Tin Grafted on Modified Alumina-Catalyzed Isomerisation of Glucose to Fructose

Article abstract of DOI:10.1016/j.apcata.2019.05.028

The present study focuses on designing a catalyst based on hot water treated alumina (Al₂O₃-HWT) for the conversion of glucose to fructose. The glucose isomerisation reactions are performed with tin incorporated on parent Al₂O₃ and Al₂O₃-HWT in methanol. 0.5 wt% Sn/Al₂O₃-HWT affords a combined yield of fructose and methylfructoside (30.4%) which is two-fold higher than that obtained with 0.5wt% Sn/Al₂O₃ (15.1%), implying the importance of hot water treatment of Al₂O₃. Al₂O₃-HWT shows a very broad peak centred around 3440 cm^-1, which could be assigned to OH stretching band of gibbsite, γ-Al(OH)₃ which significantly diminished after solid state ion-exchange with SnCl₄.5H₂O (0.5 wt% Sn/Al₂O₃-HWT). UV-Vis diffused reflectance spectrum of 0.5 wt% Sn/Al₂O₃-HWT displays a peak centered at 241 nm, which can be ascribed to the incorporation of tin into the alumina network. XRD patterns of 0.5, 3 and 5 wt% Sn/Al₂O₃-HWT show that no peak corresponding to SnO₂ is formed. Importantly, 0.5wt% SnO₂/Al₂O₃-HWT exhibits a low activity, giving 13.2% of the total yield of fructose and methylfructoside, respectively, compared to 0.5wt% Sn/Al₂O₃-HWT (30.4% fructose), signifying the role of incorporated tin into the alumina network.

Full text of DOI:10.1016/j.apcata.2019.05.028

Accepted Manuscript
Title: Tin Grafted on Modified Alumina-Catalyzed
Isomerisation of Glucose to Fructose
Authors: Muhamad Aadil Yatoo, Shunmugavel
Saravanamurugan
PII:
S0926-860X(19)30241-8
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2019.05.028
APCATA 17094
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
8 April 2019
8 May 2019
25 May 2019
Please cite this article as: Yatoo MA, Saravanamurugan S, Tin Grafted on Modified
Alumina-Catalyzed Isomerisation of Glucose to Fructose, Applied Catalysis A, General
(2019), https://doi.org/10.1016/j.apcata.2019.05.028
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Tin Grafted on Modified Alumina-Catalyzed Isomerisation of
Glucose to Fructose
Muhamad Aadil Yatoo and Shunmugavel Saravanamurugan*
Laboratory of Bioproduct Chemistry, Center of Innovative and Applied Bioprocessing, Sector
81(Knowledge City), Mohali-140306, Punjab, India.
E-mail: saravana@ciab.res.in
Abstract: The present study focuses on designing a catalyst based on hot water treated alumina
(Al2O3-HWT) for the conversion of glucose to fructose. The glucose isomerisation reactions are
performed with tin incorporated on parent Al2O3 and Al2O3-HWT in methanol. 0.5 wt% Sn/Al2O3-
HWT affords a combined yield of fructose and methylfructoside (30.4%) which is two-fold higher
than that obtained with 0.5wt% Sn/Al2O3 (15.1%), implying the importance of hot water treatment
of Al2O3. Al2O3-HWT shows a very broad peak centred around 3440 cm-1, which could be assigned
to OH stretching band of gibbsite, -Al(OH)3 which significantly diminished after solid state ion-
exchange with SnCl4.5H2O (0.5 wt% Sn/Al2O3-HWT). UV-Vis diffused reflectance spectrum of
5 wt% Sn/Al2O3-HWT displays a peak centered at 240 nm, which can be ascribed to the
incorporation of tin into the alumina network. XRD patterns of 0.5, 3 and 5 wt% Sn/Al2O3-HWT
show that no peak corresponding to SnO2 is formed. Importantly, 0.5wt% SnO2/Al2O3-HWT
exhibits a low activity, giving 13.2% of the total yield of fructose and methylfructoside,
respectively, compared to 0.5wt% Sn/Al2O3-HWT (30.4% fructose), signifying the role of
incorporated tin into the alumina network.
Keywords: glucose • isomerisation • hot water treatment • tin •alumina • alcohol
1

Highlights
 Alumina is treated under hot water to enhance the presence of hydroxyl groups.
 Tin is incorporated on hot water treated alumina (Sn/Al2O3-HWT) as evidenced by UV-
Vis analysis.
 Sn/Al2O3-HWT is employed as a catalyst for the isomerisation of glucose to fructose
(30.4%) in methanol.
 Sn/Al2O3-HWT shows two-fold higher catalytic activity than Sn/Al2O3-parent.
Graphical Abstract
1. Introduction
Glucose is one of the essential carbohydrates and abundantly available in nature in the form of
cellulose biopolymer in terrestrial lignocellulosic biomass. Obtaining monomeric glucose from
cellulose is highly valuable as it serves as a potential platform chemical to produce value-added
chemicals and fuels [1-4]. However, selective conversion of cellulose to monomeric glucose
followed by transforming into useful compounds is generally challenging [5-8]. On the other hand,
2

glucose as a starting substrate to yield a broad spectrum of chemicals including fructose via
isomerisation has been considerably studied in the past decades to produce targeted chemicals
selectively [9-12]. Isomerisation of glucose to fructose in aqueous medium maintains a
thermodynamic equilibrium (Keq= 1 at 298 K) [4,13] and is slightly endothermic in nature (∆H=3
kJ/mol)[4], making the cost-effective conversion of glucose to fructose challenging. Fructose is
the sweetest natural sugar with the lowest glycemic index [14] and has received much attention in
the food industry for the production of high fructose corn syrup (HFCS or HFS) or glucose-fructose
syrup (GFS) with varying glucose to fructose ratios [13,15,16]. Currently, HFS is being produced
through glucose isomerisation via an enzymatic approach which is one of the largest biocatalytic
processes using immobilised D-glucose/xylose isomerase (GI, EC 5.3.1.5) at 60 C, yielding a
product distribution of 50% glucose, 42% fructose and 8% other monosaccharides. This process,
however, has a lot of limitations due to the inactivation of GI at temperatures above 60 C,
maintaining an optimal pH with buffered solutions, inhibition of GI by Ca2+ present in the
feedstock, frequent replacement of catalyst due to its irreversible deactivation over time, and hence
increasing the cost of the process [4,17].
Production of fructose via isomerisation of glucose is not only useful process for making HFS
but also as a versatile platform chemical which can be subsequently converted into high-value
chemicals, such as 5-hydroxymethylfurfural (HMF), 2,5-furandicarboxylic acid (FDCA) and
levulinic acid [14-16]. To overcome the drawbacks of the enzymatic approach, an ample number
of reports for glucose isomerisation over solid acid and base catalysts have been developed [10,
14, 18-20]. In base catalysis, heterogeneous bases like hydrotalcite (HT), metallosilicates and
alkali/alkaline earth metal titanates were found to give a moderate yield of fructose [21].
Alkali/alkaline earth metal-containing titanates (SrTiO3) yielded 34% of fructose along with 10%
mannose at 60% conversion of glucose at 110 C after 1 h [21]. Magnesium impregnated NaY
zeolites gave a fructose yield of 23-32% in an aqueous medium at 100 C after 2 h but showed a
poor catalytic performance as cation leaching and coke retention led to pore blockage and loss of
crystallinity of material [22]. However, desilication of the same catalyst led to an increase of the
fructose yield to a maximum of 35%, but still, this process entailed leaching of Na and Mg, which
partly contributed homogenously to the conversion of glucose to some extent [23]. On the other
hand, in acid catalysis, Lewis acid-containing tanosilicates, especially Sn-beta, exhibited exciting
activity towards glucose isomerisation in aqueous conditions, yielding 32% fructose and 9%
3

mannose from a 10wt% glucose solution at 110 C after 30 min [17]. We reported that Lewis and
Bronsted acid-containing zeolites, such as H-USY-6, exhibited an excellent activity in terms of
fructose yield (55%) in the methanol-water solvent system at 120 C in a one-pot two-step
approach [4]. In a similar approach, Ren et al. reported that Sn-containing zeolite (Sn-SPP) showed
unusual activity in terms of fructose yield (65%) in the ethanol-water solvent mixture at 90 C in
a two-step reaction protocol [24].
Other metals than Sn in zeolite has also been reported for the isomerisation of glucose. Recently,
Haian Xia et al. replaced Sn with Fe in beta zeolite and applied it for glucose isomerisation, and
the obtained fructose yield was only 20% at 130 C after 3 h [25]. Recently, metal-organic
framework (MOF) based materials have also been reported for the isomerisation of glucose in
alcohol and/or water systems and were found to yield 20 to 60% [26, 27]. Sun et al. reported a
one-pot two-step process for the conversion of glucose to fructose in ethanol via oxidation of
glucose to D-glucosone and followed by hydrogenation to form fructose (77%) [28]. These
reported catalyst systems encounter with significant drawbacks such as longtime synthesis of
materials, expensive materials, deactivation and metal leaching. In this context, we set out to search
for cheap materials which would offer a moderate to good yield of fructose from glucose.
The focus of the present study is to employ alumina-based catalysts for the interconversion of
glucose to fructose in methanol medium as shown in Scheme 1. It is known that -alumina gets
hydrated to form boehmite and gibbsite during hot water treatment resulting in changing surface
area and acidity consequently influencing the catalytic activity of the material [29]. It has been
reported that the hydrothermal treatment of alumina could improve the total number of weak acid
sites by 15% compared to untreated -alumina [30]. Also, kaolin, an alumina-based material, was
recently reported as an efficient catalyst for the production of methyl fructoside (up to 52%) from
glucose via isomerisation and ketalisation [31]. These pioneer studies on alumina intrigued and
prompted us to design a catalyst system based on hot water treated alumina as support material for
Lewis acid metal(s), such as Sn, Zr and Ti. With regard to this, Lewis acid metals were
incorporated on pristine and hot water treated alumina, characterized and compared with respect
to their catalytic activity towards glucose isomerisation to fructose in alcohol. The reaction
parameters, such as time, temperature, concentration and metal loading were also optimised for
improving the yield of fructose.
4

Scheme 1. The isomerisation glucose to fructose with Sn/Al2O3-HWT
2. Experimental Section
2.1 Catalyst synthesis
In a round bottom flask, an appropriate amount of aluminium oxide active (neutral 70-230
mesh) was treated at 110 C in water for 4 h in a preheated oil bath. After cooling down to room
temperature, the slurry was filtered off, washed with plenty of water and dried at 80 C overnight
to obtain hot water treated alumina (Al2O3-HWT). Al2O3-HWT samples were thoroughly
grounded with various weight percentage (wt%) of metal precursors, tin chloride pentahydrate
(SnCl4.5H2O) or zirconium oxychloride octahydrate (ZrOCl2.8H2O) or titanium(IV) ethoxide
using pestle and mortar. The above mixtures were calcined at 450 C for 4 h with a ramp of 3
C/min to yield metal-containing alumina catalysts (M/Al2O3-HWT; M=Sn or Ti or Zr).
SnO2/Al2O3-HWT was also prepared by grinding the appropriate amount of SnO2 and Al2O3-HWT
followed by calcination at 450 C for 4 h with a ramp of 3 C/min.
2.2 Glucose isomerisation
All glucose isomerisation reactions were carried out in 15 ml Ace pressure tubes. In a typical
reaction protocol, 180 mg of glucose, 80 mg of catalyst, 4 g methanol were placed in the pressure
tube, tightly closed and kept for stirring in a preheated oil bath maintained at required reaction
temperature for various time. The pressure tube was quenched with cold water after the completion
of the reaction, and the aliquots of the reaction mixtures were collected for HPLC analysis (Agilent
5

1260, Rezex-RCM-Monosaccharide Ca2+ column-300 x 7.8 mm, RI detector and deionised water
as eluent). Glucose, fructose and mannose were quantified by making a series of several individual
standards. As methyl fructoside merged with fructose, the combined yield of fructose and methyl
fructoside using fructose as standard is reported.
2.3 Characterization techniques
-
The synthesised metal-containing Al2O3 HWT were characterised by UV-Vis using the
Schimadzu 2600 spectrometer in reflectance mode in the wavelength range of 200-900 cm-1. XRD
analysis was carried out using the X’Pert PRO Panalytical Netherlands XRD diffractometer with
CuKα radiation of 1.54060 Å. Diffraction patterns were collected in the range of 2θ degree of 10-
80. FTIR analyses were performed with Agilent Cary 660 series and the spectra were collected
between 500 and 4000 cm-1. Scanning electron microscopy (SEM) pictures were captured using
JEOL-6000 Benchtop instrument. NH3-temperature programmed desorption (TPD) analyses were
performed in BELCAT II for quantifying the total number of acid sites. ICP-MS analysis was
performed to quantify the amount of tin leached into the reaction mixture using an Agilent 7700X.
Results and Discussion
3.1 Characterization
The FTIR spectra of all alumina-based catalysts for specific wavenumber range (3000-4000
cm-1) and the full frequencies are given in Fig. 1 and Fig. S1, respectively. The peaks
corresponding to isolated hydroxyl groups and associated hydroxyl groups appeared at 3700-3800
cm-1 and 3200-3600 cm-1, respectively [32]. A very broad peak appeared between 2900 and 3800
cm-1 for Al2O3-HWT after treating under hot water conditions, indicating the enhancement of
presence of hydroxyl groups on the surface of alumina (Fig 1 and Fig. S1). A peak at 3451 cm-1
for pristine alumina can be ascribed to OH stretching band of gibbsite, -Al(OH)3 which broadened
with increased intensity centred at 3440 cm-1 for Al2O3-HWT, as in accordance with the previous
report [33]. In the case of Sn/Al2O3-HWT, the intensity of the peak at 3440 cm-1 significantly
reduced.
6

3451
3095
0.5wt% Sn/Al₂O₃-HWT
Al₂O₃-HWT
Al₂O₃
0.5wt% SnO₂/Al₂O₃-HWT
4000
3900
3800
3700
3600
3500
3400
3300
3200
3100
3000
Wavenumber (cm-1)
Fig. 1. FTIR spectra of Sn-containing alumina catalysts
Likewise, pristine alumina exhibited a peak around 3095 cm-1 corresponding to hydroxyl
groups [34] which, in the cases of Al2O3- HWT and 0.5 % Sn/Al2O3-HWT, shifted towards lower
wavenumber (3087 cm-1). This peak (3087 cm-1) assigned to hydroxyl groups of boehmite phase
(-AlOOH) formed [34] due to the hot water treatment, but in the case of 0.5% Sn/Al2O3-HWT,
the peak remained almost unchanged. However, in the case of SnO2/Al2O3-HWT, the FTIR
spectrum showed a peak similar to pristine alumina. The peak belonging to gibbsite, -Al(OH)3
(3451cm-1) almost disappeared for 3 and 5 wt% Sn/Al2O3-HWT. The intense peak around 2360-
2400 cm-1 ascribed to adsorbed CO2 (Fig. S1) [35, 36]. The peak at 1065-1100 cm-1 for Al2O3-
HWT can be attributed to the vsymAl-OH formed on hydrothermal treatment which was small in
the case of pristine alumina (Fig. S1) [36].
7

0.5%SnO2/Al2O3
SnO2
5%Sn/Al2O3
3%Sn/Al2O3
0.5%Sn/Al2O3
HWT-Al2O3
Al2O3
10
20
30
40
50
60
70
80
2 /
Fig. 2. XRD diffractograms of Sn-containing alumina catalysts
XRD diffractograms of Sn-containing alumina-based catalysts are presented in Fig. 2. The
XRD patterns of Al2O3-HWT revealed that the hydrothermal treatment of alumina slightly
influenced the crystallinity of alumina, which could be seen by the appearance of peaks at 2 of
around 20 and 40. These two new peaks attributed to Al(OH)3 and -AlOOH and are in line with
the previously reported article related to the hydrothermal treatment of alumina [30]. These newly
appeared peaks completely disappeared for 0.5wt% Sn/Al2O3-HWT. Moreover, it has been
reported that the formation of hydroxide phase of boehmite -AlOOH during the hydrothermal
treatment of alumina could revert upon calcination [37]. Interestingly, when increasing the loading
of Sn (up to 5 wt%) with respect to Al2O3-HWT, no peaks corresponding to SnO2 (2 of 28 and
35) was formed, as confirmed from the XRD diffractogram of SnO2 (Fig. 2).
8

0.4
0.3
0.2
0.1
0.0
0.5wt% Sn/Al O -HWT
SnO2
5wt% Sn/Al2O3-HWT
2
3
0.5wt% SnO2/Al2O3-HWT
Al2O3-HWT
3wt% Sn/Al2O3-HWT
241
233
200
300
400
500
600
700
Wavelength (nm)
Fig. 3. UV-Vis diffuse reflectance spectra of Sn-containing alumina catalysts
Fig. 3 shows UV-Vis diffused reflectance spectra of the alumina, tin oxide and tin-containing
alumina-based materials, and the corresponding deconvoluted spectra are presented in Fig. S2.
0.5wt% Sn/Al2O3-HWT exhibited a peak around 241 nm which could be ascribed to tin inserted
into the network of alumina, which was close to the previously published reports related to tin
incorporated zeolites and mesoporous Sn-Al catalysts [38-40]. Crystalline SnO2 and 0.5wt%
SnO2/Al2O3-HWT showed a significant peak around 255 nm, revealing that tin was not
incorporated into the Al2O3-HWT matrix when using SnO2 as a precursor (Fig 3 and Fig. S2).
Moreover, SnO2 and 0.5wt% SnO2/Al2O3-HWT displayed a very broad peak between 200 and 350
nm which patterns were dissimilar to 0.5wt% Sn/Al2O3-HWT. Besides, Al2O3-HWT showed
major bands at 251 and 306 nm which shifted towards lower wavelength (blue shift) for 0.5wt%
Sn/Al2O3-HWT; thus suggesting the tin incorporation. When increasing the tin content (3 and 5
wt% Sn/Al2O3-HWT), the peak shifted towards lower wavelength (239 and 231nm) from 241 nm
(0.5wt% Sn/Al2O3-HWT), substantiating the incorporation of tin into the network of Al2O3-HWT
(Fig. 3 and Fig. S2)
The SEM images of metal-containing alumina-based catalysts are shown in Fig. 4. Pristine
Al2O3 possessed large particles (Fig. 4a), and the surface of these large particles was significantly
9

affected by the hot water treatment (Fig. 4b). The significant changes in the morphology of Al2O3
- HWT could be due to the formation of other phases gibbsite (-Al(OH)3) and boehmite (-
AlOOH), as evidenced from FTIR spectra and XRD diffractograms of Al2O3 and Al2O3-HWT
materials. The SEM images of metal (Sn, Ti and Zr)-incorporated Al2O3-HWT are shown in Fig.
4 and Fig. S3.
(a)
(b)
(c)
(d)
Fig. 4. SEM Images of a) Al2O3-pristine b) Al2O3 - HWT c) 0.5 wt% Sn/ Al2O3 - HWT d) 0.5 wt% SnO2/ Al2O3 –
HWT
Al2O3-pristine, Al2O3-HWT and 0.5 wt% Sn/Al2O3-HWT were subjected to 27Al MAS
NMR analysis, and the spectra are given in Fig. 5. Al2O3-pristine showed a major and very minor
peaks at10.92 and 71.59 ppm which ascribed to octahedral (AlO6) and tetrahedral (AlO4)
aluminium, respectively [41, 42]. In the case of Al2O3-HWT, the intensity of tetrahedral (AlO4)
aluminium increased with a slight downfield shift (76.23 ppm), suggesting the hot water treatment
might have played a role for generating additional tetrahedral aluminium species. Interestingly,
the peaks of both octahedral (AlO6) and tetrahedral (AlO4) aluminium significantly shifted to 5.67
and 55.17 ppm, respectively, for 0.5 wt% Sn/Al2O3-HWT [43], inferring that Sn might have
incorporated into the alumina network; thus affecting the inner structure of Al2O3. 0.5 wt%
Sn/Al2O3-HWT was also subjected to XPS analysis and compared with crystalline SnO2 and
Al2O3-HWT. The XPS spectra of Sn 3d and Al 2p of the materials are shown in Fig. S4. 0.5 wt%
Sn/Al2O3-HWT exhibited broad peaks centred at 484.5 and 492.7 eV which could be assigned to
3d3/2 and 3d5/2, respectively. The binding energy value difference between the spin-orbit doublets
10

was 8.2 which is in good agreement with the previous reports [44]. The XPS spectrum of
crystalline SnO2 displayed two sharp peaks of Sn 3d at 483.7 and 492.2 eV with a spin-orbit
splitting of 8.5 eV. Sn 3d peaks of 0.5 wt% Sn/Al2O3-HWT significantly shifted when compared
to SnO2, implying the incorporation of Sn into the network of alumina, and the broader peaks
suggested further that tetrahedral Sn4+ or octahedral tin species present in alumina might be
overlapping. A minimal shift in binding energy value of Al 2p peak (71.9 eV) of 0.5 wt%
Sn/Al2O3-HWT compared to Al2O3-HWT (71.6 eV) was observed, inferring the incorporation of
tin into the alumina network.
Al2O3
Al2O3-HWT
0.5% Sn/Al2O3-HWT
200
100
0
-100
-200
ppm
Fig. 5. 27Al MAS NMR spectra of Al2O3-pristine, Al2O3-HWT and 0.5 wt% Sn/Al2O3-HWT
3.2 Catalytic Results
Initial experiments were performed with pristine Al2O3 and Al2O3- HWT for the catalytic
isomerisation of glucose to fructose in water and methanol, and the results are summarised in Table
1. The yield of fructose was only 7% with Al2O3-pristine when experimenting in water. The yield
of fructose (15%) increased two-fold when using Al2O3-HWT and Al2O3-pristine as catalysts in
methanol. As fructose and methyl fructoside are not able to be separated in HPLC, a combined
yield of these two products are provided. 0.5 wt% Sn on both Al2O3-pristine, and Al2O3-HWT
exhibited a significant difference in terms of fructose yield, affording 15.1 and 30.4% fructose
along with 3.3 and 11.0% mannose, respectively, implying the significance of the hot water
treatment on the activity of the material. With regard to this, the samples were subjected to NH3-
TPD analysis, and the results are presented in Table 1 and Fig. 6. Al2O3-pristine and Al2O3-HWT
11

possessed 453 and 634 µmol/g of total acid sites, respectively, suggesting that the number of acid
sites increased due to the enhancement of the presence of hydroxyl groups upon hot water
treatment. However, Al2O3-HWT did not increase the yield of fructose compared to Al2O3-pristine.
The total number of acid sites decreased to 392 µmol/g (0.5wt% Sn/Al2O3-pristine) and 522 µmol/g
(0.5wt% Sn/Al2O3-HWT) after the incorporation of tin into the alumina network. Unlike Al2O3-
HWT, 0.5wt% Sn/Al2O3-HWT displayed higher catalytic activity than Sn/Al2O3-pristine (Table
1), signifying the importance of hot water treatment before incorporating tin. Moreover, 0.5wt%
Sn/Al2O3-HWT possessed strong acid sites (> 400 C, Fig. 6) which favoured the formation of
methyl fructoside along with a minor amount of methyl glucosides.
The yield of fructose decreased to 18.3 and 14.9% when changing the metal to Zr and Ti in
place of Sn, respectively, under identical reaction conditions, suggesting that the strength of Lewis
acidity is weaker than Sn which could lower the catalytic activity, according to the previously
published articles [2,45]. Moreover, the hot water treatment of alumina caused to increase the
total acidity and surface area of the alumina [30] which further contributed to enhancing the total
activity of Sn/Al2O3-HWT.
Table 1. Glucose isomerisation in methanol with alumina-based catalystsa
Entry Catalyst
Conv.
(%)
Product distribution (%)
Glucose Fructoseb
Acid sites
(µmol/g)
Mannose
1c
2
Al2O3-pristine
20.8
20.1
18.7
22.3
79.2
79.9
81.3
77.7
6.9
-
-
Al2O3-pristine
Al2O3-HWT
14.8
14.6
15.1
3.0
2.6
3.3
453
634
392
3
4
Sn/Al2O3-
pristine
5
6
7
Sn/Al2O3-HWT
Zr/Al2O3-HWT
Ti/Al2O3-HWT
59.7
26.4
22.5
40.3
73.6
77.5
30.4
18.3
14.9
11.0
4.0
522
-
-
3.5
aReaction Conditions: 180 mg glucose, 80 mg catalyst (0.5 wt% metal with respect to
c
alumina), 4g methanol, 110 C, 4h. bincluding methylfructoside. water as a solvent.
12

Al2O3-pristine
Al2O3-HWT
0.5% Sn/Al2O3-HWT
0.5% Sn/Al2O3-pristine
3% Sn/Al2O3-HWT
5% Sn/Al2O3-HWT
200
300
400
500
600
Temperature (C)
Fig. 6. NH3-TPD profiles of alumina and tin-containing alumina catalysts
In order to have substantial evidence for the formation of fructose, methyl fructosides and
mannose from glucose, experiments were performed with isotopically labelled D-[2-13C]-glucose,
and D-[2-13C]-fructose with 0.5wt% Sn/Al2O3-HWT in MeOH-d4, and the corresponding samples
were subjected to 13C-NMR analysis (Fig. S5 and S6). Experiment with D-[2-13C]-glucose showed
three major peaks appeared at 101.53, 105.21 and 109.21 ppm corresponding to methyl-α-D-
fructopyranoside, methyl-β-D-fructofuranoside and methyl-α-D-fructofuranoside, respectively,
inferring that fructose formed from glucose was subsequently ketalised to form these tautomers
with methanol in the presence of 0.5 wt% Sn/Al2O3-HWT as in accordance with previous reports
[4, 46]. The tautomers of fructose, β-D-fructopyranose, β-D-fructofuranose and α-D-
fructofuranose, were also observed at 99.22, 103.17 and 105.42 ppm, respectively. The minor
peaks corresponding to α-D-mannopyranose and β-D-mannopyranose were also found at 73.00
and 72.65 ppm, respectively, substantiating the epimerization reaction of glucose to mannose took
place to a small degree. Two other peaks also appeared at 79.30 and 81.81 ppm which can be
13

assigned to methyl-α-D-glucofuranoside and methyl-β-D-glucofuranoside, respectively, as in line
with previous studies, [4, 47] implying that α- and β-glucopyranose forms were actalised by
methanol to form corresponding products with 0.5wt% Sn/Al2O3-HWT. Besides, two other major
peaks for unconverted glucose, α-D-glucopyranose and β-D-glucopyranose, appeared at 73.88 and
76.31 ppm, respectively.
Table 2. Influence of starting substrate on product distributiona
Substrates
Conv. (%)
Product distribution (%)
Fructoseb
Glucose
Mannose
Glucose
Fructose
Mannose
59.7
25.8
68.6
40.3
< 0.5
2.5
30.4
74.2
34.7
11.0
<0.5
31.4
aReaction Conditions: 180 mg substrate, 80 mg catalyst of 0.5 wt% Sn/Al2O3-HWT, 4g methanol, 4 h, 110 C
bincluding methylfructoside
The reverse isomerisation and epimerization of glucose were studied using fructose and
mannose as starting substrates with 0.5 wt% Sn/Al2O3-HWT in methanol. A small amount of
glucose and mannose were observed in HPLC chromatogram when using fructose as a substrate.
Likewise, when employing mannose as a substrate, 34.7 and 2.5% of fructose and glucose,
respectively, were obtained. This implies that reversible isomerisation and epimerization occurred,
and these sugars were in equilibrium in methanol to a small degree (Table 2). To prove the
reversible isomerisation of fructose, an experiment was performed with D-[2-13C]-fructose as
substrate. 13C-NMR spectrum revealed the formation of minor peaks corresponding to α- and β-
glucopyranose and α-D-mannopyranose, suggesting that the isomerisation of glucose was
reversible under these reaction conditions, apart from the three significant peaks of methyl
fructosides (Fig. S6).
14

Table 3. Influence of Sn loading on alumina on the yield of productsa
Catalyst
Conv. (%)
Product distribution (%)
Glucose Fructoseb
Mannose
2.6
Acid sites
(µmol/g)
Al2O3-HWT
18.7
59.7
51.5
61.8
81.3
40.3
48.5
38.2
14.6
30.4
26.4
25.0
634
522
508
561
0.5wt% Sn/Al2O3-HWT
3wt% Sn/Al2O3-HWT
5wt% Sn/Al2O3-HWT
11.0
16.4
18.3
b
aReaction Conditions: 180 mg glucose, 80 mg catalyst, 4g methanol, 110 C, 4 h. including
methylfructoside.
The influence of tin loading on Al2O3-HWT was investigated on the yield of products, and the
obtained results are summarized in Table 3. When increasing the amount of Sn on Al2O3-HWT
from 0.5 to 3 wt%, the fructose yield decreased from 30.4 to 26.4% but the yield of mannose
increased from 11.0 to 16.4 % with the glucose conversion of 59.7 and 51.5%, respectively. The
other byproducts methyl glucosides (<1.0%) were observed in HPLC chromatogram. However, no
significant change in the total number of acid sites was observed (Table 3). Further, increasing the
tin content from 3 to 5wt% on Al2O3-HWT resulted in slightly lowering the yield of fructose (25%)
and increasing the yield of mannose (18%), even though the acid sites increased to 561 µmol/g.
Concurrently, glucose conversion increased to 61.8% from 51.5% might be due to the formation
of a significant amount of methyl glycosides (6%). The influence of reaction time on the yield of
fructose was also investigated, and the results are presented in Fig. 7. The yields of fructose and
mannose were 20 and 4%, respectively, along with 38.5% conversion of glucose with 0.5wt%
Sn/Al2O3-HWT in methanol after 1 h. As the reaction time increased, the yields of both fructose
and mannose increased along with the increase in the conversion of glucose. The influence of
reaction temperature was also studied between 90 and 120 C with 0.5 wt% Sn/Al2O3-HWT and
the highest yield of fructose (30.4%) was obtained at 110 C after 4 h (Table 4).
15

60
50
40
30
20
10
0
Glucose
Fructose
Mannose
0
2
4
6
8
10
Time (h)
Fig. 7. The influence of reaction time on the yield of products. Reaction Conditions: 180 mg
glucose, 80 mg 0.5 wt% Sn/Al2O3-HWT, 4g methanol, 110 C. (Fructose yield includes
methylfructoside).
Table 4. Influence of reaction temperature on the yield of productsa
Temp.
Conv. (%)
Product distribution (%)
(C)
Glucose
62.1
Fructoseb
Mannose
120
110
100
90
37.9
59.7
37.9
20.3
24.5
30.4
17.8
12.4
3.9
11.0
4.8
40.3
62.1
79.7
3.3
aReaction Conditions :180 mg glucose, 80 mg 0.5 wt% Sn/Al2O3-HWT, 4g methanol, 4 h.
bincluding methylfructoside
To confirm the role of grafted tin on alumina network, a series of experiments were performed
with a tin-containing catalyst under similar reaction conditions, and the results are given in Table
16

5. Physically mixed SnO2 and Al2O3-HWT yielded only 13.2 and 3.3 % of fructose and mannose,
respectively, with 82.2% of unconverted glucose, substantiating that SnO2 formed on the surface
of alumina was not very active for glucose isomerisation. An experiment was also performed with
SnO2 alone as catalysts and exhibited no activity towards fructose and mannose formation,
indicating that tin incorporated into the alumina network could show significant activity towards
glucose isomerisation to form fructose and methyl fructosides. On the other hand, SnCl4.5H2O was
also applied as a homogeneous catalyst under similar reaction conditions and gave only 13.6% of
fructose with a large amount of mannose (22%) with glucose conversion of 93.5%. HPLC
chromatogram of this reaction mixture revealed that a large amount of methyl glucosides (MAGP
and MBGP; 42.3%) (Fig. S7) formed during the reaction. It is known that Sn incorporated in the
framework of porous materials exhibit as a strong Lewis acid catalyst [2, 48]. With regard to this,
commercially available Amberlyst 16 wet, a Bronsted acid catalyst, was employed under identical
reaction conditions, affording 52.5 and 32.4 % of MAGP and MBGP, respectively, with no yield
of fructose. Moreover, these results demonstrated that Bronsted acid sites entirely favored for the
formation of acetalised products (methyl glucosides) and manifesting further that Sn, as Lewis
acid sites, on alumina played a crucial role for the interconversion of glucose to fructose.
17

Table 5. Role of tin on the product distributiona
Product distribution (%)
Catalyst
Conv. (%)
Glucose
40.3
Fructoseb
Mannose
11.0
MAGP
MBGP
<1.0
0.5 wt% Sn/ Al2O3-HWT
59.7
17.2
30.4
13.2
-
<1.0
-
0.5 wt% SnO2/ Al2O3-HWT
82.2
99.9
6.5
-
3.3
-
-
#
SnO2
<0.1
93.5
-
-
SnCl4.5H2O*
13.6
-
22.2
4.3
31.5
52.5
10.8
32.4
Amberlyst 16 wet
>99.9
b
aReaction Conditions: 180 mg g, 80 mg catalyst, 4g methanol, 110 C, 4 h. including
methylfructoside. *#equivalent amount of tin used as present in 80 mg of 0.5 wt% Sn/Al2O3–
HWT.
Finally, recyclability of 0.5 wt% Sn/Al2O3-HWT was investigated at the lower conversion of
glucose (< 25%) for four runs, and the results are exhibited in Fig. 8. After the first run, glucose
conversion was only 20% along with 12.6% of fructose and 4% mannose. The yield of fructose
was maintained between 12 and 15 % along with 2 to 4% of mannose from the first to the fourth
run, revealing that the activity of 0.5 wt% Sn/Al2O3-HWT was retained at least up to three recycles.
The recovered catalyst after each run was washed with water and calcined at 450 C for 4 h. After
the first run, the reaction mixture was subjected to ICP-MS analysis and found that only a
negligible amount of tin (1.4 ppm) leached into the solution which could have a minimal
contribution on catalysing glucose isomerisation, thus proving the heterogeneity of Sn in alumina
network.
18

Glucose
Fructose
Mannose
80
60
40
20
0
2
Number of Reuse
1
0
3
Fig. 8. The reuse study of 0.5 wt% Sn/Al2O3-HWT on the yield of fructose. Reaction Conditions:
Sn to glucose molar ratio = 1:30, 8g methanol, 110 C, 4 h. (Fructose yield includes methyl
fructosides)
4. Conclusions
We have shown that tin on hot water treated alumina (0.5 wt% Sn/Al2O3-HWT)
displayed a two-fold increase in the catalytic activity towards glucose isomerisation to yield
fructose and methyl fructoside (30.4%) in methanol than corresponding tin on pristine
alumina (0.5 wt% Sn/Al2O3-pristine). FTIR spectrum of Al2O3-HWT revealed the
appearance of the peaks centred at 3440 and 3087 cm-1 corresponding to gibbsite (
-
Al(OH)3) and boehmite ( -AlOOH), respectively, confirming the enhancement of the

presence of hydroxyl groups during hot water treatment which facilitated the facile
incorporation of tin into the matrix of Al2O3-HWT. UV-Vis spectrum of 0.5wt% Sn/Al2O3-
HWT confirmed the incorporation of tin into the alumina matrix when comparing the
spectra of SnO2/Al2O3-HWT and SnO2. 13C NMR spectra of isotope experiments with D-
[2-13C]-glucose and D-[2-13C]-fructose in methanol-d4 with 0.5 wt% Sn/Al2O3-HWT
exhibited the formation of fructose (β-D-fructopyranose, β-D-fructofuranose and α-D-
fructofuranose) and methyl fructosides (methyl-α-D-fructopyranoside, methyl-β-D-
fructofuranoside and methyl-α-D-fructofuranoside). 0.5 wt% SnO2/Al2O3-HWT showed a
lower activity towards fructose formation (13.2%) compared to 0.5 wt% Sn/Al2O3-HWT,
revealing the role of tin incorporated into the alumina network. No formation of fructose
19

with only SnO2 as catalysts was found, implying that any SnO2 formed on Al2O3-HWT
surface was not active for the glucose isomerisation under these conditions.
Acknowledgements
Authors gratefully thank Department of Biotechnology (DBT), New Delhi, India for financial
support and MAY thanks DBT for the fellowship. SERB-DST is gratefully acknowledged for
financial support (ECR/2016/001331), New Delhi, India.
References
[1] A. Corma, S. Iborra, A. Velty, Chem. Rev. 107 (2007) 2411−2502.
[2] M. S. Holm, S. Saravanamurugan, E. Taarning, Science. 328 (2010) 602-605.
[3] W. Zhang, W. Hou, T. Meng, W. Zhuang, J. Xie, Y. Zhou, J. Wang, Catal. Sci. Technol. 7
(2017) 6050-6058.
[4] S. Saravanamurugan, M. Paniagua, J. A. Melero, A. Riisager, J. Am. Chem. Soc. 135 (2013)
5246-5249.
[5] G. D. McGinnis, S. E. Prince, C. J. Biermann, J. T. Lowrimore Carbohydr. Res. 128 (1984)
51-60.
[6] F. Jin, Z. Zhou, T. Moriya, H. Kishida, H. Higashijima, H. Enomoto, Environ. Sci. Technol. 39
(2005) 1893-1902.
[7] P. Lanzafame, D. M. Temi, S. Perathoner, A. N. Spadaro, G. Centi, Catal.Today. 179 (2012)
178–184.
[8] Z. Zhang, Z. K. Zhao, Carbohydr. Res. 344 (2009) 2069–2072.
[9] Hu Li, S. Yang, A. Riisager, A. Pandey, R. S. Sangwan, S. Saravanamurugan, R. Luque, Green
Chem. 18 (2016) 5701-5735.
[10] H. Li, S. Yang, S. Saravanamurugan, A. Riisager, ACS Catal. 7 (2017) 3010-3029.
20

[11] A. Aho, S. Roggan, K. Eränen, T. Salmi, D. Yu. Murzin, Catal. Sci. Technol. 5 (2015) 953-
959.
[12] C. A. S. Lanziano, S. F. Moya, D. H. Barrett, E. Teixeira-Neto, R. Guirardello, F. de S. da
Silva, R. Rinaldi, C. B. Rodella, ChemSusChem. 11 (2018) 872 – 880.
[13] I. Delidovich, R. Palkovits, Green Chem. 18 (2016) 5822-5830.
[14] A. A. Marianou, C. M. Michailof, A. Pineda, E. F. Iliopoulou, K. S. Triantafyllidis, A. A.
Lappas, ChemCatChem. 8 (2016) 1100 – 1110.
[15] H. Li, S. Yang, S. Saravanamurugan, A. Riisager, ACS Catal. 7 (2017) 3010−3029.
[16] C. Liu, J. M. Carraher, J. L. Swedberg, C. R. Herndon, C. N. Fleitman, J. P. Tessonnier, ACS
Catal. 4 (2014) 4295−4298.
[17] M. Moliner, Y. Román-Leshkov, M. E. Davis, Proc. Natl. Acad. Sci. U.S.A. 107 (2010) 6164.
[18] Q. Yang, W. Lan, T. Runge, ACS Sustain.Chem. Eng. 4 (2016) 4850–4858.
[19] H. Kitajima, Y. Higashino, S. Matsuda, H. Zhong, M. Watanabe, T. M. Aida, R. L. Smith Jr.,
Catal. Today. 274 (2016) 67−72.
[20] J. W. Harris, M. J. Cordon, J. R. Di Iorio, J. C. Vega-Vila, F. H. Ribeiro, R. Gounder, J. Catal.
335 (2016) 141–154.
[21] J. Ohyama, Y. Zhang, J. Ito, A. Satsuma, ChemCatChem. 9 (2017) 2864.
[22] I. Graça, D. Iruretagoyena and D. Chadwick, Appl. Catal. B: Environ. 206 (2017) 434-443.
[23] I. Graçaa, M.C. Bacariza, A. Fernandes, D. Chadwick, Appl. Catal. B: Environ. 224 (2018)
660–670.
[24] L. Ren, Q. Guo, P. Kumar, M. Orazov, D. Xu, S. M. Alhassan, K. A. Mkhoyan, M. E. Davis
and M. Tsapatsis, Angew. Chem. Int. Ed. 54 (2015) 10848-10851.
[25] S. Xu, L. Zhang, K. Xiao, H. Xia, Carbohydr. Res. 446 (2017) 48-51.
[26] M. D. de Mello and M. Tsapatsis, ChemCatChem. 10 (2018) 706 – 709.
21

[27] Q. Guo, L. Ren, P. Kumar, V. J. Cybulskis, K. A. Mkhoyan, M. E. Davis, M. Tsapatsis,
Angew. Chem. Int. Ed. 57 (2018) 4926 –4930.
[28] J. Sun, H. Li, H. Huang, Bo Wang, L. P. Xiao, G. Song, ChemSusChem. 11 (2018) 1157 –
1162.
[29] R. M. Ravenelle, J. R. Copeland, W. G. Kim, J. C. Crittenden, C. Sievers, ACS Catal. 1 (2011)
552–561.
[30] L. Jun-Cheng, X. Lan, X. Feng, W. Zhan-Wen, W. Fei, Appl. Surf. Sci. 253 (2006) 766–770.
[31] F. Pienko, C. Ochoa-Hernández, N. Theyssen, W. Leitner, ACS Sustainable Chem. Eng. 6
(2018) 8782-8789.
[32] T. H. Ballinger, J. T. Yates, Langmuir. 7 (1991) 3041.
[33] E. Balan, M. Lazzeri, G.Morin, F. Mauri, American Mineralogist, 91 (2006) 115–119.
[34] C. Liu, K. Shih, Y. Gao, F. Li, L. Wei, J. Soil Sediment, 12 (2012) 724-733.
[35] L. Qu, C. He, Y. Yang, Y. He, Z. Liu, Mater. Lett. 59 (2005) 4034– 4037.
[36] C. L. Lu, J. G. Lv, L. Xu, X. F. Guo, W.H.Hou, Y. Hu, H. Huang, Nanotechnology. 20 (2009)
215604.
[37] R.M. Mironenko, O.B. Belskaya, V.P. Talsi, T.I. Gulyaeva, M.O. Kazakov, A.I. Nizovskii,
A.V. Kalinkin, V.I. Bukhtiyarov, A.V. Lavrenov, V.A. Likholobov, Appl. Catal. A: Gen. 469
(2014) 472–482.
[38] W. N. P. van der Graaff, G. Li, B. Mezari, E. A. Pidko, E. J. M. Hensen. ChemCatChem. 7
(2015) 1152 – 1160.
[39] Y. Romn-Leshkov, M. Moliner, J. A. Labinger, M. E. Davis, Angew. Chem. Int. Ed. 49 (2010)
8954 – 8957.
[40] Z. Zhou, P. Yu, J. Qin, W. Wu, L. Xu, Z. Gu, X. Liu, J Porous. Mater. 23 (2016) 239-245.
22

[41] C. Wan, M. Y. Hu, N. R. Jaegers, D. Shi, H. Wang, F. Gao, Z. Qin, Y. Wang, J. Z. Hu,
J. Phys. Chem. C, 120 (2016) 23093–23103.
[42]T. V. Cleve, D. Underhill, M. V. Rodrigues, C. Sievers, J. W. Medlin, Langmuir, 34 (2018),
3619–3625
[43]I. Contreras-Andrade, A. Vázquez-Zavala, T. Viveros, Energy Fuels, 23 (2009), 3835–
3841.
[44] J. A. Jennings, S. Parkin, E. Munson, S. P. Delaney, J. L. Calahan, M. Isaacs, K. Honge,
M. Crocker, RSC Adv., 7 (2017) 25987-25997.
[45] E. Taarning, S. Saravanamurugan, M. S. Holm, J. Xiong, R. M. West, C. H. Christensen,
ChemSusChem. 2 (2009) 625 – 627.
[46] L. Johnson, D. L. Verraest, J. van Haveren, K. Hakala, J. A. Peters, H. van Bekkum,
Tetrahedron: Asymmetry. 5 (1994) 2475.
[47] S. Saravanamurugan, A. Riisager, E. Taarning, S. Meier, Chem. Commun. 52 (2016) 12773-
12776.
[48] A. Corma, M. E. Domine, S. Valencia, J. Catal. 215 (2003) 294–304.fig
23