Catalytic isomerization of allylic alcohols promoted by complexes [RuCl2(η6-arene)(PTA-Me)] under homogeneous conditions and supported on Montmorillonite K-10

Article abstract of DOI:10.1016/j.molcata.2012.10.025

The mononuclear arene-ruthenium(II) derivatives [RuCl₂(η ⁶-arene)(PTA-Me)] (arene = C₆H₆ (3a), p-cymene (3b), 1,3,5-C₆H₃Me₃ (3c), C₆Me ₆ (3d)), containing the ionic phosphine ligand 1-methyl-3,5-diaza-1- azonia-7-phosphaadamantane chloride (PTA-Me), have been synthesized and fully characterized. These complexes were evaluated as potential catalysts for the redox isomerization of allylic alcohols. Among them, best results in terms of activity were obtained with complex [RuCl₂(η⁶-C ₆H₆)(PTA-Me)] (3a) which, in combination with K ₂CO₃ (2.5 equiv. per Ru), was able to selectively isomerize a number of allylic alcohols RCH(OH)CHCH₂ (R = H, aryl, alkyl or heteroaryl group) into the corresponding carbonyl compounds RC(O)CH₂CH₃ in refluxing THF (TOF values up to 800 h ^-1). Complex [RuCl₂(η⁶-C₆H ₆)(PTA-Me)] (3a) was adsorbed onto the Montmorillonite K-10 clay, and the resulting solid proved also active in the isomerization of the model substrate 1-octen-3-ol. In addition, it could be easily separated from the reaction media by simple filtration and reused several times (up to 11 consecutive runs) with retention of its efficiency.

Full text of DOI:10.1016/j.molcata.2012.10.025

Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Catalytic isomerization of allylic alcohols promoted by complexes
[RuCl2(ꢀ6-arene)(PTA-Me)] under homogeneous conditions and
supported on Montmorillonite K-10
Lucía Menéndez-Rodríguez, Pascale Crochet∗, Victorio Cadierno∗
Laboratorio de Compuestos Organometálicos y Catálisis (Unidad Asociada al CSIC), Departamento de Química Orgánica e Inorgánica, Instituto Universitario de Química Organometálica
“Enrique Moles”, Universidad de Oviedo, Julián Clavería 8, 33006 Oviedo, Spain
a r t i c l e i n f o
a b s t r a c t
The mononuclear arene-ruthenium(II) derivatives [RuCl2(ꢀ6-arene)(PTA-Me)] (arene = C6H6 (3a),
p-cymene (3b), 1,3,5-C6H3Me3 (3c), C6Me6 (3d)), containing the ionic phosphine ligand 1-methyl-
3,5-diaza-1-azonia-7-phosphaadamantane chloride (PTA-Me), have been synthesized and fully
characterized. These complexes were evaluated as potential catalysts for the redox isomerization of allylic
alcohols. Among them, best results in terms of activity were obtained with complex [RuCl2(ꢀ6-C6H6)(PTA-
Me)] (3a) which, in combination with K2CO3 (2.5 equiv. per Ru), was able to selectively isomerize a
number of allylic alcohols RCH(OH)CH CH2 (R = H, aryl, alkyl or heteroaryl group) into the corresponding
carbonyl compounds RC( O)CH2CH3 in refluxing THF (TOF values up to 800 h−1). Complex [RuCl2(ꢀ6-
C6H6)(PTA-Me)] (3a) was adsorbed onto the Montmorillonite K-10 clay, and the resulting solid proved
also active in the isomerization of the model substrate 1-octen-3-ol. In addition, it could be easily sepa-
rated from the reaction media by simple filtration and reused several times (up to 11 consecutive runs)
with retention of its efficiency.
Article history:
Received 3 August 2012
Received in revised form 22 October 2012
Accepted 22 October 2012
Available online 1 November 2012
Keywords:
Ruthenium catalysts
Supported catalysts
Montmorillonite
Redox isomerizations
Allylic alcohols
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
of the naturally occurring pheromones muscone [3] and (+)-iso-
exo-brevicomin [4], the marine alkaloid (−)-brevisamide [5], the
The redox isomerization of allylic alcohols catalyzed by
transition-metal complexes has emerged in recent years as a use-
ful and straightforward synthetic route to carbonyl compounds
(Scheme 1, path a), which conveniently replaces the classical two-
step sequential oxidation/reduction reactions (Scheme 1, path b or
c). Besides step and atom economy, one advantage of the redox
isomerization is to avoid the use of toxic and/or expensive oxi-
dation and reduction reagents. Hence, considerable efforts have
been devoted to the development of efficient catalytic systems for
this process, with those based on ruthenium, rhodium and iridium
showing nowadays the best performances [1]. The transformation
is based on the well-known ability of transition-metals to assist the
migration of carbon-carbon double bonds, that is, the catalyst turns
the allylic alcohol into an enol which readily tautomerizes into the
corresponding carbonyl compound [2].
fragrance Florhydral® [6] and the antitumor agent (−)-FR182877
[7], all of them including a redox isomerization step. The interest
of the pharmaceutical industry for the large scale production of the
analgesic drugs hydromorphone and hydrocodone by catalytic iso-
merization of the naturally occurring opiates morphine and codeine
also deserves to be mentioned [8]. In addition, remarkable results
have been recently obtained in the development of asymmetric
versions of this reaction, excellent enantiomeric excesses being
achieved in some instances [3,9].
On the other hand, a very important aspect for large-scale
chemical processes is the availability of efficient methods for
catalyst separation and recycling [10]. In this context, most
efforts devoted to achieve recyclable systems for the catalytic
isomerization of allylic alcohols have focused on water-soluble
complexes, the use of water as solvent allowing the easy sep-
aration of the reaction products from the catalyst by simple
liquid–liquid extraction (or by decantation if a biphasic medium
is employed) [1]. In this way, the active species can be re-used
in further catalytic cycles. Best results in this field have been
obtained with the hydrosoluble arene-ruthenium(II) derivative
With the gradual availability of efficient and selective cata-
lysts, this relatively simple reaction is beginning to be applied in
multi-step syntheses of structurally elaborated substrates of high
added value. In this context, we must highlight the preparations
[RuCl2(ꢀ6-p-cymene){P(OCH2CH2NMe3)3}][SbF6]3, containing
a
phosphite-ammonium ligand, which could be used for at
least 10 consecutive runs without significant loss of activity
[11].
∗
Corresponding authors. Tel.: +34 985105076/985103453; fax: +34 985103446.
E-mail addresses: crochetpascale@uniovi.es (P. Crochet), vcm@uniovi.es
(V. Cadierno).
1381-1169/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2012.10.025

L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
391
Fig. 1. Structure of the ionic phosphine ligand PTA-Me (1) employed in this work.
[20], the catalytic activity of these complexes in the redox iso-
merization of allylic alcohols under homogeneous conditions, and
the heterogenization of the most active one, namely [RuCl2(ꢀ6-
C6H6)(PTA-Me)], onto the commercially available and inexpensive
Montmorillonite K-10 clay. As the reader will see, the resulting
solid proved also active in the isomerization process and showed
an excellent recyclability.
Scheme 1. The redox isomerization of allylic alcohols.
Immobilization on insoluble solids supports is also a com-
mon strategy to overcome the intrinsic separation problems
associated with homogeneous catalysts [10]. However, to date,
this method has been scarcely applied in the redox isomeriza-
tion of allylic alcohols. In this context, Majoral and co-workers
have described the grafting of the organometallic fragments
[RuCl2(ꢀ6-p-cymene)] and [RhCl(cod)] (cod = 1,5-cyclooctadiene)
on the surface of phosphine-ended dendrimers [12]. The result-
ing systems were able to isomerize selectively 1-octen-3-ol into
octan-3-one and could be recycled (up to four times), but their
activities were relatively low (TOF up to 12 h−1). PdCl2 impreg-
nated on different polymers was also employed to promote redox
isomerization processes [13]. However, the requirement of a hydro-
gen atmosphere resulted in low selectivities toward the desired
carbonyl compounds, due to the occurrence of hydrogenation and
hydrogenolysis reactions in competition [14]. A much more gen-
eral, selective, and efficient system (TOF up to 6400 h−1), consisting
of a rhodium complex immobilized onto a ligand–polymer sup-
port, has been recently reported by Bergens and co-workers [15].
Unfortunately, despite the remarkable efficiency showed by this
polymeric material, its recycling was not demonstrated.
In our group, we supported an arene-ruthenium(II) complex
on the surface of functionalized silica-coated ferrite nanoparti-
cles [16]. The resulting heterogeneous catalyst was able to convert
selectively and efficiently several allylic alcohols into the corre-
sponding aldehydes or ketones (TOF up to 253 h−1). In addition,
this magnetic material was easily separated with the aid of exter-
nal magnet and could be reused in three further catalytic runs.
However, the preparation of the support was rather tedious and
simpler alternatives are desirable. This fact, along with a recent
study by Martín-Matute and co-workers demonstrating the utility
of cationic Rh(I) complexes supported onto the mesoporous alu-
minosilicate AISBA-15 for this catalytic transformation (TOF up to
48 h−1; up to five consecutive runs after simple separation of the
catalyst by filtration) [17], prompted us to explore more readily
available inorganic solids as supports [18]. In particular, we directed
our attention to the smectite-type laminar clay Montmorillonite,
whose combination of cation exchange, intercalation and swelling
properties has been largely exploited for the immobilization of
metal catalysts in the last decades [19]. Ionic complexes are known
to be readily adsorbed on the surface of this clay through electro-
static interactions or, alternatively, they can be intercalated into its
interlaminar spacing, leading in both cases to robust heterogenized
systems.
2. Experimental
2.1. General information
All manipulations were performed under an atmosphere of
dry nitrogen using vacuum-line and standard Schlenk tech-
niques. Organic solvents were dried by standard methods and
distilled under nitrogen before use. All reagents were obtained
from commercial suppliers and used without further purifi-
cation, with the exception of compounds PTA-Me (1) [21],
[{RuCl(ꢁ-Cl)(ꢀ6-arene)}2] (2a–d) [22], 1-(2-naphthyl)-2-propen-
1-ol (4h) [23], 1-(3-methoxyphenyl)-2-propen-1-ol (4i) [24], 1-
(4-methoxyphenyl)-2-propen-1-ol (4j) [25], 1-(4-chlorophenyl)-
2-propen-1-ol (4k) [26], 1-(4-fluorophenyl)-2-propen-1-ol (4l)
[27], 1-(2-furyl)-2-propen-1-ol (4m) [26], (5R)-5,9-dimethyl-1,8-
decadien-3-ol (4n) [18b] and (E/Z)-5,9-dimethyl-1,4,8-decatrien-
3-ol (4o) [28], which were prepared by following the methods
reported in the literature. Infrared spectra were recorded on a
Perkin-Elmer 1720-XFT spectrometer. NMR spectra were recorded
on Bruker DPX300 or AV400 instruments. Chemical shifts are given
in ppm, relative to internal tetramethylsilane (1H and 13C), and
external 85% aqueous H3PO4 solutions (31P). DEPT experiments
have been carried out for all the compounds reported in this paper.
The elemental C, H, and N analyses and FAB mass spectra were
provided by Analytical Services of the University of Sevilla. ICP-
MS analyses and powder X-ray diffraction (XRD) patterns were
provided by the Scientific-Technical Services of the University
of Oviedo. GC measurements were made on a Hewlett-Packard
HP6890 equipment using a Supelco Beta-DexTM 120 (30 m length;
250 ␮m diameter) or HP-INNOWAX (30 m length; 250 ␮m diam-
eter) column. GC/MSD measurements were performed on Agilent
6890 N equipment coupled to a 5973 mass detector (70 eV electron
impact ionization) using a HP-1MS column.
2.2. Synthesis of complexes [RuCl2(ꢀ6-arene)(PTA-Me)]
(arene = C6H6 (3a), p-cymene (3b), 1,3,5-C6H3Me3 (3c), C6Me6
(3d))
A suspension of the corresponding dimeric precursor [{RuCl(ꢁ-
Cl)(ꢀ6-arene)}2] (2a–d) (0.5 mmol) in tetrahydrofuran (30 mL) was
treated with the phosphine ligand PTA-Me (0.250 g, 1.2 mmol) at
room temperature for 24 h. Concentration of the resulting solution
to ca. 10 mL followed by the addition of 50 mL of diethyl ether pre-
cipitated an orange solid, which was washed with diethyl ether (3×
10 mL) and dried in vacuo. (3a): Yield: 78% (0.357 g); Anal. Calcd
for RuC13H21Cl3N3P (457.73 g mol−1): C, 34.11; H, 4.62; N, 9.18%.
Found: C, 33.98; H, 4.73; N, 9.03%; IR (KBr) ꢂ = 422 (m), 579 (w), 569
Thus, herein we describe the preparation of mononuclear
arene-ruthenium(II) complexes [RuCl2(ꢀ6-arene)(PTA-Me)] con-
taining the ionic phosphine ligand 1-methyl-3,5-diaza-1-azonia-7-
phosphaadamantane chloride 1 (abbreviated as PTA-Me; see Fig. 1)

392
L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
(s), 743 (s), 812 (s), 834 (s), 897 (s), 929 (s), 982 (s), 1023 (s), 1087 (s),
1124 (s), 1251 (m), 1277 (m), 1301 (s), 1430 (s), 1451 (s), 1509 (w),
Acros Organics), by their fragmentation in GC/MS, and/or NMR
spectroscopy.
1
2984 (m), 3032 (m) cm−1
;
31P{ H} NMR (CD3OD) ı = −21.3 (s) ppm;
1H NMR (CD3OD) ı = 2.87 (d, 3H, 4JPH = 2.1 Hz, NMe), 4.26–5.15 (m,
2.4. General procedure for the support of complex
12H, PCH2N and NCH2N), 5.92 (d, 6H, JPH = 1.0 Hz, C6Me6) ppm;
[RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) in Montmorillonite K-10
3
13C{ H} NMR (CD3OD) ı = 48.2 (s, NMe), 48.4 (d, JPC = 19.6 Hz,
1
1
PCH2N), 55.9 (d, JPC = 15.8 Hz, PCH2N+), 69.1 (d, JPC = 6.0 Hz,
Prior to its use, Montmorillonite K-10 (Sigma–Aldrich) was dried
for 24 h at 110 ◦C. Under an inert atmosphere, the appropriate
amount of complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) (0.1, 0.2, 0.3,
0.4 or 0.5 mmol) was dissolved in 15 mL of dichloromethane. To the
resulting solution, 1 g of the clay was then added and the mixture
stirred at room temperature for 4–24 h. After this time, the suspen-
sion was filtered off and the solid washed with dichloromethane
(3× 15 mL) and vacuum-dried. The weight percentage of Ru in the
resulting heterogeneous catalysts MK10-3a-x was determined by
ICP-MS. Solids MK10-3a-x were further characterized by XRD and
IR spectroscopy (see Section 3).
1
3
NCH2N), 80.7 (d, JPC = 3.0 Hz, NCH2N+), 88.1 (d, JPC = 3.0 Hz,
C6H6) ppm; MS (FAB) m/z = 422.0 (100%) [M+]. (3b): Yield: 93%
(0.478 g); Anal. Calcd for RuC17H29Cl3N3P (513.83 g mol−1): C,
39.74; H, 5.69; N, 8.18%. Found: C, 39.66; H, 5.81; N, 8.05%; IR
(KBr) ꢂ = 481 (m), 553 (m), 575 (s), 633 (m), 675 (m), 752 (s),
772 (s), 816 (s), 849 (m), 874 (m), 898 (s), 928 (s), 989 (s), 1147
3
2
(w), 1255 (m), 1306 (s), 1366 (w), 1394 (m), 1420 (m), 1457 (m),
1
3515 (m) cm−1
;
31P{ H} NMR (CD3OD) ı = −22.5 (s) ppm; 1H NMR
(CD3OD) ı = 1.24 (d, 6H, 3JHH = 7.1 Hz, CHMe2), 2.07 (s, 3H, Me), 2.72
3
(sept, 1H, JHH = 7.1 Hz, CHMe2), 2.86 (s, 3H, NMe), 4.20–5.14 (m,
12H, PCH2N and NCH2N), 5.88 (br, 4H, CH of cym) ppm; 13C{ H}
1
2.5. General procedure for catalytic isomerization of 1-octen-3-ol
(4a) into octan-3-one (5a) using the heterogeneous catalysts
MK10-3a-x
NMR (CD3OD) ı = 15.9 (s, Me), 19.7 (s, CHMe2), 29.5 (s, CHMe2),
46.5 (s, NMe), 46.7 (d, 1JPC = 18.9 Hz, PCH2N), 54.0 (d, 1JPC = 15.9 Hz,
3
3
PCH2N+), 67.8 (d, JPC = 5.5 Hz, NCH2N), 79.0 (d, JPC = 3.3 Hz,
NCH2N+), 84.7 and 87.8 (d, JPC = 4.9 Hz, CH of cym), 96.7 and
2
In a sealed tube, under nitrogen atmosphere, the appropriate
amount of the corresponding heterogeneous catalyst MK10-3a-x
(0.5 mol% of Ru) and K2CO3 (6.9 mg, 5 mmol; 1.25 mol%) were added
to a solution of 1-octen-3-ol (4a) (0.618 mL, 4 mmol) in tetrahy-
drofuran (4 mL), and the resulting mixture stirred at 75 ◦C for the
indicated time (see Table 7). The course of the reaction was moni-
tored by taking regularly samples of ca. 10 ␮L which after dilution
with dichloromethane (3 mL) were analyzed by GC. The hetero-
geneous catalyst MK10-3a-x was then separated by filtration and
the reaction product isolated by solvent removal and chromato-
graphic work-up of the residue on silica-gel using a mixture of
EtOAc–hexane (1:10) as eluent. The identity of the resulting octan-
3-one (5a) was assessed by GC, by comparison of its retention time
with that of a commercially available pure sample.
106.3 (s, C of cym) ppm. (3c): Yield: 81% (0.405 g); Anal. Calcd
for RuC16H
27Cl3N3P (499.81 g mol−1): C, 38.45; H, 5.44; N, 8.41%.
Found: C, 38.35; H, 5.57; N, 8.31%; IR (KBr) ꢂ = 474 (m), 570 (s), 638
(m), 741 (s), 769 (m), 813 (s), 891 (m), 939 (s), 1007 (s), 1314 (s),
1377 (m), 1409 (m), 1451 (s), 1535 (m), 3343 (w), 3513 (w) cm−1
;
1
31P{ H} NMR (CD3OD) ı = −25.4 (s) ppm; 1H NMR (CD3OD) ı = 2.22
(s, 9H, Me), 2.80 (s, 3H, NMe), 4.19–4.70 (m, 6H, PCH2N), 4.90–5.10
1
(m, 6H, NCH2N), 5.46 (s, 3H, CH of mes) ppm; 13C{ H} NMR
1
(CD3OD) ı = 17.9 (s, Me), 47.8 (d, JPC = 18.9 Hz, PCH2N), 48.3 (s,
NMe), 55.8 (d, 1JPC = 15.8 Hz, PCH2N+), 69.2 (d, 3JPC = 5.3 Hz, NCH2N),
3
2
80.7 (d, JPC = 3.8 Hz, NCH2N+), 85.8 (d, JPC = 3.8 Hz, CH of mes),
103.3 (d, 2JPC = 2.3 Hz, C of mes) ppm; MS (FAB) m/z = 464.0 (100%)
[M+]. (3d): Yield: 77% (0.417 g); Anal. Calcd for RuC19H33Cl3N3P
(541.89 g mol−1): C, 42.11; H, 6.14; N, 7.75%. Found: C, 42.24;
H, 6.18; N, 7.69%; IR (KBr) ꢂ = 574 (m), 754 (m), 813 (m), 897
2.6. Recycling of the heterogeneous catalyst MK10-3a-2.5
(m), 930 (m), 982 (m), 1018 (m), 1092 (s), 1130 (m), 1245 (w),
1
1303 (m), 1389 (m), 1457 (m), 3396 (m), 3513 (w) cm−1
;
31P{ H}
NMR (CD3OD) ı = −25.4 (s) ppm; 1H NMR (CD3OD) ı = 2.01 (d, 18H,
In a sealed tube, under nitrogen atmosphere, the heterogeneous
catalyst MK10-3a-2.5 (0.2 g; 1.25 mol% of Ru) and K2CO3 (17.3 mg,
1.25 mmol; 2.5 equiv. per Ru) were added to a solution of 1-octen-
3-ol (4a) (0.618 mL, 4 mmol) in tetrahydrofuran (4 mL), and the
resulting mixture stirred at 75 ◦C until complete transformation of
4a into 5a (confirmed by GC analysis of aliquots). The reaction mix-
ture was then cooled to room temperature, the solid filtered off,
washed with hexanes (3× 5 mL) and vacuum-dried. To the solid,
THF (4 mL) and a new load of 1-octen-3-ol (4a) (0.618 mL, 4 mmol)
were then added, and the mixture heated again at 75 ◦C for the
indicated time (see Fig. 4).
4
4JPH = 1.1 Hz, C6Me6), 2.68 (d, 3H, JPH = 1.9 Hz, NMe), 4.01–4.85
1
(m, 12H, PCH2N and NCH2N) ppm; 13C{ H} NMR (CD3OD) ı = 15.0
1
(s, C6Me6), 47.0 (d, JPC = 12.1 Hz, PCH2N), 47.6 (s, NMe), 55.4
1
3
(d, JPC = 16.6 Hz, PCH2N+), 69.1 (d, JPC = 6.0 Hz, NCH2N), 80.6 (d,
3JPC = 3.0 Hz, NCH2N+), 97.6 (d, 2JPC = 1.5 Hz, C6Me6) ppm; MS (FAB)
m/z = 506.1 (100%) [M+].
2.3. General procedure for the catalytic isomerization of allylic
alcohols using complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a)
In
a sealed tube, under nitrogen atmosphere, the ruthe-
nium complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) (0.02–0.1 mmol;
0.5–2.5 mol% of Ru) and K2CO3 (0.05–0.25 mmol; 1.25–6.25 mol%)
were added to a solution of the corresponding allylic alcohol 4a–o
(4 mmol) in tetrahydrofuran (4 mL), and the resulting mixture
stirred at 75 ◦C for the indicated time (see Table 4 and Scheme 3).
The course of the reaction was monitored by taking regularly sam-
ples of ca. 10 ␮L which after dilution with dichloromethane (3 mL)
were analyzed by GC. After the reaction was finished, the mixture
was cooled to room temperature leading to the partial precipita-
tion of 3a. The solid was separated by filtration and the reaction
product isolated by solvent removal and chromatographic work-
up of the residue on silica-gel using a mixture of EtOAc–hexane
(1:10) as eluent. The identity of the resulting carbonyl compounds
5a–o was assessed by comparison of their retention times with
those of commercially available pure samples (Sigma–Aldrich or
3. Results and discussion
Our study began with the synthesis of the mononuclear ruthe-
nium(II) derivatives [RuCl2(ꢀ6-arene)(PTA-Me)] (arene = C6H6 (3a),
p-cymene (3b) [29], 1,3,5-C6H3Me3 (3c), C6Me6 (3d)). They were
generated in high yields (77–93%) through a classical chloride
bridges splitting reaction in the corresponding dimeric precur-
sors [{RuCl(ꢁ-Cl)(ꢀ6-arene)}2] (2a–d) [22] upon treatment with
2.4 equivalents of the ligand PTA-Me (1) in tetrahydrofuran, at
room temperature, for 24 h (Scheme 2). Compounds 3a–d, isolated
as air-stable orange solids, were fully characterized by means of
elemental analyses, IR, multinuclear NMR spectroscopy and mass
spectrometry (FAB technique), the data obtained being fully con-
sistent with the proposed structures (details are given in Section 2).
1
In particular, the 31P{ H} NMR spectra of these derivatives exhibit

L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
393
Scheme 2. Synthesis of the mononuclear Ru(II) complexes [RuCl2(ꢀ6-arene)(PTA-Me)] (3a–d).
a singlet resonance at ıP −21.5 to −25.4 ppm, strongly deshielded
with respect to that of the free phosphine PTA-Me (ıP = −84.6 ppm),
thus supporting the selective P-coordination of 1 to ruthenium [30].
appreciable differences in activity were observed in function of
the arene ligand present in the complex. The rate order observed
(C6H6 > p-cymene > 1,3,5-C6H3Me3 > C6Me6) indicates clearly that
higher performances are found for the less sterically demanding
and electron-rich arenes [31]. Thus, the highest rate was achieved
with the benzene derivative [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a), which
led to the quantitative formation of 5a after only 30 min of heat-
ing (TOF = 100 h−1; entry 1). At this point, we must note that,
when the same reaction was performed in the presence of free
benzene (50 equiv. per mol of Ru), the performance of [RuCl2(ꢀ6-
C6H6)(PTA-Me)] (3a) remained almost unaltered (entry 5 vs 1). This
fact confirms that release of the arene ligand does not take place
during the catalytic event, and that the required vacant sites for
coordination of substrate are generated by the expected dissocia-
tion of the chloride ligands from the metal. On the other hand, it
is also worthy of note that [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) retains
its high catalytic activity at low metal loadings. Thus, using only
0.5 mol% of this complex, quantitative formation of octan-3-one
was observed by GC after 35 min, allowing to improve the TOF of
the process up to 343 h−1 (entry 6).
1
Their 1H and 13C{ H} NMR spectra also show the expected signals
for the ꢀ6-coordinated arene groups and the PTA-Me ligand. For the
latter, characteristic N-Me singlet resonances were observed at ıH
3
2.68–2.87 (a doublet with JPH = 1.9–2.1 Hz for 3a and 3d) and ıC
46.5–48.3 ppm.
The ability of complexes [RuCl2(ꢀ6-arene)(PTA-Me)] (3a–d) to
promote the redox isomerization of allylic alcohols in homoge-
neous phase was evaluated employing 1-octen-3-ol (4a) as model
substrate. Since this catalytic reaction usually requires of a base
[1], our initial efforts focused on finding the most appropriate one
for carrying out the process. To this end, we studied the isomer-
ization of 4a with complex [RuCl2(ꢀ6-p-cymene)(PTA-Me)] (3b) in
the presence of different alkali metal carbonates, hydroxides and
tert-butoxides. Experiments were carried out at 75 ◦C employing
4 mmol of 1-octen-3-ol, 2 mol% of complex 3b, tetrahydrofuran as
the solvent (1 M solutions of the allylic alcohol) and 5 mol% of the
corresponding base (2.5 equiv. per Ru). The results obtained from
this initial screening are collected in Table 1.
The isomerization of 4a into 5a by means of complex [RuCl2(ꢀ6-
C6H6)(PTA-Me)] (3a) was also studied in other organic solvents
(1,4-dioxane, toluene, 1,2-dichloroethane, methanol and ethanol),
as well as in water. The results obtained are collected in Table 3.
Thus, although quantitative conversions were in all cases reached
using a metal loading of 0.5 mol%, none of these solvents allowed to
improve the result previously obtained in THF (entry 3). Apparently,
the performance of [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) is optimal in
aprotic media of intermediate polarity (entries 3–4). Some experi-
ments were also performed in THF at different concentrations of the
substrate (from 0.1 M to 2 M), but no marked differences in activ-
ity and selectivity were observed. In contrast, remarkably poorer
results were obtained lowering the reaction temperature (e.g. at
55 ◦C only 60% of conversion of 1-octen-3-ol was observed after 4 h
of heating).
Noteworthy, the reaction profile of this isomerization process
is characterized by an induction period of several minutes (see
Fig. 2), indicating the slow formation of the real active species
in the catalytic cycle. Moreover, deactivation of the active species
does not occur after total consumption of the substrate 1-octen-
3-ol (4a). Thus, after a regular catalytic test, a faster reaction was
observed upon addition of a second charge of 4a (quantitative for-
mation of 5a after only 20 min), with the sigmoidal shape of the
graph suggesting that generation of the active species still contin-
ues during the reaction. As the formation of metal(0) nanoparticles
is known to be particularly favored in basic alcoholic media [32],
these observations raised the question on the real homogeneous
or heterogeneous nature of the catalytic reaction. To shed light
on this point, the isomerization of 4a was performed in the pres-
ence of mercury [33] and no major difference in activity was
As expected, performing the catalytic reaction in the absence
of base did not yield appreciable amounts of the desired octan-3-
one (5a) after 2 h of heating (entry 1), thus confirming the need of
a basic media. Deprotonation of the hydroxyl group of the allylic
alcohol, thereby enhancing its coordinating ability, is a requisite
commonly proposed in the literature for this reaction [1,2]. As
shown in Table 1, the introduction of a base in the medium allows
the isomerization process to proceed, with those containing the
bulkier and less polarizing cations K+ and Cs+ being the most effec-
tive. In particular, the best results were obtained with K2CO3 (entry
4) and CsOH·H2O (entry 9), which allowed the selective and quan-
titative conversion of 1-octen-3-ol (4a) into octan-3-one (5a) after
only 45 min of heating (TOF = 67 h−1). Based on these results, inex-
pensive K2CO3 was chosen as the base for the rest of our studies.
Concerning the [Ru]:[K2CO3] ratio, we have found that the isomer-
ization reaction of 4a also proceeds when equimolar amounts of
complex 3b and K2CO3 are employed, but a longer time (3 h) is in
this case needed to completion (entry 12). As shown in entry 13,
increasing the amount of K2CO3 to 4 equiv. per Ru resulted coun-
terproductive, due probably to the partial decomposition of the
active species under these strongly basic conditions. Consequently,
a [Ru]:[K2CO3] ratio of 1:2.5 was adopted for the rest of our studies.
Using this optimal [Ru]:[K2CO3] ratio, the catalytic activity of
the four complexes synthesized, i.e. [RuCl2(ꢀ6-arene)(PTA-Me)]
(arene = C6H6 (3a), p-cymene (3b), 1,3,5-C6H3Me3 (3c), C6Me6
(3d)), in the isomerization of the model alcohol 1-octen-3-ol (4a)
was compared. As shown in Table 2, all of them were found to
be active and selective catalysts in the isomerization process, pro-
viding octan-3-one (5a) as the unique reaction product. However,

394
L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
Table 1
Isomerization of 1-octen-3-ol (4a) into octan-3-one (5a) catalyzed by complex [RuCl2(ꢀ6-p-cymene)(PTA-Me)] (3b).a
c
Entry
Base
Time
Yield (%)b
TOF (h−1
)
1
2
3
4
5
6
7
8
9
10
11
12d
13e
–
2 h
2 h
2 h
45 min
1.25 h
2 h
2 h
1 h
45 min
2 h
2 h
3 h
3 h
0
5
15
100
100
4
0
1
4
67
40
1
15
50
67
1
Li2CO3
Na2CO3
K2CO3
Cs2CO3
LiOH·H2O
NaOH
60
KOH
100
100
5
100
100
40
CsOH·H2O
NaOtBu
KOtBu
K2CO3
K2CO3
25
17
7
a
Reactions performed under N2 atmosphere at 75 ◦C using 4 mmol of 1-octen-3-ol (4a) (1 M in THF). [1-octen-3-ol]:[Ru]:[base] ratio = 100:2:5.
Yield of octan-3-one (5a) determined by GC (uncorrected GC areas).
Turnover frequencies ((mol product/mol Ru)/time) were calculated at the time indicated in each case.
Reaction performed employing a [1-octen-3-ol]:[Ru]:[K2CO3] ratio of 100:2:2.
b
c
d
e
Reaction performed employing a [1-octen-3-ol]:[Ru]:[K2CO3] ratio of 100:2:8.
Table 2
Isomerization of 1-octen-3-ol (4a) into octan-3-one (5a) catalyzed by complexes [RuCl2(ꢀ6-arene)(PTA-Me)] (3a–d).a
c
Entry
Catalyst
Time
Yield (%)b
TOF (h−1
)
1
2
3
[RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a)
[RuCl2(ꢀ6-p-cymene)(PTA-Me)] (3b)
[RuCl2(ꢀ6-1,3,5-C6H3Me3)(PTA-Me)] (3c)
[RuCl2(ꢀ6-C6Me6)(PTA-Me)] (3d)
[RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a)
[RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a)
30 min
45 min
1 h
2.5 h
35 min
35 min
100
100
100
75
100
100
100
67
50
15
86
4
5d
6e
343
a
Reactions performed under N2 atmosphere at 75 ◦C using 4 mmol of 1-octen-3-ol (4a) (1 M in THF). [1-octen-3-ol]:[Ru]:[K2CO3] ratio = 100:2:5.
Yield of octan-3-one (5a) determined by GC (uncorrected GC areas).
Turnover frequencies ((mol product/mol Ru)/time) were calculated at the time indicated in each case.
Reaction performed in the presence of 50 equiv. per Ru of free benzene.
b
c
d
e
Reaction performed employing a [1-octen-3-ol]:[Ru]:[K2CO3] ratio of 100:0.5:1.25.
Table 3
Isomerization of 1-octen-3-ol (4a) into octan-3-one (5a) catalyzed by complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) in different solvents.a
d
Entry
Solvent
εb
2.3
2.4
7.6
10.4
24.6
32.7
80.1
Time
Yield (%)c
TOF (h−1
)
1
2
3
4
5
6
7
1,4-Dioxane
Toluene
THF
1,2-Dichloroethane
Ethanol
Methanol
Water
2 h
100
100
100
100
100
100
100
100
80
2.5 h
35 min
45 min
1 h
1 h
24 h
343
267
200
200
8
a
Reactions performed under N2 atmosphere at 75 ◦C using 4 mmol of 1-octen-3-ol (4a) (1 M in the appropriate solvent). [1-octen-3-ol]:[Ru]:[K2CO3] ratio = 100:0.5:1.25.
Dielectric constant of the solvent at 25 ◦C.
Yield of octan-3-one (5a) determined by GC (uncorrected GC areas).
b
c
d
Turnover frequencies ((mol product/mol Ru)/time) were calculated at the time indicated in each case.

L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
395
Table 4
Isomerization of different allylic alcohols catalyzed by complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a).a
c
Entry
Substrate
mol% of 3a
Time
Yield (%)b
TOF (h−1
)
1
2
3
4
5
6
7
8
9
10
11
12
13
R = nPent (4a)
R = nBu(4b)
0.5
0.5
0.5
0.5
0.5
2.5
0.5
0.5
0.5
0.5
0.5
0.5
2.5
35 min
1 h
1 h
45 min
15 min
4.5 h
20 h
22 h
20 h
20 h
20 h
100 (91)
100 (90)
93 (82)
100 (89)
100 (88)
78 (65)
100 (92)
98 (87)
88 (75)
100 (93)
92 (80)
100 (91)
82 (73)
343
200
186
267
800
7
10
9
9
10
9
10
1
R = nPr (4c)
R = Et (4d)
R = Me (4e)
R = H (4f)
R = Ph (4g)
R = 2-Naphthyl (4h)
R = 3-C6H4OMe (4i)
R = 4-C6H4OMe (4j)
R = 4-C6H4Cl (4k)
R = 4-C6H4F (4l)
R = 2-Furyl (4m)
20 h
36 h
a
Reactions performed under N2 atmosphere at 75 ◦C using 4 mmol of the corresponding allylic alcohol 4a–m (1 M in THF) and the quantity of complex 3a indicated in
each case. [K2CO3]:[Ru] ratio = 2.5:1.
b
Yield of the resulting carbonyl compounds 5a–m determined by GC (uncorrected GC areas). Isolated yields after are given in brackets.
Turnover frequencies ((mol product/mol Ru)/time) were calculated at the time indicated in each case.
c
observed. This fact discards a heterogeneous process involving
metallic nanoparticles as the active species. On the other hand,
a diminution of the induction period was observed when com-
plex 3a was preheated with K2CO3 prior to the addition of the
allylic alcohol (see Fig. 2), suggesting that the dichloride complex
[RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) may be initially transformed into
the carbonate one [Ru(CO3)(ꢀ6-C6H6)(PTA-Me)] [34]. A faster sub-
stitution of the carbonate ligand by the substrate, in comparison
with that of the chlorides in 3a, could then explain the reduction of
the induction period.
To further define the scope of this catalytic transformation,
other allylic alcohols were subjected to the combined action of
[RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) and K2CO3 (Table 4). Reactions
were performed in all cases in THF (1 M solutions) at 75 ◦C using a
[K2CO3]:[Ru] ratio of 2.5:1. Thus, as observed for 1-octen-3-ol (4a),
related aliphatic substrates 4b–e could be efficiently converted into
the corresponding ketones 5b–e within 1 h of heating (entries 2–5),
leading to a maximum TOF value of 800 h−1 for the isomerization
of 3-buten-2-ol (4e) into butan-2-one (5e) (entry 5). Allylic alcohol
(4f) itself also participated in the reaction, albeit a longer reaction
time (4.5 h) and a higher metal loading (2.5 instead of 0.5 mol%
of Ru) were in this case needed to generate propionaldehyde (5f)
in high yield (entry 6). This result is not surprising since a lower
reactivity of primary vs secondary alcohols has been previously
observed with other catalytic systems [35]. The gradual decompo-
sition of the catalyst into an inactive metal-carbonyl compound by
decarbonylation of the aldehyde may be responsible of this effect
[36].
As shown in entries 7–12, complex 3a (0.5 mol%) was also effec-
tive in the isomerization of the aromatic substrates 4g–l, thus
confirming the generality of this catalytic transformation. However,
due probably to the steric congestion associated with the presence
an aromatic ring in ␣-position to the alcohol unit that disfavors
their coordination to the metal, longer reaction times (20–22 h)
were in these cases required to attain good conversions [37]. Using
a ruthenium loading of 2.5 mol%, isomerization of the heteroaro-
matic substrate 1-(2-furyl)-2-propen-1-ol (4m) was also possible,
leading to a maximum conversion of 82% after 36 h of heating. The
lower reactivity of this substrate is probably associated to the com-
petitive coordination of the furyl and olefin units onto the active
ruthenium species.
Furthermore, [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) was also success-
fully employed in the chemoselective isomerization of allylic
alcohols 4n (1:1 mixture of diastereoisomers) and 4o (1:1 mixture
of E/Z stereoisomers) derived from the naturally occurring alde-
hydes (R)-(+)-citronellal and citral, respectively, into the known
ketones 5n–o [18b,38] (Scheme 3). Interestingly, although 4o
presents two carbon-carbon double bonds in ␣-position with
respect to the alcohol group, only the isomerization of the mono-
substituted one was observed, leading to the exclusive formation of
the enone 5o as a 1:1 mixture of the corresponding E and Z isomers.
In both cases the Me2C CHCH2 unit remained unaltered.
Fig. 2. Isomerization of 1-octen-3-ol (4a) into octan-3-one (5a) catalyzed by complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) under optimized conditions (1 M solution in THF, at
75 ◦C using a [4a]:[Ru]:[K2CO3] ratio of 100:0.5:1.25): yield of octan-3-one (5a) as a function of time.

396
L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
Scheme 3. Isomerization of allylic alcohols 4n–o by complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a).
Table 5
Table 6
Ruthenium content in the solids.
Powder XRD data for Montmorillonite K-10 and solids MK10-3a-x.
Entry
Ru initiala
% Ru (exp.)b
% Ru (theor.)c
MK10-3a-xd
Entry
Solid
2Âa
db
8.8874◦
8.9658◦
8.9554◦
8.9381◦
8.9562◦
8.9220◦
9.950 A
˚
˚
1
2
3
4
5
0.1 mmol
0.2 mmol
0.3 mmol
0.4 mmol
0.5 mmol
0.7%
1.5%
2.0%
2.5%
3.1%
1.0%
1.9%
2.7%
3.4%
4.1%
MK10-3a-0.7
MK10-3a-1.5
MK10-3a-2.0
MK10-3a-2.5
MK10-3a-3.1
1
2
3
4
5
6
Montmorillonite K-10
MK10-3a-0.7
MK10-3a-1.5
MK10-3a-2.0
MK10-3a-2.5
9.863 A
˚
9.875 A
˚
9.894 A
˚
9.888 A
˚
MK10-3a-3.1
9.912 A
a
Quantity of complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) employed for the impreg-
a
nation of 1 g of Montmorillonite K-10.
The (0 0 1) diffraction line.
Interlaminar (basal) distance.
b
b
Weight percentage of ruthenium in the resulting solids determined by ICP-MS.
Maximum weight percentage of ruthenium that can be incorporated in the solid
c
(calculated from the quantity of 3a employed in each case).
d
Name given to the solid (x refers to the amount of ruthenium incorporated).
24 h of stirring. This fact suggests that saturation of the clay takes
place with this quantity of the complex. The suspensions gener-
ated in the impregnation process were then filtered off and the
resulting solids washed with dichloromethane and dried under
vacuum.
The IR spectra obtained for these solids confirmed the incorpo-
ration of complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) to the clay since,
in addition to the strong absorption bands characteristics of the
Montmorillonite K-10 support, others of lower intensity at 1437
and 749 cm−1 attributable to 3a were also observed. The weight
percentage of ruthenium present in the solids was determined by
ICP-MS measurements (see Table 5). As expected, the ruthenium
content increased with the quantity of complex 3a initially present
in solution. Thus, the weight percentage of Ru varied from 0.7%
(starting from 0.1 mmol of 3a; solid MK10-3a-0.7 in entry 1) to
Once demonstrated the synthetic potential of [RuCl2(ꢀ6-
C6H6)(PTA-Me)] (3a), we proceeded to support this complex onto
the commercial clay Montmorillonite K-10 (Sigma–Aldrich). The
immobilization process was carried out by the classical solvent-
impregnation method [39]. Thus, colored dichloromethane (15 mL)
solutions containing 0.1, 0.2, 0.3, 0.4 and 0.5 mmol of [RuCl2(ꢀ6-
C6H6)(PTA-Me)] (3a) were prepared under nitrogen atmosphere
and treated with 1 g of the clay, previously dried at 110 ◦C to elim-
inate adsorbed water, for 4 h. Almost complete discoloration of the
solutions was observed when 0.1–0.4 mmol of 3a were employed.
In contrast, starting from 0.5 mmol of 3a, the loss of the initial
orange color of the solution was much less pronounced even after
Table 7
Isomerization of 1-octen-3-ol (4a) into octan-3-one (5a) catalyzed by the heterogeneous systems MK10-3a-x.a
c
Entry
Catalyst
Time (h)
Yield (%)b
TOF (h−1
)
1
2
3
4
5
MK10-3a-0.7
MK10-3a-1.5
MK10-3a-2.0
MK10-3a-2.5
MK10-3a-3.1
8
5
4
3
9
97 (89)
100 (95)
100 (96)
100 (95)
76 (60)
24
40
50
67
17
a
Reactions performed under N2 atmosphere at 75 ◦C using 4 mmol of 1-octen-3-ol (4a) (1 M in THF). [1-octen-3-ol]:[Ru]:[K2CO3] ratio = 100:0.5:1.25.
Yield of octan-3-one (5a) determined by GC (uncorrected GC areas). Isolated yields are given in brackets.
Turnover frequencies ((mol product/mol Ru)/time) were calculated at the time indicated in each case.
b
c

L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
397
3.1% (starting from 0.5 mmol of 3a; solid MK10-3a-3.1 in entry
5). However, we must note that, in all cases, the actual content
in ruthenium is slightly lower than the maximum value calcu-
lated from the amount of 3a employed [40], indicating that part
of the added complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) has not been
attached to the solid support, well because it was not adsorbed by
the clay initially, or by a problem of leaching during the washings
with dichloromethane.
Solids MK10-3a-x were further characterized by powder X-ray
diffraction, the diffractograms obtained showing in all the cases
an insignificant shift in the (0 0 1) diffraction line with respect to
that of the pure clay (see Table 6). For illustrative purposes, XRD
patterns of samples of the Montmorillonite K-10 employed in this
study and the solid MK10-3a-0.7 are shown in Fig. 3. The small
variations of the basal distances with respect to the pure clay,
calculated from the (0 0 1) diffraction line by the Bragg equation
(Table 6), suggest that complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) is
mainly adsorbed on the external surface of the clay through elec-
trostatic or H-bonding interactions [41], rather than intercalated
into its interlaminar spacing [42].
Once characterized, the catalytic activity of MK10-3a-x in
the redox isomerization of 1-octen-3-ol (4a) was then explored
under the optimized conditions described above for [RuCl2(ꢀ6-
C6H6)(PTA-Me)] (3a) under homogeneous conditions (i.e. 1 M
solution of 1-octen-3-ol in THF, a ruthenium loading of 0.5 mol%,
1.25 mol% of K2CO3 as the base, and a working temperature of
75 ◦C). The amount of solids MK10-3a-x needed to fit with the
metal loading of 0.5 mol% was calculated according to the ICP-MS
measurements collected in Table 5.
As shown in Table 7, all of the heterogeneous systems pre-
pared were able to promote the selective isomerization of 4a into
5a in high yields. However, they were comparatively less effec-
tive than the unsupported complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a)
(TOF values of 17–67 vs 343 h−1). Thus, under heterogeneous con-
ditions, the best results were found with MK10-3a-2.5, which was
able to generate quantitatively the desired octan-3-one (5a) after
3 h of heating (entry 4). Remarkably, the TOF value attained with
this system (67 h−1) was slightly higher than that reported by
Martín-Matute and co-workers for the isomerization of 4a with a
cationic Rh(I) complex anchored onto the mesoporous aluminosili-
cate AISBA-15 in THF at 80 ◦C (48 h−1) [17]. We must also note that,
Fig. 3. Powder XRD patterns of Montmorillonite K-10 pure (top) and solid MK10-
3a-0.7 (bottom).
as a general trend, an increase of the catalytic activity of the solids
MK10-3a-x with the weight percentage of ruthenium present in
the sample was observed (see entries 1–4). However, with MK10-
3a-3.1, which features the largest content of ruthenium, this trend
failed (entry 5 vs entries 1–4). Steric hindrance as a consequence
of the proximity between the metal centers on the surface of the
clay may be responsible for this anomaly since, as commented pre-
viously, saturation of this solid was observed during impregnation
with complex 3a.
Fig. 4. Reuse of the heterogeneous system MK10-3a-2.5 in the isomerization of 1-octen-3-ol (4a) into octan-3-one (5a).

398
L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
Finally, the recycling of the most active system MK10-3a-2.5
(g) A. Varela-Álvarez, J.A. Sordo, E. Piedra, N. Nebra, V. Cadierno, J. Gimeno,
Chem. Eur. J. 17 (2011) 10583–10599;
was investigated, using again the isomerization of 1-octen-3-ol (4a)
into octan-3-one (5a) as model reaction. As shown in Fig. 4, per-
forming the catalytic process with a ruthenium loading of 1.25 mol%
[43], MK10-3a-2.5 could be effectively reused in at least eleven con-
secutive runs after a simple filtration and washing of the solid with
hexanes. In complete accord with the induction period observed
for [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a) under homogeneous conditions
(see Fig. 2), conversion of 4a into 5a proceeded faster in the second
run than in the first one (1 vs 2 h). From there, the performance of
MK10-3a-2.5 remained stable without appreciable loss of activity
until the tenth recycling cycle. In the eleventh one, more than 90%
conversion was still reached after 4 h of heating, thus leading to
a cumulative TON of 872. We also determined the Ru content in
solution after the first reaction cycle by ICP-MS, and found that
it was only ca. 28 ppm. The excellent recyclability of MK10-3a-
2.5, superior to that shown by the related heterogeneous system
Rh(I)/AISBA-15 described by Martín-Matute and co-workers (up
to 5 consecutive cycles) [17], along with the low metal leaching
observed attest for the high stability and lifetime of our Ru-based
supported catalyst.
(h) L. Bellarosa, J. Díez, J. Gimeno, A. Lledós, F.J. Suárez, G. Ujaque, C. Vicent,
Chem. Eur. J. 18 (2012) 7749–7765.
[3] M. Ito, S. Kitahara, T. Ikariya, J. Am. Chem. Soc. 127 (2005) 6172–6173.
[4] A. Bouziane, T. Régnier, F. Carreaux, B. Carboni, C. Bruneau, J.-L. Renaud, Synlett
(2010) 207–210.
[5] G. Sabitha, S. Nayak, M. Bhikshapathi, J.S. Yadav, Org. Lett. 13 (2011) 382–385.
[6] S. Bovo, A. Scrivanti, M. Bertoldini, V. Beghetto, U. Metteoli, Synthesis (2008)
2547–2550.
[7] N. Tanaka, T. Suzuki, T. Matsumara, Y. Hosoya, M. Nakada, Angew. Chem. Int.
Ed. 48 (2009) 2580–2583.
[8] A.E. Díaz-Álvarez, V. Cadierno, Recent Pat. Catal. 1 (2012) 43–50, and references
cited therein.
[9] See, for example:
(a) K. Tanaka, G.C. Fu, J. Org. Chem. 66 (2001) 8177–8186;
(b) L. Mantilli, D. Gérard, S. Torche, C. Besnard, C. Mazet, Angew. Chem. Int. Ed.
48 (2009) 5143–5147;
(c) L. Mantilli, D. Gérard, S. Torche, C. Besnard, C. Mazet, Chem. Eur. J. 16 (2010)
12736–12745;
(d) M.A. Fernández-Zumel, B. Lastra-Barreira, M. Scheele, J. Díez, P. Crochet, J.
Gimeno, Dalton Trans. 39 (2010) 7780–7785;
(e) L. Mantilli, D. Gérard, S. Torche, C. Besnard, C. Mazet, Pure Appl. Chem. 82
(2010) 1461–1469;
(f) A. Quintard, A. Alexakis, C. Mazet, Angew. Chem. Int. Ed. 50 (2011)
2354–2358;
(g) R. Wu, M.G. Beauchamps, J.M. Laquidara, J.R. Sowa Jr., Angew. Chem. Int. Ed.
51 (2012) 2106–2110;
(h) V. Bizet, X. Pannecoucke, J.-L. Renaud, D. Cahard, Angew. Chem. Int. Ed. 51
(2012) 6467–6470.
[10] M. Benaglia (Ed.), Recoverable and Recyclable Catalysts, Wiley, Chichester,
2009.
4. Conclusion
In summary, a new catalytic system for the selective iso-
merization of allylic alcohols into carbonyl compounds under
homogeneous conditions has been developed by the aid of the
ruthenium(II) complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a). This ionic
species was successfully immobilized onto the smectite-type clay
Montmorillonite K-10 and used as a heterogeneous catalyst in
the redox isomerization process. Remarkably, although the activ-
ity of the heterogeneous system was comparatively lower than
that of the unsupported complex [RuCl2(ꢀ6-C6H6)(PTA-Me)] (3a),
it could be reused for a large number of runs, with negligible loss
of the catalytic activity. Overall, our results, along with those pre-
viously reported by Martín-Matute and co-workers [17], clearly
demonstrate the great potential of inorganic solids supports for the
development of robust heterogeneous catalysts able to promote the
synthetically useful isomerization reaction of allylic alcohols.
[11] P. Crochet, J. Díez, M.A. Fernández-Zúmel, J. Gimeno, Adv. Synth. Catal. 348
(2006) 93–100.
[12] (a) P. Servin, R. Laurent, L. Gonsalvi, M. Tristany, M. Peruzzini, J.-P. Majoral, A.-
M. Caminade, Dalton Trans. (2009) 4432–4434;
(b) P. Servin, R. Laurent, H. Dib, L. Gonsalvi, M. Peruzzini, J.-P. Majoral, A.-M.
Caminade, Tetrahedron Lett. 53 (2012) 3876–3879.
[13] (a) A.K. Zharmagambetova, E.E. Ergozhin, Y.L. Sheludyakov, S.G.
Mukhamedzhanova, I.A. Kurmanbayeva, B.A. Selenova, B.A. Utkelov, J.
Mol. Catal. A: Chem. 177 (2001) 165–170;
(b) M.G. Musolino, P. De Maio, A. Donato, R. Pietropaolo, J. Mol. Catal. A: Chem.
208 (2004) 219–224.
[14] Low selectivities were also attained employing Pd or Pd/Au nanoparticles sta-
bilized by alkanethiolate ligands under H2 atmosphere:
(a) E. Sadeghmoghaddam, K. Gaïeb, Y.-S. Shon, Appl. Catal. A: Gen. 405 (2011)
137–141;
(b) E. Sadeghmoghaddam, C. Lam, D. Choi, Y.-S. Shon, J. Mater. Chem. 21 (2011)
307–312;
(c) E. Sadeghmoghaddam, H. Gu, Y.-S. Shon, ACS Catal. 2 (2012) 1838–1845.
[15] E.G. Corkum, S. Kalapugama, M.J. Hass, S.H. Bergens, RSC Adv.
3473–3476.
2 (2012)
[16] S.E. García-Garrido, J. Francos, V. Cadierno, J.-M. Basset, V. Polshettiwar, Chem-
SusChem 4 (2011) 104–111.
Acknowledgments
[17] S. Sahoo, H. Lundberg, M. Edén, N. Ahlsten, W. Wan, X. Zou, B. Martín-Matute,
ChemCatChem 4 (2012) 243–250.
[18] (a) Ruthenium complexes supported onto amorphous SiO2, showing only a
modest catalytic activity, have also been described: C.M. Standfest-Hauser, T.
Lummerstorfer, R. Schmid, K. Kirchner, H. Hoffmann, M. Puchberger, Monatsh.
Chem. 134 (2003) 1167–1175;
Financial support by the Ministerio de Economía y Competitivi-
dad of Spain (Projects CTQ2010-14796/BQU and CSD2007-00006)
is gratefully acknowledged. L.M.-R. also thanks the Spanish Govern-
ment and the European Social Fund for the award of a PhD grant
(FPI program).
(b) Ru(OH)x supported onto Al2O3 was also employed as a recyclable system in
the reduction of allylic alcohols through a tandem redox isomerization/transfer
hydrogenation reaction: J.W. Kim, T. Koike, M. Kotani, Y. Yamaguchi, N. Mizuno,
Chem. Eur. J. 14 (2008) 4104–4109.
References
[19] For representative recent examples, see:
(a) B. Vijayakumar, G.R. Rao, J. Porous Mater. 19 (2012) 491–497;
(b) A. Romero-Pérez, A. Infantes-Molina, A. Jiménez-López, E.R. Jalil, K. Sapag,
E. Rodríguez-Castellón, Catal. Today 187 (2012) 88–96;
(c) B. Vijayakumar, G.R. Rao, J. Porous Mater. 19 (2012) 233–242;
(d) M. Farias, M. Martinelli, K. Guilherme, Appl. Catal. A: Gen. 40 (2011)
119–127;
[1] For reviews and accounts on this catalytic transformation, see:
(a) R.C. van der Drift, E. Bouwman, E. Drent, J. Organomet. Chem. 650 (2002)
1–24;
(b) R. Uma, C. Crévisy, R. Grée, Chem. Rev. 103 (2003) 27–51;
(c) V. Cadierno, P. Crochet, J. Gimeno, Synlett (2008) 1105–1124;
(d) L. Mantilli, C. Mazet, Chem. Lett. 40 (2011) 341–344;
(e) N. Ahlsten, A. Bartoszewicz, B. Martín-Matute, Dalton Trans. 41 (2012)
1660–1670;
(e) R. Takagi, N. Igata, K. Yamamoto, S. Kojima, J. Mol. Catal. A: Chem. 321 (2010)
71–76.
[20] For reviews on the coordination chemistry of the PTA ligand (1,3,5-triaza-7-
phosphaadamantane) and its derivatives, see:
(f) P. Lorenzo-Luis, A. Romerosa, M. Serrano-Ruiz, ACS Catal.
1079–1086;
2 (2012)
(a) A.D. Phillips, L. Gonsalvi, A. Romerosa, F. Vizza, M. Peruzzini, Coord. Chem.
Rev. 248 (2004) 955–993;
(b) J. Bravo, S. Bolan˜o, L. Gonsalvi, M. Peruzzini, Coord. Chem. Rev. 254 (2010)
555–607.
(g) J. García-Álvarez, S.E. García-Garrido, P. Crochet, V. Cadierno, Curr. Top.
Catal. 10 (2012) 35–56.
[2] For mechanistic studies, see:
(a) B.M. Trost, R.J. Kulawiec, J. Am. Chem. Soc. 115 (1993) 2027–2036;
(b) D.V. McGrath, R.H. Grubbs, Organometallics 13 (1994) 224–235;
(c) V. Branchadell, C. Crévisy, R. Grée, Chem. Eur. J. 9 (2003) 2062–2067;
(d) V. Cadierno, S.E. García-Garrido, J. Gimeno, A. Varela-Álvarez, J.A. Sordo, J.
Am. Chem. Soc. 128 (2006) 1360–1370;
(e) N. Ahlsten, B. Martín-Matute, Adv. Synth. Catal. 351 (2009) 2657–2666;
(f) M. Batuecas, M.A. Esteruelas, C. García-Yebra, E. On˜ate, Organometallics 29
(2010) 2166–2175;
[21] F.P. Pruchnik, P. Smolenski, Appl. Organomet. Chem. 13 (1999) 829–836.
[22] (a) M.A. Bennett, A.K. Smith, J. Chem. Soc. Dalton Trans. (1974) 233–241;
(b) M.A. Bennett, T.-N. Huang, T.W. Matheson, A.K. Smith, Inorg. Synth. 21
(1982) 74–78;
(c) J.W. Hull, W.L. Gladfelter, Organometallics 3 (1984) 605–613.
[23] A. Briot, C. Baehr, R. Brouillard, A. Wagner, C. Mioskowski, J. Org. Chem. 69
(2004) 1374–1377.

L. Menéndez-Rodríguez et al. / Journal of Molecular Catalysis A: Chemical 366 (2013) 390–399
399
[24] B. Schmidt, J. Org. Chem. 69 (2004) 7672–7678.
[25] S. Laabs, W. Münch, J.W. Bats, U. Nubbemeyer, Tetrahedron 58 (2002)
1317–1334.
(a) P. Crochet, M.A. Fernández-Zúmel, J. Gimeno, M. Scheele, Organometallics
25 (2006) 4846–4849;
(b) A.E. Díaz-Álvarez, P. Crochet, M. Zablocka, C. Duhayon, V. Cadierno, J.
Gimeno, J.P. Majoral, Adv. Synth. Catal. 348 (2006) 1671–1679;
(c) V. Cadierno, J. Francos, J. Gimeno, N. Nebra, Chem. Commun. (2007)
2536–2538;
[26] R.C.D. Brown, M.L. Fisher, L.J. Brown, Org. Biomol. Chem. 1 (2003) 2699–2709.
[27] N. Marion, R. Gealageas, S.P. Nolan, Org. Lett. 9 (2007) 2653–2656.
[28] H. Chen, Y. Li, Lett. Org. Chem. 5 (2008) 467–469.
[29] The synthesis and X-ray crystal structure of complex [RuCl2(ꢀ6-p-
cymene)(PTA-Me)] (3b) has been previously described: A. Dorcier, P.J.
Dyson, C. Gossens, U. Rothlisberger, R. Scopelliti, I. Tavernelli, Organometallics
24 (2005) 2114–2123.
(d) V. Cadierno, P. Crochet, J. Francos, S.E. García-Garrido, J. Gimeno, N. Nebra,
Green Chem. 11 (2009) 1992–2000.
[38] K. Takahashi, A. Honma, K. Ogura, H. Iida, Chem. Lett. (1982) 1263–1266.
[39] See, for example:
[30] Although rare, N-coordination of PTA-based ligands to transition-metals has
been documented. See Ref. [20] and works cited therein.
(a) C.S. Chin, B. Lee, I. Yoo, T.-H. Kwon, J. Chem. Soc. Dalton Trans. (1993)
581–593;
[31] The same trend was previously observed in our group when study-
ing the catalytic behaviour of the related arene-ruthenium(II) complexes
[RuCl2(ꢀ6-arene){P(CH2OH)3}] under biphasic water/n-heptane conditions:
V. Cadierno, P. Crochet, S.E. García-Garrido, J. Gimeno, Dalton Trans. (2004)
3635–3641.
(b) M. Crocker, R.H.M. Herold, J. Mol. Catal. 141 (1993) 70–79;
(c) V.L.K. Valli, H. Alper, Chem. Mater. 7 (1995) 359–362;
(d) M. Bartók, G. Szollosi, A. Mostalir, I. Dékánmy, J. Mol. Catal. A: Chem. 139
(1999) 227–234;
(e) E. Mas-Marzá, A.M. Segarra, C. Claver, E. Peris, E. Fernández, Tetrahedron
Lett. 44 (2003) 6595–6599;
[32] See, for example:
(a) A. Roucoux, J. Schulz, H. Patin, Chem. Rev. 102 (2002) 3757–3778;
(b) W. Yu, M. Liu, H. Liu, J. Zheng, J. Colloid Interface Sci. 210 (1999) 218–221.
[33] The Hg(0)-poisoning test is the most direct method to distinguishing homo-
geneous from heterogeneous catalysis when transition metals able to form an
amalgam are employed. See, for example: J.A. Widegren, R.G. Finke, J. Mol. Catal.
A: Chem. 198 (2003) 317–341, and references therein.
[34] (a) Attempts to isolate this complex from the stoichiometric reaction of 3a with
K2CO3 failed; (b) Related carbonate complexes [Ru(CO3)(ꢀ6-arene)(NHC)][Cs]
(NHC = N-heterocyclic carbene ligand) have been described by Peris and co-
workers as active catalysts for the redox isomerization of allylic alcohols: A.
Azua, S. Sanz, E. Peris, Organometallics 29 (2010) 3661–3664.
[35] See, for example, Ref. [2d] and:
(f) M. Poyatos, F. Márquez, E. Peris, C. Claver, E. Fernández, New J. Chem. 27
(2003) 425–431;
(g) V. Lillo, E. Fernández, A.M. Segarra, Tetrahedron: Asymm. 18 (2007)
911–914.
[40] This was estimated by considering that the totality of complex 3a, including
both the organometallic cation and the chloride anion, has been incorporated
into the clay and that the totality of the anions and cations present initially in
the Montmorillonite-K10 is maintained. It is also assumed that no molecules
of solvent have been incorporated in the solid.
[41] For related examples of superficial adsorption of ionic organometallic com-
plexes in Montmorillonite, see:
(a) R. Margalef-Català, C. Claver, P. Salagre, E. Fernández, Tetrahedron Lett. 41
(2000) 6583–6588;
(b) A.M. Segarra, R. Guerrero, C. Claver, E. Fernández, Chem. Commun. (2001)
1808–1809;
(c) C. Claver, E. Fernández, R. Margalef-Català, F. Medina, P. Salagre, J.E. Sueiras,
J. Catal. 201 (2001) 70–79;
(d) A.M. Segarra, R. Guerrero, C. Claver, E. Fernández, Chem. Eur. J. 9 (2003)
191–200.
(a) Y. Takai, R. Kitaura, E. Nakatami, T. Onishi, H. Kurosawa, Organometallics 24
(2005) 4729–4733;
(b) B. Martín-Matute, K. Bogár, M. Edin, F.B. Kaynak, J.-E. Bäckvall, Chem. Eur.
J. 11 (2005) 5832–5842;
(c) T. Campos-Malpartida, M. Fekete, F. Joó, Á. Kathó, A. Romerosa, M. Saoud,
W. Wojtków, J. Organomet. Chem. 693 (2008) 468–474.
[36] This fact was experimentally demonstrated by Bianchini and co-workers
with the zwitterionic Rh(I) complex [Rh(cod)(sulphos)] (sulphos = Ä3-P,P,P-
(PPh2CH2)3CCH2-p-C6H4SO3) which, under catalytic conditions, evolves into
the inactive species [Rh(CO)2(sulphos)]: C. Bianchini, A. Meli, W. Oberhauser,
New J. Chem. 25 (2001) 11–12.
[42] Inclusion of a bulky organometallic cation into the layers of Montmorillonite
results in an increase of several armstrong of the interlaminar distance of the
solid. Examples can be found in Refs. [19,39].
[43] We decided to increase the amount of solid employed to minimize mechanical
losses of the catalyst during the processes of filtration and washing, as well as
in the taking of aliquots for the reaction monitoring by GC.
[37] Previous studies in our group had already pointed out this effect. See, for exam-
ple, Ref. [11] and: