Serendipitous syntheses of Rh(H)2Cl(PRPh2)3 complexes, and their crystal structures, where R = Me, Cy (cyclohexyl)

Article abstract of DOI:10.1016/j.ica.2007.10.044

Reactions of RhCl(cod)(THP) (cod = 1,5-cyclooctadiene; THP = P(CH₂OH)₃) with PMePh₂ or PCyPh₂ (Cy = cyclohexyl) in acetone/MeOH solution under H₂ surprisingly form the complexes cis, mer-Rh(H)₂

Full text of DOI:10.1016/j.ica.2007.10.044

Available online at www.sciencedirect.com
Inorganica Chimica Acta 361 (2008) 2123–2130
www.elsevier.com/locate/ica
Serendipitous syntheses of Rh(H)2Cl(PRPh2)3 complexes,
and their crystal structures, where R = Me, Cy (cyclohexyl)
*
Fabio Lorenzini, Brian O. Patrick, Brian R. James
Department of Chemistry, The University of British Columbia, Vancouver, British Columbia, Canada V6T 1Z1
Received 17 September 2007; received in revised form 27 October 2007; accepted 28 October 2007
Available online 6 November 2007
Abstract
Reactions of RhCl(cod)(THP) (cod = 1,5-cyclooctadiene; THP = P(CH2OH)3) with PMePh2 or PCyPh2 (Cy = cyclohexyl) in
acetone/MeOH solution under H2 surprisingly form the complexes cis,mer-Rh(H)2Cl(PRPh2)3 (R = Me or Cy); both complexes are
characterized by crystallography (the first structures in which the hydride ligands of such dihydrido-chloro-trisphosphine complexes have
1
been located), and by detailed H and 31P NMR spectroscopy. The key role of the THP in the observed chemistry is discussed.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Rhodium; Hydride complexes; Phosphine complexes; Crystallography; NMR spectroscopy
1. Introduction
of the latest findings are available in three recent volumes
[6].
The activation of molecular hydrogen by transition
metal complexes after seven decades of study remains an
extremely active research area, particularly in aspects of cat-
alytic asymmetric hydrogenation. Key advances in terms of
application in organic syntheses developed from studies in
the 1960s when the groups of Halpern [1] and Wilkinson
[2] reported, respectively, on the use of Ru and Rh species
in solution for catalytic hydrogenation of olefins. Of histor-
ical note, the first success for olefin hydrogenation catalysts
was with chlororuthenate(II) species within an aqueous sys-
tem reported in 1961 [1], but the more well-known and more
widely applicable system was that of (chloro)tris(triphenyl-
phosphine)rhodium(I) in non-aqueous media reported in
1965 [2]. The RhCl(PPh3)3 complex is invariably and cor-
rectly called the ‘‘Wilkinson catalyst’’, although its activity
for the olefin hydrogenation was independently discovered
simultaneously by Coffey [3]. The complex was also made
independently (again reported in 1965) by three other
groups [4]. The historical aspects of homogeneous hydroge-
nation can be traced through a 1973 text [5], while reviews
There are now thousands of publications dealing with
Rh–phosphine complexes (including recent ones) that still
discuss mechanistic details of olefin hydrogenations cata-
lyzed in solution by RhCl(PPh3)3 or other chloro-tertiary
phosphine analogues [7]. The essentials of the mechanism
seem clear, although the proposed species directly respon-
sible for the hydrogenation are not detected: these are
usually RhClP2S, Rh(H)2ClP2S, Rh(H)2Cl(olefin)P2, and
RhH(alkyl)ClP2, where P = PPh3 and S = solvent or coor-
dination vacancy, whereas the precursor RhClP3 and
Rh(H)2ClP3 species are isolable [8]. During our recent stud-
ies on the isolation of water-soluble RhI–THP complexes
and investigation of their chemistry (THP = tris(hydroxy-
methyl)phosphine, P(CH2OH)3) [9], we accidentally pre-
pared the complexes cis,mer-Rh(H)2Cl(PRPh2)3, where
R = Me (complex 1) or Cy (2). We were surprised to find
that there was only one other crystallographically charac-
terized RhIII-dihydride of this type, the PPh3 analogue, in
which the hydride ligands were not located [10], and that
there were limitations in the characterization of other iso-
lated Rh(H)2Cl(PR3)3 complexes, where R3 = EtPh2 [11],
tBuMePh [12], and (DBH)3 (DBH = 5-phenyl-5H-dibenzo-
phosphole) [13]. A quite recent paper has reported 1H
*
Corresponding author.
E-mail address: brj@chem.ubc.ca (B.R. James).
0020-1693/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2007.10.044

2124
F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
NMR data (using para-hydrogen-induced polarization –
the PHIP technique) for the hydride ligand of Rh(H)2-
Cl(PR3)3 species formed in situ, where R3 = Me3, Me2Ph,
22.7). 13C{1H} NMR for a mixture of 1 and 3 – see below
(CD2Cl2): d 16.14 (CH3), 128.03–133.77 (C6H5). Spectral
integrations for 1 and 3 are ‘normalized’, assuming the
absence of a mixture.
n
MePh2, Et3, Et2Ph, EtPh2 and Bu3 [14]. Our paper here
describes the serendipitously synthesized complexes 1 and
2, and their characterization, including X-ray structures
2.2.1. RhCl(PMePh2)3 (3)
1
with located hydrides, and detailed, conventional H, and
1H NMR (CD2Cl2, see Fig. 1 for labeling): d 1.36 (br s,
3H, CH3Pc), 1.68 (dd, 6H, J(HPd) @ J(HRh) = 2.4,
CH3Pd), 7.51 (m, 30H, C6H5). 31P{1H} NMR (CD2Cl2):
d 18.35 (dd, 2Pd, J(PdRh) = 139.6, J(PdPc) = 42.6 Hz),
33.72 (dt, 1Pc, J(PcRh) = 186.1, J(PcPd) = 42.6).
31P{1H} NMR data.
2. Experimental
2.1. General considerations
2.3. Synthesis of cis,mer-Rh(H)2Cl(PCyPh2)3 (2)
The precursor complex RhCl(cod)(THP) (cod = 1,5-
cyclooctadiene) was synthesized by our recently reported
method [9]; the phosphines were used as received from
Strem Chemicals, and the reactions with the Rh precursor
were carried out under Ar or H2, using standard SCHLENK
techniques, or in a J-Young NMR-tube. MeOH was dried
over Mg–I2 and distilled under N2, and acetone was dried
over K2CO3 and distilled under N2. 31P{1H}, 13C{1H}, 1H,
1H{31P} and 2D NMR spectra were measured in 1:1
CD3OD/acetone-d6 or CD2Cl2 solutions at room tempera-
ture (ꢀ300 K), on a Bruker AV400 spectrometer; the deu-
terated solvents were used as received from Cambridge
Isotope Laboratory. A residual deuterated solvent proton
(relative to external SiMe4) and external 85% aq H3PO4
were used as references (s = singlet, d = doublet, t = triplet
m = multiplet; J values given in Hz). When necessary,
atom assignments were made by means of 31P–1H (HMQC)
NMR correlation spectroscopy. Elemental analyses were
performed on a Carlo Erba 1108 analyzer.
The synthesis was exactly as described for complex 1,
but using PCyPh2 (10.4 mg, 38.0 mmol) and 6.6 mg,
17.8 mmol of RhCl(cod)(THP), and again yellow, X-ray
quality crystals of 2 Æ MeOH were deposited from the solu-
tion over a 12 h period; these were collected, washed with
Et2O and dried as for 1 (9.5 mg; yield 79% on PCyPh2).
Anal. Calc. for C55H69OClP3Rh: C, 67.59; H, 7.12. Found:
C, 67.93; H, 6.91%. MS: 908 ([MꢁClꢁ2H]+), 675
1
([Mꢁ2HꢁPCyPh2]+). H NMR (CD2Cl2; labeling corre-
sponds to that in Fig. 1): d ꢁ17.81 (m, 1H, RhH2 cis to
P atoms, J(H2Rh) = 22.3, J(H2H1) = 3.5, J(H2Pa) = 13.8,
J(H2Pb) = 9.0), ꢁ9.03 (dddt, 1H, RhH1 trans to Pb,
J(H1Rh) = 12.3, J(H1H2) = 3.5, J(H1Pa) @ 12.3, J(H1Pb) =
154.7), 0.22–2.64 (m, 33H, C6H11), 6.45–7.78 (m, 30H,
C6H5). 31P{1H} NMR (CD2Cl2): d 32.43 (m, 1Pb), 46.80
(br d, 2Pa, J(PaRh) @ 110.9). 13C{1H} NMR (CD2Cl2): d
26.71–29.86 (C6H10), 127.11–135.38 (C6H5).
2.4. Crystallographic analyses of 1 and 2
2.2. Synthesis of cis,mer-Rh(H)2Cl(PMePh2)3 (1)
X-ray data for the Rh(H)2Cl(PRPh2)3 complexes
(R = Me, 1; R = Cy, 2) were collected at 173 ( 0.1) K on
a Bruker X8 APEX diffractometer using graphite-mono-
Addition of PMePh2 (11 lL, 57.9 mmol) in acetone-d6
(0.4 mL) to a yellow CD3OD solution (0.4 mL) of RhCl-
(cod)(THP) (10 mg, 27.0 mmol) at room temperature
under Ar in a J-Young NMR-tube results in immediate
formation of a brown solution. The Ar is then removed
by evacuation, and the tube filled with H2 and shaken, this
resulting in a yellow-brown solution. Over 12 h, X-ray
quality, yellow crystals of 1 deposit from the solution; these
were filtered off, washed with 3 · 2 mL of Et2O and then
dried under vacuum overnight (10.9 mg; yield 76 % on
PMePh2). A satisfactory elemental analysis for a crystal
of 1 could not be obtained. MS for C39H41ClP3Rh: 703
˚
chromated Mo Ka radiation (k = 0.71073 A) to maximum
2h values of 56.3ꢁ (for 1) and 56.0ꢁ (for 2), in a series of /
and x scans in 0.50ꢁ oscillations with 4.0 and 10.0 s expo-
sures, respectively; the crystal-to-detector distance was
36.00 mm for both complexes. Of 29,629 and 27,487 reflec-
tions collected for 1 and 2, respectively, 8,297 and 5,982
were unique (with corresponding Rint values of 0.032 and
0.045), equivalent reflections being merged. Data were col-
lected and integrated using the Bruker SAINT software pack-
age [15], and were corrected for absorption effects using the
multi-scan technique SADABS [16], with respective minimum
and maximum transmission coefficients of 0.740 and 0.895
(for 1), and 0.835 and 0.973 (for 2). Data were corrected for
Lorentz and polarization effects, and the structures were
solved by direct methods [17]. Selected crystallographic
data for 1 and 2 are shown in Table 1, and more details
are provided in the Supporting Information. For complex
1, all non-hydrogen atoms were refined anisotropically,
while all H-atoms were placed in calculated positions,
except H1 and H2, the metal hydrides (see Figs. 1 and 2),
1
([MꢁClꢁ2H]+), 503 ([MꢁClꢁ2HꢁPMePh2]+). H NMR
(CD2Cl2, see Fig. 1 for labeling): d ꢁ17.94 (m, 1H, RhH2
cis to
P
atoms, J(H2Rh) = 22.3, J(H2H1) = 2.0,
J(H2Pa) = 13.7, J(H2Pb) = 9.2), ꢁ9.55 (dddt, 1H, RhH1
trans to Pb, J(H1Rh) = 12.0, J(H1H2) = 2.0, J(H1Pa) =
12.0, J(H1Pb) = 163.4), 1.35 (br s, 3H, CH3Pb), 1.78 (dd,
6H, J(HPa) @ J(HRh) = 2.9, CH3Pa), 7.17 (m, 10H,
C6H5Pb), 7.35 (m, 10H, C6H5Pa), 7.67 (m, 10H, C6H5Pa).
31P{1H} NMR (CD2Cl2): d 6.01 (dt, 1Pb, J(PbRh) = 89.4,
J(PbPa) = 22.7), 24.94 (dd, 2Pa, J(PaRh) = 110.9, J(PaPb) =

F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
2125
PaMePh2
Ph2MePb
Cl
H2
H1
PaMePh2
PcMePh2
PdMePh2
Ph2MePd
Cl
- H2
H2
Rh
Rh
dd, δ 24.94, Pa,
J(PaRh) = 110.9, J(PaPb) = 22.7 Hz
dd, δ 18.35, Pd,
J(PdRh) = 139.57, J(PdPc) = 42.6 Hz
dt, δ 33.72, Pc,
J(PcRh) = 186.1, J(PcPd) = 42.6 Hz
dt, δ 6.01, Pb,
J(PbRh) = 89.4, J(PbPa) = 22.7 Hz
Fig. 1. 31P{1H} NMR spectrum of a CD2Cl2 solution of cis,mer-Rh(H)2Cl(PMePh2)3 (1), showing the resulting mixture of 1 and RhCl(PMePh2)3 (3).
*Impurity, probably OPMePh2.
Table 1
Crystal data and structure refinements for cis,mer-Rh(H)2Cl(PMePh2)3 (1)
and cis,mer-Rh(H)2Cl(PCyPh2)3 (2)
Compound
1
2 Æ CH3OH
Empirical formula
Formula weight
Crystal system
Space group
C
740.99
triclinic
39H41P3RhCl
C55H69OClP3Rh
945.37
monoclinic
P21=m (#13)
ꢀ
P1 (#2)
Crystal size (mm3)
0.15 · 0.35 · 0.35 0.05 · 0.15 · 0.30
˚
a (A)
9.9056(5)
11.8266(6)
16.5794(9)
103.006(2)
91.063(3)
112.870(2)
1731.47(16)
2
1.421
7.36
29629
8297 [0.032]
408
10.0140(14)
22.598(3)
11.2538(16)
90.0
108.474(7)
90.0
2415.5(6)
2
1.342
5.47
27487
5982 [0.045]
266
0.058, 0.102
1.03
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
V (A )
Z value
Dcalc (g cmꢁ3
)
l (cmꢁ1
No. of reflections measured
Unique [Rint
)
]
No. parameters
b
R1a, wR2
0.036, 0.067
1.05
0.51/ꢁ0.40
Goodness-of-fit indicator
Max. diffraction peak/hole (e A
Fig. 2. Structure of cis,mer-Rh(H)2Cl(PMePh2)3 (1), with 50% probability
ꢁ3
˚
)
0.90/ꢁ1.38
˚
ellipsoids. Selected distances (A) and angles (ꢁ): Rh(1)–P(1), 2.2850(5);
a
Rh(1)–P(2), 2.3861(5); Rh(1)–P(3), 2.3101(5); Rh(1)–Cl(1), 2.5022(5);
Rh(1)–H(1), 1.48(2); Rh(1)–H(2), 1.45(2); P(1)–Rh(1)–P(2), 98.246(19);
P(1)–Rh(1)–P(3), 160.456(19); P(1)–Rh(1)–Cl(1), 95.156(18); P(1)–Rh(1)–
H(1), 79.5(9); P(1)–Rh(1)–H(2), 81.9(8); P(2)–Rh(1)–P(3), 99.507(18);
P(2)–Rh(1)–Cl(1), 91.124(19); P(2)–Rh(1)–H(1), 177.4(9); P(2)–Rh(1)–
H(2), 95.5(8); P(3)–Rh(1)–Cl(1), 92.609(19); P(3)–Rh(1)–H(1), 82.9(9);
P(3)–Rh(1)–H(2), 88.3(8); Cl(1)–Rh(1)–H(1), 87.8(9); Cl(1)–Rh(1)–H(2),
173.0(8); H(1)–Rh(1)–H(2), 85.4(12).
R1 = Rj(Fo) ꢁ (Fc)j/R(Fo).
2
2 1=2
b
wR2 ¼ ½RwðF 2o ꢁ F c2Þ =RwðF 2oÞ ꢂ
.
which were found in a difference map and refined isotrop-
ically. Complex 2 crystallizes with one half-molecule resid-
ing on a mirror plane perpendicular to the b-axis. All the
substituents on P2 (see Fig. 5, below) are disordered, with
two of the three positions sharing both a phenyl and cyclo-
hexyl ring, while the third position has disordered phenyls
across the mirror plane; all C-atoms of the disordered frag-
ments were refined isotropically, while all other non-hydro-
gen atoms were refined anisotropically. The metal hydrides
H1m and H2m (corresponding to H1 and H2 of complex 1,
respectively) were located in difference maps and refined
isotropically; other H-atoms were placed in calculated posi-
tions. Finally, the MeOH molecule that crystallizes in the
asymmetric unit of 2 is also disordered about the mirror
plane.

2126
F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
3. Results and discussion
3.1. The PMePh2 system
(3), with a 1:3 equilibrium ratio of ꢀ1.7 within the NMR
tube; addition of H2 at 1 atm completely regenerates 1.
The assignments of the key 31P{1H} and high-field hydride
1H resonances (for 1) are given in Section 2 and are reason-
ably straightforward. The Pa-atom gives a doublet of dou-
blets due to coupling to Pb and Rh, and Pb gives rise to a
doublet of triplets due to coupling to two Pa atoms and
Rh (Fig. 1). The H1-hydride gives a doublet of doublets
of doublets of triplets due to coupling to Rh, the
H2-hydride, and the two Pa and Pb atoms; and the
H2-hydride generates a multiplet for a second-order spec-
trum that is well simulated using the illustrated J values
for coupling to Rh, H1, Pa and Pb (Fig. 3). The assignments
of the H1 and H2 resonances of 1, and the respective Me
resonances to 1 and 3, were obtained via use of a 2D
HMQC 31P{1H}/1H NMR spectrum (Fig. 4). An
1H{31P} NMR spectrum confirms assignment of all the
1H NMR resonances, including the Me proton resonances
which appear as a doublet of doublets for Pa–Me protons
(coupled to Pa and Rh) and a broad singlet presumably
due to a rapidly exchanging Pb ligand [19] (Fig. 4). To
the best of our knowledge, this is the first comprehensive
NMR characterization of a dihydride of general formula
Rh(H)2Cl(PR3)3 (R substituents are the same or different)
by conventional NMR methods, and the high-field hydride
data determined for 1 are in excellent agreement with liter-
ature data obtained in acetone-d6 using the PHIP technique
[14]. For RhCl(PMePh2)3 (3), the 31P{1H} NMR spectrum
(Fig. 1) shows the expected doublet of triplets (for Pc) and a
doublet of doublets for the Pd atoms; the protons of the Me
groups are also assigned (Fig. 4). The various coupling con-
stants for cis and trans-disposed P-atoms at Rh, and cis-
hydrides (cis and trans to P-atoms at Rh) are consistent
with those reported for analogous cis,mer-Rh(H)2Cl(PR3)3
complexes, where the R substituents may be the same or
different [7,14,18–20].
We have reported recently [9a] on slow reactions of
RhCl(cod)(THP) with 2 mol equivalents of PMePh2 or
PCyPh2 in alcohol or acetone solution under Ar at room
temperature. After a 12 h period, the final product is a mix-
ture of isomers formed according to Scheme 1; the reac-
tions involve a THP-promoted P–C6H5 bond cleavage of
a PRPh2 ligand with co-production of benzene, the
required proton deriving from a THP-hydroxy group. In
this current work, an atmosphere of H2 was added to the
mixture in the same media essentially immediately after
mixing the reagents, and now the isolated products are cis,-
mer-Rh(H)2Cl(PRPh2)3, where R = Me (complex 1), or Cy
(2). The complexes 1 and 2 could be formed by oxidative
addition of H2 to RhCl(PRPh2)3, which is a well estab-
lished reaction and typically a reversible process within
other reported RhClP3 systems, where P is a general,
monodentate tertiary phosphine [11,14,18]. However,
NMR evidence (see below) shows the presence of cis-
Rh(H)2Cl(PRPh2)2(THP) species that would be formed
by oxidative addition of H2 to RhCl(PRPh2)2(THP) (see
Scheme 1) and, more plausibly, 1 and 2 could be formed
by displacement of the THP of cis-Rh(H)2Cl(PRPh2)2-
(THP) by PRPh2; in essence, the THP promotes the oxida-
tive addition step.
Immediately after the phosphine and RhCl(cod)(THP)
are mixed in the acetone–MeOH solvent under Ar, the
in situ 31P{1H} NMR spectrum of the brown solution
shows no signal for the Rh-cod precursor (dP 17.7, d,
J(RhP) = 145 Hz), a small amount of free PMePh2 (dP
ꢁ28.0, s) corresponding to the slight excess for a 2:1 phos-
phine:Rh reaction, and a complex mixture of products with
signals between dP ꢁ0.28 and 18.57; of note, no free THP
(dP ꢁ25.0, s) is seen. If the solution is not exposed to H2,
this spectrum changes slowly with time to generate a
roughly 1:1 mixture of the cis- and trans-isomer products
shown in Scheme 1 [9a]; the 31P{1H} spectrum of this final
mixture shows the signals for the three inequivalent
P-atoms of each isomer, each P-atom giving a doublet of
doublets of doublets, and all the resonances have been
assigned [9a]. With the immediate H2 treatment, however,
Many groups have studied such catalytically important
RhIII-dihydrides formed usually by reversible oxidative
addition of H2 to a RhI precursor. However, because of
their short lifetimes and low concentrations in solutions
[7,11–14,18–20], including decomposition in chlorinated
solvents to chloro species [21], NMR data have not gener-
ally been definitive until the utilization of the much more
1
sensitive PHIP technique (for 31P, H, and 103Rh NMR),
1
the color instantly becomes lighter and the in situ H and
which has been applied for the in situ formed R3 = Ph3,
31P{1H} NMR spectra show the major product to be cis,-
mer-Rh(H)2Cl(PMePh2)3 (1). Overnight, crystals of 1 are
deposited and X-ray analysis corresponds to this structure
(Fig. 2). The NMR data of isolated 1 in CD2Cl2 under Ar
show the partial loss of H2 to give some RhCl(PMePh2)3
Me3, Me2Ph, MePh2, Et3, Et2Ph, EtPh2 and Bu3 species
n
[14,20]. Of note, complex 1 in the solid state does not lose
H2, even under the vacuum conditions used for drying the
isolated crystals, but does so in solution under Ar but not
under H2.
Ph RP
OH
HO
PRPh
Cl
Ph2RP
P
Ph2RP
PhRP
Cl
PRPh
Cl
2 PRPh
- cod
2
2
2
Rh
Rh
+ PhH
RhCl(cod)(THP)
Rh
P
HO
OH
THP
O
O
Scheme 1. Synthesis of trans- and cis-RhCl(PRPh2)[P,P-Ph(R)POCH2P(CH2OH)2] (R = Me, Cy); trans and cis refer to the disposition of the P-atoms
with Ph substitutents.

F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
2127
H1: δ − 9.55 (dddt, 1H, RhH1, J(H1Rh) = 12.0, J(H1H2) = 2.0 J(H1Pa) = 12.0, J(H1Pb) = 163.4 Hz
H2: δ − 17.94 (m, 1H, RhH2, J(H2Rh) = 22.3, J(H2H1) = 2.0 J(H2Pa) = 13.7, J(H2Pb) = 9.2 Hz
PaMePh2
Ph2MePb
Cl
H2
H1
PaMePh2
Rh
H1
H2
Exptl. 1H NMR
Sim. 1H NMR
Exptl. 1H NMR
Sim. 1H NMR
H1
H2
Simulated 1H NMR
Fig. 3. Experimental and simulated 1H NMR spectrum, in the hydride region, of a CD2Cl2 solution of cis,mer-Rh(H)2Cl(PMePh2)3 (1).
m, δ 1.78 (CH3a)
1
m,
s, δ
1.35 (CH3b)
δ
1.68 (CH3d)
δ − 17.94, m, RhH2
3
δ − 9.55, dddt, RhH1
1
s, δ
1.36 (CH3c)
1
3
1
1
δ 6.01, dt, P
b
δ
18.35, dd, Pd
3
1
δ 24.94, dd, Pa
3
δ 33.72, dt, Pc
Fig. 4. HMQC 31P{1H}–1H NMR spectrum of a CD2Cl2 solution of cis,mer-Rh(H)2Cl(PMePh2)3 (1); enlargement of the methyl and hydride regions,
allowing for assignment of the CH3 resonances, and confirming assignment of the 31P NMR resonances, of 1 and RhCl(PMePh2)3 (3).
Except for the crystallographically characterized cis,-
mer-Rh(H)2Cl(PPh3)3 (the crystals being formed serendipi-
tously as one of seven products formed from reaction of
RhCl(PPh3)3 with catecholborane) [10], the only other iso-
lated analogues are (as mentioned in Section 1) the
the octahedral distortion that results from the bulky phos-
phines and small hydride ligands. The Rh–P bond lengths
˚
for P1 and P3 are ꢀ0.1 A shorter than the Rh–P length for
P2 trans to a hydride, and the Rh-hydride bond trans to P2
˚
is 0.03 A longer than that for the hydride trans to the
t
R3 = EtPh2 [11], BuMePh [12], and (DBH)3 complexes
chloride. Similar general features were found in the
X-ray structure of cis,mer-Rh(H)2Cl(PPh3)3, although
hydride ligands were not located for this complex
[10].
It should be noted that an aqueous solution of
RhCl(TPPTS)3 (TPPTS = P(m-C6H4SO3Na)3) reacts with
H2 to form not only cis,mer-Rh(H)2Cl(TPPTS)3, but also
a species with a cis, fac arrangement of hydrogen and phos-
phine ligands; however, this species may be cationic in
which the chloride has been replaced by an aquo ligand
[18c]. In organic solvents, the cis,mer-isomer is considered
[13]; however, the EtPh2 species was characterized only
by elemental analysis and IR data, only a tentative inter-
1
pretation of the H and 31P NMR data was given for the
tBuMePh species, and the high-field resonances for the
DBH species were not detected because of loss of H2 from
the complex in solution.
The solid state structure of 1 (Fig. 2) shows the typical
distorted octahedral coordination geometry for RhIII. The
two hydrides, the chloride and the P2 atom form an almost
perfect plane, while the P1–Rh–P3 angle of ꢀ160ꢁ reveals

2128
F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
to be the thermodynamically stable species, although the
phosphine trans to the hydride has been shown to revers-
ibly dissociate rapidly [19]. Relevant to this, the in situ
reaction of the Rh-cod precursor with PMePh2 under H2
described above does generate initially another dihydride
species as a trace by-product (<5%); however, this is again
[19]. The H1 hydride NMR resonance at dH ꢁ9.03 is again
a doublet of doublets of doublets of triplets with J values
similar to those seen in the PMePh2 complex, and the
multiplet for H2 at dH ꢁ 17.81 was completely resolved
by 1H{31P}, 1H{31Pa} and 1H{31Pb} NMR spectra of
CD2Cl2 solutions of the isolated, pure complex 2 (the shifts
and J values are similar to those seen for 1 – see Section 2).
The decoupled spectra also confirmed the assigned cou-
pling constants for the H1 resonance.
1
a cis,mer-isomer as evidenced by the high-field H NMR
spectrum, which shows an almost identical pattern to that
of complex 1, but shifted upfield by ꢀ1 ppm. Using the
same P-atom labeling of Fig. 1, there is: a doublet of dou-
blets of doublets of triplets at d ꢁ10.39 (RhH3 trans to Pb,
J(H3Rh) = 13.6, J(H3H4) @ 2.0, J(H3Pa) = 13.6, J(H3Pb) =
161.3 Hz), and a second-order multiplet resonance at d
ꢁ19.02 (RhH4 cis to Pa atoms, J(H4Rh) = 24.0, J(H4H3) @
2.0, J(H4Pa) = 13.8, J(H4Pb) = 8.0 Hz). The species could
be the THP-containing species cis-Rh(H)2Cl(PMePh2)2-
(THP), which is supported by some 2D NMR evidence
where correlations are observed between 1H resonances
in the range (d 4.45–3.86), typical of methylene protons
of THP coordinated at Rh [9], and a weak broad
31P{1H} resonance centered at d ꢀ 20.6, a signal that was
undetectable in the 1D 31P{1H} spectrum, but might refer
to a coordinated THP-phosphorus atom [9]. Some data
from the corresponding PCyPh2 system (see below) also
support such a formulation. Another possibility is that
the chloride (trans to H4) is displaced by solvent to form
cis,mer-[Rh(H)2(solv)(PMePh2)3]Cl (solv = CD3OD or
acetone-d6), but this is considered less likely since as noted
above the phosphine trans to a hydride is typically more
labile.
As in the PMePh2 system, the in situ 1H NMR spectrum
shows a minor dihydride product (initially ꢀ20%, but this
decreases as crystals of 2 are formed); this dihydride species
(a likely intermediate in formation of 2) is characterized by
two doublets of doublets of doublets of triplets in the
hydride region at dH ꢁ9.95 (RhH3 trans to Pb,
J(H3Rh) = 12.7, J(H3H4) = 4.0, J(H3Pc) = 12.7, J(H3Pd) =
147.0 Hz), and at d ꢁ 19.76 (RhH4 cis to Pa atoms,
J(H4Rh) = 35.9, J(H4H3) = 4.0, J(H4Pc) = 14.3, J(H4Pd) =
9.26 Hz). Unlike for the PMePh2 system, the 31P{1H}
NMR resonances for the minor product were now
detected: a doublet of doublets at dP 45.92 (2 equiv. trans
P-atoms), J(PtransRh) = 110.1, J(PtransPcis) = 19.3 Hz, and
a doublet of triplets at dP 35.38 (Pcis), J(PcisRh) = 98.2,
J(PcisPtrans) = 19.3 Hz, where Pcis implies cis to two trans-
phosphines. These data perhaps refer to cis-Rh(H)2-
Cl(PCyPh2)2(THP), with THP being cis to two trans
PCyPh2 ligands; the NMR data for the cis-phosphines
are very similar to those reported for fac-RhCl3(THP)3
(in CD3OD, dP = 35.5, J(PRh) = 107 Hz [22]; in D2O,
dP = 31.39, J(PRh) = 107 Hz [23]).
Reactions of RhCl(cod)(THP) with the more obvious
three mole equivalents of PMePh2 (or PCyPh2, see below)
under H2 to generate the title Rh(H)2Cl(phosphine)3 com-
plexes were attempted, but a much more complicated mix-
ture of products was formed.
The X-ray structure for complex 2 is shown in Fig. 5,
and crystal and refinement data are given in Table 1.
The distorted octahedral structure is analogous to that
of complex 1 with again the two hydrides, the chloride,
and the P-atom of one phosphine being planar, and
the distortion coming from the trans-phosphines, where
the P1–Rh–P1* angle is ꢀ152ꢁ; however, in contrast to
the structure of 1, these two phosphines are structurally
equivalent, and the overall structure is now a monoclinic,
3.2. The PCyPh2 system
The room temperature solution reaction of PCyPh2 with
RhCl(cod)(THP) (ꢀ2:1) under Ar was very similar to that
of the PMePh2 reaction: again immediately a slight excess
of free PCyPh2 was seen (dPꢁ3.03 s) and there was no free
THP, while the complicated mixture of products gave rise
to dP resonances in the range d 20.23 to 66.67 (correspond-
ing PCyPh2-containing products of those shown in Scheme
1 are again slowly formed) [9a]. Under H2, the in situ NMR
data correspond to formation of cis,mer-Rh(H)2Cl-
(PCyPh2)3 (2) as the major product. Cleaner spectra are
seen for isolated 2 dissolved in CD2Cl2. The 31P{1H} signal
for Pa (Fig. S1, see Fig. 1 for comparison) is seen as a broad
doublet at dP 45.39 (J(PaRh) @ 110.9 Hz), coupling to Pb
not being resolved, and likewise the Pb signal is a broad
symmetric multiplet at dP 31.12; both signals are
unchanged as the temperature was varied between 323
and 213 K, and J(PaPb) and J(PbRh) could not be resolved.
The broadness of the resonances probably results from
some exchange of the Pb phosphine trans to a hydride
ꢀ
P21=m system, versus the triclinic P1 system of 1. As for
˚
1, the two trans-PCyPh2 ligands are 0.08 A closer to the
metal than the PCyPh2 trans to the hydride, and the Rh–
H bond length for the hydride trans to the phosphine is
˚
0.08 A longer than the Rh–H bond for the hydride trans
to the chloride.
Surprisingly, and in contrast to 1, complex 2 does not
undergo reductive elimination of H2 when dissolved in
CH2Cl2 under Ar, as evidenced by NMR data. The differ-
ence in basicity between PCyPh2 and PMePh2 should be
small, while steric effects should be insignificant for binding
of the small H2 molecule. A more quantitative study of the
respective equilibria is needed to evaluate the factor(s)
involved. (The steric effects are reflected in the P1–Rh–P3
and P1–Rh–P1* angles for 1 and 2, respectively; the 8ꢁ
smaller angle in 2 results from interactions of the cyclo-
hexyl substituent of the trans-phosphines with that of the
mutually cis-phosphines.)

F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
2129
dence for formation of any hydrido species! [24]. An
important unanswered question in the syntheses is the fate
of the THP, which does not finish up as free THP; this is a
reactive phosphine and, as well the hydroxy-promoted
cleavage reaction shown in Scheme 1, THP is a strong
reducing agent (including the reduction of water to H2),
[9b] can readily form phosphonium salts [9a], and can
decompose with loss of formaldehyde to generate
HP(CH2OH)2 [9]. Structures of metal-THP complexes
typically reveal extensive hydrogen-bonding networks
involving oxygen and hydrogen atoms of the hydroxy sub-
stituent, and the hydrogens of the methylene groups [9]; it
is not obvious how this may be a possible factor, but the-
oretical studies have revealed interactions of the hydrido
ligands of cis,mer-Rh(H)2Cl(PMe3)3 with protons [7c].
Appendix A. Supplementary material
CCDC 661134 and 661135 contain the supplementary
crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif. Supplementary data associated with this article
can be found, in the online version, at doi:10.1016/
j.ica.2007.10.044.
Fig. 5. Structure of cis,mer-Rh(H)2Cl(PCyPh2)3 (2), with 50% probability
ellipsoids. Selected distances (A) and angles (ꢁ): Rh(1)–P(1) = Rh(1)–P(1)*,
˚
2.3186(8); Rh(1)–P(2), 2.4021(10); Rh(1)–Cl(1), 2.4938(10); Rh(1)–H(1),
1.49(4); Rh(1)–H(2), 1.41(4); P(1)–Rh(1)–P(2) = P(1)*–Rh(1)–P(2),
103.971(18); P(1)–Rh(1)–P(1)*, 152.03(4); P(1)–Rh(1)–Cl(1) = P(1)*–
Rh(1)–Cl(1), 89.475(19); P(1)–Rh(1)–H(1) = P(1)*–Rh(1)–H(1), 76.16(6);
P(1)–Rh(1)–H(2) = P(1)*–Rh(1)–H(2), 90.9(4); P(2)–Rh(1)–Cl(1), 94.92(4);
P(2)–Rh(1)–H(1), 169.1(15); P(2)–Rh(1)–H(2), 83.4(16); Cl(1)–Rh(1)–H(1),
96.0(15); Cl(1)–Rh(1)–H(2), 178.3(16); H(1)–Rh(1)–H(2), 86(2). The well
defined Cy rings are the ones labeled C1,C2.. and C1*, C2*. . .
References
[1] J. Halpern, J.F. Harrod, B.R. James, J. Am. Chem. Soc. 83 (1961)
753.
3.3. Other tertiary phosphine systems
[2] (a) F.H. Jardine, J.A. Osborn, G. Wilkinson, J.F. Young, Chem. Ind.
(Lond.) (1965) 560;
The reactivity of RhCl(cod)(THP) with PEtPh2, P(p-
tol)3, P(p-F-C6H4)3 and PnPr3 in the presence of H2 was
also investigated using experimental conditions identical
to those described above for the PMePh2 and PCyPh2 sys-
tems. However, in situ 1H and 31P{1H} NMR spectra both
revealed a complicated mixture of products, and no pure
complexes could be isolated. Why the fortuitous, close to
80% yield syntheses of 1 and 2 were successful remains a
mystery. The Rh-precursor and the phosphines are both
soluble in acetone and in MeOH, and the use of the 1:1
solvent mixture just happened to deposit crystals of the
dihydrides! Worth noting is the unusual observation of
what appears to be the presence of a two-phase solvent sys-
tem where the crystals of 1 and 2 begin to form; the pres-
ence of the water-soluble THP (a white solid in the free
state) almost certainly plays a role, since the use of a more
standard [RhCl(diene)]2 under identical conditions does
not precipitate the solid hydride complexes. Indeed,
NMR evidence supports formation of 1 and 2 via the cis-
Rh(H)2Cl(PRPh2)2(THP) intermediate. A perhaps related
phenomenon is the reproducible and remarkable observa-
tion of the immediate precipitation of crystals of
RhCl(P–N)(THP) from an acetone solution of the complex
(P–N is the P, N-chelated ligand o-PPh2(C6H4NMe2), and
THP is trans to the N-donor), when the solution is sub-
jected to an atmosphere of H2, even though there is no evi-
(b) J.F. Young, J.A. Osborn, F.H. Jardine, G. Wilkinson, Chem.
Commun. (1965) 131.
[3] R.S. Coffey, ICI, Brit. Pat. 1,121,642 (filed 1965).
[4] (a) M.A. Bennett, P.A. Longstaff, Chem. Ind. (Lond.) (1965) 846;
(b) L. Vaska, R.E. Rhodes, J. Am. Chem. Soc. 87 (1965) 4970;
(c) K.C. Dewhirst, unpublished results quoted in W. Keim,
J. Organomet. Chem. 14 (1968) 179.
[5] B.R. James, Homogeneous Hydrogenation, Wiley, New York, USA,
1973.
[6] J.G. de Vries, C.J. Elsevier (Eds.), Handbook of Homogeneous
Hydrogenation, vol. 1–3, Wiley-VCH, Weinheim, Germany, 2007.
[7] For example (a) L.A. Bengtsson, B.T. Heaton, J.A. Iggo, C. Jacob,
G.L. Monks, J. Ratnam, A.K. Smith, J. Chem. Soc., Dalton Trans.
(1994) 1857;
(b) S.B. Duckett, C.L. Newell, R. Eisenberg, J. Am. Chem. Soc. 116
(1994) 10548;
´
´
´
(c) G. Kovacs, G. Schubert, F. Joo, I. Papai, Organometallics 24
(2005) 3059.
[8] L.A. Oro, D. Carmona, in: J.G. de Vries, C.J. Elsevier (Eds.),
Handbook of Homogeneous Hydrogenation, vol. 1, Wiley-VCH,
Weinheim, Germany, 2007 (Chapter 1), and refs. therein.
[9] (a) F. Lorenzini, B.O. Patrick, B.R. James, Inorg. Chem. 46 (2007)
8998;
(b) F. Lorenzini, B.O. Patrick, B.R. James, Dalton Trans. (2007)
3224.
[10] (a) K. Burgess, W.A. van der Donk, S.A. Westcott, T.B. Marder,
R.T. Baker, J.C. Calabrese, J. Am. Chem. Soc. 114 (1992) 9350;
(b) W. Clegg, M.R.J. Elsegood, N.C. Norman, N.L. Pickett, Acta
Crystallogr., Sect. E 58 (2002) m406.
[11] A. Sacco, R. Ugo, A. Moles, J. Chem. Soc. A (1966) 1670.

2130
F. Lorenzini et al. / Inorganica Chimica Acta 361 (2008) 2123–2130
´
´
´
´
[12] C. Masters, B.L. Shaw, J. Chem. Soc. A (1971) 3679.
[13] D.G. Holah, A.N. Hughes, B.C. Hui, Can. J. Chem. 53 (1975) 3669.
[14] A. Koch, J. Bargon, Inorg. Chem. 40 (2001) 533.
[15] SAINT. Version 7.03A. Bruker AXS Inc., Madison, Wisconsin, USA,
1997–2003.
[16] SADABS. Bruker Nonius Area Detector Scaling and Absorption
Correction – V2.10, Bruker AXS Inc., Madison, WI, USA, 2003.
[17] A. Altomare, M.C. Burla, M. Camalli, G.L. Cascarano, C. Giaco-
vazzo, A. Guagliardi, A.G.G. Moliterni, G. Polidori, R. Spagna,
J. Appl. Crystallogr. 32 (1999) 115.
(g) F. Joo, J. Kovacs, A.Cs. Benyei, L. Nadasdi, G. Laurenczy, Chem.
Eur. J. 7 (2001) 193.
[19] J.M. Brown, A.R. Lucy, J. Chem. Soc., Chem. Commun. (1984) 914,
and refs. therein.
[20] For example (a) S.B. Duckett, G.K. Barlow, M.G. Partridge,
R.J. Mawby, B.A. Messerie, Bull. Magn. Res. 18 (1996) 114, and
refs. therein;
(b) B.A. Messerle, C.J. Sleigh, M.G. Partridge, S.B. Duckett,
J. Chem. Soc., Dalton Trans. (1999) 1429.
[21] For example (a) M.L.H. Green, D.J. Jones, Adv. Inorg. Chem.
Radiochem. 7 (1965) 115;
[18] For example (a) C.A. Tolman, P.Z. Meakin, D.L. Lindner,
J.P. Jesson, J. Am. Chem. Soc. 96 (1974) 2762;
(b) J.P. Rourke, G. Stringer, P. Chow, R.J. Deeth, D.S. Yulfit,
J.A.K. Howard, T.B. Marder, Organometallics 21 (2002) 429;
(c) M.A. Garralda, R. Hernandez, L. Ibarlucea, E. Pinilla,
(b) R.S. Drago, J.G. Miller, M.A. Hoselton, R.D. Farris,
M.J. Desmond, J. Am. Chem. Soc. 105 (1983) 444;
´
(c) C. Larpent, H. Patin, J. Organomet. Chem. 335 (1987) C13;
(d) G.K. Anderson, R. Kumar, Inorg. Chim. Acta 146 (1988) 89;
M.R. Torres, M. Zarandona, Organometallics 26 (2007) 1031.
[22] F. Lorenzini, B.R. James, unpublished data.
´
(e) D. Rutherford, J.J.J. Juliette, C. Rocaboy, I.T. Horvath,
[23] K. Raghuraman, N. Pillarsetty, W.A. Volkert, C. Barnes, S. Jurisson,
K.V. Katti, J. Am. Chem. Soc. 124 (2002) 7276.
[24] F. Lorenzini, B.O. Patrick, B.R. James, Inorg. Chim. Acta, in press,
doi:10.1016/j.ica.2007.11.017.
J.A. Gladysz, Catal. Today 42 (1998) 381;
(f) B. Richter, A.L. Spek, G. van Koten, B.-J. Deelman, J. Am. Chem.
Soc. 122 (2000) 3945;