Organometallic ruthenium complexes: Application in the olefination of carbonyl compounds

Article abstract of DOI:10.1021/om060508w

Ruthenium compounds of general formula Cp′RuX(PR₂R′) ₂ (Cp′ = η⁵-C₅H₅ (Cp), η⁵-C₉H₇ (Ind), η⁵C ₅(CH₃)₅ (Cp*); X = Cl, CF ₃C(O)O; R = C₆H₅ (Ph), C₆H ₄(CH₃) (m-tolyl); R′ = C₆H₅, C₆H₁₁ (Cy), C₆H₄(CH₃) (m-tolyl, o-tolyl)) are examined as catalysts for the aldehyde olefination starting from diazo compounds, phosphanes, and aldehydes. Cp*RuCl(PPh ₃)₂ is highly active for the olefmation of several aldehydes, displaying a very high E-selectivity, as well as for ketone olefination (with benzoic acid as cocatalyst). The reaction's mechanism is substantiated by the isolation of a catalytic active reaction species, namely, a mixed carbene/phosphane ruthenium complex, Cp*RuCl(=CHCO ₂Et)(PPh₃) (8). Spectroscopic studies reveal that the latter compound reacts with PPh₃ to produce the phosphorus ylide Ph₃P=CHCO₂Et, which further reacts with the aldehyde to produce the olefin.

Full text of DOI:10.1021/om060508w

302
Organometallics 2007, 26, 302-309
Organometallic Ruthenium Complexes: Application in the
Olefination of Carbonyl Compounds
Filipe M. Pedro,† Ana M. Santos,† Walter Baratta,‡ and Fritz E. Ku¨hn*,†,§
Department fu¨r Chemie der Technischen UniVersita¨t Mu¨nchen, Lichtenbergstrasse 4,
D-85747 Garching bei Mu¨nchen, Germany, Dipartimento di Scienze e Tecnologie Chimiche, UniVersita` di
Udine, Via Cotonificio 108, I-33100 Udine, Italy, and Department of Chemistry, Instituto Tecnolo´gico e
Nuclear (ITN), Estrada Nacional 10, P-2686-953 SacaVe´m, Portugal
ReceiVed June 8, 2006
Ruthenium compounds of general formula Cp′RuX(PR2R′)2 (Cp′ ) η5-C5H5 (Cp), η5-C9H7 (Ind), η5-
C5(CH3)5 (Cp*); X ) Cl, CF3C(O)O; R ) C6H5 (Ph), C6H4(CH3) (m-tolyl); R′ ) C6H5, C6H11 (Cy),
C6H4(CH3) (m-tolyl, o-tolyl)) are examined as catalysts for the aldehyde olefination starting from diazo
compounds, phosphanes, and aldehydes. Cp*RuCl(PPh3)2 is highly active for the olefination of several
aldehydes, displaying a very high E-selectivity, as well as for ketone olefination (with benzoic acid as
cocatalyst). The reaction’s mechanism is substantiated by the isolation of a catalytic active reaction species,
namely, a mixed carbene/phosphane ruthenium complex, Cp*RuCl(dCHCO2Et)(PPh3) (8). Spectroscopic
studies reveal that the latter compound reacts with PPh3 to produce the phosphorus ylide Ph3PdCHCO2-
Et, which further reacts with the aldehyde to produce the olefin.
Compounds of the general formula Cp′RuX(PR3)2 have found
widespread applications in catalysis. The ease of synthesis, the
wide array of ligands available to coordinate with the metal
center, and their general high stabilities make them a versatile
and promising class of catalysts. Baratta et al.8 studied the
decomposition of ethyl diazoacetate (EDA) with this type of
catalysts, presenting evidence for PPh3dCHCO2Et formation
in the presence of triphenylphosphane (PPh3). This prompted
us to study these compounds in the aldehyde olefination reaction
as catalysts for phosphorus ylide generation. In a preliminary
communication, we were able to confirm the activity of
organometallic ruthenium complexes, by using Cp′RuX(PR2R′)2
as efficient catalysts for the olefination of activated aldehydes.9
Introduction
The olefination of carbonyl compounds, namely, the trans-
formation of a carbonyl group into a carbon-carbon double
bond, is one of the most convenient and universal methods for
the preparation of alkenes.1 The catalytic alternative to the
classic Wittig reaction has proven to be quite promising,
avoiding the need for stepwise generation of phosphorus ylides
under basic conditions. Furthermore, the use of transition metal
complexes as catalysts also opens the possibility of efficient
asymmetric olefination.2 Accordingly, several catalytic aldehyde
and ketone olefination systems have been reported based on
Ru,3 Re,4 Rh,5 Fe,6 and Co7 complexes.
In the present study we explore the activity of these
complexes more broadly, including the olefination reaction of
nonactivated aldehydes and ketones. Investigations on the
influence of changing the nature of the reaction components
and conditions on catalytic activity are performed, in order to
attempt establishing a better understanding of the olefination
reaction catalyzed by Cp′RuX(PR2R′)2 complexes. A compre-
hensive study of the reaction mechanism confirms the presence
of a catalytically active mixed phosphane/carbene Ru species,
Cp*RuCl(dCHCO2Et)(PPh3). Subsequent NMR spectrospic
studies confirm the identity of the latter compound formed
during the catalytic cycle and elucidate the way substrates and
catalysts react to produce the desired olefin.
* Corresponding author. Current address: ITN, Estrada Nacional 10.
E-mail: fritz.kuehn@ch.tum.de or fekuhn@itn.pt.
† Technischen Universita¨t Mu¨nchen.
‡ Universita` di Udine.
§ TUM, Lichtenbergstr. 4.
(1) Kelly, S. E. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon Press: New York, 1999; p 730.
(2) (a) Rein, T.; Reiser, O. Acta Chem. Scand. 1996, 50, 369. (b) Li, A.
H.; Dai, L. X.; Aggarwal, V. K. Chem. ReV. 1997, 97, 2341. (c) Rein, T.;
Pederson, T. M. Synthesis 2002, 5, 579.
(3) (a) Fujimura, O.; Honma, T. Tetrahedron Lett. 1998, 39, 625. (b)
Lebel, H.; Paquet, V. Organometallics 2004, 23, 1187. (c) Graban, E.;
Lemke, F. R. Organometallics 2002, 21, 3823. (d) Sun, W.; Ku¨hn, F. E.
Appl. Catal. A Gen. 2005, 285, 163. (e) Sun, W.; Yu, B. S.; Ku¨hn, F. E.
Tetrahedron Lett. 2006, 47, 1993.
(4) (a) Herrmann, W. A.; Wang, M. Angew. Chem., Int. Ed. Engl. 1991,
30, 1641. (b) Herrmann, W. A.; Roesky, P. W.; Wang, M.; Scherer, W.
Organometallics 1994, 13, 4531. (c) Santos, A. M.; Roma˜o, C. C.; Ku¨hn,
F. E. J. Am. Chem. Soc. 2003, 125, 2414. (d) Zhang, X.; Chen, P. Chem.-
Eur. J. 2003, 9, 1852. (e) Chen, X. P.; Zhang, X. Y.; Chen, P. Angew.
Chem., Int. Ed. 2003, 42, 3798. (f) Santos, A. M.; Pedro, F. M.; Jogalekar,
A. A.; Lucas, I. S.; Roma˜o, C. C.; Ku¨hn, F. E. Chem. Eur. J. 2004, 10,
6313. (g) Pedro, F. M; Hirner, S.; Ku¨hn, F. E. Tetrahedron Lett. 2005, 46,
7777. (h) Ledford, B. E.; Carreira, E. M. Tetrahedron Lett. 1997, 38, 8125.
(5) (a) Lebel, H.; Paquet, V.; Proulx, C. Angew. Chem., Int. Ed. 2001,
40, 2887. (b) Grasa, G. A.; Moore, Z.; Martin, K. L.; Stevens, E. D.; Nolan,
S. P.; Paquet, V.; Lebel, H. J. Organomet. Chem. 2002, 658, 126. (c) Lebel,
H.; Paquet, V. Org. Lett. 2002, 10, 1671. (d) Lebel, H.; Paquet, V. J. Am.
Chem. Soc. 2004, 126, 320. (e) Lebel, H.; Guay, D.; Paquet, V.; Huard, K.
Org. Lett. 2004, 6, 3047. (f) Paquet, V.; Lebel, H. Synthesis-Stuttgart 2005,
1901.
(6) (a) Mirafzal, G. A.; Cheng, G.; Woo, L. K. J. Am. Chem. Soc. 2002,
124, 176. (b) Cheng, G.; Mirafzal, G. A.; Woo, L. K. Organometallics
2003, 22, 1468. (c) Chen, Y.; Huang, L.; Ranade, M. A.; Zhang, X. P. J.
Org. Chem. 2003, 68, 3714. (d) Chen, Y.; Huang, L.; Zhang, X. P. Org.
Lett. 2003, 5, 2493. (e) Chen, Y.; Huang, L.; Zhang, X. P. J. Org. Chem.
2003, 68, 5925. (f) Sun, W.; Ku¨hn, F. E. Tetrahedron Lett. 2004, 45, 7415.
(g) Aggarwal, V. K.; Fulton, J. R.; Sheldon, C. G.; de Vicente, J. J. Am.
Chem. Soc. 2003, 125, 6034. (h) Sharma, V. B.; Jain, S. L.; Sain, B. Catal.
Lett. 2004, 98, 141.
(7) Lee, M. Y.; Chen, Y.; Zhang, X. P. Organometallics 2003, 22, 4905.
(8) (a) Baratta, W.; Del Zotto, A.; Rigo, P. Chem. Commun. 1997, 2163.
(b) Baratta, W.; Del Zotto, A.; Rigo, P. Organometallics 1999, 18, 5091.
(9) Ku¨hn, F. E.; Santos, A. M.; Jogalekar, A. A.; Pedro, F. M.; Rigo, P.;
Baratta, W. J. Catal. 2004, 227, 253.
10.1021/om060508w CCC: $37.00 © 2007 American Chemical Society
Publication on Web 12/10/2006

Organometallic Ruthenium Complexes
Organometallics, Vol. 26, No. 2, 2007 303
Chart 1
C6H5), 104.0 (5 C, s, C5Me5), 59.1 (1 C, s, CH2CH3), 13.8 (1 C, s,
CH2CH3), 8.8 (5 C, s, C5Me5). 31P NMR (162 MHz, CD2Cl2,
-20 °C, standard 85% H3PO4): δ 49.6.
Ph3PdN-NdCHCO2Et (Phosphazine XIII).13 PPh3 (1.311 g,
5 mmol) was dissolved in 20 cm3 of n-pentane. Ethyl diazoacetate
(0.571 g, 0.5 mmol) was added to the mixture, and the solution
was stirred for 8 h. The resulting solid was filtered and washed
three times with cold pentane, producing a white powder (1.675 g,
89%). Anal. Calcd for C22H21N2O2P: C, 70.20; H, 5.62; N, 7.44.
1
Found: C, 69.43; H, 5.74; N, 7.12. H NMR (400 MHz, CDCl3):
δ 7.76 (1 H, d, 3J(H,P) 2.0 Hz, NdCH), 7.68-7.46 (15 H, m, C6H5),
3
3
4.19 (2 H, q, J(H,H) 7.20 Hz, CH2), 1.26 (3 H, t, J(H,H) 7.20
Hz, CH3). 13C NMR (100 MHz, CDCl3): δ 165.4 (1 C, s, C(O)),
138.1 (1 C, d, J(C,P) 47.6 Hz, NdCH), 133.5 (6 C, d, o-C6H5),
132.5 (3 C, s, p-C6H5), 128.7 (6 C, t, m-C6H5), 127.7 (3 C, d, J(C,P)
93.8 Hz, C-P), 59.7 (1 C, t, CH2), 14.4 (1 C, q, CH3). 31P NMR
(162 MHz, CDCl3, standard 85% H3PO4): δ 22.7.
Experimental Section
General Information. All reactions were carried out under an
inert gas atmosphere, using standard Schlenk techniques. Solvents
were dried following standard procedures and kept under argon.
THF was dried over Na/benzophenone and distilled before use.
n-Hexane, n-pentane, n-heptane, ethanol, and toluene were dried
over Na and stored over 4 Å molecular sieves.
General Procedure for Aldehyde Olefination. Aldehyde (2
mmol), PPh3 (0.577 g, 2.2 mmol), and Cp′RuCl(PR2R′)2 (0.040 to
0.012 mmol) were dissolved in the appropriate solvent (10 cm3)
and heated (details concerning solvent and the reaction temperature
of each run are described in Table 1). Ethyl diazoacetate (0.274 g,
2.4 mmol) was added in one portion and allowed to react while
monitoring the reaction progress by GC-MS. Afterward the solution
was cooled to room temperature and the solvent removed under
reduced pressure. The residue was taken into a small amount of
toluene and chromatographed over a silica gel column with
n-hexane/ethyl acetate (20:1), affording the olefin(s). For 4-di-
methylaminobenzaldehyde the yield was determined by GC-MS
using a previously recorded calibration curve (R2 > 0.999).
General Procedure for Ketone Olefination. Ketone (2 mmol),
PPh3 (0.577 g, 2.2 mmol), the ruthenium complex Cp*RuCl(PPh3)2
(0.016 g, 0.020 mmol), and benzoic acid (0.122 g, 1 mmol) were
dissolved in the appropriate solvent (10 cm3) and heated (details
concerning solvent and the reaction temperature of each run are
described in Table 2). Ethyldiazoacetate (0.274 g, 2.4 mmol) was
added in one portion and allowed to react while monitoring the
reaction progress by GC-MS. Afterward the solution was cooled
to room temperature and the solvent removed under reduced
pressure. The residue was taken into a small amount of toluene
and chromatographed over a silica gel column with n-hexane/ethyl
acetate (20:1), affording the olefin(s).
Mechanistic Studies. Catalysis without Aldehyde. Complex
7 (2 mg, 2.5 µmol) and PPh3 (36 mg, 0.138 mmol) were dissolved
in 0.6 cm3 of CD2Cl2. The resulting solution was transferred via
cannula to the NMR tube and cooled to -70 °C. EDA (14.3 mg,
0.125 mmol) was added and the NMR tube was kept at -70 °C
until the beginning of the measurements, which were performed at
RT.
Catalysis with Aldehyde. Complex 7 (2 mg, 2.5 µmol), PPh3
(36 mg, 0.138 mmol), and benzaldehyde (26.5 mg, 0.25 mmol)
were dissolved in 0.6 cm3 of CD2Cl2. The resulting solution was
transferred via cannula to the NMR tube and cooled to -70 °C.
EDA (14.3 mg, 0.125 mmol) was added, and the NMR tube was
kept at -70 °C until the beginning of the measurements, which
were performed at RT.
Compounds 1,10 2,8 3,8 4,8 6,11 and 712 (Chart 1) were synthesized
following literature procedures.
NMR measurements were made with a 400 MHz Bruker Avance
DPX-400 and a 400 MHz Jeol JNM-GX 400. Deuterated solvents
were dried under molecular sieves and degassed prior to use.
Syntheses and Catalysis. CpRu(OC(O)CF3)(PPh3)2 (5). CpRu-
Cl(PPh3)2 (0.298 g, 0.41 mmol) and AgOC(O)CF3 (0.1 g, 0.45 mol)
were dissolved in 20 mL of THF. The mixture was stirred for 16 h
under exclusion of light. The solution was slowly filtered followed
by partial removal of the volatiles under vacuum. Addition of 15 mL
of n-heptane and stirring afforded a yellow precipitate after a few
minutes. The solution was filtered and the CpRu(OC(O)CF3)(PPh3)2
(0.244 g, 74%) was washed twice with n-heptane. Anal. Calcd for
C43H35O2F3P2Ru: C, 64.26; H, 4.39. Found: C, 63.97; H, 4.39.
1H NMR (400 MHz, CDCl3): δ 7.22-7.08 (30 H, m, C6H5), 4.30
(5 H, s, C5H5). 13C NMR (100 MHz, CDCl3): δ 164.1 (1 C, weak
2
q, J(F,C) 35.1 Hz, OC(O)CF3), 137.3 (1 C, t, C6H5), 133.5 (2 C,
t, C6H5), 129.0 (1 C, s, C6H5), 127.7 (2 C, m, C6H5), 113.2 (1 C,
weak t, 1J(F,C) 292.7 Hz, CF3), 79.1 (5 C, s, C5H5). 31P NMR (162
MHz, CDCl3, standard 85% H3PO4): δ 41.1. 19F NMR (376 MHz,
CDCl3, standard CFCl3): δ -75.7.
Cp*RuCl(dCHCO2Et)(PPh3) (8). Cp*RuCl(PPh3)2 (0.305 g,
0.38 mmol) was dissolved in dry CH2Cl2 (30 cm3) and cooled to
-12 °C in a dry ice/isopropanol bath. Ethyldiazoacetate (0.217 g,
1.9 mmol) was added to the solution, and the mixture was stirred
between -12 and -6 °C until the evolution of nitrogen ceased
(which occurred usually within ca. 2 h). The solvent was concen-
trated in vacuum to approximately 5 cm3, maintaining the temper-
ature below -10 °C. The solution was cooled to -30 °C, and cold
n-hexane (ca. 40 cm3 at -30 °C) was added through a cannula to
the solution. Stirring the solution produced a green precipitate,
which was filtered and dried under vacuum. The product was
dissolved in cold, dry toluene (2 cm3 at -30 °C) and precipitated
by adding cold, dry n-hexane (50 cm3 at -30 °C). The precipitate
(0.162 g, 68%) was dried in vacuum. For characterization the
product was recrystallized from toluene/n-hexane (2/40 cm3) and
dried under vacuum for 2 h. Anal. Calcd for C32H36O2ClPRu: C,
1
61.98; H, 5.85. Found: C, 62.74; H, 5.60. H NMR (400 MHz,
CD2Cl2, -20 °C): δ 14.90 (1 H, d, 3J(H, P) 12.26 Hz, CH), 7.88-
Results and Discussion
3
7.15 (15 H, br m, C6H5), 3.99 (2 H, q, J(H,H) 7.35 Hz, CH2),
Compounds 1-7, CpRuCl(PPh3)2 (1), CpRuCl(PPh2Cy)2 (2),
CpRuCl(P(m-tolyl)3)2 (3), CpRuCl(PPh2(o-tolyl))2 (4), CpRu-
(CF3C(O)O)(PPh3)2 (5), IndRuCl(PPh3)2 (6), and Cp*RuCl-
(PPh3)2 (7) (see Chart 1), were tested for the aldehyde olefination
reaction as catalysts for phosphorus ylide generation in nonbasic
conditions (Scheme 1).
3
1.30 (15 H, s, C5Me5), 1.26 (3 H, t, J(H,H) 7.35 Hz, CH3). 13C
NMR (100 MHz, CD2Cl2, -20 °C): δ 264.5 (1 C, d, 2J(C,P) 14.8
Hz, RudCHCO2Et), 182.9 (1 C, s, CO), 135.7-127.7 (6 C, br m,
(10) Bruce, M. I.; Hameister, C.; Swincer, A. G.; Wallis, R. C. Inorg.
Synth. 1982, 21, 78.
(11) Oro, L. A.; Ciriano, J.; Campo, M.; Foces-Foces, C.; Cano, F. H.
J. Organomet. Chem. 1985, 289, 117.
In the aldehyde olefination leading to R,â-unsaturated esters,
both the reaction rate and E:Z selectivity have to be considered.
(12) Chinn, M. S.; Heinekey, D. M. J. Am. Chem. Soc. 1990, 112, 5166.

304 Organometallics, Vol. 26, No. 2, 2007
Table 1. Catalytic Aldehyde Olefination with Compounds 1-7 as Catalysts
Pedro et al.
catalyst
loading/% mol
entry
aldehyde
catalyst/phosphane
yielda/%
time/h
E:Z ratiob
1
2
3
4
5
6
7
8
9
10
11
12
13
I
I
I
I
I
I
I
II
I
1/PPh3
2/PPh2Cy
3/P(m-tolyl)3
4/PPh2(o-tolyl)
5/PPh3
6/PPh3
7/PPh3
7/PPh3
Fe(TPP)Cl/PPh3
Cl3ORe(PPh3)2/PPh3
7/PPh3
Fe(TPP)Cl/PPh3
Cl3ORe(PPh3)2/PPh3
2
2
2
2
2
1
1
1
1
1
0.6
0.6
0.6
42
46
63
92
29
96
98
97
97
95
96
96
42
2
2
2
2
2
1
88:12
94:6
98:2
96:4
87:13
92:8
99:1
94:6
92:8
98:2
96:4
91:9
93:7
5 min
0.5
2
2
2
I
II
II
II
2
2
b
a Isolated yields. E:Z ratios determined by GC-MS. Reaction temperature 50 °C, except entry 8, reaction conducted at 80 °C. Solvent THF, entry 8
toluene.
Table 2. Catalytic Aldehyde and Ketone Olefination with
Compound 7 as Catalyst
Chart 2
entry
aldehyde/ketone
catalyst
yielda/%
time/h
E:Z ratiob
1
2
3
4
5
6
7
8
9
10
11
12
13
II
II
7
7
7
7
95
96
61c
92
92
98
98c
85
78
92
57
84
58
3
95:5
84:16
96:4
97:3
97:3
94:6
92:8
88:12
42:58
35:65
41:59
0.25
24
1
5
24
2
III
VI
IV
V
III
VII
VIII
IX
X
7
7
7d
7
4
7d
7d
7d
7d
7d
72
44
7 days
24
48
XI
XII
b
c
a Isolated yields. E:Z ratios determined by GC-MS. Yield determined
d
by GC-MS. Benzoic acid (0.5 equiv) as cocatalyst. Solvent: toluene, except
entry 2, ethanol. Temperature: 80 °C, except entry 1, RT, and entry 2, 78
°C. Catalyst loading: 1% mol of carbonyl compound.
were tested in the aldehyde olefination reaction in this study.
The use of phosphanes with high Tolman cone angles15c has a
positive influence on the reaction outcome, since these catalysts
improve both the olefin yield and the E-isomer selectivity. The
best yield among the compounds 1-4 is obtained with complex
4 (entry 4), affording 92% olefin yield after 2 h reaction time,
and the best selectivity with complex 3, with a final E:Z ratio
of 98:2 (entry 3). For all of these catalytic runs it is essential to
use the same phosphane both as ligand and as oxygen abstractor
during the reaction, since the use of PPh3 as oxygen abstractor
in combination with another phosphane as ligand rapidly ensues
the replacement of (at least one of) the bulkier phosphanes and
would therefore change the activity of the catalyst, turning it
increasingly similar to that of compound 1 (in a control
experiment with compound 4 and 4-nba, using PPh3 as oxygen
abstractor, only 63% of olefin yield was observed).
Scheme 1
Scheme 2
Frequently, the most active catalysts present relatively low
selectivities.14 The aldehydes and ketones used as substrates are
depicted in Chart 2.
Catalyst Optimization. Compound 1 shows a comparatively
low activity in the olefination of 4-nitrobenzaldehyde (4-nba),
reaching an olefin yield of 42% (Table 1, entry 1) after 2 h
reaction time. Simultaneously the selectivity is rather poor, with
a final E:Z ratio of 88:12.
Changing the phosphane ligand of the studied complexes has
a twofold influence on the catalytic reaction: (1) It changes
the catalyst activity as a result of combined effects of the steric
It has been observed for several other processes that the
phosphane structure has a paramount effect on the catalysts
performance,15 since it allows the fine-tuning of the metal’s
reactivity and selectivity. According to previous studies, for
example, complexes 2-4, bearing bulky phosphanes, decompose
EDA at lower temperatures compared to compound 1 ligated
by PPh3.8b On the basis of this observation compounds 1-4
(15) (a) Hughes, R. P. In ComprehensiVe Organometallic Chemistry;
Wilkinson, G., Stone, F. G. A., Abel, E. W., Eds.; Pergamon Press: Oxford,
1982; Chapter 35. (b) Sharp, P. R. In ComprehensiVe Organometallic
Chemistry II; Wilkinson, G., Stone, F. G. A., Abel, E. W., Eds.; Pergamon
Press: Oxford, 1995; Vol. 8, Chapter 2. (c) Tolman, C. A. Chem. ReV. 1977,
77, 313. (d) Pignolet, L. H., Ed. Homogeneous Catalysis with Metal
Phosphane Complexes; Plenum: New York, 1983.
(13) Staudinger, H.; Meyer, J. HelV. Chim. Acta 1919, 2, 619.
(14) Ku¨hn, F. E.; Santos, A. M. Mini-ReV. Org. Chem. 2004, 1, 55.

Organometallic Ruthenium Complexes
Organometallics, Vol. 26, No. 2, 2007 305
Scheme 3
eventual influence of the catalyst in the second step cannot be
entirely discarded, the E:Z ratios improve directly with the
bulkiness of the phosphane:14c PPh3 < PPh2Cy < PPh2(o-tolyl)
< P(m-tolyl)3.
The substitution of the -Cl ligand by a more electron-
withdrawing group, namely, -OC(O)CF3 (compound 5), has a
negative effect on the outcome of the catalytic reaction,
producing only 29% yield. This is not surprising, since a more
electron-poor metal withdraws more electron density from the
attached phosphanes and coordinates them more strongly. The
E:Z ratio is within the measurement error equal to that obtained
with compound 1.
The most striking results were obtained when instead of the
cyclopentadienyl ligand bulkier ligands such as indenyl (Ind) 6
or pentamethylcyclopentadienyl (Cp*) 7 were applied. These
catalysts show an enhanced activity that allows 100% conversion
of 4-nba into olefin after 60 min (entry 6) and 5 min (entry 7)
of reaction time, for 6 and 7, respectively. For these catalysts it
was possible to decrease the catalyst loading to 1 mol % without
any activity decrease or additional side product formation. The
E:Z ratios, 92:8 for 6 and 99:1 for 7, show a substantial
selectivity improvement in comparison to 1.
On going from Cp to Ind and Cp* ligands, the catalyst
activities rise in accord with the increasing bulkiness and better
electron donation abilities of the ligands.19 The increased
bulkiness weakens the P-Ru bond,16b enabling a faster dis-
sociation of the phosphine21 (see Scheme 3). The increased
electron donation is likely to stabilize the 16e- active species
and thus favor the P-Ru bond cleavage as well.
bulk of the phosphanes (main factor) and their electron donation
abilities; (2) the generated phosphorus ylides show different
reactivity toward the aldehyde, once again, depending on both
electron donor ability and bulkiness.
Bulkier phosphanes than PPh3 are less strongly bonded to
the Ru center due to their increased steric requirements,16 and
thus removal of one phosphane ligand becomes easier in
complexes 2-4, a key factor for triggering the catalytic cycle
(see Scheme 3). On the basis of this rationalization complexes
2-4 should also show higher catalytic activity than PPh3sas
they in fact do.
Ylides of the type (p-R-C6H4)3PdCHCO2Et bearing electron-
donating groups R react faster with aldehydes than Ph3PdCHCO2-
Et does.17,18 However, bulkier ylides than Ph3PdCHCO2Et such
as Ph2(o-tolyl)PdCHCO2Et react slower with aldehydes.18 When
the electron donation of the ligands is not significantly different
(both PPh2(o-tolyl) and P(m-tolyl)3 display only slightly higher
electron donation abilities than PPh3),19 bulkiness seems to exert
a stronger effect over the ylide reactivity than the electron
donation. Therefore, when it concerns the ylides’ reactivity,
PPh2Cy should be the most effective ligand, followed by PPh2-
(o-tolyl) and P(m-tolyl)3.
Bulkier phosphanes enable an easier formation of the
phosphorus ylide, by facilitating the release of a phosphane
ligand, but also produce less reactive ylides, which slow the
second step. These observations emphasize the importance of
the ease of phosphane release in the ylide generation, since it
overcompensates, at least to a certain degree, the more sluggish
second step. The results presented here show that the optimal
result is found in a compromise between the ease of phosphane
release from the catalyst and the reactivity of the ylide, which
is accomplished in this case with PPh2(o-tolyl) and complex 4.
The increase of selectivity toward the E-isomer follows the
trend also observed in the Wittig reaction,18 where an increase
of the ligand bulkiness leads to higher E:Z ratios. Although an
The promising results obtained with compound 7 regarding
both activity and selectivity tempted us to compare its perfor-
mance to other known aldehyde olefination catalysts. Fe(TPP)-
Cl6c (TPP ) tetra(p-tolyl)porphyrin) and Cl3ORe(PPh3)24h have
been tested successfully in the olefination of several aldehydes
and presented the highest activities yet reported.
Three catalysts, Cp*RuCl(PPh3)2, Fe(TPP)Cl, and Cl3ORe-
(PPh3)2, were tested in the olefination of 4-nba under the same
reaction conditions, namely, at 50 °C in THF, using 1 mol %
catalyst loading. All three compounds reach quantitative yield
within 2 h of reaction time. The E:Z ratios are 99:1, 92:8, and
98:2, and the TOFs (determined after 5 min reaction time) 900,
890, and 1200 h-1, respectively. With a catalyst loading of 0.1%
mol none of the catalysts reach quantitative yield. The TOFs
are 1600, 3680, and 1230 h-1, respectively. The differences in
initial activity between these catalysts at low concentrations are
explained by the mechanism depicted in Scheme 3. The huge
excess of free PPh3 when low concentrations of catalyst 7 are
applied (more than 1000 times excess for a 0.1 mol % loading)
shifts the equilibrium strongly toward the inactive 18e- species,
reducing greatly the catalytic activity. This effect very likely
also occurs for Cl3ORe(PPh3)2, while Fe(TPP)Cl, having no
PPh3 ligands, is devoid of any PPh3 inhibition, and so maintains
high activity even at low concentrations. Also in the olefination
of benzaldehyde Cp*RuCl(PPh3)2 affords the highest E-isomer
selectivity for this set of catalysts (entries 12 to 14).
(16) (a) Luo, L.; Nolan, S. P. Organometallics 1994, 13, 4781. (b)
Cucullu, M. E.; Luo, L.; Nolan, S. P.; Fangan, P. J.; Jones, N. L.; Calabrese,
J. C. Organometallics 1995, 14, 289. (c) Serron, S.; Luo, L.; Li, C.; Cucullu,
M. E.; Nolan, S. P. Organometallics 1995, 14, 5290. (d) Smith, D. C.;
Haar, C. M.; Luo, L.; Li, C.; Cucullu, M. E.; Mahler, C. H.; Nolan, S. P.;
Marshall, W. J.; Jones, N. L.; Fagan, P. J. Organometallics 1999, 18, 2357.
(e) Serron, S. A.; Luo, L.; Stevens, E. D.; Nolan, S. P.; Fagan, P. J.
Organometallics 1996, 15, 5209.
None of the tested catalysts produce any olefin product when
we attempted aldehyde olefination with diethyl diazomalonate
in place of ethyl diazoacetate. By 31P NMR spectroscopy the
formation of the corresponding phosphorus ylide, Ph3PdC(CO2-
Et)2, can be observed, but this ylide is unable to react further
with 4-nba. This strongly suggests that this class of complexes
catalyzes phosphorus ylide formation from phosphanes and
(17) (a) Giese, B.; Schoch, J.; Ru¨chardt, C. Chem. Ber. 1978, 111, 1395.
(b) Ru¨chardt, C.; Panse, P.; Eichler, S. Chem. Ber. 1967, 100, 1144.
(18) Allen, D. W. Z. Naturforsch. 1980, 35b, 1455.
(19) Tolman, C. A. J. Am. Chem. Soc. 1970, 92, 2953.
(20) Campion, B. K.; Heyn, R. H.; Tilley, T. D. Organometallics 1990,
9, 1106.
(21) Gamasa, M. P.; Gimeno, J.; Gonzales-Bernardo, C.; Martin-Vaca,
B. M.; Monti, D.; Bassetti, M. Organometallics 1996, 15, 302.

306 Organometallics, Vol. 26, No. 2, 2007
Pedro et al.
diazo compounds, but not the further reaction of ylide to the
desired olefin.
reaction conditions and/or long reaction times24 to obtain reason-
able yields. Exceptions to this rule are trifluoromethyl ketones,
in which the carbonyl group is sufficiently electron deficient to
undergo fast nucleophilic attack by the phosphorus ylide.4c,25
The olefination of R,R,R-trifluoroacetophenone proceeds
smoothly at 80 °C in toluene, producing a 85% olefin yield
after a reaction time of 5 h (entry 8). With an E:Z ratio of 88:
12, the selectivity of this reaction is not as high as that observed
for aldehyde olefination, but considerably higher than that
obtained applying nonactivated ketones (see below).
The set of ketones tested included nonactivated aromatic rings
(VIII), activated aromatic rings (IX), deactivated aromatic rings
(X), and cyclic ketones (XI and XII) (using the same terminol-
ogy as applied for the aldehydes). To increase the reaction rates,
benzoic acid (0.5 equiv) was added as cocatalyst.
The yields obtained with aromatic ketones range from 92%
for 4-nitroacetophenone to 57% for 4-methoxyacetophenone.
Activated ketone IX reacts relatively fast (44 h) and shows the
best selectivity observed for the aromatic ketones under
examination, with an E:Z ratio of 35:65. Ketones VIII (entry
9) and X (entry 11) require long reaction times to obtain
moderate yields, 78% and 57%, respectively, having very similar
E:Z ratios. Moreover, the use of an initial excess of ketone VIII
(3 equiv) produces a 72% olefin yield with an E:Z ratio of 42:
58, after 72 h of reaction time, which is not an improvement
over our standard reaction conditions. The lack of electron-
withdrawing groups seems to be responsible for the lack of
reactivity of this type of ketones.
Cyclic ketones can also be olefinated by this method. Cyclo-
hexanone is olefinated with 84% yield (entry 12) within 24 h,
while for XII as substrate only a moderate yield of 58% is
achieved.
Complex 7 catalyzes efficiently the aldehyde olefination,
giving high yields and excellent selectivities in short reaction
times, while for the olefination of some ketones only moderate
yields could be obtained, even in the presence of benzoic acid
as cocatalyst. With respect to selectivity the results obtained
with Cp*RuCl(PPh3)2 for ketone olefination are the same order
of magnitude as those obtained by other good catalysts,4g,6d
unlike the case of aldehyde olefination, where compound 7 is
clearly the most selective system reported so far.
Mechanistic Studies. In a previous communication9 we
suggested a possible mechanism for the aldehyde olefination
with Cp′RuCl(PPh3)2 type complexes based on published results8
and our own 31P NMR spectroscopy studies. The proposed
mechanism involves a Cp*RuCl(dCHCO2Et)(PPh3) (8) carbene
intermediate.
Following the reaction of complex 7 with EDA by 1H NMR
and 31P NMR in CD2Cl2 at low temperatures, the existence of
complex 8 is evident. At -15 °C a doublet at δH ) 14.88 ppm
appears in the 1H NMR, suggesting the formation of a Cp*RuCl-
(dCHCO2Et)(PPh3) carbene.8 Simultaneously the intensity of
the methyl groups peak of Cp* of complex 7 (δH(Cp*) ) 0.98
ppm) decreases, while a new one appears around δH ) 1.30
ppm. At the same temperature the 31P NMR shows a steady
decrease of the peak of complex 7 (δP ) 41 ppm) while two
new peaks arise: a peak at 50.3 ppm for carbene complex 8
and a somewhat broader peak at 17.9 ppm, which splits into
Through careful tuning of the ligands coordinating to the
metal, it was possible to turn a medium-performance catalyst
(compound 1) into a good catalyst (compound 7) combining
fast reaction rates with high E-olefin selectivity. On the basis
of these results, complex 7 was tested for the olefination of
several other aldehydes and also ketones as described below.
Aldehyde and Ketone Olefination with Compound 7.
Before testing its activity toward different aldehydes and
ketones, studies were carried out on the optimization of the
reaction conditions. Taking the olefination of benzaldehyde as
the model reaction, it was shown that the best results are
obtained at 80 °C in toluene (Table 1, entry 8); the reaction
issnot surprisinglysmuch faster under these conditions than
at 50 °C in THF (Table 1, entry 12). The same reaction in
toluene at room temperature (Table 2, entry 1) reaches quantita-
tive yield after 3 h, also with an E:Z ratio of 95:5. Using ethanol
as solvent (Table 2, entry 2), the reaction is completed after
25 min, but the selectivity decreases notably to a final E:Z ratio
of 84:16.
As it seems, the solvent influences both the reaction rate and
selectivity. While in THF and toluene the selectivities are
similar, the reaction time can be considerably shortened in
toluene due to its higher boiling point. In ethanol the reaction
is slightly faster than in toluene, but the loss in selectivity
overcomes the gain in reaction speed. This decrease is also
observed for the Wittig reaction in different solvents.22 Therefore
all further olefination reactions were performed in toluene at
80 °C. The obtained results are summarized in Table 2.
Regarding the aldehyde olefination, the results present the
expected trends: activated aldehydes, i.e., compounds containing
electron-withdrawing groups such as aldehyde I and VI (entry
4), reach quantitative yields in short reaction times, typically
in less than 1 h. Deactivated aldehydes, containing electron-
donating groups, react comparatively slowly. By extending the
reaction time to 24 h high yields (98%) in the olefination of V
(entry 6) are obtained, while with substrate III even with
prolonged reaction times, only moderate yields are observed
(entry 3). A nonaromatic aldehyde, IV, reacts fairly fast under
the conditions applied, reaching 92% yield after 5 h reaction
time (entry 5).
A common feature found among all the results presented in
this work is the high E-isomer selectivity displayed by catalyst
7, being the highest selectivity reported so far.
The Wittig reaction between aldehydes17b or ketones6d,7,17b,23
and Ph3PdCHCO2Et can be strongly accelerated by submolar
quantities of benzoic acid. The acid activation is not limited to
benzoic acid, as several other acids display similar properties,23a
but it seems to have a broader efficiency regarding the ketones
and ylides. The influence of benzoic acid was therefore also
investigated in this work. The olefination of III proceeds
smoothly in the presence of 0.5 equiv of benzoic acid as
cocatalyst, reaching 100% yield after just 2 h of reaction time
(entry 7). The selectivity decreases to an E:Z ratio of 92:8,
though. When using just 0.1 equiv of benzoic acid, quantitative
olefin formation is observed within 4 h, displaying the same
E:Z ratio.
Ketone olefination with stabilized ylides, such as PPh3dCH-
CO2Et, is usually a sluggish process and requires both harsh
(24) (a) Fodor, G.; To¨mo¨sko¨zi, I. Tetrahedron Lett. 1961, 16, 579. (b)
Isaacs, N. S.; El-Din, G. N. Tetrahedron Lett. 1987, 28, 2191. (c) Spinella,
A.; Fortunati, T.; Soriente, A. Synlett 1997, 1, 93. (d) Wu, J.; Wu, H.; Wei,
S.; Dai, W.-M. Tetrahedron Lett. 2004, 45, 4401.
(25) (a) Dull, D. L.; Baxter, I.; Mosher, H. S. J. Org. Chem. 1967, 32,
1622. (b) Tadashi, E.; Tetsuya, A.; Katsumi, K. Tetrahedron Lett. 1992,
33, 5545.
(22) Vedejs, E.; Fleck, T. J. J. Am. Chem. Soc. 1989, 111, 5861.
(23) (a) Ru¨chardt, C.; Eichler, S.; Panse, P. Angew. Chem. 1963, 75,
858. (b) Flisza´r, S.; Hudson, R. F.; Salvadori, G. HelV. Chim. Acta 1964,
47, 159.

Organometallic Ruthenium Complexes
Organometallics, Vol. 26, No. 2, 2007 307
Chart 3
two peaks after 20 min at -10 °C. The latter peaks, at δP )
18.4 and δP ) 16.7 ppm, correspond to the cis and trans isomers
of the phosphorus ylide,26 as further confirmed by the 1H NMR.
At -10 °C the conversion of complex 7 is quite fast since in
1
less than 30 min it is completely absent from the H and 31P
NMR. Performing the same procedure in deuterated toluene or
THF produces very similar results, with the exception that
complex 7 is never fully consumed due to its lower solubility
in these solvents.
The reactive complex 8 is isolated in 68% yield by performing
all operations at low temperature (-5 °C or below), because at
higher temperatures a carbene-carbene coupling reaction easily
occurs, leading to diethyl maleate, 7, and other uncharacterized
ruthenium complexes. The properties of 8 can be compared with
those of the related compounds 98b and 1027 (Chart 3), which
have previously been reported by Baratta et al. Thus, complex
8 is more stable than compound 9, which was observed only in
solution, due to the stabilizing effect of the Cp* moiety, but
less stable than complex 10, containing a stabilizing N-
heterocyclic carbene, which was isolated at RT and decomposes
in solution after many hours. Complex 8 displays hindered
rotation of the phosphane and carbene ligands. At -20 °C the
phenyl peaks are broad (1H and 13C-NMR), while at -70 °C
the peaks are resolved and the fine structure of the phenyl
carbons is observed. The peaks of the carbene ligand change
from definite peaks to broad peaks on going from -20 to
-90 °C, indicating also a rotation barrier. The rotation restriction
is more pronounced for the phosphane ligand, in agreement with
its larger bulk.
Figure 1. 1H NMR of the reaction of complex 8 with PPh3 in
CD2Cl2 at -30 °C: (A) complex 8; (B) complex 8 after addition
of less than 1 equiv of PPh3; (C) complex 8 with 2.5 equiv of PPh3.
The aromatic region is excluded for simplicity purposes.
Complex 8 reacts, as expected, with PPh3 (Figure 1): the
outcome of adding less than 1 equiv of phosphane is a mixture
of complex 8, complex 7, and Ph3PdCHCO2Et; when using
more than 2 equiv of PPh3, complex 7 is the only Ru species
present plus the phosphorus ylide (in a predictable 1:1 ratio)
and the excess phosphane.
Figure 2. 31P NMR of a catalytic benzaldehyde olefination with
complex 7 at RT in CD2Cl2: (A) phosphorus ylide; (B) phosphazine
XIII; (C) OdPPh3; (D) PPh3.
Noteworthy is the fact that this reaction occurs even at
-50 °C. The mixture reacts quantitatively with 2 equiv of 4-nba
within 20 min (starting at -30 °C then proceeding to RT) to
produce the corresponding olefin with a final E:Z ratio of 97:3
(determined by GC-MS). For comparison purposes the olefi-
nation of 4-nba, using complex 7 as catalyst in CH2Cl2 at RT,
produces the same resultsswithin the measurement error of the
GC-MS techniquesthe final E:Z ratio being 96:4. The agree-
ment between the results renders it likely that complex 8 is the
active intermediate of the catalytic cycle.
of aldehyde in less than 5 min the quantitative formation of
phosphorus ylide and the phosphazine Ph3PdN-NdCHCO2-
1
Et (XIII) in an 86:14 ratio (by integration of H NMR peaks,
available as Supporting Information) is observed.
The 31P NMR has three main peaks besides the free PPh3
(excess): a singlet at δP ) 21.1 ppm and two broad peaks
corresponding to the geometrical isomers of the ylide (δP )
18.0 and 16.4 ppm), which, as the temperature increases, almost
coalesce to a single broad peak at δP )17.9 ppm.26 It is
necessary to wait 24 h until the phosphazine is almost entirely
consumed, leaving ylide as the main product, observable in both
1
Following a catalytic run by H and 31P NMR in both the
absence and the presence of benzaldehyde in CD2Cl2, the
identity of the species formed during the catalytic cycle becomes
clearer (see Experimental Section for details). In the absence
1
the 31P and H NMR spectra.
The same reaction in the presence of benzaldehyde (Figure
2) shows some particular features: the ylide has a single sharp
peak (δP ) 18.1 ppm) in the 31P NMR, a result of its interaction
with the aldehyde;26 the formation of OdPPh3 (δP ) 28.0 ppm)
is fast and it is present almost from the beginning of the
catalysis; the ylide is consumed completely within 15 min,
during which phosphazine XIII (δP ) 22.6 ppm) scarcely reacts.
After 150 min at RT the reaction is complete, as there is only
(26) (a) Kayser, M. H.; Hooper, D. L. Can. J. Chem. 1990, 68, 2123,
and references therein. (b) Kayser, M. H.; Hatt, K. L.; Hooper, D. L. Can.
J. Chem. 1991, 69, 1929. (c) Crews, P. J. Am. Chem. Soc. 1968, 90, 2961.
(d) Crouse, D. M.; Wehman, A. T.; Schweizer, E. E. Chem. Commun. 1968,
866.
(27) (a) Baratta, W.; Herdtweck, H.; Herrmann, W. A.; Rigo, P.; Schwarz,
J. Organometallics 2002, 21, 2101. (b) Baratta, W.; Herrmann, W. A.; Rigo,
P.; Schwarz, J. J. Organomet. Chem. 2000, 489, 593-594.

308 Organometallics, Vol. 26, No. 2, 2007
Pedro et al.
the peak of OdPPh3 in the 31P NMR. These observations lead
to two conclusions: (a) the ylide formation is much faster than
the phosphazine XIII formation and XIII is also less reactive;
(b) the phosphazine reaction is promoted by the presence of
aldehyde.
Bestmann and Go¨thlich28 proved that in solution resonance-
stabilized phosphazines exist in equilibrium with the original
phosphane and diazo starting compounds (Scheme 2). This
suggests that even in the presence of phosphazine the actual
reaction partner is EDA.
A mixture of phosphazine and complex 7 (ratio 2:1) in CD2-
Cl2 does not show any changes in the 31P or 1H NMR between
-30 and 10 °C. At 15 °C a small doublet at δH ) 14.9 ppm in
the 1H NMR can be observed, evidencing the presence of
complex 8 in solution. At 20 °C there is a small peak at δP )
51 ppm in the 31P NMR (complex 8). However, it is necessary
to let the sample react for 20 min at 20 °C to observe the first
peak of Ph3PdCHCO2Et in the 31P NMR.
Figure 3. 1H NMR spectra of the phosphorus ylide: (A) dried
over CaH2; (B) after addition of ca. 0.1 equiv of benzoic acid to
A.
The conversion of phosphazine into phosphorus ylide should
be faster as the equilibrium is shifted to the starting materials.
This explains why the presence of aldehyde in solutionswhich
pushes the reaction forward through the formation of the olefins
or a temperature increase enhances the conversion rate since
they drive the equilibrium toward the right side of Scheme 2.
For highly active catalysts such as complex 7, the phosphazine
formation is strongly suppressed and should become significant
only at very low catalyst loadings; therefore, the main reaction
pathway for complex 7 should be the one described in Scheme
3.
Therefore, the best catalysts promote the direct formation of
Ph3PdCHCO2Et from EDA and PPh3, avoiding the production
of phosphazine XIII, a comparatively sluggish reactant. The
first step, release of phosphane and formation of complex 8 by
reaction with EDA, occurs at temperatures above -10 °C. The
second step, formation of the phosphorus ylide through reaction
of complex 8 with PPh3, is considerably easier and occurs even
at -50 °C. The third step, reaction of the phosphorus ylide with
the aldehyde, was considered above in terms of ylide structure
and reactivity. For this class of ruthenium complexes the
different E:Z ratios obtained with catalysts 1, 6, and 7 (Table
1, entries 1, 6, and 7, respectively) show that the catalyst might
play a certain role in the third step, influencing the reaction’s
selectivity. The E:Z selectivity obtained with complex 7 is
(within the experimental error) for several tested aldehydes (I,
II, and III) equal to the selectivity obtained in the noncatalyzed
Wittig reaction using similar reaction conditions. The NMR
studies also did not show any noticeable interaction between
this catalyst and the phosphorus ylide, so that it may be
concluded that the role of complex 7 in the third step is
negligible to the reaction outcome.
Figure 4. 31P NMR spectra of the phosphorus ylide: (A) dried
over CaH2; (B) after addition of ca. 0.1 equiv of benzoic acid to
A.
One of the main features is the collapse of the peaks of the
ylide’s cis and trans isomers. This pattern is similar to that
observed on addition of benzaldehyde to the ylide.26a,b The
collapse of the peaks indicates that the interconversion between
cis and trans isomers is very fast (on the NMR time scale).
Hooper et al.26b suggested that during the Wittig reaction
neither the cis nor trans isomers of the phosphorus ylide seem
to readily react with the carbonyl group. Instead the reaction in
carried out by an intermediate present during the interconversion
of the two isomers, which bears a negative charge on the
methine carbon (Scheme 4).26b In the presence of traces of water
or acidic protons (to promote the protonation of the isomers),
the carbonyl of benzaldehyde easily “traps” this reactive
intermediate, and thereby the peaks collapse.
Mixing acetophenone with the ylide (1:1) no such collapse
is observed, indicating that the ketone cannot “trap” the reactive
intermediate due to its low electrophilic character. The benzoic
acid, however, greatly increases the concentration of the
“reactive ylide” in solution, while not reacting with it (Scheme
4), making it available to nucleophilic attack on a carbonyl
group. Still the carbonyl group of ketones is much less
electrophilic than the carbonyl of aldehydes, and concomitantly
the reactions are considerably slower with ketones than with
aldehydes. Nevertheless the trend is clear, the more electrophilic
the ketone’s carbonyl, the higher the reaction rate.
To the best of our knowledge, the reaction rate enhancement
effected by benzoic acid was never fully understood. Some
authors suggest the carbonyl to be protonated by the acid,
making it more electrophilic and prone to a nucleophilic attack
by the ylide.6d
1
A H and 13C NMR of a mixture of acetophenone/benzoic
acid (1:0.5) resemble the arithmetic sum of the spectra of the
individual compounds, therefore excluding a significant interac-
tion between the compounds and less likely any activation
toward the olefination reaction.
Addition of benzoic acid to the phosphorus ylide changes
their NMR spectra considerably (Figure 3 and 4).
Other interesting features observed in Figure 3 are the position
of the methine proton and its lack of coupling with the
(28) Bestmann, H. J.; Go¨thlich, L. Liebigs Ann. Chem. 1962, 655, 1.

Organometallic Ruthenium Complexes
Organometallics, Vol. 26, No. 2, 2007 309
Scheme 4. “Activation” of Phosphorus Ylide by Benzoic
Acida
compounds, phosphanes, and aldehydes. The phosphane struc-
ture has a strong effect on the catalyst performance, bulkier
phosphanes leading to better yields and higher selectivities. The
replacement of the Cl ligand by a more electron-withdrawing
group such as -OC(O)CF3 has a negative effect on the catalyst
performance. On the other hand, the replacement of the Cp
ligand by bulkier π-ligands such as Cp* or indenyl has a
remarkable positive effect on the catalyst activity. Particularly
Cp*RuCl(PPh3)2 leads to excellent yields and particularly high
E:Z ratios, the latter being the best reported so far. This
compound was also found to be a good catalyst for the
olefination of activated ketones and a moderate catalyst for the
olefination of nonactivated ketones using benzoic acid as
cocatalyst. Investigations on the mechanism show the existence
of a mixed carbene/phosphane ruthenium intermediate, complex
8, which was isolated and characterized. The catalytic cycle is
triggered by EDA, and the ylide Ph3PdCHCO2Et is formed by
reaction of complex 8 with PPh3. Under catalytic conditions
phosphazine XIII, a less reactive reaction partner, is also formed
and exists in equilibrium with its precursor compounds. Highly
active catalysts such as complex 7 largely suppress phosphazine
formation.
a
Partially adapted from ref 26b.
phosphorus. While the positions of the methylene and methyl
protons appear to be the average of the initial positions and
relative amounts of each isomer existing before the addition of
the benzoic acid, the methine proton is considerably shifted
toward lower fields. Further experiments using different amounts
of benzoic acid unveil that this peak shifts to lower fields as
the relative amount of benzoic acid increases. We conclude that
the proton must be switching between the two ylide isomers
and the benzoic acid, and the resulting peak is an average of
these three contributions. A calculation of the expected position
of this peak simply based on the amounts of each component
and the peaks’ original positions leads to δH(CH) ) 3.48 ppm
as the expected chemical shift of the methine proton and is in
very good agreement with the experimental result δH(CH) )
3.42 ppm (Figure 3), thus confirming the above-described
suggestion. The lack of coupling with phosphorus of the
resulting peak stems from this same process, since the proton
while coordinated to benzoic acid does not couple with the
phosphorus.
Benzoic acid promotes the catalytic reaction with both
aldehydes and ketones, by facilitating the transition between
cis and trans isomers of the phosphorus ylide and hence
increasing the presence of the “reactive ylide” in solution,
making it available to react with otherwise unreactive ketones.
Acknowledgment. Dr. Klaus Ruhland is gratefully acknowl-
edged for his help on the NMR studies. F.M.P. thanks the
Fundac¸a˜o para a Cieˆncia e a Tecnologia for a Ph.D. scholarship;
A.M.S. thanks the Alexander von Humboldt foundation for a
postdoctoral research fellowship. The authors are also grateful
to the Fonds der Chemischen Industrie, the Leonhard Lorenz-
foundation, and the DAAD (Vigoni) for financial support.
Conclusions
Supporting Information Available: This material is available
free of charge via the Internet at http://pubs.acs.org.
Several ruthenium compounds bearing phosphane ligands act
as catalysts for the aldehyde olefination starting from diazo
OM060508W