Selective hydrogenation of cinnamaldehyde and other αβ- Unsaturated sustrates catalyzed by rhodium and ruthenium complexes

Article abstract of DOI:10.1021/om900223p

The complexes [Rh(PhBP₃)(cod)] (1) and [{Ru(PhBP ₃)(μ-CI)}₂] (8), containing the tripodal phosphanoborate ligand [PhB(CH₂pph₂)₃] ^-, are efficient catalysts for the selective hydr

Full text of DOI:10.1021/om900223p

Organometallics 2009, 28, 3193–3202
3193
Selective Hydrogenation of Cinnamaldehyde and Other
r,ꢀ-Unsaturated Substrates Catalyzed by Rhodium and Ruthenium
Complexes
Sonia Jime´nez,†,‡ Jose´ A. Lo´pez,‡ Miguel A. Ciriano,‡ Cristina Tejel,*,‡ Alberto Mart´ınez,†
and Roberto A. Sa´nchez-Delgado*,†
Instituto de Ciencia de Materiales de Arago´n (ICMA), Departamento de Qu´ımica Inorga´nica,
CSIC-UniVersidad de Zaragoza, E-50009 Zaragoza, Spain, and Chemistry Department, Brooklyn College
and The Graduate Center, The City UniVersity of New York, 2900 Bedford AVenue,
Brooklyn, New York 11210
ReceiVed March 24, 2009
The complexes [Rh(PhBP3)(cod)] (1) and [{Ru(PhBP3)(µ-Cl)}2] (8), containing the tripodal phospha-
noborate ligand [PhB(CH2PPh2)3]-, are efficient catalysts for the selective hydrogenation of cinnamaldehyde
to the corresponding allyl alcohol. Complex 8 is one of the most efficient catalysts reported to date for
this reaction, in terms of activity (TOF 527 h-1) and selectivity (g97%) under mild reaction conditions
(6.8 atm H2, 75 °C). The rhodium system also displays good catalytic features in the hydrogenation of
cinnamaldehyde (TOF 219 h-1), particularly a high selectivity (81%) for this metal in the reduction of
the CdO bond. Crotonaldehyde can also be reduced, although the selectivities are not as high as for
cinnamaldehyde; 2-cyclohexenone is rapidly and specifically reduced to cyclohexanone by both catalysts.
The ruthenium catalyst 8 operates via heterolytic activation of hydrogen, involving monohydride
intermediates and possibly ionic hydrogen transfer, while the rhodium complex 1 involves oxidative
addition of dihydrogen to form dihydride intermediates and follows a substrate route. Indeed, complex
1 reacts with hydrogen in acetonitrile to give the dihydride complex [Rh(PhBP3)(H)2(NCMe)] (3), while
protonation of one of the phosphane arms of the ligand occurs on treatment of complex 1 with HBF4 to
give the cationic species [Rh{PhB(PH)P2}(cod)]BF4. The hydride ligands in 3 are easily removed as
molecular hydrogen by reaction with CO under atmospheric pressure to give the rhodium(I) complex
[Rh(PhBP3)(CO)2]. In this reaction, the replacement of acetonitrile by CO takes place previously to the
elimination of hydrogen, which indicates that substrates can coordinate to the metal in 3 by replacement
of the labile acetonitrile ligand. Under an atmosphere of argon, complex 3 reacts with chloroform to give
an equimolecular mixture of the cis and trans isomers of [{Rh(PhBP3)(H)(µ-Cl)}2] and, eventually, complex
[Rh(PhBP3)Cl2] in one day.
Scheme 1. Possible Hydrogenation Pathways for trans-
Introduction
Cinnamaldehyde
The chemoselective reduction of R,ꢀ-unsaturated aldehydes
is of considerable importance in the synthesis of fine chemicals,
particularly intermediates for the fragrance and pharmaceutical
industries.1 The possible reaction pathways and products that
are available in the hydrogenation of trans-cinnamaldehyde are
depicted in Scheme 1; although the unsaturated alcohol (cin-
namyl alcohol) is highly desirable, its formation is often
accompanied by the products of CdC bond hydrogenation
(hydrocinnamaldehyde) and/or total reduction (3-phenylpro-
panol). Selective reduction of the CdO bond can be brought
about stoichiometrically with boron and aluminum hydrides,2
* To whom correspondence should be addressed. E-mail: rsdelgado@
brooklyn.cuny.edu; ctejel@unizar.es.
† The City University of New York.
‡ ICMA, CSIC-Universidad de Zaragoza.
but such methods are inadequate for large-scale applications,
particularly in view of the increasing demand for greener
processes involving more atom-economic catalytic reactions and
more straightforward workup procedures. Selective catalytic
hydrogenation with molecular hydrogen is a more attractive
synthetic method in terms of atom economy, but reduction of
CdC bonds is often thermodynamically favored over CdO
reduction in the intramolecular competition between olefinic and
(1) (a) Ullman’s Encyclopedia of Technical Chemistry, Electronic
Release, 6th ed.; Wiley-VCH: Weinheim, 2000. (b) Kraft, P.; Bajgrowicz,
J. A.; Denis, C.; Fra´ter, G. Angew. Chem., Int. Ed. 2000, 39, 2980–3010.
(c) Bauer, K.; Garbe, D.; Surburg, H. Common Fragrance and FlaVor
Materials: Preparation, Properties and Uses, 3rd ed.; Wiley-VCH: Wein-
heim, 1977.
(2) (a) Fukuzawa, S. I.; Fujinami, T.; Ymauchi, S.; Sakai, S. J. Chem.
Soc., Perkin Trans. 1986, 1929–1932. (b) Gemal, A. L.; Luche, J. L. J. Am.
Chem. Soc. 1981, 103, 5454–5459.
10.1021/om900223p CCC: $40.75
2009 American Chemical Society
Publication on Web 05/11/2009

3194 Organometallics, Vol. 28, No. 11, 2009
Jime´nez et al.
carbonyl groups,3 and consequently hydrogenation of the CdC
bond is often achieved under mild conditions with high
specificity.4,5
Nevertheless, the design of suitable catalysts can open
kinetically viable pathways for the hydrogenation of the carbonyl
group in R,ꢀ-unsaturated carbonyl compounds, and therefore
the search for new selective systems remains an important
challenge. Reduction with molecular hydrogen may be achieved
by use of heterogeneous catalysts,4,6 which are not always
selective enough or tolerant to functional groups. Soluble
transition metal complexes are known to promote the chemose-
lective reduction of R,ꢀ-unsaturated aldehydes under hydrogen
transfer conditions with primary or secondary alcohols serving
as hydrogen donors;7,8 some examples are known of homoge-
neous catalysts promoting the preferential formation of unsatur-
ated alcohols by use of molecular hydrogen,9-11 perhaps the
most notable being Noyori’s ruthenium-based systems including
diamino and diphosphine ligands,8 water-soluble catalysts
containing ruthenium and sulfonated triphenylphosphine,12 and
bis-phosphine ruthenium(II) arene complexes.13 Nevertheless,
the development of new selective catalysts capable of promoting
this important transformation under mild reaction conditions
continues to be an interesting challenge.
Figure 1. Structure (ORTEP at 50% level) of complex 1 (only the
ipso carbons from the phenyl groups are shown for clarity).
phosphane ligands, including neutral tripodal ligands such as
[CH3C(CH2PPh2)3] and [PhP(CH2PPh2)3],18,19 to our knowledge
no applications of “PhBP3M” systems in catalysis have been
previously reported.
The anionic tripodal phosphanoborate ligand [PhB(CH2-
PPh2)3]- (henceforth referred to as PhBP3) displays interesting
structural and electronic properties that have been exploited in
the synthesis of new complexes of Fe, Ru, Co, Ir, Rh, and Ni
with interesting reactivity.14-17 In particular, iridium complexes
bearing PhBP3 can activate C-H and Si-H bonds;16 the unusual
ground state geometry of the pseudotetrahedral complex
[Co(PhBP3)I] can correlate with novel modes of reactivity15 that
could be exploited in unusual catalytic transformations. The
chemistry of complexes of rhodium with this anionic type of
ligand has been scarcely explored.16a,17 Although a good number
of hydrogenation catalysts are based on group VIII metals with
In this paper we describe the synthesis and some reactions
of rhodium complexes with the anionic PhBP3 ligand as well
as the regioselective hydrogenation of trans-cinnamaldehyde by
[{Ru(PhBP3)(µ-Cl)}2] and [Rh(PhBP3)(cod)], to yield prefer-
entially the allylic alcohol. The effects of various reaction
parameters on the hydrogenation activity and selectivity are also
discussed. Finally, applications to other R,ꢀ-unsaturated carbonyl
compounds (crotonaldehyde, 2-cyclohexenone) are also examined.
Results and Discussion
Synthesis and Reactivity Studies of [Rh(PhBP3)(cod)]
(1). Complex 1 was prepared straightforwardly in high yield
by metathesis of [{Rh(µ-Cl)(cod)}2] (cod ) 1,5-cyclooctadiene)
with [Li(tmen)][PhBP3] (tmen ) N,N,N′,N′-tetramethylethane-
1,2-diamine) in dichloromethane. The X-ray structure of 1
(Figure 1) shows a pentacoordinated rhodium center bonded to
the three phosphorus atoms of the anion [PhBP3]- and a
chelating 1,5-cyclooctadiene ligand. The geometry around the
rhodium can be described as a distorted trigonal bipyramid with
(3) Mertens, P. G. N.; Poelman, H.; Ye, X.; Vankelecom, I. F. J.; Jacobs,
P. A.; De Vos, D. E. Catal. Today 2007, 122, 352–360.
(4) (a) Rylander, P. N. Hydrogenation Methods; Academic Press:
London, 1985; p 70. (b) Rylander, P. N. Catalytic Hydrogenation in Organic
Synthesis; Academic Press: London, 1979; p 74.
(5) (a) Nuithitikul, K.; Winterbottom, M. Chem. Eng. Sci. 2004, 59,
5439–5447. (b) Lo´pez-Linares, F.; Agrifoglio, G.; Labrador, A.; Karam,
A. J. Mol. Catal. A: Chem. 2004, 207, 117–122. (c) Kostas, I. D. J.
Organomet. Chem. 2001, 634, 90–98.
(6) (a) Serrano-Ruiz, J. C.; Lo´pez-Cudero, A.; Solla-Gullo´n, J.; Sepu´lveda-
Escribano, A.; Aldaz, A.; Rodriguez-Reinoso, F. J. Catal. 2008, 253, 159–
166. (b) Meric, P.; Yu, K. M. K.; Kong, A. T. S.; Tsang, S. C. J. Catal.
2006, 237, 330–336. (c) Ma¨ki-Arvela, P.; Ha´jek, J.; Salmi, T.; Murzin, D. Y.
Appl. Catal. A: Gen. 2005, 292, 1–49. (d) Zhao, F.; Ikushima, Y.; Chatterjee,
M.; Shirai, M.; Arai, M. Green Chem. 2003, 5, 76–79. (e) Yu, W.; Lui, H.;
Liu, M.; Liu, Z.; Milone, C. React. Funct. Polym. 2000, 44, 21–29, and
references therein.
(7) (a) Farnetti, E.; Nardin, G.; Graziani, M. J. Chem. Soc., Chem.
Commun. 1989, 1264–1265. (b) Visintin, M.; Spogliarich, R.; Kaspar, J.;
Graziani, M. J. Mol. Catal. 1985, 32, 349–351.
(8) (a) Ohkuma, T.; Utsumi, N.; Tsutsumi, K.; Murata, K.; Sandoval,
C.; Noyori, R. J. Am. Chem. Soc. 2006, 128, 8724–8725. (b) Noyori, R.;
Ohkuma, T. Angew. Chem., Int. Ed. 2001, 40, 40–73.
(14) (a) Hou, H.; Rheingold, A. L.; Kubiak, C. P. Organometallics 2005,
24, 231–233. (b) Thomas, J. C.; Peters, J. C. Polyhedron 2004, 23, 489–
497. (c) MacBeth, C. E.; Thomas, J. C.; Betley, T. A.; Peters, J. C. Inorg.
Chem. 2004, 43, 4645–4662. (d) Brown, S. D.; Peters, J. C. J. Am. Chem.
Soc. 2004, 126, 4538–4539. (e) Betley, T. A.; Peters, J. C. J. Am. Chem.
Soc. 2004, 126, 6252–6254. (f) Jenkins, D. M.; Peters, J. C. J. Am. Chem.
Soc. 2003, 125, 11162–11163. (g) Betley, T. A.; Peters, J. C. Inorg. Chem.
2003, 42, 5074–5084. (h) Betley, T. A.; Peters, J. C. J. Am. Chem. Soc.
2003, 125, 10782–10783. (i) Jenkins, D. M.; Betley, T. A.; Peters, J. C.
J. Am. Chem. Soc. 2002, 124, 11238–11239. (j) Brown, S. D.; Betley, T. A.;
Peters, J. C. J. Am. Chem. Soc. 2003, 125, 322–323.
(15) Jenkins, D. M.; Di Bilio, A. J.; Allen, M. J.; Betley, T. A.; Peters,
J. C. J. Am. Chem. Soc. 2002, 124, 15336–15350.
(9) Clarke, M. L.; Roff, G. J. In Handbook of Homogeneous Hydrogena-
tion; de Vries, J. G., Elsevier, C. J., Eds.; Wiley-VCH: Weinheim, 2007;
Vol. 1, p 413.
(10) Li, R.-X.; Wong, N.-B.; Li, X.-J.; Mak, T. C. W.; Yang, Q.-C.;
Tin, K.-C. J. Organomet. Chem. 1998, 571, 223–229.
(16) (a) Turculet, L.; Feldman, J. D.; Tilley, T. D. Organometallics 2004,
23, 2488–2502. (b) Feldman, J. D.; Peters, J. C.; Tilley, T. D. Organome-
tallics 2002, 21, 4050–4064. (c) Feldman, J. D.; Peters, J. C.; Tilley, T. D.
Organometallics 2002, 21, 4065–4075. (d) Peters, J. C.; Feldman, J. D.;
Tilley, T. D. J. Am. Chem. Soc. 1999, 121, 9871–9872.
(11) (a) Hotta, K. J. Mol. Catal. 1985, 29, 105–107. (b) Farnetti, E.;
Kasˇpar, J.; Spogliarich, R.; Graziani, M.; Pesce, M. J. Chem. Soc., Chem.
Commun. 1986, 746–747.
(12) (a) Herna´ndez, M.; Kalck, Ph. J. Mol. Catal. A: Chem. 1997, 116,
131–146. (b) Grosselin, J. M.; Mercier, C.; Allmang, G.; Grass, F.
Organometallics 1991, 10, 2126–2133. (c) Joo´, F. Acc. Chem. Res. 2002,
35, 738–745.
(17) Tejel, C.; Ciriano, M. A.; Jime´nez, S.; Passarelli, V.; Lo´pez, J. A.
Angew. Chem., Int. Ed. 2008, 47, 2093–2096.
(18) de Vries, J. G., Elsevier, C. J., Eds. Handbook of Homogeneous
Hydrogenation; Wiley-VCH: Weinheim, 2007; Vol. 1.
(19) (a) Hierso, J.-C.; Amardeil, R.; Bentabet, E.; Broussier, R.;
Gautheron, B.; Meunier, P.; Kalck, P. Coord. Chem. ReV. 2003, 236, 143–
206. (b) Simpson, M. C.; Cole-Hamilton, D. J. Coord. Chem. ReV. 1996,
155, 163–207.
(13) Chaplin, A. B.; Dyson, P. J. Organometallics 2007, 26, 4357–4360.

SelectiVe Hydrogenation of Cinnamaldehyde
Organometallics, Vol. 28, No. 11, 2009 3195
Table 1. Bond Distances (Å) and Angles (deg) for Complex 1a
Scheme 3. Formation of the Dicarbonyl Complex 4 through
the Dihydride Intermediate 5
Rh-P1
2.4464(10)
2.3236(10)
2.3910(10)
2.179(4)
2.153(3)
2.046(4)
2.230(3)
2.271(3)
2.140(4)
1.417(5)
1.383(5)
P1-Rh-Ct1
P2-Rh-Ct1
P3-Rh-Ct1
P1-Rh-Ct2
P2-Rh-Ct2
P3-Rh-Ct2
Ct1-Rh-Ct2
P1-Rh-P2
P1-Rh-P3
P2-Rh-P3
120.19(10)
89.36(10)
148.38(10)
103.18(10)
167.21(10)
96.95(10)
83.12(14)
89.53(3)
Rh-P2
Rh-P3
Rh-C46
Rh-C47
Rh-Ct1
Rh-C50
Rh-C51
Rh-Ct2
C46-C47
C50-C51
90.92(3)
85.04(3)
replaced by CO under atmospheric pressure to give the
rhodium(I) complex [Rh(PhBP3)(CO)2] (4) (Scheme 3). Alter-
natively, complex 4 was straightforwardly prepared from the
reaction of [{Rh(µ-Cl)(CO)2}2] with [Li(tmen)][PhBP3] (see
Experimental Section). Complex 4 was characterized according
to its analytical and spectroscopic data as having two carbonyl
groups, which give two µ(CO) bands at 1966 and 2037 cm-1
in the IR spectrum, and a triphosphane ligand coordinated in a
fac fashion. The equivalence of the phosphorus nuclei, which
give a doublet at δ 20.7 ppm (JP-Rh ) 99 Hz) in the 31P{1H}
NMR spectrum, can be again attributed to a fluxional behavior
of the pentacoordinated species.
a Ct1 and Ct2 are the middle points of C46-C47 and C50-C51,
respectively.
Scheme 2. Protonation of Compound 1 to Produce the Cationic
Complex 2
A confirmation of the way in which this reaction proceeds
was obtained by monitoring the uptake of 13CO by NMR
spectroscopy. The intermediate [Rh(PhBP3)(H)2(13CO)] (5)
(Scheme 3), resulting from the replacement of acetonitrile by
CO in a first step, was detected and characterized spectroscopi-
cally. Complex 5 subsequently eliminates hydrogen (detected
P2 and one of the CdC bonds (that formed by C50-C51) at
the apical positions (see Table 1) and two P atoms (P1 and P3)
and the other CdC bond at the equatorial plane forming
P-Rh-P angles close to 90°, probably due to steric require-
ments of the tripodal ligand. The Rh-P bond distances are in
the normal range, with the phosphorus at the apical position
(Rh-P2) being the shortest.
1
by H NMR) in the presence of CO to give the dicarbonyl
complex 4. A consequence of this reaction pathway is the
evidence that the acetonitrile ligand in complex 3 is labile, which
opens the possibility to coordinate substrates at the site provided
by replacement of acetonitrile in 3.
In solution, complex 1 was found to be a fluxional species
showing a doublet in the 31P{1H} NMR spectrum. The fluxional
behavior that makes the three phosphorus nuclei equivalent
could be attributed to an easy turnstile rotation movement,
although the dissociation of one of the phosphane arms leading
to a square-planar transient species cannot be excluded. In fact,
one of the phosphane arms of the ligand is protonated on
treatment of complex 1 with HBF4 to give the square-planar
cationic species [Rh{PhB(PH)P2}(cod)]BF4 (2) (Scheme 2). The
The structure of complex 5 depicted in Scheme 3 is consistent
with the spectroscopic features of the complex labeled with
13CO. The mutually cis equivalent hydride ligands give a high-
field resonance centered at δ -7.41 ppm in the 1H NMR
spectrum, while the phosphane arms (PB) in trans position
produce a double doublet of doublets in the 31P{1H} NMR
spectrum by coupling with rhodium, one phosphorus nucleus
(PA), and the carbonyl group. Meanwhile, the fine structure of
the signal due to the carbonyl group in the 13C{1H} NMR
spectrum is consistent with its disposition trans to one of the
phosphane arms and cis to the other two.
Complex 3 reacts readily with chloroform in minutes to
produce the replacement of hydride by chloride. Monitoring the
reaction of 3 in CDCl3, the spectra corresponding to complexes
6a and 6b (Figure 2) in equimolecular amounts along with
CHDCl2 were observed a few minutes after mixing. This
reaction proceeds further to produce the full replacement of the
hydrido ligands to give [Rh(PhBP3)Cl2] (7) quantitatively in one
day. Replacement of hydride by chloride by using a chlorocar-
bon such as carbon tetrachloride is a well-known reaction of
metal hydrides,20 but as the degree of chlorination of the
chlorocarbon decreases, this reaction is less frequent or disap-
pears. Therefore, we have studied the progress of the reaction
of 3 in CD2Cl2 since some of the catalytic studies were carried
1
phosphanium proton in 2 was clearly detected by H NMR
spectroscopy, giving rise to a doublet of triplets at δ ) 6.29
ppm due to the coupling with the phosphorus nucleus (JH-P
)
481 Hz) and two methylenic protons (JH-H ) 7.1 Hz), while
one of the two signals in the 31P{1H} NMR spectrum lacks the
coupling with the rhodium atom, as expected.
In spite of containing a coordinative and electronically
saturated metal center, complex 1 was found to be a reactive
species undergoing oxidative-addition reactions with, for
example, oxygen and hydrogen. The product of the reaction
with oxygen is the previously reported dinuclear peroxo complex
[{Rh(PhBP3)(O)2}2],17 while the reaction with hydrogen in a
coordinating solvent such as acetonitrile produces the mono-
nuclear dihydride complex [Rh(PhBP3)(H)2(NCMe)] (3). Com-
plex 3 was isolated as an air-stable yellow solid with two
equivalent cis-hydrido ligands and the anionic triphosphane
ligand coordinated in a fac mode. This structure is consistent
1
with a signal at high field in the H NMR spectrum and two
resonances (AM2X spin system, X ) 103Rh) in the 31P{1H} NMR
spectrum. Noticeably, the signal corresponding to the phospho-
rus atoms trans to the hydride ligands (PB) is shifted to high
field (see Experimental Section). Spin-lattice relaxation time
measurements for the hydride resonance of 3 at δ -7.08 ppm
gave a T1 min value of 321(16) ms (5 °C), in accordance with
the expected hydride nature of these ligands.
Although complex 3 does not eliminate hydrogen spontane-
ously, the hydride ligands are easily removed as hydrogen and
Figure 2. Cis (6a) and trans (6b) isomers of complex [{Rh(PhBP3)(µ-
Cl)(H)}2].

3196 Organometallics, Vol. 28, No. 11, 2009
Jime´nez et al.
Table 3. Hydrogenation of trans-Cinnamaldehyde Catalyzed by
[{Ru(PhBP3)(µ-Cl)}2] (8)a
[subst] [NEt3] TOF select.b
entry
solvent
[cat] (mM)
(M)
(M) (h-1
)
(%)
1
2
3
4
5
6
7
8
1,2-dichloroethane
1,2-dichloroethane
THF
ethyl acetate
toluene
THF
1,2-dichloroethane
toluene
0.6
0.6
0.6
0.6
0.6
0.2
0.2
0.2
1.33
1.33
1.33
1.33
1.33
0.88
0.88
0.88
0
31
81
77
92
90
90
95
94
51
0.6
0.6
0.6
0.6
0.3
0.3
0.3
258
203
488
448
114
70
51
a 95 °C; 13.6 atm H2. b Selectivity: % of trans-cinnamyl alcohol in
total product.
fluxional, showing a single signal in the 31P{1H} NMR spectrum.
The coupling constant JP-Rh (108 Hz) was found to be similar
to that observed for the Rh(I) derivatives (105 and 99 Hz for
complexes 1 and 4, respectively), apparently being insensitive
to the oxidation state of the metal.
Hydrogenation of trans-Cinnamaldehyde with [{Ru(PhBP3)(µ-
Cl)}2] (8). The dinuclear complex [{Ru(PhBP3)(µ-Cl)}2] (8),
previously reported by Peters,14g was found to be a very efficient
catalyst for the reduction of trans-cinnamaldehyde in a variety
of solvents at 95 °C and 13.6 atm H2, as shown by the data
contained in Table 3. A marked enhancement of the catalytic
activity (measured as TOF) is accomplished by addition of
triethylamine (compare entries 1 and 2); this is a good indication
that this catalyst operates by a mechanism involving heterolytic
hydrogen splitting, something rather common for ruthenium(II)
complexes containing chloride ligands.18
Figure 3. Structure (ORTEP at 50% level) of complex 7 (only the
ipso carbons from the phenyl groups are shown for clarity).
Table 2. Bond Distances (Å) and Angles (deg) for the Two
Independent Molecules of Complex 7
Rh1-P1 2.2332(13), 2.2342(13) P1-Rh1-Cl1 113.88(5), 121.67(5)
Rh1-P2 2.3290(14), 2.3193(13) P2-Rh1-Cl1 90.27(5), 86.82(4)
Rh1-P3 2.2729(14), 2.2647(13) P3-Rh1-Cl1 156.18(5), 151.62(5)
Rh1-Cl1 2.3715(13), 2.3646(12) P1-Rh1-Cl2 98.17(5), 96.73(5)
Rh1-Cl2 2.3460(13), 2.3497(13) P2-Rh1-Cl2 171.40(5), 171.46(5)
P3-Rh1-Cl2 90.20(5), 91.60(5)
Cl1-Rh1-Cl2 86.33(4), 86.78(4)
P1-Rh1-P2
P1-Rh1-P3
P2-Rh-P3
90.42(5), 91.45(5)
89.94(5), 86.68(5)
89.74(5), 91.34(5)
out in this solvent. The mixture of complexes 6 can be detected
by NMR in solutions of 3 in CD2Cl2, but this reaction proceeds
more slowly than in CDCl3. The half-life of 3 to give 6 is ca.
5 h, while the reaction is completed in 24 h. Moreover, the
mixture of isomers 6 is scarcely soluble in dichloromethane,
and it precipitates as formed, in such a way that the replacement
of hydride by chloride in 6 to give 7 does not take place in
dichloromethane. Spectroscopic data of 6a and 6b (see Experi-
mental Section) were found to be similar to those of the related
dicationic analogous complex [{Rh(triphos)(µ-Cl)(H)}2](BPh4)2
containing the neutral tripodal ligand triphos (MeC-
(CH2PPh2)3).21
The most striking feature of this system is the very high
selectivity toward trans-cinnamyl alcohol observed in all cases,
which seems to be favored in more dilute solutions in polar
solvents to reach values as high as 94% and 95% in 1,2-
dichloroethane and THF, respectively, although at some expense
of reaction rate (compare entries 2/7 and 3/6, Table 3). Further
optimization, together with rapid and controlled internal stirring
of the reaction mixture, led to a maximum TOF value of 527
h-1 in the hydrogenation of trans-cinnamaldehyde in 1,2-
dichloroethane under mild conditions (75 °C, 6.8 atm H2, [Ru]
) 0.32 mM; [substrate] ) 1.28 M) with 97% selectivity for
the reduction of the CdO bond. It is important to note that this
remarkable selectivity was maintained up to 100% conversion.
Such high efficiency in the hydrogenation of R,ꢀ-unsaturated
aldehydes is very seldom encountered for homogeneous transi-
tion metal catalysts, with the notable exception of Noyori’s Ru-
phosphane-diamine system, which functions very well at room
temperature and 4 atm H2, producing unsaturated alcohols with
100% selectivity, although it requires also the addition of a
KOH/i-PrOH mixture as an activator.8 A more atom-economic
and therefore more desirable approach is the selective reduction
using molecular hydrogen, for which only a few examples have
been reported. The complex [Ru(bdna)(CO)(H)2(PPh3)] (bdna
) 1,8-bis(diphenylphosphinomethyl)naphthalene) reduces citral
and trans-cinnamaldehyde to the corresponding unsaturated
alcohols with >95% selectivity under rather drastic conditions
(50 atm H2 and 70-80 °C).10 Similarly, arene-ruthenium(II)
complexes with PPh3 and P(p-tol)3 ligands reduce trans-
cinnamaldehyde selectively to trans-cinnamyl alcohol with TOF
values around 250-400 h-1 and selectivities >99% under 50
atm H2 at 50 °C.13 Ruthenium complexes of water-soluble
phosphines (tppms, tppts, triphenylphosphine monosulfonate or
trisulfonate, respectively) also reduce the CdO bond of trans-
cinnamaldehyde with selectivities over 90% in an aqueous
biphasic medium under 20 atm of H2.13 Other known catalysts
Complex [Rh(PhBP3)Cl2] (7) was fully characterized (see
Experimental Section) including an X-ray structure determina-
tion. Figure 3 shows the molecular structure of 7, while Table
2 lists selected bond distances and angles.
The rhodium atom in 7 shows an approximately square-
pyramidal coordination polyhedron with one of the phosphorus
atoms, P1, at the apical position and the other two phosphorus
atoms and the two chloride ligands at the base. The tripodal
PhBP3- ligand coordinates to the rhodium in a fac mode similar
to that described above for 1. Nevertheless, the average Rh-P
bond distance in 7 is shorter than in [Rh(PhBP3)(cod)] (1), as
expected for the higher oxidation state of the metal in the former.
Interestingly, complex 7 is mononuclear, while related com-
plexes with anionic fac ligands such as [{Rh(TpMe2)Cl(µ-
Cl)}2]22a and [{Rh(C5Me5)Cl(µ-Cl)}2]22b were found to be
dinuclear chloride-bridged species. In solution, complex 7 is
(20) Tejel, C.; Ciriano, M. A.; Millaruelo, M.; Lo´pez, J. A.; Lahoz, F. J.;
Oro, L. A. Inorg. Chem. 2003, 42, 4750–4758.
(21) Bianchini, C.; Meli, A.; Laschi, F.; Ramirez, J. A.; Zanello, P.;
Vacca, A. Inorg. Chem. 1988, 25, 4429–4435.
(22) (a) May, S.; Reinsalu, P.; Powell, J. Inorg. Chem. 1989, 19, 1582–
1589. (b) Churchill, M. R.; Julis, S. A.; Rotella, F. J. Inorg. Chem. 1977,
16, 1137–1141.

SelectiVe Hydrogenation of Cinnamaldehyde
Organometallics, Vol. 28, No. 11, 2009 3197
Table 4. Hydrogenation of trans-Cinnamaldehyde Catalyzed by
under different H2 pressures at 95 °C. Within the range 6.8-34
[Rh(PhBP3)(cod)] (1)a
atm H2 the activity increased in an approximately linear fashion
TOF select.b
with the reaction pressure (Figure 4a) up to a value of 430 h-1
,
[subst]
(M)
additive
(M)
entry
solvent
no solvent
1,2-dichloroethane
(h-1
)
(%)
while a slight increase in the selectivity for the production of
trans-cinnamyl alcohol from 76% to 83-85% was observed
(Figure 4b). A log P(H2) vs log TOF plot shows that the reaction
rate is close to first-order in hydrogen pressure within the range
studied.
1
2
3
4
5
6
7
8
7.95
1.33
0
0
150
130
51
81
64
7
12
56
69
51
1,2-dichloroethane/HBF4 1.33 0.74 × 10-3 138
1,2-dichloroethane/HBF4 1.33 1.7 × 10-2
24
42
80
47
71
1,2-dichloroethane/NEt3
ethyl acetate
THF
1.33 0.6
1.33
1.33
1.33
0
0
0
The effect of catalyst concentration on the hydrogenation
activity and selectivity is reported in Figure 5. TOF values
increase linearly with catalyst concentration up to 219 h-1 at
0.89 × 10-3 M, beyond which no further increase in the reaction
rate was observed (Figure 5a); the selectivity also increased from
71% to 81% with increasing catalyst concentration up to the
saturation point (Figure 5b). The reason for this saturation value
may be that at high concentrations the catalyst could be involved
in dimerization equilibria to form dinuclear species of the type
[{Rh(PhBP3)(µ-H)H}2], similar to iridium complexes described
by Peters.16b Such dimerization processes would not allow a
higher effective concentration of the monomeric active species
that operates in the hydrogenation cycle.
Figure 6 shows the effect of the substrate concentration on
the TOF values and the selectivity toward trans-cinnamyl
alcohol. The selectivity remained unchanged at ca. 81% within
the range of trans-cinnamaldehyde concentrations employed,
while the TOF values increase with increasing substrate
concentration up to a maximum value of 219 h-1 (at 1.3 M
trans-cinnamaldehyde and beyond).
The experimental data are suggestive of saturation kinetics,
in agreement with a rapid equilibrium being established between
the substrate and the catalyst, followed by an irreversible rate-
determining reaction with H2 (Scheme 4).
This behavior is very common for enzyme-catalyzed reactions
(Michaelis-Menten kinetics) and corresponds to the classical
substrate route that predominates in a number of homogeneous
catalytic hydrogenation reactions with other rhodium com-
plexes.18
Direct Comparison of the Properties of [Rh(PhBP3)(cod)]
(1) and [{Ru(PhBP3)(µ-Cl)}2] (8) in the Hydrogenation of
trans-Cinnamaldehyde. It is interesting to compare both catalytic
systems under similar reaction conditions. As shown in Table
5, both complexes display good reaction rates and high
selectivity for the formation of trans-cinnamyl alcohol, which
is the most valuable product of the hydrogenation. It is important
to note that the high selectivity is maintained at conversions as
high as 99% and 90% for complexes 8 and 1, respectively. At
75 °C and 6.8 atm H2 the hydrogenation proceeds with a much
higher turnover frequency for the ruthenium catalyst 8 (527 h-1)
as compared to the rhodium complex 1 (TOF ) 43 h-1). The
selectivity is also considerably higher in the case of ruthenium
(compare entries 1/2). By using the optimal reaction conditions
developed for the rhodium catalyst 1 (95 °C, 13.6 atm H2, entry
3) both the activity and selectivity are enhanced but not to the
same level as for 8.
In general, ruthenium complexes are more selective for the
reduction of the CdO bond of R,ꢀ-unsaturated aldehydes, while
rhodium catalysts tend to hydrogenate the CdC bond prefer-
entially, although a few examples are known of Rh complexes
that are moderately selective for CdO bond reduction.23 In this
study we have found that the rhodium complex 1 containing
the rigid anionic tripodal ligand [PhBP3]- displays very good
catalytic features in the hydrogenation of trans-cinnamaldehyde,
toluene
a 95 °C; 13.6 atm H2; [Rh] ) 0.74 × 10-3 M. b Selectivity: % of
cinnamyl alcohol in total product.
are in general less active, are less selective, or require consider-
ably higher temperatures and pressures to operate than
complex 8.9
Hydrogenation of trans-Cinnamaldehyde with [Rh(PhBP3)-
(cod)] (1). Table 4 displays the results of hydrogenating trans-
cinnamaldehyde with the rhodium complex [Rh(PhBP3)(cod)]
(1) in different solvents at 95 °C and 13.6 atm H2. The best
combination of activity and selectivity for CdO bond reduction
was obtained in 1,2-dichloroethane (TOF 130 h-1; 81% trans-
cinnamyl alcohol, 6% hydrocinnamaldehyde, 13% 3-phenyl-
propanol). Other polar solvents (ethylacetate, THF) produce
lower activity and selectivity, perhaps by competing with the
substrate for coordination sites. In toluene, the catalyst shows
moderate activity for the production of the unsaturated alcohol
TOF (71 h-1) and low selectivity (51%). We also performed
solventless reactions, which are generally preferred as “greener”
alternatives. In this case, although the catalytic activity is
somewhat enhanced when pure trans-cinnamaldehyde is em-
ployed (TOF 150 h-1), the selectivity for trans-cinnamyl alcohol
is drastically reduced (51%) with respect to the reaction in 1,2-
dichloroethane. The higher selectivity observed in dilute solution
suggests that the CdO bond binds preferentially to the rhodium
center, but when the pure substrate is used, coordination of the
metal to both the CdO and CdC bonds is likely to be enhanced
in a more competitive manner, thereby resulting in a selectivity
drop. Addition of triethylamine has a detrimental effect on both
the activity and selectivity of complex 1 for the hydrogenation
of trans-cinnamaldehyde, which indicates that this catalyst does
not operate through heterolytic activation of hydrogen, as is the
case of the ruthenium derivative 8. Thus, oxidative addition of
H2 seems a more plausible reaction pathway, to form an 18-e
Rh(III) catalytically active species of the type [Rh(PhBP3)(H)2L],
similar to [Rh(PhBP3)(H)2(NCMe)] (3), where L is a substrate
molecule.
Addition of 1 equiv of HBF4 to the catalytic mixture in 1,2-
dichloroethane results in a slight improvement of the activity
(TOF 138 h-1) together with a drop in the selectivity toward
cinammyl alcohol from 81% to 64%. This is due to protonation
of one of the phosphorus atoms of the PhBP3 ligand to form
[Rh{PhB(PH)P2}(cod)]BF4 (2) with an additional vacant site
in the catalytically active species, which would thereby become
less constrained, making coordination of the sterically hindered
CdC bond of cinnamaldehyde more facile. If acid in excess
over the stoichiometric amount is added to the reaction medium,
both the activity and selectivity are decreased, which suggests
the decomposition of the catalyst under strongly acidic conditions.
In an attempt to optimize the catalytic performance of
complex 1 to a level comparable to that obtained with complex
8, a systematic variation of various reaction parameters was
undertaken. Figure 4 shows the results of hydrogenating trans-
cinnamaldehyde with complex 1 in 1,2-dichloroethane solution
(23) Nomura, K. J. Mol. Catal. A: Chem. 1998, 130, 1–28.

3198 Organometallics, Vol. 28, No. 11, 2009
Jime´nez et al.
Figure 4. Effect of the hydrogen pressure on the rate and selectivity of the hydrogenation of cinnamaldehyde by complex 1 (1,2-dichloroethane
95 °C, [cat] ) 0.89 × 10-3 M, [subst] ) 1.4 M).
Figure 5. Effect of catalyst concentration on the rate and selectivity in the hydrogenation of cinnamaldehyde by complex 1 (in 1,2-
dichloroethane, 95 °C, 13.6 atm H2, [subst] ) 1.4 M).
Table 5. Catalytic Hydrogenation of trans-Cinnamaldehyde in
1,2-Dichloroethane
[M]
P (H2)
T
TOF select.
entry
catalyst
(mM) (atm) (°C) (h-1
)
(%)a
1
2
3
[{Ru(PhBP3)(µ-Cl)}2]
[Rh(PhBP3)(cod)]
[Rh(PhBP3)(cod)]
0.30
0.71
0.71
6.8
6.8
13.6
75
75
95
527
43
219
97
78
81
a Selectivity: % of cinnamyl alcohol in total product.
and selectivity (g97%) under very mild reaction conditions (6.8
atm H2, 75 °C).
Hydrogenation of Other r,ꢀ-Unsaturated Aldehydes and
Ketones with [Rh(PhBP3)(cod)] (1) and [{Ru(PhBP3)(µ-Cl)}2]
(8). The results presented above encouraged us to extend our
studies to other R,ꢀ-unsaturated aldehydes and ketones (see
Table 6). In the case of crotonaldehyde the selectivity when
using the ruthenium complex 8 is lower than the one observed
for trans-cinnamaldehyde but still attractive; the rhodium
complex 1, on the other hand, was found to be rather unselective
in the reduction of crotonaldehyde. A possible explanation for
the lower selectivities observed lies in the fact that the CdC
bond in crotonaldehyde is much less sterically constrained than
the one in trans-cinnamaldehyde, and this probably results in a
poorer discrimination of the metal between coordination of the
CdO bond against the CdC bond; electronic effects, however,
cannot be ruled out. Previous reports involving other ruthenium
and rhodium catalysts that compared cinnamaldehyde and
crotonaldehyde follow a similar trend.9
Figure 6. Effect of initial substrate concentration on the rate of
hydrogenation of trans-cinnamaldehyde by complex 1 (in 1,2-
dichloroethane, 95 °C, 13.6 atm H2, [cat] ) 0.89 × 10-3 M).
Scheme 4. Possible Steps in the Hydrogenation of
trans-Cinnamaldehyde Catalyzed by 1
particularly an unusually high selectivity for this metal in the
reduction of the CdO bond. Nevertheless, it is clear that the
ruthenium complex 8 is superior to 1 and no doubt one of
the most efficient catalysts reported to date for the hydrogenation
of trans-cinnamaldehyde, in terms of activity (TOF 527 h-1
)

SelectiVe Hydrogenation of Cinnamaldehyde
Organometallics, Vol. 28, No. 11, 2009 3199
Table 6. Hydrogenation of Other r,ꢀ-Unsaturated Substrates Catalyzed by Complexes 1 and 8
catalyst
substrate
[M] (mM)
[subst] (M)
T (°C)
P(H2) (atm)
TOF (h-1
)
select. (%)
[{Ru(PhBP3)(µ-Cl)}2]
[Rh(PhBP3)(cod)]
[{Ru(PhBP3)(µ-Cl)}2]
[Rh(PhBP3)(cod)]
crotonaldehyde
crotonaldehyde
2-cyclohexenone
2-cyclohexenone
0.41
0.74
0.40
0.74
1.14
2.03
1.68
1.75
75
95
75
95
6.8
13.6
6.8
443
74
1475
1153
68a
15b
100c
100c
6.8
a Selectivity is % of crotyl alcohol in total product. Other products were butyraldehyde (0.5%) and butanol (16.4%). b Selectivity is % of crotyl
alcohol in total product. Other products were butyraldehyde (14%) and butanol (9%). c Selectivity is % of cyclohexanone in total product.
Scheme 5. Possible Mechanistic Pathways in the Hydrogenation of trans-Cinnamaldehyde Catalyzed by Complex 8
In the case of 2-cyclohexenone, the CdO bond is not readily
reduced with either one of these two catalysts under mild
reaction conditions. Instead, both complexes rapidly and specif-
ically reduce the CdC bond to yield cyclohexanone as the sole
product of the reaction.
Ru(IV) species C and subsequent reductive elimination of the
allyl alcohol regenerates A and completes the catalytic cycle.25
Alternatively, the electron-deficient alkoxy intermediate B could
react with molecular hydrogen by a heterolytic mechanism (path
b, Scheme 5) in which the coordinated oxygen atom acts as the
base,8,12,24 as in species D. An intermolecular ionic mechanism
can also be envisaged (path c, Scheme 5), involving protonolysis
of the alkoxy intermediate B to liberate the product and a highly
reactive cationic Ru(II) intermediate E, capable of rapidly
interacting with dihydrogen via heterolytic splitting (aided by
triethylamine) to regenerate A and a proton that would be
captured by triethylamine.24 All these possibilities have prece-
dent in the chemistry of ruthenium, and any one of them could
be consistent with our data; at present, we do not have enough
information to distinguish a preferred pathway.
In the case of the rhodium catalyst the experimental data point
to a dihydride active species and a classical “substrate mech-
anism” (Scheme 6); this is in agreement with the saturation
kinetics observed with respect to substrate concentration, the
close to first-order dependence on the hydrogen pressure, and
a wealth of knowledge on rhodium-catalyzed hydrogenation
reactions.26
Mechanistic Considerations. Although the experimental data
available do not allow us to accurately describe the mechanistic
details of the hydrogenation of trans-cinnamaldehyde with
complexes 1 and 8, it is interesting to consider the possible
routes available to each catalyst that are consistent with the
experimental results. In the case of ruthenium, it is clear that
the presence of a base is required for the activation of the
catalyst, which points to a heterolytic splitting of dihydrogen
together with a bridge cleavage reaction to produce a reactive
14-electron monomeric monohydride active species A, which
enters the catalytic cycle depicted in Scheme 5.24
Interaction of the substrate with the metal center probably
takes place preferentially through the CdO bond, in accord with
the high selectivity observed, particularly for trans-cinnamal-
dehyde; in the case of the less sterically demanding crotonal-
dehyde, coordination of the CdC bond is not as disfavored.
Stepwise or concerted hydride transfer to the CdO bond leads
to the alkoxide intermediate B, which may generate the product
by several reaction pathways: In a “classical” mechanism (path
a, Scheme 5), oxidative addition of hydrogen to generate the
The high selectivity for the reduction of trans-cinnamaldehyde
again suggests that the substrate must likely bind to the metal
(25) Sa´nchez-Delgado, R. A.; Rosales, M. Coord. Chem. ReV. 2000,
196, 249–280.
(24) Stanley, G. G. ComprehensiVe Organometallic Chemistry III;
Crabtree, R. H.; Mingos, D. M., Eds.; Elsevier: New York, 2007; Vol. 1,
pp 119-140.
(26) Oro, L. A.; Carmona, D. In Handbook of Homogeneous Hydro-
genation; de Vries, J. G., Elsevier, C. J., Eds.; Wiley-VCH: Weinheim,
2007; Vol. 1, pp 3-30.

3200 Organometallics, Vol. 28, No. 11, 2009
Jime´nez et al.
Scheme 6. Possible Mechanism for the Hydrogenation of
case of trans-cinnamaldehyde. 2-Cyclohexenone was rapidly
and specifically reduced at the CdC bond with both systems.
The ruthenium catalyst 8 operates via heterolytic activation
of hydrogen, involving monohydride intermediates and possibly
ionic hydrogen transfer, while the rhodium complex 1 follows
a substrate route and involves oxidative addition of dihydrogen
to form dihydride intermediates.
trans-Cinnamaldehyde Catalyzed by 1
Experimental Section
Starting Materials and Physical Methods. All the operations
were carried out under an argon atmosphere using standard Schlenk-
tube technique. Solvents were purified by standard procedures.
[Li(tmen)][PhB(CH2PPh2)3],33 [{Rh(µ-Cl)(cod)}2],34 and [{Ru(Ph-
BP3)(µ-Cl)}2]14g were prepared according to the literature methods.
HBF4 (Aldrich, 54% w/w in Et2O) was commercially available and
was used as received. Elemental analyses were performed using a
Perkin-Elmer 2400 microanalyzer. Mass spectra (MALDI) were
recorded with a Bruker MicroFlex spectrometer using DCTB (1,1-
dicyano-4-tert-butylphenyl-3-methylbutadiene) as matrix and in a
VG Autospec double-focusing mass spectrometer operating in the
FAB+ mode. NMR spectra were recorded on a Bruker AV 300
through the CdO bond; although an end-on coordination is
represented for species F in Scheme 6, η2-CdO side binding is
also possible. A less sterically demanding situation, as in the
case of crotonaldehyde or of protonation of one phosphorus
atom, would lead to a loss in selectivity, as observed. Oxidative
addition of hydrogen forms the rhodium(III) intermediate G.
Subsequently, transfer of the two hydrides to the CdO bond
yields the allylic alcohol as the major product and regenerates
the catalyst. An analogous set of the reactions on a CdC
coordinated substrate would produce the saturated aldehyde,
which is always observed with the rhodium system. However,
an alternative explanation for the presence of hydrocinnamal-
dehyde in the final reaction mixture could be the isomerization
of cynnamyl alcohol, but this point was not investigated in detail.
Finally, it should be indicated that, in the absence of a hydrogen
atmosphere, complex 3 decarbonylates trans-cinnamaldehyde
to styrene probably through the intermediate F.
1
spectrometer operating at 300.13 MHz, respectively, for H. IR
spectra were recorded with a Nicolet 550 spectrophotometer.
Conductivities were measured in acetone solutions using a Philips
PW 9501/01 conductimeter.
Synthesis of the Complexes: [Rh(PhBP3)(cod)] (1). Solid
[{Rh(µ-Cl)(cod)}2] (133 mg, 0.27 mmol) was added to a solution
of [Li(tmen)][PhB(CH2PPh2)3] (436 mg, 0.54 mmol) in dichlo-
rometane (8 mL) to give an orange solution and a white solid in
1 h at rt. The solvent was evaporated under vacuum, the residue
was extracted with a 1:1 mixture of toluene/dichloromethane (30
mL), and the LiCl was removed by filtration. Evaporation of the
extract to ca. 5 mL and addition of hexane (20 mL) produced the
crystallization of 1 as a yellow solid. The solid was washed with
1
hexane (2 × 5 mL) and vacuum-dried. Yield: 380 mg, 80%. H
Conclusion
NMR (300 MHz, CDCl3, 25 °C): δ 7.61 (d, 7.6 Hz, 2H; Ho), 7.28
(t, 7.6 Hz, 2H; Hm) and 7.09 (t, 7.6 Hz, 1H; Hp) PhB, 7.19 (m, 18
H, Ho+p) and 7.06 (t, 37.3 Hz, 12 H; Hm) Ph2P, 3.70 (br s, 4H;
HCd), 2.65 (m, 4H; H2C) and 2.28 (m, 4H; H2C) cod, 1.46 (br s,
6H; H2C-P). 31P{1H} NMR (300 MHz, CDCl3, 25 °C): δ 17.8 (d,
JP-Rh ) 105 Hz). MS (Maldi): m/z (%): 788 (100) [M+ - cod].
Anal. Calcd (%) for C53H53BP3Rh: C 71.00, H 5.96. Found: C 71.03;
H 6.08.
We have demonstrated that the complexes [Rh(PhBP3)(cod)]
(1) and [{Ru(PhBP3)(µ-Cl)}2] (8), containing the anionic tripodal
phosphanoborate ligand [PhB(CH2PPh2)3]-, are efficient cata-
lysts for the regioselective hydrogenation of trans-cinnamalde-
hyde to the corresponding allyl alcohol. Complex 8 turned out
to be one of the most active and selective catalysts reported so
far for the mild hydrogenation of the CdO bond of trans-
cinnamaldehyde using molecular H2, exhibiting higher activity
and selectivity than the rhodium complex 1, provided that NEt3
is added to the reaction mixture. Side reactions that often take
place concurrently with the hydrogenation of aldehydes, such
as decarbonylation, which is promoted by a number of phos-
phane rhodium27-29 or ruthenium30 complexes, or condensation
reactions that are catalyzed by rhodium31 and ruthenium32
phosphane complexes, are not observed in this case. Crotonal-
dehyde can also be reduced with catalysts 1 and 8, although
the selectivity for the unsaturated alcohol is lower than in the
[Rh{PhB(PH)P2}(cod)]BF4 (2). HBF4 · Et2O (21 mL, 0.13 mmol)
was added to a solution of [Rh(PhBP3)(cod)] (1) (116 mg, 0.13
mmol) in dichloromethane (4 mL) to produce an orange solution.
Addition of hexane (10 mL) led to the crystallization of an orange
solid corresponding to 2. The solid was washed with hexane (2 ×
1
5 mL) and vacuum-dried. Yield: 110 mg, 87%. H NMR (CDCl3,
25 °C): δ 7.8-6.6 (35H, PhB+Ph2P), 6.29 (dt, JH-P ) 481 Hz,
JH-H ) 7.1 Hz, 1H, PH), 4.34 (br, 2H, dCH), 4.10 (br, 2H, dCH),
2.27 (m, 2H, CH2), 2.15 (m, 2H, CH2), 2.11 (br, 4H, CH2), 1.67
(m, δA, 2H) and 1.40 (m, δB, 2H, JA-B ) 12.5 Hz, CH2PRh), 0.68
(dd, JH-H ) 16 Hz, JH-P ) 7.1 Hz, 2H, CH2PH). 31P{1H} NMR
(CDCl3, 25 °C): δ 24.0 (dd, JP-Rh ) 140 Hz, JP-P ) 5 Hz, 2P),
13.2 (t, JP-P ) 5 Hz, 1P). MS (FAB+, NBA): m/z (%): 897 (53)
[M+]. Anal. Calcd (%) for C53H54B2F4P3Rh · 1.5CH2Cl2: C 58.87,
H 5.16. Found: C 58.51; H 5.25. ΛM (5.16 × 10-4 M in acetone)
(27) (a) Tsuji, J.; Ohno, K. Tetrahedron Lett. 1965, 6, 3969–3971. (b)
Ohno, K.; Tsuji, J. J. Am. Chem. Soc. 1968, 90, 99–107. (c) Tsuji, J.; Ohno,
K. J. Am. Chem. Soc. 1968, 90, 94–98.
(28) Abu-Hasanayn, F.; Goldman, M. E.; Goldman, A. S. J. Am. Chem.
Soc. 1992, 114, 2520–2524.
(29) Beck, C. M.; Rathmill, S. E.; Park, Y. J.; Chen, J.; Crabtree, R. H.;
Liable-Sands, L. M.; Rheingold, A. L. Organometallics 1999, 18, 5311–
5317.
(30) Van Der Sluys, L. S.; Kubas, G. J.; Caulton, K. G. Organometallics
1991, 10, 1033–1038.
) 69 S · mol-1 · cm-1
[Rh(PhBP3)(H)2(NCMe)] (3).
.
A
yellow suspension of
[Rh(PhBP3)(cod)] (1) (116 mg, 0.13 mmol) in acetonitrile was
stirred at 70 °C under an atmosphere of hydrogen at 3 bar overnight
(15 h) to give a white precipitate after the starting material had
(31) Burkhardt, E. R.; Bergman, R. G.; Heathcock, C. H. Organome-
tallics 1990, 9, 30–34.
(32) Ito, T.; Horino, H.; Koshiro, Y.; Yamamoto, A. Bull. Chem. Soc.
Jpn. 1982, 55, 504–512.
(33) Feldman, J. D.; Peters, J. C.; Tilley, T. D. Organometallics 2002,
21, 4050–4064.
(34) Giordano, G.; Crabtree, R. H. Inorg. Synth. 1990, 28, 88–90.

SelectiVe Hydrogenation of Cinnamaldehyde
Organometallics, Vol. 28, No. 11, 2009 3201
Table 7. Crystallographic Data for Compounds 1 · 2CH2Cl2 and 7
become a solution. The suspension was concentrated to ca. 1 mL,
and the solid was isolated by filtration under argon, washed with
cold diethyl ether (2 × 5 mL), and vacuum-dried. Yield: 86 mg,
1 · 2CH2Cl2
7
chem formula
fw
C55H57BCl4P3Rh
1066.44
C45H41BCl2P3Rh
859.31
1
80%. H NMR (C6D6, 25 °C): δ 8.17 (d, 7.2 Hz, 2H, Ho, PhB),
8.06 (m, 4H, Ho, PhB -PB), 7.94 (m, 4H, Ho, PhA2-PA), 7.68 (t, 7.2
2
cryst syst
space group
a [Å]
b [Å]
c [Å]
R [deg]
ꢀ [deg]
γ [deg]
V [Å3]
Z
triclinic
P1 (no. 2)
triclinic
P1 (no. 2)
Hz, 2H, Hm, PhB), 7.42 (t, 7.2 Hz, 1H, Hp, PhB), 7.14 (t, 7.8 Hz,
j
j
4H, Ho, PhB -PB), 6.92 (m, 6H, Hm+p, PhB -PB), 6.38 (m, 12H, Hm+p
,
12.0712(8)
13.5583(9)
15.4578(11)
91.3980(10)
94.1190(10)
103.3790(10)
2452.7(3)
2
1.444
0.702
0.8587/0.9391
1100
9.6633(7)
18.6168(13)
22.0862(15)
104.400(1)
93.131(1)
90.810(1)
3841.1(5)
4
1.486
0.742
0.7951/0.9638
1760
2
2
PhB -PB + Hm+p, PhA2-PA), 2.11 (m, δA, 2H) and 1.81 (m, δB, JA-B
2
) 12.2 Hz; CH2PB), 1.59 (d, JH-P ) 13.6 Hz; CH2PA), 0.27 (s, 3H,
B
MeCN), - 7.08 (m, JH-P ) 157 Hz, JH-Rh ) 15 Hz, JH-H ) 4
Hz, JH-P ) - 2 Hz). 31P{1H} NMR (C6D6, 25 °C): δ 53.8 (dt,
A
JP-Rh ) 129 Hz, JP-P ) 28 Hz, PA), 24.0 (dd, JP-Rh ) 79 Hz, JP-P
) 28 Hz, PB). IR (KBr, cm-1): 1946 (s), 1918 (s). Anal. Calcd (%)
for C47H46NBP3Rh: C 67.89, H 5.58, N 1.68. Found: C 67.36; H
5.80, N 1.53.
Fcalcd [g cm-3
µ(Mo KR) [mm-1
min/max transmn
F(000)
]
]
[Rh(PhBP3)(CO)2] (4). Solid [{Rh(µ-Cl)(CO)2}2] (23 mg, 0.06
mmol) was added to a solution of [Li(tmen)][PhB(CH2PPh2)3] (93
mg, 0.12 mmol) in toluene (5 mL) to give a yellow solution, from
which a yellow solid crystallized in 20 min. Toluene (10 mL) was
added to extract the solid, and then, the suspension was filtered
over kieselguhr to remove LiCl. Concentration of the filtrate and
addition of hexane gave complex 4 as yellow crystals after two
days at -25 °C. The solid was washed with cold hexane (2 × 5
cryst size [mm]
θ range [deg]
reflns collected
indep reflns
Rint
0.16 × 0.11 × 0.09 0.36 × 0.05 × 0.05
2.27-25.39
29 190
2.22-26.02
42 014
10 649
15 011
0.0449
0.0753
reflns [F2> 2σ(F2)]
data/restraints/params
R(F) [F2 > 2σ(F2)]
wR(F2) [all data]
8168
10 784
10 649/7/590
0.0478
15 011/0/937
0.0587
1
mL) and vacuum-dried. Yield: 90 mg, 93%. H NMR (C6D6, 25
0.1122
0.1149
°C): δ 8.05 (d, 7.2 Hz, 2H, Ho, PhB), 7.64 (t, 7.2 Hz, 2H, Hm,
PhB), 7.38 (m, 1 + 12H, Hp, PhB + Ho, Ph2P), 6.80 (m, 18H,
Hm+p, Ph2P), 1.75 (d, JH-P ) 10.8 Hz, 6H, CH2P). 31P{1H} NMR
(C6D6, 25 °C): δ 20.7 (d, JP-Rh ) 99 Hz). IR (Nujol, cm-1): 2037
(s), 1966 (s). MS (FAB+, NBA): m/z (%) 788 (100) [M+ - 2CO].
Anal. Calcd (%) for C47H41BO2P3Rh: C 66.85, H 4.89. Found: C
66.54; H 4.79.
GOF(F2) [all data]
0.977
1.051
largest diff peak/hole [e Å-3
]
0.886/-1.187
0.950/-0.796
liquid and washing with diethyl ether. Yield: 96 mg, 90%. 1H NMR
(C6D6, 25 °C): δ 7.63 (d, 7.2 Hz, 2H, Ho, PhB), 7.50 (dd, 12H, Ho,
Ph2P), 7.37 (t, 7.6 Hz, 2H, Hm, PhB), 7.19 (t, 4.8 Hz, 1H, Hp, PhB),
7.16 (t, 6H, 7.2 Hz, Hp, Ph2P), 6.98 (t, 12H, 7.6 Hz, Hm, Ph2P),
1.57 (d, JH-P ) 7.6 Hz, 6H, CH2P). 31P{1H} NMR (C6D6, 25 °C):
δ 52.3 (d, JP-Rh ) 108 Hz). MS (FAB+, NBA): m/z (%) 823 (100)
[M+ - Cl]. Anal. Calcd (%) for C45H41BCl2P3Rh · 0.5CHCl3: C
59.50, H 4.55. Found: C 60.02; H 4.27.
Structural Analysis of 1 · 2CH2Cl2 and 7. A summary of crystal
data and refinement parameters is reported in Table 7. X-ray data
were collected with a Bruker Smart Apex CCD diffractometer, with
graphite-monochromated Mo KR radiation (λ ) 0.71073 Å) using
ω scans (0.3°). The crystals were covered with inert oil, mounted
on glass fibers, and cooled to the data collection temperature (100
K). Data were corrected for absorption using a multiscan method
applied with the SADABS program.35 The structures were solved
by direct methods and refined by full-matrix least-squares on F2,
with the program SHELX9736 in the WINGX37 package. All non-
hydrogen atoms except those of a disordered solvent molecule were
refined with anisotropic displacement parameters. Hydrogen atoms
were included in calculated positions refined riding on the corre-
sponding atom with a isotropic displacement parameter related to
that of the bonded atom. The program ORTEP-338 was used for
diagrams.
Procedure for Catalytic Reactions. Solvents (analytical grade,
Aldrich) were dried and degassed by use of an Innovative
Technology solvent purification unit; substrates and other reagents
(Aldrich) were used as received. In preliminary experiments the
catalyst, the substrate, the solvent, any additive desired, and a
stirring bar were placed in a 4793 Parr reactor (100 mL) inside a
glovebox. The reactor was purged with hydrogen three times, then
charged to the desired pressure and immersed in an oil bath pre-
equilibrated at the desired temperature. The reaction was stopped
at the appropriate time by placing the reactor in ice; excess hydrogen
Monitoring the Reaction of Complex 3 with 13CO by NMR.
A solution of [Rh(PhBP3)(H)2(NCMe)] (3) in C6D6 was exposed
to a 13CO atmosphere for 20 min in a NMR tube. Examination of
the solution by NMR revealed the complete transformation of the
starting material into hydrogen and a mixture of the complexes
[Rh(PhBP3)(H)2(13CO)] (5) and [Rh(PhBP3)(13CO)2] (4′) in a 1:2.5
molar ratio. Further exposure to an atmosphere of 13CO gave
eventually a solution of [Rh(PhBP3)(13CO)2] and hydrogen. Selected
spectroscopic data of 5: 31P{1H} NMR (C6D6, 25 °C): δ 38.6 (ddt,
JP-C )109 Hz, JP-Rh ) 97 Hz, JP-P ) 35 Hz), 21.6 (ddd, JP-Rh
)
80 Hz, JP-P ) 35 Hz, JP-C )6 Hz). 13C{1H} NMR (300 MHz,
C6D6, 25 °C): δ 194.1 (ddt, JC-P ) 109 Hz, JC-Rh ) 53 Hz, JC-P
)6 Hz).
Reaction of Complex 3 with CD2Cl2. [Rh(PhBP3)(H)2(NCMe)]
(3) (25 mg, 0.030 mmol) was dissolved in CD2Cl2 (0.5 mL) in a
NMR tube with a small amount of silicone as internal standard.
Examination of the solution by 1H and 31P{1H} NMR revealed slow
transformation of 3 into the complexes 6a/6b. A pale yellow solid
precipitates from this solution in one day. After centrifugation of
the tube, the solvent was removed and the residue was dried under
vacuum. Yield: 18 mg, 73%. 1H NMR (C6D6, 25 °C): δ 8.11-6.58
(35H, PhB+Ph2P), 2.19 (d, JH-P ) 11.6 Hz, 2H, CH2P), 1.96 (d,
JH-P ) 11.9 Hz, 2H, CH2P), 1.89 (m, δA, 1H) and 1.70 (t, δB, JH-P
) 11.1 Hz, JA-B ) 14.6 Hz, 1H, CH2P), 1.85 (m, δA, 1H) and 1.53
(Brt, δB, JA-B ) 12.0 Hz, 1H, CH2P), -5.34 (dm, JH-P ) 205.0
Hz, 1H, H-Rh), -6.29 (dm, JH-P ) 210.1 Hz, 1H, H-Rh). 31P{1H}
NMR (C6D6, 25 °C): δ 48.9 (dd, JP-Rh ) 126 Hz, JP-P ) 22 Hz),
46.6 (dd, JP-Rh ) 124 Hz, JP-P ) 20 Hz), 1,98 (dt, JP-Rh ) 43 Hz,
JP-P ) 22 Hz), -5.61 (dt, JP-Rh ) 40 Hz, JP-P ) 20 Hz). IR (KBr,
cm-1): 2014 (m), 1995 (m). Anal. Calcd (%) for C45H42BClP3Rh ·
3CH2Cl2: C 58.63, H 4.76. Found: C 58.85; H 5.32.
(35) Sheldrick, G. M. SADABS Version 2.03; University of Go¨ttingen:
Germany, 2002.
(36) Sheldrick, G. M. SHELX97, Programs for Crystal Structure
Analysis, Release 97-2; Institu¨t fu¨r Anorganische Chemie der Universita¨t:
Go¨ttingen, Germany, 1998.
(37) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(38) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837–838.
[Rh(PhBP3)Cl2] (7). A solution of complex 3 (103 mg, 0.124
mmol) in chloroform (5 mL) was left for 1 day at room temperature.
Evaporation of the solvent to ca. 2 mL and addition of hexane
rendered complex 8 as an orange crystalline solid in two days in
the freezer. The solid was isolated by decantation of the mother

3202 Organometallics, Vol. 28, No. 11, 2009
Jime´nez et al.
was removed at room temperature, the reactor was opened, and
the products were immediately analyzed by GC.
to ensure reproducibility of the results. Average values of two or
more runs are reported in the tables and figures.
Once the desired reaction conditions were determined, further
hydrogenation experiments were performed by use of a 5513 Parr
reactor (100 mL) fitted with internal stirring and thermocouple, a
sampling valve, and a high-pressure buret and coupled to a Parr
4836 controller. The reactor was loaded with the catalyst, the
substrate, any additive desired, and the solvent, flushed three times
with hydrogen, charged to the desired pressure, and placed in an
electric oven pre-equilibrated at the desired temperature. Samples
of the reaction mixture were periodically withdrawn through the
sampling valve; the pressure was readjusted to maintain a constant
value throughout the experiment, and the samples were immediately
analyzed by GC, using a Shimadzu 2010 chromatograph with an
FID detector. Each experiment was repeated at least twice in order
Acknowledgment. We thank The Petroleum Research
Fund of the American Chemical Society (Grant #47472-AC3
to R.A.S.-D.), the MICINN/FEDER (Project CTQ2008-
03860), and Gobierno de Arago´n (ARMOIN, Group E70)
for generous financial support. S.J. thanks GA for a
fellowship.
Supporting Information Available: Full ORTEP drawings for
complexes 1 · 2CH2Cl2 and 7. Single-crystal X-ray diffraction data
in CIF format for complexes 1 · 2CH2Cl2 and 7 are available free
of charge via the Internet at http://pubs.acs.org.
OM900223P