. Soc., Dalton Trans., 1997, 2257–2262.
43 J. N. L. Dennett, M. Bierenstiel, M. J. Ferguson, R. McDonald and M.
Cowie, Inorg. Chem., 2006, 45, 3705–3717.
1 A. M. Sargeson, Pure Appl. Chem., 1984, 56, 1603–1619.
2 S. J. Archibald, Annu. Rep. Prog. Chem., Sect. A, 2007, 103, 264–286.
3 F. T. Edelmann, Angew. Chem., Int. Ed., 2001, 40, 1656–1660.
4 M. Albrecht and G. van Koten, Angew. Chem., Int. Ed., 2001, 40,
3750–3781.
44 N. D. Jones, P. Meessen, M. B. Smith, U. Losehand, S. J. Rettig, B. O.
Patrick and B. R. James, Can. J. Chem., 2002, 80, 1600–1606.
45 N. D. Jones and B. R. James, Adv. Synth. Catal., 2002, 344, 1126–1134.
46 S. J. L. Foo, N. D. Jones, B. O. Patrick and B. R. James, Chem. Commun.,
2003, 988–989.
5 M. M. Taqui Khan and A. E. Martell, Inorg. Chem., 1974, 13, 2961–
2966.
6 C. A. Bessel, P. Aggarwal, A. C. Marschilok and K. J. Takeuchi, Chem.
Rev., 2001, 101, 1031–1066.
7 J.-C. Hierso, R. Smaliy, R. Amardeil and P. Meunier, Chem. Soc. Rev.,
2007, 36, 1754–1769.
8 J. E. Huheey, E. A. Keiter and R. L. Keiter, Inorganic Chemistry:
Principles of Structure and Reactivity, HarperCollins, New York, 4th
edn, 1993, pp. 522–524.
9 R. J. Puddephatt, Chem. Soc. Rev., 1983, 12, 99–127.
10 B. Chaudret, B. Delavaux and R. Poilblanc, Coord. Chem. Rev., 1988,
86, 191–243.
47 N. D. Jones, S. J. L. Foo, B. O. Patrick and B. R. James, Inorg. Chem.,
2004, 43, 4056–4063.
48 N. D. Jones, P. Meessen, U. Losehand, B. O. Patrick and B. R. James,
Inorg. Chem., 2005, 44, 3290–3298.
49 S. E. Clapham, A. Hadzovic and R. H. Morris, Coord. Chem. Rev.,
2004, 248, 2201–2237.
50 R. Noyori, Angew. Chem., Int. Ed., 2002, 41, 2008–2022.
51 P. Maire, T. Bu¨ttner, F. Breher, P. Le Floch and H. Gru¨tzmacher, Angew.
Chem., Int. Ed., 2005, 44, 6318–6323.
11 A. L. Balch, L. A. Fossett, R. R. Guimerans and M. M. Olmstead,
Organometallics, 1985, 4, 781–788.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 4213–4226 | 4225
©

View Article Online
52 G. Giordano and R. H. Crabtree, Inorg. Synth., 1979, 19, 218–220.
53 J. A. McCleverty, G. Wilkinson, L. G. Lipson, M. L. Maddox and H. D.
Kaesz, Inorg. Synth., 1990, 28, 84–86.
54 H. P. Fritz, I. R. Gordon, K. E. Schwarzhans and L. M. Venanzi,
J. Chem. Soc., 1965, 5210–5216.
55 H. Gilman and I. Banner, J. Am. Chem. Soc., 1940, 62, 344–345.
56 Spectral simulations were carried out using SpinWorks v. 2.5.5: K.
Marat, SpinWorks, University of Manitoba, Winnipeg, MB, Canada,
2006.
57 Programs for diffractometer operation, unit cell indexing, data collec-
tion, data reduction and absorption correction were those supplied by
Bruker.
62 G. M. Sheldrick, CELL_NOW version 2008/2, University of
Go¨ttingen, Germany, 2008.
63 G. M. Sheldrick, SAINT version 7.06A, Bruker AXS Inc., Madison,
WI, USA, 2003.
64 A. B. van Oort, P. H. M. Budzelaar, J. H. G. Frijns and A. G. Orpen,
J. Organomet. Chem., 1990, 396, 33–47.
65 A. R. Katritzky, W.-Q. Fan and K. Akutagawa, Tetrahedron, 1986, 42,
4027–4034.
66 C. J. Wu, S. H. Lee, S. T. Yu, S. J. Na, H. Yun and B. Y. Lee,
Organometallics, 2008, 27, 3907–3917.
67 C. J. Wu, S. H. Lee, H. Yun and B. Y. Lee, Organometallics, 2007, 26,
6685–6687.
58 G. M. Sheldrick, Acta Crystallogr., Sect. A, 2008, 64, 112–122.
59 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giacov-
azzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori and R. Spagna,
J. Appl. Crystallogr., 1999, 32, 115–119.
60 P. T. Beurskens, G. Beurskens, R. de Gelder, S. Garcia-Granda, R. O.
Gould, R. Israel and Jan M. M. Smits, The DIRDIF-99 program system,
Crystallography Laboratory, University of Nijmegen, The Netherlands,
1998.
68 T. B. Rauchfuss and D. M. Roundhill, J. Am. Chem. Soc., 1974, 96,
3098–3105.
69 A. D. Bain, Prog. Nucl. Magn. Reson. Spectrosc., 2003, 43, 63–103.
70 P. Suomalainen, S. Ja¨a¨skela¨inen, M. Haukka, R. H. Laitinen, J.
Pursiainen and T. A. Pakkanen, Eur. J. Inorg. Chem., 2000, 2607–2613.
71 N. D. Jones, PhD thesis, University of British Columbia, 2001.
72 F. Lorenzini, K. T. Hindle, S. J. Craythorne, A. R. Crozier, F. Marchetti,
C. J. Martin, P. C. Marr and A. C. Marr, Organometallics, 2006, 25,
3912–3919.
61 P. van der Sluis and A. L. Spek, Acta Crystallogr., Sect. A, 1990, 46,
194–201; A. L. Spek, PLATON—a multipurpose crystallographic tool,
Acta Crystallogr., Sect. A, 1990, 46, C34.
73 Y. V. Zefirov and P. M. Zorkii, Russ. Chem. Rev. (Engl. Transl.), 1989,
58, 421–440.
4226 | Dalton Trans., 2009, 4213–4226
This journal is
The Royal Society of Chemistry 2009
©