Catalytic dehydration of levoglucosan to levoglucosenone using Br?nsted solid acid catalysts in tetrahydrofuran

Article abstract of DOI:10.1039/c9gc01526d

We studied the production of levoglucosenone (LGO) via levoglucosan (LGA) dehydration using Br?nsted solid acid catalysts in tetrahydrofuran (THF). The use of propylsulfonic acid functionalized silica catalysts increased the production of LGO by a factor of two compared to the use of homogeneous acid catalysts. We obtained LGO selectivities of up to 59% at 100% LGA conversion using solid Br?nsted acid catalysts. Water produced during the reaction promotes the solvation of the acid proton reducing the activity and the LGO production. Using solid acid catalysts functionalized with propylsulfonic acid reduces this effect. The hydrophilicity of the catalyst surface seems to have an effect on reducing the interaction of water with the acid site, improving the catalyst stability.

Full text of DOI:10.1039/c9gc01526d

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: O. Oyola-Rivera,
J. He, G. Huber, J. Dumesic and N. Cardona-Martínez, Green Chem., 2019, DOI: 10.1039/C9GC01526D.
Volume 20
This is an Accepted Manuscript, which has been through the
Number
May 2018
Pages 1919-2160
9
7
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Page 1 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
Catalytic dehydration of levoglucosan to levoglucosenone using Brønsted solid acid
catalysts in tetrahydrofuran
Oscar Oyola-Riveraa, Jiayue Heb, George W. Huberb, James A. Dumesicb, and Nelson
Cardona-Martíneza*
E-mail addresses: oscar.oyola@upr.edu (O. Oyola-Rivera), jiayue.he@wisc.edu (Jiayue
He), jdumesic@wisc.edu (James A. Dumesic), gwhuber@wisc.edu (George W. Huber),
and nelson.cardona@upr.edu (N. Cardona-Martínez)
Abstract
We studied the production of levoglucosenone (LGO) via levoglucosan (LGA)
dehydration using Brønsted solid acid catalysts in tetrahydrofuran (THF). The use of
propylsulfonic acid functionalized silica catalysts increased the production of LGO by a
factor of two compared to the use of homogeneous acid catalysts. We obtained LGO
selectivities of up to 59% at 100% LGA conversion using solid Brønsted acid catalysts.
Water produced during the reaction promotes the solvation of the acid proton reducing the
activity and the LGO production. Using solid acid catalysts functionalized with
propylsulfonic acid reduces this effect. The hydrophilicity of the catalyst surface seems to
have an effect in reducing the interaction of water with the acid site, improving the catalyst
stability.
aDepartment of Chemical Engineering, University of Puerto Rico, Mayagüez, PR, 00681-
9000, Puerto Rico
bDepartment of Chemical and Biological Engineering, University of Wisconsin, Madison,
WI, 53706, United States
*Corresponding author.
1

Green Chemistry
Page 2 of 41
View Article Online
DOI: 10.1039/C9GC01526D
1. Introduction
Lignocellulosic biomass is one of the most abundant carbon sources in the world
and has the potential of becoming an important source for the renewable production of
commodity chemicals and transportation fuels.1–5 Biomass may be obtained from forestry
resources, agricultural residues, energy crops, and industry wastes.6,7 Levoglucosenone
(LGO) is a platform molecule that can be produced from cellulose by different catalytic
routes.5,8–11 LGO is an anhydrous sugar that has been used as a feedstock for the synthesis
of pharmaceuticals, 1,6-hexanediol (a polymer precursor), and CyreneTM (a
solvent).5,8,10,12–14 LGO is commonly produced from the catalytic pyrolysis of cellulose and
the yield obtained for LGO is typically low (less than 20%).10,11,15 Recently, Huber and co-
workers, obtained yields of 38%, for LGO from cellulose in tetrahydrofuran (THF) in the
absence of water using sulfuric acid as catalyst.8 However, the LGO yield decreased from
38% to less than 15% as the cellulose loading increased from 1 to 10 wt%.8
LGO is produced from the dehydration of levoglucosan (LGA) during the
conversion of cellulose in the presence of an acid catalyst and in the absence of water.5,8,16,17
LGA is an anhydrous sugar that is mainly produced from cellulose by fast pyrolysis, at
yields of up to 45%.18 The production of LGO has been mainly studied via acid-catalyzed
pyrolysis of LGA and cellulose using acid catalysts.11,15,19–21 Two routes have been
proposed for the acid-catalyzed conversion of LGA to LGO. LGA may be directly
dehydrated to LGO or iso-LGO, or it can be dehydrated to 1,4:3,6-dianhydro-a-D-
glucopyranose (DGP) followed by the dehydration of DGP to LGO.5,16,17
Cao et al., studied the production of LGO from LGA using 20 mM sulfuric acid in
tetrahydrofuran (THF) and water.5 They obtained LGO yields of up to 50% starting with
2

Page 3 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
0.4 wt% LGA in pure THF.5 By adding 2.7 wt% water in THF, the LGO yield decreases
from 50% to around 22% due to the hydrolysis of LGA to glucose and the production of
5-hydroxymethylfurfural (HMF) from glucose.5 These observations show that the presence
of water adversely affects the production of LGO from LGA. However, water production
during the reaction is unavoidable since LGO is produced from LGA through a dehydration
reaction. Therefore, we used an acid catalyst, with similar properties found for sulfuric acid
such as propylsulfonic acid, supported on silica with hydrophilic silanol groups that may
help to reduce the effects of water during the reaction.22 Using a solid catalyst also
facilitates the separation of products and reduces exposure during the process to corrosive
liquid species such as sulfuric acid.
In this study, we explored the use of solid Brønsted acid functionalized silica
catalysts for the dehydration of LGA to LGO in THF. Using acid functionalized silica
catalysts, we obtained LGO selectivities of up to 68% at 58% LGA conversion. We
observed that the silica surface silanol groups interact with the water produced during the
reaction, reducing the interaction of water with the acid sites and catalyst deactivation. We
demonstrated that the surface properties of acid functionalized silica catalysts such as
hydrophilicity and hydrophobicity affect LGO production.
2. Experimental
2.1. Catalysts preparation
Propyl sulfonic-functionalized mesoporous silica (PS-SBA-15) was synthesized
using the one-step synthesis procedure reported by Stucky and co-workers.23 In this
synthesis, 15 g of Pluronic P123 (Sigma-Aldrich) were dissolved in 480 mL of 2.0 M HCl
(Acros Organics, 37% solution in water) and 24 mL of DI water. The mixture was heated
3

Green Chemistry
Page 4 of 41
View Article Online
DOI: 10.1039/C9GC01526D
to 313 K and stirred for 3 h. Then 25 mL of tetraethyl orthosilicate (TEOS) (Acros
Organics, 98% purity) were added dropwise to the P123 solution and pre-hydrolyzed for
45 min at 313 K under stirring. Following the pre-hydrolysis of TEOS, 4.8 mL of (3-
mercaptopropyl)trimethoxysilane (MPTMS) (Alfa Aesar, 95% purity) and 7.3 mL of 30
wt% H2O2 (Fisher Chemical, 30% certified ACS) solution were added dropwise to the
suspension. The mixture was aged at 313 K for 20 h under stirring. The temperature of the
mixture was increased to 373 K and aged for an additional 20 h under stirring. The
precipitated solid was filtered and washed with DI water and dried overnight at room
temperature under vacuum. The copolymer template (P123) was removed by refluxing
with 4 L of ethanol for 24 h. Following the removal of the copolymer template, the solid
was filtered and washed with 2 L of fresh ethanol. Finally, the solid was dried for 24 h at
353 K under vacuum.
The non-functionalized SBA-15 was synthesized using the same procedure
described above for PS-SBA-15 without adding the MPTMS and H2O2. The aged
precipitated solid was washed with DI water and dried at 373 K overnight under vacuum.
The dried solid was calcined for 6 h at 823 K in air with a heating ramp of 3 K/min.
We used commercial acid-functionalized catalysts SiliaBond® from SiliCycle as
received. These catalysts are amorphous silica gels functionalized with propylsulfonic acid
(SiliaBond®-PSA) and arenesulfonic acid (SiliaBond®-Tosic-a).24 We used commercial
available non-endcapped (SiliaBond®-PSA, R51430B) and endcapped (SiliaBond®-PSA,
R51230B and SiliaBond®-Tosic-a, R60530B) samples. The non-endcapped catalysts have
acid functionality and residual silanol groups (surface hydroxyl groups that did not react
during the functionalization).24 The endcapped catalysts have acid functionality and
4

Page 5 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
trimethylsilyl groups instead of silanol groups.24 The synthesis procedure for these
catalysts consists of the functionalization of a fraction of the silanol groups with the acid
precursor followed by the functionalization of the residual silanol groups with
trimethylsilychloride to form trimethylsilyl groups.24 The endcapping reduces the acidity
and polarity on the silica surface and should make it more stable.24
2.2. Catalyst characterization
Nitrogen physisorption isotherms at 77 K were obtained using an accelerated
surface area and porosimetry system Micromeritics ASAP 2020 unit. The BET surface
areas were determined using adsorption data at relative pressures between 0.10 and 0.25.
The total pore volume was estimated from the amount of N2 adsorbed at a relative pressure
of 0.99. The micro and mesopore volumes were determined using the αs-plot method.25–27
The mesopore diameter was estimated from a pore size distribution (PSD) analysis
including the correction of the Kruk-Jaroniec-Sayari (KJS) for the statistical film thickness
in the Barret-Joyner-Halenda (BJH) method.28 The pore wall thickness was estimated by
calculating the difference between the lattice cell parameter and the pore diameter.
The powder X-ray diffraction studies were performed using a Bruker D8 Discover
X-Ray diffractometer equipped with cross beam optics and a Cu Kα target, operating at 50
kV and 1000 μA to study the ordered mesoporous structure of the samples. Powder
diffraction patterns using small angle mode were determined from 0 to 5o 2θ angles.
The acid loading of 0.10 g of propyl sulfonic acid functionalized SBA-15 was
determined by ion exchange with 20 mL of an aqueous solution of 2 M NaCl (Fisher
Chemical, certified ACS) for 2 h and titration with an aqueous solution of 0.01 M NaOH
(Acros Organics, 98.5% purity).
5

Green Chemistry
Page 6 of 41
View Article Online
DOI: 10.1039/C9GC01526D
The samples were chemically analyzed by inductively coupled plasma optical
emission spectroscopy (ICP-OES) using a Varian Vista-MPX ICP-OES to obtain the sulfur
loadings on the catalysts. The samples were pretreated by acid digestion at 393 K
overnight. In a typical experiment, around 15 mg of catalyst were digested with a mixture
of 1 mL of 37% hydrochloric acid solution (ACS reagent, Sigma-Aldrich), 1 mL of 70%
nitric acid solution (ACS reagent, Sigma-Aldrich) and 1 mL 48% hydrofluoric acid
solution (99.99% trace metal basis, Sigma-Aldrich).
2.3. Catalytic performance
The catalytic conversion of levoglucosan was studied using a 100 mL Parr
Instrument model 4565 high-pressure batch reactor. The reactor was heated using a 500 W
rigid heating mantle. The temperature and stirring were controlled using a Parr Instrument
model 4848 closed loop temperature controller. The temperature was measured using an
Omega JQSS-M30G-300 Type J thermocouple of 1/8 in diameter. The reactor was stirred
using a four-blade impeller with a 1/8 HP motor. The levoglucosan (CarboSynth, 98%
purity), catalysts and reactor were dried at 358 K overnight before the reactions. The LGA
solutions were prepared using 139, 349 and 867 mg of LGA to obtain concentrations of 14,
36 and 89 mM LGA in THF (Alfa Aesar, ACS 99% stabilized with BHT), respectively.
The solid acid catalyst loading was adjusted to 0.08 mmol H+ in 60 mL of THF. In a typical
experiment, the desired amount of LGA and solid catalyst, previously dried, were added to
the reactor followed by the addition of 60 mL of THF. Then, the reactor was closed and
purged with He (99.999%, Airgas) 5 times. The reactor was pressurized to 34 bar and
heated to the desired temperature and pressurized with He to 69 bar. The stirring rate was
maintained at 700 rpm. The zero time was set to the point when we started to heat the
6

Page 7 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
reactor. Samples of around 0.5 mL were taken periodically via dip tube and quenched in a
dry ice-bath and filtered with 0.22 μm syringe filter (Restek, PTFE
(polytetrafluoroethylene)). The reactor was quenched in an ice-water bath after reaction.
2.4. Thermal stability tests
The thermal stability of propyl sulfonic acid functionalized SBA-15 in THF was
investigated by exposing the catalyst sample to THF at 483 K and 69 bar of He for various
treatment times (40, 70 and 100 min). Then each sample was filtered and washed with THF
to remove the physisorbed species on the catalyst surface. Finally, the solid was dried at
353 K under vacuum overnight. The solid was then characterized using XRD, BET, ICP
and acid-base titration to determine any change in the structural and acid properties of the
material.
2.5. Analytical methods
The soluble products were analyzed using a high-performance liquid
chromatograph (Shimadzu LC-20AT) equipped with an UV (UV-vis SPD-20AV) and
refractive index (RID-10A) detectors. The reaction products were separated with a packed
carbohydrates analysis column (Bio-Rad Aminex HPX-87H) using 5 mM H2SO4 aqueous
solution as the mobile phase at a flow rate of 0.6 mL/min and a column temperature of 303
K. The LGA concentration was measured using a refractive index detector. HMF and
furfural were identified and quantified using their UV spectra at 280 nm. LGO was
identified and quantified using its UV spectrum at 370 nm. No other products were detected
using refractive index and UV detectors. The sample injection volume was 1 μL.
The conversion of LGA was calculated as follows:
7

Green Chemistry
Page 8 of 41
View Article Online
DOI: 10.1039/C9GC01526D
퐶퐿퐺퐴, 0 ― 퐶퐿퐺퐴
푥 100(%)
푋퐿퐺퐴(%) =
퐶퐿퐺퐴, 0
where
is the initial carbon concentration of LGA and
is the carbon
퐶퐿퐺퐴
퐶퐿퐺퐴, 0
concentration of LGA at each sampling time. The yield and selectivity corresponding to
each detectable product i were calculated as follows:
퐶푖
푌푖(%) =
푥 100(%)
퐶퐿퐺퐴, 0
퐶푖
퐶퐿퐺퐴, 0 ― 퐶퐿퐺퐴
푆푖(%) =
푥 100(%)
where 푖 is the carbon concentration of each detectable product i. The turnover frequency
퐶
(TOF) towards LGO based on the Brønsted acid concentrations (CH+) were estimated as
follows:
푑퐶푖
1
푇푂퐹푖 =
+
퐶퐻 푑푡
using the change in LGO concentration with time after the reaction reached 483 K at 40
min of reaction. The TOF of LGA consumption based on the Brønsted acid concentrations
were estimated as follows:
푑퐶퐿퐺퐴
1
푇푂퐹퐿퐺퐴
=
+
퐶퐻 푑푡
using the change in LGA concentration with time after reaching 483 K at 40 min of
reaction.
3. Results and discussion
3.1. Catalyst Characterization
Table 1 summarizes the physicochemical properties for the functionalized SBA-15
as well as for the commercial silica catalysts. The presence of Brønsted acid sites on PS-
8

Page 9 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
SBA-15 confirms the incorporation and the oxidation of MPTMS on the catalysts surface.
The N2 adsorption isotherms obtained for SBA-15 and PS-SBA-15 catalysts are both type
IV with an H1 hysteresis that correspond to average pore diameters inside the mesoporous
region. The functionalized PS-SBA-15 exhibits a lower average pore diameter compared
to that obtained for non-functionalized SBA-15. This difference may be attributed to the
incorporation of the organofunctional groups on the pore surface.29 The surface area of
functionalized PS-SBA-15 is lower than that obtained for non-functionalized SBA-15. The
wall thickness for the PS-SBA-15 is higher than for SBA-15. The difference in the surface
area and pore wall thickness is consistent with the presence of the propylsulfonic groups
on the catalyst pore surface. The external surface area for PS-SBA-15 is almost four times
higher than that obtained for SBA-15. It is possible that the particles of PS-SBA-15 are
smaller than the SBA-15 particles. The micropore volume for PS-SBA-15 is lower than
that obtained for SBA-15. The micropores channels may be blocked by the MPTMS groups
or by residual copolymer template P123. The copolymer template was removed by
calcination for the non-functionalized SBA-15. On the other hand, for PS-SBA-15, the
copolymer template was removed using a reflux of ethanol. The calcination treatment is
typically more effective in removing the copolymer template completely from micropores,
obtaining a less dense material and a free path through the micropores channels.
9

Green Chemistry
Page 10 of 41
Table 1. Physicochemical properties of solid acid catalystsa
b
c
c
d
d
CH+
SBET
SExt
Dp
(nm)
Wp
VP/Po=0.99
(cm3/g)
Vp
Vm
(mmol/g)
(m2/g)
(m2/g)
(nm)
3.7
5.6
9.3
9.4
8.6
(cm3/g)
(cm3/g)
Catalyst
SBA-15
-
855
734
279
420
238
6
26
9
7.9
5.8
7.8
7.2
7.2
0.05
0.63
0.43
0.58
0.28
0.03
0.01
PS-SBA-15
0.96
0.71
0.62
0.61
0.59
0.00
SiliaBond®-PSA non-endcapped
SiliaBond®-PSA endcapped
SiliaBond®-Tosic-a endcapped
0.42
0.00
15
0.57
0.00
4
0.28
0.00
aCH+ is the content of Brønsted acid sites on the catalyst surface per gram of catalyst, SBET is the BET surface area, SExt is the external
surface area, Dp is the mesopore diameter, WP is the pore wall thickness, VP/Po=0.99 is the total pore volume estimated from the amount
of N2 adsorbed at a relative pressure of 0.99, Vp is the mesopore volume, and Vm is the micropore volume.
bDetermined by titration.
cThe pore wall thickness was calculated using the distance between pores (a) calculated based on the d100 interplanar spacing and
assuming hexagonal geometry.
dCalculated using the αs-plot method using the standard parameters corresponding to mesoporous silica.
10

Page 11 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
The XRD patterns for non-functionalized and acid-functionalized silica are shown
in Figure 1. The XRD pattern corresponding to non-functionalized SBA-15 exhibits the
(100), (110) and (200) peaks that correspond to an ordered hexagonal array, p6mm.29,30
The acid functionalized PS-SBA-15 only shows the (100) peak at 2θ = 0.9 confirming the
presence of an ordered hexagonal array. The (100) peak for PS-SBA-15 is not as sharp as
that obtained for SBA-15. This may be attributed to the incorporation of organofunctional
groups on PS-SBA-15.30 The absence of the long-ordered reflection peaks (110) and (200)
for PS-SBA-15 may be attributed to a less ordered mesostructure resulting from the co-
condensation synthesis method.31,32
The physicochemical properties of SiliaBond catalyst are summarized in Table 1.
It has been reported that SiliaBond catalysts have a non-ordered porous structure.24 We
observed that the average pore diameter for the SiliaBond catalyst is between 7.2 and 7.8
nm. The N2 adsorption isotherms for these catalysts are of type IV, which is characteristic
of a mesoporous structure. The micropore volume for these catalysts was zero, confirming
the absence of pores with diameter ≤ 2 nm. Based on these results we conclude that the
porosity of these materials is mainly within the mesoporous range. The XRD patterns for
SiliaBond catalysts exhibit only one broad diffraction peak around 2θ = 0.8, as shown in
Figure 1. This peak may be attributed to the low order of the pores, which is consistent
with the manufacturer’s description.
[Figure 1 near here]
11

Green Chemistry
Page 12 of 41
View Article Online
DOI: 10.1039/C9GC01526D
3.2. Catalytic Performance
3.2.1. Homogeneous catalytic dehydration of LGA
Huber and co-workers studied the use of sulfuric acid for the conversion of
cellulose to LGO in polar aprotic solvents.5,8 They compared the catalytic performance of
sulfuric, phosphoric, formic, and hydrochloric acids for the conversion of cellulose to LGO
in THF.8 The highest yield for LGO was obtained when H2SO4 was used as the catalyst.
They also studied the conversion of cellulose to LGO using H2SO4 in different polar aprotic
solvents. The highest yield obtained for LGO was in THF; this was attributed to a lower
rate of LGO degradation in THF compared to the other polar aprotic solvents.8 We studied
the conversion of LGA to LGO using sulfuric acid to have a reference of the LGO yields
that can be obtained with the catalyst that showed a better performance for the production
of LGO from cellulose in THF reported in the literature. For comparison we also studied
the conversion of LGA to LGO using 1-propanesulfonic. 1-Propanesulfonic acid is a liquid
acid analogous to the functional groups present on PS-SBA-15. This comparison will help
to understand the effect of having those functional groups on the surface of a solid acid
catalyst.
Figure 2 shows the LGA conversion and products selectivity as a function of
reaction time obtained using homogeneous acid catalysts. The amount of catalyst was fixed
to have the same concentration of Brønsted acid in each experiment. When using sulfuric
acid as the catalyst, two regions with different reaction rates were observed. Initially, the
average TOF from 40 to 100 min was 0.7 ks-1. After 100 min of reaction the average TOF
decreases to 0.1 ks-1 and the maximum conversion obtained was 51.5%. For 1-
propanesulfonic acid, three regions with different reaction rates were observed. Initially, a
12

Page 13 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
slow LGA conversion rate was observed with an average TOF of 0.1 ks-1 from 40 to 50
min, followed by an increment in the TOF to 1.4 ks-1 from 50 to 90 min. At 90 min the
TOF decreases to 0.4 ks-1, this behavior is similar to that observed for sulfuric acid. The
maximum LGA conversion obtained using 1-propanesulfonic acid was 54.5% at 160 min.
These observations suggest that the acid proton may have been stabilized during the
reaction reducing its reactivity. This effect may be attributed to the formation of water
during the reaction. Mellmer et al., proposed that the lower reactivity of an acid proton in
water can be attributed to high proton solvation by water molecules.22 They also proposed
that the reactivity of an acid proton may be reduced due to a higher stabilization of the
proton in water than in a polar aprotic solvent.22 Hence, we speculate that the reactivity of
the sulfuric acid protons decreases as the water content increases during the reaction
reducing the rate of LGA conversion.
[Figure 2 near here]
The LGO selectivity using 1-propanesulfonic acid (33%) and sulfuric acid (33%)
were almost the same at a similar LGA conversion (around 49%). The HMF and furfural
selectivities were 1% and 0% for 1-propanesulfonic acid and 7% and 1% for sulfuric acid
at 49% conversion. The selectivity for non-detectable products at the same conversion was
65% and 59% for 1-propanesulfonic and sulfuric acids, respectively. For sulfuric acid the
production of non-detectable products increases as the LGO, HMF and furfural selectivities
decrease with time, suggesting that the main reaction routes for their production comes
primarily from the degradation of the detectable products. The average rate of LGA
consumption at 40 min, when the 483 K temperature set point was reached, is 7 times
higher for sulfuric acid than for 1-propanesulfonic acid. These results indicate that sulfuric
13

Green Chemistry
Page 14 of 41
View Article Online
DOI: 10.1039/C9GC01526D
acid is more active for the conversion of LGA compared to 1-propanesulfonic acid. The
higher production of HMF and furfural using sulfuric acid may also be attributed to its
higher activity (between 40 and 60 min) compared to 1-propanesulfonic acid and can
promote the conversion of LGO to HMF, furfural and non-detectable products. Sulfuric
acid is a stronger Brønsted acid catalyst (pKa = -3.0) compared to 1-propanesulfonic acid
(pKa = -1.5).22 The difference in the catalyst activity at the beginning of the reaction can
thus be attributed to a higher proton solvation in THF when sulfuric acid is used as catalyst
compared to 1-propanesulfonic acid. These observations are in agreement with the
conversion of sugars in polar aprotic solvents previously studied by Mellmer et al.22 It has
to be pointed out that this observation only applies to the initial reaction rates. For our
experimental conditions we calculate the initial reaction rates at 40 min of reaction because
that is the time when we reach the temperature set point. Before 40 min of reaction, the
temperature is increasing and a reliable estimation of the TOF cannot be obtained. Also, at
40 min of reaction the production of water is still low and the stabilization effect of water
on the acid proton is lower. After 50 min of reaction the TOF for 1-propanesulfonic acid is
higher compared to that obtained for sulfuric acid. This difference can be related to the
stabilization of sulfuric acid proton due to the production of water, as explained above.
From this point, we believe that the low production of HMF and furfural, and the higher
production of degradation products using 1-propanesulfonic acid compared to sulfuric acid
is related to its higher activity. The degradation of HMF and furfural to non-detectable
products using 1-propanesulfonic acid may be higher, showing low selectivities towards
HMF and furfural.
14

Page 15 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
3.2.2. LGA dehydration using acid-functionalized silica
The dehydration of LGA was studied using functionalized ordered porous SBA-15
and commercial functionalized non-ordered porous silica (SiliaBond). The conversion of
LGA using non-functionalized SBA-15 was around 2% after 190 min of reaction. LGO
was not produced using non-functionalized SBA-15. Trace amounts of HMF and furfural
were observed during the reaction, with selectivities of up to 14% and 3% for HMF and
furfural, respectively. In the absence of a catalyst, no LGA conversion was detected. It has
been suggested that the silanol groups on SBA-15 surface can act as acid sites with similar
acid properties to those of acetic acid.29 The low production of HMF and furfural may be
attributed to the presence of weak acidic hydroxyl groups on SBA-15.
Figure 3 shows the conversion and product selectivities obtained for LGA
dehydration for different LGA concentrations using PS-SBA-15. The reaction time to
obtain full conversion increased from 70 min for 14 mM LGA to 80 min for 36 mM LGA.
The maximum conversion achieved was 83% for 89 mM LGA feed at a reaction time of
160 min. The LGO selectivities, at similar LGA conversion, were obtained using initial
LGA concentrations of 14, 36 and 89 mM were 68% (XLGA=58%), 39% (XLGA=65%), and
37% (XLGA=66%), respectively. Increasing the concentration of LGA results in a decrease
in the LGO selectivity. After reaching the maximum LGA conversion, the LGO and HMF
selectivities start to decrease as shown in Figure 3a and 3b. However, the selectivity
towards furfural increases after the LGA was completely consumed, suggesting that the
furfural produced comes from LGO or HMF and it is promoted by the presence of the acid
catalyst. The average TOF for the consumption of LGA, when the reaction set point was
reached, rises as the initial LGA concentration increases, as shown in Figure 7. These
15

Green Chemistry
Page 16 of 41
View Article Online
DOI: 10.1039/C9GC01526D
results suggest that the apparent reaction order for the dehydration of LGA is near 1. The
selectivity towards LGO decreases as the TOF for LGA consumption increases, while the
selectivity towards non-detectable products increases. As the catalyst dehydration activity
increases the conversion of LGO to degradation products increases.
[Figure 3 near here]
LGA was not completely consumed when the feed was 89 mM LGA, with a
maximum LGO selectivity (37%) obtained before the LGA conversion reached a plateau.
After that time, the LGO selectivity decreased from 37% to approximately 33% and
remained constant until a new plateau at around 24% was observed. The decrease in LGO
selectivity with an initial LGA concentration of 89 mM (approximately 2% decrease) is
lower compared to the decrease in LGO selectivity using 14 and 36 mM LGA initial
concentrations (approximately 33% to 36%). We attribute this behavior to the deactivation
of the catalyst due to the production of water during the reaction. The rise in water
concentration can promote an interaction between water and the acid proton, thus
stabilizing it and reducing its catalytic activity, similar to the behavior observed for sulfuric
acid. Based on the LGA consumed during the reaction, we estimated the amount of water
produced during the reaction, as shown in Figure 4a. Contrary to the selectivity, the amount
of LGO produced is higher as the initial concentration of LGA increases. Hence, the
production of water increases as the initial LGA concentration increases. These results
suggest that the catalyst is deactivating due to a higher production of water when an initial
concentration of 89 mM LGA is used. The results also suggest that the consumption of
LGO starting with 14 and 36 mM LGA may be attributed to the presence of the remaining
active proton sites on the catalyst once the LGA is completely consumed. From these
16

Page 17 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
observations, we also conclude that the LGO degradation in the presence of water and THF
at 483 K is low in the absence of active acid species.
[Figure 4 near here]
The conversion and selectivities obtained for the dehydration of LGA using
commercial non-endcapped SiliaBond®-PSA are shown in Figure 5. The maximum LGO
selectivity obtained with non-endcapped SiliaBond®-PSA was 52%. The TOF for LGA
consumption and the TOF for LGO were 2.7 ks-1 and 1.2 ks-1, respectively. The TOF for
LGA consumption obtained using non-endcapped SiliaBond®-PSA (2.7 ks-1) is similar to
that obtained for PS-SBA-15 (2.3 ks-1). This result suggests that the structural properties
of the silica support do not have a significant effect on the acid site activity. The selectivity
towards LGO obtained using non-endcapped SiliaBond®-PSA was 49% at 100%
conversion. This selectivity is lower compared to that obtained using PS-SBA-15 (59%) at
100% conversion. While the selectivity towards non-detectable products at 100%
conversion is higher (44%) compared to that obtained for PS-SBA-15 (31%). Also, after
reaching full LGA conversion, the LGO selectivity decreases from 49% to 21%, while
when using PS-SBA-15 the LGO selectivity decreases from 59% to 34%. The degradation
of LGO using non-endcapped SiliaBond®-PSA is higher compared to that for PS-SBA-15.
This behavior may be related to the lower concentration of hydroxyl groups on non-
endcapped SiliaBond®-PSA (3 mmol OH/g) surface compared to on PS-SBA-15 (8 mmol
OH/g). The hydroxyl groups are hydrophilic species that may attract the water produced
during the reaction to the catalyst surface. Hence, increasing the hydroxyl groups
concentration can lead to the interaction of more water molecules with the catalyst surface
reducing the interaction with LGO and reducing its degradation. Additionally, we note that
17

Green Chemistry
Page 18 of 41
View Article Online
DOI: 10.1039/C9GC01526D
the production of furfural increases after LGA was completely consumed while the LGO
concentration decreases using non-encapped SiliaBond®-PSA, as observed using PS-SBA-
15, confirming that a considerable fraction of the furfural produced comes from LGO.
[Figure 5 near here]
The results obtained for the dehydration of 14 mM LGA using endcapped
SiliaBond®-PSA are shown in Figure 6a. The maximum selectivity obtained for LGO was
59%. This selectivity is higher than that obtained for non-endcapped SiliaBond®-PSA
(49%). The average TOF at 40 min for the consumption of LGA (2.7 ks-1) is the same
compared to that obtained using the non-endcapped catalysts and similar to that for PS-
SBA-15. These results suggest that the trimethylsilyl functionalization does not have a
significant effect on the acid site activity. However, the incorporation of trimethylsilyl
groups on the catalysts surface reduces the presence of hydroxyl groups on the catalyst
surface. The silyl ether groups may reduce the interaction of water with the catalyst surface
due to an increase in the hydrophobicity on the catalyst surface.34 These properties may
help to reduce the interaction of water with the acid sites and LGO reducing its degradation
and improving its selectivity.
[Figure 6 near here]
The results obtained for LGA dehydration using different LGA initial
concentrations with the endcapped SiliaBond®-PSA are shown in Figures 6a, 6b and 6c.
The selectivity for LGO decreases with an increase in LGA initial concentration. The
decrease in the LGO selectivity with the increase in LGA initial concentration is higher for
endcapped SiliaBond®-PSA compared with those obtained using PS-SBA-15. Full LGA
conversion was achieved starting with 14 and 36 mM LGA, while with 89 mM LGA the
18

Page 19 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
maximum conversion reached was 58%. The LGA conversion starting with 89 mM LGA
reaches a plateau at 70 min. After 70 min of reaction the catalyst shows no activity for
LGA consumption, the average TOF after 70 min is 0.0 ks-1. While PS-SBA-15 shows a
TOF of 0.0 ks-1 after 130 min at a conversion of 82% starting with 89 mM LGA. A
comparison of Figures 4a and 4b shows that the production of water is almost the same
using endcapped SiliaBond®-PSA or PS-SBA-15 when the initial concentration of LGA is
14 or 36 mM. However, the production of water is 34% lower for endcapped SiliaBond®-
PSA starting with 89 mM LGA. This difference suggests that the deactivation of the
endcapped SiliaBond®-PSA is faster compared to PS-SBA-15 and less water is required
for its complete deactivation. As mentioned above, the presence of silyl ether groups makes
the endcapped SiliaBond®-PSA hydrophobic. PS-SBA-15 has a surface with a large
concentration of hydroxyl groups that makes it hydrophilic. Hence, the difference in the
deactivation of endcapped SiliaBond®-PSA and PS-SBA-15 starting with 89 mM LGA
may be attributed to the presence of hydroxyl groups on the SBA-15 surface. The hydroxyl
groups promote an interaction of water with the catalyst surface reducing its interplay with
the acid sites that delay the catalyst deactivation.
Figure 7 shows the effect of the initial LGA concentration on the TOF for the
consumption of LGA and the TOF for LGO production using PS-SBA-15 and endcapped
SiliaBond®-PSA as catalysts. The TOFs for LGA and LGO both increase as the LGA
concentration increases. Increasing the LGA concentration increases the amount of LGO
and water produced during the reaction. We speculate that the difference in the TOFs for
LGA and LGO at high LGA concentrations between PS-SBA-15 and endcapped
SiliaBond®-PSA can be attributed to an increase in the water concentration promoting an
19

Green Chemistry
Page 20 of 41
View Article Online
DOI: 10.1039/C9GC01526D
interaction of water with the acid sites on the endcapped SiliaBond®-PSA. The presence of
silanol groups on PS-SBA-15 may promote an interaction between water and the catalyst
surface, reducing the interaction of water with the acid site at high local water
concentrations. On the endcapped SiliaBond®-PSA surface, the absence of the silanol
groups reduces the interaction of water with the catalyst surface, but the interaction of water
with the acid sites may be higher at high local water concentrations.
[Figure 7 near here]
The effect of the sulfonic group linker was also evaluated comparing the catalytic
performance of two endcapped SiliaBond® functionalized catalysts, one with
propylsulfonic acid (SiliaBond-PSA) and the other with arenesulfonic acid (SiliaBond®-
Tosic-a) as shown in Figures 8 and 6a. The LGO selectivity at 100% LGA conversion
obtained using endcapped SiliaBond®-Tosic-a was lower (52%) compared to the LGO
selectivity (59%) with endcapped SiliaBond®-PSA. The selectivity for non-detectable
products at 100% LGA conversion obtained with the endcapped SiliaBond®-Tosic-a was
42%, and this selectivity was 34% using endcapped SiliaBond®-PSA. The average TOF
for the consumption of LGA was 3.27 ks-1 for the endcapped SiliaBond®-Tosic-a and equal
to 2.7 ks-1 for the endcapped SiliaBond®-PSA. These results indicate that the activity for
LGA conversion of the endcapped SiliaBond®-Tosic-a is higher compared to that for the
endcapped SiliaBond®-PSA. The difference in the LGO and non-detectable products
selectivities can be attributed to the difference in catalyst’s activity. The endcapped
SiliaBond®-Tosic-a has a higher activity and promotes the formation of non-detectable
products affecting the selectivity towards LGO. The higher activity for LGA consumption
observed for SiliaBond®-Tosic-a can be attributed to a higher pKa of the arenesulfonic acid
20

Page 21 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
functional group compared to propylsulfonic acid. The terminal sulfonic acid group on
SiliaBond®-Tosic-a is similar to benzenesulfonic acid which has a pKa of -6.5, while the
terminal sulfonic acid group on SiliaBond®-PSA is similar to methanesulfonic acid which
has a pKa of -2.35 Based on those acid strengths the acid sites on SiliaBond®-Tosic-a should
be stronger than those on SiliaBond®-PSA endcapped, and as a consequence more active.
However, this higher acid strength may also promote the degradation of LGO.
[Figure 8 near here]
In summary, the absence of products using non-functionalized SBA-15 and the
preferential production of LGO using 1-propanesulfonic acid shows that the propylsulfonic
acid on the functionalized PS-SBA-15 catalyzes the conversion of LGA to LGO.
Propylsulfonic functionalized silica catalysts have twice the rate of LGA conversion
compared to sulfuric acid. The dehydration activity of the catalysts affects the selectivity
towards LGO; catalysts with high activity promote the production of non-detectable
products. An increase in the initial concentration of LGA causes a reduction in the LGO
selectivity. However, the TOF for the consumption of LGA and the TOF for LGO
production increase with an increment in the LGA initial concentration. This behavior is
higher using PS-SBA-15 compared to the commercial endcapped SiliaBond®-PSA but both
catalysts seem to deactivate at some point when the reaction starts with 89 mM LGA. The
deactivation of the catalysts is related to the production of water during the reaction. The
deactivation of endcapped SiliaBond®-PSA is faster and requires less water compared to
PS-SBA-15. This difference suggests that the catalyst surface hydrophilicity reduces the
interaction of water with the acid sites retarding the catalyst deactivation.
21

Green Chemistry
Page 22 of 41
View Article Online
DOI: 10.1039/C9GC01526D
3.3. Thermal stability of propylsulfonic functionalized silica in THF
It has been reported that acid functionalized silica catalysts are not thermally stable
in the presence of water.29,36 It has also been demonstrated that the incorporation of
hydrophobic species on silica surfaces helps to improve its stability and reduce the
physicochemical changes in water environments.34 To measure the stability of
propylsulfonic acid functionalized silica catalysts in THF, the PS-SBA-15 and SiliaBond®-
PSA catalysts were exposed to a series of thermal treatments in THF at 483 K and 69 bar
He. Table 2 shows a summary of the physicochemical properties obtained for PS-SBA-15
and SiliaBond® catalysts after thermal treatments in THF at 483 K. A decrease in the acid
sites concentration was observed when the samples were treated for 40 and 70 min. When
the samples were treated for 100 min the acid concentration was almost the same to that
obtained at 70 min. The acid concentration on PS-SBA-15 and SiliaBond®-PSA catalysts
decreases around 15% for PS-SBA-15 and 10% for the non-endcapped and endcapped
SiliaBond®-PSA catalysts. This loss can be attributed to acid site leaching during the
thermal treatment in THF. The non-endcapped and the endcapped SiliaBond®-PSA
catalysts have almost the same loss in acidity. This result suggests that the hydrophobic
functional groups on the endcapped catalysts do not protect the acid from leaching in THF.
22

Page 23 of 41
Green Chemistry
Table 2. Physicochemical properties of solid acid catalysts after treatment in THF at 483 K.a
d
e
e
e
Timeb S:Sic
(min)
CH+
SBET
SExt
Dp
(nm)
VP/Po=0.99
Vp
Vm
Catalyst
(mmol/g)
(m2/g)
(m2/g)
26
48
24
13
15
6
(cm3/g) (cm3/g) (cm3/g)
PS-SBA-15
0
40
70
100
0
0.06
-
0.96
734
1,439
621
362
420
223
167
342
5.8
5.8
5.8
5.8
7.2
7.5
7.3
7.2
0.631
1.261
0.551
0.321
0.581
0.318
0.250
0.499
0.587
1.170
0.511
0.299
0.573
0.323
0.253
0.509
0.001
0.013
0.004
0.001
0.000
0.000
0.000
0.000
PS-SBA-15
0.87
PS-SBA-15
0.07 0.80 (0.82)f
PS-SBA-15
0.07
0.02
-
0.84
0.62
0.56
0.55
0.56
SiliaBond®-PSA endcapped
SiliaBond®-PSA endcapped
SiliaBond®-PSA endcapped
SiliaBond®-PSA endcapped
40
70
100
-
5
0.02
6
SiliaBond®-PSA non-
endcapped
0
0.03
0.71
279
9
7.8
0.427
0.425
0.000
SiliaBond®-PSA non-
endcapped
70
0.03
0.64
338
5
5.8
0.535
0.535
0.000
aCH+ is the content of Brønsted acid sites on the catalyst surface per gram of catalyst, SBET is the BET surface area, SExt is the external
surface area, Dp is the mesopore diameter, VP/Po=0.99 is the total pore volume estimated from the amount of N2 adsorbed at a relative
pressure of 0.99, Vp is the mesopore volume, and Vm is the micropore volume.
bTime for which the catalyst was exposed to 483 K in THF at 69 bar He.
cSulfur to silicon molar ratio obtained from the elemental composition determined by ICP.
dDetermined by titration.
eCalculated using αs-plot method using the standard parameters corresponding to mesoporous silica.
fThe second value corresponds to the acidity determined for a sample treated removing 25 mL of the liquid phase during the experiment.
23

Green Chemistry
Page 24 of 41
View Article Online
DOI: 10.1039/C9GC01526D
Figures 9 and 10 show the XRD patterns and the physisorption isotherms obtained
for propylsulfonic acid functionalized silica catalysts after treatment in THF at 483 K. The
physicochemical properties of the catalysts change with its exposure to THF at 483 K. For
PS-SBA-15, the BET and external surface area and porosity increase after 40 min of
treatment in THF, as shown in Table 2. Afterwards, the BET and external surface area and
porosity decrease with an increase in the treatment time. However, the average pore
diameter remains constant during all treatments. The ratio of sulfur to silicon atoms
increases after 70 min treatment and remains constant at 100 min treatment. These results
suggest that the increase in surface area and porosity are related to a loss in silica during
the thermal treatment in THF at 483 K. These physicochemical changes do not cause
significant changes to the mesostructure order, as shown in Figures 9a and 10a. The XRD
pattern and the type of hysteresis obtained for the treated catalyst remains almost the same
to those obtained for the untreated PS-SBA-15.
[Figure 9 near here]
[Figure 10 near here]
The surface areas, average mesopore diameter and volume vary with the treatment
time for the endcapped and non-endcapped SiliaBond®-PSA catalysts. However, the ratio
of sulfur to silicon remains constant in contrast to PS-SBA-15. Apparently, the silicon of
the SiliaBond® catalysts is not dissolving in the THF. The XRD pattern and the type of
hysteresis obtained for the treated SiliaBond®-PSA catalysts remains almost the same to
those obtained for the untreated SiliaBond®-PSA. These results suggest that these
physicochemical changes do not cause significant changes to the partial order of the pores,
as observed in PS-SBA-15.
24

Page 25 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
PS-SBA-15 was also treated in THF for 70 min at 483 K partially removing the
solvent during the treatment. The acid concentration of this catalyst was measured after
each treatment. This experiment was performed to corroborate that the functional groups
remain on the catalyst surface during the reaction and were not removed at the reaction
temperature and readsorbed on the catalyst surface during the ice bath quench. The total
THF removed during the treatment was 25.5 mL. The acid concentration after this
treatment was 0.82 mmol H+/g, almost the same obtained for the catalyst treated without
removing the THF during treatment (0.80 mmol H+/g). This result confirms that the
functional groups remain on the catalyst surface during reaction and the functional group
is not present in the solvent and acting as a homogeneous catalyst.
4. Conclusions
The dehydration of LGA to LGO was studied using Brønsted solid acid catalysts in
tetrahydrofuran (THF). LGO selectivities of up to 68% are obtained using propylsulfonic
acid functionalized catalysts. The production of water during the reaction affects the
catalytic behavior of the acid group due to a stabilization of the acid proton. Using a solid
acid catalyst with hydroxyl groups on the surface such as PS-SBA-15 reduces this effect,
allowing the achievement of higher LGA conversion and LGO selectivity. The structural
properties and surface hydrophobicity and hydrophilicity do not have an effect on the acid
site activity. The acidity of the PS-SBA-15 and SiliaBond® catalysts in THF at the reaction
temperature appears to be stable. The PS-SBA-15 physicochemical changes can be mainly
attributed to the loss of Si during the temperature treatment in THF. This study shows that
the use of Brønsted solid acid catalysts improves the production of LGO from LGA and
25

Green Chemistry
Page 26 of 41
View Article Online
DOI: 10.1039/C9GC01526D
provides insights regarding the catalytic properties needed for the selective production of
LGO from LGA.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work was supported through the NSF CREST: Nanotechnology Center for
Biomedical, Environment and Sustainability Applications - Phase II (Award HRD-
1345156) and the DOE EERE: Office of Energy Efficiency and Renewable Energy (Award
DE-EE0006878). We wish to also acknowledge partial support from the NSF funded RII
Track-2 FEC: Center for a Sustainable Water, Energy, and Food Nexus (Award OIA-
1632824).
References
1
2
D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem., 2010, 12, 1493.
C.-H. Zhou, X. Xia, C.-X. Lin, D.-S. Tong and J. Beltramini, Chem. Soc. Rev.,
2011, 40, 5588.
3
S. G. Wettstein, D. M. Alonso, E. I. Gürbüz and J. A. Dumesic, Curr. Opin.
Chem. Eng., 2012, 1, 218–224.
4
5
M. J. Climent, A. Corma and S. Iborra, Green Chem., 2014, 16, 516.
F. Cao, T. J. Schwartz, D. J. McClelland, S. H. Krishna, J. A. Dumesic and G. W.
Huber, Energy Env. Sci, 2015, 8, 1808–1815.
6
7
Biomass Explained, http://www.eia.gov/energyexplained/?page=biomass_home.
O. Oyola-Rivera, A. M. González-Rosario and N. Cardona-Martínez, Biomass
Bioenergy, 2018, 119, 284–292.
8
9
J. He, M. Liu, K. Huang, T. W. Walker, C. T. Maravelias, J. A. Dumesic and G.
W. Huber, Green Chem, 2017, 19, 3642–3653.
A. M. Sarotti, R. A. Spanevello and A. G. Suárez, Green Chem., 2007, 9, 1137.
26

Page 27 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
10
11
12
13
Z. Wang, Q. Lu, X.-F. Zhu and Y. Zhang, ChemSusChem, 2011, 4, 79–84.
S. Kudo, Z. Zhou, K. Norinaga and J. Hayashi, Green Chem., 2011, 13, 3306.
M. Ostermeier and R. Schobert, J. Org. Chem., 2014, 79, 4038–4042.
C. Müller, M. A. Gomez-Zurita Frau, D. Ballinari, S. Colombo, A. Bitto, E.
Martegani, C. Airoldi, A. S. vanꢀNeuren, M. Stein, J. Weiser, C. Battistini and F.
Peri, ChemMedChem, 2009, 4, 524–528.
14
M. De bruyn, J. Fan, V. L. Budarin, D. J. Macquarrie, L. D. Gomez, R. Simister,
T. J. Farmer, W. D. Raverty, S. J. McQueen-Mason and J. H. Clark, Energy
Environ. Sci., 2016, 9, 2571–2574.
15
16
17
18
A. Broido, M. Evett and C. C. Hodges, Carbohydr. Res., 1975, 44, 267–274.
Y. Halpern, R. Riffer and A. Broido, J. Org. Chem., 1973, 38, 204–209.
F. Shafizadeh, R. H. Furneaux and T. T. Stevenson, Carbohydr. Res., 1979, 71,
169–191.
C. M. Lakshmanan and H. E. Hoelscher, Ind. Eng. Chem. Prod. Res. Dev., 1970,
9, 57–59.
19
20
H. Kawamoto, S. Saito and S. Saka, J. Anal. Appl. Pyrolysis, 2008, 82, 78–82.
J. Zandersons, A. Zhurinsh, G. Dobele, V. Jurkjane, J. Rizhikovs, B. Spince and
A. Pazhe, J. Anal. Appl. Pyrolysis, 2013, 103, 222–226.
21
22
23
24
D. Carpenter, T. L. Westover, S. Czernik and W. Jablonski, Green Chem, 2014,
16, 384–406.
M. A. Mellmer, C. Sener, J. M. R. Gallo, J. S. Luterbacher, D. M. Alonso and J.
A. Dumesic, Angew. Chem. Int. Ed., 2014, 53, 11872–11875.
D. Margolese, J. A. Malero, S. C. Christiansen, B. F. Chmelka and G. D. Stucky,
Chem. Mater., 2000, 12, 2448–2459.
V. Pandarus, G. Gingras, F. Béland, R. Ciriminna and M. Pagliaro, Catal. Sci.
Technol., 2011, 1, 1600.
25
26
27
A. Sayari, P. Liu, M. Kruk and M. Jaroniec, Chem. Mater., 1997, 9, 2499–2506.
M. Kruk, M. Jaroniec and A. Sayari, Chem. Mater., 1999, 11, 492–500.
M. Kruk, M. Jaroniec, C. H. Ko and R. Ryoo, Chem. Mater., 2000, 12, 1961–
1968.
27

Green Chemistry
Page 28 of 41
View Article Online
DOI: 10.1039/C9GC01526D
28
29
M. Jaroniec and L. A. Solovyov, Langmuir, 2006, 22, 6757–6760.
D. Reyes-Luyanda, J. Flores-Cruz, P. J. Morales-Pérez, L. G. Encarnación-
Gómez, F. Shi, P. M. Voyles and N. Cardona-Martínez, Top. Catal., 2012, 55,
148–161.
30
31
S. Ravi, R. Roshan, J. Tharun, A. C. Kathalikkattil and D. W. Park, J. CO2 Util.,
2015, 10, 88–94.
S. Hamoudi and S. Kaliaguine, Microporous Mesoporous Mater., 2003, 59, 195–
204.
32
33
R. B. Babou Kammoe and S. Hamoudi, J. Environ. Qual., 2014, 43, 1032.
K. Leung, I. M. B. Nielsen and L. J. Criscenti, J. Am. Chem. Soc., 2009, 131,
18358–18365.
34
35
G. Morales, G. Athens, B. Chmelka, R. Vangrieken and J. Melero, J. Catal.,
2008, 254, 205–217.
E. P. Serjeant, B. Dempsey and International Union of Pure and Applied
Chemistry, Eds., Ionization constants of organic acids in aqueous solutions,
Pergamon Press, Oxfordꢁ; New York, 1979.
36
M. H. Tucker, A. J. Crisci, B. N. Wigington, N. Phadke, R. Alamillo, J. Zhang, S.
L. Scott and J. A. Dumesic, ACS Catal., 2012, 2, 1865–1876.
28

Page 29 of 41
Green Chemistry
View Article Online
DOI: 10.1039/C9GC01526D
List of Figures
Figure 1. Powder XRD patterns for silica catalysts functionalized with sulfonic acid
groups.
Figure 2. Dehydration of 14 mM LGA using homogeneous acid catalysts: (a) H2SO4 and
(b) 1-propanesulfonic acid in THF at 483 K and 69 bar He with 0.08 mmol H+.
Figure 3. Effect of LGA concentration on the dehydration of LGA using PS-SBA-15
catalyst in THF at 483 K and 69 bar He and LGA concentrations of (a) 14 mM, (b) 36 mM,
and (c) 89 mM. The acid loading was kept to 0.08 mmol H+.
Figure 4. Water production during the dehydration of LGA for different LGA initial
concentrations in THF at 483 K and 69 bar He and catalysts: (a) PS-SBA-15 and (b)
SiliaBond-PSA endcapped.
Figure 5. Dehydration of 14 mM LGA using non-endcapped SiliaBond®-PSA catalyst in
THF at 483 K and 69 bar He with an acid loading of 0.08 mmol H+.
Figure 6. Effect of LGA concentration on the dehydration of LGA using endcapped
SiliaBond®-PSA catalyst in THF at 483 K and 69 bar He. The acid loading was kept at
0.08 mmol H+. The LGA concentrations were: (a) 14 mM, (b) 36 mM and (c) 89 mM.
29