Transition metal-tungsten bimetallic catalysts for the conversion of cellulose into ethylene glycol

Article abstract of DOI:10.1002/cssc.200900197

Tungsten-based bimetallic catalysts (W M(8,9,10); where M(8,9,10) is Ni, Pd, Pt, Ir, Ru, or Rh) are found to be highly active and selective for the formation of ethylene glycol from cellulose. The cooperation between C-C cracking reactions over metallic tungsten and the hydrogenation of unsaturated intermediates over the transition metals M(8,9,10) results in a particularly high selectivity towards ethylene glycol, up to 75%.

Full text of DOI:10.1002/cssc.200900197

DOI: 10.1002/cssc.200900197
Transition Metal–Tungsten Bimetallic Catalysts for the Conversion of
Cellulose into Ethylene Glycol
Ming-Yuan Zheng, Ai-Qin Wang, Na Ji, Ji-Feng Pang, Xiao-Dong Wang, and Tao Zhang*[a]
Currently, the human race faces two important issues: energy
and the environment. Obtaining energy and chemicals in a
sustainable and environmentally friendly manner has been the
target of many chemists. Among many potential approaches,
biomass conversion has been considered as one of the most
important ones because of its renewable and carbon-neutral
properties. Cellulose is the most abundant source of biomass
and generally accounts for 30–60 wt% of dried plants. Much
effort has been devoted to the development of a green and ef-
fective process for cellulose conversion, such as fermentation
with enzymes to produce ethanol,[1] thermo-pyrolysis to bio-
oils[2] and syn-gas,[3] and hydrolysis with dilute acids in ionic liq-
uids to yield oligomers,[4] glucose,[5] and HMF.[6]
conversion of cellulose into EG? With the Ni-modified W2C cat-
alysts, a remarkable synergistic effect was observed. Which
effect caused this synergy in the conversion process? Herein,
based on the element W, we develop a series of bimetallic cat-
alysts for the cellulose conversion, including Ni–W, Pd–W, Pt–
W, Ru–W, and Ir–W, supported on different carriers. A remark-
able synergy is observed between tungsten and metals of
groups 8, 9, and 10 [M(8,9,10)]. Moreover, it is found that the
product selectivity can be tuned effectively by changing the
weight ratio of W to M(8,9,10). The maximum yield of EG,
75.4%, was obtained on a Ni–W/SBA-15 catalyst.
The reaction results for the conversion of cellulose into poly-
ols over monometallic and tungsten-based bimetallic catalysts
are summarized in Table 1. Interestingly, cellulose was com-
pletely depolymerized over the W/AC catalyst after 30 min.
However, the EG yield was negligibly low (2.2%). This is differ-
ent from the W2C/AC catalyst we reported previously, which
produced EG at a yield of 27%.[10] Other polyols, such as sorbi-
tol, mannitol, erythritol, and 1,2-propylene glycol, were not de-
tectable in the liquid product either. The total organic carbon
analysis (TOC; see Supporting Information, Table S1) showed
that 74.0% of the total carbon put into the reactor remained
in the liquid product, demonstrating that the cellulose was
predominantly converted into many unknown liquid products.
Gas analysis (Supporting Information, Table S2) showed that
CO2 (0.23%v/v) and CO (0.02%v/v) were formed during the re-
action, which accounted for 3.8 wt% of the total carbon in the
feedstock. The sum of carbon in gas and liquid was 77.8%.
The discrepancy to the carbon balance of 22.2% should be at-
tributed to CO2 formation during the reaction, which was read-
ily dissolved in the aqueous solution at high pressures but
then released at ambient pressure, so that it could not be
quantified accurately by TOC analysis. The freshly produced
liquid was light yellow and then turned dark brown after expo-
sure to air for several days. Fehling’s tests revealed that a large
amount of unsaturated compounds was contained in the
liquid. This is consistent with the CO chemisorption result
(Table 1) that the W/AC had a negligible CO uptake and might
be inactive for hydrogenation of the unsaturated intermediates
derived from cellulose degradation.
An alternative green process for cellulose conversion is to
use solid catalysts to convert cellulose into polyols in an aque-
ous solution. Yan et al.[7] found that cellubiose, the simplest
model molecule for cellulose, can be converted into sorbitol
by using a Ru nanocatalyst in an aqueous solution. Fukuoka
and Dhepe,[8] and Luo et al.,[9] disclosed that cellulose can be
converted into hexitols under hydrothermal and hydrogen
conditions over noble-metal catalysts. These studies opened
up a green route for the conversion of cellulose. On the other
hand, the yields of polyols in these processes were not very
high; generally less than 40%. Furthermore, expensive noble-
metal catalysts were used. Therefore, the development of less-
expensive but efficient catalysts for cellulose conversion is
highly desirable.
Recently, we successfully achieved the direct conversion of
cellulose into ethylene glycol (EG) by a heterogeneous catalytic
reaction under hydrogen atmosphere and hydrothermal condi-
tions.[10] Over an active-carbon-supported tungsten carbide cat-
alyst (W2C/AC), cellulose was completely depolymerized and
EG was produced with a yield of 27%. More importantly, upon
the addition of a small amount of Ni to the W2C catalyst, the
EG yield was significantly increased to as high as 61%. EG is an
important bulk chemical and sourced dominantly from petrole-
um resources.[11] Thus, this work provided a novel route for EG
production as well as for cellulose conversion.
However, there are some very intriguing questions about
this novel process yet to be answered. For example, besides
the W2C catalysts, are there any other catalysts effective for the
Different from the W/AC catalyst, other monometallic
M(8,9,10) catalysts such as Ni/AC, Pd/AC, Pt/AC, Ru/AC, and Ir/
AC could not degrade cellulose completely. The cellulose con-
versions over these monometallic catalysts were in the range
of merely 60–70%. On the other hand, various polyols includ-
ing EG, sorbitol, mannitol, erythritol, and 1,2-propylene glycol
were formed. EG was produced at yields of ca. 10% over the
Pt/AC, Ni/AC, and Ir/AC catalysts. Meanwhile, a notable yield of
hexitols (20.5%) was obtained on the Pt/AC catalyst, which is
consistent with the results reported in Ref. [8]. However, no
[a] Dr. M.-Y. Zheng, Prof. A.-Q. Wang, N. Ji, J.-F. Pang, Prof. X.-D. Wang,
Prof. T. Zhang
State Key Laboratory of Catalysis
Dalian Institute of Chemical Physics, Chinese Academy of Sciences
P.O. Box 110, Dalian 116023 (PR China)
Fax: (+86)411-84691570
E-mail: taozhang@dicp.ac.cn
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.200900197.
ChemSusChem 2010, 3, 63 – 66
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
63

over the Ru–W/AC catalyst,
whereas only a small amount of
EG was obtained on either the
W/AC or Ru/AC catalysts.
Table 1. Results of cellulose conversion and yields of main products.[a]
Catalyst
CO uptake [mmolgÀ1
]
Yield [%]
Ery
Conv.[%]
EG
2.2
9.2
7.6
11.6
2.2
9.8
59.6
56.9
61.7
50.6
45.3
36.8
75.4
46.6
1,2-PG
M
S
To make sure that this syner-
gistic effect did not originate
from the influence of a particular
supports, the Ni–W catalysts
were prepared and tested with
different carriers. Although the
Ni–W/SiC and Ni–W/TiO2 cata-
lysts had much lower surface
areas (13 m2 gÀ1 and 9 m2 gÀ1, re-
spectively) than the Ni–W/SBA-
15 catalyst (417 m2 g), the EG
yields produced with them were
45.3% and 36.8%, respectively,
both of which are significantly
higher than those obtained with
monometallic Ni or W catalysts.
W/AC
Ni/AC
Pd/AC
Pt/AC
Ru/AC
Ir/AC
Pd–W/AC
Pt–W/AC
Ru–W/AC
Ir–W/AC
Ni–W/SiC
Ni–W/TiO2
Ni5–W25/SBA-15
Ni/AC+W/AC
1.1
386.7
167.8
60.7
333.5
331.1
5.9
19.1
490
547
–
0
0
1.0
4.2
0
2.0
4.9
5.8
7.9
2.5
4.8
1.9
4.0
2.7
0
4.3
0
2.9
0
7.2
3.8
3.3
3.2
2.5
1.2
1.5
4.1
8.0
0
1.5
0
5.8
0
2.2
2.9
0
3.1
0
1.0
0.9
1.3
5.9
0
100
71.9
67.2
63.5
74.1
63.8
100
97.0
100
100
92.0
100
100
100
10.1
1.2
14.7
0
5.9
7.8
3.6
4.1
0
2.3
2.8
3.1
2.5
–
7.9
–
[a] Ery: erythritol; EG: ethylene glycol; 1,2-PG: 1,2 propylene glycol; M: mannitol; S: sorbitol.
particularly high selectivity towards EG formation was ob-
served with the Pt/AC catalyst. Surprisingly, nearly no polyols
were detected in the products with the Ru/AC catalyst, except
very little EG (2.2% yield). A gas product analysis (Supporting
Information, Table S2) disclosed that significant amounts of
CH4 (3.52%v/v), C2H6 (0.16%v/v), and C3H8 (0.04% v/v) were
formed, which accounted for 58.4 wt% of the total carbon in
the feedstock (Supporting Information, Table S1). Thus, the
main products of the cellulose conversion over the Ru/AC cata-
lyst were gases, in contrast to much less gases being produced
over other metal catalysts. Fehling’s tests for products of the
monometallic M(8,9,10) catalysts showed that the colorless liq-
uids contained no detectable levels of aldehydes or reducing
sugars, suggesting that the unsaturated intermediates pro-
duced in the reactions were hydrogenated completely over
these transition-metal catalysts.
For the bimetallic catalysts, ligand or strain effects are often
considered to be important for enhanced catalytic perfor-
mance.[12–15] An examination of the Ni–W/SBA-15 catalyst by X-
ray diffraction (XRD) showed the formation of a Ni4W alloy
phase, besides the metallic tungsten and nickel phases. This is
slightly different from other bimetallic catalysts, for which only
metallic phases of tungsten and the noble metal were identi-
fied (Supporting Information, Figure S1). To determine whether
this new alloy phase was responsible for the excellent perfor-
mance of the Ni–W bimetallic catalyst, we mixed the Ni/AC
and the W/AC catalysts mechanically and then tested for cellu-
lose conversion. After 30 min, the cellulose was converted
completely. More importantly, a notable amount of EG was
produced over the mixed catalysts with a yield of 43.0%, in
contrast to the EG yields of less than 10% over the individual
Ni/AC and W/AC catalysts. This result strongly suggests that
the synergistic effect between Ni and W does not necessarily
require intimate contact between the two metallic compo-
nents. Probably, many reactions are involved in the process of
converting cellulose into EG, sequentially taking place on dif-
ferent types of catalytic sites and proceeding independently.
To further probe the respective roles of M(8,9,10) and W in
the EG production, different weight ratios of Ni and W were
employed for the Ni–W/SBA-15 catalysts and investigated for
the catalytic conversion of cellulose. As shown in Table 2, the
cellulose conversions over all four Ni–W/SBA-15 catalysts were
100%. However, the product distribution, particularly the EG
yield, varied noticeably with the Ni/W weight ratio. For the
Ni5–W25/SBA-15 and Ni5–W15/SBA-15 catalysts, the EG yields
were as high as ca. 75%, and the yields of other polyols were
all less than 10%, demonstrating the unique high selectivity of
the two Ni–W catalysts towards the EG formation. Further in-
creasing the Ni/W weight ratio to 2:3 and 1:1 led to the EG
yield dropping remarkably, down to 51.0% and 31.8%, respec-
tively. Meanwhile, the yields of other polyols, especially erythri-
tol (C4 sugar alcohol) and hexitols (mannitol and sorbitol) in-
creased sharply. For example, over the Ni15–W15/SBA-15 cata-
Because the W/AC catalyst was very effective for the depoly-
merization of cellulose while the monometallic M(8,9,10) cata-
lysts were efficient for hydrogenation and even moderately se-
lective for polyol production, we designed bimetallic catalysts
comprising tungsten and another transition metal, aiming to
transform the cellulose into EG or other desired polyols. As
shown in Table 1, the cellulose conversions over these
M(8,9,10)–W bimetallic catalysts were significantly higher than
those over the monometallic M(8,9,10) catalysts, reaching
100% except for the Pt–W/AC and Ni–W/SiC catalysts. More in-
triguingly, different from the results of the monometallic cata-
lysts, significantly high EG yields were obtained with these bi-
metallic catalysts. The EG yields were most often in the range
of 50–70%. Particularly on Ni–W/SBA-15, the EG yield was as
high as 75.4%; an increase by more than 14% in comparison
to the Ni–W2C/AC catalyst we reported previously.[10] Evidently,
besides W2C catalysts, bimetallic catalysts comprising tungsten
and transition metals are also very efficient for the production
of EG from cellulose. A synergistic effect apparently exists be-
tween tungsten and the other transition metal in this reaction.
For example, EG was produced with a yield as high as 61.7%
64
www.chemsuschem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2010, 3, 63 – 66

carbide was improved substan-
tially, thus leading to a large en-
hancement of the EG yield.
Table 2. Cellulose conversion over Ni–W/SBA-15 catalysts with different Ni/W weight ratios and metal loadings.
Catalyst[a]
Ni/W weight ratio CO uptake [mmolgÀ1
]
Yield [%]
Ery EG
Conv. [%]
1,2-PG
4.0 75.4 4.1
M
S
In summary, we have devel-
oped a series of M(8,9,10)–W bi-
metallic catalysts which are
highly active and selective for
the formation of EG from cellu-
lose. In particular for the Ni5–
W25/SBA-15 and Ni5–W15/SBA-
15 catalysts, the EG yield reaches
Ni5–W25/SBA-15
Ni5–W15/SBA-15
Ni10–W15/SBA-15 2:3
Ni15–W15/SBA-15 1:1
1:5
1:3
7.9
15.6
1.3
2.2
3.1 100
5.8 100
6.2 20.8 100
10.1 30.3 100
6.0 76.1 3.2
18.5 51.0 1.8
17.2 36.8 2.0
151.3
438.4
[a] The number besides the metal represents the metal loading of the catalyst. For example, in Ni5–W25/SBA-
15 the loadings of Ni and W are 5 wt% and 25 wt%, respectively.
lyst (Ni/W=1:1), the hexitols yield reached 40.4% and the er-
ythritol yield was as high as 17.2%.
values as high as ~75%, which is the highest value reported
for the cellulose transformation so far. In this type of
M(8,9,10)–W bimetallic catalysts, W is found to be the key com-
ponent for degradation of cellulose, i.e., the CÀC cracking reac-
tions; while the M(8,9,10) transition metals are mainly responsi-
ble for the hydrogenation reactions of unsaturated intermedi-
ates. Because the cellulose transformation under our condi-
tions involves many reactions, including hydrolysis, hydrogena-
Nickel catalysts are well-known for their hydrogenation reac-
tivity,[16] and the number of catalytically active sites for hydro-
genation can be determined by CO chemisorption.[10,17] As
measured by CO chemosorption on Ni–W/SBA-15 (Table 2),
there is a more than 50-fold increase in CO uptake with an in-
crease of the Ni/W weight ratio from 1:5 to 1:1, implying that
more active hydrogenation sites are present on the Ni–W/SBA-
15 catalysts at higher Ni/W ratios. Correlated to the results that
more hexitols and C4 sugar alcohol were produced at the ex-
pense of the EG yield upon increasing the Ni/W ratio, one can
speculate that there are several reaction routes that take place
and compete with the EG formation reaction. We suppose that
the reactions include at least three types, that is, hydrolysis,
cracking, and hydrogenation. The cracking reaction happens
most likely over the W sites, as suggested by the complete
conversion of cellulose and the formation of a large amount of
unsaturated compounds over W/AC. Meanwhile, the hydroge-
nation reaction probably takes place on the Ni and noble-
metal active sites due to their high reactivities for hydrogena-
tion reactions. In the case of bimetallic catalysts, glucose de-
rived from cellulose hydrolysis is cracked over the W active
sites and subsequently hydrogenated over the metallic
M(8,9,10) active sites to form EG and other polyols. A too large
number of hydrogenating sites will promote the hydrogena-
tion reactions and make them predominate over the cracking
reactions, leading to increases in the yields of hexitols and er-
ythritol but at the expense of EG formation. The hexitols (man-
nitol and sorbitol) were found to be stable and did not form
EG when they were used as feedstock. On the contrary, very
few or even no hydrogenating sites on the catalysts will result
in small amounts of polyols being formed, just like the case of
the W/AC catalyst. Thus, there is an optimal regime for the Ni/
W ratio, in which the W active sites and Ni sites cooperate
properly to obtain a high EG yield.
tion, and CÀC cracking,
a
good balance between
hydrogenation and CÀC cracking will determine the final prod-
uct distribution. Therefore, by changing the weight ratio of
M(8,9,10) to W, one can effectively tune the competition be-
tween hydrogenation and CÀC cracking capabilities, and ac-
cordingly the selectivity towards ethylene glycol. This rule will
guide the design of more-efficient catalysts for the catalytic
conversion of cellulose into a desired polyol.
Experimental Section
The M(8,9,10)-W bimetallic catalysts including Ru–W/AC, Ir–W/AC,
Pd–W/AC, Pt–W/AC, Ni–W/SiC, and Ni–W/TiO2 were prepared by in-
cipient wetness impregnation, dried at 393 K overnight, followed
by reduction in a hydrogen flow at 973 K for 1 h. In detail, for the
Ir–W and Pt–W catalysts the AC support was first impregnated
with an aqueous solution of [H2IrCl4] or [H2PtCl4] and dried at 393 K
overnight, and then impregnated with an aqueous solution of am-
monia metatungstate (AMT). The Ru–W, Pd–W, and Ni–W catalysts
were prepared by co-impregnation with an aqueous solution of
AMT and RuCl3, PdCl3, or Ni(NO3)2·6H2O. For all of the above cata-
lysts, 25 wt% of tungsten and 5 wt% of the second metal were
employed in the preparation.
The Ni–W/SBA–15 catalysts with different Ni/W weight ratios were
also prepared by the co-impregnation method described above.
The SBA-15 support was synthesized by the method reported in
Ref. [24]. Based on the nickel and tungsten weight loadings, the
catalysts are denoted as Ni5–W25/SBA-15, Ni5–W15/SBA-15, Ni10–
W15/SBA-15, and Ni15–W15/SBA-15, respectively.
The monometallic catalysts were prepared by impregnation with
the corresponding metal precursors mentioned above, dried at
393 K overnight, and reduced in hydrogen flow at different tem-
peratures, that is, 973 K for W/AC, 723 K for Ni/AC, 673 K for the
Ru/AC, Ir/AC, Pd/AC, and Pt/AC catalysts. The metal loadings were
fixed at 5 wt% for the noble metal catalysts, 30 wt% for the W/AC
catalyst, and 20 wt% for the Ni/AC catalyst.
The CO chemisorption experiments were performed with a Micro-
meritics Autochem 2910 chemisorber. Prior to CO adsorption, cata-
lysts were reduced at 673 K for the M(8,9,10)/AC catalysts, 773 K
for the M(8,9,10)–W catalysts, and 973 K for the W/AC catalyst for
30 min.
For the tungsten carbide catalysts we reported previously,[10]
it has been known that transition metal carbides have good
hydrogenation properties, similar to the noble metals.[18–23] This
means that the tungsten carbide itself comprises multifunc-
tional catalytic sites, including hydrogenating and cracking
sites, for the cellulose conversion and as a result, exhibits
unique selectivity for EG production. Nevertheless, the ratio of
the two types of active sites in the tungsten carbide may not
be at an optimum value. Upon promotion with a small
amount of nickel, the hydrogenation ability of the tungsten
ChemSusChem 2010, 3, 63 – 66
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemsuschem.org
65

[8] A. Fukuoka, P. L. Dhepe, Angew. Chem. 2006, 118, 5285–5287; Angew.
Chem. Int. Ed. 2006, 45, 5161–5163.
[9] C. Luo, S. A. Wang, H. C. Liu, Angew. Chem. 2007, 119, 7780–7783;
Angew. Chem. Int. Ed. 2007, 46, 7636–7639.
[10] N. Ji, T. Zhang, M. Y. Zheng, A. Q. Wang, H. Wang, X. D. Wang, J. G.
Chen, Angew. Chem. 2008, 120, 8638–8641; Angew. Chem. Int. Ed. 2008,
47, 8510–8513.
[11] ICIS Chemical Business Weekly 2008, 352.
[12] J. R. Kitchin, J. K. Norskov, M. A. Barteau, J. G. Chen, J. Chem. Phys. 2004,
120, 10240–10246.
[13] J. K. Norskov, P. Liu, Phys. Chem. Chem. Phys. 2001, 3, 3814–3818.
[14] J. G. Chen, C. A. Menning, M. B. Zellner, Surf. Sci. Rep. 2008, 63, 201–
254.
Catalytic conversion of cellulose was carried out in a stainless-steel
autoclave (Parr Instrument Company, 300 mL) typically at 6 MPa H2
pressure (measured at room temperature) and 518 K for 30 min.
For each reaction, 1.0 g cellulose (Merck, microcrystalline), 0.3 g
catalyst, and 100 mL water were put into the reactor and stirred at
a rate of 1000 rpm. After the reaction, the liquid-phase products
were analyzed by HPLC. Cellulose conversions were determined by
the change of cellulose weight before and after the reaction. The
yield of polyols was calculated as: yield (%)=(weight of polyol in
the products)/(weight of cellulose put into the reactor)ꢁ100%.
[15] M. S. Y. Gauthier, S. Padovani, E. Lundgren, V. Bus, G. Kresse, J. Redinger,
P. Varga, Phys. Rev. Lett. 2001, 87, 036103.
Acknowledgements
[16] C. T. Campbell, D. W. Goodman, Surf. Sci. 1982, 123, 413–426.
[17] Y. Y. Shu, S. T. Oyama, Carbon 2005, 43, 1517–1532.
[18] S. T. Oyama, The Chemistry of Transition Metal Carbides and Nitrides,
Blackie, Glasgow 1996.
This work was supported by grants from the 973 Program of
China (2009CB226102) and the National Science Foundation of
China (20903089).
[19] J. G. Chen, Chem. Rev. 1996, 96, 1477–1498.
[20] C. H. Liang, L. Ding, A. Q. Wang, Z. Q. Ma, J. S. Qiu, T. Zhang, Ind. Eng.
Chem. Res. 2009, 48, 3244–3248.
[21] H. Wang, A. Q. Wang, X. D. Wang, T. Zhang, Chem. Commun. 2008,
2565–2576.
Keywords: biomass · ethylene glycol · supported catalysts ·
transition metals · tungsten
[22] J. Sun, M. Y. Zheng, X. D. Wang, A. Q. Wang, R. Cheng, T. Li, T. Zhang,
Catal. Lett. 2008, 123, 150–155.
[23] X. W. Chen, T. Zhang, P. L. Ying, M. Y. Zheng, W. C. Wu, L. G. Xia, T. Li,
X. D. Wang, C. Li, Chem. Commun. 2002, 288–289.
[24] D. Y. Zhao, J. L. Feng, Q. S. Huo, N. Melosh, G. H. Fredrickson, B. F.
Chmelka, G. D. Stucky, Science 1998, 279, 548–552.
[1] H. Fukuda, A. Kondo, S. Tamalampudi, Biochem. Eng. J. 2009, 44, 2–12.
[2] H. B. Goyal, D. Seal, R. C. Saxena, Renewable Sustainable Energy Rev.
2008, 12, 504–517.
[3] M. Asadullah, T. Miyazawa, S. I. Ito, K. Kunimori, M. Yamada, K. Tomish-
ige, Appl. Catal. A: Gen. 2004, 267, 95–102.
[4] R. Rinaldi, R. Palkovits, F. Schꢂth, Angew. Chem. 2008, 120, 8167–8170;
Angew. Chem. Int. Ed. 2008, 47, 8047–8050.
[5] C. Z. Li, Q. Wang, Z. K. Zhao, Green Chem. 2008, 10, 177–182.
[6] J. B. Binder, R. T. Raines, J. Am. Chem. Soc. 2009, 131, 1979–1985.
[7] N. Yan, C. Zhao, C. Luo, P. J. Dyson, H. C. Liu, Y. Kou, J. Am. Chem. Soc.
2006, 128, 8714–8715.
Received: August 23, 2009
Revised: October 27, 2009
Published online on December 8, 2009
66
www.chemsuschem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2010, 3, 63 – 66