Three-step cascade over a single catalyst: Synthesis of 5-(ethoxymethyl)furfural from glucose over a hierarchical lamellar multi-functional zeolite catalyst

Article abstract of DOI:10.1039/c8ta01242c

The synthesis of hierarchical lamellar zeolites with a controlled meso-/microporous morphology and acidity is an expanding area of research interest for a wide range of applications. Here, we report a one-step synthesis of a hierarchical meso-/microporous

Full text of DOI:10.1039/c8ta01242c

View Article Online
View Journal
Journal of
M at erials Chemistry A
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Y. Bai, L. Wei, M.
Yang, H. Chen, S. Holdren, C. Yao, R. Sun, Y. Pan and D. Liu, J. Mater. Chem. A, 2018, DOI:
1
0.1039/C8TA01242C.
Volume
4
Number
1
7
January 2016 Pages 1–330
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry A
Accepted Manuscripts are published online shortly after
Materials for energy and sustainability
www.rsc.org/MaterialsA
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
author guidelines.
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, outlined
in our author and reviewer resource centre, 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 2050-7488
PAPER
Kun Chang, Zhaorong Chang et al.
Bubble-template-assisted synthesis of hollow fullerene-like
MoS nanocages as a lithium ion battery anode material
2
rsc.li/materials-a

Page 1 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
1
2
3
4
5
6
7
8
9
Threeꢀstep cascade over one single catalyst: synthesis of 5ꢀ(ethoxymethyl)furfural from
glucose over hierarchical lamellar multiꢀfunctional zeolite catalyst
a,b,c
a
d
d
e
c
Yuanyuan Bai , Wei Lu , Mengfei Yang , Huiyong Chen , Scott Holdren , Chunli Yao ,
c
f
*a
Runcang Sun , Yanbo Pan and Dongxia Liu
a
Department of Chemical and Biomolecular Engineering, University of Maryland, College Park,
MD, 20742, USA
b
National Engineering Lab for Pulp and Paper, China National Pulp and Paper Research
1
1
1
1
1
1
1
1
1
0
1
2
3
4
5
6
7
8
Institute Co. Ltd., Beijing, 100102, China
c
Beijing Key Laboratory of Lignocellulosic Chemistry, Beijing Forestry University, Beijing,
100083, China
d
School of Chemical Engineering, Northwest University, Xi’an, Shaanxi, 710069, China
e
Department of Chemistry and Biochemistry, University of Maryland, College Park, MD, 20742,
USA
f
Department of Chemical and Biomolecular Engineering, University of Akron, Akron 44325,
USA
1
2
2
2
2
2
9
0
1
2
3
4
*Corresponding author:
Prof. Dongxia Liu
Email: liud@umd.edu
Phone: (+1) 301ꢀ405ꢀ3522
Fax: (+1) 301ꢀ405ꢀ0523
2
2
5
6
1

Journal of Materials Chemistry A
Page 2 of 40
View Article Online
DOI: 10.1039/C8TA01242C
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
Abstract: The synthesis of hierarchical lamellar zeolites with controlled mesoꢀ/microporous
morphology and acidity is an expanding area of research interest for a wide range of
applications. Here, we report a oneꢀstep synthesis of hierarchical mesoꢀ/microporous lamellar
MFIꢀSn/Al zeolite (i.e., containing both Lewis acidic Snꢀ and Alꢀsites and Brønsted acidic Al–
O(H)–Si site and its catalytic application for conversion of glucose into 5ꢀ(ethoxymethyl)furfural
(EMF). Particularly, the MFIꢀSn/Al zeolite was prepared under the assistance of a diquaternary
+
+
2ꢀ
ammonium ([C H ꢀN (CH ) ꢀC H ꢀN (CH ) ꢀC H ]Br , C22ꢀ6ꢀ6) template in a composition of
22
45
3 2
6
12
3 2
6
13
100SiO /5C
/18.5Na O/xAl O /ySnO /2957H O (x=0.5, 1, and 2, y=1 and 2, respectively).
2 2 3 2 2
2
22ꢀ6ꢀ6
The MFIꢀSn/Al zeolites innovatively feature dual mesoꢀ/microporosity and dual Lewis and
Brønsted acidity, which enabled threeꢀstep reaction cascade for EMF synthesis from glucose in
the ethanol solvent. The reaction proceeded via the isomerization of glucose to fructose over
Lewis acidic Snꢀsites and dehydration of fructose to 5ꢀhydroxymethylfurfural (HMF) and then
etherification of HMF and ethanol to EMF over the Brønsted acidic Al–O(H)–Si sites. The coꢀ
existence of multiple acidities in one single zeolite catalyst enabled oneꢀpot cascade reactions for
carbohydrate upgrading. The dual mesoꢀ/microporosity in the MFIꢀSn/Al zeolites facilitated
mass transport in processing of bulky biomass molecules. The balance of both types of acidity
and mesoꢀ/microporosity realized EMF yield as high as 44% from the glucose reactant.
Keywords: 5ꢀEthoxymethylfurfural; Glucose; Biomass; Zeolites; Mesoꢀ/microporosity;
Hierarchical structure.
2

Page 3 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
5
0
5
5
5
5
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
Introduction
Over the past century, the world society has been heavily relied on fossil resources, such as coal,
1
ꢀ3
gas, and oil, to produce human beings’ daily needed fuels and chemicals. The carbon dioxide
produced from usage of these nonꢀrenewable fossil resources has been significantly changing the
2
climate on earth, and the heavy reliance on these fossil resources makes them more expensive
and less abundant. As a widely available and sustainable carbon source, biomass has the
potential to replace the nonꢀrenewable fossil resources for the production of fuels and
4
5
chemicals. Lignocellulosic biomass, consisting of 60~80% carbohydrate component, is the
most promising renewable carbon source for production of both fuels and platform chemicals.
Many strategies have been explored for the conversion of carbohydrates in lignocellulosic
6
ꢀ8
biomass into chemicals and fuels. The use of an appropriate catalyst plays a crucial role in
achieving high conversion and high selectivity to the end fuel and/or chemical alternative
9
,10
products in a green and economic process.
5ꢀ(Ethoxymethyl)furfural (EMF), a liquid with boiling point of 508 K and an energy density
−
1
−1
−1
of 30.3 MJ L , close to that of diesel (33.6 MJ L ), gasoline (31.1 MJ L ) and ethanol (23.5
−
1
11ꢀ13
MJ L ), is an attractive and potential end biofuel alternative product.
The synthesis of EMF
from biomass feedstock has been practiced via a single step etherification of 5ꢀ
1
4
15
hydroxymethylfurfural (HMF), twoꢀstep cascade of dehydrationꢀetherification of fructose,
1
6
threeꢀstep cascade of decompositionꢀdehydrationꢀetherification of disaccharide (e.g., sucrose)
17
or oligosaccharide (e.g., inulin) , and threeꢀstep cascade of isomerizationꢀdehydrationꢀ
1
2
etherification of monosaccharide such as glucose in ethanol solvent. The singleꢀstep synthesis
5
of EMF from HMF can occur over a variety of acidic catalysts with > 90% yield. The high cost
3

Journal of Materials Chemistry A
Page 4 of 40
View Article Online
DOI: 10.1039/C8TA01242C
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
and nonꢀstability of HMF, however, restrict its wide application in EMF production. The twoꢀ
step and threeꢀstep cascade reactions of fructoseꢀbased feedstocks on acidic catalysts overcome
the shortcomings in EMF synthesis via the singleꢀstep approach, but these saccharides are less
abundant in nature. In contrast, glucose or glucoseꢀbased oligosaccharide and polysaccharide is
the most abundant carbohydrate in lignocellulosic biomass.
The synthesis of EMF from glucose feedstock is desired, but requires multiple types of
catalysts for different steps in the reaction cascade. For example, a combination of SnꢀBEA and
1
2
Amberlystꢀ15 catalysts has been used for glucose conversion to EMF , in which the former
catalyst acts as Lewis acid for isomerization of glucose to fructose and the latter one catalyzes
the dehydration and etherification of fructose to EMF. In this threeꢀstep reaction cascade,
transport of reaction intermediates from one catalyst to the other in the reactor is needed, which
might impede the efficiency in EMF production from the process point of view. The usage of
multiple catalysts also complicates the catalyst preparation and/or separation steps if the catalyst
recycling is needed. Multifunctional catalyst, i.e., a single catalyst comprised of structurally
and/or compositionally different active components within nanoscale distance, can enable
multiple reaction steps successively and efficiently in the catalysis cascade. The precise position
1
8ꢀ
of active centers in the multifunctional catalysts, like the enzymatic catalyst systems in nature,
2
0
can also facilitate cooperative catalysis. The synthesis of EMF from glucose or glucoseꢀbased
abundant carbohydrates is desired to take place over such a multifunctional catalyst, but such a
catalyst system has been rarely reported.
In the present study, we report a hierarchical lamellar multifunctional MFIꢀSn/Al zeolite
catalyst that contains dual mesoꢀ/microporosity and dual Lewis and Brønsted acidity for glucose
conversion into the EMF product. The MFIꢀSn/Al catalyst was prepared via a singleꢀstep
4

Page 5 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
9
9
9
9
6
7
8
9
hydrothermal crystallization method under the assistance of a diquaternary ammonium template
+
+
2ꢀ
([C H ꢀN (CH ) ꢀC H ꢀN (CH ) ꢀC H ]Br , C
) in a composition of 100SiO /5C
2 22ꢀ6ꢀ
22
45
3 2
6
12
3 2
6
13
22ꢀ6ꢀ6
/18.5Na O/xAl O /ySnO /2957H O (x=0.5, 1, and 2, y=1 and 2, respectively). The coꢀexistence
6
2
2
3
2
2
of Sn and Al elements in the MFIꢀSn/Al zeolite endows both Lewis acidic Snꢀ and Alꢀsites and
Brønsted acidic Al–O(H)–Si sites for the threeꢀstep reaction cascade of isomerizationꢀ
dehydrationꢀetherification in the EMF production (Scheme 1). Tetravalent Sn site has showed
good performance for catalyzing the isomerization of a series of pentose and hexose sugars with
1
1
1
1
1
1
1
1
1
1
1
00
01
02
03
04
05
06
07
08
09
10
1
9,20
activities comparable to biological processes
by a mechanism similar to enzymatic
1
8
catalysts. Therefore, the Snꢀsites in the MFIꢀSn/Al catalyst were used to catalyze the
isomerization step in the reaction cascade. The Brønsted acidic Al–O(H)–Si sites catalyze the
subsequent dehydration of fructose to HMF and etherification of HMF to EMF product. The
presence of hierarchical mesoꢀ and microporosity in the MFIꢀSn/Al catalyst allows for facile
transport of bulky molecules, and thereby further improving the catalyst performance in this
reaction cascade. A high EMF yield (44%) has been achieved from glucose conversion over the
multifunctional MFIꢀSn/Al catalyst.
HO
OH
O
OC H
2
5
O
HO
OH
OH
O
O
O
O
MFIꢀSn/Al
MFIꢀSn/Al
MFIꢀSn/Al
OH
Isomerization
OH
OH
Dehydration
Etherification
OH
OH
OH
Glucose
Fructose
5ꢀ(Hydroxymethyl)furfural
5
ꢀ(Ethoxymethyl)furfural
1
1
11
12
Scheme 1 Schematic illustration of EMF synthesis from glucose via a threeꢀstep reaction
1
13
cascade over single MFIꢀSn/Al zeolite catalyst.
1
1
14
15
Experimental
Materials
5

Journal of Materials Chemistry A
Page 6 of 40
View Article Online
DOI: 10.1039/C8TA01242C
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Aluminum isopropoxide (Al[OCH(CH ) ] , 99.99+% purity), tetraethyl orthosilicate (TEOS,
3 2 3
98%), tetrapropylammonium hydroxide (TPAOH, 40 wt%), D (+)ꢀglucose (98%), Dꢀfructose
(98%), sucrose (98%) and inulin (98%) were purchased from Alfa Aesar. Tin chloride
pentahydrate (SnCl ꢁ5H O, 98%), sodium hydroxide (NaOH, ≥ 97.0%), and ammonium nitrate
4
2
(NH NO , ≥99.0%) were supplied by SigmaꢀAldrich. Deionized (DI) water was used throughout
4
3
+
+
the experiment. Diquaternary ammonium template ([C H ꢀN (CH ) ꢀC H ꢀN (CH ) ꢀ
2
2
45
3 2
6
12
3 2
2
1
C H ]Br , (C22ꢀ6ꢀ6)) was synthesized following the method reported by Choi et al and the
6
13
2
2
2,23
detailed synthesis procedure has been described in our previous work.
Catalyst preparation
The MFIꢀSn/Al zeolite catalysts were synthesized in a oneꢀstep hydrothermal crystallization
process using the following recipe: 100SiO /5C /18.5Na O/xAl O /ySnO /2957H O, where
2
22ꢀ6ꢀ6
2
2
3
2
2
x=0.5, 1, and 2, y=1 and 2, respectively. Typically, the synthesis started by dissolving 0.41 g
NaOH in 11.50 g DI water, followed by addition of desired amount of SnCl ⸱ 5H O and
4
2
Al[OCH(CH ) ] , separately. After a clear solution was formed, 5.78 g TEOS was added and the
3
2 3
resultant mixture was stirred vigorously at room temperature for 8 h. A C22ꢀ6ꢀ6 solution was
prepared by dissolving 0.95 g C22ꢀ6ꢀ6 in 2.83 g of DI water at 333 K and then added into above
mixture. The resultant zeolite synthesis gel was continuously stirred at room temperature for 2 h.
Finally, the synthesis gel was transferred into a Teflon vessel in a stainless steel autoclave. The
autoclave was tumbled vertically at 30 rpm and 423 K in an oven for 14 days. After the
hydrothermal synthesis, the zeolite products were collected by centrifugation at 6000 rpm for 15
min and then washed by dispersing in DI water. The water washing and centrifugation steps were
repeated 6 times. A vacuum oven was utilized to dry the wet zeolite product at room temperature
ꢀ1
overnight. Finally, the zeolite samples were calcined in a flowing air (1.67 mL s , ultrapure air,
6

Page 7 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
Airgas) at 873 K for 6 h. The obtained samples were named as MFIꢀSn/Al (Si/Sn molar ratio,
Si/Al molar ratio). According to the synthesis recipe, the following samples, MFIꢀSn/Al
(100/100), MFIꢀSn/Al (100/50), MFIꢀSn/Al (100/25), MFIꢀSn/Al (50/50), MFIꢀSn/Al (∞/100),
have been prepared, respectively.
For comparison purpose, the conventional microporous MFI zeolite containing both Sn and
2
4
Al (designated as CꢀMFIꢀSn/Al (100/100)) was prepared according to the work of Mal et al.,
and the synthesis procedure was as follows. Firstly, 0.26 g of SnCl ꢁ5H O and 0.15 g of
4
2
Al[OCH(CH ) ] were dissolved with 19.00 g of DI water in a polyethylene bottle. Then, 15.61 g
3
2 3
TEOS was added and the resultant mixture was stirred magnetically for 30 min. 16.50 g TPAOH
solution was then added and the mixture was continuously stirred for 1 h. The molar ratio of the
CꢀMFIꢀSn/Al (100/100) synthesis gel was 100SiO /SnO /0.5Al O /44TPAOH/3430H O. Finally,
2
2
2
3
2
the mixture was transferred into a Teflonꢀlined autoclave and kept at 433 K for 2 days under
static condition. After the synthesis, the sample was collected by centrifugation, washed by DI
water and calcined using the procedures described for preparation of MFIꢀSn/Al zeolites above.
Prior to catalysis tests, all the zeolite samples were ionꢀexchanged three times using 1.0 M
aqueous NH NO solution (weight ratio of zeolite to NH NO solution = 1:10) for 3 hours at 353
4
3
4
3
K. The sample was collected by centrifugation, washed with DI water, and dried at 353 K for 12
+
ꢀ1
h. All zeolite samples in their NH ꢀform were treated in air (1.67 mL s , ultrapure air, Airgas)
4
ꢀ1
by increasing the temperature from ambient to 823 K at 0.167 K s rate and holding them at this
temperature for 4 h.
Catalyst characterization
The scanning electron microscopy (SEM) images taken from a Hitachi Suꢀ70 electron
microscope were used to detect the morphologies of the synthesized zeolite samples. The powder
7

Journal of Materials Chemistry A
Page 8 of 40
View Article Online
DOI: 10.1039/C8TA01242C
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
Xꢀray diffraction (XRD) patterns were recorded using a Bruker D8 Advance Lynx Powder
Diffractometer (LynxEye PSD detector, sealed tube, Cu Kα radiation with Ni βꢀfilter) in the 2 θ
ranging from 1.2° to 40°. The N adsorptionꢀdesorption isotherms were measured at 77 K with
2
an AutosorbꢀiQ analyzer (Quantachrome Instruments) to characterize the textual properties of
the zeolite samples. Prior to the measurement, all the zeolite samples were evacuated overnight
at 623 K and 1mm Hg pressure. The solid state magic angle spinning NMR (MAS NMR) spectra
2
9
27
29
were recorded at a field of 7 T (Bruker DSX 300) for Si and Al. Si MAS NMR spectra were
recorded at 59.64 MHz using 4 mm rotors at a spinning speed of 10 kHz, a dwell time of 20 ꢂs, a
π/2 pulse of 5.0 ꢂs, and a recycle delay of 60 s. The spectra were referenced with respect to 3ꢀ
2
7
(trimethylsilyl)ꢀ1ꢀpropanesulfonic acid salt. Al MAS NMR spectra were recorded at 78.22
MHz using 4 mm rotors at 10 kHz spinning speed, a dwell time of 1.0 ꢂs, a selective π/10 pulse
of 0.6 ꢂs, and a recycle delay of 4 s. An aqueous solution of aluminum nitrate (0.1 M) was used
as the reference. The ultravioletꢀvisible diffuse reflectance spectroscopy (UVꢀVisꢀDRS) was
performed using a PerkinꢀElmer Lambda 650S model to understand the structure of the Snꢀsites
in zeolite samples. The integrating sphere is 60 mm and is made by LabSphere. The coating in
the LabSphere 60 mm integrating attachment is Spectrolong. Typical scan rate was 1 nm/second
and the scan range was from 200 nm to 800 nm. All the spectra were recorded in ambient
condition.
Composition and acidity measurement
The Si, Al, and Sn contents of zeolite samples were determined by inductively coupled plasma
atomic emission spectroscopy (ICPꢀAES, iCAP 6500 dual view). The concentration of active
Brønsted acid sites in each zeolite sample was determined via the reactive gas chromatography
(RGC) method using isopropylamine (CH CH(NH )CH ) as the probe molecule. The selective
3
2
3
8

Page 9 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
decomposition of isopropylamine adsorbate on Brønsted acid site (CH CH(NH )CH ꢁꢁꢁAlO(H)Si)
3 2 3
in zeolites via Hoffmann elimination forms propylene and ammonia, CH CH(NH )CH +
3
2
3
2
5,26
AlO(H)Si ꢀ CH CH(NH )CH ꢁꢁꢁAlO(H)Si ꢀ CH =CHCH + NH + AlO(H)Si.
The
3
2
3
2
3
3
quantification of propylene by a GC instrument determined the number of AlO(H)Si sties in each
zeolite sample. The experimental setup and reaction conditions for chemical titration of Brønsted
2
6
acid site were the same as those reported by Abdelrahman, et al.
The types of acid sites were identified by the recorded Fourier transform infrared (FTIR)
spectra on a Nicolet iSꢀ50R spectrometer with an associated Harrick Scientific Praying Mantis
optical accessory and high temperature reaction chamber (HVCꢀDRPꢀ5). In a typical run, the
DRIFTS sample cell was filled with a zeolite sample that was preꢀtreated at 823 K for 6 h under
−
1
following helium (He, 0.083 mL s ) in a furnace. The zeolite sample was then activated at 823
−
1
K under the He flow (0.083 mL s ) for 1 h in the DRIFTS cell. After cooling the sample to 393
K, a spectrum was recorded against the background spectrum of KBr powder measured under the
same conditions. Afterwards, adsorption of pyridine was performed by flowing a mixed pyridine
and He gas stream through the catalyst. Following adsorption, the sample was heated to 523 K
under a flowing He to desorb the physisorbed pyridine, cooled to the temperature of 393 K, and
ꢀ1
then FTIR spectra were recorded in the range of 4000−600 cm with 128 scans at an effective
ꢀ1
resolution of 2 cm . The asꢀrecorded FTIR spectra of adsorbed pyridine were used to determine
the type and concentration of Brønsted and Lewis acid sites in the catalyst. Alternatively, the
activated zeolite sample in the DRIFT cell was cooled to 303 K and then exposed to a mixed
trideuteroacetonitrile (CD3CN) and argon (Ar) gas stream using a bubbler with Ar as the carrier
gas. After the excess CD3CN was flushed out from the zeolite sample using a flowing Ar gas
9

Journal of Materials Chemistry A
Page 10 of 40
View Article Online
DOI: 10.1039/C8TA01242C
−
1
ꢀ1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
(0.083 mL s ), the spectra were collected at 2 cm resolution. The recorded FTIR spectra of
CD3CN were used to analyze the Snꢀsite in each zeolite sample.
The acidity of the zeolite was further evaluated by ammonia temperature programmed
desorption (NH ꢀTPD) using an AutosorbꢀiQ instrument (Quantachrome, ASIQM000ꢀ4)
3
equipped with a thermal conductivity detector (TCD). In the NH ꢀTPD measurement, the zeolite
3
–
1
sample (0.10 g) was loaded into a quartz cell and heated to 873 K at a rate of 0.05 K s under He
–
1
(0.67 mL s ) and maintained at this temperature for 1 h. After cooling to 423 K under the He
–
1
stream, the sample was exposed to NH stream (0.5 mL s , ultrapure, Airgas) for 0.5 h, and then
3
–
1
back to the helium flow (0.5 mL s ) for 2 h to remove NH residue in the system. Lastly, the
3
–
1
catalyst sample was ramped to 823 K at a ramp rate of 0.167 K s , and the NH desorption
3
profile was recorded.
Catalytic reaction for EMF synthesis from carbohydrates
In a typical catalytic reaction, a 20 mL thickꢀwalled glass reactor sealed with crimp tops
(PTFE/silicone septum) was charged with glucose (0.090 g), zeolite catalyst (0.074 g), and
ethanol (3 mL), respectively. The glass reactor was then heated to a desired temperature and the
magnetic stirring was controlled at 600 rpm in all the catalysis experiment. After a certain
reaction time, the reactor was quenched in an ice bath and the reaction mixture was sampled for
composition analysis. A high performance liquid chromatography (HPLC, Agilent 1100)
equipped with an Aminex HPXꢀ87H column connected to an autoꢀsampler and a refractive index
detector was used to calibrate and separate the reactant and products. During the measurement,
ꢀ1
the column was kept at 413 K with 0.005 M sulfuric acid at a flow rate of 0.5 mL min as the
mobile phase. For comparison purpose, the same reaction and composition analysis conditions
were applied to the fructose, sucrose, and inulin reactant in ethanol solvent, respectively. The
10

Page 11 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
2
2
30
31
effect of types of reactants on the EMF production over MFIꢀSn/Al zeolite catalysts was thus
examined.
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Results and Discussion
Morphology and textural properties of MFIꢀSn/Al zeolite catalysts
Fig. 1 shows XRD patterns of the synthesized hierarchical lamellar MFIꢀSn/Al zeolite catalysts.
For comparison purpose, the XRD patterns of lamellar MFI that only contains Al sites (MFIꢀ
Sn/Al (∞/100)) and conventional microporous MFI that comprises of both Sn and Al sites (Cꢀ
MFIꢀSn/Al (100/100)) are included. As indicated in Fig. 1, all the MFIꢀSn/Al zeolite samples
exhibit similar diffraction patterns, resembling the characteristics of crystalline MFI
21,27,28
zeolite.
This result confirms the successful synthesis of hierarchical lamellar MFI in the
presence of both Sn and Al species in the synthesis gels. The absence of smallꢀangle diffraction
peaks in Fig. 1 suggests that longꢀrange ordering of the layered zeolitic structure does not exist in
the hierarchical lamellar MFIꢀSn/Al zeolites. A comparison for the peak intensity between the
MFIꢀSn/Al samples and CꢀMFIꢀSn/Al (100/100) shows that the diffraction peak intensities of the
former samples are all relatively lower than the latter one. This is mainly caused by the small
framework thickness along the bꢀaxis direction in the hierarchical lamellar MFIꢀSn/Al zeolites
20
compared to that in the three dimensional microporous CꢀMFIꢀSn/Al (100/100) zeolite.
11

Journal of Materials Chemistry A
Page 12 of 40
View Article Online
DOI: 10.1039/C8TA01242C
2
2
47
48
Fig. 1 XRD patterns of (a) MFIꢀSn/Al (∞/100), (b) MFIꢀSn/Al (50/50), (c) MFIꢀSn/Al (100/25),
2
49
(d) MFIꢀSn/Al (100/50), (e) MFIꢀSn/Al (100/100) and (f) CꢀMFIꢀSn/Al (100/100), respectively.
2
2
2
2
2
2
2
2
2
2
2
50
51
52
53
54
55
56
57
58
59
60
The morphologies of the hierarchical lamellar MFIꢀSn/Al zeolites together with CꢀMFIꢀ
Sn/Al (100/100) and MFIꢀSn/Al (∞/100) samples are shown by the SEM images in Fig. 2 and
Fig. S1 in the supplementary information. All the MFIꢀSn/Al zeolites (Figs. 2(a)ꢀ(e)) are
comprised of interweaved plateletꢀlike structure that formed flowerꢀlike aggregates with irregular
sizes (Fig. S1), while the CꢀMFIꢀSn/Al (100/100) sample is composed of uniform but irregular
shaped particles, whose sizes are in the range of 150 – 200 nm (Fig. 2(f)). A closer examination
on the MFIꢀSn/Al zeolite samples shows that the width and thickness of the plateletꢀlike
structure decreases with increasing Al content in the synthesis recipe (Figs. 2aꢀbꢀc). Moreover,
the plateletꢀlike structures change from curved to straight ones with increasing Al content.
Similarly, the increase in Sn content in the synthesis recipe decreases the thickness of the zeolite
plateletꢀlike structure in the MFIꢀSn/Al samples (refer to Figs. 2bꢀd). The effect of Si/Al and
12

Page 13 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
2
2
2
2
2
2
2
61
62
63
64
65
66
67
Si/Sn ratios in the synthesis recipe on the morphology of derived MFIꢀSn/Al zeolites is the same
2
9
as that observed by Machoke et al on synthesis of multilamellar MFI under static condition.
The introduction of aluminum and/or tin into the lamellar zeolite synthesis recipe leads to defects
in zeolite nanosheet structures, which consequently affect the growth of nanosheets along the aꢀc
plane. Therefore, more randomly oriented and smaller sized zeolite nanosheets are formed in the
crystallization process, which results in the thinner and straighter zeolite plates from the
synthetic gels containing higher aluminum and/or tin contents.
2
2
68
69
Fig. 2 SEM images of (a) MFIꢀSn/Al (100/100), (b) MFIꢀSn/Al (100/50), (c) MFIꢀSn/Al
(100/25), (d) MFIꢀSn/Al (50/50), (e) MFIꢀSn/Al (∞/100) and (f) CꢀMFIꢀSn/Al (100/100),
respectively.
2
2
70
71
2
2
2
2
2
72
73
74
75
76
N adsorptionꢀdesorption isotherms were used to study the textural properties of the MFIꢀ
2
Sn/Al zeolite samples. The textural parameters of each sample are included in Table S1 of the
supplementary information. Fig. 3A illustrates that all the synthesized hierarchical lamellar MFIꢀ
Sn/Al zeolites exhibit typeꢀIV isotherms with hysteresis loops corresponding to capillary
2
1,30
condensation in mesopores.
Particularly, at low relative pressures (P/P < 0.45), the
0
13

Journal of Materials Chemistry A
Page 14 of 40
View Article Online
DOI: 10.1039/C8TA01242C
2
2
2
2
2
2
2
2
2
2
77
78
79
80
81
82
83
84
85
86
isotherms of these zeolites are similar and comparable to that of CꢀMFIꢀSn/Al (100/100) zeolite.
This suggests all the synthesized MFIꢀSn/Al samples have similar micropore structures. At high
relative pressures (P/P > 0.45), the isotherms of the MFIꢀSn/Al samples show hysteresis loops,
0
different from that of CꢀMFIꢀSn/Al (100/100). For all the MFIꢀSn/Al samples, the hysteresis
loops in their isotherms are relatively narrow and flat, which indicates these materials have
31,32
disordered mesopores and broad pore size distributions.
Fig. 3B shows the nonlocal density
functional theory (NLDFT) pore size distributions of all MFIꢀSn/Al samples. As suggested from
the shape of hysteresis loops in their isotherms, the MFIꢀSn/Al (100/100), MFIꢀSn/Al (50/50),
MFIꢀSn/Al (100/50), and MFIꢀSn/Al (100/25) zeolites have broad mesopore size distributions. In
contrast, the CꢀMFIꢀSn/Al (100/100) zeolite does not have significant mesoporosity.
2
2
87
88
Fig. 3 N isotherms (A) and the corresponding NLDFT pore size distributions (B) as determined
2
2
2
2
89
90
91
from adsorption branch of (a) CꢀMFIꢀSn/Al (100/100), (b) MFIꢀSn/Al (50/50), (c) MFIꢀSn/Al
(100/25), (d) MFIꢀSn/Al (100/50), (e) MFIꢀSn/Al (100/100) and (f) MFIꢀSn/Al (∞/100),
respectively.
2
2
92
93
Solid state NMR was employed to investigate the local bonding environment of Si and Al
2
9
species in the MFIꢀSn/Al zeolite catalysts. Figs. 4A and 4B show the Si single pulse (SP) and
14

Page 15 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
2
7
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
Al NMR spectra of the MFIꢀSn/Al (100/100), MFIꢀSn/Al (50/50), MFIꢀSn/Al (100/50), and
MFIꢀSn/Al (100/25) zeolites, respectively. In Fig. 4A, one wellꢀresolved peak at ꢀ113 ppm and
one shoulder peak at ꢀ103 ppm can be observed. These two peaks correspond to the
4
n
33ꢀ36
3
crystallographically nonequivalent Q tetrahedral sites (Q stands for X Si[OSi] )
and Q
respectively. The similarity of
Si SP NMR spectra of MFIꢀSn/Al zeolites suggests similar local bonding environments exist in
4
ꢀn
n
2
3,37
sites that arise from the silanol groups on the zeolite surface,
2
9
27
these samples. In Fig. 4B, the two peaks (at 55 ppm and 0 ppm) in the Al MAS NMR spectra
were observed. The peak at 55 ppm is due to the tetrahedrally coordinated framework aluminum
(Al ), whereas the peak around 0 ppm is due to an octahedral coordination typical of extraꢀ
F
3
8ꢀ40
framework Al (AlEF).
The percentages of the AlEF sites in MFIꢀSn/Al (50/50), MFIꢀSn/Al
(100/25), MFIꢀSn/Al (100/50), and MFIꢀSn/Al (100/100) samples were calculated from the
intensities of the peak around 0 ppm over the peak around 55 ppm, which are 3.6%, 8.1%, 4.4%
and 2.0%, respectively. Obviously, the higher Al content in the synthesis led to higher
proportions of Al in MFIꢀSn/Al zeolites. When the concentration of Al was the same, the one
EF
with higher Sn content in the synthesis recipe resulted in a higher proportion of AlEF in the asꢀ
obtained MFIꢀSn/Al samples.
3
10
15

Journal of Materials Chemistry A
Page 16 of 40
View Article Online
DOI: 10.1039/C8TA01242C
2
9
27
3
3
3
11
12
13
Fig. 4 Solidꢀstate Si MAS NMR (A) and Al MAS NMR (B) spectra of hierarchical lamellar
MFIꢀSn/Al zeolite samples: (a) MFIꢀSn/Al (50/50), (b) MFIꢀSn/Al (100/25), (c) MFIꢀSn/Al
(100/50), and (d) MFIꢀSn/Al (100/100), respectively.
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Composition and acidity of MFIꢀSn/Al zeolite catalysts
The tin (Sn) and aluminum (Al) contents of the MFIꢀSn/Al zeolite samples were analyzed by
ICPꢀAES technique, represented by Si/Sn and Si/Al ratios, as shown in Table S2 of the
supplementary information. The Sn and Al concentrations of the synthesized MFIꢀSn/Al samples
are generally higher than those determined from the synthesis recipe. Table 1 lists the
concentrations of Al and Sn sites determined by elemental analysis and the concentration of
Brønsted acid sites measured by RGC method. The concentration of Al sites is higher than
Brønsted acid sites, suggesting that there is a considerable amount of nonꢀframework Al in the
synthesized MFIꢀSn/Al zeolite samples. If we assume all the Al species that could not be
detected by the RGC method as Lewis acid site (L(Al)), the ratio of the Brønsted to Lewis (Al)
acid site can be calculated, as shown by entry 6 in Table 1. The overall Brønsted to Lewis acid
site ratio (entry 7 in Table 1) in each zeolite sample can also evaluated by division of Brønsted
acid sites with the sum of Lewis (Al) acid and Sn sites. Apparently, the increase in Al or Sn
content in the hydrothermal synthesis increased more Lewis acid sites compared to Brønsted acid
sites in the MFIꢀSn/Al samples.
Table 1 Concentration of Sn, Al and acid sites in MFIꢀSn/Al zeolite catalysts.
a
a
Zeolite
Sn
Al
Brønstedb
Lewis
(Al) acid
site (L(Al)
(mmol/g)
Brønsted/Lewis Brønsted/Lewis Brønsted/Lewis
f
(
mmol/g)
(mmol/g) acid site
(Al) acid site
(Al + Sn) acid
site ratio
acid site ratio
(I /I )
c
,
d
e
(
mmol/g)
0.141
0.239
0.274
)
ratio
B L
(I /IL(Al))
B
(I /IL (Al+Sn))
B
MFIꢀSn/Al
0.219
0.317
0.299
0.245
0.505
0.794
0.104
1.351
0.895
0.528
0.436
0.410
0.334
0.989
0.893
0.744
(
100/100)
MFIꢀSn/Al
100/50)
MFIꢀSn/Al
100/25)
0.266
0.520
(
(
16

Page 17 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
MFIꢀSn/Al
50/50)
MFIꢀSn/Al
0.523
0
0.438
0.228
0.121
0.159
0.317
0.069
0.379
ꢀ
0.144
ꢀ
0.464
ꢀ
(
(
∞/100)
a
b
3
3
3
3
3
30
31
32
33
34
Calculated from elemental analysis (ICPꢀAES); Determined from RGC by selective decomposition of isopropyl
c
amine on Brønsted acid sites; Determined from deduction of concentration of Brønsted acid sites determined by
d
RGC from Al concentration measured by elemental analysis; Calculated by ratio of data in column 4 relative to
that in column 5; Calculated by ratio of data in column 4 relative to the sum of data in columns 2 and 5; Evaluated
by FITR spectra of adsorbed pyridine. I /I is the ratio of peak intently of 1545 cm relative to that of 1450 cm .
e
f
ꢀ1
ꢀ1
B
L
3
35
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
To understand the types of acidity, the FTIR spectra of the OHꢀstretching mode (ν(OH)) and
adsorbed pyridine in the MFIꢀSn/Al samples were recorded, as presented in Fig. 5(A) and (B),
ꢀ1
ꢀ1
ꢀ1
respectively. In Fig. 5(A), there are four peaks centered around 3739 cm , 3676 cm , 3605 cm ,
ꢀ1
and 3501 cm , in sequence, which are associated with terminal silanol (SiꢀOH) groups, Lewis
acidic or nonꢀacidic extraframework AlꢀOH, Bronsted acid sites (SiꢀO(H)ꢀAl), and hydrogenꢀ
4
1ꢀ45
bonded internal silanol (SiꢀOH) groups.
In the absence of Al in MFIꢀSn/Al (100/∞) zeolite,
ꢀ1
the peak of SiꢀO(H)ꢀAl does not exist, in contrast to the 3605 cm peak in all other MFIꢀSn/Al
samples. The same trend is observed in AlꢀOH sites in all these samples. The increase in Al
(sample (b)ꢀ(d) or Sn (sample (a) and (c)) content increased the AlꢀOH and (SiꢀO(H)ꢀAl) sites, as
ꢀ1
ꢀ1
indicated by increasing intensity of 3676 cm , 3605 cm peaks. The control sample, commercial
ꢀ1
MFI, and the conventional microporous MFIꢀSn/Al (100/∞) both contain a big peak at 3500cm ,
indicating the presence of internal SiꢀOH groups in these samples. The hierarchical MFIꢀSn/Al,
apparently, have quite low amount of such internal groups. Fig. 5(B) presents the FTIR spectra
of adsorbed pyridine on zeolites, which hints the relative quantity of Brønsted to Lewis acid sites
4
6
in the MFIꢀSn/Al samples. MFIꢀSn/Al (50/50), MFIꢀSn/Al (100/25), MFIꢀSn/Al (100/50) and
ꢀ1
MFIꢀSn/Al (100/100) all showed characteristic peaks for Brøsnted (1545 cm ) and Lewis (1450
ꢀ1
cm ) acid sites. MFIꢀSn/Al (100/∞) only showed the peak of Lewis acid sites due to the absence
3
53
of Al in the structure. The difficulty in determining accurate molar extinction coefficients of
17

Journal of Materials Chemistry A
Page 18 of 40
View Article Online
DOI: 10.1039/C8TA01242C
3
3
3
3
3
3
3
3
3
54
55
56
57
58
59
60
61
62
solid acid materials in DRIFT mode restricts our capability of calculating the accurate number of
4
7,48
Brøsnted and Lewis acid sites in each sample.
The intensities of the signals in Fig. 5(B), due
to both types of acid sites, are proportional to their concentrations. Therefore, we conducted
semiꢀquantitative analyses of the ratio of Brønsted to Lewis acid sites in MFIꢀSn/Al zeolites by
calculating the intensity ratio of Brønsted relative to the Lewis (I /I ) acid sites in these samples.
B
L
As listed in Table 1, the I /I decreases with increasing Al content from samples (b)ꢀ(d). The
B
L
increase in Sn content from samples (a) and (c) enhanced the Lewis acidity. These results are
consistent with the type of acidity determination by the combination of elemental analysis and
RGC measurement in Table 1.
3
3
3
3
3
3
3
3
3
3
3
3
3
3
63
64
65
66
67
68
69
70
71
72
73
74
75
76
FTIR spectroscopy of adsorbed deuterated acetonitrile can be used for characterization of
4
9,50
51
the framework and extraꢀframework Lewis acid Sn sites in SnꢀBeta
or SnꢀMFI zeolites. We
also used this technique to understand the structure of Sn sites existing in the synthesized MFIꢀ
ꢀ1
Sn/Al zeolites. As shown in Fig. 5(C), three peaks at 2266, 2273 and 2316 cm are observed,
4
9
which are assigned to the CꢀN stretching vibration of the adsorbed deuterated acetonitrile . The
ꢀ1
band at 2266 and 2273 cm corresponds to the deuterated acetonitrile physisorbed and adsorbed
ꢀ1
on silanol groups. The band at 2316 cm can be assigned to the stronger deuterated acetonitrile
49
50
–1
adsorption on Lewis acid sites . According to the previous report , the features near 2316 cm
are clearly associated with framework Sn sites, which confirmed the presence of framework Sn
in the zeolite structures. Furthermore, the structure of Sn sites was evaluated by the UVꢀvisꢀDRS
spectra, as shown in Fig. 5(D). The absorbance around 210 nm was observed for all the MFIꢀ
4
+
Sn/Al samples, showing the presence of isolated, tetrahedral Sn species within the zeolite
4
+
framework. The increase in Sn content in the synthesis increased the Sn sites in the resultant
MFIꢀSn/Al zeolites, as indicated by the sample (a) and (c) in Fig. 5(D). The increase in Al
18

Page 19 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
3
3
3
3
77
78
79
80
content in zeolite synthesis affected the incorporation of Sn sites into the framework, as reflected
by the increase in 210 nm peak intensity from sample (b), (c) and (d). Apparently, no Sn oxide is
observed in all MFIꢀSn/Al zeolite samples, as indicated by the absence of characteristic peak of
SnO sample.
2
(
A)
-1
(B)
3
739cm
MFI-Sn/Al (100/ )
-1
-
1
-
1
-1
3676cm
1545cm
1450cm
-
1
3501cm
3
605cm
(
d)
c)
(
(d)
(c)
(b)
(
b)
a)
Commercial MFI
800 3700 3600 3500 3400 3300
(
a)
(
MFI-Sn/Al (100/ )
3
1575 1550 1525 1500 1475 1450 1425
ꢀ1
ꢀ1
cm
Wavenumber
(cm
)
Wavenumber
(
)
3
81
-
1
-1
(D)
(
C)
2273cm 2266cm
210nm
0.4
(a)
(
b)
(
-1
2316cm
0.2
c)
(d)
SnO2
(
d)
(c)
(
b)
(
a)
0.0
2360
2320
2280
2240
2200
200 250 300 350 400 450 500
ꢀ
1
Wavenumber
(
cm
)
Wavelength (nm)
3
3
82
83
Fig. 5 (A) DRIFTS spectra of MFIꢀSn/Al zeolites; FTIR spectra of pyridine (B) and deuterated
acetonitrile (C) adsorbed on the MFIꢀSn/Al zeolites; and (D) DRꢀUVꢀvis spectra of MFIꢀSn/Al
samples, respectably. (Denotation in each figure: (a) MFIꢀSn/Al (50/50), (b) MFIꢀSn/Al (100/25),
(c) MFIꢀSn/Al (100/50) and (d) MFIꢀSn/Al (100/100). MFIꢀSn/Al (100/∞) stands for
3
3
3
84
85
86
19

Journal of Materials Chemistry A
Page 20 of 40
View Article Online
DOI: 10.1039/C8TA01242C
3
3
3
87
88
89
conventional microporous MFI comprised of only Sn sites. The commercial MFI with Si/Al ratio
of 40 (Supplier No. 45882ꢀ36) in Fig. 5(A) and SnO (Supplier No. 44606ꢀ18) in Fig. 5(D) were
2
purchased from Alfa Aesar.)
3
3
3
3
3
3
3
3
3
3
4
4
4
90
91
92
93
94
95
96
97
98
99
00
01
02
The acidity of the MFIꢀSn/Al zeolite catalysts were further characterized by the NH ꢀTPD
3
measurement, and the results are shown in Fig. 6. Two desorption peaks located in the
temperature range of 400ꢀ550 K (low temperature desorption peak) and 550ꢀ750 K (high
temperature desorption peak), respectively, are observed in all the tested zeolite samples. It has
been generally reported that the low temperature peak is associated with the physical and Lewis
acid site adsorption of NH , while the high temperature desorption peak is assigned to the
3
5
2ꢀ54
adsorption of NH on the Brøsnted acid sites in the catalysts
although caution must be paid
,
3
since desorption of NH from nonꢀacidic sites in catalysts scan take place over a wide
3
5
5,56
temperature range.
Additionally, the intensity is correlated to the quantity of acid sites in the
5
7
zeolite catalysts. Fig. 6 shows that the low and high temperature desorption peaks are increased
with increasing Al sites from MFIꢀSn/Al (100/∞), MFIꢀSn/Al (100/100), MFIꢀSn/Al (100/50) to
MFIꢀSn/Al (100/25) materials. The increase in Sn sites mainly increases the low temperature
peak, as the zeolite samples transit from MFIꢀSn/Al (∞/100) to MFIꢀSn/Al (100/100) and MFIꢀ
4
4
03
04
Sn/Al (50/50) to MFIꢀSn/Al (100/50). The results suggest that Al in the MFIꢀSn/Al zeolites led
to both Lewis and Brønsted acidity, while Snꢀsite predominately contributes to the Lewis acidity.
20

Page 21 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
MFI-Sn/Al ( /100)
MFI-Sn/Al (100/
MFI-Sn/Al (50/50)
)
MFI-Sn/Al (100/100)
MFI-Sn/Al (100/50)
MFI-Sn/Al (100/25)
400
500
600
700
800
Temperature (K)
4
05
4
4
4
06
07
08
Fig. 6 NH ꢀTPD profiles of MFIꢀSn/Al zeolites with different Si/Al and Si/Sn ratios. (MFIꢀSn/Al
3
(∞/100) and MFIꢀSn/Al (100/∞) stand for the conventional microporous MFI zeolites containing
only Al and Sn sites, respectively.)
4
4
4
4
4
4
4
4
4
4
4
09
10
11
12
13
14
15
16
17
18
19
Reaction cascade for EMF synthesis from carbohydrates over MFIꢀSn/Al zeolite catalysts
As shown in Scheme 1, the EMF synthesis from glucose carbohydrate occurs via a threeꢀstep
reaction cascade, in which the first step is isomerization of glucose into fructose and the next two
steps are dehydration of fructose to HMF and etherification of HMF to EMF final product. The
first step reaction can be catalyzed by the Lewis acidic Sn sites and the last two steps can take
place over the Brønsted acidic Al–O(H)–Si sites in the MFIꢀSn/Al zeolite catalysts. The acidity
characterizations in Section 3.2 indicate the coꢀexistence of both Sn and Al sites, i.e., both Lewis
and Brønsted acid sites, in the MFIꢀSn/Al zeolites. We expect that they are efficient catalysts for
this threeꢀstep reaction cascade. Moreover, the morphological and textural property
characterizations in Section 3.1 indicates the presence of mesoporosity in the synthesized MFIꢀ
Sn/Al zeolites, which could be beneficial for the reaction cascade compared to the conventional
21

Journal of Materials Chemistry A
Page 22 of 40
View Article Online
DOI: 10.1039/C8TA01242C
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
microporous zeolite analogue. We, therefore, studied the performance of the MFIꢀSn/Al zeolite
catalysts in EMF synthesis from carbohydrate reactants. The effects of catalyst structure, reaction
temperature, catalyst acidity and carbohydrate reactant type on EMF synthesis were examined in
sequence via a series of catalytic reactions. The reaction conditions used for EMF synthesis from
glucose reactant over MFIꢀSn/Al (100/100) catalyst at 413 K were used as the base case for
comparison. Individual experimental variables were changed, and the effects of changing these
variables were investigated by comparing the catalysis data to that obtained with the standard
conditions, as discussed below.
EMF synthesis from glucose over MFIꢀSn/Al (100/100), CꢀMFIꢀSn/Al (100/100) and MFIꢀ
Sn/Al (∞/100) catalysts
The synthesis of EMF from glucose was firstly tested over the MFIꢀSn/Al (100/100), CꢀMFIꢀ
Sn/Al (100/100), and MFIꢀSn/Al (∞/100) zeolite catalysts. Figs. 7aꢀc show the glucose
conversion and product yields versus the reaction time over each catalyst. It should be noted that
the products are mainly comprised of fructose, HMF and EMF in the glucoseꢀtoꢀEMF conversion,
and thus only these major products were analyzed in the present study. Fig. 7d shows the EMF
yield at the reaction time of 9 h over each tested catalyst. As demonstrated in Figs. 7aꢀc, the
glucose conversion was higher in MFIꢀSn/Al (100/100) than CꢀMFIꢀSn/Al (100/100), so did the
EMF yield (Fig. 7d). A closer examination on the kinetic data (e.g., reaction time: < 5 h) in Figs.
7a and b shows that the EMF formation took place at a much faster rate in MFIꢀSn/Al (100/100)
than CꢀMFIꢀSn/Al (100/100) catalyst. These results indicate that mesoporosity in the hierarchical
lamellar MFI zeolite plays an important role in enhancing this threeꢀstep reaction cascade. As
5
8
reported previously, the Stokes diameter of glucose is 0.73 nm, larger than the micropore sizes
of MFI zeolite. As a result, the reaction mainly takes place on the external surface of
22

Page 23 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
4
4
4
4
4
4
4
4
4
4
4
4
43
44
45
46
47
48
49
50
51
52
53
54
conventional CꢀMFIꢀSn/Al (100/100) zeolite. The presence of mesopores in MFIꢀSn/Al (100/100)
apparently promotes the access of reactant to acid sites in interior of MFIꢀSn/Al (100/100)
zeolite. The mesopores also enhance the transport of reactant onto or product (or reaction
intermediate) away from the active sites in the catalyst in these three steps in the reaction cascade.
The sole presence of Al sites in the hierarchical lamellar MFI zeolite (MFIꢀSn/Al (∞/100)) also
enabled glucose conversion, but the conversion was lower than MFIꢀSn/Al (100/100) and
produced nearly zero EMF production (Figs.7c and d). Both glucose conversion and EMF yield
suggest the efficiency of Lewis Sn sites in catalysts for glucose isomerization, as well as the
dehydration and etherification steps in the threeꢀstep reaction cascade. Additionally, the
comparison between the performances of MFIꢀSn/Al (100/100) and MFIꢀSn/Al (∞/100) also
suggests the efficiency of the cooperative catalysis offered from both types of acid sites in this
threeꢀstep reaction cascade.
23

Journal of Materials Chemistry A
Page 24 of 40
View Article Online
DOI: 10.1039/C8TA01242C
4
4
55
56
Fig. 7 Catalytic conversion of glucose to EMF over (a) MFIꢀSn/Al (100/100), (b) CꢀMFIꢀSn/Al
(100/100) and (c) MFIꢀSn/Al (∞/100), respectively, versus reaction time. (d) shows the EMF
yield at reaction time of 9 h across these three zeolite catalysts. (reaction condition: glucose:
0.090 g; zeolite: 0.074 g; temperature: 413 K; ethanol: 3 mL).
4
4
4
57
58
59
4
4
4
4
4
60
61
62
63
64
Effect of reaction temperature on EMF synthesis from glucose over MFIꢀSn/Al (100/100)
catalyst
The effect of reaction temperature on glucose conversion to EMF over the MFIꢀSn/Al zeolite
catalysts was examined by employing the MFIꢀSn/Al (100/100) catalyst at the reaction
temperature of 398 K, 406 K, and 413 K, respectively. Fig. 8 shows the glucose conversion and
24

Page 25 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
4
4
4
4
4
4
4
4
4
4
4
4
65
66
67
68
69
70
71
72
73
74
75
76
product yields as a function of the reaction time (Figs. 8aꢀc) together with the EMF yield at the
reaction time of 9 h (Fig. 8d) at these three reaction temperatures. By increasing the reaction
temperature from 398 K to 413 K, the rate of glucose conversion increased, so did the fructose
conversion and EMF production (refer to reaction time of 0ꢀ5 h in Figs. 8aꢀc). The changes of
HMF yields follow the same trend of fructose. This indicates the successive threeꢀstep reaction
nature in glucoseꢀtoꢀEMF reaction cascade, besides the endothermic nature of these three
reaction steps. The maximum EMF yields in the course of this reaction are slightly increased
with an increase in the reaction temperature. As shown in Fig. 8d, the EMF yield was increased
from 23.2% to 24.5% to 26.0% when the reaction temperature was increased from 398 K to 406
K and then to 413 K. These results indicate that higher temperature increases the conversion of
glucose reactant and the successive step of dehydration of fructose in the threeꢀstep reaction
cascade.
25

Journal of Materials Chemistry A
Page 26 of 40
View Article Online
DOI: 10.1039/C8TA01242C
4
4
77
78
Fig. 8 Effect of reaction temperature ((a) 398 K, (b) 406 K, and (c) 413 K, respectively) on
catalytic conversion of glucose to EMF over MFIꢀSn/Al (100/100). (d) shows the maximum
EMF yield at the reaction process at these three temperatures. (reaction condition: glucose: 0.090
g; zeolite: 0.074 g; ethanol: 3 mL).
4
4
4
79
80
81
4
4
4
4
4
4
82
83
84
85
86
87
Effect of zeolite acidity on EMF synthesis from glucose over MFIꢀSn/Al catalysts
The influences of Sn and Al contents (i.e., acidity, represented by Si/Sn and Si/Al ratios) in MFIꢀ
Sn/Al zeolite catalysts on their catalytic performances in the glucoseꢀtoꢀEMF reaction was
studied at 413 K. Fig. 9 shows the product yields from the reaction over MFIꢀSn/Al (100/100),
MFIꢀSn/Al (100/50), MFIꢀSn/Al (100/25), and MFIꢀSn/Al (50/50) catalysts, respectively. As
indicated in this figure, all the MFIꢀSn/Al zeolites enabled EMF, HMF and fructose production.
26

Page 27 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
The yields of these three products follow the order of EMF > fructose > HMF in each catalyst,
suggesting the etherification of HMF to EMF happened much faster than dehydration of fructose
in this reaction cascade. Fig. 9 also shows that the HMF and EMF yields from this reaction
cascade increases in the order of MFIꢀSn/Al (100/100) < MFIꢀSn/Al (100/50) < MFIꢀSn/Al
(100/25) << MFIꢀSn/Al (50/50). The fructose yield also follows the same order as that of EMF
and HMF over the MFIꢀSn/Al (100/100), MFIꢀSn/Al (100/50) and MFIꢀSn/Al (100/25) catalysts
that have similar number of Sn sites but different Al sites. The increase in Al content increases
Brønsted acidity of the MFIꢀSn/Al catalysts, which benefits etherification of HMF to EMF
reaction step. Additionally, this acidity enhances dehydration of fructose to HMF and
isomerization of glucose to fructose reaction steps. It should be noted that the glucose conversion
happened at a slower rate with increasing Al content in MFIꢀSn/Al catalysts (refer to Figs. S2aꢀ
bꢀc in Section S2 of the Supplementary Information). The controversy between the glucose
conversion and fructose/HMF/EMF formation with increasing Al content in the MFIꢀSn/Al
catalysts suggests that the increased Brønsted acidity affected the performance of Lewis acidity
from Sn sites, i.e., promoted fructose formation from glucose isomerization and suppressed
glucose conversion to side products and thus glucose conversion rate, which overall might be the
consequences of cooperative catalysis between both types of acidity in a single catalyst for this
threeꢀstep reaction cascade.
The comparison for glucose conversion and product yields between MFIꢀSn/Al (100/50)
and MFIꢀSn/Al (50/50) catalysts in Fig. 9 and Fig. S2 (Section S1 in supplementary information)
shows the effect of Sn sites (i.e., Lewis acidity) on the performance of the hierarchical lamellar
MFIꢀSn/Al catalysts. The increase in Sn sites promoted the glucose conversion as well as HMF
and EMF formation, while the fructose yield was reduced. It is suggested that Sn sites in zeolite
27

Journal of Materials Chemistry A
Page 28 of 40
View Article Online
DOI: 10.1039/C8TA01242C
5
9
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
function as Lewis acidity that is beneficial for isomerization of glucose to fructose product, and
thus the higher glucose conversion and fructose formation are expected with increasing Sn sites
while keeping Al site concentration in the MFIꢀSn/Al catalysts. The discrepancy for observed
lower fructose formation (or yield) and higher glucose conversion as well as higher HMF and
EMF formation in this threeꢀstep reaction cascade hints that the cooperative catalysis between
the Brønsted acidic Al–O(H)–Si sites and Lewis acidic Sn sites might exist for this observed
catalytic results. The nearly complete conversion of glucose over each catalyst (Section S2 of the
Supplementary Information) suggests that Sn sites in each MFIꢀSn/Al catalyst are efficient for
isomerization of glucose reactant in this reaction cascade. Overall, the increase in Sn or Al
content in the catalyst facilitates the particular reaction step in the reaction cascade, which
consequently influences the whole reaction network and exhibits different catalytic performances.
A good balance between the two types of acid sites can be optimized to realize cooperative
catalysis for further optimizing the EMF yield in glucose conversion in the reaction cascade. As
shown in Fig. 9, the MFIꢀSn/Al (50/50) zeolite showed best performance, with as high as ~44%
EMF yield, among these four MFIꢀSn/Al catalyst formulations.
28

Page 29 of 40
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C8TA01242C
5
5
26
27
Fig. 9 Effect of Al and Sn contents in hierarchical lamellar MFIꢀSi/Al zeolites on conversion of
glucose to EMF at 413 K. (reaction condition: glucose: 0.090 g; zeolite: 0.074 g; time: 9 h;
ethanol: 3 mL).
5
5
28
29
5
5
5
5
5
5
5
5
5
5
30
31
32
33
34
35
36
37
38
39
Effect of carbohydrate type on EMF synthesis over MFIꢀSn/Al (100/100) catalyst
Glucose is one type of monosaccharides and simple carbohydrates in lignocellulosic biomass,
which has been shown the feasibility of conversion into EMF over the MFIꢀSn/Al zeolite
catalysts. In reality, the natural biomass produces various types of saccharides, including other
monosaccharides, disaccharides and polysaccharides. To investigate the applicability of the
hierarchical lamellar MFIꢀSn/Al in enabling reaction cascades for conversion of other types of
biomass feedstock into more valuable EMF product, we purposely studied the performance of
the MFIꢀSn/Al (100/100) catalyst in EMF synthesis from fructose monosaccharide, sucrose
disaccharide and inulin polysaccharides (a polymer of ~24 fructose and 1 glucose units),
respectively. The effect of temperature on conversion of these carbohydrate feedstocks for EMF
29