A tripodal sulfur ligand for the selective ruthenium-catalysed hydrogenation of dimethyl oxalate

Article abstract of DOI:10.1039/b601216g

The first example of a catalyst utilising a sulfur-based ligand [MeC(CH₂SBu)₃] for the selective hydrogenation of dimethyl oxalate to methyl glycolate is reported. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b601216g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A tripodal sulfur ligand for the selective ruthenium-catalysed
hydrogenation of dimethyl oxalate{
Brian Boardman, Martin J. Hanton,* Hendrik van Rensburg and Robert P. Tooze
Received (in Cambridge, UK) 26th January 2006, Accepted 13th April 2006
First published as an Advance Article on the web 2nd May 2006
DOI: 10.1039/b601216g
much improved binding of metal ions in comparison to acyclic
The first example of a catalyst utilising a sulfur-based ligand
[MeC(CH2SBu)3] for the selective hydrogenation of dimethyl
oxalate to methyl glycolate is reported.
thioethers.15 It is noteworthy, that simple thioether-type sulfur
ligands have previously found application in hydrogenation
catalyst systems in combination with Rh, Pd, Ir and Pt, for
unsaturated substrates, such as (functionalised) alkenes and
oxygenates such as ketones, but never esters.14" Regarding
tripodal sulfur ligands, MeC(CH2SMe)3 specifically,17 has found
widespread application in the coordination chemistry of transition
metals,18 including ruthenium.19
The hydrogenation of esters to alcohols is a conversion of
industrial importance, being employed in the production of fatty
alcohols,1,2 and being a potential route to ethane-1,2-diol via
dimethyl oxalate.2,3 Currently, all commercial fatty ester hydro-
genation plants employ heterogeneous catalysts,1,2 a suitable
homogeneous alternative having yet to be identified. Indeed, the
homogeneously-catalysed hydrogenation of esters to yield alcohols
is a notoriously difficult transformation to effect, illustrated by the
relative sparsity of catalysts reported for this conversion.4–12 This
prior art has been discussed previously.11{
Recently, a Ru-containing system based upon a tripodal
phosphine ligand, MeC(CH2PPh2)3 (TriPhosPh), was described
that represented a significant step forwards, providing near
quantitative conversion of dimethyl oxalate (DMO) through to
ethanediol (ED) at a significantly faster rate than previously
reported systems (Fig. 1).10,11 This system has also found wider
application in industry.13 However, catalysts that give conversion
to methyl glycolate (MG), do so at a much reduced rate.4–7
Elsevier et al. examined a number of different ligands for this
transformation with ruthenium, and concluded that a facially-
capping tripodal phosphine ligand was the optimal choice.10 In
fact, phosphines are the only class of ligand reported to date, to
generate an active ruthenium catalyst for this reaction.4–13§ Our
own examination of the literature lead us to conclude that in
addition to the nature of the binding mode, soft, electron-rich
donor moieties are preferential.4–13 To this end, we undertook a
study of facially-capping sulfur ligands for this chemistry.
The commercially available sulfur macrocycles 1,3,5-trithiane
and 1,4,7-trithiacyclononane were identified as suitable target
ligands, whilst a tripodal scaffold was accessed via the synthesis of
MeC(CH2SBun)3 (TriSulfBu); using a standard coupling reaction
between alkyl chloride and thiol in the presence of base.16 In an
unoptimised synthesis, MeC(CH2Cl)3, excess BunSH and NaOH
were stirred in EtOH at 70 uC for 36 d (Fig. 2). The synthesis was
inefficient, but returned the desired compound in good yield
(20.3 g, 83%). Great difficulty in achieving substitution at all three
arms was observed, in line with the observation of a low yield in
the original report of the synthesis of MeC(CH2SMe)3.17
Dimethyl oxalate has been the most widely examined substrate
to feature in literature reports of ester hydrogenation,4–7 and so
can be seen to represent a benchmark reaction in this type of
catalysis (see Fig. 1). Furthermore, its hydrogenation to ethane-
1,2-diol is of industrial interest.3 Initial tests with the sulfur
macrocycles 1,3,5-trithiane and 1,4,7-trithiacyclononane were
unsuccessful, no conversion being observed (see Table 1, entries
1 and 2). Notably, no decomposition of the DMO substrate
occurred, but the metal was lost as ‘ruthenium-black’, indicating
an inability of these ligands to stabilise ruthenium under the
reaction conditions, in contrast to what may have been expected.18
However, upon application of the TriSulfBu ligand in concert with
a ruthenium source, successful hydrogenation was achieved
through to methyl glycolate (Table 1, entries 3–6). In all cases
with the TriSulfBu ligand, upon opening the autoclave a
transparent yellow solution was observed, with no precipitate in
evidence, indicating the improved ability of the tripodal sulfur
ligand to stabilise ruthenium(II) under these conditions, compared
to the sulfur macrocycles.
Although less widely employed than phosphines, sulfur-based
ligands also have a proven track record in catalysis,14 and in the
case of sulfur macrocycles it has been suggested that these offer
Fig. 1 The hydrogenation of dimethyl oxalate (DMO).
Sasol Technology UK, St Andrews Laboratory, Purdie Building, North
Haugh, St Andrews, Fife, UK KY11 8FJ.
E-mail: martin.hanton@uk.sasol.com; Fax: +44(0)1707 383 647;
Tel: +44(0)1334 460 830
{ Electronic supplementary information (ESI) available: Experimental
details, synthesis and catalysis, calculation details, kinetics. See DOI:
10.1039/b601216g
Fig. 2 The synthesis of TriSulfBu
.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2289–2291 | 2289

View Article Online
Table 1 Results of sulfur-based ruthenium catalysts in the hydrogenation of DMOa
Additive Induction Run
Entry Ru/mmol Ligand
(%)
period/min Substrate time/h Conversion (%) TONb TOFc k/mol dm23 h21
0
1
2
3
4
5
6
7
8
212
53
53
53
212
212
212
212
212
53
None
0.3
0.3
0.3
—
0.3
0.3
0.3
0.3
0.3
0.3
0.1
—
—
—
.360
y180
y180
.200
—
DMO
DMO
DMO
DMO
DMO
DMO
DMO
MG
72
20
20
136
23
24
69
0
0
0
—
—
—
100
36.9
32.2
87.2
—
—
—
—
0.74
2.6
3.4
3.2
—
—
—
—
—
0.019
0.024
0.023
—
1,3,5-Trithiane
1,4,7-Trithiacyclononane
TriSulfBu
100 (MG)
36.9 (MG)
32.2 (MG)
87.2 (MG)
0
TriSulfBu
TriSulfBu
TriSulfBu
TriSulfBu
48
TriPhosPh
P(n-Oct)3
30
DMO
DMO
CODe
5.7 100 (ED)
304
28
200
100
1415.5 50.6
50.3
0.3
0.355
—
—
n/md
—
100 (MG)
51.6 (c-C8H16
38.4 (c-C8H14
d
9
10
53
TriSulfBu
)
)
a
General conditions: 100 uC, 80 bar H2, MeOH (30 mL), Ru(acac)3, ligand = 1.3 equiv. to Ru, catalyst (Ru) loading 1% in all cases, Zn
additive (%DMO). (mol ester moiety)(mol Ru)21
1,5-Cyclooctadiene, catalyst loading 0.1%.
.
(mol ester moiety)(mol Ru)21 h21
.
6 equiv. of ligand to Ru; n/m = not measured.
=
b
c
d
e
Ruthenium-based hydrogenation catalysts of this class, are
generally observed to exhibit an induction period, between the
attainment of reaction conditions and the onset of catalysis; this is
widely believed to correspond to a reduction of the RuIII precursor
to a RuII species.10 The sulfur ligand-based catalyst system
appeared to form with or without the presence of zinc, however
the additive did serve to reduce the length of the induction period,
and increase the rate (Table 1, entry 3 vs. 4–6). This correlates with
The wider applicability of this catalyst to substrates aside from
esters, is illustrated by its application to the hydrogenation of 1,5-
cyclooctadiene (Table 1, entry 10). At a catalyst loading of 0.1%,
the substrate was converted to cyclooctane (51.6%) and cyclo-
octene (38.4%) in 28 h; the reaction being stopped due to time
constraints rather than the catalyst losing activity. This represents
a TON of 1415.5 (mol olefin moiety)(mol Ru)21 and a TOF of
50.6 (mol olefin moiety)(mol Ru)21 h21, illustrating that the
TriSulfBu system hydrogenates olefins significantly faster than
esters, as may have been expected.
a
similar observation for the TriPhosPh-based systems.10,11
Significantly, the catalyst seems unable to hydrogenate DMO
further than MG, as even after complete conversion to MG has
occurred, prolonged reaction times see no formation of ED. To
verify this, a fresh batch of catalyst was exposed to MG as
substrate with no hydrogenation being evident after 48 h (Table 1,
entry 7).
In conclusion, the first example of a homogeneous{{ ester
hydrogenation catalyst utilising a sulfur ligand has been described,
which is notable for being the first example of such a catalyst not
based upon phosphine ligands. The TriSulfBu system provides the
most active catalyst to date that is selective towards the formation
of methyl glycolate. Furthermore, this represents the first example
of a simple thioether ligand in combination with ruthenium as a
hydrogenation catalyst.
Analysis of the gas uptake data for these experiments reveals
good reproducibility (Table 1, entries 4–6) and a reaction that is
zero-order in substrate. This leads to an average zero-order rate
constant of 2.2(¡0.2) 6 1022 mol dm23 h21 which equates to an
average TOF of 3.1(¡0.2) (mol ester moiety)(mol Ru)21 h21. In
order to provide a comparison with the phosphine-based systems,
runs were performed using P(n-Oct)3 and TriPhosPh ligands
(Table 1, entries 8 and 9) under our experimental conditions. It can
be seen that the latter system gave complete conversion to ED in
under 6 h, again with zero-order kinetics, leading to a calculated
rate constant of 3.55 6 1021 mol dm23 h21; TOF of 50.3 (mol
ester moiety)(mol Ru)21 h21.I However, with the monodentate
phosphine conversion only as far as MG was achieved, after
approximately 300 h, corresponding to a TOF of 0.3 (mol ester
moiety)(mol Ru)21 h21. Clearly, the TriPhosPh system is far more
active, but does not stop at MG. This leads to a more meaningful
comparison between TriSulfBu and P(n-Oct)3 both of which are
selective to MG; the former being more active. A comparison with
other phosphine ligands reported in the literature,10** also reveals
lower rates of DMO hydrogenation to MG, than that with
TriSulfBu, [e.g. TOF: PhP(C2H4PPh2)2, 2.5; (CH2PPhC2H4PPh2)2,
2.2; PPh3, 0.9].13 The ruthenium complexes of the general type
Ru(CO)2(CO2Me)(PR3)2 reported by Bianchi et al., also
show similar or lower rates of hydrogenation to MG than the
TriSulfBu system [e.g. TOF: R = Bun, 3.0;6 R = Pri, 0.67]{{
however these are not selective to MG, hydrogenating further to
ED subsequently.5–7
Notes and references
{ The authors note that during the submission of this work a report from
Milstein et al.,20 described a catalyst system that represents a step-change in
performance in homogeneous ester hydrogenation. Unactivated esters are
hydrogenated in a timely fashion, with good conversions at low hydrogen
pressures (5.3 atm).
§ A TON of 3 is reported for the tridentate nitrogen ligand, tris(pyr-
azolyl)borate,10,11 but given a blank run also showed a single turnover, this
result is not considered.
" The use of sulfoxide-based sulfur ligands with ruthenium for asymmetric
hydrogenations of functionalised alkenes has been described.14
I The TOF of 50.3 (mol ester moiety)(mol Ru)21 h21 determined here
correlates well with that found by the original authors of 53.5 (mol ester
moiety)(mol Ru)21 h21 under similar conditions.11
** Very similar reaction conditions were employed: 12 mL MeOH, p(H2) =
80 bar, T = 120 uC, Ru(acac)3 y20 mmol, 0.3 mol% Zn, DMO substrate,
catalyst loading y2%. The zero-order in substrate nature of this
transformation allows a direct comparison of TOFs at differing catalyst
loadings.
{{ Values calculated from the data reported in ref. 6 and ref. 7, for the
DMO to MG hydrogenation step specifically.
{{ No evidence of catalysis by Ru nanoclusters has been detected. Several
blank experiments (see ESI){ were performed in the absence of ligand,
colloidal Ru black being observed post-run, yet no hydrogenation activity
was observed. In successful runs using the TriPhosPh or TriSulfBu ligands,
Ru black formation was normally not present, a yellow/orange
homogeneous solution being observed on opening the autoclave.
2290 | Chem. Commun., 2006, 2289–2291
This journal is ß The Royal Society of Chemistry 2006

View Article Online
1 R. F. Modler, R. Gubler and Y. Inoguchi, Detergent Alcohols, CEH
Marketing Research Report, 2004; W. S. Fong, Natural Detergent
Alcohols by a Vapour Phase Ester Hydrogenation Process, PEP Review
93-2-1, 2004.
9 Y. Hara, H. Inagaki, S. Nishimura and K. Wada, Chem. Lett., 1992,
1983–1986.
10 M. C. van Engelen, H. T. Teunissen, J. G. De Vries and C. J. Elsevier,
J. Mol. Catal. A: Chem., 2003, 206, 185–192.
2 N. W. Cant, D. L. Trimm and T. Turek, Catal. Rev. Sci. Eng., 1994, 36,
645–683.
3 J. Lacson, Mono-, Di- and Tri-ethylene Glycols, CEH Product Review,
2003.
11 H. T. Teunissen and C. J. Elsevier, Chem. Commun., 1997, 667–668.
12 H. T. Teunissen and C. J. Elsevier, Chem. Commun., 1998, 1367–1368.
13 M. Kilner, D. V. Tyers, S. P. Crabtree and M. A. Wood, Davy Process
Technology, WO 03/093208 A1, 2003.
4 R. A. Grey, G. P. Pez and A. Wallo, J. Am. Chem. Soc., 1981, 103,
7536–7542; R. A. Grey, G. P. Pez, J. Corsi and A. Wallo, J. Chem. Soc.,
Chem. Commun., 1980, 783–784; R. A. Grey and G. P. Pez, Allied
Chemicals, US 4232170, 1980; R. A. Grey and G. P. Pez, Fundam. Res.
Homogeneous Catal., 1984, 4, 97–116; R. A. Grey and G. P. Pez, Prepr.
- Am. Chem. Soc., Div. Pet. Chem., 1980, 25, 399–403.
5 U. Matteoli, G. Menchi, M. Bianchi, F. Piacenti and P. Frediani, J. Mol.
Catal., 1984, 22, 353–362; U. Matteoli, G. Menchi, M. Bianchi,
P. Frediani and F. Piacenti, J. Mol. Catal., 1985, 29, 269–270;
U. Matteoli, G. Menchi, M. Bianchi and F. Piacenti, J. Mol. Catal.,
1988, 44, 347–355; U. Matteoli, G. Menchi, M. Bianchi, F. Piacenti,
S. Ianelli and M. Nardelli, J. Organomet. Chem., 1995, 498, 177–186.
6 U. Matteoli, G. Menchi, M. Bianchi and F. Piacenti, J. Organomet.
Chem., 1986, 299, 233–238.
14 J. C. Bayo´n, C. Claver and A. M. Masdeu-Bulto´, Coord. Chem. Rev.,
1999, 193–195, 73–145.
15 S. J. A. Pope, N. R. Champness and G. Reid, J. Chem. Soc., Dalton
Trans., 1997, 1639–1644.
16 M. Konrad, F. Meyer, K. Heinze and L. Zsolnai, J. Chem. Soc., Dalton
Trans., 1988, 199–205.
17 R. Ali, S. J. Higgins and W. Levason, Inorg. Chim. Acta, 1984, 84,
65–69.
18 A. J. Barton, J. Connolly, W. Levason, A. Mendia-Jalon, S. D. Orchard
and G. Reid, Polyhedron, 2000, 19, 1373–1379; W. Levason, B. Patel
and G. Reid, Inorg. Chim. Acta, 2004, 357, 2115–2120; N. J. Hill,
W. Levason and G. Reid, Inorg. Chem., 2002, 41, 2070–2076;
P. J. Schebler, C. G. Riordan, I. A. Guzei and A. L. Rheingold,
Inorg. Chem., 1998, 37, 4754–4755.
19 W. Levason, S. D. Orchard and G. Reid, J. Chem. Soc., Dalton Trans.,
2000, 4550–4554.
20 J. Zhang, G. Leitus, Y. Ben-David and D. Milstein, Angew. Chem., Int.
Ed., 2006, 45, 1113–1115.
7 U. Matteoli, G. Menchi, M. Bianchi and F. Piacenti, J. Mol. Catal.,
1991, 64, 257–267.
8 K. Nomura, H. Ogura and Y. Imanishi, J. Mol. Catal. A: Chem., 2001,
166, 345–349; K. Nomura, H. Ogura and Y. Imanishi, J. Mol. Catal. A:
Chem., 2002, 178, 105–114.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 2289–2291 | 2291