Influence of Silanol Defects of ZSM-5 Zeolites on Trioxane Synthesis from Formaldehyde

Article abstract of DOI:10.1007/s10562-019-03040-x

Abstract: The silanol defects in ZSM-5 zeolite have been recognized as an important factor in catalytic activity. Here, ZSM-5 zeolites with different amounts of silanol defect sites were synthesized from synthetic gels containing fluoride medium and were applied as catalysts for trioxane synthesis. The results of XRD, SEM, NH₃-TPD, Py-IR, OH-IR, ²⁷Al MAS NMR, ¹H MAS NMR, and TG indicated that all ZSM-5 zeolites showed similar crystal size, relative crystallinity, porosity, and the number of Br?nsted acid sites. However, the silanol defects reduced obviously and the number of Lewis acid sites reduced correspondingly when a little NH₄F was added in the synthesis gels and both of them decreased slightly with increase of F/Si ratio. Compared with ZSM-5 zeolite prepared in hydroxide medium, ZSM-5 zeolite prepared in fluoride medium displayed higher selectivity to trioxane and increased gradually with increase of F/Si ratio. Moreover, the lifetime of ZSM-5 zeolite prepared in fluoride medium was longer than that of prepared in hydroxide medium. Thus, ZSM-5 zeolite prepared in fluoride medium which contained the few silanol defects and low Lewis acid sites is an efficient catalyst for trioxane synthesis. Graphic Abstract: [Figure not available: see fulltext.]

Full text of DOI:10.1007/s10562-019-03040-x

Catalysis Letters
https://doi.org/10.1007/s10562-019-03040-x
Infuence oꢀ Silanol Deꢀects oꢀ ZSM‑5 Zeolites on Trioxane Synthesis
ꢀ
rom Formaldehyde
Yuling Ye1,2,3 · Mengqin Yao1,3 · Honglin Chen · Xiaoming Zhang
1
1
Received: 11 October 2019 / Accepted: 8 November 2019
©
Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
The silanol defects in ZSM-5 zeolite have been recognized as an important factor in catalytic activity. Here, ZSM-5 zeolites
with diꢀerent amounts of silanol defect sites were synthesized from synthetic gels containing ꢁuoride medium and were
2
7
1
applied as catalysts for trioxane synthesis. The results of XRD, SEM, NH -TPD, Py-IR, OH-IR, Al MAS NMR, H MAS
3
NMR, and TG indicated that all ZSM-5 zeolites showed similar crystal size, relative crystallinity, porosity, and the number
of Brønsted acid sites. However, the silanol defects reduced obviously and the number of Lewis acid sites reduced cor-
respondingly when a little NH F was added in the synthesis gels and both of them decreased slightly with increase of F/Si
4
ratio. Compared with ZSM-5 zeolite prepared in hydroxide medium, ZSM-5 zeolite prepared in ꢁuoride medium displayed
higher selectivity to trioxane and increased gradually with increase of F/Si ratio. Moreover, the lifetime of ZSM-5 zeolite
prepared in ꢁuoride medium was longer than that of prepared in hydroxide medium. Thus, ZSM-5 zeolite prepared in ꢁuo-
ride medium which contained the few silanol defects and low Lewis acid sites is an eꢂcient catalyst for trioxane synthesis.
Graphic Abstract
Keywords ZSM-5 zeolite · Fluoride · Silanol defects · Lewis acid · Trioxane synthesis
*
Honglin Chen
hlchen@cioc.ac.cn
Extended author information available on the last page of the article
Vol.:(0
1
123456789)
3

Y. Ye et al.
1
Introduction
induced by the framework Al atoms in tetrahedral sites; the
Lewis acid sites came from the presence of extra-framework
Al and/or framework perturbed Al species. The extra-frame-
work Al Lewis acid sites formed at the expenses of adjacent
Brønsted acid sites, however, such extra-framework Al spe-
cies were in very small quantity in the HZSM-5 zeolites
with high SiO2/Al2O3 ratio of 20 [21]. The formation of
framework Al Lewis acid sites came from the dehydroxyla-
tion of Brønsted acid sites which originated from perturbed
framework Al atoms [22, 23]. By using a DFT calculation, it
was conꢃrmed that the defects coming from the dehydroxy-
lation of vicinal silanol defect sites and Brønsted acid sites
were the primary candidate for the framework Lewis acid
sites in zeolites [24]. So the relative concentration of the
vicinal silanol defect sites next to Brønsted acid sites would
determine the formation of framework Lewis acid sites.
As reported, zeolites synthesized in ꢁuoride medium
exhibited low concentration of silanol defect sites compared
with that of synthesized in hydroxide medium [25–28]. For
example, ZSM-5 zeolite with a high SiO2/Al2O3 ratio of
90 which synthesized in ꢁuoride medium showed both few
structural defects and Lewis acid sites [29]. However, little
attention has been paid to the eꢀects of silanol defects on
TOX formation. In order to increase selectivity to TOX, the
eꢀects of silanol defects in ZSM-5 zeolites on TOX for-
mation from formaldehyde were investigated in this study.
Thus, ZSM-5 zeolites with diꢀerent amounts of silanol
defect sites and Lewis acid sites were synthesized in hydrox-
ide and ꢁuoride medium with diꢀerent F/Si ratio. The cata-
lytic performances for the reaction of formaldehyde to TOX
and the stability of the synthesized ZSM-5 zeolites were
investigated.
Trioxane (TOX), an important substitute for anhydrous for-
maldehyde, is mainly used for the synthesis of polyoxym-
ethylene (POM) and poly(oxymethylene) dimethyl ethers
(
PODE) [1–5]. POM has been widely used as a traditional
metal substitute due to its high mechanical strength, excel-
lent abrasion resistance, fatigue resistance and moldability
[
6, 7]. The worldwide consumption of POM resins is more
than 840 kt/a and continues to grow [8, 9]. PODE is consid-
ered as a promising diesel additive, which can signiꢃcantly
reduce emissions of hazardous exhaust gases such as CO ,
x
NO during diesel combustion [2]. Therefore, developing an
x
eꢂcient synthesis route of TOX has attracted much attention
in recent years.
Commercial TOX is mainly produced by cyclization of
about 60% aqueous formaldehyde with acid catalysts (Eq. 1)
[
10, 11]. Besides, several side reactions were also occurred.
Methanol (MeOH) and formic acid (HCOOH) were formed
via Cannizzaro reaction (Eq. 2); methyl formate (MF) came
from esteriꢃcation of MeOH and HCOOH (Eq. 3); dimeth-
oxymethane (DMM) generated from the reaction of formal-
dehyde and MeOH (Eq. 4) [10–13].
+
H
(
(
(
(
1)
2)
3)
4)
3
CH O ⇌ (CH O)
2 2 3
Cat.
2
CH O + H O ⇌ CH OH + HCOOH
2 2 3
+
H
CH OH + HCOOH ⇌ HCOOCH + H O
3
3
2
+
H
CH O + 2CH OH ⇌ CH OCH OCH + H O
2
3
3
2
3
2
2
Experimental
Homogeneous catalysts such as sulfuric acid [11] and ionic
liquid [14] are eꢀective for TOX synthesis. However, the
diꢂculty of homogeneous catalysts is to separate HCOOH
from the mixture of feed and catalyst [15]. H-type zeolites
are a good option to avoid the problems of homogeneous
catalysts. In the previous study, ZSM-5, β and mordenite
zeolites which had nanoparticle size and high SiO /Al O
2.1 Materials
Formaldehyde (60 wt%) was prepared by concentrating a 37
wt% industrial formaldehyde solution supplied by Chengdu
(
China) Weite Plastic Co. Tetraethyl orthosilicate (TEOS,
2
2
3
9
8%), aluminum nitrate (Al(NO ) ·9H O, 99%), ammo-
3 3 2
ratio (≥45) showed comparable selectivity of TOX with sul-
furic acid [16]. The eꢀects of acid types of ZSM-5 zeolite
on TOX synthesis were also investigated in our laboratory. It
was found that ZSM-5 zeolite with the low concentration of
Lewis acid sites exhibited a high selectivity to TOX. Some
literature also reported that Cannizzaro reaction took place
in the presence of Lewis acid catalysts [17–19]. Thus, to
increase the selectivity of TOX catalyzed by zeolites, the
Lewis acid sites should be controlled in a low concentration.
For zeolites [20], the Brønsted acid sites originated from
the presence of protons balancing the negative charge which
nium ꢁuoride (NH F, 98%) and hydrochloric acid (HCl,
4
3
7%) were supplied by Chron Chemical Industrial Corpo-
ration (Chengdu, China). Tetrapropylammonium hydroxide
(
TPAOH, 20% aqueous solution) and chromatographically
pure 1,4-dioxane (99.99%) were supplied by Aladdin Indus-
trial Corporation (Shanghai, China).
2.2 Catalyst Preparation
The synthesis of ZSM-5 zeolites was performed from a gel
with the composition: 1.0SiO :0.01Al O :0.377TPAOH:
2
2
3
1
3

Inꢀuence of Silanol Defects of ZSM-5 Zeolites on Trioxane Synthesis from Formaldehyde
extinction coeꢂcients: 1.88 cm μmol and 1.42 cm μmol−1
−1
(
0–0.15)NH F:27H O. First, TEOS was added slowly into
4 2
the TPAOH aqueous solution and the resultant mixture was
kept for 3 h to hydrolyze TEOS. The aluminum nitrate solu-
for Brønsted and Lewis acid sites respectively.
Solid-state magic-angle spinning nuclear magnetic reso-
2
7
1
tion was added under vigorously stirring. NH F was added
nance (MAS NMR) spectra of Al and H were performed
on a Bruker AVANCE III NMR spectrometer at a magnetic
4
before the aluminum nitrate solution. The mother gels were
stirred at room temperature for 2 h and then hydrothermally
treated at 170 °C for 7 days. After ꢃltration and washing
with deionized water to neutral, all of the as-synthesized
ZSM-5 zeolites were dried overnight at 110 °C, and cal-
cined for 4 h in an oven at 550 °C to remove the template.
Then H-types zeolites were obtained directly. The samples
synthesized in hydroxide medium and in ꢁuoride medium
were designated as ZSM-5-OH and ZSM-5-F-x respectively,
where x represents the F/Si mole ratio.
2
7
ꢃeld of 14.1 T. The Al MAS NMR spectra were recorded
at a resonance frequency of 600 MHz, spinning rate of
14 kHz, and ꢁip angle of 15°. The pulse width was 0.65 μs,
which corresponded to π/12, and the recycle delay was 1 s.
1
The samples for H MAS NMR spectra were dehydrated at
1
350 °C for 4 h under vacuum. The H MAS NMR meas-
urements were performed at 400 MHz with a π/2 pulse of
3.5 μs, spinning rate of 14 kHz and a recycle delay of 2 s.
Aqueous solutions of Al(NO ) and tetramethylsilane (TMS)
3
3
2
7
were used as standard references for chemical shifts of Al
1
and H, respectively.
2
.3 Catalyst Characterization
2.4 Catalytic Reaction
X-ray diꢀraction (XRD) patterns were collected using a
Bruker D8 ADVANCE X-ray diꢀractometer with Cu Kα
radiation in the 2θ range from 5 to 50° at a scanning rate of
TOX was prepared from formaldehyde via reaction distilla-
tion, which involved a 250 ml glass reactor and a distillation
column. A condenser, which was added to the top cover of
the packer tower to collect the reaction gas mixture, was
heated with warm water at 50 °C to avoid the vapor phase
precipitated on the inner face. The reaction mixture com-
posed of formaldehyde (100.0 g) and solid catalyst (5.0 g)
was introduced into the reactor and heated in an oil bath with
continuous stirring. A Pt/100 monitoring temperature sensor
was inserted into the reactor to control the reaction tem-
perature with an uncertainty of± 0.1 °C. After the reaction
time was recorded when the temperature reached ca. 100 °C,
the formaldehyde was continuously fed to the reactor and
the feed rate of the formaldehyde was adjusted to keep a
constant level of the reaction mixture. The continuous reac-
tion proceeded for 3 h to compare catalytic performances of
diꢀerent zeolite. For the stability test of the ZSM-5-OH and
ZSM-5-F-0.15 zeolites, the distillate was analyzed every 8 h
and the continuous experiment was ꢃnished until the conver-
sion decreased by 10%. The distillate was analyzed by gas
chromatography with a thermal conductivity detector, using
1, 4-dioxane as internal standard material. The formaldehyde
conversion and selectivity of products were calculated by
carbon molar quantity in feed using Eqs. 5, 6:
−1
3
° min . The actual SiO /Al O ratio was analyzed by X-ray
2 2 3
ꢁ
uorescence (XRF, S4 Explorer, Bruker AXS). N adsorp-
2
tion/desorption measurements were performed with a Kubo
X-1000 (Bjbuilder). Speciꢃc surface areas (SSA ) were
BET
determined from the BET equation. The total volume was
calculated from nitrogen adsorbed volume at P/P =0.99 and
0
the external surface areas (SSA ) was calculated from iso-
ext.
therms by t-plot method. Electron micrographs were taken
on an FEI Inspect F50 (FSEM) scanning electron micro-
scope (SEM) equipped with a ꢃeld emission gun. Thermo-
gravimetric measurements (TG) was carried out on TGA
apparatus (Pyris Diamond, Perkin-Eimer, heating rate 10 °C/
min, helium ꢁow 30 ml/min).
Temperature-programmed desorption of ammonia
(
NH -TPD) experiments were carried out on a chemisorp-
3
tion apparatus (Micromeritics, AutoChem II 2920). First, the
samples were activated at 550 °C for 30 min, then cooled to
1
00 °C to absorb NH for 30 min. After purging with helium
3
gas for 1 h to remove physisorbed ammonia, NH -TPD pro-
3
−
1
ꢃles were acquired by heating samples at 10 °C min from
1
00 to 550 °C.
Fourier transform infrared (FTIR) spectra were recorded
with a Nicolet FTIR 6700 spectrometer in accordance with
a previously reported procedure [30]. Zeolite samples were
directly pressed into self-supporting wafers and evacu-
ated under vacuum at 450 °C for 3 h. OH-IR spectra were
recorded at room temperature and 150 °C. Then, the samples
were saturated with pyridine for 1 h at room temperature.
After evacuated for 1 h to remove the physically adsorbed
pyridine, the Py-IR spectra were recorded at 150 °C. The
amounts of Brønsted and Lewis acid sites were determined
by using an integrated area of a given band with the molar
all products carbon molar quantity in feed
ConversionHCHO(%) =
× 100%
(5)
HCHO molar quantity in feed
product carbon molar quantity
Selectivityproduct(%) =
× 100%
all products carbon molar quantity in feed
(6)
1
3

Y. Ye et al.
3
Results and Discussion
total pore volume in all samples were also identiꢃed by N2
adsorption/desorption measurements (Table 1). Thus, the
presence of NH4F in the gel had little inꢁuence on the par-
ticle size and porosity properties.
3.1 General Characterization of ZSM‑5 Zeolites
XRD patterns of zeolites prepared in hydroxide and ꢁuo-
ride medium showed that all samples were identiꢃed as
MFI structure with a high crystallinity (Fig. 1, Table 1).
The SiO /Al O ratio of the ZSM-5-OH zeolite was 102,
3.2 Acidic Properties of ZSM‑5 Zeolites
The total acid amounts of HZSM-5-OH and ZHSM-5-F with
diꢀerent F/Si ratios were measured by NH3-TPD (Fig. 3a,
Table 2). There were two desorption peaks at 200–240 °C
and 350–420 °C, which were ascribed to weak and strong
acid sites, respectively [34]. The total acid amounts
decreased slightly with increase of F/Si ratio. This trend
was consistent with the results of the SiO2/Al2O3 ratios of
zeolites. There were two absorption peaks located at 1540
2
2
3
which was close to the initial gel SiO /Al O ratio of 100.
2
2
3
For the ZSM-5-F zeolites, the SiO /Al O ratios were little
2
2
3
higher than the initial gels and slightly increased from 110
to 114 with the increase of F/Si ratio from 0.02 to 0.15. This
was consistent with the results of previous studies [31, 32],
since the presence of ꢁuoride ions in the gel could complex
aluminum atoms to inhibit the incorporation of aluminum in
ZSM-5 zeolite framework [33]. SEM micrographs showed
that ZSM-5-OH zeolite was round particles and ZSM-5-F
zeolites exhibited hexagonal particles, but the diameters
of all samples were about 230–260 nm (Fig. 2, Table 1).
Similar speciꢃc surface areas, external surface areas and
−1
and 1450 cm in the Py-IR spectra (Fig. 3b), which attribut-
ing to Brønsted acid sites and Lewis acid sites respectively
[30, 35, 36]. The amounts of Brønsted acid sites and Lewis
acid sites were also quantitatively calculated and listed in
Table 2. The amount of Brønsted acid sites of diꢀerent sam-
ples was very similar. Thus, the amounts of Brønsted acid
had been little aꢀected by adding NH F in the gel. Compared
4
with ZSM-5-OH, however, the amount of Lewis acid sites
decreased obviously in the samples of ZSM-5-F and it was
decreased gradually with increase of F/Si ratio. This illus-
trated that the amount of Lewis acid sites could be reduced
by synthesizing in ꢁuoride medium, which was consistent
with previous reports [27, 29].
Since the major diꢀerence in acidity between ZSM-5-OH
and ZSM-5-F with diꢀerent F/Si ratio was the amount of
Lewis acid sites, the reason for the decrease of the amount
of Lewis acid sites in the presence of ꢁuoride ions was of
particular importance. Lewis acid sites came from extra-
framework Al and/or framework Al in zeolites, therefore, the
2
7
Al MAS NMR spectra were used to analysis the complex
environments of Al in the synthesized zeolites (Fig. 4). The
peak at ca. 56 ppm was contributed to tetrahedrally coordi-
nated framework Al. Another one at ca. 0 ppm corresponded
to hexacoordinated extra-framework Al. All samples had a
sharp peak at ca. 56 ppm but the peak at ca. 0 ppm was
Fig. 1 XRD patterns of ZSM-5-OH and ZSM-5-F with diꢀerent F/Si
ratio
Table 1 Relative crystallinity,
2
a
2
b
Vtotal (cm3/g)c
Catalyst
Relative
SiO /Al O Particle SSABET (m /g) SSAext·(m /g)
2 2 3
SiO /Al O ratio, particle size,
2
2
3
crystallinity
size (nm)
and porosity of the studied
(
%)
ZSM-5 zeolites
ZSM-5-OH
98.5
102
110
113
114
230
234
255
260
480
459
458
450
217
216
216
210
0.367
0.357
0.357
0.351
ZSM-5-F-0.02 94.0
ZSM-5-F-0.08 96.6
ZSM-5-F-0.15 100
a
BET method
t-plot method
Volume absorbed at p/p =0.99
b
c
0
1
3

Inꢀuence of Silanol Defects of ZSM-5 Zeolites on Trioxane Synthesis from Formaldehyde
Fig. 2 SEM micrographs of ZSM-5-OH and ZSM-5-F with diꢀerent F/Si ratio: a ZSM-5-OH; b ZSM-5-F-0.02; c ZSM-5-F-0.08; d ZSM-
5
-F-0.15
Fig. 3 a NH -TPD proꢃles of ZSM-5-OH and ZSM-5-F with diꢀerent F/Si ratio; b Py-IR spectra of ZSM-5-OH and ZSM-5-F with diꢀerent F/
3
Si ratio
1
3

Y. Ye et al.
Table 2 Catalytic acidity of ZSM-5-OH and ZSM-5-F with diꢀerent
F/Si ratio
Catalyst
Total acid
amounts
Acid amounts (mmol/g)b
a
Brønsted
Lewis
L/B
(
mmol/g)
ZSM-5-OH
0.305
0.301
0.290
0.286
0.161
0.163
0.164
0.159
0.062
0.040
0.035
0.033
0.385
0.245
0.213
0.208
ZSM-5-F-0.02
ZSM-5-F-0.08
ZSM-5-F-0.15
a
Determined by NH -TPD
Determined by Py-IR
3
b
Fig. 5 Infrared spectra of ZSM-5-OH and ZSM-5-F with diꢀerent F/
Si ratio
were intense and broad in the sample of ZSM-5-OH. How-
ever, these two bands became sharpen and shifted back to
higher wavenumbers in the samples of ZSM-5-F but did not
change signiꢃcantly as the F/Si ratio increased. The band
−
1
at 3500 cm in the samples of ZSM-5-F also became less
intense than that of ZSM-5-OH. Thus, the amounts of the
silanol groups including internal silanol sites and external
silanol sites in the samples of ZSM-5-F were generally
higher than that of ZSM-5-OH but increased slightly as
the F/Si ratio increased. This result was in accord with the
amounts of Lewis acid sites of HZSM-5-OH and ZHSM-5-F
with diꢀerent F/Si ratios. Therefore, the amount of frame-
work Al Lewis acid sites, which originating from the per-
turbation of tetrahedrally coordinated framework Al, was
related to silanol defects [22, 24, 39, 40].
Fig. 4 27Al MAS NMR spectra of ZSM-5-OH and ZSM-5-F with dif-
ferent F/Si ratio
nearly unobserved, meaning that almost all Al was inserted
in the framework and there was very little extra-framework
Al in the synthesized ZSM-5 zeolites. This was mainly
because the SiO /Al O ratios of the synthesized ZSM-5
1
H MAS NMR was performed to further examine the
silanol sites information of the ZSM-5 zeolite samples
1
(Fig. 6). The assignments and relative distributions of H
2
2
3
zeolites were much higher than 20 [21]. Thus, combined
with the results of Py-IR above, the Lewis acid sites were
mainly originated from framework Al in this study.
MAS NMR resonances were also achieved by simula-
tion using the Gaussian Deconvolution method (Fig. 6,
Table 3). There were ꢃve distinct resonance peaks in the
proꢃles, locating at 1.7, 2.0, 2.5, 4.0 and 5.0 ppm respec-
tively. The peak at 1.7 ppm was assigned to external silanol
The FT-IR spectra in the region of OH stretching vibra-
tions were recorded in Fig. 5. There were several bands in
−
−
1
1
the OH-stretch vibrations region. The ꢃrst one at 3500 cm
correspond to silanol groups, the second one at 3720 cm
groups (SiOH ), the one at 2.0 ppm was ascribed to silanol
ext.
groups in defect sites (SiOH ), the peak at around 2.5 ppm
def.
was attributed to internal silanol sites, the third band at
belonged to extra-framework Al–OH [37, 41]. The rather
broadband at around 3.2–6 ppm, which was decomposed
into two peaks at 4.0 and 5.0 ppm, corresponded to bridg-
ing Si(OH)Al groups [41]. As seen from Table 3, the rela-
tive peak area of bridging Si(OH)Al groups in two sam-
ples kept constant. Compared with ZSM-5-OH, however,
ZSM-5-F-0.15 showed that the relative peak area of silanol
groups in defect sites was reduced signiꢃcantly. This result
was consistent with the FT-IR analysis above. It also further
−
1
3
736 cm was assigned to external silanol sites, and two
−1
−1
shoulder bands at 3610 cm and 3667 cm were associated
to strongly acidic bridging Si(OH)Al groups (Brønsted acid
sites) and low acidity perturbed Al–OH groups respectively
−
1
[
37, 38]. The weak intensity of two bands at 3610 cm and
−
1
3
667 cm were similar in all samples, since the low Al
content in the zeolite with high SiO /Al O molar ratio. It
2
2
3
−
1
−1
should be noted that the bands at 3720 cm and 3736 cm
1
3

Inꢀuence of Silanol Defects of ZSM-5 Zeolites on Trioxane Synthesis from Formaldehyde
formaldehyde catalyzed by ZSM-5-F-0.15 only reduced by
.03%. However, the selectivity of TOX increased obviously
3
when catalyzed by ZSM-5-F zeolites. For ZSM-5-F-0.02
zeolite, the selectivity to TOX was 93.36%, which was
higher than that of ZSM-5-OH by 5.76%; the selectivity
to HCOOH, MeOH, MF, and DMM were 0.58%, 3.35%,
0
5
.41%, and 2.30%, which were lower than that of ZSM-
-OH by 0.87%, 1.47%, 1.6%, and 1.82%, respectively. By
continuing to increase the F/Si molar ratio, the selectivity
to TOX increased slightly and selectivities to all byprod-
ucts also decreased gradually. Since the Lewis acid sites
initiated the side reaction of Cannizzaro [17–19], HCOOH
and MeOH were formed (Eq. 2). Then, the esteriꢃcation
(
Eq. 3) [42] of HCOOH and MeOH and the acetalization
Eq. 4) [43, 44] of formaldehyde and MeOH took place with
(
Brønsted acidic catalysts to generate MF and DMM. With-
out the byproducts of MeOH and HCOOH, the reactions of
esteriꢃcation of HCOOH and MeOH and the acetalization
of formaldehyde and MeOH would not occur in this sys-
tem. Thus, the increase of TOX selectivity and the decrease
of byproducts selectivities were mainly ascribed to the less
silanol defect sites and thus of the Lewis acid sites in ZSM-
Fig. 6 1H MAS NMR spectra of ZSM-5-OH and ZSM-5-F-0.15
5
-F zeolites.
Table 3 Relative peak area and assignments of OH groups in 1
H
Lifetime experiments were performed to examine the
MAS NMR spectra
stability of the ZSM-5-OH and ZSM-5-F-0.15 zeolites on
the reaction of formaldehyde to TOX (Fig. 7). Similarly as
above, ZSM-5-F-0.15 showed higher selectivity of TOX
but lower conversion of formaldehyde and lower selectivi-
ties of all byproducts including HCOOH, MeOH, MF, and
DMM than ZSM-5-OH. Moreover, ZSM-5-F-0.15 exhibited
a longer lifetime of 186 h, while ZSM-5-OH only had a life-
time of 146 h. Two zeolites after reaction were analyzed by
TG to determine their coke deposition (Fig. 8). The weight
loss of the ZSM-5-F-0.15 zeolite was ca. 10%, which was
lower than that of ZSM-5-OH (ca. 13%). This suggested that
the low concentration of silanol defect sites would prevent
the coke formation, so had a long lifetime.
Catalyst
Relative peak area (%)
SiOHext.
SiOHdef.
Al–OH
Si(OH)Al
ZSM-5-OH
ZSM-5-F-0.15
16.64
28.89
27.72
18.51
12.16
8.82
43.48
43.78
conꢃrmed that the amount of framework Al Lewis acid sites
was related to silanol groups in defect sites.
3.3 Catalytic performance
The eꢀects of ZSM-5-OH and ZSM-5-F with diꢀerent F/Si
ratio on the conversion of formaldehyde and selectivity of
products in the formation of TOX are listed in Table 4. For
ZSM-5-OH, the conversion of formaldehyde was 32.61%.
For ZSM-5-F, the conversion of formaldehyde decreased
from 32.26 to 29.58% with increase of F/Si ratio from 0.02
to 0.15. Compared with ZSM-5-OH, the conversion of
4 Conclusions
ZSM-5 zeolites which had similar crystal size, relative
crystallinity and porosity porosities were synthesized in
hydroxide medium and ꢁuoride medium. The amounts of
Table 4 Conversion of TOX
Catalyst
Conversion (%)
Selectivity (%)
and selectivities of TOX and
byproducts over ZSM-5-OH
and ZSM-5-F with diꢀerent F/
Si ratio
TOX
HCOOH
MeOH
MF
DMM
ZSM-5-OH
32.61
32.26
31.63
29.58
87.60
93.36
93.77
94.69
1.45
0.58
0.53
0.50
4.82
3.35
3.30
2.82
2.01
0.41
0.28
0.25
4.12
2.30
2.12
1.74
ZSM-5-F-0.02
ZSM-5-F-0.08
ZSM-5-F-0.15
1
3

Y. Ye et al.
Fig. 7 a Conversions of formaldehyde with ZSM-5-OH and ZSM-5-F-0.15; b selectivities of TOX and byproducts catalyzed by ZSM-5-OH; c
selectivities of TOX and byproducts catalyzed by ZSM-5-F-0.15
Brønsted acid sites were nearly kept constant in the sam-
ples of ZSM-5-OH and ZSM-5-F zeolites. However, while
adding NH F in the gels, the amounts of silanol defect
4
sites and Lewis acid sites reduced obviously and both of
them decreased slightly with increase of F/Si ratio. This
results suggested that the formation of silanol defect sites
contributed to the formation of framework Lewis acid sites
and it could be inhibited by adding NH F in the synthetic
4
gel. In the system of TOX synthesis from formaldehyde,
the selectivity of TOX catalyzed by ZSM-5-F zeolite was
higher than that of ZSM-5-OH and increased gradually
with increase of F/Si ratio. Moreover, ZSM-5-F zeolite
showed a longer lifetime than that of ZSM-5-OH.
Acknowledgements This work was supported by the National Key
R&D Program of China (Grant Number 2018YFB0604902). The fund-
ing source has no role in study design; in the collection, analysis and
interpretation of data; in the writing of the report; and in the decision
to submit the article for publication.
Fig. 8 TG measurement of ZSM-5-OH and ZSM-5-F-0.15 after con-
tinuous reactions
1
3

Inꢀuence of Silanol Defects of ZSM-5 Zeolites on Trioxane Synthesis from Formaldehyde
2
2
2. Brus J, Kobera L, Schoefberger W, Urbanova M, Klein P, Sazama
P, Tabor E, Sklenak S, Fishchuk AV, Dedecek J (2015) Angew
Chem Int Ed 54:541–545
Compliance with Ethical Standards
Conꢀlict oꢀ interest The authors declare no conꢁict of interest.
3. Yarulina I, De Wispelaere K, Bailleul S, Goetze J, Radersma M,
Abou-Hamad E, Vollmer I, Goesten M, Mezari B, Hensen EJM,
Martinez-Espin JS, Morten M, Mitchell S, Perez-Ramirez J, Ols-
bye U, Weckhuysen BM, Van Speybroeck V, Kapteijn F, Gascon
J (2018) Nat Chem 10:804–812
References
2
2
2
2
2
2
4. Sokol AA, Catlow CRA, Garces JM, Kuperman A (2000) Adv
Mater 12:1801–1805
1
2
3
4
5
6
.
.
.
.
.
.
Wu Q, Li WJ, Wang M, Hao Y, Chu TH, Shang JQ, Li HS, Zhao
Y, Jiao QZ (2015) RSC Adv 5:57968–57974
5. Prodinger S, Shi H, Wang HM, Derewinski MA, Lercher JA
Baranowski CJ, Bahmanpour AM, Krocher O (2017) Appl Catal
B Environ 217:407–420
(
2018) Appl Catal B Environ 237:996–1002
6. Jo C, Park W, Ryoo R (2017) Microporus Mesoporus Mater
Wu JB, Zhu HQ, Wu ZW, Qin ZF, Yan L, Du BL, Fan WB, Wang
JG (2015) Green Chem 17:2353–2357
2
39:19–27
7. Kalvachev Y, Jaber M, Mavrodinova V, Dimitrov L, Nihtianova D,
Valtchev V (2013) Microporous Mesoporous Mater 177:127–134
8. Nabavi MS, Zhou M, Mouzon J, Grahn M, Hedlund J (2019)
Microporous Mesoporous Mater 278:167–174
Wang RY, Wu ZW, Li ZK, Qin ZF, Chen CM, Zheng ZF, Wang
GF, Fan WB, Wang JG (2019) Appl Catal A Gen 570:15–22
Mu YB, Jia MC, Jiang W, Wan XB (2013) Macromol Chem Phys
2
14:2752–2760
9. Li JJ, Liu M, Guo XW, Dai CY, Song CS (2018) J Energy Chem
Luftl S, Archodoulaki VM, Seidler S (2006) Polym Degrad Stabil
2
7:1225–1230
9
1:464–471
3
3
0. Wu WQ, Weitz E (2014) Appl Surf Sci 316:405–415
1. Losch P, Pinar AB, Willinger MG, Soukup K, Chavan S, Vincent
B, Pale P, Louis B (2017) J Catal 345:11–23
7
8
.
.
Zhao XW, Ye L (2011) Mater Sci Eng, A 528:4585–4591
Luftl S, Visakh PM, Chandran S (2014) Polyoxymethylene hand-
book: structure, properties, applications and their nanocomposites.
Blackwell, Oxford
3
3
3
2. Aiello R, Crea F, Nigro E, Testa F, Mostowicz R, Fonseca A,
Nagy JB (1999) Microporous Mesoporous Mater 28:241–259
3. Louis B, Kiwi-Minsker L (2004) Microporous Mesoporous Mater
9
.
Hoꢀmann M, Bizzarri C, Leitner W, Muller TE (2018) Catal Sci
Technol 8:5594–5603
7
4:171–178
1
1
1
1
1
0. Maiwald M, Grutzner T, Strofer E, Hasse H (2006) Anal Bioanal
Chem 385:910–917
4. Rodriguez-Gonzalez L, Simon U (2010) Meas Sci Technol
2
1(2):027003
1. Grutzner T, Hasse H, Lang N, Siegert M, Strofer E (2007) Chem
Eng Sci 62:5613–5620
3
3
3
5. Jin F, Li YD (2009) Catal Today 145:101–107
6. Isernia LF (2013) Mater Res Ibero-am J Mater 16:792–802
7. Qin ZX, Lakiss L, Tosheva L, Gilson JP, Vicente A, Fernandez
C, Valtchev V (2014) Adv Funct Mater 24:257–264
2. Masamoto J, Hamanaka K, Yoshida K, Nagahara H, Kagawa K,
Iwaisako T, Komaki H (2000) Angew Chem Int Ed 39:2102–2104
3. Ma WT, Hu YF, Qi JG, Wei LH, Zhang XM, Yang ZY, Jiang SQ
3
3
4
4
4
4
4
8. Chakarova K, Drenchev N, Mihaylov M, Nikolov P, Hadjiivanov
K (2013) J Phys Chem C 117:5242–5248
(
2017) Ind Eng Chem Res 56:6910–6915
4. Zhao YM, Hu YF, Qi JG, Ma WT (2016) Chin J Chem Eng
9. Bordiga S, Lamberti C, Bonino F, Travert A, Thibault-Starzyk F
2
4:1392–1398
(
2015) Chem Soc Rev 44:7262–7341
1
1
1
5. Yin LY, Hu YF, Wang HY (2016) Petrol Sci 13:770–775
6. Ishida H, Akagishi K (1996) Nippon Kagaku Kaishi, pp 290–297
7. Busca G, Lamotte J, Lavalley JC, Lorenzelli V (1987) J Am Chem
Soc 109:5197–5202
0. Janiszewska E, Kowalska-Kus J, Gora-Marek K, Szymocha A,
Nowinska K, Kowalak S (2019) Appl Catal A Gen 581:1–10
1. Muller M, Harvey G, Prins R (2000) Microporous Mesoporous
Mater 34:281–290
1
1
2
8. Chen MT, Lin YS, Lin YF, Lin HP, Lin JL (2004) J Catal
2. Indu B, Ernst WR, Gelbaum LT (1993) Ind Eng Chem Res
2
28:259–263
3
2:981–985
9. Russell AE, Miller SP, Morken JP (2000) J Org Chem
3. Oestreich D, Lautenschutz L, Arnold U, Sauer J (2017) Chem Eng
Sci 163:92–104
6
5:8381–8383
0. Stosic D, Auroux A (2013) Characterization of acid-base sites in
zeolites. In: Auroux A (ed) Calorimetry and thermal methods in
catalysis, vol 154. Springer, Berlin, pp 353–384
4. Sun RY, Delidovich I, Palkovits R (2019) ACS Catal 9:1298–1318
Publisher’s Note Springer Nature remains neutral with regard to
2
1. Yi XF, Liu KY, Chen W, Li JJ, Xu ST, Li CB, Xiao Y, Liu
HC, Guo XW, Liu SB, Zheng AM (2018) J Am Chem Soc
jurisdictional claims in published maps and institutional aꢂliations.
1
40:10764–10774
Aꢀliations
Yuling Ye1,2,3 · Mengqin Yao1,3 · Honglin Chen · Xiaoming Zhang
1
1
1
2
3
Chengdu Institute of Organic Chemistry, Chinese Academy
of Sciences, No. 9, Section 4, Renmin South Road,
Chengdu 610041, Sichuan, People’s Republic of China
University of Chinese Academy of Sciences, Beijing 100049,
People’s Republic of China
College of Chemical Engineering, Sichuan University
of Science & Engineering, Zigong 643000, Sichuan,
People’s Republic of China
1
3