Efficient intermolecular [2 + 2 + 2] alkyne cyclotrimerization in aqueous medium using a ruthenium(IV) precatalyst

Article abstract of DOI:10.1021/ja066552k

The dimeric bis(allyl)-ruthenium(IV) complex [{Ru(η³:η³-C₁₀H₁₆)(μ-Cl)Cl}₂] (C₁₀H₁₆ = 2,7-dimethylocta-2,6-diene-1,8-diyl) was found to catalyze efficiently the [2 + 2 + 2] cyclization of terminal and internal alkynes in aqueous medium. Copyright

Full text of DOI:10.1021/ja066552k

Research Article
pubs.acs.org/acscatalysis
Imidazole Based Ruthenium(IV) Complexes as Highly Efficient
Bifunctional Catalysts for the Redox Isomerization of Allylic Alcohols
in Aqueous Medium: Water as Cooperating Ligand
Josefina Díez,† Jose Gimeno,*,† Agustí Lledos,*,‡ Francisco J. Suarez,† and Cristian Vicent§
́
́
́
†Laboratorio de Química Organometal
Instituto Universitario de Química Organometal
́
ica y Catal
́
isis (Unidad Asociada al C.S.I.C.), Departamento de Química Organ
ica “Enrique Moles”, Universidad de Oviedo, 33071 Oviedo, Spain
oma de Barcelona, Cerdanyola del Valles, 08193, Barcelona, Spain
Científica, Universitat Jaume I, Avenida Sos Baynat s/n, 12071 Castello, Spain
́ ́
ica e Inorganica,
́
‡Departament de Química, Universitat Auton
̀
̀
§Serveis Centrals d’Instrumentacio
́
́
S
* Supporting Information
ABSTRACT: Bis-allyl ruthenium(IV) complexes containing 1,3 azole β-N-H
protic ligands [Ru(η3:η3-C10H16)Cl2L] (C10H16 = 2,7-dimethylocta-2,6-diene-
1,8-diyl) (L = imidazole (1a), benzimidazole (1b)), and N-methylimidazole (1c)
are highly active precatalysts in the redox isomerization of allylic alcohols into
carbonyl compounds in aqueous medium and in the absence of base. A wide
series of primary and secondary allylic alcohols can be isomerized into the
corresponding carbonyl compounds. Remarkably, complex 1b has been found to
be the most efficient catalyst reported to date for the isomerization of 1-octen-3-
ol in water leading to a turnover frequency (TOF) value of 60 000 h−1. More-
over, catalyst 1b can be recycled remaining active up to seven cycles. Density
functional theory (DFT) calculations give evidence that the hydroxo complexes
derived from 1a−c species can be formed in aqueous solution and that they can
act as the catalytic active species in a bifunctional catalyzed process. This study
demonstrate that in water the participation of the β-N-H protic group of the 1,3-azole ligands in the bifunctional catalysis is not
required, provided that a water molecule can act as cooperating ligand.
KEYWORDS: allylic alcohols, aqueous medium, cooperating ligands, DFT calculations, redox isomerization, ruthenium catalysts
Chart 1
INTRODUCTION
■
Transition metal catalyzed isomerization of allylic alcohols has
become a very useful synthetic methodology of carbonyl deriva-
tives, much more attractive than those procedures based on
classic stoichiometric reactions which involve a two-step se-
quential oxidation and reduction processes using highly reactive
reagents. During the past decade a large number of highly
efficient synthetic routes have been reported disclosing genuine
examples of atom economy processes with both academic and
industrial interest.1−5 Among the most striking developments,
catalytic systems working in aqueous medium are particularly
relevant especially those using water as unique solvent since they
are able to perform more efficient transformations vs those using
organic solvents.6−23
During the past few years, our research group has been
engaged in the development of redox isomerization of allylic
alcohols in aqueous medium catalyzed by ruthenium(II) and
ruthenium(IV) complexes which proceed in excellent yields
under mild reaction conditions.3,24−38 More recently, we have
reported that ruthenium(IV) complexes [Ru(η3:η3-C10H16)-
Cl2(L)]) containing pyrazole based and other analogous ligands
containing α-N-H protic groups (see Chart 1) are highly active
catalysts in water including the transformation at 35 °C and in
the absence of base.39,40
It has been shown (by Electrospray Ionization Mass Spec-
trometry (ESI-MS)) that aqua-complexes of the type [Ru(η3:η3-
C10H16)Cl(H2O)(pyrazole)]+ (A) and [Ru(η3:η3-C10H16)(H2O)-
(OH)(pyrazole)]+ (B) (see Chart 2) are formed in solution and
play an important role on the catalytic activity.40
Remarkably, theoretical studies (density functional theory
(DFT)) reveal that hydroxo species (B and C) generated by
Received: June 11, 2012
Revised: August 22, 2012
Published: August 27, 2012
© 2012 American Chemical Society
2087
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Chart 2
Chart 3
catalysts, which lost any activity when the key NH sites are
methylated,41 1c shows a remarkable catalytic activity in water.
EXPERIMENTAL SECTION
■
deprotonation of the corresponding aqua-complexes are acting as
metal−ligand bifunctional catalysts, disclosing concerted outer-
sphere mechanisms via transition states [TS1] and [TS2], which
favor overall lower energy barriers (Scheme 1). The isomerization
implies two hydrogen-transfer steps from the substrate to the
catalyst and subsequent transfer back to the substrate. Motivated by
these findings we wondered whether other 1,3-azole β-N-H protic
heterocycles, that is, imidazole derivatives, might also provide
cooperative effects based on analogous ligand-substrate hydrogen
interactions enabling alternative favored outer-sphere mechanisms.
In these derivatives there is no accessible NH site able to take part in
Noyori-type metal−ligand bifunctional catalysis.
Herein, we describe that bis-allyl ruthenium(IV) complexes
[Ru(η3:η3-C10H16)Cl2(L)] (L = imidazole (1a), benzimidazole
(1b)) (Chart 3) are also highly efficient catalysts for the isomeri-
zation of allylic alcohols into carbonyl compounds in aqueous
media and in the absence of base. In particular, the following
remarkable novel features have been found by using these
catalysts: (i) the efficiency of catalyst 1b is much higher than that
shown by analogous pyrazole partner (TOF = 60000 vs 3000 h−1)
which, as far as we know, is the most efficient catalyst described to
date in water and in the absence of base; (ii) it is active at room
temperature; (iii) ESI mass spectra reveal the existence of
dissociation processes and the presence of cationic aqua-complexes
analogous to A and B which are the actual active catalytic species.
On the basis of these experimental results the catalytic mech-
anism has been analyzed theoretically. Different competitive
pathways reveal outer-sphere mechanisms enabled by the forma-
tion of aqua-complexes and their deprotonated species. The
most remarkable feature relies on the fact that the uncoordinated
N-H group of imidazole remains unchanged acting as a spectator
ligand during the whole catalytic cycle. This fact seems to suggest
that the use of N-substituted 1,3-azole ligands should also give rise to
active catalytic species. Experimental confirmation of this theoretical
prediction was obtained by studying the catalytic activity of the
complex [Ru(η3:η3-C10H16)Cl2(L)] (L = N-methylimidazole) (1c)
in water. In marked contrast to the usual behavior of Noyori
The manipulations were performed under an atmosphere of dry
nitrogen using vacuum-line and standard Schlenk techniques.
Solvents were dried by standard methods and distilled under
nitrogen before use. All reagents were from suppliers and used
without further purification with the exception of compounds
[{Ru(η3:η3-C10H16)(μ-Cl)Cl}2] (1),42−44 [{Ru(η6-arene)(μ-
Cl)Cl}2] (arene = C6H6, p-cymene, C6Me6),45,46 [Ru(η6-p-
cymene)Cl2(benzimidazole)] (2b)47,48 (conductivity in water,
20 °C: 173 Ω−1 cm2 mol−1), [Ru(η6-C6Me6)Cl2(benzimidazole)]
(2c)46,47 (conductivity in water, 20 °C: 180 Ω−1 cm2 mol−1), which
were prepared by following the methods reported in the literature.
Infrared spectra were recorded on a Perkin-Elmer 1720-XFT spec-
trometer. The conductivities were measured at room temperature,
in about 10−3 mol dm−3 water solutions, with a Jenway PCM3 con-
ductimeter. pH measurements have been performed with a WTW
Microprocessor pH-meter with a SenTix50T electrode. The C and
H analyses were carried out with a Perkin-Elmer 2400 micro-
analyzer. GC measurements were made on Hewlett-Packard
HP6890 equipment using a HP-INNOWAX cross-linked poly-
(ethyleneglycol) (30 m, 250 μm) or a Supelco Beta-Dex 120 (30 m,
250 μm) column. GC/MS measurements were performed on
Agilent 6890N equipment coupled to a 5973 mass detector (70 eV
electron impact ionization) using a HP-1MS column. NMR spectra
were recorded on a Bruker DPX-300 instrument at 300 MHz (1H)
or 75.4 MHz (13C) using SiMe4 as standard. DEPT experiments
have been carried out for all of the compounds reported. The
numbering for protons and carbons of the 2,7-dimethylocta-2,6-
diene-1,8-diyl skeleton is as follows:
Electrospray Ionization Mass Spectrometry. A Quattro
LC (QhQ quadrupole−hexapole−quadrupole) mass spectrometer
Scheme 1. Mechanism of the Isomerization of Allylic Alcohols Catalyzed by Pyrazole-Based Ruthenium(IV) Complexes in Water40
2088
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
with an orthogonal Z-spray-electrospray interface (Waters,
Manchester, U.K.) was used. The drying gas as well as nebulizing
gas was nitrogen at a flow of 300 and 50 L/h, respectively. The
temperature of the source block was set to 100 °C and the
interface to 120 °C. The capillary voltage was set at 3.5 kV in the
positive scan mode, and the cone voltage was adjusted to a low
value (typically Uc = 5−15 V) to control the extent of fragmenta-
tion in the source region. Sample solutions (ca. 5 × 10−4 M) in
the appropriate solvent were introduced through a fused-silica
capillary to the ESI source via syringe pump at a flow rate of
10 μL/min. The chemical composition of each peak obtained in
the full scan mode was assigned by comparison of the isotope
experimental and theoretical patterns using the MassLynx
4.0 program. Collision induced dissociation (CID) experiments
were performed with argon at various collision energies, ranging
from Elab = 3−15 eV. The most intense precursor peak of interest
was mass-selected with Q1 (isolation width of 1 Da), interacted
with argon in the hexapole cell while scanning Q2 to monitor
the ionic fragments. Typically, water speciation of compounds
1a−1c was investigated in triplicate where reproducibility was
observed. The temporal evolution of the ESI mass spectra of
compounds 1a−1c in water was investigated at different time
intervals encompassing t = 0 h to t = 1 h.
(CD2Cl2): δ 2.36 (s, 6H, 2xCH3), 2.41 (m, 2H, H4 and H6), 3.03
(m, 2H, H5 and H7), 3.71 (s, 3H, N−CH3), 4.28 (s, 2H, H2 and
H10), 4.45 (s, 2H, H1 and H9), 5.09 (m, 2H, H3 and H8), 6.64 (s,
1H, H5 L), 7.56 (s, 1H, H4 L), 8.11 (s, 1H, H2 L) ppm. 13C{1H}
NMR (CD2Cl2): δ 20.8 (s, 2xCH3), 34.5 (s, N-CH3),), 36.4 (s,
C4 and C5), 77.6 (s, C1 and C8), 94.7 (s, C3 and C6), 120.2 (s, C5
L), 128.9 (s, C2 and C7), 131.6 (s, C4 L), 141.5 (s, C2 L) ppm.
Conductivity (water, 20 °C): 75 Ω−1 cm2 mol−1.
Synthesis of Complex [RuCl2(η6-C6H6)(benzimidazole)]
(2a). A mixture of [{RuCl(μ-Cl)(η6-C6H6)}2] (250 mg,
0.5 mmol) and benzimidazole (1 mmol) in methanol (20 mL)
was stirred for 1 h. Then the solvent was removed under vacuum,
and the resulting orange solid residue was recrystallized in dichloro-
methane/diethylether and dried in vacuo. Yield 76% (279 mg).
Anal. Calcd for RuC13H12N2Cl2: C, 42.40; H, 3.28; N, 7.61.
Found: C, 42.55; H, 3.32; N, 7.57. IR (KBr, cm−1): 618 (w),744
(vs), 775 (w), 844 (s), 879 (w), 977 (w), 1006 (m), 1108 (w),
1140 (w), 1195 (w), 1247 (s), 1269 (m), 1305 (m), 1408 (s),
1432 (s), 1458 (m), 1485 (m), 1496 (w), 1595 (w), 3062 (m),
3125 (m), 3164 (vs), 3434 (s). 1H NMR (CDCl3): δ 5.76 (s, 6H,
C6H6), 7.10−7.70 (m, 4H, H4, H5, H6, H7 L), 8.37 (s, 1H, H2 L),
9.16 (s, 1H, NH) ppm. Conductivity (water, 20 °C): 190 Ω−1
cm2 mol−1.
Synthesis of [Ru(η3:η3-C10H16)Cl2(L)] (L = imidazole (1a),
benzimidazole (1b), N-methylimidazole (1c)). A stoichio-
metric amount of 1,3-azole (0.64 mmol) was added, at room
temperature, to a solution of complex [{Ru(η3:η3-C10H16)-
(μ-Cl)Cl}2] (1) (0.200 g; 0.32 mmol) in 20 mL of dichloro-
methane. After the mixture was stirred for 10 min, the solvent
was removed under vacuum, and the resulting orange solid
residue was washed with hexane (3 × 10 mL) and dried in vacuo.
1a: Yield 63% (0.152 g). Anal. Calcd for RuC13H20N2Cl2: C,
41.49; H, 5.36; N, 7.44. Found: C (41.40), H (4.49), N (7.56).C,
47.89; H, 5.27; N, 6.46. IR (KBr, cm−1): ν 611 (s), 653 (s), 753
(s), 952 (m), 1023 (m), 1060 (vs), 1100 (m), 1319 (m), 1380
General Procedure for the Catalytic Isomerization of
Allylic Alcohols into Carbonyl Compounds. In a sealed tube
4 mmol of the corresponding allylic alcohol, the ruthenium
catalyst precursor, and cocatalyst (if required) were dissolved in
tetrahydrofuran (THF) or water (20 mL), and the reaction
mixture was stirred at 75 °C for the indicated time. The course of
the reaction was monitored by regular sampling and analysis by
gas chromatography every five minutes. Only turnover frequency
(TOF) values at quantitative transformations are reported (otherwise,
it is indicated properly). The identity of the resulting saturated
carbonyl compounds was assessed by comparison with commer-
cially available pure samples (Aldrich Chemical Co. or Acros
Organics) and by their fragmentation in GC/MS. Catalyst recycling
is carried out by liquid−liquid extraction of the product with diethyl
ether (3 × 5 mL) and subsequent addition of substrate.
X-ray Crystal Structure Determination of Complexes
1a, 1b, and 1c. Crystals suitable for X-ray diffraction analysis
were obtained, for 1a and 1c, by slow diffusion of hexane into a
saturated solution of the complexes in dichloromethane, and for
1b, were obtained from tetrahydrofuran/diethylether solvent
system. The most relevant crystal and refinement data are
collected in the Supporting Information, Table S1.
For 1a and 1c, diffraction data were recorded on a Nonius
KappaCCD single crystal diffractometer, using Mo−Kα radiation
(α = 0.71073 Å). In both cases images were collected at a 35 mm
fixed crystal-detector distance, using the oscillation method, with 1°
oscillation and 50 s exposure time per frame. Data collection strategy
was calculated with the program Collect.49 Data reduction and cell
refinement were performed with the programs HKL Denzo and
Scalepack.50 A semiempirical absorption correction was applied
using the program SORTAV.51
For 1b data collection was performed on a Oxford Diffraction
Xcalibur Nova single crystal diffractometer, using Cu−Kα radia-
tion (α = 1.5418 Å). Images were collected at a 65 mm fixed
crystal-detector distance, using the oscillation method, with 1°
oscillation and (1.5−10) s variable exposure time per image. Data
collection strategy was calculated with the program CrysAlis Pro
CCD.52 Data reduction and cell refinement was performed with
the program CrysAlis Pro RED.52 An empirical absorption
1
(m), 1455 (m), 1540 (m), 2916 (m), 3143 (m), 3270 (s). H
NMR (CD2Cl2): δ 2.33 (s, 6H, 2xCH3), 2.42 (m, 2H, H4 and
H6), 3.06 (m, 2H, H5 and H7), 4.31 (s, 2H, H2 and H10), 4.47 (s,
2H, H1 and H9), 5.11 (m, 2H, H3 and H8), 7.09 (s, 1H, H5 L),
7.64 (s, 1H, H4 L), 8.29 (s, 1H, H2 L), 9.70 (s, 1H, NH) ppm.
13C{1H} NMR (CD2Cl2): δ 20.8 (s, 2xCH3), 36.4 (s, C4 and C5),
77.6 (s, C1 and C8), 94.9 (s, C3 and C6), 115.7 (s, C5 L), 129.1 (s,
C2 and C7), 130.8 (s, C4 L), 139.3 (s, C2 L) ppm. Conductivity
(water, 20 °C): 108 Ω−1 cm2 mol−1. 1b: Yield 76% (0.210 g).
Anal. Calcd for RuC17H22N2Cl2: C, 47.89; H, 5.20; N, 6.57.
Found: C, 47.93; H, 5.17; N, 6.63. IR (KBr, cm−1): ν 613 (w),
701 (w), 750 (vs), 763 (w), 887 (w), 962 (m), 1012 (m), 1031
(m), 1111 (m), 1151 (m), 1248 (m), 1261 (m), 1302 (m), 1361
(m), 1383 (m), 1418 (s), 1456 (m), 1494 (m), 1618 (w), 2907
(m), 3136 (s), 3197 (s). 1H NMR (d8-THF): δ 2.42 (m, 2H, H4
and H6), 2.49 (s, 6H, 2xCH3), 3.11 (m, 2H, H5 and H7), 4.54 (s,
2H, H2 and H10), 4.97 (s, 2H, H1 and H9), 5.22 (m, 2H, H3 and
H8), 7.30−7.40 (m, 2H, H6 and H7 L), 7.56 (d, 1H, JHH = 8.4 Hz,
H8 L), 8.48 (d, 1H, JHH = 8.1 Hz, H5 L), 8.86 (s, 1H, H2 L), 9.71
(s, 1H, NH) (s, 1H, NH) ppm. 13C{1H} NMR (d8-THF): δ 21.4
(s, 2xCH3), 37.0 (s, C4 and C5), 78.2 (s, C1 and C8), 95.4 (s, C3
and C6), 103.3 (s, C9 L), 112.7 (s, C5 L), 122.6 (s, C6 L), 123.3 (s,
C7 L), 123.8 (s, C8 L), 129.7 (s, C2 and C7), 134.7 (s, C4 L), 148.9
(s, C2 L) ppm. 1c: Yield 72% (0.180 g). Anal. Calcd for
RuC13H20N2Cl2: C, 43.08; H, 5.68; N, 7.18. Found: C, 43.27; H,
5.85; N, 7.25. IR (KBr, cm−1): ν 619 (m), 658 (m), 753 (f), 861
(m), 1023 (m), 1093 (mf), 1110 (f), 1238 (m), 1379 (m), 1423
1
(m), 1456 (m), 1536 (m), 2916 (m), 3118 (f). H NMR
2089
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Scheme 2. Synthesis of Complexes 1a−c and 2a−c
correction was applied using the SCALE3 ABSPACK algorithm
as implemented in the program CrysAlis Pro RED.52
The energies in solution (ΔEsol) of all the species were
estimated adding to the potential energy the contribution of the
Gibbs energy of solvation by the application of single-point
calculations, using the polarizable continuum model (PCM) with
standard solvation spheres69−72 in H2O, (ε = 78.4) at the gas-
phase optimized stationary points. An extended basis set, with a
6-311++G(d,p) description of all the atoms73 but the ruthenium,
described as LANL2DZ+f, was employed in these calculations.
The ΔEsol values include the Gibbs energy effects of the solvent,
but not those of the solute. The Gibbs energies in solution
(ΔGsol) were calculated adding the gas-phase Gibbs energy cor-
rections of the solute to the energies in solution. All the energies
collected in the text are ΔGsol. ΔEsol profiles are shown in the
Supporting Information.
In all cases the software package WINGX53 was used for space
group determination, structure solution, and refinement. The
structures of complexes 1a and 1c were solved by Patterson
interpretation and phase expansion using DIRDIF.54 For 1b the
structure was solved by direct methods using SIR92.55
Isotropic least-squares refinement on F2 was performed using
SHELXL9756. During the final stages of the refinements, all the
positional parameters and the anisotropic temperature factors of
all the non-H atoms were refined. The H atoms for 1a and 1c
(except H1A,B for 1a and H1A,B, H10A,B, H24A,B for 1c) were
geometrically located, and their coordinates were refined riding
on their parent atoms. For 1b, H1A,B for 1a and H1A,B, H10A,B,
H24A,B for 1c, the coordinates of H atoms were found from
different Fourier maps and included in a refinement with
RESULTS AND DISCUSSION
■
2
isotropic parameters. The function minimized was ([∑w(Fo
−
Synthesis of Imidazole-Based Ruthenium Complexes
and Speciation in Aqueous Medium. We have investigated
the catalytic activity of ruthenium(IV) [Ru(η3:η3-C10H16)-
Cl2(L)] (L = imidazole (1a), benzimidazole (1b), N-methy-
limidazole (1c)) and ruthenium(II) [Ru(η6-arene)Cl2L] com-
plexes (2a−c) for comparative purposes. All of them have been
synthesized in good yield from the commercially available
dimeric precursors [{Ru(η3:η3-C10H16)Cl2}2] and [{RuCl(μ-Cl)-
(η6-arene)}2] by reaction with the appropriate ligand following
procedures reported in the literature for analogous 1,2-azoles39,40
and for 2b,c47,48 (Scheme 2).
All of them have been isolated as air-stable orange solids
soluble in dichloromethane and acetone and insoluble in hexane
and diethyl ether. Although conductivity measurements of
acetone solutions show that they are no electrolytes, the molar
conductivities found in water (177 (1a), 75(1c) and 190 (2a)
Ω−1 cm2 mol−1) indicate that dissociation processes to give
cationic derivatives take place in solution (low solubility in water
of 1b precludes conductivity measurements). Moreover, aqueous
solutions of 1a and 1c are acid (pH = 5.75). These data seem to
indicate that water-chloride exchange processes along with
proton dissociations may occur.
Detailed insights on the chemical speciation of 1a−1c in
aqueous media were obtained from in situ ESI mass spectrometry
(following a protocol previously reported),40 which are in accord
with the conductivity data. Aqueous solutions of compounds
1a−1c were stirred and aliquots were extracted at different time
intervals, diluted with water to a final concentration of 5 × 10−4 M
(based on the initial Ru concentration), and directly introduced
to the mass spectrometer. As a representative example of the ESI
mass of complex 1a are provided in Supporting Information,
Fc2)/∑w(Fo )]1/2 where w = 1/[σ2(Fo ) + (aP)2 + bP] (a and b
2
2
values are collected in Supporting Information, Table S1) with
σ2(Fo ) from counting statistics and P = (max (Fo , 0) + 2Fc2)/3.
Atomic scattering factors were taken from the International
Tables for X-ray Crystallography International.57 Geometrical
calculations were made with PARST.58 The crystallographic
plots were made with PLATON.59
2
2
Computational Details. All the theoretical calculations have
been performed using the density functional theory (DFT). The
full catalytic cycles with complexes 1a−c were computed with
the Becke’s three-parameter exchange functional (B3)60,61 in
conjunction with the Lee−Yang−Parr correlation functional
(LYP)62,63 as implemented in Gaussian 03.64 To obtain more
accurate values of the energy barriers of the hydrogen transfer
steps these barriers were recalculated, with reoptimization of
intermediates and transition states, using the M06 functional65
with an ultrafine grid, as implemented in the Gaussian 09 program.66
In all geometry optimizations the effective core potential LANL2DZ
along with its associated basis set, complemented by a series of f
polarization functions was employed for Ru,67,68 6-31G(d) basis set
was used for all the non-hydrogen atoms, 6-31G(d,p) for the
hydrogen atoms of the aqua and pyrazole ligands, as well as the
substrates, and 6-31G for the rest of hydrogen atoms.68 Vibra-
tional frequencies were computed to characterize the transitions
states and the minima, as well as for the calculation of gas-phase
Gibbs energies. The connection between transition states and
their corresponding reactants and products was checked by fully
optimizing structures derived from the transition states with
a small displacement following the transition vector in both
directions.
2090
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Scheme 3. Formation Equilibrium of Cationic Species Detected by ESI-MS of 1a Aqueous Solution at t = 0 and t = 1.5 h
Figure S1 illustrating the temporal evolution of a freshly prepared
aqueous solution at different time intervals.
Catalytic Activity of Imidazole Based Ruthenium
Complexes for the Isomerization of Allylic Alcohols In
Water. Figure 1 shows the catalytic activity of ruthenium(IV)
At t = 0 h, the ESI mass spectrum shows a dominant species
identified as [1a − Cl]+ (m/z 341) accompanied by a minor peak
assigned to [1a − Cl + H2O]+ (m/z 359) that formally corre-
spond to the Ru−Cl cleavage and subsequent water coordination
in the vacant site. This experimental evidence in combination
with conductivity measurements strongly suggest that water
speciation of 1a at the early stages is dominated by singly charged
[1a − Cl]+ and [1a − Cl +H2O] + species. An additional minor
species at m/z = 305 is also observed and assigned to [1a − Cl −
HCl]+ on the basis of its m/z value as well as its isotopic
distribution (Scheme 3). After 1 h, new species are detected in
the ESI mass spectrum, arising from the evolution of the former
aqua-species. Cationic chloride free species, namely, [1a − Cl −
HCl]+ (m/z 305), their corresponding mono- and disolvated
species, namely, [1a − Cl − HCl + H2O]+ (m/z 323) and [1a −
Cl − HCl +2H2O]+ (m/z 341), were also detected and their
relative intensity increase relative to the initial [1a − Cl]+ (m/z
341). Note that species [1a − Cl]+ and [1a − Cl − HCl +2H2O]+
are overlapped; however, they could be unambiguously identified
on the basis of their distinctive collision induced dissociation
(CID) spectra.40 Let us note that the versatility of the ESI-MS
technique and its tandem version in transferring intact ions from
the reaction solution to the gas phase is well-documented in the
context of homogeneous catalytic reactions74−77 as further
illustrated in the present work. Aqueous chemical speciation for
compound 1b and 1c was comparable to that found for 1a as
judged by identical ESI-MS analysis (see Supporting Information).
Spectroscopic data (IR and NMR spectra) and elemental
analysis for compounds 1a−c and 2a are in agreement with the
proposed formulations (see Experimental Section for details). In
particular, 1H and 13C{1H} NMR spectra of 1a−c and 2a show
the signals for the η3:η3 2,7- dimethylocta-2,6-diene-1,8-diyl and
η6-C6H6 ligands respectively, which compare well with those
previously reported for analogous complexes.40,78−81 As
expected, proton and carbon resonances of the 2,7-dimethyl-
octa-2,6-diene-1,8-diyl group in the NMR spectra are in
accordance with the formation of a simple equatorial adduct
with a C2 symmetry in which the two halves of the bis-allyl ligand
are in equivalent environments.82 In addition, the structures of
ruthenium(IV) complexes 1a−c were unequivocally confirmed
by a single crystal X-ray diffraction study. Drawings of the
molecular structures including selected bond distances and
angles are depicted in the Supporting Information, Figure S2.83
Figure 1. Isomerization of 1-octen-3-ol into octa-3-one catalyzed by
complexes 1a−c and 2a−c in water. Reactions performed under
N2 atmosphere using 4 mmol of 1-octen-3-ol (0.2 M). Substrate/Ru =
500/1. Yield of octan-3-one determined by GC. Turnover frequencies
((mol product/mol Ru)/time) were calculated at the time indicated in
each case.
(1a−c) and ruthenium(II) (2a−c) complexes for the isomer-
ization of 1-octen-3-ol into octa-3-one in water and in the absence of
base as a reaction model. Under the optimized reaction conditions
previously reported for analogous catalysts (i.e., 0.2 M of 1-octen-3-ol,
0.2 mol % in Ru, heating at 75 °C in water)24−29,39,40 all the cata-
lysts turned out to be active, although ruthenium(IV) precata-
lysts show much higher activity than ruthenium(II) derivatives84
reaching quantitative transformations in only 5−20 min vs 30−
180 min for the latter. We were pleased to find that complex 1b
displays much higher catalytic activity than 1a and 1c, leading to
quantitative transformations in less than 5 min. This outstand-
ing catalytic performance is maintained at room temperature
(100 min; >99% yield)85 and most interestingly under a lower
catalyst loading (0.02 mol %) reaching an excellent yield of
3-octanone (>99%) in only 5 min. (TOF = 60,000 h−1).86 This
methodology is also amenable for practical applications in a
multigram scale in which 20 mmol (2.68 g) of α-vinylbenzyl
alcohol is isomerized by 1b (1 mol %) into 1-phenylpropan-1-
one after 15 min at 75 °C (isolated yield 89%).
2091
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Complex 1b is also very efficient in the isomerization of a range
of allylic alcohols proving the wide scope and limitations of this
catalytic transformation. The results are shown in Table 1. There
for pyrazole-based complexes. The previous study pointed out a
crucial role of hydroxo species generated by deprotonation of the
corresponding aqua-complexes (Chart 2). Thus, hydroxo species
with imidazole (OH-a-I0), benzimidazole (OH-b-I0), and
methylimidazole (OH-c-I0) will be considered as the active species.
For all the hydroxo complexes we have found two pathways with
relative cis- (c-) and trans- (t-) arrangement of the hydroxo and
imidazole ligands. The catalytic cycle has been investigated with
both isomers for the three hydroxo-complexes.
Table 1. Isomerization of Substituted Allylic Alcohols
a
Catalyzed by Complex 1b in Water at 75 °C
The previous study involving pyrazole catalysts showed that
the overall isomerization implies two-hydrogen transfer steps:
from the substrate to the catalyst (dehydrogenative step) and
subsequent transfer back to the substrate (acrolein hydro-
genation) disclosing an outer-sphere bifunctional catalyzed
mechanism for both processes (Scheme 1). In the first step
(path A) hydrogen transfer to the Ru/OH moiety occurs while a
cooperative action of the metal and the pyrazole ligand takes
place in the second step (hydrogenation from the Ru−H/
pyrazole N-H moiety; path B). Calculations also demonstrated
that the second half of the reaction has a considerably lower
barrier when the acrolein hydrogenation takes place via 1,4-
addition, forming initially the enol (1-propen-1-ol) instead of a
direct CC reduction. The enol can be easily isomerized to
propanal in water. With this in mind, we have explored the iso-
merization mechanism with the imidazole-based catalysts con-
sidering the water ligand formed by protonation of the hydroxo
ligand in the first step is acting as the proton transfer agent in the
hydrogenation in a similar way as described for pyrazole com-
plexes. Schemes 4 and 5 show the catalytic cycles investigated with
the trans- (trans-hydroxo route) and cis- (cis-hydroxo route) isomers
of the hydroxo-imidazole active species t-OH-a-I0 and c−OH-a-I0,
respectively (for catalytic cycles with the corresponding active
species of the benzimidazole and N-methylimidazole complexes see
the Supporting Information, Schemes S1−S4).
a
Reactions performed under N2 atmosphere using 1 mmol of allylic
b
c
alcohol (0.2 M). Yields determined by GC. [Cycle] and Accumula-
tive TON (mol product/mol Ru): [1st] 99; [2nd] 198; [3rd] 297;
[4th] 396; [5th] 495; [6th] 592; [7th] 691. 12% of α,β-unsaturated
Accessibility of Hydroxo-Imidazole Active Species. An
essential point in the proposed mechanism is the existence in
water of the hydroxo active species cis- and trans- hydroxo com-
plexes (Scheme 6). To discuss more quantitatively the accessibility
of such species from the neutral dichloro complexes we have
estimated their relative Gibbs energies in water using eq 1.
d
aldehyde (cinnamaldehyde).
is a strong dependence upon the substitution of the carbon−
carbon double bond, a behavior previously reported with other
catalytic systems.6−29 Thus, monosubstituted aliphatic olefins 1-
hepten-3-ol, 1-hexene-3-ol, 1-pentene-3-ol, 3-buten-2-ol, and 2-
propen-1-ol are readily isomerized in 15−30 min (90−99% yield,
entries 1−5).87 In contrast, when either 1,1- or 1,2-disubstituted
allylic alcohols are used, longer reaction times and higher catalyst
loadings (1−10 mol %) are required to achieve high conversions
(entries 7−13). Remarkably, the catalyst 1b has shown a high capacity
to be recycled remaining active up to seven cycles (entry 1c).
Bifunctional Catalysis without Participation of N-H
Protic Groups: Mechanistic Considerations. Our previous
study on allylic alcohol isomerization in water catalyzed by
pyrazole-based ruthenium complexes (Scheme 1) raised doubts
about the need for an NH moiety in the ligand to perform a
metal−ligand cooperating bifunctional catalysis of such process.
The experimental results reported above show that similar or even
improved catalytic efficiencies can be obtained with imidazole and
benzimidazole complexes 1a and 1b in which the orientation of the
β-N-H unit of the ligands make them unsuitable to participate in
proton transfer to/from the substrate. To confirm the key role that
water ligands can play as alternative to N-H Noyori type ligands in
bifunctional catalyzed processes in water, we have carried out a DFT
study of the isomerization of 2-propen-1-ol by imidazole-based
complexes 1a, 1b, and 1c, using a similar approach to that employed
To calculate these values, experimentally derived values for the
aqueous solvation Gibbs energy of a proton (−264.0 kcal mol−1)88
and a chloride ion (−74.6 kcal mol−1)89 have been used. The calcu-
lated relative energies should be considered an estimation given the
inherent difficulties in calculating solvation energies of charged
species. For all the hydroxo complexes both cis- and trans- isomers
have very similar energies (Scheme 6).
The hydroxo complexes are found in the range from about 13
to 16 kcal mol−1 above the neutral saturated dichloro complexes.
These values are lower than those calculated for the analogous
cis- and trans- hydroxo pyrazole complexes (18.3 and 20.1
kcal mol−1, respectively). Moreover, whereas in the pyrazole-
based system the cis-aqua/pyrazolide complex is 3.8 kcal mol−1
2092
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Scheme 4. Proposed Catalytic Cycle for the Isomerization of 2-Propen-1-ol into Propanal in Water Promoted by t-OH-a-I0 (trans-
hydroxo route for 1a)
Scheme 5. Proposed Catalytic Cycle for the Isomerization of 2-Propen-1-ol into Propanal in Water Promoted by c−OH-a-I0 (cis-
hydroxo route for 1a)
Scheme 6. Relative Gibbs Energies in Water (kcal mol−1), with Respect to the Parent Dichloro Complexes, of the cis- and trans-
Hydroxo Complexes
more stable than the trans-hydroxo/pyrazole active species, in
the imidazole-based complexes the cis-aqua/imidazolide is
8.5 kcal mol−1 less stable than the trans-hydroxo/imidazole
and the aqua/benzimidazolide and hydroxo/benzimidazole
complexes are isoenergetic. Overall the data point out an
easier formation and higher concentration of the hydroxo
active species in the imidazole-based complexes than in the
pyrazole-based partners.
To further assess the presence of hydroxo complexes in
aqueous solutions of 1a, 1b, and 1c we have calculated, using the
M06 functional,65 the pKa of the coordinated water in the
cationic species formed by water-chloride exchange and that have
been detected by ESI-MS (for instance, for 1a: [1a − Cl − HCl +
H2O]+ m/z 323, see Scheme 3). To this end we have estimated
the Gibbs energy of deprotonation in water (ΔG°solv,deprot) of the
aqua-complexes water-a-I0, water-b-I0, and water-c-I0 (Scheme 7)
2093
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
(Scheme 6) behave as active species. The Gibbs energy profiles in
water for the isomerization catalyzed by trans- and cis-hydroxo
complexes t-OH-a-I0 and c−OH-a-I0 (trans- and cis-hydroxo
routes, catalytic cycles outlined in Schemes 4 and 5, respectively)
are presented in Figure 2. The catalytic cycles and Gibbs energy
profiles for the isomerization reaction catalyzed by the cis- and
trans- hydroxo derivatives formed in aqueous solution from
benzimidazole (1b) and N-metlylimidazole (1c) complexes are
very similar to those of the related imidazole species and are
collected in the Supporting Information, Figures S11 and S13,
respectively.91
Scheme 7. Aqua-Complexes Considered in the pKa
Calculations
The isomerization takes place through two sequential
concerted hydrogen transfers with a cooperative action of the
metal and the hydroxo ligand. The imidazole ligand is first
accepting a proton from the 2-propen-1-ol substrate (dehydro-
genation step) and then the aqua ligand formed in this way is
transferring back the proton to the acrolein. (hydrogenation
step). At the same time a hydride is transferred to and from the
metal, giving rise to the enol form of the product. Transition
states for the hydrogen transfer steps with selected structural
parameters are shown in Figure 3.
The Gibbs energy barriers for the dehydrogenative step are
16.6 kcal mol−1 in the trans-hydroxo route and 21.3 kcal mol−1 in
the cis-hydroxo route. At this point acrolein has been formed, and
the catalyst has achieved its reduced form, which will transfer two
hydrogens to acrolein in the second half of the reaction. Acrolein
hydrogenation is markedly more asynchronic than 1-propen-1-ol
hydrogenation, hydride transfer being much more advanced than
the proton transfer in the transition state of this step.92 Regard-
ing the reduced form of the catalyst, the cis-aqua ruthenium
hydride is considerably more stable than the parent trans-aqua
(10.5 kcal mol−1). This fact places the energy profile for the
acrolein hydrogenation in the trans-hydroxo route above that of
the cis-hydroxo pathway (Figure 2). However the energy barriers
for this step, computed from the respective intermediates (t-OH-
a-I2 and c−OH-a-I2) are very similar: 11.6 and 12.0 kcal mol−1
for the trans- and cis-routes, respectively. Overall the isomer-
ization catalyzed by the trans-hydroxo active species is favored by
3.6 kcal mol−1 over that involving the cis-hydroxo species (global
barriers93 of 17.6 and 21.3 kcal mol−1).
using the thermodynamic cycle we have already employed for the
estimation of pKa's of organometallic complexes (see Supporting
Information).90 From this value, calculation of pKa is straight-
forward (pKa = ΔG°solv,deprot/2.303RT). Details of the pKa calcu-
lations are given in the Supporting Information. We checked the
suitability of such an approach for calculating the pKa of the free
imidazole and benzimidazole molecules. The obtained results
(14.3 and 12.5, respectively) are in very good agreement with the
experimental values (14.2 and 12.75). The ΔG°solv,deprot of the
aqua-complexes are 13.7 kcal mol−1 (L = imidazole), 16.0 kcalmol−1
(L = benzimidazole), and 13.5 kcal mol−1 (N-methyl imidazole).
Accordingly, the calculated pKa for the coordinated water in aqua-
complexes water-a-I0, water-b-I0, and water-c-I0 are 10.1
(imidazole), 11.9 (benzimidazole), and 10 (methylimidazole).
Comparing these values with that of a water molecule in pure water
(15.7), it is apparent that a significant increase in the acidity of water
(about 5 pKa units) can be recognized. The pKa decrease places
coordinated water in an acidity strength similar to that of phenol
(pKa = 9.9), for which phenolate anion exists in appreciable
concentration in water. Moreover, the presence of a basic iminic N
atom of the ligand allows the proton dissociation and subsequent
protonation of the resulting neutral hydroxo complex (proton
transfer). This leads to a more stable product, which displaces the
overall deprotonation-protonation equilibrium toward the formation
of the hydroxo complexes proposed to be the active species (eq 2).
The results discussed above stress the feasibility of the bi-
functional catalysis with cooperative action of a water ligand
for the ruthenium-catalyzed isomerization of allylic alcohols in
water. However, hydrogen-bonding interactions between the
reactive system and the water medium, not accounted for with a
continuum description of the solvent, could affect the energy
barriers.94 To check whether by improving the solvent model
could affect our results, we have recalculated the energy barriers
of the two hydrogen transfer steps (transition states TS1 and
TS2) in the trans-hydroxo route with an improved description of
solvent effects. In this model, explicit water molecules (from one
to four) able to establish hydrogen-bonding interactions with the
catalyst and the substrate have been included, in addition to the
continuum description of the solvent. The calculated Gibbs energy
barriers, depending on the number of explicit water molecules, are
collected in Table 2. Geometries of the transition states with explicit
water molecules are presented at the Supporting Information and
Scheme 8 (three water molecules as illustration).
ESI-MS detected a [1a − Cl − HCl + H2O]+ species (m/z
323). Both the aqua/imidazolide and the hydroxo/imidazole
formulations are compatible with this m/z value. Calculations
support the hydroxo-imidazole nature of such species. Therefore
it seems very likely that the presence in aqueous solution of the
hydroxo-complexes derived from 1a−c species, proposed to be
the active species in the isomerization process.
Catalytic Efficiency in Water of the β-N-H Protic
Imidazole Based and N-methylimidazole Ruthenium(IV)
Complexes. We have computed the full catalytic cycles for the
isomerization of allylic alcohol in water assuming the cationic
trans- and cis-hydroxo complexes derived from 1a, 1b, and 1c
As it is sketched in Scheme 8 for the model with three explicit
water molecules, a network of hydrogen bonds is connecting the
substrate and the hydroxo ligand in the transition states of the
first and second hydrogen transfer, and a water molecule is also
2094
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Figure 2. Gibbs energy profiles in water for the isomerization of 2-propen-1-ol into propanal assuming trans-[Ru(η3:η3-C10H16)(OH)(imidazole)]+
(t-OH-a-I0) and cis-[Ru(η3:η3-C10H16)(OH)(imidazole)]+ (c−OH-a-I0) are the active species (trans- and cis-hydroxo route for 1a). The values are in
kcal mol−1 and referred to the separated reactants t-OH-a-I0 and 2-propen-1-ol.
hydrogen bonded to the β-N-H protic group of the 1,3-azole
ligand. Although the presence of explicit water molecules
modifies the energy barriers, the global barriers obtained with
the most accurate models (including three or four explicit solvent
molecules) differ less than 1 kcal mol−1 from that obtained with
the purely continuum model. Thus, we can conclude that the
feasibility of a bifunctional-catalyzed mechanism with participa-
tion of an aqua ligand for hydrogen-transfer processes in aqueous
medium is kept when a cluster of explicit solvent molecules is
included in the calculations. Note that the results with only one
explicit water molecule deviate notably from those with larger
water clusters, pointing out that smaller water cluster could give
skewed results.
The calculated barriers for the two hydrogen transfer steps
assuming the hydroxo complexes of the three imidazole based
ruthenium(IV) complexes take part as the active species in the
isomerization of 2-propen-1-ol into propanal are gathered in
Table 3. The B3LYP results outline lower barriers for the trans-
hydroxo routes than for the cis-hydroxo ones, as well as lower
barriers for the second hydrogen transfer (hydrogenation step)
than for the first one (dehydrogenation step). The overall
barriers for the favored trans-hydroxo routes with the three
imidazole based complexes although similar follow the order
1b < 1a < 1c (imidazole: 17.6 kcal mol−1; benzimidazole: 17.0
kcal mol−1; N-methylimidazole: 18 kcal mol−1). We have also
included in Table 3 the barriers already reported for the pyrazole
complex.40 The global barrier for the 2-propen-1-ol isomer-
ization with this catalyst is the lowest one, in seeming con-
tradiction with its lowest catalytic efficiency. However there is an
important feature of the hydroxo catalysts not reflected in Table 3:
its relative energy with respect the parent aqua complex initially
formed by chloride dissociation. This issue determines the
effective concentration of the active species in the reaction
medium. Whereas for the hydroxo/imidazole-based complexes
the cis-aqua-imidazolide is less stable than the cis-hydroxo/
imidazole (isoenergetic for the corresponding benzimidazole
complexes), the hydroxo-pyrazole active species is found 3.8
kcal mol−1 above its aqua-pyrazolide isomer.40 Thus, adding the
energy required to forming the active species to the global barrier
the value obtained for the pyrazole complex is 19.1 kcal mol−1,
which is the highest one in the series. Therefore, calculations
point out an important role of the availability of the hydroxo
species in the catalytic efficiency of these species.
The B3LYP computed barriers seem to be too high, according
to the experimentally determined TOF. Because recent studies
have established that the M06 functional predicts the activation
energies more accurately than B3LYP,65 we have recalculated the
Gibbs energy barriers in water for catalysts 1a−1c, as well as for
the pyrazole complex, using the M06 functional. The geometries
of the transition states for both hydrogen transfers and their most
stable preceding intermediates have been reoptimized with this
functional and are included in the Supporting Information.
Although changes in the geometries are minor, notable changes
in the barriers take place when changing the B3LYP functional by
the M06, which describe more accurately non covalent inter-
actions such as those occurring in the transition states of the
hydrogen transfer steps. As shown in Table 3, the M06 global
barriers are about 3.5 kcal mol−1 lower than the B3LYP ones, in
better agreement with the catalytic activity of the imidazole based
ruthenium complexes. The M06 global barriers for the three
imidazole based complexes are very similar ((imidazole: 13.8 kcal
mol−1; benzimidazole: 13.9 kcal mol−1; N-methylimidazole:
13.6 kcal mol−1) suggesting that the slight differences in the
barriers, which are at the origin of the differences in TOF
between 1a−1c complexes, are out of the accuracy of the calcu-
lations. M06 calculations also point out much closer barriers for
the cis- and trans-hydroxo routes and for the first and second
hydrogen transfers.
2095
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Figure 3. Transition states for the hydrogen-transfer steps in the isomerization of 2-propen-1-ol into propanal catalyzed by 1a in the cis- and trans-
hydroxo routes. Hydrogen atoms in the bis(allyl) ligand have been omitted for clarity.
Table 2. Gibbs Energy Barriers (kcal mol−1) for the Two
Hydrogen Transfer Steps in the Isomerization of 2-Propen-1-
ol into Propanal in Water Catalyzed by 1a with Explicit
Solvent Molecules Included in the Calculated System
Scheme 8. Transition States for the Hydrogen-Transfer Steps
in the trans-Hydroxo Route Catalyzed by t-OH-a-I0 with
Three Additional Water Molecules
number of explicit water
molecules
1st hydrogen
2nd hydrogen
global
b
a
a
transfer
transfer
barrier
0
1
2
3
4
16.6
12.4
16.6
16.7
18.4
11.6
13.3
14.2
12.7
9.8
17.6
14.4
18.8
16.7
18.4
a
ΔG†
barriers from the most stable preceding intermediate.
Defined as the Gibbs energy difference between the highest and the
water
b
lowest structures along the whole catalytic cycle, ref 93.
Overall, the barriers found with the three imidazole based
complexes underline the viability of a mechanism relying on the
key role of a water ligand acting as cooperating ligand along the
catalytic cycle. In our previous report, using pyrazole-based
ruthenium complexes, we gave evidence that the hydroxo com-
plex could take part in the dehydrogenative half of the reaction,
but protonation from the pyrazole ligand was favored over
protonation from the aqua ligand (see Scheme 1).40 The
present results demonstrate that in water the participation of
the N-H group of the azole ligand in the bifunctional catalysis
process is not required, provided that a water ligand could play
the same role. In this sense, the catalytic action of the
N-methylimidazole complex, where both nitrogen atoms of the
ring are substituted, is particularly appealing. Indeed, the TOF
obtained with the benzimidazole complex 1b (much higher
than that shown by the partner pyrazole complex) prove that a
higher catalytic efficiency can be achieved without direct inter-
vention of the azole ligands.
2096
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Our study shows that, despite that these ligands featuring β-N-
H protic groups are not required for the hydrogen transfer
reactions, they play an important role in the fine-tuning of the
ruthenium environment for the catalyst to attain its optimal
efficiency. Efficient catalysis of the allylic alcohol isomerization
entails both hydrogen transfers to and from the catalyst. Thus, a
good catalyst should be able to perform efficiently both processes,
which could require contrasting properties (for instance,
formation of a too strong M-H bond in the first step would
hamper the hydride transfer back to the substrate in the second
step or a too acidic water ligand would make difficult the protona-
tion of the hydroxo ligand in the second half of the process). Two
key issues are related with the donor properties of the N-ligand:
the acidity of the water ligand, which affects the concentration of
the hydroxo active species, and the Lewis acid properties of the
ruthenium center, which make it suitable for accepting the
hydride. The donor properties of the benzimidazole ligand fit
very well with the system requirements to reach an optimal
performance in the allylic alcohol isomerization process. The
results reported in the present work pave the way for the design
of new bifunctional catalyzed processes in water in which aqua
ligands act as the cooperating ligand.
Table 3. Calculated Gibbs Energy Barriers for Each Hydrogen
Transfer Step and Global Barrier for the Isomerization of
2-Propen-1-ol into Propanal in Water Using Different
Pre-Catalysts with 1,3-Azole Ligands (kcal mol−1; in italics,
M06 values)
a
ASSOCIATED CONTENT
* Supporting Information
■
a
S
Defined as the Gibbs energy difference between the highest and the
b
lowest structures along the whole catalytic cycle, ref 93. ΔG†water bar-
riers from the most stable preceding intermediate. Stepwise process
ESI mass spectra of aqueous solutions of compound 1a. Crys-
tallographic structures and data for compounds of 1a−c. Proposed
catalytic cycle and Gibbs energy profiles for the isomerization of 2-
propen-1-ol promoted by 1b and 1c complexes. Energy profiles in
water (ΔEwater). Optimized structures of all the minima and
transition states. Details of the pKa calculations. Complete authors’
list for references 64 and 66. Cartesian coordinates and absolute
energies and Gibbs energies (in hartrees) of all the calculated
species. This material is available free of charge via the Internet at
http://pubs.acs.org.
c
with two transition states (first hydride addition, then protonation).
Ref 40.
d
CONCLUSIONS
■
In this work, we have shown that the bis-allyl ruthenium(IV)
complexes containing 1,3 azole β-N-H protic ligands [Ru(η3:
η3-C10H16)Cl2L] (C10H16 = 2,7-dimethylocta-2,6-diene-1,8-diyl)
(L = imidazole (1a), benzimidazole (1b)), N-methylimidazole
(1c), are highly efficient precatalysts for the redox isomerization
of allylic alcohols into carbonyl compounds in aqueous medium
and in the absence of base. In particular, complex 1b displays
much higher catalytic activity than 1a and 1c leading to
quantitative transformations of 1-octen-3-ol in less than 5 min.
This outstanding catalytic performance is maintained at room
temperature and most interestingly under a low catalyst loading
(0.02 mol %) reaching a quantitative yield in only 5 min (TOF =
60,000 h−1). To the best of our knowledge, this is the most
efficient catalyst reported to date in water. Moreover, catalyst 1b
can be recycled, remaining active up to seven cycles. These
results show that the presence of a N-H group is not required to
catalyze the isomerization. In this sense, the catalytic action of the
N-methylimidazole complex, where both nitrogen atoms of the
ring are substituted, is particularly appealing.
AUTHOR INFORMATION
Corresponding Author
*E-mail: jgh@uniovi.es (J.G.), agusti@klingon.uab.es (A.Ll.).
■
Funding
Financial support from the Spanish MICINN (projects CTQ2011-
23336, CTQ2006-08485/BQU, and ORFEO Consolider-Ingenio
2010 CSD2007-00006) is gratefully acknowledged.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors are grateful to the Serveis Centrals d’Instrumentacio
́
́
Cientifica (SCIC) of the Universitat Jaume I for providing us
with mass spectrometry facilities. CESCA is acknowledged for
providing computational resources.
DFT calculations give evidence that the hydroxo-complexes
derived from 1a−c species can be formed in aqueous solution.
The theoretical study has also addressed the mechanism of action
of a bifunctional catalyst in water in which the uncoordinated
N-H group of imidazole remains unchanged, acting as a spectator
ligand during the whole catalytic cycle. In this mechanism the
hydroxo complex is the catalytic active species, acting as co-
operative ligand in the same way that α-N-H groups do in mech-
anisms of conventional bifunctional catalysis. Our results demon-
strate that in water the participation of the β-N-H protic group of
the 1,3-azole ligands in the bifunctional catalysis is not required,
provided that a water molecule could play the same role.
REFERENCES
■
(1) Van der Drift, R. C.; Bouwman, E.; Drent, E. J. Organomet. Chem.
2002, 650, 1−24.
́ ́
(2) Uma, R.; Crevisy, C.; Gree, R. Chem. Rev. 2003, 103, 27−51.
(3) Cadierno, V.; Crochet, P.; Gimeno, J. Synlett 2008, 1105−1124.
(4) For more recent reviews see: Mantilli, L.; Mazet, C. Chem. Lett.
2011, 40, 341−344 , and refs 5, 6.
(5) Ahlsten, N.; Bartoszewicz, A.; Martín-Matute, B. Dalton Trans.
2012, 1660−1670.
(6) Lorenzo-Luis, P.; Romerosa, A.; Serrano-Ruiz, M. ACS Catal. 2012,
2, 1079−1086.
2097
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
(7) Joo,
The Netherlands, 2001.
́
F. In Aqueous Organometallic Catalysis; Kluwer: Dordrecht,
(43) Salzer, A.; Bauer, A.; Podewils, F. In Synthetic Methods of
Organometallic and Inorganic Chemistry; Herrmann, W. A., Ed.; Thieme
Verlag: Stuttgart, Germany, 2000; Vol. 9, p 36.
(8) Aqueous Phase Organometallic catalysis; Cornils, B., Herrmann, W.
A., Eds.; Wiley-VCH: Weinheim, Germany, 2004.
(9) Multiphase Organometallic Catalysis; Cornils, B., Herrmann, W. A.,
Vogt, I., Eds.; Wiley-VCH: Weinheim, Germany, 2005.
(10) Li, C.-J. Chem. Rev. 2005, 105, 3095−3165.
(11) Chen, L.; Li, C.-J. Adv. Synth. Catal. 2006, 348, 1459−1184.
(12) Organic Reactions in Water; Lindstrom, U. M., Ed.; Blackwell
Publishing: Oxford, U.K., 2007.
(44) Salzer, A.; Bauer, A.; Geyser, S.; Podewils, F.; Turpin, G. C.; Ernst,
R. D. Inorg. Synth. 2004, 34, 59−65.
(45) Bennett, M. A.; Smith, A. K. J. Chem. Soc., Dalton Trans. 1974,
233−241.
(46) Bennett, M. A.; Huang, T. N.; Matheson, T. W.; Smith, A. K.
Inorg. Synth. 1982, 21, 74−78.
̈
(47) Çetinkaya, B.; Ozdemir, I.; Bruneau, C.; Dixneuf, P. H. Eur. J.
Inorg. Chem. 2000, 29−32.
(13) McGrath, D. V.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc.
(48) Vock, C. A.; Scolaro, C.; Phillips, A. D.; Scopelliti, R.; Sava, G.;
Dyson, P. J. J. Med. Chem. 2006, 49, 5552−5561.
1991, 113, 3611−3613.
(14) Karlen, T.; Ludi, A. Helv. Chim. Acta 1992, 75, 1604−1605.
(15) Bricout, H.; Monflier, E.; Carpentier, J. F.; Mortreux, A. Eur. J.
Inorg. Chem. 1998, 1739−1744.
(49) Collect, data collection software; Bruker AXS: Delft, The
Netherlands, 2004.
(50) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307−
326.
(51) Blessing, R. H. Acta Crystallogr. 1995, A51, 3−58.
(52) CrysAlisPro CCD, CrysAlisPro RED; Oxford Diffraction Ltd.:
Abingdon, Oxfordshire, U.K., 2008.
(16) Schumann, H.; Ravindar, V.; Meltser, L.; Baidossi, W.; Sasson, Y.;
Blum, J. J. Mol. Catal. A 1997, 118, 55−61.
(17) de Bellefon, C.; Caravieilhes, S.; Kuntz, E. G. C. R. Acad. Sci., Ser.
IIc: Chim. 2000, 3, 607−614.
(18) Bianchini, C.; Meli, A.; Oberhauser, W. New J. Chem. 2001, 25,
11−12.
(53) WINGX; Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837−838.
(54) Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.;
García-Granda, S.; Gould, R. O.; Smits, J. M. M.; Smykalla, C. The
DIRDIF Program System; Technical Report of the Crystallographic
Laboratory; University of Nijmegen: Nijmegen, The Netherlands, 1999.
(55) SIR-92; Altomare, A.; Cascarano, G.; Giacovazzo, C.; Gualardi, A.
J. Appl. Crystallogr. 1993, 26, 343−350.
(19) Knight, D. A.; Schull, T. L. Synth. Commun. 2003, 33, 827−831.
́
(20) Fekete, M.; Joo, F. Catal. Commun. 2006, 7, 783−786.
(21) Campos-Malpartida, T.; Fekete, M.; Joo,
́ ́
F.; Katho, A.; Romerosa,
, W. J. Organomet. Chem. 2008, 693, 468−474.
A.; Saoud, M.; Wojtkow
(22) Ahlsten, N.; Lundberg, H.; Martín-Matute, B. Green Chem. 2010,
́
12, 1628−1633.
(56) Sheldrick, G. M. SHELXL97, Program for the Refinement of Crystal
(23) Liu, P. N.; Ju, K. D.; Lau, C. P. Adv. Synth. Catal. 2011, 353, 275−
Structures; University of Gottingen: Gottingen, Germany, 1997.
̈
̈
280.
(57) Tables for X-Ray Crystallography; Kynoch Press: Birmingham,
U.K., 1974; Vol. IV (present distributor: Kluwer Academic Publishers;
Dordrecht, The Netherlands).
(24) Cadierno, V.; Crochet, P.; García-Garrido, S. E. Dalton Trans.
2004, 3635−3641.
(25) Cadierno, V.; García-Garrido, S. E.; Gimeno, J.; Varela-Alvarez,
́
(58) PARST; Nardelli, M. Comput. Chem. 1983, 7, 95−98.
(59) Spek, A. L. PLATON: A Multipurpose Crystallographic Tool;
University of Utrecht: Utrecht, The Netherlands, 2007.
(60) Becke, A. D. J. Chem. Phys. 1993, 98, 1372−1377.
(61) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(62) Lee, C.; Parr, R. G.; Yang, W. Phys. Rev. B 1988, 37, 785−789.
(63) Stephens, P. J.; Devlin, J. F.; Chabalowski, C. F.; Frisch, M. J. J.
Phys. Chem. 1994, 98, 11623−11627.
A.; Sordo, J. A. J. Am. Chem. Soc. 2006, 128, 1360−1370.
́
(26) Crochet, P.; Díez, J.; Fernandez-Zumel, M. A.; Gimeno, J. Adv.
Synth. Catal. 2006, 348, 93−100.
́
(27) Díaz-Alvarez, A. E.; Crochet, P.; Zablocka, M.; Duhayon, C.;
Cadierno, V.; Gimeno, J.; Majoral, J. P. Adv. Synth. Catal. 2006, 348,
1671−1679.
́
(28) Crochet, P.; Fernandez-Zumel, M. A.; Gimeno, J.; Scheele, M.
Organometallics 2006, 25, 4846−4849.
(64) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, J. A.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K. ; Fukuda, R.; Hasegawa,
J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.;
Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q. ; Baboul, A. G. ;
Clifford, S. ; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03, Revision C.02; Gaussian, Inc.: Pittsburgh, PA, 2004.
(65) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−241.
(66) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.;
Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.;
Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, L.; Hada, M.; Ehara,
M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.;
Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.;
Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
(29) Lastra-Barreira, B.; Díez, J.; Crochet, P. Green Chem. 2009, 11,
1681−1686.
(30) For other ruthenium catalized processes in aqueous medium see
refs 30−38 and 3: Cadierno, V.; García-Garrido, S. E.; Gimeno, J. Chem.
Commun. 2004, 232−233.
(31) Cadierno, V.; García-Garrido, S. E.; Gimeno, J.; Nebra, N. Chem.
Commun. 2005, 4086−4088.
(32) Cadierno, V.; García-Garrido, S. E.; Gimeno, J. J. Am. Chem. Soc.
2006, 128, 15094−15095.
(33) Cadierno, V.; Francos, J.; Gimeno, J. N. Chem. Commun. 2007,
2536−2538.
(34) Cadierno, V.; Gimeno, J.; Nebra, N. Chem.Eur. J. 2007, 13,
6590−6594.
́
(35) Díaz-Alvarez, A. E.; Crochet, P.; Zablocka, M.; Duhayon, C.;
Cadierno, V.; Majoral, J.-P. Eur. J. Inorg. Chem. 2008, 786−791.
(36) Cadierno, V.; Crochet, P.; Francos, J.; García-Garrido, S. E.;
Gimeno, J.; Nebra, N. Green Chem. 2009, 11, 1992−2000.
(37) Cadierno, V.; Francos, J.; Gimeno, J. Green Chem. 2010, 12, 135−
142.
(38) Cadierno, V.; Francos, J.; Gimeno, J. Chem. Commun. 2010, 46,
4175−4177.
(39) Díez, J.; Gimeno, J.; Merino, I.; Rubio, E.; Suar
́
ez, F. J. Inorg. Chem.
2011, 50, 4868−4881.
(40) Bellarosa, L.; Díez, J.; Gimeno, J.; Lledos
́
, A.; Suarez, F. J.; Ujaque,
́
G.; Vicent, C. Chem.Eur. J. 2012, 18, 7749−7765.
(41) Noyori, R.; Hashiguchi, S. Acc. Chem. Res. 1997, 30, 97−102.
(42) Porri, L.; Gallazzi, M. C.; Colombo, A.; Allegra, G. Tetrahedron
Lett. 1965, 47, 4187−4189.
2098
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099

ACS Catalysis
Research Article
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich,
S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J. Gaussian 09, Revision A.1; Gaussian, Inc.: Wallingford, CT,
2009.
(67) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270−283.
(68) Ehlers, A. W.; Boehme, M.; Dapprich, S.; Gobbi, A.; Hoellwarth,
A.; Jonas, V.; Koehler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 111−114.
(69) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin. Trans. 1993, 2,
̈
̈
799−805.
(70) Andzelm, J.; Kolmel, C.; Klamt, A. J. Chem. Phys. 1995, 103,
̈
9312−9320.
(71) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995−2001.
(72) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem.
2003, 24, 669−681.
(73) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(74) Chen, P. Angew. Chem., Int. Ed. 2003, 42, 2832−2847.
(75) Takats, Z.; Nanita, S. C.; Cooks, R. G. Angew. Chem., Int. Ed. 2003,
42, 3521−3523.
(76) Santos, L. S.; Metzger, J. O. Angew. Chem., Int. Ed. 2006, 45, 977−
981.
(77) Guo, H.; Qian, R.; Liao, Y.; Ma, S.; Guo, Y. J. Am. Chem. Soc. 2005,
127, 13060−13064.
(78) Toerien, J. G.; van Rooyen, P. H. J. Chem. Soc., Dalton Trans. 1991,
1563−1568.
(79) Steed, J. W.; Tocher, D. A. J. Chem. Soc., Dalton Trans. 1992,
2765−2773.
(80) Jones, C. J.; McCleverty, J. A.; Rothin, A. S. J. Chem. Soc., Dalton
Trans. 1986, 109−111.
(81) García, M. P.; Portilla, A.; Oro, L. A.; Foces-Foces, C.; Cano, F. H.
J. Organomet. Chem. 1987, 322, 111−120.
(82) For a general review on the coordination chemistry and catalytic
applications of complex 1, see: Cadierno, V.; Crochet, P.; García-
Garrido, S. E.; Gimeno, J. Curr. Org. Chem. 2006, 10, 165−183.
(83) The distance and angle values found in 1a for H···Cl1i and N2-H-
Cl1i of 2.66 Å and 148°, respectively, and in 1b for H···Cl2i and N2-
H···Cl2i of 2.46 Å, and 158°, respectively, suggest the presence of an
intermolecular weak hydrogen interaction.
(84) The analogous ruthenium(II) pyrazole-precatalysts also show a
lower catalytic activity (see ref 40).
(85) Complex 1a is also active at 35 °C (60 min; 99%).
(86) It is interesting to note that the catalyst is scarcely soluble in water,
the reaction proceeding in a biphasic manner.
(87) Activity probably being dependent on the different miscibility of
the alcohols in water.
(88) Tissandier, M. D.; Cowen, K. A.; Feng, W. Y.; Gundlach, E.;
Cohen, M. H.; Earhart, A. D.; Coe, J. V.; Tuttle, T. R. J. Phys. Chem. A
1998, 102, 7787−7794.
(89) Pliego, J. R., Jr.; Riveros, J. M. Phys. Chem. Chem. Phys. 2002, 4,
1622−1627.
(90) Kovac
Organometallics 2010, 29, 5121−5131.
́ ́
s, G.; Rossin, A.; Gonsalvi, L.; Lledos, A.; Peruzzini, M.
(91) In the cis-hydroxo route of the benzilimidazole complex the
acrolein hydrogenation is stepwise, hydride transfer preceding proton
transfer. However, the two transition states located have very similar
energies and are separated by a very shallow minimum (see ref 92).
(92) The asynchronicity in ketone hydrogenation by bifunctional
ruthenium catalysts has been recently analyzed: Chen, Y.; Tang, Y.; Lei,
M. Dalton Trans. 2009, 2359−2364.
(93) Described as the energetic span of the catalytic cycle: Kozuch, S.;
Martin, J. M. L. ChemPhysChem 2011, 12, 1413−1418.
(94) Stabilization of the transition state of asymmetric transfer
hydrogenation of ketones by hydrogen bonding has been reported: Wu,
X.; Liu, J.; Di Tommaso, D.; Iggo, J. A.; Catlow, C. R. A.; Bacsa, J.; Xiao,
J. Chem.Eur. J. 2008, 14, 7699−7715.
2099
dx.doi.org/10.1021/cs300369j | ACS Catal. 2012, 2, 2087−2099