Thiobenzamidato and thiobenzoato complexes of ruthenium, osmium and iridium

Article abstract of DOI:10.1039/a607011f

Thiobenzamide reacted with the complexes [MCl₂(PPh₃)₃], [MH(X)(CO)(PPh₃)₃] (M = Ru or Os, X = H or Cl) and mer-[IrH₃(PPh₃)₃] in boiling toluene to afford the products [M{S(NH)CPh}₂(PPh₃)₂], [MX{S(NH)CPh}(CO)(PPh₃)₂] and [IrH₂{S(NH)CPh}(PPh₃)₂] respectively. The crystal structure of [Os{S(NH)CPh}₂(PPh₃)₂] confirms the expected bis(N,S-chelate) formulation and has established the presence of cis-phosphorus, cis-nitrogen and trans-sulfur donor atom pairs within a highly distorted octahedral co-ordination sphere. Thiobenzoic acid reacted with [MCl₂(PPh₃)₃] and [MH(X)(CO)(PPh₃)₃] under similar conditions to yield [M{S(O)CPh}₂(PPh₃)₂] and [M{S(O)CPh}₂(CO)(PPh₃)₂] respectively. The reactions of thiobenzoic acid with mer-[IrH₃(PPh₃)₃] under conditions of increasing severity give [IrH₂{S(O)CPh}(PPh₃)₃], [IrH{S(O)CPh}₂(PPh₃)₂] and [Ir{S(O)CPh}₃(PPh₃)₂] the last two of which on carbonylation afford [IrH{S(O)CPh}₂(CO)(PPh₃)₂] and [Ir{S(O)CPh}₃(CO)(PPh₃)₂] respectively. The crystal structure of [IrH{S(O)CPh}₂(PPh₃)₂] has established the presence of trans phosphorus and trans sulfur donor atom pairs, with the η¹- and η²-thiobenzoate ligands sharing a meridional set of three co-ordination sites. However, the NMR spectra of this complex show two high-field ¹H triplets and two ³¹P-{¹H} singlets. Possible explanations for this anomalous NMR behaviour, which is also found for the corresponding thioacetate, are advanced.

Full text of DOI:10.1039/a607011f

View Article Online / Journal Homepage / Table of Contents for this issue
Thiobenzamidato and thiobenzoato complexes of ruthenium, osmium
and iridium*
Stephen D. Robinson,a Arvind Sahajpala and Derek A. Tocher b
a
Department of Chemistry, Kings College London, Strand, London WC2R 2LS, UK
b Department of Chemistry, University College London, 20 Gordon Street,
London WC1H 0AJ, UK
Thiobenzamide reacted with the complexes [MCl2(PPh3)3], [MH(X)(CO)(PPh3)3] (M = Ru or Os, X = H or Cl) and
mer-[IrH3(PPh3)3] in boiling toluene to afford the products [M{S(NH)CPh}2(PPh3)2], [MX{S(NH)CPh}(CO)-
(PPh3)2] and [IrH2{S(NH)CPh}(PPh3)2] respectively. The crystal structure of [Os{S(NH)CPh}2(PPh3)2] confirms
the expected bis(N,S-chelate) formulation and has established the presence of cis-phosphorus, cis-nitrogen and
trans-sulfur donor atom pairs within a highly distorted octahedral co-ordination sphere. Thiobenzoic acid reacted
with [MCl2(PPh3)3] and [MH(X)(CO)(PPh3)3] under similar conditions to yield [M{S(O)CPh}2(PPh3)2] and
[M{S(O)CPh}2(CO)(PPh3)2] respectively. The reactions of thiobenzoic acid with mer-[IrH3(PPh3)3] under
conditions of increasing severity give [IrH2{S(O)CPh}(PPh3)3], [IrH{S(O)CPh}2(PPh3)2] and [Ir{S(O)CPh}3-
(PPh3)2] the last two of which on carbonylation afford [IrH{S(O)CPh}2(CO)(PPh3)2] and [Ir{S(O)CPh}3(CO)-
(PPh3)2] respectively. The crystal structure of [IrH{S(O)CPh}2(PPh3)2] has established the presence of trans
phosphorus and trans sulfur donor atom pairs, with the η1- and η2-thiobenzoate ligands sharing a meridional set
1
of three co-ordination sites. However, the NMR spectra of this complex show two high-field H triplets and two
31P-{1H} singlets. Possible explanations for this anomalous NMR behaviour, which is also found for the
corresponding thioacetate, are advanced.
In the course of work on heteroallylic complexes of the plat-
inum metals we have recently observed reactions between cer-
tain ruthenium triphenylphosphine complexes and organic
amides RC(O)NH2 (R = Ph, CF3, C2F5 or C6F5) leading to
formation of binuclear amidato-bridged products.1,2 This
behaviour contrasts sharply with that observed for amidines3
and carboxylic acids4,5 which react with the same ruthenium
precursors to generate more conventional mononuclear prod-
ucts containing chelating heteroallyl ligands. In order further to
examine the possibility of a link between the asymmetric nature
of the amidate ligands and their tendency to adopt a bridging
mode of bonding we have now investigated the reactions of the
asymmetric heteroallyls, thiobenzamide and thiobenzoic acid,
with a similar series of platinum metal precursors. Unfortu-
nately the products obtained during the course of this study are
all mononuclear and therefore lend no support to the hypo-
thesis that in our systems asymmetric heteroallyl ligands
have a greater tendency to adopt bridging co-ordination modes
than do their symmetric analogues. However the study did
afford an iridium() derivative [IrH{S(O)CPh}2(PPh3)2] which
shows NMR evidence compatible with the presence of a
temperature-dependent mixture of two isomeric or solvated
forms.
reduced pressure to leave oils. These were crystallised from
CH2Cl2–MeOH to afford the pure crystalline materials.
[Ru{S(NH)CPh}2(PPh3)2]ؒ0.5CH2Cl2. Dichlorotris(triphenyl-
phosphine)ruthenium (0.6 g, 0.62 mmol) and thiobenzamide
(0.34 g, 2.48 mmol) were heated under reflux in toluene (40
cm3) for ca. 6 h to give a dark red-brown solution. Standard
work-up gave dark orange-brown crystals (0.31 g, 63%), m.p.
170–172 ЊC (Found: C, 64.0; H, 4.95; N, 2.9; S, 7.25. Calc. for
C50H42N2P2RuS2ؒ0.5CH2Cl2: C, 64.5; H, 4.6; N, 3.0; S, 6.8%).
[RuH{S(NH)CPh}(CO)(PPh3)2]ؒ0.5CH2Cl2. Carbonyldihy-
dridotris(triphenylphosphine)ruthenium (0.8 g, 0.87 mmol) and
thiobenzamide (0.48 g, 3.5 mmol) were heated under reflux in
toluene (40 cm3) for ca. 30 min to give a dark red solution.
Standard work-up gave light brown microcrystals (0.4 g, 59%),
decomp. 175–177 ЊC (Found: C, 64.35; H, 4.95; N, 1.7. Calc. for
C44H37NOP2RuSؒ0.5CH2Cl2: C, 64.2; H, 4.6; N, 1.7%).
[RuCl{S(NH)CPh}(CO)(PPh3)2]ؒ0.5CH2Cl2.
Carbonyl-
chlorohydridotris(triphenylphosphine)ruthenium (0.6 g, 0.63
mmol) and thiobenzamide (0.32 g, 2.3 mmol) were heated
under reflux in toluene (40 cm3) for ca. 6 h to give a dark red-
brown solution. Standard work-up gave orange microcrystals
(0.28 g, 55%), decomp. 208–210 ЊC (Found: C, 62.15; H, 4.3; N,
1.45. Calc. for C44H36ClNOP2RuSؒ0.5CH2Cl2: C, 61.6; H, 4.3;
N, 1.6%).
Experimental
Thiobenzamide, thiobenzoic acid and thioacetic acid were
obtained from the Aldrich Chemical Co. and used as supplied.
Other reagents, experimental techniques and instrumentation
were as described in the preceding paper in this series.1 Spectro-
scopic data are given in Table 1.
[Os{S(NH)CPh}2(PPh3)2].
Dichlorotris(triphenylphos-
phine)osmium (0.5 g, 0.49 mmol) and thiobenzamide (0.28 g,
2.04 mmol) were heated under reflux in degassed toluene (40
cm3) for ca. 5 h to give a dark red-brown solution. Standard
work-up gave dark maroon crystals (0.26 g, 54%), m.p. 174–
176 ЊC (Found: C, 59.85; H, 4.05; N, 2.65. Calc. for C50H42N2-
OsP2S2: C, 60.85; H, 4.3; N, 2.85%).
Preparations
All products described were isolated and purified by a stand-
ard procedure. The crude reaction solutions, after cooling to
ambient temperature, were filtered and then concentrated under
[OsH{S(NH)CPh}(CO)(PPh3)2]ؒ0.5CH2Cl2. Carbonyldihy-
dridotris(triphenylphosphine)osmium (0.4 g, 0.38 mmol) and
thiobenzamide (0.23 g, 1.56 mmol) were heated under reflux in
toluene (40 cm3) for ca. 6 h to give a dark red-brown solution.
* Complexes of the platinum metals. Part 48.1
J. Chem. Soc., Dalton Trans., 1997, Pages 757–762
757

View Article Online
Standard work-up gave yellow microcrystals (0.19 g, 58%),
decomp. 178–180 ЊC (Found: C, 58.75; H, 4.4; N, 1.45. Calc.
for C44H37NOOsP2Sؒ0.5CH2Cl2: C, 58.0; H, 4.2; N, 1.5%).
ard work-up gave white microcrystals (0.26 g, 46%), m.p. 203–
205 ЊC (Found: C, 63.45; H, 4.45. Calc. for C61H52IrOP3Sؒ
0.5CH2Cl2: C, 63.65; H, 4.6%).
[OsCl{S(NH)CPh}(CO)(PPh3)2]ؒ0.5CH2Cl2.
Carbonyl-
[IrH{S(O)CPh}2(PPh3)2]. mer-Trihydridotris(triphenylphos-
phine)iridium (0.35 g, 0.44 mmol) and thiobenzoic acid (0.124
g, 0.90 mmol) were heated under reflux in toluene (40 cm3) for
ca. 40 min to give a yellow solution. Standard work-up gave
yellow microcrystals (0.3 g, 69%), m.p. 235–236 ЊC (Found: C,
60.35; H. 4.05. Calc. for C50H41IrO2P2S2: C, 60.5; H, 4.15%).
chlorohydridotris(triphenylphosphine)osmium (0.6 g, 0.58
mmol) and thiobenzamide (0.32 g, 2.33 mmol) were heated
under reflux in toluene (40 cm3) for ca. 15 h to give a dark red-
brown solution. Standard work-up gave orange microcrystals
(0.27 g, 52%), m.p. 225–227 ЊC (Found: C, 56.95; H, 3.75; N,
1.05. Calc. for C44H36ClNOOsPSؒ0.5CH2Cl2: C, 55.85; H, 3.9;
N, 1.45%).
[IrH{S(O)CMe}2(PPh3)2]. mer-Trihydridotris(triphenylphos-
phine)iridium (0.4 g, 0.51 mmol) and thioacetic acid (0.1 g,
1.31 mmol) were heated under reflux in toluene (40 cm3) for ca.
40 min to give a yellow solution. Standard work-up gave pale
yellow microcrystals (0.26 g, 59%), m.p. 197–199 ЊC (Found: C,
55.0; H, 4.2. Calc. for C40H37IrO2P2S2: C, 55.3; H, 4.3%).
[IrH2{S(NH)CPh}(PPh3)2].
mer-Trihydridotris(triphenyl-
phosphine)iridium (0.4 g, 0.51 mmol) and thiobenzamide (0.1
g, 0.72 mmol) were heated under reflux in toluene (40 cm3) for
ca. 75 min to give a yellow solution. Standard work-up gave
very pale yellow microcrystals (0.21 g, 61.4%), m.p. 198–200 ЊC
(Found: C, 60.15; H, 4.5; N, 1.85. Calc. for C43H38IrNP2S: C,
60.4; H, 4.5; N, 1.65%).
[Ir{S(O)CPh}3(PPh3)2].
Hydridobis(thiobenzoato)bis(tri-
phenylphosphine)iridium (0.35 g, 0.35 mmol) and thiobenzoic
acid (0.15 g, 1.09 mmol) were heated under reflux in toluene (30
cm3) for ca. 2 h to give an orange-yellow solution. Standard
work-up gave yellow-orange microcrystals (0.2 g, 51%),
decomp. 272–274 ЊC (Found: C, 60.5; H, 4.1. Calc. for C57H45-
IrO3P2S3: C, 60.65; H, 4.0%).
[Ru{S(O)CPh}2(PPh3)2]. Dichlorotris(triphenylphosphine)-
ruthenium (0.8 g, 0.84 mmol) and thiobenzoic acid (0.5 g, 3.62
mmol) were heated under reflux in toluene (40 cm3) for ca. 4.5 h
to give a dark red-brown solution. Standard work-up gave
orange microcrystals (0.65 g, 87%), m.p. 196–198 ЊC (Found: C,
65.95; H, 4.6. Calc. for C50H40O2P2RuS2: C, 66.7; H, 4.5%).
[IrH{S(O)CPh}2(CO)(PPh3)2]ؒ0.5CH2Cl2. Hydridobis(thio-
benzoato)bis(triphenylphosphine)iridium (0.15, 0.14 mmol)
was dissolved in toluene (30 cm3) and heated under reflux for
ca. 20 min with slow passage of carbon monoxide. Standard
work-up of the yellow solution gave very pale yellow micro-
crystals (0.12 g, 84%), m.p. 222–224 ЊC (Found: C, 58.2; H,
3.85. Calc. for C51H41IrO3P2S2ؒ0.5CH2Cl2: C, 58.2; H, 3.95%).
[Ru{S(O)CPh}2(CO)(PPh3)2]. Method A. Carbonyldihydrido-
tris(triphenylphosphine)ruthenium (0.8 g, 0.87 mmol) and
thiobenzoic acid (0.5 g, 3.62 mmol) were heated under reflux in
toluene (40 cm3) for ca. 90 min to give an orange-red solution.
Standard work-up gave orange microcrystals (0.75 g, 93%),
m.p. 242–245 ЊC (Found: C, 65.5; H, 4.6. Calc. for C51H40O3-
P2RuS2: C, 66.0; H, 4.35%).
[Ir{S(O)CPh}3(CO)(PPh3)2]ؒ0.5CH2Cl2. Tris(thiobenzoato)-
bis(triphenylphosphine)iridium (0.08 g, 0.071 mmol) in toluene
(25 cm3) was heated under reflux for ca. 1 h with slow passage
of carbon monoxide. The solution was then cooled and evapor-
ated under reduced pressure to yield an oil which on crystallis-
ation from CH2Cl2–MeOH afforded orange microcrystals (0.07
g, 87%), m.p. 246–248 ЊC (Found: C, 58.2; H, 3.75. Calc. for
C58H45IrO4P2S3ؒ0.5CH2Cl2: C, 58.6; H, 3.85%).
Method B. Carbonylchlorohydridotris(triphenylphosphine)-
ruthenium (0.6 g, 0.63 mmol) and thiobenzoic acid (0.35 g, 2.53
mmol) were heated in boiling toluene (40 cm3) for ca. 5 h to give
an orange solution. Standard work-up gave orange micro-
crystals (0.55 g, 95%) identical to those obtained by method A
(Found: C, 65.65; H, 4.15. Calc. for C51H40O3P2RuS2: C, 66.0;
H, 4.35%).
[Os{S(O)CPh}2(PPh3)2]. Dichlorotris(triphenylphosphine)-
osmium (0.5 g, 0.49 mmol) and thiobenzoic acid (0.28 g, 2.0
mmol) were heated under reflux in degassed toluene (40 cm3)
for ca. 4.5 h to give a dark red-brown solution. After evapor-
ation to dryness under reduced pressure the residue was
dissolved in dichloromethane and the solution diluted with
methanol whereupon a small amount of impure material
deposited. This was filtered off and the filtrate set aside to crys-
tallise. After 2 d the major product had deposited as dark
brown microcrystals (0.28 g, 57%), m.p. 135–137 ЊC (Found: C,
62.2; H, 3.85. Calc. for C50H40O2OsP2S2: C, 60.7; H, 4.1%).
Crystallography
A red single crystal of [Os{S(NH)CPh}2(PPh3)2] of approxi-
mate size 0.60 × 0.20 × 0.06 mm was mounted on a glass fibre.
All geometric and intensity data were taken from this sample at
293 K using an automated four-circle diffractometer (Nicolet
R3mV) equipped with Mo-Kα radiation (λ = 0.710 73 Å).
The lattice vectors were identified by the application of
the automatic indexing routine of the diffractometer to the
positions of 38 reflections taken from a rotation photograph
and centred by the diffractometer. The ω–2θ technique was
used to measure 7887 reflections (7458 unique) in the range
5 < 2θ < 50Њ. Three standard reflections (remeasured every 97
scans) showed no significant loss in intensity during data col-
lection. The data were corrected for Lorentz and polarisation
effects, and empirically for absorption. The 6536 unique data
with I у 2.0σ(I) were used to solve and refine the structure in
[Os{S(O)CPh}2(CO)(PPh3)2]. Method A. Carbonyldihydri-
dotris(triphenylphosphine)osmium (0.25 g, 0.25 mmol) and
thiobenzoic acid (0.15 g, 1.08 mmol) were heated under reflux
in toluene (40 cm3) for ca.18 h to give a dark red-brown solu-
tion. Standard work-up gave brown microcrystals (0.18 g, 72%)
identical to an authentic specimen prepared by method B.
Method B. As for method A, but using carbonylchloro-
hydridotris(triphenylphosphine)osmium in place of carbonyl-
dihydridotris(triphenylphosphine)osmium. Product isolated as
chocolate brown crystals (92%), m.p. 257–259 ЊC (Found: C,
59.9; H, 4.15. Calc. for C51H40O3OsP2S2: C, 60.2; H, 3.95%).
¯
the triclinic space group P1.
Crystal data. C50H42N2OsP2S2, M = 987.12, triclinic, space
¯
group P1, a = 10.746(2), b = 11.077(2), c = 18.152(4) Å, α =
88.29(3), β = 85.79(3), γ = 81.53(3)Њ, U = 2131 Å3, Z = 2, Dc =
1.54 g cmϪ3, F(000) = 988, µ(Mo-Kα) = 32.03 cmϪ1
.
[IrH2{S(O)CPh}(PPh3)3]ؒ0.5CH2Cl2. mer-Trihydridotris(tri-
phenylphosphine)iridium (0.4 g, 0.5 mmol) and thiobenzoic
acid (0.07 g, 0.51 mmol) were heated under reflux in benzene
(40 cm3) for ca. 30 min to afford a pale yellow solution. Stand-
The structure was solved by Patterson methods and
developed by using alternating cycles of least-squares refine-
ment and Fourier-difference synthesis. The non-hydrogen atoms
were refined anisotropically while the hydrogens were placed in
758
J. Chem. Soc., Dalton Trans., 1997, Pages 757–762

View Article Online
idealised positions (C᎐H 0.96 Å) and assigned a common iso-
tropic thermal parameter (U = 0.08 Å2). The final cycle of least-
squares refinement included 514 parameters for 6536 variables
and did not shift any parameter by more than 0.001 times its
standard deviation. The final R, RЈ values were 0.0334 and
0.0800 [weighting scheme wϪ1 = σ2Fo2 + (0.0452P)2 + 3.91P]
and the final Fourier-difference synthesis was featureless with
no peaks greater than 0.80 e ÅϪ3
.
The structure solution used the SHELXL 936 program
package on a micro Vax II computer. The molecular structure is
shown in Fig. 1, selected bond lengths and angles in Table 2.
A
yellow single crystal of [IrH{S(O)CPh}2(PPh3)2] of
approximate size 0.40 × 0.35 × 0.35 mm was mounted on a
glass fibre. Data were collected as above.
The lattice vectors were identified by application of the
automatic indexing routine of the diffractometer to the posi-
tions of 34 reflections taken from a rotation photograph and
centred by the diffractometer. The ω–2θ technique was used
to measure 7990 reflections (7609 unique) in the range
5 р 2θ р 50Њ. Three standard reflections (remeasured every 97
scans) showed no significant loss in intensity during data collec-
tion. The data were corrected as above. The 6480 unique data
with I у 3.0σ(I) were used to solve and refine the structure in
Fig. 1 Molecular structure of [Os{S(NH)CPh}2(PPh3)2]
¯
the triclinic space group P1.
Ru{µ-NH(O)CC6H4}{µ-NH(O)CPh}(µ-H)Ru(CO)(PPh3)2] ob-
tained from the parallel reaction with benzamide.2 The NMR
spectra for [RuH{S(NH)CPh}(CO)(PPh3)2] reveal the presence
of two geometrical isomers. For each a 31P-{1H} singlet and a
Crystal data. C50H41IrO2P2S2, M = 992.161, triclinic, space
¯
group P1, a = 12.353(3), b = 13.661(3), c = 14.497(3) Å,
α = 87.55(2), β = 68.89(2), γ = 72.04(2)Њ, U = 2165 Å3, Z = 2,
1
high-field H triplet [2JH P(cis) ca. 20 Hz] are consistent with trans
Dc = 1.52 g cmϪ3, F(000) 992, µ(Mo-Kα) 32.7 cmϪ1
.
phosphines stereochemistry. However, the presence of an
The structure was solved, developed and refined as above.
The hydridic ligand was not located and was omitted from the
refinement. Nevertheless its position is readily apparent in the
fully refined structure. The final cycle of least-squares refine-
ment (based on F2) included 514 parameters for 6480 variables
and did not shift any parameter by more than 0.001 times its
standard deviation. The final R, RЈ values were 0.0305 and
0.0348 [weighting scheme wϪ1 = σ2(F) + 0.001 54F2] and the
final Fourier-difference map was featureless with no peaks
1
additional coupling [3JH H Ј = 2.6 Hz] in the high-field H NMR
spectrum of the major isomer allows it to be positively iden-
tified as the one with hydride trans to the NH group of the
thiobenzamide. It follows that the minor isomer (ca. 10%) has
the hydride ligand trans to sulfur.
The reaction of [RuH(Cl)(CO)PPh3)3] with thiobenzamide
affords an air-stable orange crystalline product of stoichio-
metry [RuCl{S(NH)CPh}(CO)(PPh3)2]. The 31P-{1H} NMR
spectrum again reveals the presence of two isomers each giving
rise to a singlet, indicative of trans phosphines. Structures with
the chloride ligand trans to the N- and S-donor atoms respect-
ively of the thiobenzamidate ligand can be assumed. However
there is no evidence to indicate which is the major isomer.
The reaction of [OsCl2(PPh3)3] with thiobenzamide in boiling
toluene affords the osmium bis(chelate) complex [Os{S(NH)-
CPh}2(PPh3)2] as dark maroon air-stable crystals. As was the
case with the ruthenium analogue there are insufficient spectro-
scopic data to permit a stereochemical assignment. Therefore a
crystal structure determination was undertaken. The molecular
structure is shown in Fig. 1, selected bond length and angle data
in Table 2. The molecule has a highly distorted octahedral
geometry reflecting the small bite angles, S᎐Os᎐N 65.77(12)
and 65.57(12)Њ, of the thioamidate ligands. The bulky triphenyl-
phosphine ligands are mutually cis as are the N-donor atoms of
the thioamidate ligands. The two S-donor atoms are essentially
trans to each other with S᎐Os᎐S 152.72(5)Њ. The bond
lengths are typical for octahedral osmium().
greater than 0.80 e ÅϪ3
.
The structure solution was as above. The molecular structure
is shown in Fig. 2, selected bond lengths and angles in Table 3.
Atomic coordinates, thermal parameters, and bond lengths
and angles have been deposited at the Cambridge Crystallo-
graphic Data Centre (CCDC). See Instructions for Authors,
J. Chem. Soc., Dalton Trans., 1997, Issue 1. Any request to the
CCDC for this material should quote the full literature citation
and the reference number 186/360.
Results and Discussion
Thiobenzamidato complexes
The reactions of thiobenzamide with various ruthenium pre-
cursors afford simple mononuclear derivatives; there is no evi-
dence of the formation of binuclear species analogous to those
previously obtained with benzamide.2 Thus with [RuCl2(PPh3)3]
in boiling toluene thiobenzamide forms the mononuclear
bis(chelate) complex [Ru{S(NH)CPh}2(PPh3)2] in good yield as
air-stable orange-brown crystals. There are five possible geo-
metric isomers for a bis(chelate) complex of this form and the
rather sparse spectroscopic evidence available does not permit
an unambiguous stereochemical assignment. However, by ana-
logy with the osmium analogue which has been characterised
by X-ray diffraction methods (see below), we propose a cis-
PPh3, cis-N, trans-S stereochemical arrangement similar to that
previously reported for [Ru(C5H4NS-2)2(PPh3)2].7
Thiobenzamide reacts with the osmium complex [OsH2-
(CO)(PPh3)3] to give [OsH{S(NH)CPh}(CO)(PPh3)2] which, in
contrast to its ruthenium analogue, is obtained in an isomeri-
cally pure state. The NMR spectra comprising a 31P-{1H} sing-
1
let and a high-field H triplet with additional coupling to the
thiobenzamidato NH group (3JH H Ј = 2.0 Hz) clearly establish a
trans-phosphines stereochemistry with hydride trans to the N-
donor atom of the thiobenzamidate ligand. A similar reaction
with the osmium complex [OsH(Cl)(CO)(PPh3)3] afforded a
mixture of the two trans-phosphines isomers of [OsCl{S-
(NH)CPh}(CO)(PPh3)2].
The reaction of [RuH2(CO)(PPh3)3] with thiobenzamide
gives rise to a mononuclear species [RuH{S(NH)CPh}(CO)-
(PPh3)2]. Again we have a sharp contrast between this product
and the binuclear cyclometallated complex [(Ph3P)(OC)-
In contrast to benzamide which reacts with mer-[IrH3(PPh3)3]
to form a monodentate N-bonded derivative [IrH2{NH(O)-
J. Chem. Soc., Dalton Trans., 1997, Pages 757–762
759

View Article Online
Table 1 Spectroscopic data
IR a/cmϪ1
NMR (δ, J/Hz)b
Complex
ν(MH)
ν(CO)
1H
31P-{H}
(a) Thiobenzamide derivatives [L = NH(S)CPh]
[RuL2(PPh3)2]
—
—
—
37.66 (br)
50.96 (s)
46.11 (s)*
33.68 (s)
41.22 (s)*
Ϫ0.5 (s)
22.21 (s)
11.77 (s)
Ϫ2.95 (s)*
21.02 (s)
[RuH(L)(CO)(PPh3)2]c
Broad unresolved
ca. 1900
Broad unresolved
ca. 1850
Ϫ12.29 (t of d, 2JH P 20.5; 3JH H Ј 2.6)
Ϫ2.58 (t, 2JH P 10.5)
—
[RuCl(L)(CO)(PPh3)2]c
[OsL2(PPh3)2]
[OsH(L)(CO)(PPh3)2]
2050
—
1875
1885
1925
Ϫ14.04 (t of d, 2JH P 17.8; 3JH H Ј 2.0)
[OsCl(L)(CO)(PPh3)2]c
[IrH2L(PPh3)2]
2105
2185
Ϫ21.22 (t of d, 2JH P 16.0; 2JH H 6.5)
Ϫ21.04 (t of d of d, 2JH P 17.7; 2JH H 6.5;
3JH H Ј 3.7)
(b) Thiobenzoate and thioacetate derivatives [L = O(S)CPh, LЈ = O(S)CMe]
[RuL2(PPh3)2]
[RuL2(CO)(PPh3)2]
[OsL2(PPh3)2]
—
—
—
1945
—
—
60.46 (s)
43.55 (s)
44.0 (s)
[OsL2(CO)(PPh3)2]
[IrH2L(PPh3)3]
—
2085
2185
1930
—
—
—
17.74 (s)
Ϫ11.08 (d of t of d, 2JH P 128 and 19.5; 2JH H 4.7)
6.88 (d, 2JPP 11.4)
1.52 (t, 2JPP 11.4)
9.06 (s) 253 K
18.33 (s)
Ϫ15.08 (d of t of d, 2JH P 10.4 and 19.5; 2JH H 4.7)
Ϫ25.3 (t, 2JH P 13)
Ϫ29.38 (t, 2JH P 13)
Ϫ25.6 (br)
[IrHL2(PPh3)2]d
2220
2240
8.33 (s) 293 K
17.82 (s)
16.92 (s) 353 K
373 K
9.85 (s) 256 K
18.80 (s)
8.67 (br) 296 K
18.22 (br)
Ϫ29.2 (br)
Ϫ29.3 (br)
Ϫ29.33 (t, 2JH P 14)
Ϫ25.71 (t, 2JH P 13.4)
Ϫ29.83 (t, 2JH P ≈13)
Ϫ26.0 (br)
[IrHLЈ2(PPh3)2]d
Ϫ29.9 (br)
Ϫ29.95 (s)
16.79 (s) 380 K
4.08 (s)
3.37 (s)
[IrL3(PPh3)2]
—
—
[IrHL2(CO)(PPh3)2]
[IrL3(CO)(PPh3)2]
2140
—
2030
2030
Ϫ11.15 (t, 2JH P 12.5)
—
Ϫ13.05 (s)
a
Nujol mulls. b Spectra run in CDCl3 solution at 298 K unless otherwise specified. c Isomer mixtures (data for minor isomer indicated by asterisk).
d Variable-temperature NMR data recorded in toluene solution.
CPh}(PPh3)3],2 thiobenzamide affords an N,S-chelate product
[IrH2{S(NH)CPh}(PPh3)2] which can be isolated in good yield
as pale yellow crystals. This difference in behaviour no doubt
reflects the greater affinity of iridium() for S- as opposed
to O-donor ligands. The expected trans-phosphine stereo-
chemistry is confirmed for [IrH2{S(NH)CPh}(PPh3)2] by the
NMR data, and the presence of an additional coupling
(3JH H Ј = 3.7 Hz) allows the high-field 1H signal at δ Ϫ21.04 to be
attributed to the hydride ligand trans to the NH group.
tical with those previously reported for the corresponding
monothioacetate complex.10 We therefore conclude that the
monothiobenzoate complex has the same trans-P trans-S ligand
stereochemistry as that previously determined for the mono-
thioacetate complex. Presumably the thiobenzoate like the
latter complex displays rapid interchange of η1- and η2-
thiocarboxylate groups in solution. However, in the absence of
a suitable NMR label it has not been possible to confirm this
hypothesis. The reaction of [OsCl2(PPh3)3] with thiobenzoic
acid in refluxing toluene affords [Os{S(O)CPh}2(PPh3)2] as dark
brown air-stable crystals. The complex [Os{S(O)CPh}2-
(CO)(PPh3)2] was similarly obtained from [OsH2(CO)(PPh3)3]
or [OsH(Cl)(CO)(PPh3)3] as air-stable brown crystals, and is
tentatively assigned the same stereochemistry as that of its
ruthenium analogue.
The reaction of mer-[IrH3(PPh3)3] with thiobenzoic acid
under increasingly vigorous conditions successively affords
three products [IrH2{S(O)CPh}(PPh3)3], [IrH{S(O)CPh}2-
(PPh3)2] and [Ir{S(O)CPh}3(PPh3)2]. The first of these was
reluctant to carbonylate under mild conditions (refluxing ben-
zene) and gave a mixture of products when more vigorous
treatment was used. However the other two complexes carbon-
ylated smoothly to yield [IrH{S(O)CPh}2(CO)(PPh3)2] and
[Ir{S(O)CPh}3(CO)(PPh3)2] respectively. For four of these
products the spectroscopic data are consistent with their formu-
lation as rigid, octahedral iridium() species. However com-
plete stereochemical assignment has been possible in only two
cases, [IrH2{S(O)CPh}(PPh3)3] which has one hydride trans
to S-bonded thiobenzoate, and [IrH{S(O)CPh}2(CO)(PPh3)2]
Thiobenzoato (and thioacetato) complexes
The monothiocarboxylato complexes obtained are also mono-
nuclear and, in several instances, are identical with or similar to
products previously reported by other groups.4,5,8,9 Discussion is
therefore kept to a minimum. However one pair of iridium()
complexes [IrH{S(O)CR}2(PPh3)2] (R = Me or Ph) display
anomalous NMR spectra and merit fuller attention.
The reaction of thiobenzoic acid with [RuCl2(PPh3)3] in
refluxing toluene affords [Ru{S(O)CPh}2(PPh3)2] as orange air-
stable crystals. This complex has previously been prepared from
[RuCl2(PPh3)3] and sodium monothiobenzoate in refluxing
acetone, and assigned an all-trans structure on the basis of
chemical and spectroscopic evidence.8 Thiobenzoic acid reacts
with [RuH2(CO)(PPh3)3] or with [RuH(Cl)(CO)(PPh3)3] in
refluxing toluene to afford [Ru{S(O)CPh}2(CO)(PPh3)2] as air-
stable orange microcrystals, again there is no evidence for the
formation of any binuclear products. The spectroscopic data
for [Ru{S(O)CPh}2(CO)(PPh3)2] (see Table 1) are virtually iden-
760
J. Chem. Soc., Dalton Trans., 1997, Pages 757–762

View Article Online
Table 2 Selected bond lengths (Å) and angles (Њ) for [Os{S(NH)C-
Ph}2(PPh3)2]
Os᎐N(2)
Os᎐N(1)
Os᎐P(1)
Os᎐P(2)
Os᎐S(2)
Os᎐S(1)
P(1)᎐C(41)
P(1)᎐C(31)
2.129(4)
2.130(4)
2.298(2)
2.309(2)
2.4225(13)
2.4292(14)
1.827(6)
1.845(5)
P(1)᎐C(21)
P(2)᎐C(71)
P(2)᎐C(51)
P(21)᎐C(61)
S(1)᎐C(1)
S(2)᎐C(11)
N(1)᎐C(1)
N(2)᎐C(11)
1.868(6)
1.840(5)
1.852(5)
1.856(5)
1.741(5)
1.728(5)
1.303(7)
1.297(6)
N(2)᎐Os᎐N(1)
N(2)᎐Os᎐P(1)
N(1)᎐Os᎐P(1)
N(2)᎐Os᎐P(2)
N(1)᎐Os᎐P(2)
P(1)᎐Os᎐P(2)
N(2)᎐Os᎐S(2)
N(1)᎐Os᎐S(2)
P(1)᎐Os᎐S(2)
P(2)᎐Os᎐S(2)
N(2)᎐Os᎐S(1)
N(1)᎐Os᎐S(1)
P(1)᎐Os᎐S(1)
P(2)᎐Os᎐S(1)
85.6(2)
S(2)᎐Os᎐S(1)
152.72(5)
113.7(2)
124.3(2)
112.8(2)
117.3(2)
124.7(2)
108.3(2)
81.0(2)
159.45(12)
90.79(12)
87.87(12)
160..23(13)
101.62(5)
65.57(12)
94.70(12)
94.68(5)
C(41)᎐P(1)᎐Os
C(31)᎐P(1)᎐Os
C(21)᎐P(1)᎐Os
C(71)᎐P(2)᎐Os
C(51)᎐P(2)᎐Os
C(61)᎐P(2)᎐Os
C(1)᎐S(1)᎐Os
C(11)᎐S(2)᎐Os
C(1)᎐N(1)᎐Os
C(11)᎐N(2)᎐Os
N(1)᎐C(1)᎐S(1)
N(2)᎐C(11)᎐S(2)
81.2(2)
99.51(5)
104.3(3)
104.2(3)
108.8(4)
109.1(4)
Fig.
(PPh3)2]
2
Molecular structure of [IrH{η1-S(O)CPh}{η2-S(O)CPh}-
92.91(12)
65.77(12)
104.01(5)
96.03(5)
This suggests the presence of a substantial kinetic barrier ren-
dering the underlying process slow on the NMR time-scale.
Finally the chemical shifts recorded for both of the high-field
resonances are consistent with hydride trans to an O- rather
than an S-donor ligand. Confirmation of our NMR observ-
ations was provided by the corresponding monothioacetate
complex [IrH{S(O)CMe}2(PPh3)2] which displayed essentially
similar temperature-dependent NMR data. As a first step
towards elucidating the nature of the complexes [IrH{S-
(O)CR}2(PPh3)2] (R = Me or Ph) and explaining their NMR
spectra an X-ray diffraction study was performed on the mono-
thiobenzoate. The molecular structure is shown in Fig. 2;
selected bond lengths and angles are collated in Table 3. The
molecule contains an unexceptional octahedral iridium()
centre co-ordinated to a trans pair of triphenylphosphine lig-
ands, an η1-thiobenzoate and an η2-thiobenzoate. The angle
subtended at the iridium centre by the small ‘bite’ η2-
thiobenzoate ligand is 65.0(1)Њ. The hydride ligand was not
located but its presence is indicated by a ‘vacancy’ in the co-
ordination sphere. The bond lengths and angles are typical of
octahedral iridium() and the structure offers no immediate
explanation of the NMR data. We must therefore seek an
answer elsewhere. Four possible options, geometric isomers,
solvation, rotamers and hydride–carboxylate hydrogen bond-
ing, appear worthy of consideration.
Table 3 Selected bond lengths (Å) and angles (Њ) for [IrH{η1-S(O)-
CPh}{η2-S(O)CPh}(PPh3)2]
Ir᎐P(1)
Ir᎐S(1)
Ir᎐O(1)
P(1)᎐C(31)
P(2)᎐C(51)
P(2)᎐C(71)
S(2)᎐C(11)
O(2)᎐C(11)
C(11)᎐C(12)
2.326(1)
2.388(1)
2.303(4)
1.836(4)
1.834(7)
1.843(5)
1.740(6)
1.230(6)
1.488(8)
Ir᎐P(2)
Ir᎐S(2)
2.342(1)
2.352(1)
1.834(4)
1.846(6)
1.833(4)
1.719(6)
1.251(6)
1.485(6)
P(1)᎐C(21)
P(1)᎐C(41)
P(2)᎐C(61)
S(1)᎐C(1)
O(1)᎐C(1)
C(1)᎐C(2)
P(1)᎐Ir᎐P(2)
P(2)᎐Ir᎐S(1)
P(2)᎐Ir᎐S(2)
P(1)᎐Ir᎐O(1)
S(1)᎐Ir᎐O(1)
Ir᎐P(1)᎐C(21)
Ir᎐P(2)᎐C(51)
Ir᎐S(1)᎐C(1)
Ir᎐O(1)᎐C(1)
S(1)᎐C(1)᎐C(2)
S(2)᎐C(11)᎐O(2)
174.4(1)
92.6(1)
86.9(1)
98.0(1)
65.0(1)
113.1(2)
114.8(2)
82.5(2)
97.1(3)
122.5(4)
123.3(5)
P(1)᎐Ir᎐S(1)
P(1)᎐Ir᎐S(2)
S(1)᎐Ir᎐S(2)
86.4(1)
94.7(1)
173.3(1)
86.5(1)
108.3(1)
119.2(2)
113.7(2)
115.0(2)
115.7(2)
106.4(2)
115.4(3)
122.1(5)
117.3(4)
P(2)᎐Ir᎐O(1)
S(2)᎐Ir᎐O(1)
Ir᎐P(1)᎐C(31)
Ir᎐P(1)᎐C(41)
Ir᎐P(2)᎐C(61)
Ir᎐P(2)᎐C(71)
Ir᎐S(2)᎐C(11)
S(1)᎐C(1)᎐O(1)
O(1)᎐C(1)᎐C(2)
S(2)᎐C(11)᎐C(12)
O(2)᎐C(11)᎐C(12) 119.4(5)
Geometric isomers involving
a temperature-dependent
which has trans phosphine and trans S-bonded thiobenzoate
ligands.
equilibrium of cis and trans phosphines, structures I and II,
both of which would be expected to display rapid η1/η2 thiocar-
boxylate interchange, are considered improbable. The second
possibility, rotamers arising from restricted rotation of the η1-
thiocarboxylate groups about the Ir᎐S bonds on II, seems
unlikely since the related complex [IrH{S(O)CPh}2(CO)-
(PPh3)2] which has a similar stereochemical arrangement of
hydride, thiobenzoate and triphenylphosphine ligands does not
display anomalous NMR behaviour. An equilibrium between
unsolvated II and solvated (aquated?) species III appears more
feasible. Structures involving intramolecular hydrogen bonding
between co-ordinated water and carboxylate are well estab-
lished.11 However the crystal structure of the complex reveals
no solvent molecules co-ordinated or otherwise. Furthermore
spectroscopic and crystallographic examination of samples
crystallised from wet solvent yielded no evidence of solvation.
The final option involves an equilibrium between the flux-
ional structure II and the hydrogen-bonded form IV. According
to this scenario the η1/η2 interchange process would maintain
structure II at higher temperatures. At lower temperatures when
the fluxional process was frozen out the hydrogen-bonding
The fifth iridium thiobenzoate complex, [IrH{S(O)CPh}2-
(PPh3)2], did not conform to expectations. At ambient tempera-
ture the high-field 1H NMR spectrum showed two slightly
broad and featureless peaks at ca. δ Ϫ25.6 and Ϫ29.2 while the
31P-{1H} spectrum contained two singlets at ca. δ 8.3 and 17.8.
Moreover, the spectra were temperature dependent; on cooling
1
to 253 K the H and 31P-{1H} signals at ca. δ Ϫ25.6 and 8.3
respectively grow at the expense of those at ca. δ Ϫ29.2 and 17.8
respectively. At the same time the high-field proton resonances
gradually sharpen into well defined triplets (2JH P ca. 13 Hz).
1
Conversely as the temperature is raised the H and 31P-{1H}
signals at ca. δ Ϫ29.2 and 17.8 respectively grow at the expense
of those at ca. δ Ϫ25.6 and 8.3 until at 373 K the latter pair have
disappeared completely. Again this change is accompanied by
sharpening of the remaining high-field signal which becomes a
sharp well defined triplet. There is little evidence of coalescence
1
in either the H or 31P-{1H} spectra; throughout the tempera-
ture range 253–373 K the chemical shifts of the individual
resonances remain essentially constant (variation ±0.7 ppm).
J. Chem. Soc., Dalton Trans., 1997, Pages 757–762
761

View Article Online
However, if the equilibrium process II
IV is operating
suppression of the η1/η2 thiobenzoate exchange process on
crystallisation should cause the molecule to adopt the
hydrogen-bonded configuration IV, and the crystal structure
shows no evidence of this occurring. Indeed the oxygen atom of
the η1-thiobenzoate ligand is directed away from the hydride
ligand. Unless this arrangement can be attributed to crystal-
packing factors it must be taken as strong evidence against the
hydride–carboxylate hydrogen-bonding explanation. We are
therefore reluctantly forced to conclude that, on the basis of
currently available evidence, a definitive explanation of the
observed NMR phenomenon is not forthcoming.
References
1 Part 47, S. D. Robinson, A. Sahajpal and D. A. Tocher, J. Chem.
Soc., Dalton Trans., 1995, 3497.
2 M. B. Hursthouse, M. A. Mazid, S. D. Robinson and A. Sahajpal,
J. Chem. Soc., Dalton Trans., 1993, 2835.
3 M. B. Hursthouse, M. A. Mazid, S. D. Robinson and A. Sahajpal,
J. Chem. Soc., Dalton Trans., 1994, 3615.
4 S. D. Robinson and M. F. Uttley, J. Chem. Soc., Dalton Trans., 1973,
1912.
5 A. Dobson, S. D. Robinson and M. F. Uttley, J. Chem. Soc., Dalton
Trans., 1975, 370.
6 G. M. Sheldrick, SHELXL 93, University of Göttingen, 1993.
7 S. R. Fletcher and A. C. Skapski, J. Chem. Soc., Dalton Trans., 1972,
635.
8 R. O. Gould, T. A. Stephenson and M. A. Thomson, J. Chem. Soc.,
Dalton Trans., 1978, 769.
9 J. D. Gilbert and G. Wilkinson, J. Chem. Soc. A, 1969, 1749.
10 P. B. Critchlow and S. D. Robinson, Inorg. Chem., 1978, 17, 1902.
11 P. M. Boyer, C. P. Roy, J. M. Bielski and J. S. Merola, Inorg. Chim.
Acta, 1996, 245, 7 and refs. therein.
12 J. C. Lee, E. Peris, A. L. Rheingold and R. H. Crabtree, J. Am.
Chem. Soc., 1994, 116, 11014.
13 J. C. Lee, A. L. Rheingold, B. Muller, P. S. Pregosin and R. H.
Crabtree, J. Chem. Soc., Chem. Commun., 1994, 1021.
14 E. Peris, J. C. Lee, J. R. Rambo, O. Eisenstein and R. H. Crabtree,
J. Am. Chem. Soc., 1995, 117, 3485.
15 A. J. Lough, S. Park, R. Ramachandran and R. H. Morris, J. Am.
Chem. Soc., 1994, 116, 8356.
16 S. J. Park, R. Ramachandran, A. J. Lough and R. H. Morris,
J. Chem. Soc., Chem. Commun., 1994, 2201.
17 D. Milstein, J. C. Calabrese and I. D. Williams, J. Am. Chem. Soc.,
1986, 108, 6387.
interaction between the hydride ligand and the oxygen atom of
the monodentate thiocarboxylate ligand would become the
dominant force leading to adoption of the alternative structure
IV. The hydrogen-bonded structure IV has a very feasible
geometry with an unstrained five-membered ring. However,
from an electronic viewpoint hydrogen bonding is a less attract-
ive proposition. Transition-metal hydrides are known to par-
ticipate in hydrogen-bonding interactions but these are usually
of the form M᎐H ؒ ؒ ؒ H᎐X (X = N or O) and involve the hydride
ligand acting as a source rather than an acceptor of electron
density.12–18 An interaction of the form M᎐H ؒ ؒ ؒ O similar to the
one postulated here has recently been convincingly demon-
strated for [WH3(O2CMe)(dppe)2] (dppe = Ph2PCH2CH2PPh2)
by X-ray diffraction methods [WH ؒ ؒ ؒ O 2.33(6) Å].19 However,
in that instance the hydride involved resonated at much lower
field (δ 2.92) indicating much less shielding and hence lower
electron density than is the case for the hydride ligands in the
complexes [IrH{S(O)CR}2(PPh3)2]. Finally it is interesting that
the shift of ca. 4 ppm to lower field observed for the hydride
ligand in going from the high- to the low-temperature form
matches a similar shift to lower field observed for the hydrogen-
bonded hydride ligand relative to its non-hydrogen-bonded
neighbours in [WH3(O2CMe)(dppe)2].
18 R. C. Stevens, R. Bau, D. Milstein, O. Blum and T. F. Koetzle,
J. Chem. Soc., Dalton Trans., 1990, 1429.
19 S. A. Fairhurst, R. A. Henderson, D. L. Hughes, S. K. Ibrahim and
C. J. Pickett, J. Chem. Soc., Chem. Commun., 1995, 1569.
Received 14th October 1996; Paper 6/07011F
762
J. Chem. Soc., Dalton Trans., 1997, Pages 757–762