A general route for the synthesis of hydrido-carboxylate complexes of the type MH(CO)(κ3-OCOR)(PPh3)2 [M = Ru, Os; R = CH3, CH2Cl, C6H5, CH(CH3)2] and their use as precatalysts for hydrogenation and hydroformylation reactions

Article abstract of DOI:10.1016/j.poly.2007.10.007

The synthesis and solid-state IR, ¹H and ³¹P{¹H} NMR spectroscopic characterization of complexes of the type MH(CO)(κ³-OCOR)(PPh₃)₂ [M = Ru, Os; R = CH₃, CH₂Cl, C₆H₅ and CH(CH₃)₂] are reported in this paper. These compounds were obtained by reaction of the respective cationic complex [MH(CO)(NCMe)₂(PPh₃)₂]BF₄ with the sodium salt of the corresponding carboxylic acid in a 1:1 v/v dichloromethane/methanol solution at room temperature. The spectroscopic data of these complexes and some DFT calculations reveal an octahedral geometry with a bidentated carboxylate, two equivalent triphenylphosphines in a mutually trans positions, a linear hydride and a linear carbonyl both in the cis-positions of the coordination sphere. The catalytic results indicate that these complexes are efficient and regioselective precatalysts for the quinoline hydrogenation and for the hydroformylation of 1-hexene, under mild reaction conditions (130 °C and 4 atm H₂ and 120 °C and 15 atm H₂/CO, respectively). For benzothiophene hydrogenation, the osmium complexes showed low activities whereas the analogous ruthenium complexes were catalytically inactive under somewhat more drastic reaction conditions to those of the quinoline hydrogenation (140 °C and 10 atm H₂).

Full text of DOI:10.1016/j.poly.2007.10.007

Available online at www.sciencedirect.com
Polyhedron 27 (2008) 530–536
www.elsevier.com/locate/poly
A general route for the synthesis of hydrido-carboxylate complexes
of the type MH(CO)(j3-OCOR)(PPh3)2 [M = Ru, Os; R = CH3,
CH2Cl, C6H5, CH(CH3)2] and their use as precatalysts
for hydrogenation and hydroformylation reactions
a,
a
b
c
Merlın Rosales *, Beatriz Alvarado , Federico Arrieta , Carlos De La Cruz ,
´
a
a
a
a
´
´
Angel Gonzalez , Karely Molina , Otto Soto , Yslamar Salazar
a
La Universidad del Zulia (L.U.Z.), Facultad Experimental de Ciencias, Departamento de Quı´mica, Laboratorio de Quı´mica Inorga´nica,
Maracaibo, Venezuela
b
Laboratorio de Quı´mica Inorga´nica Teorica, Maracaibo, Venezuela
Laboratorio de Espectroscopı´a Molecular y Ato´mica, Maracaibo, Venezuela
c
Received 28 April 2007; accepted 4 October 2007
Available online 26 November 2007
Abstract
The synthesis and solid-state IR, 1H and 31P{1H} NMR spectroscopic characterization of complexes of the type MH(CO)(j3-OCOR)-
(PPh3)2 [M = Ru, Os; R = CH3, CH2Cl, C6H5 and CH(CH3)2] are reported in this paper. These compounds were obtained by reaction of
the respective cationic complex [MH(CO)(NCMe)2(PPh3)2]BF4 with the sodium salt of the corresponding carboxylic acid in a 1:1 v/v
dichloromethane/methanol solution at room temperature. The spectroscopic data of these complexes and some DFT calculations reveal
an octahedral geometry with a bidentated carboxylate, two equivalent triphenylphosphines in a mutually trans positions, a linear hydride
and a linear carbonyl both in the cis-positions of the coordination sphere. The catalytic results indicate that these complexes are efficient
and regioselective precatalysts for the quinoline hydrogenation and for the hydroformylation of 1-hexene, under mild reaction conditions
(130 ꢀC and 4 atm H2 and 120 ꢀC and 15 atm H2/CO, respectively). For benzothiophene hydrogenation, the osmium complexes showed
low activities whereas the analogous ruthenium complexes were catalytically inactive under somewhat more drastic reaction conditions
to those of the quinoline hydrogenation (140 ꢀC and 10 atm H2).
ꢁ 2007 Elsevier Ltd. All rights reserved.
Keywords: Ruthenium; Osmium; Carboxylate; Quinoline; Benzothiophene; Hydrogenation; 1-Hexene; Hydroformylation
1. Introduction
compounds [1,2]. Perhaps the Vaska and Diluzio’s complex
RuHCl(CO)(PPh3)3 [3] is the most thoroughly studied com-
pound of this series [2]. Interaction of this compound, or
alternatively of RuH2(CO)(PPh3)3, with carboxylic acids
produces RuCl(j3-OCOR)(CO)(PPh3)2 and RuH(j3-
OCOR)(CO)(PPh3)2, where R = CH3 [4,5], C2H5, C6H5
[4], C(CH3)3, C6H11, CH2Cl, CF2Cl, CF3 [5]. Examples of
analogous osmium-complexes are rather scarce in the liter-
ature [6] and, to our knowledge, complexes of the type
OsH(j3-OCOR)(CO)(PPh3)2 have not been synthesized.
The chemistry of ruthenium and osmium complexes con-
taining hydrido, carbonyl and trivalent phosphorus co-
ligands has attracted continuous attention because of the
wide variety of structures and reactivities exhibited by these
*
Corresponding author. Fax: +58 261 7598125.
E-mail address: merlin2002@cantv.net (M. Rosales).
0277-5387/$ - see front matter ꢁ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2007.10.007

M. Rosales et al. / Polyhedron 27 (2008) 530–536
531
Furthermore, interest in hydrido-carboxylate complexes of
transition-metals have been further enhanced by their cata-
lytic activity in a number of industrially important chemical
reactions [7].
In the present paper, we reported the synthesis and spec-
troscopic characterization of a group of complexes with
general formulae MH(j3-OCOR)(CO)(PPh3)2 [M = Ru
and Os; R = CH3, CH2Cl, C6H5 and CH(CH3)2]; these
were tested as precatalysts for the homogeneous hydroge-
nation of polynuclear heteroaromatic nitrogen and sulfur
compounds (quinoline and benzothiophene) and for the
olefin hydroformylation (1-hexene).
1950 cmꢀ1. Comparison with the reported data for RuH-
(CO)(j3-OCOCMe3)(PPh3)2 and OsHCl(CO)(PPh3)3 [8]2
1
allowed us to assign the main bands in these wavenumber
regions to the m(MH) and m(CO) vibrations of terminal
MH and linear MCO groups, respectively [9]. The C–O dis-
˚
tance is estimated to be 1.15 0.02 A for the carbonyls of
these complexes [9b]. The wide wavenumber range found
for the m(MH) vibration relative to the m(CO)-range shows
the high sensitivity of the metal–hydrogen bond to changes
in electron density around the central metal atom [9b]. This
is affected by the type of metal in question [i.e. m(RuH) =
2010–2070 cmꢀ1 whereas m(OsH) = 2090–2133 cmꢀ1] and
by charge donation or withdrawal of the other co-ligands
[e.g. m(RuH) = 2019 or 2036 cmꢀ1 for RuH(CO)(j3-
OCOCH3)(PPh3)2 or RuH(CO)(j3-OCOCH2Cl)(PPh3)2,
respectively].
2. Results and discussion
2.1. Synthesis and characterization of complexes
MH(CO)(j3-OCOR)(PPh3)2
The infrared spectra of the new complexes (with the
exception of the Os-complex VII) also show the two
expected, symmetry-allowed, infrared-active absorptions
of the carboxylate group at 1560–1521 [masym(OCO)] and
1457–1425 [msym(OCO)] cmꢀ1. These values indicate biden-
tated coordination between the two equivalent oxygen
atoms of the OCOR groups and the Ru or Os atom. Indeed
when the carboxyl group is monodentated, one of the C–O
bonds should have enhanced double bond character and
therefore give rise to a high frequency band in the 1590–
1650 cmꢀ1-region which is not observed in our spectra [10].
High field 1H NMR spectra showed triplets between
ꢀ16.3 and ꢀ16.9 ppm with coupling constants close to
20 Hz for the ruthenium complexes, whereas for the corre-
sponding osmium ones the triplets appear between ꢀ19.3
and ꢀ20.0 ppm (2JHP close to 16 Hz). These values are con-
sistent with a hydride ligand coupled to two equivalent tri-
phenylphosphines in mutually trans positions [5]. This is
further supported by the spectra from the 31P{1H}NMR
experiments which show singlets between 41.8 and
46.4 ppm and between 23.3 and 25.2 ppm for the Ru and
Os compounds, respectively.
Our results and interpretations are summarized in Table
1. It clearly indicates an octahedral geometry with four
monodentated ligands (hydride, carbonyl and two PPh3)
and one bidentated carboxylate group coordinated to the
central Ru or Os atom.
A more detailed analysis of the infrared spectra in the
2100–1850 cmꢀ1 region (Fig. 1) indicates the presence of
two isomers for the above octahedral complexes in solid
state, as may be observed by the presence of two hydride
bands. For complexes with the general formula RuH(CO)-
(j3-OCOR)(PPh3)2, four geometrical isomers could be
drawn, which are shown in Fig. 2; for convenience we refer
to these below as the trans-PPh3/trans-PPh3 (A), trans-CO/
trans-PPh3 (B), trans-H/trans-PPh3 (C) and trans-H/trans-
CO (D) isomers.
The complexes MH(CO)(j3-OCOR)(PPh3)2 (III–X)
were synthesized in high yields (>70%) by reaction of the
respective cationic complex [MH(NCMe)2(CO)(PPh3)2]-
BF4 (Ru = I and Os = II) with the sodium salt of the cor-
responding carboxylic acid in a 1:1 v/v dichloromethane/
methanol solution, at room temperature (see Eq. (1)).
O
PPh3
PPh3
M
BF4
R
R
C
O
C
NCMe
NCMe
ONa
+ NaBF4
OC
M
O
OC
MeOH, CH2Cl2
H
H
PPh3
PPh3
M = Ru; R = CH3 (III), CH2Cl (IV), CH(CH3)2 (V) and C6H5 (VI)
M = Os; R = CH3 (VII), CH2Cl (VIII), CH(CH3)2 (IX) and C6H5
(X)
ð1Þ
The products III–X were white unstable solids that had
to be manipulated and stored strictly under N2 atmo-
sphere. Three of the hydrido-carboxylate ruthenium com-
plexes (R = CH3, CH2Cl and C6H5) were reported early
by Dobson and Robinson [4] by the reaction of RuH2-
(CO)(PPh3)3 with the corresponding carboxylic acid and
´
by Sanchez-Delgado et al. [5] by slight modifications of
the Robinson‘s method. However, this method is not of
general use for Ru complexes, in the sense that the reaction
time, the amount of the carboxylic acid added and the
solvent used depend of the carboxylate ligand. On the other
hand, the synthetic route that we are reporting in this
article is of general application and allowed us to prepare
the osmium analogues which, to our knowledge, were
unknown to the date. No satisfactory analytical data could
be obtained for these complexes, which darken after a few
hours even under inert atmosphere at room temperature.
Structural characterization of these complexes had to be
based on their FT-IR and NMR spectroscopic properties
and by comparison of those with the spectra from well-
known complexes.
The FT-IR spectra of all complexes III–X showed med-
1
For RuH(CO)(OCO(CH3)3)(PPh3)2; IR: m(Ru–H) = 2000 cmꢀ1 and
ium-to-weak absorptions between 2150 and 2000 cmꢀ1
,
m(CO) = 1925 cmꢀ1
.
2
For OsHCl(CO)(PPh3)3. IR: m(Os–H) = 2097 and m(CO) = 1912 cmꢀ1
.
and very strong and broad bands between 1900 and

532
M. Rosales et al. / Polyhedron 27 (2008) 530–536
Table 1
Selected FT-IRS, 1H and 31P{1H} NMR data for [MH(g3-OCOR)(CO)(PPh3)2] {M = Ru and R = CH3 (V), CH2Cl (VI), CH(CH3) (VII), Ph (VIII);
M = Os and R = CH3 (IX), CH2Cl (X) CH(CH3)2 (XI) and Ph (XII)}
Complex
FT-IRS
1H NMRa
31P{1H} NMRb
Wavenumbers (cmꢀ1) and relative intensitiesc
Chemical shifts (ppm) and coupling constants (Hz)d
m(MH)
m(CO)
masym(OCO)
m
sym(OCO)
d(RCO2)
d(MH)
J(PH)
d(PPh3)
III
IV
V
2019 (m)
2036 (m)
2070 (sh)
2010 (m)
2110 (br)
2133 (m)
2090 (sh)
2105 (m)
1929 (vs)
1933 (vs)
1909 (vs)
1925 (vs)
1928 (s)
1925 (vs)
1909 (vs)
1915 (vs)
1531 (s)
1560 (s, br)
1523 (s)
1521 (s)
Obscured by other bands
1550 (s, br)
1556 (s)
1551 (s)
1457 (s)
1432 (s, br)
1425 (s)
1.7 (s)
2.4 (s)
ꢀ16.7 (t)
ꢀ16.9 (t)
ꢀ16.3 (t)
ꢀ16.4 (t)
ꢀ19.6 (t)
ꢀ20.0 (t)
ꢀ19.3 (t)
ꢀ19.5 (t)
20
20
21
20
16
16
17
16
46.4 (s)
42.8 (s)
41.8 (s)
46.0 (s)
24.7 (s)
25.2 (s)
23.3 (s)
24.2 (s)
1.2 (sept) 0.1 (d, J = 7.1)
7.6 (m)
1.7 (s)
2.0 (s)
1.1 (sept) 0.1 (d, J = 6.3)
7.1 (m)
VI
VII
VIIIe
IX
X
1425 (s)
1432 (s)
1433 (s, br)
1431 (s)
a
Negative values are upfield from the signal of external SiMe4.
Positive values are downfield from the signal of external 85% H3PO4.
FT-IRS abbreviations: m = bond-stretching, vs = very strong, s = strong, m = medium, w = weak, vw = very weak, br = broad and sh = shoulder.
NMR abbreviations: s = singlet, d = doublet, t = triplet, sept = septuplet, m = multiplet.
The infrared spectrum also shows bands at 3480 (s, br), 2200 (vw, br), 2050 (vw, br), 1910 (s, br), 1610 (vs), 1490 (m), 1410 (s), 1380 (s), 1325 (m) and
b
c
d
e
1100 (vs) cmꢀ1. These can be assigned to vibrational modes of water, N„C M–H, C@O, P–Ph3 and BF4ꢀ. They indicate a mixture of small amounts of
[OsH(NCMe)2(CO)(PPh3)2]BF4 and OsH(OCOCH2Cl)(CO)(PPh3)2 (VIII) which probably adsorbed water during the pressing-disc procedure.
1.8
A
B
(V)
PR3
M
PR3
M
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
R´
R´
O
C
O
C
(IV)
(V)
R3P
O
OC
O
(IV)
H
H
CO
PR3
(III)
(VI)
C
D
(VI)
(III)
H
PR3
M
R´
R´
O
C
O
C
R3P
R3P
M
O
R3P
O
OC
2100
2050
2000
1950
1900
1850
WAVENUMBER cm-1
CO
H
Fig. 1. Possible isomers for complexes MH(CO)(j3-OCOR)(PPh3)2
(M = Ru, Os).
Fig. 2. FTIR for complexes RuH(CO)(j3-OCOR)(PPh3)2 in the 2100–
1850 cmꢀ1 region.
Table 2
Optimized total energies (hartrees) of the isomers of the model compound
Fe(H)(CO)(OOCH)(PH3)2
In other to determine the stability of these isomers, the-
oretical DFT-calculations were performer for these four
isomers with the FeH(CO)(OOCH)(PH3)2 models using
the STO/6-31+G(d,p) extended basis; in Table 2 are listed
the calculated energies of these four isomers. As may be
observed the isomer A, which present the two phosphine
ligands in a mutually trans position was the more stable
isomer, having an energy 20.1, 12.0 and 24.4 kJ molꢀ1 les-
ser than those of B, C and D, respectively. Isomers B and D
were the least stable of the four isomers. The order of sta-
bility was A < C < B < D. In view of these results, we pro-
posed that, in solid state, isomer A was the major isomer
whereas C was the minor one. Halide complexes of the type
MX(CO)(j3-OCOR)(PPh3)2 have been found to have sim-
ilar structure to that of A (M = Ru, X = Cl) or C (M = Os,
X = Br) isomers.
Isomer
Energy
A
B
C
D
ꢀ2253.089846
ꢀ2253.082210
ꢀ2253.085265
ꢀ2253.080594
Although small amounts of the isomer C of complexes
III–X can be presented (less than 10%) in the solid state,
the spectroscopic data in solution are in agreement with a
stereochemistry in which the two triphenylphosphine are
in a mutually trans position, the carboxylate ligand is
bonded in a bidentated fashion and the hydride and the
carbonyl ligand occupy the cis positions of the coordina-
tion sphere (isomer A).

M. Rosales et al. / Polyhedron 27 (2008) 530–536
533
Table 3
2.2. Catalytic activities of complexes MH(CO)
(j3-OCOR)(PPh3)2
Quinoline hydrogenation catalyzed by MH(CO)(j3-OCOR)(PPh3)2
R
pKaa
107 ri (msꢀ1
)
The new Rh and Os complexes MH(CO)(j3-
OCOR)(PPh3)2 were tested as catalytic homogeneous pre-
cursors for the regioselective hydrogenation of quinoline
(Q) to 1,2,3,4-tetrahydroquinoline (THQ) and of benzo-
thiophene (BT) to 2,3-dihydrobenzothiophene (DHBT)
under mild temperatures and low hydrogen pressures; see
Eqs. (2) and (3).
Ru
Os
CH2Cl
C5H6
CH3
2.86
4.20
4.76
4.86
6.13 0.12
6.94 0.15
9.24 0.09
8.00 0.05
2.05 0.06
3.51 0.08
4.23 0.08
2.65 0.05
CH(CH3)2
Conditions:
[cat] = 1.0 · 10ꢀ3 M,
[Q] = 0.17 M,
p(H2) = 4 atm,
T = 130 ꢀC, solvent = xylene, ri (initial rate) are expressed in mean-
s
SEM; P = 0.05.
a
pKa of the acid from which the carboxylate ligand is derived. Yield:
cat
70–80% for Ru complexes and 40–50% for Os ones.
+ 2 H2
ð2Þ
130ºC, 4 atm
N
propose that the equilibrium between these two species
(Eq. (4)) is very fast and, therefore, is not the rate-deter-
mining step of the catalytic reaction. Preliminary kinetic
studies indicate that the hydrogenation rate is independent
of the amount of quinoline present in the reaction mixture;
these results are consistent with the opening of the carbox-
ylate ligand and the coordination of Q molecule being a
very fast step.
N
H
cat
+ H2
ð3Þ
140ºC, 10 atm
S
S
The hydrogenation of the heterocyclic ring of polynu-
clear heteroaromatic nitrogen and sulfur compounds is
an area of considerable interest in connection with the
industrially important hydrodenitrogenation (HDN) and
hydrodesulfurization (HDS) reactions, respectively [11].
Although some Rh- and Ir-complexes have been used for
the hydrogenation of the heterocyclic ring of Q and related
compounds, the only ruthenium and osmium precatalysts
used in this reaction were RuHCl(PPh3)3 [12], [MH(CO)-
(NCMe)2(PPh3)2]+ (M = Ru or Os) [13] and RuCl2-
(NCMe)4/tris(pyrazolyl)borate [14]. Recently, Borowski
et al. [15] reported the regioselective hydrogenation of the
non-heterocyclic rings of Q, by using [RuH2-(g2-H2)2-
(PCy3)2] (Cy = cyclohexyl) as the precatalyst. For the BT
hydrogenation, the complexes [Ru(NCMe)3(triphos)]+2
[16] and [RuH2(g2-H2)2(PCy3)2] [17] were the only precur-
sors that showed very high activities for this reaction.
The hydrogenation of Q to THQ was achieved by all of
the ruthenium and osmium hydrido-carboxylate complexes
with high catalytic activities; the results are collected in
Table 3 together with the pKa values of the R groups of
the carboxylate ligand of the Ru and Os complexes. As
may be observed, the ruthenium complexes were twice
more active than the corresponding osmium ones. It is
clear that the hydrogenation activities of these compounds
(except for the hydrido-isobutyrate cases where steric
effects seem to be important) are affected by the electron-
releasing character of the carboxylate substituents. Thus,
for example, a plot of logri versus pKa of the acid from
which the carboxylate ligand is derived shows an excellent
correlation [logri(Ru) = ꢀ6.48 + 0.09pKa, r2 = 0.99], as
may be observed in Fig. 3; for the osmium complexes, a
O
C
PPh3
M
PPh3
M
R
O
C
O
R
OC
O
+ Q
OC
N
H
H
PPh3
PPh3
ð4Þ
On the other hand, the hydrogenation of BT to DHBT
was observed only with the osmium hydrido-carboxylate
complexes (Table 4), the activities being very low (less than
10% conversion); the analogous ruthenium complexes were
catalytically inactive for this reaction. Because of the close
and very small activities of these osmium systems, no con-
clusions may be extracted on the relation between activity
and the donor character of the carboxylate substituent.
-6.04
-6.08
-6.12
-6.16
-6.20
similar tend was observed [logri(Os) = ꢀ7.17 + 0.17pKa,
2
3.0
3.5
4.0
4.5
´
r = 0.99]. Sanchez-Delgado et al. [7] found the contrary
pKa
effect for the propionaldehyde hydrogenation, which was
explained through a mechanism involving equilibrium
between species containing the carboxylate ligand coordi-
nated in a bi- and monodentated mode. In our case, we
Fig. 3. Hydrogenation of quinoline catalyzed by RuH(CO)-
(j3-OCOR)(PPh3)2 showing the relation between initial rate (ri) and the
nature of the carboxylate ligand. Conditions as in Table 3.

534
M. Rosales et al. / Polyhedron 27 (2008) 530–536
Table 5
In order to find new and lesser expensive precatalysts
Hydroformylation of hex-1-ene catalyzed by MH(CO)(j3-OCOR)(PPh3)2
(M = Ru, Os)
than the rhodium ones for the alkene hydroformylation,
in the present work we used the hydrido-carboxylate com-
plexes MH(CO)(j3-OCOR)(PPh3)2 as catalyst precursors
for this reaction. This reaction is a well-known synthetic
tool for a wide range of organic molecules of commercial
value and it is also one of the largest scale applications of
homogeneous catalysis in industry. The major capacity is
based on the processes using phosphine-modified cobalt
[18] and rhodium carbonyl systems [19] but some ruthe-
nium precatalysts have been used for this reaction, such
as RuH2(CO)2(PPh3)2 [20], [RuH(CO)(NCMe)2(PPh3)2]-
BF4 [21], [Ru(CO)2(g5-Cp)]2 [22] and RuH2(CO)2(TPP-
MS)2 [23].
R
pKaa
TF (l/b)
Ru
Os
CH2Cl
C5H6
CH3
2.86
4.20
4.76
4.86
38 (2.8)
80 (3.5)
155 (2.5)
180 (3.0)
10 (1.6)
15 (0.8)
22 (1.2)
27 (0.9)
CH(CH3)2
Conditions: [cat] = 3.3 · 10ꢀ3 M, [1-hexene] = 1 M, p(H2) = 15 atm,
T = 120 ꢀC, solvent = toluene, TF = mole of product per mole of catalyst
per hour, l/b = linear to isoaldehyde ratio.
a
pKa of the acid from which the carboxylate ligand is derived.
The results found both for the Q and BT hydrogenation
reactions as for the 1-hexene hydroformylation are promis-
ing for two reasons. First, although the Ru precatalysts
synthesized in this work are close to one magnitude order
less active than Rh ones, ruthenium is a metal less expen-
sive than rhodium and, on the other hand, it is possible
through the variation of the electronic properties of the
carboxylate and phosphine ligands to increase the catalytic
activities of our systems. Finally, it is convenient to men-
tion that Fe complexes have shown to have less catalytic
activity for both reactions. For example, Fe(CO)5 is
inactive for quinoline hydrogenation and Fe(CO)4(PBu3)
presented low activity (1% in 5 h) for hydrogenation of
5,6-benzoquinoline under drastic water-gas shift reaction
conditions [27], whereas the activities of the Fe systems
for olefin hydroformylation are four magnitude order
lower than those of the ruthenium ones [18c].
The ruthenium and osmium hydrido-carboxylate com-
plexes were efficient and regioselective precatalysts for the
hydroformylation of 1-hexene towards heptanal and 2-
methyl-hexanal (Eq. (5)), under mild reaction conditions
(120 ꢀC and 15 atm H2/CO) in toluene as the solvent; the
results in turnover frequency (TF) are showed in Table 5.
O
H
cat
+
+ H2/CO
O
120ºC, 15 atm
H
ð5Þ
The ruthenium complexes were more active than the
corresponding osmium ones and similar to the Q hydroge-
nation described before, the complexes with carboxylates
presenting donor groups were more active than those
containing electron-withdrawing substituents. Again, this
effect may be explained through a mechanism which did
not involve the opening of the carboxylate ligand as the
rate-determining step. For the ruthenium complexes, a
plot of logTF versus pKa of the acid from which the
carboxylate ligand is derived shows a very good correla-
tion, logTF(Ru) = 0.61 + 0.32pKa, r2 = 0.96 (Fig. 4); a
similar graph may be constructed for the osmium ana-
logues [logTF(Ru) = 0.41 + 0.20pKa, r2 = 0.92]. Finally,
the regioselectivity find for ruthenium complexes, mea-
sured as the linear to branch ratio (l/b), is ranging between
2.5 and 3.5, which is somewhat higher that those found for
other ruthenium complexes; the l/b ratio for the osmium
analogues varied from 0.8 to 1.6.
More detailed kinetic and coordination chemistry stud-
ies of these catalytic reactions (hydrogenation of Q and BT
and olefin hydroformylation) are in progress, aimed at bet-
ter understanding these systems and at improving even
more both the activity as the selectivity properties of the
hydroformylation reaction.
2.2
2.1
2.0
1.9
1.8
1.7
1.6
Table 4
Benzothiophene hydrogenation catalyzed by OsH(CO)(j3-OCOR)(PPh3)2
R
Conversion (%)
CH2Cl
C5H6
CH3
8.9
6.2
8.7
9.1
3.0
3.5
4.0
4.5
pKa
CH(CH3)2
Fig. 4. Hydroformylation of 1-hexene catalyzed by RuH(CO)-
(j3-OCOR)(PPh3)2 showing the relation between turnover frequency
(TF) and the nature of the carboxylate ligand. Conditions as in Table 5.
Conditions: [cat] = 1.0 · 10ꢀ3 M, [BT] = 0.1 M, p(H2) = 15 atm, T = 140 ꢀC,
t = 8 h, solvent = xylene.

M. Rosales et al. / Polyhedron 27 (2008) 530–536
535
3. Experimental
even in inert atmosphere. The infrared spectrum of this
solid as KBr pellet revealed a solid mixture of small
amounts of compound II and OsH(OCOCH2Cl)-
(CO)(PPh3)2 (VIII) (see footnote e of Table 1). So increas-
ing amounts of the Na salt and a longer period of reaction
time (8 h) were needed to obtain white microcrystals of
complex VIII. 1H- and proton-decoupled 31P NMR spectra
indicated that only compound VIII was present in the
CDCl3 solutions.
3.1. General considerations
All manipulations were performed under argon using
standard Schlenck techniques. Solvents, MCl3 Æ 3H2O
(M = Ru, Os) for the home-made preparation of
[MH(NCMe)2(CO)(PPh3)2]BF4, M = Ru (I) [21] and Os
(II) [13c], quinoline, benzothiophene and 1-hexene were
obtained as high purity compounds from Aldrich. Prior
to use solvents, quinoline and 1-hexene were distilled over
appropriate standard agents under nitrogen. Research
grade H2 and CO gases were obtained from Praxair Vene-
zuela. Before use, these gases were dried by passing them
through a column containing CaSO4.
3.4. Computational details
The optimized energy calculations of complexes MH-
(j3-OCOR)(CO)(PPh3)2 were carried out using Fe as
metal, R = H and the simplified PH3 analogues in order
to have a better calculation level. These were performed
with the GAUSSIAN-98 computational package by using the
B3LYP [24] hybrid method of density functional density
theory (DFT); the optimizations were carried out with
the 6-31+G(d,p) basis sets [25].
3.2. Analytical procedures
FT-IR spectra of dried discs (i.e. solid samples + KBr)
were obtained using a 13-mm-die, 13-mm disk holder, a
Caver hydraulic press (>12 ton), a vacuum pump and a
1
Shimadzu FTIR-8300 interferometer. High resolution H
3.5. Details of the catalytic runs
and 31P{1H} NMR spectra of the samples dissolved in
CDCl3were obtained at 300 MHz and 121 MHz, respec-
tively, in a Bruker AC-300 instrument at room tempera-
The apparatus and the reaction procedure for the cata-
lytic runs of hydrogenation [13] and hydroformylation reac-
tions [26] have been described in details elsewhere. Briefly, a
solution of the precatalyst, substrate, internal standard and
solvent were placed in a high pressure reactor which was
flushed three times with argon. When this system reached
thermal equilibrium, H2 (or an H2/CO mixture) was admit-
ted into the reactor at the desired pressure. The reactions
were followed by both measuring the gas pressure-drop as
a function of time and by gas chromatography. Each
reaction was repeated at least twice in order to ensure
reproducibility of the results. All straight lines were fitted
by conventional linear regression programs to r2 > 0.98.
1
ture; H- and 31P-chemical shifts are expressed in part per
million down-field and up-field from SiMe4 and H3PO4,
respectively.
Gas chromatographic analyses were performed in a
3300 Series VARIAN instrument fitted with a flame ioniza-
tion detector (FID) and a 2 m 20% SP-2100 on a 0.1% car-
bowax 100/120 Supelcoport column, using N2 as carrier
gas and naphthalene (for the Q and BT hydrogenation)
and n-heptane (for hydroformylation of 1-hexene) as inter-
nal standards. The results were quantified with a VARIAN
4400 microcomputer.
3.3. Synthesis of complexes
Acknowledgements
RuH(j3-OCOR)(CO)(PPh3)2 [R = CH3 (III), CH2Cl (IV),
CH(CH3)2 (V), and Ph (VI)] were prepared by the following
method. To a solution of [RuH(NCMe)2(CO)(PPh3)2]BF4
(I) (0.5 mmol) in a 1:1 v/v mixture of CHCl2 and MeOH
(25 mL) was added the corresponding carboxylic Na salt
(1.5 mmol). The mixture was stirred vigorously at room
temperature for 3–4 h, giving a pale yellow solution which
was evaporated under vacuum to about, one-third of its ini-
tial volume, obtaining a white precipitate. This solid was
washed with methanol and diethylether, and then dried in
vacuum. It affords white solids (yields: 70–80%).
This work was supported from Consejo de Desarrollo
´
´
Cientıfico y Humanıstico through Project (CONDES-
0174-02) and for the acquisition of a gas chromatograph.
´
We thank Dr. Jose Ortega for his valuable help to record
the NMR spectra.
References
[1] (a) C.A. McAuliffe (Ed.), Transition Metal Complexes of Phospho-
rus, Arsenic and Antimony Ligands, MacMillan, London, 1973, pp.
72–86;
The corresponding OsH(j3-OCOR)(CO)(PPh3)2 com-
plexes [where R = CH3 (VII), CH(CH3)2 (IX), Ph (X)] were
synthesized by a similar method but with a reaction time
of approximately 7 h. All products were white solids
(yields: 70–80%). However, reaction of [OsH(NCMe)2-
(CO)(PPh3)2]BF4 (II) with Na[OCOCH2Cl] afforded a
amourphous, white solid which darken after few minutes
(b) C.A. McAuliffe, W. Levanson (Eds.), Phosphine, Arsine and
Stibine Complexes of the Transition Elements, Elsevier, Amsterdam,
1979, pp. 112–126.
´
[2] R.A. Sanchez-Delgado, M. Rosales, Trends Inorg. Chem. 3 (1993)
479.
[3] L. Vaska, J.W. Diluzio, J. Am. Chem. Soc. 83 (1961) 1262.
[4] (a) A. Dobson, S.D. Robinson, Platinum Met. Rev. 20 (1976) 36;
(b) A. Dobson, S.D. Robinson, Inorg. Chem. 16 (1977) 321.

536
M. Rosales et al. / Polyhedron 27 (2008) 530–536
´
´
´
´
[5] R.A. Sanchez-Delgado, U. Thewalt, N. Valencia, A. Andriollo, R.-L.
(c) M. Rosales, A. Gonzalez, J. Navarro, H. Soscun, J. Zarraga,
´
Marquez-Silva, J. Puga, H. Scho¨llhorn, H.-P. Klein, B. Fontal, Inorg.
Inorg. Chim. Acta 257 (1997) 131;
´
Chem. 25 (1986) 1097.
(d) M. Rosales, F. Arrieta, J. Castillo, A. Gonzalez, J. Navarro, R.
[6] (a) J.E. Amstrong, W.R. Robinson, R.A. Walton, J. Chem. Soc.,
Chem. Commun. (1981) 1120;
Vallejo, Stud. Surf. Sci. Catal. 130D (2000) 3357;
´
´
(e) M. Rosales, J. Castillo, A. Gonzalez, L. Gonzalez, K. Molina, J.
(b) J.E. Amstrong, W.R. Robinson, R.A. Walton, Inorg. Chem. 22
(1983) 1301;
Navarro, I. Pacheco, Transition Met. Chem. 29 (2004) 221.
´
[14] Y. Alvarado, M. Busolo, F. Lopez-Linares, J. Mol. Catal. 142 (1999)
(c) D.C. Moore, A.S. Alves, G. Wilkinson, J. Chem. Soc., Chem.
Commun. (1981) 1164;
(d) D.A. Tocher, M.D. Walkinshaw, J. Organomet. Chem. 232 (1982)
C51;
163.
[15] A.F. Borowski, S. Sabo-Etienne, B. Donnadieu, B. Chaudret,
Organometallics 22 (2003) 1630.
[16] C. Bianchini, A. Meli, S. Moneti, W. Oberhauser, F. Vizza, V.
´
(e) T. Behing, G. Wilkinson, T.A. Stephenson, D.A. Tocher, M.D.
Walkinshaw, J. Chem. Soc., Dalton Trans. (1983) 2109;
(f) A.R. Chakravarty, F.A. Cotton, D.A. Tocher, J. Chem. Soc.,
Chem. Commun. (1984) 501;
Herrera, A. Fuentes, R.A. Sanchez-Delgado, J. Am. Chem. Soc. 121
(1999) 7071.
[17] A.F. Borowski, S. Sabo-Etienne, B. Donnadieu, B. Chaudret,
Organometallics 22 (2003) 4803.
(g) F.A. Cotton, A.R. Chakravarty, D.A. Tocher, T.A. Stephenson,
Inorg. Chim. Acta 87 (1984) 115;
(h) S.M. Tetrick, V.T. Coombe, G.A. Heath, T.A. Stephenson, R.A.
Walton, Inorg. Chem. 23 (1984) 4567;
(i) A.R. Chakravarty, F.A. Cotton, D.A. Tocher, Inorg. Chem. 23
(1984) 4697;
[18] (a) J.P. Collman, L.S. Hegedus, J.R. Norton, R.G. Finke, Principles
and Applications of Organotransition Metal Chemistry, University
Science Book, Mill Valley, 1988, pp. 564–567;
(b) B. Cornils, W. A Herrmann (Eds.), Applied Homogeneous
Catalysis with Organometallic Compounds, VCH, Weinhein, 1996,
pp. 571–618;
(j) A. Dobson, S.D. Robinson, M.F. Uttley, J. Chem. Soc., Dalton
Trans. (1975) 370;
(k) D. Robinson, M.F. Uttley, J. Chem. Soc., Dalton Trans. (1973)
1912.
(c) S. Bhaduri, D. Mukesh, Homogeneous Catalysis. Mechanisms and
Industrial Applications, Wiley, New York, 2000, pp. 55–89.
[19] P. Van Leeuwen, C. Claver (Eds.), Rhodium Catalyzed Hydrofor-
mylation, Kluwer Academic Publishers, Dordrecht, 2000.
´
´
´
[7] R.A. Sanchez-Delgado, N. Valencia, R.-L. Marquez-Silva, A.
[20] R.A. Sanchez-Delgado, J.S. Bradley, G. Wilkinson, J. Chem. Soc.,
Andriollo, M. Medina, Inorg. Chem. 25 (1986) 1106.
Dalton Trans. (1976) 309.
´
[8] L. Vaska, J. Am. Chem. Soc. 88 (1966) 4100.
[21] R.A. Sanchez-Delgado, M. Rosales, A. Andriollo, Inorg. Chem. 30
[9] (a) C. De La Cruz, Norman Sheppard, J. Mol. Struct. 22 (1990) 141;
(b) C. De La Cruz, Act. Cient. Venezolana 43 (1992) 269;
(c) S.L. Morrison, J.J. Turner, J. Mol. Struct. 317 (1994) 39.
[10] K. Nakamoto, Infrared Spectra of Inorganic and Coordination
Compounds, 4th ed., Wiley–Interscience, New York, 1986.
[11] (a) H. Topsoe, B.S. Clausen, F.E. Massoth, Hydrotreating Catalysis,
Springer, Berlin, 1996;
(1991) 1170.
[22] E. Cesarotti, A. Fusi, R. Lugo, G.M. Zanderighi, J. Mol. Catal. 4
(1978) 205.
[23] P.J. Baricelli, E. Lujano, M. Modrono, A.C. Marrero, Y.M. Garcıa,
´
˜
´
A. Fuentes, R.A. Sanchez-Delgado, J. Organomet. Chem. 689 (2004)
3782.
[24] (a) A.D. Becke, J. Chem. Phys. 98 (1993) 5648;
(b) C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785;
(c) B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 157
(1989) 200.
[25] (a) J.B. Collins, P.V.R. Schleyer, J.S. Binkley, J.A. Pople, J. Chem.
Phys. 64 (1967) 5142;
´
(b) R.A. Sanchez-Delgado, in: B.R. James, P.W.N.M. van Leeuwen
(Eds.), Organometallic Modeling of the Hydrodesulfurization and
Hydrodenitrogenation Reactions, Kluwer Academic Publishers,
Dordrecht, 2000.
[12] (a) R.H. Fish, J.L. Tan, A.D. Thormodsen, J. Org. Chem. 49 (1984)
4500;
(b) K.D. Dobbs, W.J. Hehre, J. Comput. Chem. 8 (1987) 880;
(c) M.S. Gordon, Chem. Phys. Lett. 76 (1980) 163;
(d) K. Raghavachari, G.W. Trucks, J. Chem. Phys. 91 (1989)
1062.
´
´
(b) R.A. Sanchez-Delgado, E. Gonzalez, Polyhedron 8 (1989) 1431.
´
[13] (a) M. Rosales, Y. Alvarado, M. Boves, R. Rubio, R. Sanchez-
Delgado, H. Soscu´n, Transition Met. Chem. 20 (1995) 246;
(b) M. Rosales, J. Navarro, L. Sa´nchez, A. Gonza´lez, Y. Alvarado, R.
Rubio, C. De La Cruz, T. Rajmankina, Transition Met. Chem. 21
(1996) 11;
[26] M. Rosales, A. Gonza´lez, Y. Guerrero, I. Pacheco, R.A. Sa´nchez-
Delgado, J. Mol. Catal. A 270 (2007) 141.
[27] R.H. Fish, Ann. N. Y. Acad. Sci. 415 (1983) 292.