Novel phosphinite-ruthenium(II) complexes covalently bound on silica: Synthesis, characterization, and catalytic behavior versus oxidation reactions of alcohols into aldehydes

Article abstract of DOI:10.1021/om020876+

The novel tridentate phosphinite-Ru complex 5 has been prepared, characterized, and bound on silica via the spacer arm 16. The efficiency of this heterogenized catalyst for the oxidation of alcohols into aldehydes with N-methylmorpholine oxide has been examined, in comparison to the efficiency of the corresponding heterogenized monodentate phosphinite-Ru complex (without 7 and with the spacer arm 11) and the corresponding homogeneous catalysts 3 and 5, respectively. The heterogenized monodentate catalysts were 2 and 3 times more active than the soluble phosphinite-Ru complex, while the heterogenized tridentate catalyst showed almost the same activity. All catalysts are endowed with good to excellent selectivities. The heterogenized monodentate catalysts could not be efficiently recycled; on the other hand, the heterogenized tridentate catalyst 16 worked for four cycles. However, the major problem was a significant metal leaching, after the first cycle.

Full text of DOI:10.1021/om020876+

804
Organometallics 2003, 22, 804-811
Novel P h osp h in ite-Ru th en iu m (II) Com p lexes Cova len tly
Bou n d on Silica : Syn th esis, Ch a r a cter iza tion , a n d
Ca ta lytic Beh a vior ver su s Oxid a tion Rea ction s of
Alcoh ols in to Ald eh yd es
Eric Dulie`re,† Michel Devillers,‡ and J acqueline Marchand-Brynaert*,†
Unite´ de Chimie Organique et Me´dicinale and Unite´ de Chimie des Mate´riaux inorganiques et
organiques, Universite´ Catholique de Louvain, Baˆtiment Lavoisier, place L. pasteur 1,
B-1348 Louvain-la-Neuve, Belgium
Received October 22, 2002
The novel tridentate phosphinite-Ru complex 5 has been prepared, characterized, and
bound on silica via the spacer arm 16. The efficiency of this heterogenized catalyst for the
oxidation of alcohols into aldehydes with N-methylmorpholine oxide has been examined, in
comparison to the efficiency of the corresponding heterogenized monodentate phosphinite-
Ru complex (without 7 and with the spacer arm 11) and the corresponding homogeneous
catalysts 3 and 5, respectively. The heterogenized monodentate catalysts were 2 and 3 times
more active than the soluble phosphinite-Ru complex, while the heterogenized tridentate
catalyst showed almost the same activity. All catalysts are endowed with good to excellent
selectivities. The heterogenized monodentate catalysts could not be efficiently recycled; on
the other hand, the heterogenized tridentate catalyst 16 worked for four cycles. However,
the major problem was a significant metal leaching, after the first cycle.
In tr od u ction
reusable without significant loss of activity or selectiv-
ity.3 The situation is totally different considering oxida-
tion processes,6 most probably because the available
homogeneous catalysts7 are not so numerous, but also
(and mainly) because the oxidation catalysts based on
transition-metal complexes are usually less stable and
more sensitive toward the reaction conditions.8 Accord-
ingly, most of the examples of heterogenizing homoge-
neous oxidation catalysts concern Sharpless- and J a-
cobsen-type catalysts used for olefin dihydroxylation9
and epoxidation.10 On the other hand, oxidation of
alcohols into aldehydes (ketones) catalyzed by supported
ruthenium complexes appears to be scarcely studied at
the present time.11-13
Numerous efforts have been devoted to the develop-
ment of new supported catalysts combining the advan-
tages of conventional homogeneous and heterogeneous
catalysts.1 Application to the field of transition-metal
catalysis was particularly addressed,2 since in this case
recovery and reuse of the catalyst could bring valuable
economical and environmental advantages. Function-
alized cross-linked polystyrene3 and silica4 were the
most currently used supports, on which organic ligands
could be immobilized either by covalent linkage or by
hydrogen-bonding interaction.5 The grafting of transi-
tion-metal complexes on such solid supports has been
well illustrated with rhodium2c and, to a lesser extent,
ruthenium complexes,5a useful in various reduction
processes such as hydrogenation and hydrogen transfer
reactions. The supported catalysts were stable and
In this paper, we describe our attempts to construct
heterogeneous ruthenium catalysts which would operate
as if they were in solution. In preliminary experiments,14a
we have already selected phosphinite derivatives as
metal ligands because they appear to be more stable
* To whom correspondence should be addressed. Fax: +32 10
474168. E-mail: marchand@chim.ucl.ac.be.
† Unite´ de Chimie Organique et Me´dicinale.
(6) Ley, S. V.; Baxendale, I. R.; Breams, R. N.; J ackson, P. S.; Leach,
A. G.; Longbottom, D. A.; Nesi, M.; Scott, J . S.; Storer, R. I.; Taylor, S.
J . J . Chem. Soc., Perkin Trans. 1 2000, 3815.
(7) Sheldon, R. A., Arends, I. W. C. E.; Dijksman, A. Catal. Today
2000, 157.
‡ Unite´ de Chimie des Mate´riaux inorganiques et organiques.
(1) (a) Sherrington, D. C., Supported Catalysts and Their Applica-
tions; Springer: New York, 2001. (b) Gao, H.; Angelici, R. J . Organo-
metallics 1999, 18, 989 and references therein.
(2) (a) Doyle, M. P.; Timmons, D. J .; Tumonis, J . S.; Gau, H.-M.;
Blossey, E. C. Organometallics 2002, 21, 1747. (b) Kingsbury, J . S.;
Garber, S. B.; Giftos, J . M.; Gray, B. L.; Okamoto, M. M.; Farrer, R.
A.; Fourkas, J . T.; Hoveyda, A. H. Angew. Chem., Int. Ed. 2001, 40,
4251. (c) Sandee, A. J .; Reek, J . N. H.; Kamer, P. C. J .; van Leeuwen,
P. W. N. M. J . Am. Chem. Soc. 2001, 123, 8468. (d) Reger, T. S.; J anda,
K. D. J . Am. Chem. Soc. 2000, 122, 6929.
(8) Sanderson, W. R. Pure Appl. Chem. 2000, 72, 1289.
(9) (a) Song, C. E.; Yang, J . W.; Ha, H. T. Tetrahedron: Asymmetry
1997, 8, 841. (b) Song, C. E.; Yang, J . W.; Ha, H. J .; Lee, S. G.
Tetrahedron: Asymmetry 1996, 7, 645.
(10) Song, Y.; Yao, X.; Chen, H.; Pan, G.; Hu, X.; Zheng, Z. J . Chem.
Soc., Perkin Trans. 1 2002, 870.
(11) Leadbeater, N. E. J . Org. Chem. 2001, 66, 2168.
(12) Dalal, M. K.; Upadhyay, M. J .; Ram, R. N. J . Mol. Catal. A:
Chem. 1999, 142, 325.
(3) Clapham, B.; Reger, T. S.; J anda, K. D. Tetrahedron 2001, 57,
4637.
(4) Clark, J . H.; Macquarrie, D. J . Chem. Commun. 1998, 853.
(5) (a) Bianchimi, C.; Barbaro, P.; Dal Santo, V.; Gobetto, R., Meli,
A.; Oberhauser, W.; Psaro, R.; Vizza, F. Adv. Synth. Catal. 2001, 343,
41. (b) Bianchini, C.; Dal Santo, V.; Meli, A.; Oberhauser, W.; Psaro,
R.; Vizza, F. Organometallics 2000, 19, 2433.
(13) J orna, A. M. J .; Boelrijk, A. E. M.; Hoorn, H. J .; Reedijk, J .
Reactive Functional Polym. 1996, 29, 101.
(14) (a) Dulie`re, E.; Tinant, B.; Schanck, A.; Devillers, M.; Marchand-
Brynaert, J . Inorg. Chim. Acta 2000, 311, 147. (b) Tinant, B.; Dulie`re,
E.; Devillers, M.; Marchand-Brynaert, J . Z. Kristallogr. 2003, 218, 1.
10.1021/om020876+ CCC: $25.00 © 2003 American Chemical Society
Publication on Web 01/18/2003

Novel Phosphinite-Ruthenium(II) Complexes
Organometallics, Vol. 22, No. 4, 2003 805
Sch em e 1. P r ep a r a tion of Hom ogen eou s Ca ta lysts
Sch em e 2. P r ep a r a tion of Heter ogen ized Ca ta lyst
7
tern of Ru centered at m/e 808.8 (M - Cl). X-ray
diffraction analysis14b of a monocrystal (obtained by slow
evaporation from a chloroform solution) showed a
dimeric structure in the solid state, very similar to that
of compound 3.14a On the other hand, 31P NMR spec-
troscopy revealed two broad signals at 142.8 and 153.7
ppm, assigned respectively to equatorial and apical
phosphorus in the monomeric complex 5. The dimer of
5, which presents a local symmetry of order 3 around
the axis linking the two Ru atoms, should give only one
signal in 31P NMR; thus, in chloroform solution, a
significant part of the complex exists as the monomeric
form.
toward self-oxidation than the traditional phosphine
ligands. The selected support was silica, a material
chemically stable to oxidation, which can be easily
derivatized by standard reactions.15 Thus, a novel
tripodal phosphinite-Ru complex has been prepared
and bound on silica via a spacer arm. The efficiency of
this material as catalyst for the oxidation of alcohols
into aldehydes has been examined, comparatively to the
corresponding homogeneous catalyst. The possibility of
recycling the heterogenized catalyst and, more particu-
larly, the extent of Ru leaching under reaction condi-
tions were also considered.
Syn t h esis a n d Ch a r a ct er iza t ion of H et er og-
en ized Ca ta lysts. Three systems of increasing com-
plexities were considered: (i) catalyst 7 (Scheme 2) with
diphenylphosphine ligands directly bound on the OH
groups of the silica surface, (ii) catalyst 11 (Scheme 3)
with a monopodal phosphinite ligand bound on the
inorganic support via a flexible spacer arm, and (iii)
catalyst 16 (Scheme 4) with a tripodal phosphinite
ligand bound via the same spacer as above. Complex 3
(Scheme 1b) can be considered as the soluble model of
heterogeneous catalysts 7 and 11, while complex 5
(Scheme 1c) mimics the catalytic site of the heterog-
enized catalyst 16.
Resu lts a n d Discu ssion
Syn t h esis a n d Ch a r a ct er iza t ion of H om oge-
n eou s Ca ta lysts (Refer en ce System s). Dichlorotris-
(triphenylphosphine)ruthenium(II) (1) (Scheme 1a) is a
commercially available catalyst, very popular for alcohol
oxidation with N-methylmorpholine oxide (NMMO);16
this will constitute our reference system. As previously
stated,14 we used phosphinite complexes, which are
more convenient for our purposes. Thus, starting with
dichlorotetrakis(dimethyl sulfoxide)ruthenium(II) (2) as
a practical source of Ru(II),17 we prepared dichlorotris-
(methoxydiphenylphosphine)ruthenium(II) (3)14a by re-
action with methoxydiphenylphosphine in refluxing
methanol (Scheme 1b). Similarly, the corresponding
tridentate complex 5 was obtained from 2 and 1,1,1-
tris(((diphenylphosphino)oxy)methyl)ethane (4) (Scheme
1c) by reaction of the triol precursor with chlorodi-
phenylphosphine in the presence of triethylamine in
dichloromethane. Compound 5 was purified by chroma-
tography on Sephadex gel and well identified by mass
spectrometry, which displayed the typical isotopic pat-
We used a silica support displaying about 2.5 mmol/g
of hydroxyl functions. Reaction with chlorodiphenylphos-
phine and triethylamine in refluxing dichloroethane
(DCE) gave the phosphinite-derivatized silica species 6,
which was directly transformed into Ru(II) complex 7
(Scheme 2) by treatment with dichlorotetrakis(dimethyl
sulfoxide)ruthenium(II) (2) as described in the case of
homogeneous catalysts. After a 2 day extraction with
CH2Cl2 and ether, the modified silica was characterized
by thermogravimetry: we found 0.44 mmol/g of grafted
ligand (i.e. theoretically 0.15 mmol/g of Ru complex),
corresponding to about 18% of surface derivatization.
Fixed ruthenium has been experimentally assayed by
atomic absorption spectrometry as follows: the catalyst
was successively treated with perchloric acid (10% in
water) and sodium hydroxide (5% in water) in order to
recover all the ruthenium species in solution; after
filtration of silica, the solution was concentrated to a
standard volume and analyzed by atomic absorption
(the apparatus being previously calibrated with refer-
ence solutions of ruthenium chloride). We found 0.08
mmol/g of Ru; this value corresponds to a P/Ru ratio of
(15) (a) Lasaicherre, M.-L.; Uttamchandani, M.; Chen, G. Y. J .; Yao,
S. Q. Bioorg. Med. Chem. Lett. 2002, 12, 2079. (b) Ayadim, M.; Habib-
J iwan, J .-L.; Desilva, A. P.; Soumillion, J .-P. Tetrahedron Lett. 1996,
37, 7039.
(16) Vijayasri, K.; Rajaram, J .; Kuriacose, J . Proc. Indian Acad. Sci.-
Chem. Sci. 1986, 97, 125.
(17) Evans, I. P.; Spencer, A.; Wilkinson, G. J . Chem. Soc., Dalton
Trans. 1973, 204.

806 Organometallics, Vol. 22, No. 4, 2003
Sch em e 3. P r ep a r a tion of Heter ogen ized Mon od en ta te Ca ta lyst 11
Dulie`re et al.
Sch em e 4. P r ep a r a tion of Heter ogen ized Tr id en ta te Ca ta lyst 16
5.5 (the theoretical P/Ru value was 3 for structure 7;
Table 1, entry 2).
The spacer arm 8 required for the construction of
catalyst 11 was prepared by reacting (p-amino)phen-

Novel Phosphinite-Ruthenium(II) Complexes
Organometallics, Vol. 22, No. 4, 2003 807
Ta ble 1. Ch a r a cter iza tion of Der iva tized Su p p or ts
amt of grafted
ligand, mmol/g
amt of fixed Ru,
mmol/g
entry
support
support no.
% surface derivat
P/Ru
1
2
3
4
5
6
7
Si-OH (native support)
Si-OPPh2Ru
-
7
9
11
14
16
17
2.50
0.44
0.50
0.40
0.45
0.33
n.d.
18
60
48
54
40
n.d.
0.08
0.16
5.5
2.5
Si-spacer-OH
Si-spacer-OPPh2Ru
Si-spacer-(CH2OH)3
Si-spacer-(CH2OPPh2)3Ru
PS-PPh2Ru11
0.62
0.80
1.6
n.d.
ethyl alcohol with 3-isocyanato-1-(triethoxysilyl)propane
in tetrahydrofuran (THF) at reflux (Scheme 3a). The
complete chemoselectivity of this reaction was proved
by the NMR and IR analysis of the product. Compound
8 was immediately adsorbed on silica in suspension in
THF at 20 °C (3 h) and then grafted by heating the
mixture for 3 days. After several washings (see the
Experimental Section), the derivatized silica 9 (Scheme
3b) was characterized by thermogravimetry: we deter-
mined the presence of 11% of organic materials, i.e. 0.5
mmol/g of grafted silanols. Thus, assuming that each
molecule 8 occupies three silanol functions of the native
silica, we could consider that about half of the available
Si-OH groups has reacted (maximum value 1/3 of 2.5
mmol/g ) 0.83 mmol/g). The modified silica 9 was
further treated with chlorodiphenylphosphine and tri-
ethylamine in refluxing DCE (Scheme 3c) to furnish
silica 10, immediately transformed into Ru catalyst 11,
as described for 7 (Scheme 3d). Thermogravimetry and
elemental analysis gave very concordant results: silica
11 contains 0.4 mmol/g of ligand, corresponding to 48%
of surface derivatization. The measurement of fixed
ruthenium by atomic absorption, as previously de-
scribed, confirmed the presence of 0.16 mmol Ru/g,
corresponding to a P/Ru atomic ratio of 2.5 (theoretical
value 3 for catalyst 11; Table 1, entry 4). The charac-
teristics of 11 are in good agreement with those of its
precursor 9 (Table 1, entry 3); thus, the chemical yields
of the reactions performed on the solid support (Scheme
3, steps c and d) are nearly quantitative.
The preparation of catalyst 16 made use of the triol
13 as precursor of the tripodal ligand. The synthesis of
this molecule has been fully described elsewhere.18 The
solvents which solubilize 13 (H2O, CH3OH, pyridine, ...)
are not compatible with the stability of 3-isocyanato-1-
(triethoxysilyl)propane. Therefore, this last reagent was
first grafted on silica in refluxing THF (Scheme 4a), and
the resulting material 12 (1.2 mmol/g of isocyanate
reactive termini, most probably contaminated with
unreactive carbamate termini resulting from EtOH
addition) was coupled to 13 in a mixture of THF and
pyridine (Scheme 4b). After 3 days of heating, the
modified silica 14 was washed, dried, and analyzed as
before. From thermogravimetry and elemental analysis,
we found 0.45 mmol/g of fixed ligand, i.e. 54% of surface
derivatization, considering that three silanol functions
are used to immobilize one molecule of ligand (Table 1,
entry 5). The next step, i.e. reaction of triol 14 with
chlorodiphenylphosphine (Scheme 4c), appeared some-
what arduous. In the presence of triethylamine (as
before), diazabicyclooctane (DABCO), diazabicycloun-
decene (DBU), or (dimethylamino)pyridine (DMAP), the
yields of substitution were less than 20% (thermogravi-
metric analysis). Therefore, we first deprotonated the
triol 14 by treatment with lithium hexamethyldisilaza-
nate (LiHMDS) in THF at -70 °C and then added
chlorodiphenylphosphine at -40 °C, in the presence of
DMAP. The mixture was further heated for 1 day to
achieve the nucleophilic substitution. The resulting
silica 15 (Scheme 4c) was washed and directly reacted
with RuCl2(DMSO)4 (2) in refluxing DCM to furnish the
complex 16 (Scheme 4d). Thermogravimetric analysis,
confirmed by the elemental analysis, gave 0.33 mmol/g
of grafted catalyst (about 80% yield from 14) corre-
sponding to 40% of surface derivatization (Table 1, entry
6). The amount of ruthenium, measured by atomic
absorption, was 0.62 mmol/g. Accordingly, the P/Ru
atomic ratio was 1.6, instead of 3 (theoretical value);
thus, an excess of ruthenium is fixed on the modified
silica 16, which could not be removed by extraction
(Soxhlet) with DCM.
Ca ta lytic P r op er ties of Hom ogen eou s Ca ta lysts.
The activity and selectivity of the new catalysts 3 and
5 (Scheme 1) have been evaluated comparatively to
catalyst 1 (commercially available) in classical oxidation
reactions of alcohols with N-methylmorpholine oxide
(NMMO).14,16 Benzyl alcohol, 1-octanol, trans-2-hexen-
1-ol, cis-2-hexen-1-ol, 2-octanol, and 1-phenylethanol,
respectively, dissolved in DCM (100 µL/5 mL) were
treated at room temperature with NMMO (2 equiv) in
the presence of the tested catalyst (2 mol %). The crude
reaction mixtures were quantitatively analyzed by gas
chromatography (GC) or high-performance liquid chro-
matography (HPLC). The conversions (percentages of
transformed alcohol) and selectivities to aldehydes
(percentages of aldehyde with respect to alcohol con-
verted), after 6 h of reaction for catalysts 3 and 5
(phosphinite ligands) and 1.5 h of reaction for catalyst
1 (Ph3P ligand), are given in Table 2. The only side
product observed, in very few cases, was the corre-
sponding carboxylic acid.
The selectivities of the three catalysts are high and
quite similar, though the selectivity of the Ru-phos-
phinite complexes appears slightly better (see entry 2,
oxidation of 1-octanol). On the other hand, the Ru-
phosphine catalyst 1 is significantly more active (short
reaction time and high conversion) than the correspond-
ing Ru-phosphinite catalyst 3, but some activity is
restored in the case of catalyst 5 built from the tripodal
phosphinite ligand. The catalytic performances of 5 were
considered as good enough to envisage its heterogeni-
zation. Moreover, a great advantage of 5 over 1 was its
stability in an air atmosphere, allowing easy manipula-
tions for further catalyst recovery. A solution of 5 in
CDCl3 was stable for several hours, while complex 1
transformed into triphenylphosphine oxide within 90
min (31P NMR analysis).
(18) Dulie`re, E.; Marchand-Brynaert, J . Synthesis 2002, 39.

808 Organometallics, Vol. 22, No. 4, 2003
Dulie`re et al.
Ta ble 2. Ca ta lytic P er for m a n ces of Va r iou s Hom ogen eou s Ca ta lysts in Alcoh ol Oxid a tion (Alcoh ol
Con ver sion , Selectivity to Ald eh yd e) a t 20 °C w ith NMMO a s Oxid a n t (2 m ol % Ru )
cat. 3a
cat. 5a
cat. 1b
entry
alcohol
conversn, %
select, %
conversn, %
select, %
conversn, %
select, %
1
2
3
4
5
6
benzyl alcohol
1-octanol
2-hexen-1-ol (trans)
2-hexen-1-ol (cis)
2-octanol
32
22
13
30
23
24
94
100
100
100
73
58
50
87
66
82
91
90
100
100
94
97
90
100
100
76
93
82
100
100
1-phenylethanol
a
b
After 6 h of reaction. After 1.5 h of reaction.
Ta ble 4. Ru Lea ch in g fr om th e Su p p or ts d u r in g
th e Recyclin g Assa ysa
Sch em e 5. P r ep a r a tion of Heter ogen eou s
Refer en ce Ca ta lyst 17
cat. 7
[Ru],
cycle mmol/g
cat. 11
[Ru],
mmol/g
0.077 100 0.158
cat. 16
[Ru],
mmol/g
cat. 17
[Ru],
mmol/g
%
%
%
%
0
1
2
3
4
5
6
100 0.618
48 0.353
0.021
100 0.800
57 0.119
100
15
6
0.053
0.008
69 0.076
10 <0.005
<0.005
3
1
0.048
0.027
0.008
0.008
<0.005
0.006
<0.005
<0.005
3
1
1
Ta ble 3. Ca ta lytic P er for m a n ces of Va r iou s
Heter ogen ized Ca ta lysts in th e Oxid a tion of
1-Octa n ol w ith NMMO a s Oxid a n t (T ) 20 °C; t ) 6
h ; 2 m ol % Ru )a
a
[Ru] was measured by atomic absorption analysis of the
solutions, after filtration of the catalyst; values are expressed per
g of catalyst and as the percentage of released Ru from the native
catalyst.
cat. 7
cat. 11
cat. 16
cat. 17
X, S,
X, S,
X, S,
X, S,
run
%
%
A
%
%
A
%
%
A
%
%
A
one;14,19 (ii) catalyst immobilization reduces mobility and
accessibility of the partners and, hence, the reactivity
of the system.
The heterogenized catalysts 11 and 16 were found to
be stable under an air atmosphere, as the corresponding
soluble complexes. After prolonged storage (3-6 months)
at room temperature, our silica-bound catalysts showed
the same features (activity and selectivity) as the freshly
prepared catalysts in the standard oxidation reaction
of 1-octanol (run 1, Table 3).
1
2
3
4
5
6
48 100 27 75
0
85 34 46
100
85 14 48
87 21
0
5
0
4
0
22 100 16 26 100 14
40 100 32 27 100 17
20 100 16 14 100
9
0
0
15 100 10
0
0
a
The definitions of X, S, and A are as follows. X ) conversion:
GC analysis (percentage of transformed 1-octanol). S ) selectivity
toward octanal: GC analysis (percentage of octanal versus trans-
formed 1-octanol). A ) activity expressed as mmol of alcohol
transformed per mmol of Ru effectively present in the catalyst (see
Table 4).
Recyclin g of Heter ogen ized Ca ta lysts. After their
first use, all the catalysts were filtered off from the
solutions and engaged again in successive oxidation
reactions of 1-octanol under the standard conditions.
Catalyst 7 was totally deactivated after the first
utilization (Table 3, run 2); thus, Ru complexes with
phosphinite ligands directly bound to silanol groups of
the silica support, entities susceptible to be formed
during the synthesis of catalysts 11 and 16, could not
be responsible for the activity of these catalysts, after
the first run. Catalysts 11, 16, and 17 are still opera-
tional in the second use, with a total selectivity (Table
3, run 2). However, the monodentate catalyst 11 induced
a very low conversion ratio, comparatively to the tri-
dentate catalyst 16 and the reference 17; this catalyst
became totally inactive in the third use (Table 3, run
3). On the other hand, the tridentate catalyst 16 and
the reference 17 worked until the fifth and sixth uses,
respectively (Table 3, runs 4-6).
Ca ta lytic P r op er ties of Heter ogen ized Ca ta lysts.
The performances of heterogenized catalysts 7 (Scheme
2), 11 (Scheme 3), and 16 (Scheme 4) (Table 1) have been
evaluated comparatively to the reference catalyst 17,
formed by ligand exchange between RuCl2(PPh3)3 (1)
and polystyrene-supported triphenylphosphine (Scheme
5) according to a recent literature procedure.11 In all
cases, the reaction considered was the oxidation of
1-octanol with NMMO under the standard conditions
previously described.
The first use of the heterogenized catalysts 7, 11, and
16 (Table 3, first run) gave conversion ratios different
from those of the corresponding homogeneous catalysts
3 and 5 (Table 2, entry 2); monodentate catalysts 7 and
11 were more active than their soluble model 3 (48%
and 75% of conversion versus 22%), and the use of a
spacer arm to anchor the ligand brought a significant
advantage. The tridentate catalyst 16 was slightly less
active than its soluble model 5 (46% of conversion versus
58%), while the reference system 1 (phosphine ligands
instead of phosphinite ones) became sluggish when
immobilized on PS: 48% of conversion after 6 h (17)
versus 97% after 1.5 h (1). These observations could
result from two antagonistic effects: (i) catalyst hetero-
genization favors the monomeric form of the ruthenium
complex, which should be more active than the dimeric
During the recycling assays, leaching of ruthenium
from the supports occurred; this has been measured by
atomic absorption analysis of the solutions, after re-
moval of the catalysts upon filtration (Table 4). Taking
into account the loss of metal, a “normalized catalytic
(19) (a) Rhodes, L. F.; Sorato, C.; Venanzi, L. M.; Bachechi, F. Inorg.
Chem. 1988, 27, 604. (b) Sime, W. J .; Stephenson, T. A. J . Organomet.
Chem. 1978, 161, 245.

Novel Phosphinite-Ruthenium(II) Complexes
Organometallics, Vol. 22, No. 4, 2003 809
Ta ble 5. Effects of th e Ca ta lyst P r ein cu ba tion (17
h a t 20 °C) in Solven t, Nea t, a n d Con ta in in g On e of
th e Rea ction P a r tn er sa
petitive ligands such as NMMO (or NMM). After this
first step, the further slow deactivation could result from
a very gradual metal leaching, progressive poisoning,
and modification of the catalytic sites, mechanical
degradation of the silica particles, or other undefined
factors.
1-octa- octa- octanoic
cat.
DCM
nol
nal
acid
NMMO NMM
7
X, %
S, %
Ru loss, % 15
44
100
31
100
13
41
87
1
33
92
1
23
100
19
45
87
4
38
93
<1
25
100
19
58
90
6
41
88
<1
<5
n.d.
98
22
100
61
23
100
15
11
100
51
55
84
18
38
93
5
Con clu sion
16 X, %
S, %
66
89
2
35
100
We have succeeded in the preparation of novel silica-
bound ruthenium catalysts based on phosphinite ligands.
In particular, the tridentate complex 16 was designed
to be stable under air and oxidative conditions and,
hence, to be reusable. This catalyst (16) has been
compared to the corresponding silica-bound monoden-
tate catalysts 11, constructed with the same spacer arm,
and 7, devoid of spacer arm. With respect to the related
soluble catalysts 3 and 5, the heterogenized monoden-
tate catalysts 7 and 11 were 2 and 3 times more active,
while the tridentate complex 16 showed almost the
same activity. All catalysts were endowed with good to
excellent selectivities.
Ru loss, %
17 X, %
S, %
Ru loss, % <1
a
For definitions of X and S, see Table 3.
activity” could be calculated after each cycle: i.e., the
ratio of 1-octanol transformed per mmol of Ru actually
present in the catalyst. We are aware that such an
activity is difficult to interpret, since a lot of weakly
bound ruthenium is lost in the first run. However, the
values collected in Table 3 (column A for each catalyst)
showed that Ru leaching is well responsible for the
deactivation of the tridentate catalyst (16): after exten-
sive loss of about 60% of metal in the first cycle, the
catalyst works with almost constant activity (and 100%
selectivity) during three further cycles. On the other
hand, the loss of about 50% of the metal after the first
cycle could not be the only reason for the dramatic
deactivation of the corresponding monodentate catalyst
(11). Only small quantities of Ru were lost from the
polymer-bound catalyst (17), most probably because the
catalytic sites are mainly buried into the material bulk;
leaching from the more active surface catalytic sites
caused an important deactivation.
The recycling experiments were somewhat disap-
pointing: the monodentate catalysts 7 and 11 could not
be reused and the tridentate catalyst 16 refused to work
again after four cycles. We identified the metal leaching
due to the presence of NMMO/NMM as being the main
factor of catalyst deactivation. We thus tried to replace
this oxidant with oxygen (of air), hydrogen peroxide, or
tert-butyl hydroperoxide, but without success: no cata-
lytic effect on the oxidation of 1-octanol could be
observed in the presence of our silica-bound Ru com-
plexes. Also, the TEMPO/O2 system,20 recently de-
scribed, did not work under our experimental conditions.
The challenging search for recoverable oxidation
catalysts based on Ru complexes has been studied by
other groups, leading to similary mitigated achieve-
ments. Ram et al.12 reported the grafting of an Ru-
Schiff base complex on cross-linked polystyrene and the
use of the resulting catalyst for the oxidation of benzyl
alcohol with molecular oxygen. The problem of metal
leaching (35%) after the first cycle was mentioned by
the authors; the catalyst was found to be stable for four
cycles. Leadbeater11 prepared the resin-bound Ru-
triphenylphosphine catalyst 17 for transfer hydrogena-
tion and hydrocarbon oxidation reactions. This material
catalyzed the oxidation of cyclohexanol with tert-butyl
hydroperoxide in yield comparable to that of the soluble
catalyst 1. The supported catalyst could be reused, but
the number of cycles was not reported by the authors.
In our hands, catalyst 17 was working for five cycles in
the NMMO oxidation of 1-octanol. Heterogenization of
a Ru-2-(phenylazo)pyridine complex on silica was
described by Reedijk et al.13 This catalyst was found to
be active for stilbene epoxidation and alcohol oxidation,
but with a very low selectivity. Recycling experiments
performed on the catalytic oxidation of n-butanol with
NaBrO3 revealed the loss of 30% of the initial Ru after
the first cycle; the catalyst kept its activity over four
cycles. Accordingly, our silica-bound catalyst 16, con-
taining an original Ru-triphosphinite complex, ranges
among the best catalysts which could be actually
We have no explanation for the sudden break in
activity of catalysts 7 (cycle 2), 11 (cycle 3), 16 (cycle 5),
and 17 (cycle 6), although significant quantities of Ru
are still present on the supports (about 20%, 50%, 40%,
and 75%, respectively; see Table 4).
Sta bility of Heter ogen ized Ca ta lysts. Since the
synthesis procedure of all silica-bound catalysts involved
a final washing with DCM, this solvent could not be
solely responsible for the important Ru leaching during
the first catalytic cycle. We therefore investigated the
individual effects of each component of the oxidation
mixture on this process. Fresh catalysts 7, 16, and 17
were incubated into DCM solutions containing respec-
tively 1-octanol, octanal, octanoic acid, N-methylmor-
pholine N-oxide (NMMO), and N-methylmorpholine
(NMM) (2 mol % Ru). After 17 h at 20 °C, soluble
ruthenium was assayed by atomic absorption analysis
(Table 5). The pretreated catalysts were then recovered
by filtration and used in the standard oxidation proce-
dure of 1-octanol (see conversions and selectivities in
Table 5).
An important metal leaching occurred in the presence
of NMMO and NMM, to a lesser extent. The catalytic
features (activity and selectivity) of catalysts 7, 16, and
17 after preincubation were similar to those of the
corresponding fresh catalysts, except after pretreatment
with NMMO; in this case, the situation was similar to
the second use of catalysts described above (see Table
3, run 2). Accordingly, the deactivation of our heterog-
enized catalysts is mainly due to an important loss of
weakly bound ruthenium under the influence of com-
(20) Dijksman, A.; Marimo-Gonzalez, A.; Mairata i Payeras, A.;
Arends, I. W. C. E.; Sheldon, R. A. J . Am. Chem. Soc. 2001, 123, 6826.

810 Organometallics, Vol. 22, No. 4, 2003
Dulie`re et al.
mL/min) and UV detection at 240 nm. Calibration was carried
out with standard solutions.
obtained by heterogenizing a homogeneous Ru complex
active for alcohol oxidation. However, better results
were obtained by Ley et al. when immobilizing perru-
thenate ions (RuO4-) on an ion exchange resin (Am-
berlyst A-26)21 or on a mesoporous silicate (MCM-41)
derivatized with ammonium groups.22
Dich lor otetr akis(dim eth yl su lfoxide)r u th en iu m (II) (2).
A solution of RuCl3‚nH2O (5.32 g; about 0.02 mol) in DMSO
(25 mL) was refluxed for 5 min. The dark red solution was
concentrated under vacuum, until precipitation occurred. After
addition of acetone (20 mL) and filtration, a yellow powder
was recovered which was washed with acetone (3 × 10 mL)
and ether (1 × 10 mL). After several days, a second fraction
of yellow crystals was formed in the mother liquor and washed
as above. Total yield: 8.69 g (85%). Mp: 198-200 °C (lit. 193
°C dec). Anal. Found (calcd) for C8H24Cl2O4RuS4: C, 19.92
(19.83); H, 4.90 (4.99). IR (KBr, cm-1): 3019, 2917 (C-H),
1119-1096 (br, st O-S-Ru), 921 (st S-O-Ru). 1H NMR
(CDCl3, δ): 3.53, 3.51, 3.44, and 3.32 (Me of DMSO, S-Ru
linked), 2.74 (Me of DMSO, O-Ru linked), 2.63 (Me of free
DMSO); ratio S-Ru:O-Ru 4:1. 13C NMR (CDCl3, δ): 47.6,
46.6, 44.4, and 44.2 (Me of DMSO, S-Ru linked), 38.8 (Me of
DMSO, O-Ru linked). MS (m/e (%)): 485.3 (2, M + 1), 449
(15), 407 (6), 371 (35, M - Cl - DMSO), 329 (50, M - 2DMSO),
309 (100).
Exp er im en ta l Section
Ma ter ia ls. Reagents (quality 99+%) and solvents (analyti-
cal grade) were purchased from Acros, Aldrich, or Fluka and
used without purification. The reactions were performed under
an argon atmosphere, in flame-dried glassware. Dichloro-
methane and -ethane (DCM and DCE) were distilled over
CaH2; tetrahydrofuran (THF) was distilled over Na.
Silica used for the preparation of catalysts was obtained
from Aldrich (reference number 40,360-1; specific area 300 m2/
g; 2.5 mmol of silanol/g; 100 Å; 70-230 mesh). The material
has been washed with methanol (Soxhlet; 3 days) and ether
(Soxhlet; 2 days) and then dried under vacuum (70 °C; 5 ×
10-3 mbar).
1-((Dip h en ylp h osp h in o)oxy)-2,2-b is(((d ip h en ylp h os-
p h in o)oxy)m eth yl)p r op a n e (4). Chlorodiphenylphosphine
(0.70 g, 3.0 mmol, 3 equiv) was added dropwise to a suspension
of 1,1,1-tris(hydroxymethyl)ethane (0.13 g, 1.0 mmol, 1 equiv)
and triethylamine (0.4 mL, 2.8 mmol, 3 equiv) in CH2Cl2 (5
mL). The mixture was refluxed for 2 h under an argon
atmosphere and then filtered. Concentration of the filtrate
under vacuum gave crude 4 as a yellow oil (sensitive toward
air oxidation) that was not further purified but directly
engaged in the synthesis of complex 5. Yield: 0.74 g (100%).
Reference catalyst 17 (dichlorotris(triphenylphosphine)-
ruthenium(II) (1) immobilized on polystyrene (PS)) has been
prepared according to ref 11; its ruthenium content was
determined by TG analysis under air, from the difference of
weight losses at 850 °C between PS-PPh2 (94.4%) and PS-
PPh2RuCl2(PPh3)2 (84.4%), by assuming that the final residue
corresponds to RuO2 (0.8 mmol of Ru/g of catalyst).
Meth od s. Melting points (uncorrected) were determined
with an electrothermal apparatus. Elemental analyses were
carried out at University College London (Christopher Ingold
Laboratory). 1H (200 or 300 MHz) and 13C (50 or 75 MHz) NMR
spectra were recorded on Varian Gemini spectrometers with
CDCl3 as the solvent and SiMe4 as internal reference. Chemi-
cal shifts are reported as δ values downfield from TMS;
coupling constants J are given in hertz. 31P NMR spectra were
recorded on a Bruker WM250 spectrometer at 250 MHz with
H3PO4 as external reference. IR spectra were taken with a Bio-
Rad FTS 135 apparatus. Mass spectra were obtained with a
Finnigan MAT TSQ-70 spectrometer in the FAB mode (Ion
Tech, 8 keV).
TGA measurements were carried out with a microbalance
from Mettler Toledo (TGA/SDTA 851e) under air flux (100 mL/
min), between 25 and 850 °C, at a heating rate of 10 °C/min.
Atomic absorption analyses were performed on a Perkin-
Elmer 5000 instrument using an air/acetylene flame. The
samples were prepared as follows: an aliquot of the catalyst
was incubated overnight in 30% aqueous HClO4 (2 mL) and
then filtered and reincubated for 10 min in 10% aqueous
NaOH (2 mL). Silica was filtered and washed with water
(HPLC grade, 3 × 2 mL). The filtrates were collected and
diluted to 50 mL with water.
3
1H NMR (CDCl3, δ): 7.5-7.2 (m, 30 H, Ar), 3.77 (d, 6 H, J HP
) 6.5, O-CH2), 1.0 (s, 3 H, CH3-C). 13C NMR (CDCl3, δ): 128.2
2
3
(d, J CP ) 7, C-o, PPh2), 129.1 (s, C-p, PPh2), 130.2 (d, J CP
)
)
1
2
22, C-m, PPh2), 140.0 (d, J CP ) 18, C-i, PPh2), 71,6 (d, J CP
19, CH2O), 43.5 (s, Me-C), 17.1 (d, 3J CP ) 7, CH3-C). MS (m/e
(%)): 672.3 (4, M - 1; C41H39O3P3), 595.2 (12), 471.1 (28), 200.9
(100, OPPh2).
Dich lor o[1-((d ip h e n ylp h osp h in o)oxy)-2,2-b is(((d i-
p h en ylp h osp h in o)oxy)m et h yl)p r op a n e]r u t h en iu m (II)
(5). To a refluxing solution of 2 (2.30 g, 5 mmol, 1 equiv) in
CH2Cl2 (15 mL) was added a solution of 4 (3.96 g, 5 mmol, 1
equiv) in CH2Cl2 (10 mL). The yellow solution became dark
red. After 30 min of reflux, the solvent was evaporated under
vacuum. The solid residue (6.2 g) was washed with water and
purified by exclusion chromatography on Sephadex gel. Yield:
64%. Mp: 183.2-184.6 °C (recrystallization from CHCl3; orange
solid). IR (KBr, cm-1): 3055, 2927, 1435, 1070, 1019. 31P NMR
(CDCl3, δ): 153.7 (br s), 142.8 (br s). MS (m/e (%)): isotopic
massif centered at 808.8 (96, M - Cl; C41H39Cl2O3P3Ru), 200.9
(100, OPPh2). X-ray data: λ ) 0.710 69 Å; monoclinic; a )
21.985 Å; b ) 25.644 Å; c ) 23.103 Å; R ) 90°; â ) 114.54°; γ
) 90°.
Gel permeation chromatography was performed on a home-
made glass column packed with a cyclodextrin gel (LH 20
Sephadex), swollen, and eluted with methanol.
N -(3-(Tr ie t h oxysilyl)p r op yl)-N ′-(4-(h yd r oxye t h yl)-
p h en yl)u r ea (8). A solution of 4-aminophenethyl alcohol (0.14
g, 1 mmol, 1 equiv) in THF (5 mL) was added dropwise to a
solution of 3-isocyanato-1-(triethoxysilyl)propane (0.29 mL, 1.1
mmol, 1.1 equiv) in THF (5 mL). The mixture was stirred for
3 h at 20 °C and then for 2 h at reflux. Solvent evaporation
led the product to crystallize as brown needles. Crude 8 was
directly used for the preparation of the silica-grafted material
GC analyses (oxidation solutions of 1-octanol, 2-octanol, and
cis- and trans-2-hexen-1-ol) were carried out on a GC 8000 TOP
Interscience gas chromatograph equipped with a flame ioniza-
tion detector and a fused silica capillary column (Optima 5
from Macherey-Nagel; 30 m × 25 mm). Temperature program-
ming was as follows: 60 °C for 3 min and then an increase of
5 °C/min up to 110 °C and 20 °C/min up to 250 °C. HPLC
analyses (oxidation solutions of benzyl alcohol and 1-phenyle-
thanol) were performed on a Beckman instrument (System
Gold; Model 128 Solvent Module and Model 168 Detector)
equipped with an Alltima C18 column (5 mm; 250 mm × 4.6
mm) from Alltech. We used H2O-MeOH (60:40) as eluent (1
1
9. Yield: 100% (crude). IR (KBr, cm-1): 3471, 2976, 1650. H
NMR (CDCl3, δ): 7.30 (d, 2 H, 3J HH ) 8, Ar), 7.24 (d, 2 H, 3J HH
3
) 8, Ar), 6.54 (s, 1 H, NH-Ar), 5.12 (t, 1 H, J HH ) 6, NH-
3
3
CH2), 3.90 (q, 6 H, J HH ) 5, OCH2), 3.38 (t, 1 H, J HH ) 7,
3
OH), 3.32 (td, 2 H, CH2-OH), 2.90 (t, 2 H, J HH ) 6, CH2-
Ar), 1.8-1.7 (m, 4 H, CH2CH2NH), 1.30 (t, 9 H, 3J HH ) 7, CH3),
0.79-0.70 (m, 2 H, Si-CH2). 13C NMR (CDCl3, δ): 156.3 (Cd
O), 137.3 and 133.6 (CAr), 129.6 and 121.1 (CHAr), 63.5 (CH2-
OH), 58.4 (CH2O-Si), 42.7 (CH2-NH), 38.5 (CH2-Ar), 18.2
(21) Hinzen, B.; Lenz, R.; Ley, S. V. Synthesis 1998, 977.
(22) Bleloch, A.; J ohnson, B. F. G.; Ley, S. V.; Price, A. J .; Shephard,
D. S.; Thomas, A. W. J . Chem. Soc., Chem. Commun. 1999, 1907.

Novel Phosphinite-Ruthenium(II) Complexes
Organometallics, Vol. 22, No. 4, 2003 811
(CH3-CH2O), 7.7 (CH2-Si). MS (m/e (%)): 385.2 (20, M + 1;
aminophenyl)-2,2-bis(hydroxymethyl)propan-1-ol (13;18 0.27 g,
1.28 mmol, 1 equiv) in pyridine (3 mL) was added to a
suspension of grafted silica 12 (2 g, 2.4 mmol of NCO, 1.9
equiv) in THF (50 mL). The mixture was refluxed for 2 days.
The grafted silica 14 was filtered off as a yellow powder and
washed in a Soxhlet apparatus with CH2Cl2 (1 day). Yield of
functionalization: 0.45 mmol of tripodal ligand/g. TGA (air):
loss of water 1.6%; loss of organic materials 13.4%. Thus, 1 g
of silica 14 contains 0.134 g (0.45 mmol) of C15H23N2O4. Anal.
Found (calcd) for (SiO2)nC15H23N2O4‚3H2O: C, 6.38 (6.55); H,
1.13 (1.06).
C
18H32N2O5), 339.1 (100, M - EtOH).
Silica Gr a fted w ith Dich lor otr is(d ip h en ylp h osp h in e)-
r u th en iu m (II) (7). Chlorodiphenylphosphine (0.49 g, 2.2
mmol, 1 equiv) was added dropwise to a suspension of silica
(1.03 g, 2.5 mmol, 1.1 equiv of silanol groups) in CH2Cl2 (10
mL) containing triethylamine (0.3 mL, 2.2 mmol, 1 equiv). The
mixture was refluxed for 2 h. The grafted silica 6 was filtered
off, washed twice with CH2Cl2 (2 × 5 mL), and placed again
in fresh CH2Cl2 (10 mL). After addition of the Ru complex 2
(0.51 g, 1.0 mmol, 0.45 equiv), the mixture was refluxed for
90 min. The yellowish silica-grafted catalyst 7 recovered by
filtration was washed with CH2Cl2 in a Soxhlet apparatus for
2 days and then with ether, in the same manner. Yield of
grafting: 0.435 mmol of derivatized silanols/g. TGA (air): loss
of water 2.1%; loss of organic materials 8.6%. Thus, 1 g of silica
7 contains 0.086 g (0.145 mmol) of C36H30Cl2P3Ru. Atomic
absorption: 0.077 mmol of Ru/g.
Silica Gr a fted w ith N-P r op yl-N′-(4-(2-h yd r oxyeth yl)-
p h en yl)u r ea (9). A solution of triethoxysilane 8 (0.49 g, 1.3
mmol, 1 equiv) in THF (30 mL) was added dropwise to a
suspension of silica (0.63 g, 1.6 mmol of silanol groups, 1.2
equiv) in THF (30 mL). The mixture was stirred for 3 h at 20
°C and then refluxed for 3 days. The grafted silica 9, recovered
by filtration as a white powder, was washed in a Soxhlet
apparatus, successively with methanol (2 days), THF (2 days),
CH2Cl2 (2 days), and ether (2 days), and then dried under
vacuum (5 × 10-3 mbar) at 70 °C for 4 h. Yield of grafting: 0.5
mmol of derivatized silanols/g. TGA (air): loss of water 2%;
loss of organic materials 11%. Thus, 1 g of silica 9 contains
0.11 g (0.50 mmol) of C12H17N2O2. Anal. Found (calcd) for
(SiO2)nC12H17N2O2‚2H2O: C, 7.18 (7.28); H, 1.18 (1.07).
Silica Gr a fted w ith Dich lor otr is[N-p r op yl-N′-(4-(((d i-
p h e n y lp h o s p h i n o )o x y )e t h y l)p h e n y l)u r e a ]r u t h e n -
iu m (II) (11). Chlorodiphenylphosphine (0.27 g, 1.2 mmol, 5
equiv) was added dropwise to a suspension of grafted silica 9
(0.52 g, 0.24 mmol, 1 equiv of alcohol functions) in CH2Cl2 (10
mL) containing triethylamine (0.14 mL, 1.0 mmol, 4 equiv).
The mixture was refluxed for 5 h. The grafted silica (10) was
filtered off, washed twice with CH2Cl2 (2 × 5 mL), and then
poured again into CH2Cl2 (10 mL). The mixture was refluxed,
and a solution of Ru complex 2 (0.14 g, 0.28 mmol, 1.4 equiv)
in CH2Cl2 (5 mL) was added in one fraction. After 2 h, the
silica-grafted catalyst 11 was filtered off and washed in a
Soxhlet apparatus, successively with CH2Cl2 (2 days) and ether
(2 days). Yield: 80% ()ratio of alcohol conversion) of a pale
yellow powder. TGA (air): loss of water 1.6%; loss of organic
materials 16.8%. Thus, 1 g of silica 11 contains 0.168 g (0.13
mmol) of C72H78N6O6P3RuCl2: i.e., 0.39 mmol/g of grafted
ligand. Anal. Found (calcd) for (SiO2)nC72H78N6O6P3RuCl2‚
10H2O: C, 10.68 (10.64); H, 1.42 (1.22); P, 1.14 (1.14). Atomic
absorption: 0.158 mmol of Ru/g.
Silica Gr a ft ed w it h [N-P r op yl-N′-(4-(2,2,2-(((d ip h en -
ylp h osp h in o)oxy)m e t h yl)e t h yl)p h e n yl)u r e a ]r u t h e n -
iu m (II) (16). Lithium hexamethyldisilazide (0.24 g, 1.42
mmol, 6 equiv) and (dimethylamino)pyridine (0.26 mL, 2.09
mmol, 9.1 equiv) were added to a stirred suspension of grafted
silica 14 (0.51 g, 0.23 mmol of triol, 1 equiv) in THF (20 mL)
at -65 °C. After 30 min, the temperature was raised to -40
°C, and chlorodiphenylphosphine (0.61 g, 2.75 mmol, 12.0
equiv) was added dropwise. The temperature was slowly raised
to 20 °C. The mixture was then refluxed for 20 h. The grafted
silica 15 was filtered off from the hot solution, rinsed succes-
sively with acetone (2 × 5 mL) and CH2Cl2 (2 × 5 mL), and
then poured into fresh CH2Cl2 (20 mL) and refluxed. After 5
min, the Ru complex 2 (0.22 g, 0.45 mmol, 2 equiv) was added
and the mixture was further refluxed for 24 h. After filtration,
the recovered silica-grafted catalyst 16, as a yellow powder,
was rinsed with acetone and washed in a Soxhlet apparatus
with CH2Cl2 (1 day). Yield: 82% (ratio of alcohol conversion).
TGA (air): loss of water 1.7%; loss of organic materials 28.1%.
Thus, 1 g of silica 16 contains 0.281 g (0.33 mmol) of C51H50
-
Cl2N2O4P3Ru. Anal. Found (calcd) for (SiO2)nC51H50Cl2N3O4P3-
Ru‚H2O: C, 8.79 (9.68); H, 1.18 (0.99); N, 0.52 (0.44). Atomic
absorption: 0.618 mmol Ru/g.
Ca ta lytic Oxid a tion of Alcoh ols: Gen er a l P r oced u r e.
The Ru catalyst (2 mol %, 0.02 equiv of Ru) was introduced
into a solution of alcohol (1 mmol, 1 equiv) and N-methylmor-
pholine N-oxide (2 mmol, 2 equiv) in CH2Cl2 (5 mL). The
mixture was stirred at 20 °C, in an air atmosphere, for 6 h.
Under these experimental conditions, Ru-catalyzed slow deg-
radation of the solvent was not an embarrassing side reaction.
The evolution of the reaction was monitored (30 min; 1 h 30
min; 3 h; 6 h) by GC with tetradecane as internal standard
(for the oxidation of 1-octanol, 2-octanol, cis and trans 2-hexen-
1-ols) or by HPLC (for the oxidation of benzyl alcohol and
1-phenylethanol). Heterogenized Ru catalysts were separated
by filtration and reused under the same conditions. The
filtrates were analyzed by atomic absorption to determine the
amount of released Ru (Table 4).
Eva lu a tion of th e Ca ta lyst Sta bility: Gen er a l P r oce-
d u r e. A suspension of heterogenized catalyst ((2 mol %, 0.02
equiv of Ru) in CH2Cl2 (5 mL) containing one of the partners
of the oxidation reaction mixture (1 mmol, 1 equiv) was stirred
overnight at room temperature into a closed flask. After
filtration, the catalyst was reused in the standard oxidation
procedure. The filtrate was analyzed by atomic absorption
spectrometry (assay of released Ru; Table 5).
Silica Gr a fted w ith P r op yl Isocya n a te (12). 3-(Trieth-
oxysilyl)propyl isocyanate (4 g, 16 mmol, 2.1 equiv) was added
dropwise to a suspension of dried silica (3 g, 7.5 mmol, 1 equiv
of silanol functions) in THF (150 mL). The mixture was stirred
for 3 h at 20 °C and then refluxed for 20 h. The grafted silica
12 was filtered off, washed with THF (2 × 50 mL), and directly
used for the coupling of 13. Yield (white powder) of function-
alization: 1.2 mmol of isocyanate/g. TGA (air): loss of water
1.2%; loss of organic materials 10.5%. Thus, 1 g of silica 12
contains 0.105 g (1.2 mmol) of C4H6NO.
Ack n ow led gm en t. We gratefully acknowledge the
UCL and FNRS for financial support. J .M.-B. is a senior
research associate of the FNRS (Belgium).
Silica Gr a ft ed w it h N-P r op yl-N′-[4-(2,2-b is(h yd r oxy-
m eth yl)p r op a n -1-ol)p h en yl]u r ea (14). A solution of 3-(4-
OM020876+