A continuous process for glyoxal valorisation using tailored Lewis-acid zeolite catalysts

Article abstract of DOI:10.1039/c3gc42353k

The aqueous-phase heterogeneously catalysed isomerisation of bio-oil derived glyoxal is herein introduced as a novel route for the sustainable production of glycolic acid. While commercial ultra-stable Y zeolites displayed only moderate performance, their evaluation enabled us to highlight the crucial role of Lewis acidity in the reaction. Gallium incorporation into these zeolites boosted the glycolic acid yield, although the best catalytic results were obtained over tin-containing MFI-type zeolites, reaching 91% yield of the desired product at full conversion. These materials comprised hydrothermally-synthesised Sn-MFI as well as a novel catalyst obtained by the introduction of tin into silicalite-1 by means of a simpler and more scalable method, i.e. alkaline-assisted metallation. In-depth spectroscopic characterisation of these systems uncovered a substantial similarity of the tin centres obtained by the top-down and bottom-up synthetic approaches. NMR spectroscopic studies gave evidence that the reaction follows a 1,2-hydride shift mechanism solely catalysed by Lewis-acid sites. The Sn-MFI analogue could be reused in 5 cycles without the need for intermediate calcination, did not evidence any tin leaching, and demonstrated suitability for utilisation under continuous-flow operation. The tin-based zeolites exhibited remarkable performance also in alcoholic solvents, leading to the one-pot production of relevant alkyl glycolates.

Full text of DOI:10.1039/c3gc42353k

Volume 16 Number 3 March 2014 Pages 927–1650
Green
Chemistry
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
ISSN 1463-9262
PAPER
Cecilia Mondelli, Javier Pérez-Ramírez et al.
A continuous process for glyoxal valorisation using tailored Lewis-acid
zeolite catalysts

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
A continuous process for glyoxal valorisation using
tailored Lewis-acid zeolite catalysts†
Cite this: Green Chem., 2014, 16,
1176
Pierre Y. Dapsens, Cecilia Mondelli,* Bright T. Kusema, René Verel and
Javier Pérez-Ramírez*
The aqueous-phase heterogeneously catalysed isomerisation of bio-oil derived glyoxal is herein intro-
duced as a novel route for the sustainable production of glycolic acid. While commercial ultra-stable Y
zeolites displayed only moderate performance, their evaluation enabled us to highlight the crucial role of
Lewis acidity in the reaction. Gallium incorporation into these zeolites boosted the glycolic acid yield,
although the best catalytic results were obtained over tin-containing MFI-type zeolites, reaching 91% yield
of the desired product at full conversion. These materials comprised hydrothermally-synthesised Sn-MFI
as well as a novel catalyst obtained by the introduction of tin into silicalite-1 by means of a simpler and
more scalable method, i.e. alkaline-assisted metallation. In-depth spectroscopic characterisation of these
systems uncovered a substantial similarity of the tin centres obtained by the top-down and bottom-up
synthetic approaches. NMR spectroscopic studies gave evidence that the reaction follows a 1,2-hydride
shift mechanism solely catalysed by Lewis-acid sites. The Sn-MFI analogue could be reused in 5 cycles
without the need for intermediate calcination, did not evidence any tin leaching, and demonstrated suit-
ability for utilisation under continuous-flow operation. The tin-based zeolites exhibited remarkable per-
formance also in alcoholic solvents, leading to the one-pot production of relevant alkyl glycolates.
Received 16th November 2013,
Accepted 13th December 2013
DOI: 10.1039/c3gc42353k
www.rsc.org/greenchem
the preparation of bio-based LA.4 In contrast, the production
of GA from bio-derived feedstocks has been investigated to a
1 Introduction
Nowadays, chemical companies are increasingly investing in limited extent. Still, its demand reached 40 ktons in 2011 and
the manufacture of polymers derived from renewable resources is expected to double by 2018, considering an annual growth
(so-called biopolymers) in view of reducing their carbon foot- rate estimated at 10.1%.3
print as well as expanding their product portfolio.1 In this
GA is currently manufactured by the carbonylation of form-
context, α-hydroxy acids (AHA) such as lactic acid (LA, C3 mole- aldehyde in the presence of H2SO4. Disadvantageously, this
cule) and glycolic acid (GA, C2 molecule) comprise monomers process requires elevated temperatures and pressures (473 K,
of primary relevance. Besides the plastics obtained from the 300 bar) to attain appreciable yields of GA. Nonetheless, sig-
individual AHAs, that is, polylactide (PLA) and polyglycolide nificant amounts of formic acid by-products and unreacted
(PGA), a mixed polymer, poly(lactic-co-glycolic) acid (PLGA), formaldehyde are still contained in the final mixture.5 Other
can be derived. PLA constitutes a novel functional material processes for GA production have been reported such as the
mainly used in the packaging industry, while PGA and PLGA saponification of chloroacetic acid, the enzymatic conversion
are extensively employed in the medical field owing to their of glycolonitrile, and the electrolytic reduction of oxalic acid.6
biocompatibility and exceptional biodegradability.2 GA is also These routes are impractical at an industrial scale due to
widely employed in pharmaceuticals and cosmetics as skin energy intensiveness, environmental issues, long reaction
care agent.3 Significant research efforts have been devoted to times, and limited selectivity. Thus, based on the present scen-
ario, the availability of an efficient, affordable, and scalable
chemocatalytic route to transform biomass into glycolic acid
under mild conditions would be highly valuable. A few
Institute for Chemical and Bioengineering, Department of Chemistry and Applied
attempts have been made in this direction. Zhang et al. have
recently investigated the direct conversion of cellulose into GA
over Mo-containing heteropolyacids at 453 K and 6 bar O2,
reaching a maximum yield of 49% at 91% conversion.7 Besides
for the sub-optimal selectivity of this catalytic system, the use
of heteropolyacids remains problematic separation-wise. An
Biosciences, ETH Zurich, Vladimir-Prelog-Weg 1, CH-8093 Zurich, Switzerland.
E-mail: jpr@chem.ethz.ch, cecilia.mondelli@chem.ethz.ch; Fax: +41 44 6331405;
Tel: +41 44 6337120
†Electronic supplementary information (ESI) available: XRD patterns and N2 iso-
therms of Sn-MFI, Ti-MFI, CBV720, and CVB720-ATGa, UV-Vis spectrum of Ti-
MFI, and characterisation data of the USY zeolites used in this study. See DOI:
10.1039/c3gc42353k
1176 | Green Chem., 2014, 16, 1176–1186
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
alternative route to GA encompasses oxidation of ethylene with the reaction scheme for the aldoketose isomerisation of
glycol, which can be produced from xylan hemicellulose by sugars and with the conversion of pyruvaldehyde into LA.17
hydrolysis to xylose, hydrogenation of the sugar to xylitol, and Tin-containing porous materials (e.g. Sn-BEA, Sn-MWW, Sn-
hydrogenolysis of the polyol.8 The drawbacks of this approach MCM-41, and Sn-CSM) exhibit excellent performances in the
are the multiplicity of steps, the requirement of expensive latter reactions, greatly surpassing commercial aluminosilicate
metals such as Ru and Pt in their majority, and the low selecti- zeolites. Their remarkable activity and selectivity are mainly
vity (ca. 33% at any conversion level) to ethylene glycol in the associated with the sole presence of strong and well-defined
hydrogenolysis step. Another possible pathway is represented Lewis-acid centres enabling high conversion rates with
by the oxidation of glycerol. Still, at present this chemocataly- minimal by-product formation.17,18 Alternatively, post-syn-
tic process yields a maximum of 17% GA over Nb2O5 and 40% thetic modification of commercial zeolites has been uncovered
GA and 60% glycolate over Au-based catalysts.9
as a suitable tool to create tin-free zeolites with tailored pro-
Herein, we put forward a novel synthetic route for GA, perties. In particular, the alumination of MFI-type zeolites and
which solely comprises the heterogeneously catalysed isomeri- galliation of USY zeolites in alkaline media have enabled the
sation of glyoxal (GLY) (Scheme 1) in water. In addition to the introduction of unique tetra-coordinated Lewis-acid Al and Ga
low temperature (363 K) and pressure (ambient) required, our sites into the zeolitic frameworks, which have been demon-
interest in this reaction, demonstrated years back at the labora- strated to be highly selective for the production of LA and its
tory level with homogeneous bases or acids,10 derived from the ester derivatives.19
fact that GLY represents an undesired oxygenated molecule
Based on these premises, the aqueous-phase isomerisation
contained in bio-oil.11 The occurrence of this compound is of glyoxal to glycolic acid was initially studied over commer-
shown to be independent of the nature of the lignocellulosic cially available USY zeolites to distinguish the role of the type
feedstock that undergoes the pyrolytic treatment and corre- of acidity. Then, state-of-the-art Lewis-acid catalysts were evalu-
sponds to ca. 2–3 wt% of the obtained liquid mixture.12 As ated, which indicated the superiority of hydrothermally-
exemplified by the case of Finland, where installations with an synthesised Sn-MFI to gallium-based systems. Following this
annual production capacity of pyrolysis oil of 50 ktons per year result, a novel catalyst was designed by the incorporation of tin
are being erected, a significant amount of glyoxal is expected into silicalite-1 through a simpler and more scalable method,
to be available in the coming years.13 Furthermore, GLY recov- i.e. alkaline treatment in the presence of a soluble tin salt. The
ery from this renewable source can be easily accomplished by tin species in the solid were characterised in depth by means
water extraction followed by chromatographic purification.14 of spectroscopic methods in terms of acidity and structure to
Thus, the isomerisation of GLY to GA appears to be an attrac- explain why similar catalytic results were obtained over the two
tive route to valorise a waste product along the biofuel manu- differently prepared catalysts. NMR spectroscopy was applied to
facture chain. This approach follows the biorefinery concept, provide experimental support to the postulated 1,2-hydride shift
which is one way to render biofuel production more economi- route for GLY isomerisation. Industrially-relevant aspects of the
cal and sustainable in the long term, i.e. through the extra reaction over the modified silicalite-1 catalyst were evaluated,
profit and higher efficiency provided by the co-generation of such as catalyst recyclability, tin leaching, the use of higher feed
chemicals. Last but not least, since GA does not possess chiral concentrations, and the possibility of a continuous-flow oper-
centres, there is no obvious drawback in using a solid catalyst. ation. Finally, it was explored whether the one-pot production of
This contrasts with the production of LA, where it has so far prominent alkyl glycolates, applied as plasticisers and chemical
not been possible to obtain the desired L-isomer selectively intermediates,20 would be possible over the tin-containing zeolites
over a heterogeneous system.15
when using an alcohol as a reaction medium instead of water.
Recent DFT studies by Oshima et al. have indicated that a
mechanism based on a Lewis-acid catalysed 1,2-hydride shift
is energetically more favourable than a pathway relying on a
Brønsted-acid catalysed dehydration–tautomerisation for the
isomerisation of GLY to GA.16 This prediction is in analogy
2 Experimental
2.1 Catalyst synthesis
USY zeolites (CBV300, CBV500, CBV600, CBV712, CBV720,
CBV760, and CBV780) were purchased from Zeolyst Inter-
national, whereas high-silica USY (HSZ390 HUA) and silicalite-
1 (HSZ890 H0A) were supplied by Tosoh Corporation. Prior to
use, the as-received zeolites were converted into their protonic
form by calcination in static air at 823 K (5 K min−1) for 5 h.
The composition and textural properties of these samples are
reported in Table SI1.† Reference catalysts comprised Amber-
lyst-15 (Dow Chemical Company) and SnO2 (Sigma-Aldrich,
99.9%).
Ti-MFI was synthesised following the protocol reported by
Scheme 1 Strategy for the production of glycolic acid and alkyl glyco-
lates based on the Lewis-acid catalysed isomerisation of glyoxal.
Huang et al.21 Thus,
a
mixture containing tetraethyl
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1176–1186 | 1177

View Article Online
Paper
Green Chemistry
orthosilicate (TEOS, 24.5 cm3, Sigma-Aldrich, 98%), tetrapropyl- materials were digested in an acidic mixture (HCl–HNO3–HF)
ammonium hydroxide (TPAOH, 17.6 cm3, Alfa Aesar, 35 wt%), under heating (343 K, overnight). Nitrogen sorption at 77 K
and deionised H2O (28.3 cm3) was allowed to react under was performed using a Quantachrome Quadrasorb-SI analyser
stirring at room temperature for 1 h. Thereafter, the solution on degassed samples (10−1 mbar, 573 K, 3 h). Powder X-ray
was cooled down to 273 K in an ice bath and a solution of diffraction (XRD) was measured using a PANalytical X’Pert
titanium(IV) butoxide (TBOT, 1.2 cm3, TCI, 98%) in PRO-MPD diffractometer with Ni-filtered Cu Kα radiation (λ =
EtOH (12.5 cm3, Scharlau, 99.9%) was added drop-wise under 0.1541 nm). Data were recorded in the 5–70° 2θ range with an
vigorous stirring. After further stirring for 1 h at the same angular step size of 0.05° and a counting time of 7 s per step.
temperature, the mixture was heated to 333 K for 1 h to enable Transmission electron microscopy (TEM) was undertaken
the hydrolysis of TBOT. This resulted in the formation of a sol using a FEI Tecnai F30 microscope operated at 300 kV (field
with a molar composition of SiO2 : 0.033TiO2 : 0.2TPAOH : 2 emission gun). The samples were prepared by depositing a few
ethanol : 20H2O. The sol was transferred into a 100 cm3 Teflon- droplets of zeolite suspension in methanol onto a carbon-
lined autoclave, which was placed in an oven and heated at coated copper grid, followed by evaporation at room tempera-
443 K for 4 days under static conditions. The obtained solid ture. Fourier transform infrared (FTIR) spectroscopy measure-
was recovered by filtration, washed thoroughly with deionised ments of adsorbed pyridine/d3-acetonitrile were carried out
water, dried overnight at 338 K, and finally calcined in static using a Bruker IFS66 spectrometer equipped with a liquid N2-
air at 823 K (5 K min−1) for 5 h to ensure the complete removal cooled mercury cadmium telluride (MCT) detector. Self-sup-
of the organic pore directing agent.
porting zeolite wafers (20 mg, 5 ton cm−2, 1 cm2) were pre-
Sn-MFI was synthesised according to the method developed treated at 10−3 mbar and 693 K for 4 h. After cooling down to
by Mal et al.22 TEOS (93.7 g, Sigma-Aldrich, 98%) was mixed room temperature, the sample was saturated with pyridine/d3-
with a solution of SnCl4·5H2O (1.26 g, Sigma-Aldrich, 98%) in acetonitrile vapour and then evacuated at room temperature
deionised H2O (30 g) and allowed to react for 30 min under (15 min). In the case of pyridine, subsequent desorption steps
stirring. Then, TPAOH (198 g, Alfa Aesar, 20 wt%) was added at 473 K (30 min), 523 K (20 min), 573 K (20 min), and 623 K
drop-wise and the mixture was stirred for another hour. (20 min) were performed. Spectra were recorded in the
Finally, 84 g of deionised H2O were added and the solution 650–4000 cm−1 range with a resolution of 4 cm−1 by co-
was stirred for 30 min. The final molar ratio of the synthesis addition of 32 scans. UV-Vis spectroscopy was performed
gel was 0.008SnO2 : 1SiO2 : 0.44TPAOH : 34.3H2O. The sol was using an Ocean Optics UV-Vis spectrometer in diffuse reflec-
transferred into a 500 cm3 Teflon-lined autoclave, which was tance mode using barium sulfate for the background collection.
heated in an oven at 433 K for 3 days under static conditions. Prior to the analysis, the samples were dried at 573 K for 3 h
The resulting crystals were separated by filtration and sub- under vacuum (10−1 mbar). X-ray photoelectron spectroscopy
jected to the same work-up as that described for Ti-MFI.
(XPS) was carried out under UHV (10−8 mbar) using a VG
CBV720 and HSZ890 H0A (hereafter denoted as MFI), Thermo Escalab 220i-XL spectrometer equipped with a high-
referred to as parent (code P), were used for post-synthetic intensity twin-anode (Al Kα radiation) source. The spectra were
modification, that is, alkaline-assisted metallation. The latter charge-corrected with respect to the C 1s signal at 284.5 eV.
consisted of an alkaline treatment (AT) in an aqueous 0.3 M
2.3 NMR spectroscopy
NaOH solution (30 cm3 per gram of dried zeolite), containing
either 0.05 M of Ga(NO3)3 (Sigma-Aldrich, 99.9%) or of Sn 13C and DEPT-13523 nuclear magnetic resonance (NMR)
(SO4)2 (Acros, 98%) at 338 K for 30 min in an Easymax™ 102 spectroscopic studies were applied to unravel the mechanism
reactor (Mettler Toledo). In the case of the alkaline solution of GLY isomerisation to GA. Measurements were carried out
containing tin sulfate, owing to the rather low solubility of the using a Bruker Avance III HD spectrometer (Bruker Biospin)
1
tin hydroxide formed, the mixture was heated at 318 K for 1 h operating at a H Larmor frequency of 500.13 MHz (125 MHz
and filtered to remove the precipitate prior to the addition of for 13C), using a BBFO probe. The 13C experiment was a stan-
the zeolite. The resulting modified metallosilicates were con- dard broadband 1H decoupled sequence using a WALTZ-16,24 in
verted into their protonic form by three consecutive ion which 128 scans were co-added. For the DEPT spectrum, 2048
exchanges in an aqueous solution of ammonium nitrate (0.1 scans were co-added using the standard pulse sequence (also
1
M NH4NO3, 6 h, 298 K, 100 cm3 per gram of dried zeolite) fol- with WALTZ-16 H decoupling) with a recycle delay of 2 s. Both
lowed by calcination in static air at 823 K (5 K min−1) for 5 h. experiments used a spectral width of 236.5 ppm centered at
These samples are denoted by the codes CBV720-ATGa and 100 ppm and consisted of 65 000 points. The phase of the DEPT
MFI-ATSn, respectively.
spectrum was adjusted using the CH3 signal of TSP-d4, which
was added to the samples to serve as a reference. The spectra
were processed using the Bruker Topspin 3.2 software.
2.2 Characterisation
The content of Si, Al, Ga, and Sn in the catalysts and in the fil-
trate obtained after catalyst preparation was determined by
2.4 Catalytic tests
inductively coupled plasma optical emission spectroscopy Batch catalytic tests were performed under autogenous
(ICP-OES) using a Horiba Ultra 2 instrument equipped with a pressure in 15 cm3 thick-walled glass vials (Ace, pressure
photomultiplier tube detector. Prior to the measurements, the tubes, front seal) dipped in an oil bath heated at the desired
1178 | Green Chem., 2014, 16, 1176–1186
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
reaction temperature (363 or 373 K). The vials were loaded before heating to 473 K (30 K min−1). The alkyl glycolate yields
with 200 mg of glyoxal (GLY, Sigma-Aldrich, 95%, trimer), were determined using diethyl malonate (Sigma-Aldrich, 99%)
80 mg of catalyst, and 3.80 g of solvent. The latter comprised as an internal standard. The GLY conversion was calculated as
water, deuterated water (D2O, Sigma-Aldrich, 100%, for the the mole of GLY reacted divided by the mole of GLY fed, and
NMR spectroscopic studies), methanol, ethanol (both pur- the GA/alkyl glycolate yields as the mole of GA/alkyl glycolate
chased from Scharlau, 99.9%), or butan-1-ol (Merck, 99.5%). formed divided by the initial moles of GLY. Both conversion
The mixture was allowed to react under vigorous stirring for and yields are expressed as a percentage.
0.5 to 48 h. Then, the reaction was quenched using an ice bath
and the catalyst was removed by means of a Chromafil Xtra
0.25 μm syringe filter. Continuous-flow catalytic tests were con-
ducted in a H-Cube Pro™ reactor system (Thales Nanotechnol-
3 Results and discussion
3.1 Commercial USY zeolites and galliated samples
ogy) equipped with
a fixed-bed micro-reactor (Catalyst
Based on the mechanistic information mentioned in the intro-
duction and in order to evaluate the role of the type of acidity
in the GLY isomerisation, the reaction was initially investigated
over commercially available USY zeolites. These alumino-
silicates possess Brønsted as well as Lewis acidity, the latter
originating from the extra-framework aluminium (EFAl)
species generated by steaming and acid leaching of the pris-
tine Y. Various USY zeolite catalysts were selected in order to
span a wide range of EFAl amounts. The evolution of the con-
centration of the components of the reaction mixture was first
studied over CBV712 (H-USY, Si/Al = 6), a material character-
ised by substantial Lewis acidity. The conversion of GLY pro-
gressively increased with time attaining a value of 66% after
48 h (Fig. 1). Simultaneously, the GA yield rose up to 60%.
Remarkably, no by-products were detected, in contrast to the
case of the aqueous-phase isomerisation of dihydroxyacetone
in which the pyruvaldehyde intermediate, i.e. the GLY ana-
logue, was prone to degradation into acetol, pyruvic acid, and
formaldehyde over Brønsted-acid sites.25 Impurities present in
the substrate, including formic acid, acetic acid, and ethylene
glycol, account for the remaining fraction to close the carbon
balance. Since separation of bio-based molecules is often
regarded as a strong contributor to the process cost, the
absence of by-products is extremely advantageous from the
perspective of a large-scale implementation. In order to
compare the performance of the zeolites, 18 h was selected as
the reaction time as it appeared as a good compromise
between the yield of GA and the time needed for its
cartridge CatCart®, h = 70 mm, i.d. = 3 mm) heated by an iso-
thermal heating jacket controlled by a Peltier unit (Scheme 2).
The catalyst (269 mg, sieve fraction = 0.2–0.4 mm) was loaded
into the reactor and heated at 363 K. Thereafter, the reaction
was started by admitting a 2 wt% GLY aqueous solution to the
catalytic bed by means of an on-line HPLC pump at a flow rate
of 0.3 cm3 min−1
.
GLY and glycolic acid (GA) were isolated by high-perform-
ance liquid chromatography (HPLC) in a Merck LaChrom
system equipped with a Biorad Aminex HPX-87H column
heated at 308 K and a refractive index detector (Hitachi Chro-
master model 5450) set at 303 K using an aqueous eluent of
0.005 M H2SO4 (pH 2.30) flowing at 0.6 cm3 min−1. Quantifi-
cation was attained by integration of their respective peaks
using butan-2-one (Merck, 99.5%) as an internal standard.
Alkyl glycolates were analysed using a gas chromatograph
(GC, HP 6890) equipped with an HP-5 capillary column and a
flame ionisation detector. Methyl glycolate (Sigma-Aldrich,
98%), ethyl glycolate (Sigma-Aldrich, 98%), and n-butyl glyco-
late (Sigma-Aldrich, 90%) were employed as references.
Helium was used as the carrier gas (flow rate = 4.3 cm3 min−1
,
pressure = 1.4 bar) and an injection volume of 0.3 μL was
applied. The initial temperature of 328 K was held for 2 min
Scheme 2 Illustration of the catalytic set-up used for the continuous-
flow isomerisation of glyoxal.
Fig. 1 Concentration profiles of the reactant and product during the
conversion of GLY in water over CBV712 at 363 K.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1176–1186 | 1179

View Article Online
Paper
Green Chemistry
production. Before 18 h, the gain in the glycolic acid yield per zeolites already reported in the literature as well as a novel
unit time is considerable while it becomes progressively less material were investigated. First, post-synthetic modification of
and less pronounced thereafter (Fig. 1). While the GA selecti- the USY framework by alkaline-assisted galliation was con-
vity remained unaffected (above 90%), the GLY conversion and sidered, which rendered highly selective catalysts for the iso-
the GA yield increased with the EFAl content (Fig. 2), reaching merisation of dihydroxyacetone in alcoholic media.19b,c The
a maximum of 62 and 57%, respectively, over the severely best performing galliated zeolite reported in that work,
steamed and most EFAl-rich CBV600 (Table 1, entry 3). obtained by treatment of CBV720-P (Table 1, entry 4) with
Notably, the EFAl-free CBV760, CBV780, and HSZ390 converted 0.2 M NaOH and 0.04 M Ga(NO3)3, displayed improved results:
GLY to GA to the level of only 2–6%. These results demonstrate 74% GLY conversion and 71% GA yield. Further tuning of the
Lewis-acid sites as the catalytic centres for the reaction, while concentration of both the base (0.3 M) and the gallium salt
the cooperative effect of Brønsted-acid sites seems negligible. (0.05 M) in the zeolite treatment enabled us to increase the
This is further supported by (i) the comparable trend observed amount of gallium incorporated in the USY framework from
for the Lewis-acid catalysed production of ethyl lactate over a ca. 6 to 7.5 wt%, still largely preserving the original crystalli-
similar set of USY zeolites,19b and (ii) the inactivity of a purely nity of the sample (Fig. SI1†). This optimised CBV720-ATGa
Brønsted-acidic material, Amberlyst-15 (Table 1, entry 2). This catalyst (Table 1, entry 5) exhibited enhanced GLY conversion
conclusion corroborates the DFT prediction of a more favour- and GA yield, i.e. 80 and 77%, respectively, which thus sur-
able isomerisation mechanism based on a 1,2-hydride shift passed by ca. 20% those obtained over the best commercial
rather than dehydration–tautomerisation.16 The reaction aluminosilicate identified (CBV600). Remarkably, raising the
pathway is demonstrated by NMR spectroscopic studies in the temperature by only 10 K enabled us to reach almost full GLY
following paragraphs.
conversion and a GA yield as high as 91% (Table 1, entry 6).
Based on these findings, with the aim of attaining an opti-
mised catalytic system for the production of GA, Lewis-acidic
3.2 Ti- and Sn-based MFI zeolites
As the next step, tin- and titanium-based zeolites were con-
sidered. Isomorphously-substituted zeolites prepared hydro-
thermally and containing these metals are expected to be
superior catalytic materials owing to the intrinsically stronger
Lewis acidity and structural uniformity of the tin and titanium
species, which have already enabled very fast conversion rates
of trioses, pentoses, and hexoses.17 With regard to tin-
containing zeolites, Sn-BEA synthesised in fluoride media
exhibited unrivalled performances in these Lewis-acid cata-
lysed isomerisations.17 Nevertheless, since its complex syn-
thesis was demonstrated to be difficult to reproduce even at
the lab-scale level, materials based on the MFI framework,
which can be obtained in a much more straightforward
manner, were preferred for the evaluation in the trans-
formation of GLY. Furthermore, even if a similar or higher
activity of Sn-BEA compared to Sn-MFI for the isomerisation of
C2 substrates cannot be excluded, there are indications in the
Fig. 2 GLY conversion and GA yield over various commercially available
USY zeolites at 363 K and 18 h. The EFAl amount in the samples was
derived from ref. 26.
Table 1 Composition, textural properties, and the main catalytic data obtained after 18 h at 363 Ka
Si/Alc
Metal contentc
(wt%)
Smeso
Vmicro
GLY
conversion (%)
GA yield
(%)
GA selectivity
(%)
d
d
Entry
Catalystb
(mol mol−1
)
(m2 g−1
)
(cm3 g−1
)
1
2
3
4
5
6
7
8
Blank
Amberlyst-15
CBV600
CBV720-P
CBV720-ATGa
CBV720-ATGa [373 K]
Ti-MFI
Sn-MFI
Sn-MFI [6 h]
MFI-P
MFI-ATSn
MFI-ATSn [6 h]
—
—
2.6
17
10
10
—
—
—
961
—
—
—
—
—
—
7.5
7.5
n.d.
1.70
1.70
—
—
41
76
128
186
186
40
41
41
—
0
1
1
1
1
100
100
92
91
96
94
91
91
94
0.23
0.29
0.20
0.20
0.14
0.14
0.14
0.17
0.10
0.10
62
11
80
97
65
99
95
4
57
10
77
91
59
91
89
3
9
10
11
12
53
118
118
100
94
95
2.55
2.55
96
73
90
69
a Reaction conditions: Wcat = 80 mg, WH O = 3.80 g, WGLY = 0.20 g. b P = parent, AT = alkaline treatment with 0.3 M NaOH in the presence of a
gallium (Ga) or tin (Sn) source. c Determi2ned by ICP-OES. Metals comprise Sn or Ga. d Determined by the t-plot method.
1180 | Green Chem., 2014, 16, 1176–1186
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
literature that the latter zeolite has a better isomerisation filtrate after synthesis indicated the incorporation of 2.55 wt%
ability than the former for smaller bio-derived substrates (C3 tin into the zeolite and a tin insertion efficiency upon the alka-
molecules).27 The textural properties, chemical composition, line-assisted process equal to 13%. The uptake at mid-high
and catalytic performance of the Ti-MFI and Sn-MFI prepared relative pressures in the N2 isotherms indicated the develop-
are reported in Table 1 (entries 7 and 8) and in Fig. S1†. Sur- ment of intra-particle mesoporosity (118 m2 g−1, pore size dis-
prisingly, the GLY conversion and the GA yield over the former tribution centred at 8 nm), which was further evidenced
solid only reached 65% and 59% respectively, thus being by transmission electron microscopy (Fig. 3a and b). No
similar to the performance of CBV600. The limited activity of additional phase to the zeolite could be visualised by the latter
this catalyst might relate to the additional presence of titanium technique, which would suggest the presence of tin as SnO2.
in the form of extra-framework TiOx species, which are unavoid- X-ray diffraction analysis evidenced the retention of the MFI
ably generated to some extent during synthesis and calcination, structure upon metallation and the lack of diffraction peaks
as confirmed by UV-Vis spectroscopy of our sample (Fig. SI2†). specific to tin oxide, supporting the absence of this phase, at
In contrast, Sn-MFI yielded 90% of GA at full GLY conversion least in the form of large particles (Fig. 3c). Remarkably,
after only 6 h (Table 1, entry 9), therefore confirming the excep- MFI-ATSn exhibited a nearly identical performance with
tional characteristics of tin-based materials.
respect to Sn-MFI, i.e. 96% GLY conversion and 90% GA yield.
Due to the typically long synthesis time and the need for
Since this result was attained after a longer reaction time
HF, no industrially applicable protocol has been developed so (18 versus 6 h), the catalytic properties of the tin sites intro-
far for the preparation of tin-containing zeolites (e.g. Sn-BEA duced by post-synthetic metallation appeared to be slightly
and Sn-MFI).17d,28 Therefore, the availability of an analogous inferior to those of the hydrothermally-synthesised catalyst.
material obtained via an alternative and more scalable route This difference was more adequately quantified based on the
would be highly relevant. We first wondered whether the extra- turnover frequency (TOF), calculated as the mole of product
polation of recently reported protocols for the preparation of a formed per hour and the mole of Sn after 30 min of reaction,
highly active Sn-beta could be instrumental for the purpose. and the catalyst productivity, determined as the g of product
Such methods comprise dealumination of an aluminium-rich formed per hour and per kg of catalyst at a conversion level of
zeolite beta followed by tin incorporation by either solid-state 20%. Lower values of TOF and productivity were indeed
ion exchange or wet impregnation with a SnCl4 solution in iso-
propanol.29 As the demetallation step, common to the two
routes, requires the use of concentrated and, thus, highly cor-
rosive nitric acid and the solid-state reaction may not be appli-
cable on a large scale, we rather conceived to attempt the
preparation of Sn-MFI by alkaline-assisted metallation. This
approach involves the creation of defects in the zeolite struc-
ture by the action of hydroxide ions and their healing with
tin cations in one pot under mild conditions and holds
brighter promises for industrial viability, as post-synthetic
alkaline-based treatments of commercial zeolites have already
successfully demonstrated at the kilogram scale.30 Never-
theless, it should be underlined that the chemical challenge
associated with this strategy is considerable, as only the in-
corporation of trivalent metal ions (M = Al3+ and Ga3+), existing
−
as negatively charged M(OH)4 hydroxide species in strong
alkaline media, has been demonstrated so far.19 Not only tin
cations would possess a different charge (either 2 or 4), but
also the hydroxides correspondingly formed in the basic solu-
tion are very poorly soluble and instantaneously form a pre-
cipitate. Still, taking into account the relatively small quantity
of tin required to generate a highly active catalyst (1.7 wt% Sn
q−
in Sn-MFI), it was thought that even a tiny amount of Sn(OH)n
in solution could be a sufficient source of the metal. In order
to enable a direct comparison with Sn-MFI, the MFI-type silica-
lite-1 zeolite (denoted hereafter as MFI-P) was selected as the
parent material. Modification was performed in the presence
of a 0.3 M NaOH solution saturated with Sn(OH)nq−, prepared
using Sn(SO4) as the precursor. The composition and textural
Fig. 3 (a) N2 sorption isotherms, (b) transmission electron micrographs
and (c) X-ray diffraction patterns of MFI in parent form and after alkaline
properties of the resulting sample, MFI-ATSn, are summarised
treatment in the presence of Sn(SO4). The inset in (a) depicts the pore
in Table 1 (entry 11). Elemental analysis of the solid and the size distribution of the two samples.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1176–1186 | 1181

View Article Online
Paper
Green Chemistry
obtained for MFI-ATSn (ca. 65 h−1 and 1060 g kg−1 h−1) than pyridine desorption at 473 and 573 K, in line with previous
for Sn-MFI (ca. 191 h−1 and 2842 g kg−1 h−1). An in-depth reports.29b Values of 0.36 and 0.31 were obtained for MFI-ATSn
investigation of the main features of the novel catalyst was per- and Sn-MFI, respectively, thus pointing to a relatively similar
formed aiming at deriving a fundamental understanding of Lewis-acid strength of the Sn centres derived from the two syn-
the differences and similarities between the incorporation of thetic methods. In order to corroborate this result, the
tin by bottom-up and top-down strategies.
samples were probed with a molecule with weaker basicity, d3-
FTIR spectroscopy of adsorbed pyridine was applied to acetonitrile (Fig. 4b). Upon chemisorption on a Lewis-acid site,
provide insights into the amount and the strength of the the band associated with the stretching of the nitrile bond
Lewis-acid sites introduced. Fig. 4a depicts the infrared spectra undergoes a shift in position, which has been reported to
of MFI-ATSn and Sn-MFI after pyridine adsorption at room strongly depend on the strength of the site.28,31,32 In agree-
temperature and subsequent desorption at 473, 523, 573, and ment with previous evidence, the signal for d3-acetonitrile che-
623 K. The amount of Lewis-acid sites in the samples was misorbed on Sn-MFI appeared at 2310 cm−1, indicating the
calculated by integration of the signal at 1454 cm−1 in the presence of framework Sn atoms.28 The band obtained for
spectrum collected after degassing at 473 K, as this tempera- MFI-ATSn was very similar and was positioned at only 2 cm−1
ture is sufficient to remove the physisorbed pyridine and the lower frequency. This evidence points to the equality of the
pyridine bonded to the silanols while preserving the pyridine acid strength of the tin sites in the two catalysts.
chemisorbed on the tin centres. A 1.8 times higher concen-
As the difference in catalytic activity of the two materials
tration of acid sites was found for MFI-ATSn compared to Sn- cannot be rationalised based on their acidic properties, diffuse
MFI (71 versus 39 μmol g−1), which is in good agreement with reflectance UV-Vis spectroscopy and XPS were employed to
the 1.5 times higher tin content in the former catalyst, consid- shed light onto the structure of the tin species located in these
ering the experimental errors associated with the two charac- zeolites (Fig. 5). The UV-Vis spectrum of MFI-ATSn displays a
terisation techniques. This finding implies that all of the Sn peak at 217 nm, generally assigned to an electronic transition
centres are accessible to pyridine and, thus, that each metal of isolated Sn(IV) species coordinated to four oxygen atoms in
atom can potentially act as a catalytic site. The strength of the the zeolite framework.33,34 No evidence for hexa-coordinated
Sn sites was determined by evaluating the ratio of the integrals polymeric Sn species was found, which would typically give
of the same absorption band in the spectra collected after rise to a band at 280 nm, as observed for the reference material
SnO2.33,34 The peak related to tetra-coordinated species
Fig. 4 FTIR spectra of MFI-ATSn and Sn-MFI after (a) pyridine adsorp-
tion followed by evacuation at different temperatures and (b) d3-aceto-
Fig. 5 (a) Diffuse reflectance UV-Vis and (b) Sn 3d core-level XPS
nitrile adsorption followed by evacuation at room temperature.
spectra of SnO2, Sn-MFI, and MFI-ATSn.
1182 | Green Chem., 2014, 16, 1176–1186
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
appeared to be sharper in the case of Sn-MFI,35 highlighting 1,2-hydride shift occurs on the monohydrate form. A recent
the greater uniformity of the Lewis-acid sites incorporated study indicated that both hydrate forms are present in rela-
through hydrothermal synthesis and the larger distribution in tively similar quantities in water,38 although 1H NMR spec-
the environments of the Lewis-acid centres introduced via troscopy experiments evidenced the presence of only one peak
post-synthetic metallation. When considering the Sn 3d core- corresponding to the monohydrate compound.16 The dehy-
level XPS spectra, hydrothermally-synthesised Sn-MFI pro- dration mechanism comprises the formation of a double bond
duced a signal at ca. 486.5 eV, which matches that obtained by syn-elimination of water and subsequent tautomerisation of
from hydrothermally-synthesised Sn-beta.36 The sample pre- the resulting enol intermediate into its keto-isomer, that is,
pared by alkaline-assisted metallation displayed a broader GA. The 1,2-hydride shift mechanism relies on the concomi-
spectral feature comprising the same contribution as that tant oxidation of the α-hydroxyl and reduction of the carbonyl
found for Sn-MFI as well as at least another component at group, which is referred to as an intramolecular Meerwein–
higher binding energy (ca. 487.9 eV), suggesting that Sn(IV) Ponndord–Verley reduction Oppenauer oxidation (MPVO
atoms in the material have non-equivalent environments. It redox). Consequently, while the dehydration mechanism is
can be speculated that a fraction of the tin centres is totally expected to introduce deuterium into the carbon backbone, no
integrated in the framework (alike in Sn-MFI), while a fraction deuterium incorporation should occur via the 1,2-hydride
might be only partially incorporated. The presence of extra- shift. Fig. 6 represents the 1D 13C and DEPT-135 NMR spectra
framework tin species in an amount below the detection limit of the mixture obtained upon conversion of GLY over
of the techniques used cannot be discarded. Besides for the MFI-ATSn after 18 h. The 13C NMR spectrum displays two
differences arising from the distinct preparation routes, it main peaks centred at 62.5 and 178.5 ppm, respectively,
might be possible that silicalite-1 could not fully accommodate associated with the secondary and quaternary carbon of GA. A
the tin atoms in framework positions due to the higher metal small peak at 93.4 ppm, due to the dihydrate form of GLY, is
loading in the final catalyst compared to Sn-MFI. The hetero- also visible. Interestingly, the peak at 62.5 ppm does not show
1
geneity of the Sn sites in MFI-ATSn is expected to determine any triplet structure caused by a JDH-coupling to a directly
the slight inferiority of MFI-ATSn in GLY isomerisation. None- attached deuterium nucleus (see the inset in Fig. 6, top). The
theless, the major similarity of the tin sites obtained by the absence of deuterium in the backbone was further evidenced
two methods remains quite striking.
by the 1D 13C DEPT-135 spectrum, in which the methyl groups
of sodium 3-(trimethylsilyl)propionoate-2,2,3,3-d4 produce a
reference peak to correctly assign the polarity of the signals.
The fact that the peak centred at 62.5 ppm is negative undoubt-
edly confirms the exclusive presence of CH2 species, and thus
the sole occurrence of the 1,2-hydride shift mechanism.
3.3 Elucidation of the reaction mechanism by NMR
spectroscopy
In order to unravel the reaction mechanism, the conversion of
GLY was performed over MFI-ATSn in deuterated water and the
mixture was analysed by 13C NMR spectroscopy. As stated
above, the isomerisation of GLY to GA can theoretically follow
two main routes, being a Brønsted-acid catalysed dehydration
or a Lewis-acid catalysed 1,2-hydride shift (Scheme 3). Even
though the nature of the centres in MFI-ATSn is substantially
Lewis-acidic, and thus the latter mechanism should obviously
be favoured, the presence of a minor component of Brønsted
acidity, and therefore the occurrence of the former mechan-
ism, cannot be fully excluded. In fact, it was demonstrated that
Lewis-acid tin sites in Sn-BEA are partially hydrolysed, i.e. Sn-
OH, in the presence of water.37 While the dehydration route
involves the reaction of the dihydrate form of GLY, the
3.4 Assessment of industrially relevant aspects of Sn-based
zeolites
In order to assess the potential of MFI-ATSn for practical appli-
cation, its stability was evaluated by means of recyclability and
hot-filtration studies. In the former case, the GLY conversion
and the GA yield were monitored upon direct reuse of the cata-
lyst, i.e. without any intermediate regeneration, in five sub-
sequent catalytic runs (Fig. 7a). Remarkably, no loss of activity
Scheme 3 Possible reaction pathways for the catalytic isomerisation of
Fig. 6 13C NMR spectra of the reaction mixture obtained at 363 K and
GLY to GA.
18 h over MFI-ATSn.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1176–1186 | 1183

View Article Online
Paper
Green Chemistry
was observed and the GA yield decreased by only 5% by the
fifth run. The original performance could be fully recovered by
calcination (823 K, 5 K min−1, 5 h, static air) of the sample at
this stage, as the outcome of the sixth experiment indicates.
To exclude that tin leaching took place, the catalyst was filtered
off at the reaction temperature from the mixture obtained after
1.5 h and the filtrate was allowed to react for 4.5 h. As illus-
trated in Fig. 7b, no further production of GA was detected
after removal of the catalyst, thus confirming the truly hetero-
geneous nature of the catalyst.
Based on these results, the extrapolation from a batch to
continuous-flow mode was investigated, since the latter is
widely preferred for the production of bulk chemicals, such as
GA. A 2 wt% GLY aqueous solution was allowed to flow at
0.3 cm3 min−1 during 8 h through a catalytic bed (0.5 cm3) of
MFI-ATSn heated at 363 K. Under these conditions, both GLY
conversion and GA selectivity remained stable at the remark-
able levels of 40 and 96%, respectively (Fig. 8). We would like
to stress the relevance of this result, as the wide majority of
chemocatalytic systems for the production of high-volume
chemicals in the field of biomass conversion so far demon-
strated have been limited to the batch operation.
Fig. 8 GLY conversion and GA selectivity in a continuous-flow experi-
ment over MFI-ATSn at 363 K.
concentration was raised up to five-fold, i.e. 20 wt%. As shown
in Fig. 9, the GA yield progressively decreased to 45% upon
utilisation of less and less diluted solutions. Nevertheless,
raising the reaction temperature by only 10 K enabled us to
substantially restore the GA yield (81%) with the use of the
most concentrated feed.
Finally, the possibility to attain a more industrially signifi-
cant process productivity was explored by evaluating the cata-
lyst response to the application of a higher feed concentration.
Tests were conducted batch-wise in which the catalyst amount
was kept constant (ca. 80 mg) each time, while the GLY
3.5 Production of alkyl glycolates over Sn-based zeolites
As the substitution of the water medium by an alcohol enabled
the direct conversion of dihydroxyacetone into the corres-
ponding ester derivatives of lactic acid,19b,c we approached the
preparation of methyl, ethyl, and butyl glycolates following the
same strategy (Scheme 1). While the latter two compounds
have found widespread application as plasticisers, methyl gly-
colate represents an intermediate in the industrial manufac-
ture of ethylene glycol.20 Table 2 reports the performances of
MFI-ATSn and Sn-MFI in the three alcoholic solvents after a
reaction time of 18 and 6 h, respectively. The yield of the alkyl
glycolates reached a maximum of 83% for the methyl ester,
thus being overall lower compared to the case of GA (91%).
Specifically, the higher the length of the alcohol, the lower the
yield of the corresponding alkyl glycolate. Sn-MFI was slightly
superior to MFI-ATSn in methanol, whereas an inverse trend
was observed in ethanol and n-butanol. The latter is likely a
Fig. 7 (a) GLY conversion and GA yield at 363 K and 18 h upon reuse in
consecutive catalytic runs and (b) a hot-filtration test over MFI-ATSn.
Fig. 9 GLY feed concentration versus GA yield over MFI-ATSn.
1184 | Green Chem., 2014, 16, 1176–1186
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 2 GLY conversion over MFI-ATSn and Sn-MFI in various
alcoholsa
Acknowledgements
This work was supported by the Swiss National Science
Foundation (project number 200021-140496). The Electron
Microscopy Centre of ETH Zurich is acknowledged for the use
of their facility. Dr S. Mitchell is thanked for TEM analysis and
Dr I. Czekaj for XPS measurements.
Solvent
Catalyst
GLY conversion (%)
Glycolate yield (%)
Methanol
Ethanol
n-Butanol
MFI-ATSn
Sn-MFI
MFI-ATSn
Sn-MFI
MFI-ATSn
Sn-MFI
85
85
73
70
53
49
82
83
69
68
53
45
a Reaction conditions: Wcat = 80 mg, Walcohol = 3.80 g, WGLY = 0.20 g,
t = 6 or 18 h for Sn-MFI or MFI-ATSn, respectively.
Notes and references
1 (a) P. N. R. Vennestrøm, C. M. Osmundsen,
C. H. Christensen and E. Taarning, Angew. Chem., Int. Ed.,
2011, 50, 10502; (b) P. Y. Dapsens, C. Mondelli and J. Pérez-
Ramírez, ACS Catal., 2012, 2, 1487.
2 H. Seyednejad, A. H. Ghassemi, C. F. van Nostrum,
T. Vermonden and W. E. J. Hennink, Controlled Release,
2011, 152, 168.
3 Available at: http://www.transparencymarketresearch.com/
glycolic-acid-market.html#sthash.x0g3VLpI.dpuf, retrieved
September 2013.
consequence of the more pronounced benefit of the presence
of a mesoporous network in MFI-ATSn when dealing with
bulkier molecules. In line with the essentially pure Lewis-acid
character of the catalysts, no evidence for the Brønsted-acid
catalysed alkylation of glyoxal was found. The competitive for-
mation of pyruvic aldehyde dialkyl acetals was also suppressed
in the case of dihydroxyacetone isomerisation over tin-based
zeolites.
4 M. Dusselier, P. Van Wouwe, A. Dewaele, E. Makshina and
B. F. Sels, Energy Environ. Sci., 2013, 6, 1415.
5 D. J. Loder, US Patent 2152852, 1939.
4 Conclusions
A chemocatalytic process for the valorisation of glyoxal, an
undesired oxygenated component of pyrolysis oil, was investi-
gated in this study. Its aqueous-phase isomerisation over
Lewis-acid zeolites was put forward as an efficient way to
renewably produce glycolic acid, a relevant skin care agent and
building block for polymer manufacture. Spectroscopic investi-
gations indicated that the reaction mechanism relies on a 1,2-
hydride shift. Tin-containing materials demonstrated the best
catalysts for this reaction, yielding more than 90% of the
target product in short times at only 363 K without forming
any by-products. Besides confirming the extraordinary Lewis-
acid properties of hydrothermally-synthesised Sn-MFI, our
work uncovered a novel zeolite displaying similar character-
istics, which was generated by simple alkaline treatment of sili-
calite-1 in the presence of a tin salt. The catalyst was shown to
be recyclable and to act in a truly heterogeneous manner. Fur-
thermore, excellent results were obtained in its application to
the continuous-flow production of glycolic acid. The gene-
ration of a set of industrially relevant alkyl glycolates by iso-
merisation of glyoxal in different alcoholic solvents was also
successfully achieved on the tin-based zeolites. Overall, the
6 (a) R. DiCosimo, A. Panova, D. P. O’Keefe, J. Thompson
and M. S. Payne, US Patent 7473545, 2009; (b) F. Ebmeyer,
H. Haberlein, H. H. Haberlein, J. T. Haberlein,
M. C. Haberlein and H. Mohn, US Patent 5723662;
(c) B. Scharbert and P. Babusiaux, US Patent 5395488.
7 J. Zhang, X. Liu, M. Sun, X. Ma and Y. Han, ACS Catal.,
2012, 2, 1698.
8 (a) P. Maki-Arvela, T. Salmi, B. Holmbom, S. Willfor and
D. Yu Murzin, Chem. Rev., 2011, 111, 5638;
(b) D. K. Mishra, A. A. Dabbawala and J. S. Hwang, J. Mol.
Catal. A: Chem., 2013, 376, 63; (c) M. Yadav, D. K. Mishra
and J. S. Hwang, Appl. Catal., A, 2012, 425, 110;
(d) Z. Huang, F. Li, B. Chen, F. Xue, Y. Yuan, G. Chen and
G. Yuan, Green Chem., 2011, 13, 3414; (e) J. Sun and H. Liu,
Green Chem., 2011, 13, 135.
9 (a) M. Ziolek, I. Sobczak, P. Decyk and L. Wolski, Catal.
Commun., 2013, 37, 85; (b) A. Takagaki, A. Tsuji,
S. Nishimura and K. Ebitani, Chem. Lett., 2011, 40, 150;
(c) M. Sankar, N. Dimitratos, D. W. Knight, A. F. Carley,
R. Tiruvalam, C. J. Kiely, D. Thomas and G. J. Hutchings,
ChemSusChem, 2009, 2, 1145.
proposed process combines: (i) the use of a renewable and 10 (a) T. A. Geissman, Org. React., 1944, 2, 94; (b) H. Wittcoff,
waste-type feedstock, (ii) the application of mild conditions,
thus limiting the energy requirements, (iii) the use of green
solvents, (iv) the absence of by-products, and (v) the intrinsic
Org. Synth., 1963, 4, 907; (c) R. M. Kellogg, in Comprehensive
Organic Synthesis, ed. B. M. Trost and I. Fleming, Pergamon
Press, Oxford, 1991, vol. 8.
advantages of a heterogeneous catalyst, which is prepared 11 D. Mohan, C. U. Pittman Jr. and P. H. Steele, Energy Fuels,
according to an industrially scalable protocol and thus 2006, 20, 848.
appears to hold a great potential to replace or substantiate the 12 M. Balat, M. Balat, E. Kirtay and H. Balat, Energy Convers.
current environmentally-unfriendly process for the production Manage., 2009, 50, 3147.
of glycolic acid and to establish a one-pot methodology for the 13 Available at: http://www.biofuelstp.eu/bio-oil.html, retrieved
generation of its alkyl ester derivatives.
in September 2013.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 1176–1186 | 1185

View Article Online
Paper
Green Chemistry
14 K. Sipilä, E. Kuoppala, L. Fagernäs and A. Oasmaa, Biomass 24 A. J. Shaka, J. Keeler, T. Frenkiel and R. Freeman, J. Magn.
Bioenergy, 1998, 14, 103. Reson., 1983, 52, 335.
15 P. Van Wouwe, M. Dusselier, A. Basic and B. F. Sels, Green 25 C. Rae, S. J. Berners-Price, B. T. Bulliman and P. W. Kuchel,
Chem., 2013, 15, 2817. Eur. J. Biochem., 1990, 193, 83.
16 T. Ohshima, Y. Yamamoto, U. Takaki, Y. Inoue, T. Saeki, 26 P. P. Pescarmona, K. P. F. Janssen, C. Delaet,
K. Itou, Y. Maegawa, T. Iwasakia and K. Mashima, Chem.
Commun., 2009, 2688.
17 (a) E. Taarning, S. Saravanamurugan, M. S. Holm, J. Xiong,
C. Stroobants, K. Houthoofd, A. Philippaerts, C. De
Jonghe, J. S. Paul, P. A. Jacobs and B. F. Sels, Green
Chem., 2010, 12, 1083.
R. M. West and C. H. Christensen, ChemSusChem, 2009, 2, 27 C. M. Lew, N. Rajabbeigi and M. Tsapatsis, Microporous
625; (b) Y. Román-Leshkov, M. Moliner, J. A. Labinger and Mesoporous Mater., 2012, 153, 55.
M. E. Davis, Angew. Chem., Int. Ed., 2010, 49, 8954; 28 C. M. Osmundsen, M. S. Holm, S. Dahl and E. Taarning,
(c) V. Choudhary, A. B. Pinar, S. I. Sandler, D. G. Vlachos Proc. R. Soc. London, Ser. A, 2012, 468, 2000.
and R. F. Lobo, ACS Catal., 2011, 1, 1724; (d) M. Moliner, 29 (a) C. Hammond, S. Conrad and I. Hermans, Angew. Chem.,
Dalton Trans., 2014, DOI: 10.1039/C3DT52293H.
Int. Ed., 2012, 51, 11736; (b) J. Dijkmans, D. Gabriëls,
M. Dusselier, F. de Clippel, P. Vanelderen, K. Houthoofd,
A. Malfliet, Y. Pontikes and B. F. Sels, Green Chem., 2013,
15, 2777.
18 (a) M. S. Holm, S. Saravanamurugan and E. Taarning,
Science, 2010, 328, 602; (b) F. de Clippel, M. Dusselier,
R. Van Rompaey, P. Vanelderen, J. Dijkmans, E. Makshina,
L. Giebeler, S. Oswald, G. V. Baron, J. F. M. Denayer, 30 J. Pérez-Ramírez, S. Mitchell, D. Verboekend, M. Milina,
P. P. Pescarmona, P. A. Jacobs and B. F. Sels, J. Am. Chem.
Soc., 2012, 134, 10089; (c) L. Li, C. Stroobants, K. Lin,
N.-L. Michels, F. Krumeich, N. Marti and M. Erdmann,
ChemCatChem, 2011, 3, 1731.
P. A. Jacobs, B. F. Sels and P. P. Pescarmona, Green Chem., 31 A. Corma, L. T. Nemeth, M. Renz and S. Valencia, Nature,
2011, 13, 1175; (d) Q. Guo, F. Fan, E. A. Pidko, W. N. P. van 2001, 412, 423.
der Graaff, Z. Feng, C. Li and E. J. M. Hensen, Chem- 32 M. Boronat, P. Concepción, A. Corma, M. Renz and
SusChem, 2013, 6, 1352. S. Valencia, J. Catal., 2005, 234, 111.
19 (a) P. Y. Dapsens, C. Mondelli and J. Pérez-Ramírez, Chem- 33 R. Bermejo-Deval, R. Gounder and M. E. Davis, ACS Catal.,
SusChem, 2013, 6, 831; (b) P. Y. Dapsens, M. J. Menart, 2012, 2, 2705.
C. Mondelli and J. Pérez-Ramírez, Green Chem., 2014, DOI: 34 S. Roy, K. Bakhmutsky, E. Mahmoud, R. F. Lobo and
10.1039/C3GC40766G; (c) P. Y. Dapsens, B. T. Kusema, R. J. Gorte, ACS Catal., 2013, 3, 573.
C. Mondelli and J. Pérez-Ramírez, J. Mol. Catal. A: Chem., 35 Y. K. Hwang, T. H. Jin, J. M. Kim, Y.-U. Kwon, S.-E. Park
2014, DOI: 10.1016/j.molcata.2013.09.032. and J.-S. Chang, J. Nanosci. Nanotechnol., 2006, 6, 1.
20 K. Miltenberger, in Ullmann’s Encyclopedia of Industrial 36 H. Y. Luo, L. Bui, W. R. Gunther, E. Min and Y. Román-
Chemistry, VCH, Weinheim, 2000, vol. 18, pp. 481–492. Leshkov, ACS Catal., 2012, 2, 2695.
21 D.-G. Huang, X. Zhang, B.-H. Chen and Z.-S. Chao, Catal. 37 R. Bermejo-Deval, R. S. Assary, E. Nikolla, M. Moliner,
Today, 2010, 158, 510.
22 N. K. Mal, V. Ramaswamy, P. R. Rajamohanan and
A. V. Ramaswamy, Microporous Mater., 1997, 12, 331.
Y. Román-Leshkov, S. Hwang, A. Palsdottir, D. Silverman,
R. F. Lobo, L. A. Curtiss and M. E. Davis, Proc. Natl. Acad.
Sci. U. S. A., 2012, 109, 9727.
23 M. R. Bendall, D. M. Doddrell and D. T. Pegg, J. Am. Chem. 38 G. Yu, A. R. Bayer, M. M. Galloway, K. J. Korshavn, C. G. Fry
Soc., 1981, 103, 4603.
and F. N. Keutsch, Environ. Sci. Technol., 2011, 45, 6336.
1186 | Green Chem., 2014, 16, 1176–1186
This journal is © The Royal Society of Chemistry 2014