Silicalite-1 zeolite acidification by zinc modification and its catalytic properties for isobutane conversion

Article abstract of DOI:10.1039/c8ra02467g

A series of Zn-modified Silicalite-1 (S-1) zeolites (Zn_x/S-1) were prepared by the wetness-impregnation method and applied in the catalytic conversion of isobutane. The structure and location of Zn species in Zn_x/S-1 were investigate

Full text of DOI:10.1039/c8ra02467g

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Silicalite-1 zeolite acidification by zinc modification
and its catalytic properties for isobutane
conversion†
Cite this: RSC Adv., 2018, 8, 18663
a
a
Guodong Liu,
Jiaxu Liu,
*
Ning He,a Cuilan Miao,a Jilei Wang,a Qin Xinb
a
and Hongchen Guo
A series of Zn-modified Silicalite-1 (S-1) zeolites (Znx/S-1) were prepared by the wetness-impregnation
method and applied in the catalytic conversion of isobutane. The structure and location of Zn species in
Znx/S-1 were investigated using UV-Vis and N2 physical adsorption. The acidity and origin of the acid
sites in Znx/S-1 were studied by NH3-temperature programmed desorption and Fourier-transform
infrared analysis. The catalytic performance of Znx/S-1 for isobutane conversion was investigated in
a fixed-bed microreactor. In the experiments, the acidity of S-1 zeolite was dramatically increased by
modification with Zn, with both Lewis and Brønsted sites identified in Znx/S-1. The relationship between
Brønsted acid sites and Zn–OH groups on ZnO clusters of Znx/S-1 was also revealed for the first time.
Furthermore, Znx/S-1 catalysts exhibited excellent catalytic performances in both isobutane
dehydrogenation and butene isomerization reactions. A high selectivity of total butene products ranging
from 84.6 to 97.2 was achieved on the catalysts with different Zn loadings. Moreover, the linear
correlation between isobutane conversion and the acid amount (determined by NH3-TPD) confirmed
that the weak-to-medium acid sites in Znx/S-1 should play a key role in isobutane conversion.
Received 21st March 2018
Accepted 14th May 2018
DOI: 10.1039/c8ra02467g
rsc.li/rsc-advances
and aromatization.7–11 This clearly indicates that the acidity of S-
1 zeolite needs to be increased to broaden its catalytic
1 Introduction
applications.
Silicalite-1 (S-1) zeolite is an aluminum-free crystalline silicate
with the same MFI topological framework as the Al-containing
ZSM-5 zeolite. Generally, crystalline S-1 zeolites are chemically
inert because their interior structure contains only unreactive
siloxane (Si–O–Si) bridges.1 The presence of defective sites in
their lattice can facilitate the formation of weak acidic centers
(derived from silanol groups), which have been reported as
active Beckmann rearrangement catalysts.2–4 Although the
amount of weak acid sites and, to some extent, the acidic
strength of S-1 zeolite can be modulated by changing the crys-
tallization conditions or through post-synthesis treatment,2,5,6
its catalytic properties remain limited. In contrast, ZSM-5 is
a strongly acidic zeolite that can catalyze several reactions,
including olen oligomerization, aromatic alkylation, toluene
disproportionation, methanol-to-hydrocarbon conversions
(including MTG, MTO, MTP), and light hydrocarbon cracking
The conversion of light alkanes, such as propane and
isobutane, into their corresponding value-added alkenes has
received much attention. Currently, Pt/Al2O3 and CrxO3/Al2O3
catalysts are employed industrially to produce propylene and
isobutene from the non-oxidative dehydrogenation of propane
and isobutane, respectively.12 However, these catalysts have
drawbacks because Pt is expensive and has limited availability,
while CrxO3/Al2O3 is not environmentally friendly owing to its
toxicity.13 Therefore, the development of novel catalysts with
superior catalytic efficiency for light alkane non-oxidative
dehydrogenation is highly desirable. It has been accepted that
Zn-containing catalysts were active in the dehydrogenation of
alkanes.14–16,19–21 Zn-modied HZSM-5 zeolites (Zn/ZSM-5) have
been reported to be effective catalysts for low-temperature
activation of light alkanes, with the modied Zn species play-
ing a key role in light alkane dehydrogenation.14–16 However,
their commercial use in such processes is impeded by their poor
alkene selectivity with respect to the formation of large amounts
of cracked dry gas byproducts (C1 and C2 hydrocarbons),
resulting from the strong acidic character of HZSM-5 zeolite.17,18
Moreover, zinc titanate catalysts in the form of hydrogels or
mixed oxides have also been mentioned in the dehydrogenation
of light alkanes.19,20 Zinc titanate is an environmentally friendly
material and of low cost. However, the overall activity of the
aDepartment of Catalytic Chemistry and Engineering, State Key Laboratory of Fine
Chemicals, Dalian University of Technology, Dalian 116012, P. R. China. E-mail:
hongchenguo@dlut.edu.cn
bState Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian 116023, P. R. China
† Electronic supplementary information (ESI) available: XRF, SEM and TEM
images, N2 adsorption and desorption isotherms, the relative crystallinity of
catalysts, XRD patterns, the acid amount of catalysts. See DOI: 10.1039/c8ra02467g
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 18663–18671 | 18663

View Article Online
RSC Advances
Paper
catalyst was found to be limited or the selectivity declined Zn-modied HZSM-5 zeolite (crystal size, 100–200 nm; SiO2/
rapidly as a function of reaction time. Recently, isolated Zn2+ Al2O3 molar ratio, 26 : 1) was also prepared using the same
species deposited on a silica support were found to be active in impregnation method (denoted as Zn6.0/ZSM-5).
propane dehydrogenation.21 The high propylene selectivity was
attributed to the absence of Brønsted acid sites on the Zn/SiO2
catalyst. However, the catalyst is not so stable enough that its
2.2 Catalyst characterization
activity only remained half of the initial activity in 12 h.
Powder X-ray diffraction (XRD) patterns were collected on
The activation of propane over Zn-modied S-1 zeolite a Rigaku D/max-2004 diffractometer with a Cu Ka radiation
(denoted as Zn/S-1) has been studied by employing in situ 13C source (40 kV, 100 mA). Measurements were conducted in the 2q
MAS NMR and IR spectroscopic techniques in 2006.15 It was range of 5–80ꢀ with a scanning speed of 2ꢀ minꢁ1. X-Ray Fluo-
proposed that propane was activated through dissociative rescence spectrometry (XRF) was conducted on a Bruker SRS
adsorption over zinc oxide species, followed by propene evolu- 3400 instrument. Field emission scanning electron microscopy
tion and recombinative desorption of dihydrogen. On the other (FE-SEM) images of crystal morphologies were obtained using
hand, the acidity of Zn-modied HZSM-5 zeolite (Zn/HZSM-5) a Hitachi S-4800 microscope. The distribution of Zn species on
was reported to play a key role in the transformation of light the modied catalysts was examined by UV-Vis spectropho-
alkanes. However, few study related to the effect of acid sites of tometry, conducted on a JASCO V-550 spectrometer with BaSO4
Zn/S-1 on alkane dehydrogenation was reported, probably as the reference. High resolution transmission electron
because the acidity of S-1 zeolite is negligible compared with microscopy (HRTEM) measurements were taken on a JEM-
that of HZSM-5 zeolite. Moreover, the inuence of Zn modi- 2100F electron microscope with an acceleration voltage of 200
cation on the acidity of S-1 zeolite and the role of these acid sites kV.
in light alkane conversion is yet unclear. Therefore, a series of
N2 physical adsorption–desorption isotherms were
Zn-modied S-1 catalysts was prepared in this work. The acidity measured on a Micromeritics ASAP 2020 instrument at ꢁ196 ꢀC
and origin of acidity of these Zn/S-1 catalysts were characterized to determine the textural properties. Prior to measurement, the
by NH3-temperature programmed desorption (TPD) and samples (sieve fraction, 380–830 mm) were degassed under
ꢀ
Fourier-transform infrared (FT-IR) spectroscopy analysis. vacuum at 350 C for 6 h. The specic surface area was calcu-
Furthermore, the catalytic properties of these Zn/S-1 catalysts in lated by the Brunauer–Emmett–Teller (BET) method using the
isobutane conversion were investigated using a xed-bed adsorption branch in the P/P0 range 0.10–0.15. The pore
microreactor.
volumes were estimated at the P/P0 point of 0.99. The micro-
porosity and mesoporosity of the samples was discriminated
using the t-plot method.
NH3-TPD was performed on a Quantachrome ChemBet 3000
chemisorb instrument to investigate the total acidity of the
2 Experimental
2.1 Catalyst preparation
S-1 zeolite was synthesized according to a reported procedure catalysts. Sample pellets (150 mg; sieve fraction, 380–830 mm)
for the synthesis of TS-1,22 except that no titanium source was were pretreated under helium (He) at 600 ꢀC for 1 h. Aer
used. Typically, tetraethyl orthosilicate (TEOS; Sinopharm cooling to 100 ꢀC, a mixed gas of 5% NH3 in He was passed
Chemical Reagent Co., Ltd), tetrapropylammonium hydroxide through the catalyst bed to allow ammonia adsorption for
(TPAOH; 25 wt% aqueous solution, Sinopharm), and distilled 30 min. The cell was then purged with He ow (50 mL minꢁ1
)
deionized water were mixed to form a clear solution with a SiO2/ for 30 min to remove all NH3 not chemically adsorbed. The NH3-
TPAOH/H2O ratio of 1 : 0.3 : 30. The solution was hydrolyzed at TPD proles were recordedꢁb1y heating the cell to 600 ꢀC at
ꢀ
60 ꢀC in a water bath for 3 h with magnetic stirring to give a ramping rate of 16 C min
.
a homogeneous sol that was then transferred to a Teon-lined
FT-IR of adsorbed NH3 (FT-IR-NH3) was used to determine
autoclave and crystallized at 170 ꢀC for 24 h. The as- the Lewis and Brønsted acidities of the catalysts. FT-IR-NH3
synthesized precipitate was collected by ltration, washed spectra were collected on a Nicolet 6700 FT-IR spectrometer
several times with deionized water, and dried at 110 ꢀC over- (4 cmꢁ1 optical resolution with one level of zero-lling for the
night. The zeolite products were then calcined at 550 ꢀC for 6 h Fourier transform). First, about 15 mg of sample was pressed
for further characterization or use.
into a self-supporting wafer (diameter, 13 mm) and then treated
Zn-modied zeolite catalysts were prepared using a wetness- in a cuvette in situ under vacuum (10ꢁ3 Pa) at 400 ꢀC for 4 h. NH3
impregnation method. The calcined S-1 zeolite was impreg- adsorption was conducted with admittance of an ammonia ow
nated with an aqueous solution of Zn(NO3)2$6H2O (Sinopharm of 30 mL minꢁ1 into the cuvette for 30 min at ambient
Chemical Reagent Co., Ltd) in a water bath at 80 ꢀC for 3 h. The temperature. The wafer was then subjected to vacuum treat-
impregnated zeolite was then dried at 110 ꢀC for 12 h and ment at 150, 300, and 450 ꢀC for 30 min each. Finally, the
calcined at 550 ꢀC for 6 h to obtain the Zn/S-1 samples. The measured spectra were collected.
desired Zn loadings were achieved by adjusting aqueous solu-
To determine the catalyst acidity of the surface hydroxyl
tion of Zn(NO3)2$6H2O with the appropriate concentration. The groups, FT-IR spectra of surface hydroxyl groups with ammonia
modied S-1 zeolites with different Zn loadings were denoted as absorption (OH-FT-IR-NH3) were recorded using a Nicolet iS10
Znx/S-1, where x is the Zn loading amount in wt%. For FT-IR spectrometer. The samples were pressed into self-
comparative catalytic tests in isobutane conversion, a 6.0 wt% supporting thin wafers (approximately 15 mg) and
18664 | RSC Adv., 2018, 8, 18663–18671
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 1 XRD patterns (A) and the magnified patterns (B) of S-1 (a), Zn1.0/S-1 (b), Zn3.0/S-1 (c), Zn6.0/S-1 (d), Zn8.0/S-1 (e) and Zn12.0/S-1 (f). (The
asterisk labels show the diffraction peaks of ZnO crystal with wurtzite structure).
ꢁ3
ꢀ
decontaminated at 400 C under vacuum (10 Pa) for 4 h in
a quartz IR cell equipped with CaF2 windows. An ammonia gas
stream (30 mL minꢁ1) was passed through the sample cell at
150 ꢀC for 30 min to allow ammonia adsorption. Evacuation
treatment (10ꢁ3 Pa) was conducted for 30 min at the same
temperature. Aer cooling to room temperature, OH-FT-IR-NH3
spectra were recorded by subtracting the blank background
spectrum (recorded using an empty IR cell without sample).
Spectra were recorded in the range 4000–400 cmꢁ1 with a reso-
3 Results and discussion
3.1 Catalyst characterization
SEM, XRD, and N2 physical adsorption–desorption isotherms
were measured to assess the effect of the Zn dopant on the
structures of the Znx/S-1 catalysts. The SEM images (see
Fig. S1†) showed that the mean size of parent S-1 zeolite was
around 100 nm. The introduction of Zn species by impregna-
tion has little inuence on the crystal size and the morphology
of parent S-1 zeolite. Structural changes in Znx/S-1 catalysts were
also analyzed using XRD patterns. As shown in Fig. 1, ve
characteristic diffraction peaks of MFI structure were found at
2q ¼ 7.9, 8.8, 23.1, 23.8 and 24.3 in all catalysts. No diffraction
signal for Zn species was observed at Zn content below 8.0 wt%,
indicating that Zn species were highly dispersed. Nevertheless,
characteristic peaks of the ZnO phase with a wurtzite structure23
(denoted with asterisks in Fig. 1A and B) were presented in
Zn12.0/S-1. When the amount of Zn added was too high, Zn
species seemed not to be uniformly dispersed in the zeolite
channels, but formed large wurtzite-structured ZnO particles on
the zeolite surface. The average particle size of ZnO particles
was estimated to be 4.8 nm by using the Scherrer's equation.
Fig. S2† shows the HRTEM images of Zn6.0/S-1 and Zn12.0/S-1
catalysts. There are a number of ZnO particles reside on the
external surface of Zn12.0/S-1 catalyst, while no clear pattern of
lution of 4 cmꢁ1
.
2.3 Catalytic tests
Isobutane conversion tests were performed in a xed-bed
microreactor (i.d., 6 mm) under atmospheric pressure. The
catalyst loading for each run was about 200 mg. Prior to activity
tests, catalysts (dp ¼ 20–40 mesh) were activated in situ at 550 ꢀC
for 1 h under nitrogen ow, and then prereduced under 5% H2
in N2 gas ow (50 mL minꢁ1, standard temperature and pres-
sure) at 500 ꢀC for 30 min. The isobutane feedstock was diluted
with nitrogen (employed as a calibrating gas) to give a 50 : 50 (v/
v) mixture. The ow rates of isobutane and nitrogen were
controlled separately using mass ow meters. For catalytic
activity tests, the mixing gas was fed at a gas hourly space
velocity (GHSV) of about 1200 hꢁ1, with the reaction conducted
at 550 ꢀC and atmospheric pressure. Effluent hydrocarbon
products were analyzed using on-line gas chromatography (GC;
Shimadzu GC-14B, OV-1 capillary column 50 m ꢂ 0.20 mm,
ame ionization detector).
ZnO particles has been observed on the outside surface of Zn6.0
/
S-1. Moreover, in comparison with parent S-1 zeolite, the rela-
tive crystallinity of Znx/S-1 catalysts gradually decreased (see
Table S2†), which was probably attributed to the high X-ray
absorption coefficient of impregnated metal species.24,25 The
textural properties of Znx/S-1 catalysts were determined from N2
physical adsorption–desorption, as summarized in Table 1. At
low Zn loadings (1.0 wt% and 3.0 wt%), the specic surface area
of Znx/S-1 decreased slightly with increasing Zn loading. This
indicated that subnanometric ZnO clusters inside the zeolitic
pores or channels only had a small effect on the textural prop-
erties of the S-1 zeolite. However, at high Zn loadings (above
The isobutane conversion (Cisobutane) and product selectivity
(Si) were calculated using the following equations:
P
P
Cisobutane ¼ ( Ai ꢁ Aisobutane)/ Ai ꢂ 100%
(1)
(2)
P
Si ¼ Ai/( Ai ꢁ Aisobutane) ꢂ 100%
where Ai and Aisobutane are the corrected chromatographic areas
of a specic compound and residual isobutane, respectively.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 18663–18671 | 18665

View Article Online
RSC Advances
Paper
Table 1 N2 adsorption–desorption data of Znx/S-1 catalysts
Zn loading
(wt%)
SBET (m2 gꢁ1
)
SMicro (m2 gꢁ1
)
Sextern (m2 gꢁ1
)
Vpores (cm3 gꢁ1
)
Vmicro (cm3 gꢁ1
)
Vmeso (cm3 gꢁ1
)
0.0
1.0
3.0
6.0
8.0
12.0
390
383
370
325
290
269
239
239
228
245
217
195
150
143
141
80
72
73
0.27
0.27
0.26
0.20
0.20
0.21
0.11
0.11
0.10
0.10
0.10
0.09
0.16
0.16
0.15
0.10
0.10
0.12
6.0 wt%), the specic surface area of Znx/S-1 declined rapidly is shown in Table S3.† For the parent S-1 used herein, the
with increasing Zn loading, mainly due to decreases in external maximum Zn loading was around 6.0–8.0 wt%. When the Zn
surface area (from 150 m2 gꢁ1 to 72 m2 gꢁ1) and mesoporosity loading was low (below 3.0%), the increase in Zn loading
(from 0.16 m3 gꢁ1 to 0.10 m3 gꢁ1). Furthermore, the micro- dramatically increased the acidity of the S-1 zeolite. However,
structure characteristics were slightly disrupted. These data when the Zn loading was high (in the range 3.0–8.0 wt%), the
showed that large ZnO particles were mainly deposited on the overall acid amount in the modied S-1 zeolite reached its
external surface and blocked mesopore openings in S-1 zeolite. maximum value, with the increasing trend with Zn loading
UV-Vis spectra of Znx/S-1 catalyst are shown in Fig. 2. becoming marginal.
Generally, the modication of S-1 with Zn leads to the appear-
The acidity of Znx/S-1 zeolites was further studied using
ance of two new bands at about 265 and 365 nm, corresponding ammonia-adsorbed FT-IR spectra. As shown in Fig. 5A, the
to subnanometric ZnO clusters inside zeolitic pores and large strong absorption peak at 1620 cmꢁ1 was attributed to asym-
ZnO crystals on the external zeolite surface, respectively.15,23 For metric deformation of ammonia coordinated to Lewis acid
Zn1.0/S-1 and Zn3.0/S-1 samples, the band at 265 nm was sites, which usually appears in the bending vibration region of
signicant, while the band at 365 nm was almost nonexistent, 1700–1300 cmꢁ1 and indicates that Zn modication resulted in
suggesting that the two catalysts mainly comprised sub- the formation of an appreciable amount of Lewis acid sites in
nanometric ZnO clusters. In contrast, both bands were present Znx/S-1.26–29 The variation trend in the intensity of this peak was
in samples with Zn loadings above 6.0 wt%. Increasing Zn in agreement with the change in the NH3-TPD results. However,
loading in these catalysts enhanced both bands, with the the existence of Brønsted acid sites in Znx/S-1 required further
intensity of the band at 365 nm increasing much faster. These conrmation because the IR bands corresponding to ammo-
observations showed that the Znx/S-1 samples with higher Zn nium ion deformation vibrations were obscure. In contrast, as
loadings (x ¼ 6.0–12.0) contained not only ZnO clusters inside shown in Fig. 5A and the magnied range of 3500–3000 cmꢁ1 in
S-1 zeolitic pores, but also large ZnO crystals located on the Fig. 5B, in addition to the apparent bands in the range 3350–
external surface. The increased Zn loading mainly facilitated 3200 cmꢁ1, which were attributed to ammonia coordination on
the formation of large ZnO crystals.
Lewis acid sites, two weak yet clear absorption peaks at 3220
and 3190 cmꢁ1 emerged in the N–H stretching region.26,28 It was
clear that Zn modication not only produced a number of Lewis
3.2 Acidity enhancement by Zn modication
The acid amount and strength of S-1 and Zn-modied sample
Zn6.0/S-1 were investigated using NH3-TPD. For comparison, the
acidities of HZSM-5 (SiO2/Al2O3 molar ratio, 26 : 1) and Zn6.0
/
HZSM-5 were also tested. As shown in Fig. 3A, the whole NH3
desorption prole for the parent S-1 zeolite was approximately
a straight line. Almost no NH3 molecules were adsorbed on S-1,
demonstrating its weak acidic character compared with that of
HZSM-5. Notably, the NH3 desorption peak of the Zn6.0/S-1
sample was quite strong, although the acid amount remained
much lower than that of HZSM-5 and Zn6.0/HZSM-5. This clearly
showed that impregnation with Zn signicantly enhanced the
acidity of the S-1 zeolite. We also compared the NH3-TPD
proles of the Znx/S-1 catalysts, which exhibited two distinct
desorption peaks at 267 and 375 ꢀC (see Fig. 3B), corresponding
to weak and medium acidic sites, respectively. Furthermore, it
was also feasible to gradually increase the amount of acid sites
in the S-1 zeolite by increasing the loading of Zn until the
maximum value was reached, as shown in Fig. 4. Detailed
information on the calculated statistical results of acid amount Fig. 2 UV-Vis spectra of Znx/S-1 (x from 0.0 to 12.0).
18666 | RSC Adv., 2018, 8, 18663–18671
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 3 NH3-TPD profiles of (A) S-1, Zn6.0/S-1, HZSM-5 and Zn6.0/HZSM-5, and (B) the series Znx/S-1 (x from 0.0 to 12.0) samples.
acid sites, but also raised some Brønsted acid sites in Znx/S-1
catalysts, as disclosed by the two N–H stretching signals.
Furthermore, we further detected variations in the spectra with
heating and evacuation procedures, as shown in Fig. 6. The
bands at 3290, 3220, and 3190 cmꢁ1 almost disappeared aer
evacuation at 450 ꢀC for 30 min, suggesting the acid sites of Znx/
S-1 were mostly weak to medium. This was in good agreement
with the NH3-TPD results.
3.3 Origin of Znx/S-1 acidity
3.3.1 Lewis acid sites in Znx/S-1. It is generally accepted
that Zn species14–16,30,31 can provide Lewis acid sites in Zn-
modied ZSM-5 or beta zeolite. Similarly, we speculated that
the Lewis acid sites in Znx/S-1 were also associated with their Zn
species. UV-Vis spectra and FT-IR spectroscopy of adsorbed CO
results showed that Zn/S-1 contained two Zn species: small ZnO
Fig. 4 The relevance of Zn loading and acid amount of S-1 (a), Zn1.0/S-
clusters inside the pore and large ZnO crystals on the external
1 (b), Zn3.0/S-1 (c), Zn6.0/S-1 (d), Zn8.0/S-1 (e) and Zn12.0/S-1 (f),
surface.15 Therefore, the Lewis acid sites of Znx/S-1 should be
respectively.
provided by ZnO clusters or large ZnO crystals. The UV-Vis
Fig. 5 Ammonia-adsorbed FT-IR spectra (A) and the magnified spectra in the NH-stretching region (B) of the series Zn6.0/S-1 Znx/S-1 samples:
(a) S-1, (b) Zn1.0/S-1, (c) Zn3.0/S-1, (d) Zn6.0/S-1, (e) Zn8.0/S-1 and (f) Zn12.0/S-1.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 18663–18671 | 18667

View Article Online
RSC Advances
Paper
ammonia-adsorbed FT-IR spectra clearly indicated a similar
change in the quantity of Lewis acid sites of Znx/S-1 with the
change in peak intensity at 1618 cmꢁ1. The Lewis acid amount
in Znx/S-1 increased dramatically with increasing Zn loading at
low loadings (below 3.0 wt%). This increasing trend slowed as
the Zn loading was increased from 3.0 wt% to 8.0 wt%, and then
began to decrease when the Zn loading reached 12.0 wt%. From
both UV-Vis and ammonia-adsorbed FT-IR spectra, it was easy
to conclude that subnanometer ZnO clusters inside S-1 zeolite
pores were the major contributors of Lewis acid sites in Znx/S-1
catalysts.
3.3.2 Brønsted acid sites in Znx/S-1. As shown in the
hydroxyl stretching vibration region of the FT-IR spectra (Fig. 7),
S-1 zeolite contained isolated silanol groups exposed on the
external surface (3741 cmꢁ1), terminal silanol groups located in
the internal cavities (3728 cmꢁ1), abundant H-bonded silanols
in internal defects as linear chains and/or rings of hydroxylated
species (3500 cmꢁ1), and vicinal silanols inside internal pores
(3684 cmꢁ1).1,2 The appearance of these silanol groups indicated
the presence of a large number of defect sites in S-1 zeolite.
When S-1 zeolite was modied with Zn, these internal silanol
groups (bands at 3728, 3684, and 3500 cmꢁ1) and some of the
isolated silanol groups (3741 cmꢁ1) exposed at the external
surface of S-1 zeolite were found to be dramatically diminished,
suggesting that some Zn species in Znx/S-1 strongly interacted
with the defect sites. Furthermore, a new band at 3678 cmꢁ1 was
observed in Zn3.0/S-1, and redshied to 3670 cmꢁ1 when the Zn
loading was increased from 6.0 to 12.0, which was likely due to
the formation of external Zn–OH groups on the ZnO clusters.15
FT-IR spectra of surface hydroxyl groups with ammonia
absorption were also measured to probe the Brønsted acid sites
of Znx/S-1. To avoid disturbance from hydrogen-bonded species
formed by the interaction of NH3 and intrinsic hydroxyl groups
in S-1 zeolite, as shown in Scheme 1,1 the ammonia-adsorbed
Fig. 6 Ammonia-adsorbed FT-IR spectra of Zn6.0/S-1 in the NH-
stretching region after evacuation at 150, 300 and 450 ꢀC for 30 min.
ꢀ
catalysts were evacuated at 150 C for 30 min. Aer Zn modi-
cation, both Lewis and Brønsted acid sites were found in Znx/S-1
(x ¼ 1.0–6.0), as shown in the (NH) stretching region in Fig. 8B.
This result was consistent with the ammonia-adsorbed FT-IR
spectra (Fig. 5). In contrast, it was interpreted from Fig. 8A
that the residual signals (n¼ 3741, 3728, and 3500 cmꢁ1) of
silicon hydroxyl groups in Znx/S-1 would hardly be disturbed
aer NH3 adsorption, indicating that they had no effect on the
Fig. 7 Infrared hydroxyl stretching region of S-1 (a), Zn1.0/S-1 (b),
Zn3.0/S-1 (c), Zn6.0/S-1 (d), Zn8.0/S-1 (e) and Zn12.0/S-1 (f). Catalysts
have been dehydrated at 400 ꢀC for 4 h.
spectra (Fig. 2) showed that subnanometer ZnO clusters inside
zeolite pores were the primary Zn species in Znx/S-1 at low Zn
loadings. In addition to ZnO clusters, most large ZnO crystals
were generated on the external surface when the Zn loadings
were much higher. In contrast, as shown in Fig. 5A, the
acidity of Znx/S-1. Notably, the bands at 3678 and 3670 cmꢁ1
,
corresponding to Zn–OH groups on ZnO clusters, were signi-
cantly consumed aer NH3 adsorption. The consumption of
these silanol group bands at 3678 and 3670 cmꢁ1 clearly
demonstrated that the Zn–OH groups were associated with
Scheme 1
18668 | RSC Adv., 2018, 8, 18663–18671
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
Fig. 8 (A) IR spectra in the n(OH) and n(NH) stretching region of Znx/S-1 (x from 0.0 to 6.0) before and after the adsorption of NH3, respectively;
(B) corresponding magnified IR spectra in the n(NH) stretching region.
Brønsted acid sites of Znx/S-1. To our knowledge, this rela- experimental conditions is close to that of the commercial
tionship between Brønsted acid sites and Zn–OH groups on Cr2O3/Al2O3 catalyst (48–65%).
ZnO clusters of Zn/S-1 is reported here for the rst time.
Among the products of isobutane transformation over Znx/S-
1 catalysts, isobutene was the primary component (about 60%),
indicating that Zn modication imparted the catalysts impres-
sive dehydrogenation functions. Furthermore, the presence of
substantial amounts of 1-butene and cis/trans-2-butene among
the products indicated that the Brønsted acid sites provided by
the Zn–OH groups in S-1 zeolite also played an important role in
isobutene isomerization. A high selectivity of isobutene and its
isomers ranging from 84.6 to 97.2 was achieved on the catalyst
with different Zn loading. Moreover, a small amounts of
aromatic products (less than 8%), due to aromatization over the
Zn species, and cracked lower hydrocarbon products (C1–C3,
much less than 8%) were formed because the Brønsted acid
strength of Znx/S-1 was still relatively weak. In comparison,
3.4 Catalytic performance
The catalytic conversion of isobutane was tested in a microscale
xed-bed reactor and the results are shown in Table 2. As an
effective strategy, Zn modication signicantly enhanced the
catalytic activity of S-1 zeolite. At 550 ꢀC and atmospheric
pressure, the parent S-1 zeolite was completely inactive for
isobutane transformation. However, under the same reaction
conditions, Znx/S-1 catalysts exhibited promoted isobutane
conversion rates. In particular, the Zn1.0/S-1 and Zn3.0/S-1
catalysts led to 27.5% and 60.2% isobutane conversion,
respectively. As the Zn content was increased (from 3.0 to
8.0 wt%), the isobutane conversion continued increasing,
reaching a maximum isobutane conversion of 66.7%. However,
a slight downward tendency in isobutane conversion was
observed when the Zn loading reached 12.0 wt%. Despite this,
the Zn12.0/S-1 catalyst still afforded a good isobutane conversion
of 65.5%. Moreover, according to the literature,32 the iso-butane
activity of Znx/S-1 (27.5–66.7%) obtained under our
when isobutane transformation was conducted over the Zn6.0
/
HZSM-5 catalyst, the isobutane conversion rated was nearly
100%, with aromatics obtained as the major product (more
than 50%). This reaction also produced a large amount of
undesired cracking components (over 40%), which comprised
dry gas (C1 + C2) and propane in a 2 : 1 ratio. In contrast, the
Table 2 The catalytic performance of iso-butane transformations over S-1, Znx/S-1 and Zn6.0/ZSM-5 catalystsa
Selectivity (%)
Cracking
C1 + C2
Dehydrogenation
Aromatization
Total C]4
Catalyst
Conv. (%)
C3
iC]4
nC]4 + cisC4] + transC]4
Aromatic products
S-1
—
—
—
—
—
—
—
Zn1.0/S-1
Zn3.0/S-1
Zn6.0/S-1
Zn8.0/S-1
Zn12.0/S-1
Zn6.0/HZSM-5
27.5
60.2
61.3
66.7
65.5
99.8
0.6
1.4
2.1
2.7
2.7
27.3
1.6
2.7
4.0
5.1
5.1
13.7
83.8
68.2
64.8
59.8
61.7
0.4
13.4
22.8
22.9
24.8
23.6
1.3
97.2
91.0
87.7
84.6
85.4
1.6
1.0
5.0
6.2
7.6
6.8
57.3
a
Catalyst loading 200 mg, reactions temperature 550 ꢀC, GHSV 1200 hꢁ1, iso-butane feedstock compositions: iso-butane : N2 ¼ 50 : 50 (v/v), the
results were obtained aer 1 h of testing.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 18663–18671 | 18669

View Article Online
RSC Advances
Paper
Fig. 9 (A) Conversion of iso-butane and (B) selectivity of butene as a function of reaction time over Zn6.0/ZSM-5 and Zn6.0/S-1 catalysts.
industrial applications of Zn/SiO2 catalyst since the activity can
only maintain nearly 50% of the initial value during 12 h test.
Therefore, Znx/S-1 catalysts exhibited both higher activity and
excellent stability in comparison with zinc titanate and Zn/SiO2
catalysts in the dehydrogenation of alkanes.
By correlating the isobutane conversion rates with the acid
amount of Znx/S-1 (determined by the NH3-TPD method), we
obtained a linear dependence of isobutane conversion on the
total acid amount of Znx/S-1 (Fig. 10). This meant that the acid
sites of Znx/S-1 played a critical role in isobutane conversion.
The yield of undesired dry gases was signicantly suppressed,
perhaps due to the absence of strong acid sites. Furthermore,
the true role of different categories of acid sites of Znx/S-1 in
isobutane conversion still needs to be identied, with investi-
gations currently underway in our laboratory.
Fig. 10 The relevance between iso-butane conversion and the acid
amount of S-1 (a), Zn1.0/S-1 (b), Zn3.0/S-1 (c), Zn6.0/S-1 (d), Zn8.0/S-1 (e)
and Zn12.0/S-1 (f), respectively. (The acid amounts of catalysts were
determined by NH3-TPD).
4 Conclusions
In summary, the inert S-1 zeolite was signicantly acidied by
modication with Zn. Both Lewis acid sites and a small amount
of Brønsted acid sites were identied in Znx/S-1 catalysts. Using
sum of dry gas and propane byproducts was efficiently sup-
UV-Vis and FT-IR analysis, we found that the Lewis and
pressed over appropriately modied Znx/S-1 catalysts.
Brønsted acid sites of Znx/S-1 were associated with small ZnO
The stability of Zn/S-1 catalyst in the conversion of iso-
clusters and external Zn–OH groups on ZnO clusters, respec-
butane was further investigated. As shown in Fig. 9A, slight
tively. Furthermore, the linear relationship between isobutane
deactivation was observed over Zn6.0/S-1 catalyst during the
conversion and total acid amount suggested that the acid sites
lifespan of 12 h. However, for Zn6.0/HZSM-5 catalyst, the
of Znx/S-1 played a crucial role in isobutene conversion. Znx/S-1
conversion decreased from 99.1% to 72.8% aer 12 h. On the
catalysts exhibited excellent isobutane dehydrogenation and
other hand, the selectivity of total butene was well-preserved
butene isomerization activities, and the yield of undesired dry
over Zn6.0/S-1 catalyst during the 12 h examination (Fig. 9B).
gas (methane and ethane) was signicantly suppressed because
These results indicated that Znx/S-1 catalysts obtain extraordi-
strong acid sites were absent. A high selectivity of total butene
nary stability in the iso-butane dehydrogenation system.
products ranging from 84.6 to 97.2 was achieved on the catalysts
Furthermore, as previously reported,19–21 zinc titanate and Zn/
with different Zn loading. Therefore, Zn-modied S-1 zeolites
SiO2 catalysts also had been considered as the candidates for
can be applied to the highly efficient transformation of light
light alkanes dehydrogenation. However, the zinc titanate
alkanes.
catalyst deactivated rapidly when both the selectivity and
conversion were kept at a higher level.19,20 Zn/SiO2 catalyst
exhibited high propylene selectivity in the dehydrogenation of
propane.21 However, the stability is also a big challenge for the
Conflicts of interest
The authors have no conicts of interest to declare.
18670 | RSC Adv., 2018, 8, 18663–18671
This journal is © The Royal Society of Chemistry 2018

View Article Online
Paper
RSC Advances
16 Y. G. Kolyagin, I. I. Ivanova and Y. A. Pirogov, Solid State
Nucl. Magn. Reson., 2009, 35, 104–112.
Acknowledgements
17 G. Caeiro, R. H. Carvalho, X. Wang, M. A. N. D. A. Lemosa,
F. Lemosa, M. Guisnet and F. R. Ribeiro, J. Mol. Catal. A:
Chem., 2006, 255, 131–158.
This work was supported by National Natural Science Founda-
tion of China (No. 21603023).
18 Y. J. Ren, F. Zhang, W. M Hua, Y. H Yue and Z. Gao, Catal.
Today, 2009, 148, 316–322.
Notes and references
1 V. Bolis, C. Busco, S. Bordiga, P. Ugliengo, C. Lamberti and
A. Zecchina, Appl. Surf. Sci., 2002, 196, 56–70.
19 A. W. Aldag and B. Okla, US pat. 4,524,144, 1985.
20 Z. X. Chen, A. Derking, W. Koot and M. P. van Dijk, J. Catal.,
1996, 161, 730–741.
¨
2 G. P. Heitmann, G. Dahlhoff and W. F. Holderich, J. Catal.,
1999, 186, 12–19.
21 N. M. Schweitzer, B. Hu, U. Das, H. Kim, L. A. Curtiss,
P. C. Stair, T. Miller and A. S. Hock, ACS Catal., 2014, 4,
1091–1098.
22 L. Q. Wang, X. S. Wang and X. W. Guo, Chin. J. Catal., 2001,
22, 513–514.
3 H. Ichihashi and H. Sato, Appl. Catal., A, 2001, 221, 359–366.
4 H. Ichihashi, M. Ishida, A. Shiga, M. Kitamura, T. Suzuki,
K. Suenobu and K. Sugita, Catal. Surv. Asia, 2003, 7(4),
261–270.
5 B. Bonelli, L. Forni, A. Aloise, J. B. Nagy, G. Fornasari,
23 J. Chen, Z. Feng, P. Ying and C. Li, J. Phys. Chem. B, 2004, 108,
12669–12676.
`
E. Garrone, A. Gedeon, G. Giordano and F. Triro,
Microporous Mesoporous Mater., 2007, 101, 153–160.
6 E. Janiszewska, A. Macario, J. Wilk, A. Aloise, S. Kowalak,
J. B. Nagy and G. Giordano, Microporous Mesoporous Mater.,
2013, 182, 220–228.
24 Y. M. Ni, W. Y. Peng, A. M. Sun, W. L. Mo, J. L. Hu, T. Li and
G. X. Li, J. Ind. Eng. Chem., 2010, 16, 503–505.
25 H. A. Zaidi and K. K. Pant, Catal. Today, 2004, 96, 155–160.
26 F. Yin, A. L. Blumenfeld, V. Gruver and J. J. Fripiat, J. Phys.
Chem. B, 1997, 101, 1824–1830.
´
7 M. Bernaue, E. Tabor, V. Pashkova, D. Kauck´y, Z. Sobalık,
´
B. Wichterlova and J. Dedecek, J. Catal., 2016, 344, 157–172.
27 W. Zhang, P. G. Smirniotis, M. Gangoda and R. N. Bose, J.
Phys. Chem. B, 2000, 104, 4122–4129.
28 V. S. Escribanoa, T. Montanarib and G. Busca, Appl. Catal., B,
2005, 58, 19–23.
29 G. V. A. Martins, G. Berlie, C. Bisio, S. Coluccia, H. O. Pastore
and L. Marchese, J. Phys. Chem. C, 2008, 112, 7193–7200.
30 E.-M. El-Malki, R. A. van Santen and W. M. H. Sachtler, J.
Phys. Chem. B, 1999, 103, 4611–4622.
8 L. B. Young, S. A. Butter and W. W. Kaeding, J. Catal., 1982,
76(2), 418–432.
9 S. Ilias and A. Bhan, J. Catal., 2014, 311, 6–16.
10 Y. Ni, A. Sun, X. Wu, G. Hai, J. Hu, T. Li and G. Li,
Microporous Mesoporous Mater., 2011, 143(2), 435–442.
11 M. Guisnet, N. S. Gnep, D. Aittaleb and Y. J. Doyemetet, Appl.
Catal., A, 1992, 87(2), 255–270.
12 J. J. H. B. Sattler, J. Ruiz-Martinez, E. Santillan-Jimenez and
B. M. Weckhuysen, Chem. Rev., 2014, 114(20), 10613–10653.
13 T. Otroshchenko, J. Radnik, M. Schneider, U. Rodermerck,
D. Linke and E. V. Kondratenko, Chem. Commun., 2016, 52,
8164–8167.
31 S. S. Arzumanov, A. A. Gabrienko, D. Freude and
A. G. Stepanov, Solid State Nucl. Magn. Reson., 2009, 35,
113–119.
32 K. J. Caspary, H. Gehrke, M. Heinritz-Adrian and
M. Schwefer, Handbook of Heterogeneous Catalysis, Wiley,
2nd edn, 2008, pp. 3206–3229.
14 J. A. Biscardi, G. D. Meitzner and E. Iglesia, J. Catal., 1998,
179, 192–202.
15 Y. G. Kolyagin, V. V. Ordomsky, Y. Z. Khimyak, A. I. Rebrovc,
F. Fajulad and I. I. Ivanova, J. Catal., 2006, 238, 122–133.
This journal is © The Royal Society of Chemistry 2018
RSC Adv., 2018, 8, 18663–18671 | 18671