Vapor-phase catalytic methylation of 1,1,1,3,3,3-hexafluoroisopropanol for the mass production of 1,1,1,3,3,3-hexafluoroisopropyl methyl ether

Article abstract of DOI:10.1016/j.jfluchem.2020.109673

With the phasing-out of chlorofluorocarbons and hydrochlorofluorocarbons required by the Montreal and Kyoto Protocols, hydrofluoroethers (HFEs) are now considered to be promising alternatives due to their zero ozone-depletion and low global-warming potentials, and their significant capacities for use in heat-pump and cleaning-agent applications. However, the pollution-free and large-scale synthesis of HFEs has been a long-standing challenge. To address the issue, we previously reported a novel synthetic method for the large-scale production of 1,1,1,3,3,3-hexafluoroisopropyl methyl ether (HFE-356mmz), a representative HFE, through the vapor-phase methylation of 1,1,1,3,3,3-hexafluoroisopropanol using metal fluorides as catalysts. In this work, mixed oxides of Mg and Al with various Mg/Al₂ ratios were employed as alternative catalysts; their abilities to promote the reaction were determined and the methylation mechanism was explored. All Mg-Al mixed oxides promoted the production of HFE-356mmz, albeit with different efficiencies, which were found to be determined by the surface acid-base properties of the catalysts. The results agree well with those obtained using metal fluorides as catalysts and provide new mechanistic evidence. Our study not only offers further evidence of the reaction mechanism, but also affords a more universal and operable process that uses more-common and less-expensive catalysts.

Full text of DOI:10.1016/j.jfluchem.2020.109673

Journal Pre-proof
Vapor-phase catalytic methylation of 1,1,1,3,3,3-hexafluoroisopropanol
for the mass production of 1,1,1,3,3,3-hexafluoroisopropyl methyl ether
Wei Li, Fengniu Lu, Xiaoling Zhang
PII:
S0022-1139(20)30371-7
DOI:
https://doi.org/10.1016/j.jfluchem.2020.109673
FLUOR 109673
Reference:
To appear in:
Journal of Fluorine Chemistry
Received Date:
Revised Date:
Accepted Date:
23 May 2020
12 October 2020
12 October 2020
Please cite this article as: Li W, Lu F, Zhang X, Vapor-phase catalytic methylation of
,1,1,3,3,3-hexafluoroisopropanol for the mass production of 1,1,1,3,3,3-hexafluoroisopropyl
1
methyl ether, Journal of Fluorine Chemistry (2020),
doi: https://doi.org/10.1016/j.jfluchem.2020.109673
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
©
2020 Published by Elsevier.

Vapor-phase catalytic methylation of 1,1,1,3,3,3-hexafluoroisopropanol for the mass
production of 1,1,1,3,3,3-hexafluoroisopropyl methyl ether
a,b
a
a
Wei Li , Fengniu Lu *, Xiaoling Zhang *
aSchool of Chemistry and Chemical Engineering, Beijing Institute of Technology, 5 South Zhongguancun Street, Beijing,
1
00081, China
bState Key Lab for Fluorine Greenhouse Gases Replacement and Control Treatment, Zhejiang Research Institute of
Chemical Industry, Hangzhou 310023, PR China
*
Corresponding Author: E-mail: feng.lyu666@hotmail.com, zhangxl@bit.edu.cn, Tel. + 86-10-88875298
Graphical Abstract
Vapor-phase pollution-free and mass production of 1,1,1,3,3,3-hexafluoroisopropyl methyl ether
using mixed oxides of Mg and Al as catalysts
Highlights

HFE-356mmz was prepared by vapor-phase methylation of HFiP over Mg-Al oxides
1



Among the various Mg-Al oxides, 3/2Mg-Al
the highest catalytic activity
2
with medium acidity and medium basicity exhibits
This study provides new evidence of the acid-base synergism reaction mechanism
-
1
-2


The conversion/total activity for HFiP over 3/2Mg-Al
2
were 89.4%/0.94 mmol h m
Mg-Al oxides is a good substitute of metal fluorides to catalyze the methylation of HFiP
Abstract: With the phasing-out of chlorofluorocarbons and hydrochlorofluorocarbons required by the
Montreal and Kyoto Protocols, hydrofluoroethers (HFEs) are now considered to be promising alternatives
due to their zero ozone-depletion and low global-warming potentials, and their significant capacities for
use in heat-pump and cleaning-agent applications. However, the pollution-free and large-scale synthesis
of HFEs has been a long-standing challenge. To address the issue, we previously reported a novel
synthetic method for the large-scale production of 1,1,1,3,3,3-hexafluoroisopropyl methyl ether
(
HFE-356mmz), a representative HFE, through the vapor-phase methylation of
,1,1,3,3,3-hexafluoroisopropanol using metal fluorides as catalysts. In this work, mixed oxides of Mg
and Al with various Mg/Al ratios were employed as alternative catalysts; their abilities to promote the
1
2
reaction were determined and the methylation mechanism was explored. All Mg-Al mixed oxides
promoted the production of HFE-356mmz, albeit with different efficiencies, which were found to be
determined by the surface acid-base properties of the catalysts. The results agree well with those obtained
using metal fluorides as catalysts and provide new mechanistic evidence. Our study not only offers
further evidence of the reaction mechanism, but also affords a more universal and operable process that
uses more-common and less-expensive catalysts.
Keywords: 1,1,1,3,3,3-hexafluoroisopropylmethyl ether, vapor-phase methylation, Mg–Al mixed oxide
2

1. Introduction
To protect the stratospheric ozone layer and address global warming issues, the consumption and
production of chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs) was phased out by the
Montreal and Kyoto Protocols [1]. Thereafter, hydrofluoroethers (HFEs) were considered as promising
substitutes for CFCs and HCFCs as they have zero ozone-depletion and low global-warming potentials,
short atmospheric lifetimes, and are widely applicable [2-4]. Recently,
1
,1,1,3,3,3-hexafluoroisopropylmethyl ether (HFE-356mmz), a member of the HFE family, has been
attracting growing interest due to its significant capacity for use in high-temperature heat-pump fluids,
detergents, solvents, and foam-blowing agents [5-7]. However, HFE-356mmz is very challenging to
prepare on the large-scale, which has greatly hindered the promotion and application of this compound.
Although HFE-356mmz was able to be prepared by the O-methylation of
1
,1,1,3,3,3-hexafluoroisopropanol (HFiP) using (CH
3
)
2
SO
4
, CH X (X = Cl, Br, I), or diazomethane as the
3
methylating agent over an inorganic base [8-10], the inability to separate the base from the product, and
the large amount of waste water produced hinders its industrial production.
To mass-produce HFE-356mmz, we developed a new synthetic method that involves the
vapor-phase catalytic methylation of HFiP with dimethyl carbonate (DMC) as the methylating reagent
and a metal fluoride as the catalyst [11]. This pollution-free, recyclable, and continuous process is very
suitable for the industrial-scale mass production of HFE-356mmz. We found that the catalytic activity of
the metal fluoride (AlF
of its surface and the total amount of acidic sites on the surface. Only MgF
acidic and basic, exhibited excellent catalytic activity. Based on the acid-base effects of the catalysts and
3
, MgF
2
, CaF
2
, SrF
2
, and BaF
2
) was determined by the Lewis acid-base properties
2
, which is moderately Lewis
3

the reported methylation reaction mechanism using DMC [12,13], we proposed a reaction mechanism for
the catalytic methylation of HFiP.
Herein, to further verify the reaction mechanism and to explore the effectiveness of other catalysts
that are less expensive and easier to synthesize, we prepared a series of mixed Mg-Al oxides with
different Mg/Al ratios as alternatives to metal fluoride catalysts, and studied their reactivities during the
2
methylation of HFiP (Scheme 1). Mixed Mg-Al oxides were selected because they are inexpensive, easy
to prepare, and have readily adjustable acid-base properties [14]. Moreover, Mg-Al mixed oxides are
widely used in the catalytic methylation of phenol and its derivatives [13,15,16]. In addition, it was found
that HFiP has good stability under the optimum reaction temperature (eg. 240 C). Therefore, Mg-Al
oxides catalyst can be used in this reaction. Since the acid-base properties of Mg-Al oxides are
determined by both the chemical composition and the preparation procedure [17,18], we prepared the
Mg-Al oxides in this study using the co-precipitation method but with various Mg/Al
ratios. The effect of the Mg/Al ratio on the surface acid-base properties of each catalyst and its catalytic
performance were systematically investigated. A metal oxide with moderate acidity and basicity (3:2
Mg/Al ratio) exhibited the best activity, in good agreement with the proposed mechanism. We believe
2
(MgO:Al
2
3
O )
2
2
that our study supports the proposed reaction mechanism. Replacing metal fluorides by metal oxides led
to a more universal and operable reaction that can be practically used to mass-produce HFE-356mmz.
Scheme 1. Synthesis of HFE-356mmz through the vapor-phase methylation of HFiP using DMC over a
Mg-Al-oxide catalyst.
4

2. Experimental Section
2
.1 Preparing the Catalysts
The Mg-Al mixed oxides were prepared by the co-precipitation method [19]. Taking the Mg-Al
mixed oxide with a 2:3 Mg/Al
Mg(NO ·6H O (25.6 g, 0.1 mol, Aladdin, > 99%) and Al(NO
9%) in 500 mL of distilled water and stirring for 1 h, and then (2) slowly adding 5% ammonia solution
2
ratio as an example, a typical preparation process includes: (1) dissolving
3
)
2
2
3
)
3
·9H O (112.5 g, 0.3 mol, Aladdin, >
2
9
to the solution at room temperature with constant stirring until pH 9.5 was reached. The resulting
suspension was collected by vacuum filtration, washed with distilled water, and then dried at 120 ℃for

1
1
2 h. The as-obtained solid was calcined at 400 ℃(10 ℃min ) for 4h and then at 500 ℃for 4 h in a
muffle furnace, after which it was ground and pelletized. MgO, Al , and the other Mg-Al mixed oxides
were all prepared following the same preparation procedure but with different molar ratios of
Mg(NO ·6H O to Al(NO ·9H O. The as-prepared Mg-Al mixed oxides are labeled as “a/b Mg-Al
where “a/b” represents the molar ratio of MgO and Al in the mixed oxides; for example, “2/3Mg-Al
. The catalysts prepared in this manner were used
2
O
3
3
)
2
2
3
)
3
2
2
”,
2
O
3
2
”
refers to the catalyst with a 2:3 ratio of MgO and Al
directly in the vapor-phase methylation reaction.
2
O
3
2
.2 Characterizing the Catalysts
The crystalline structures of all catalysts were examined by X-ray diffractometry (XRD) in the 10–90°
θ range on a Bruker D8 Advance diffractometer with Ni-filtered Cu Kα radiation (λ = 1.540 Å). The
2
Mg/Al
2
molar ratios were determined by inductively coupled plasma-optical emission spectrometry
(ICP-OES) on an Agilent ICP-OES 730 instrument.
5

Brunauer–Emmett–Teller (BET) surface areas were determined by N
2
physisorption using a
Micromeritics TriStar II 3020 porosity analyzer. Prior to N physisorption, each sample was degassed for 3
2
h at 250 °C. Surface areas were calculated according to the BET equation.
Catalyst acidities were determined by the temperature programmed desorption of NH
3
(NH -TPD)
3
using a Quantachrome Autosorb-iQ-C (USA) instrument. Prior to the measurement, all the samples were
vacuum dried at 400 °C for 8 h. After calcination, each sample (100 mg) was first heated in He (30 cm3

1
min ) at 250 °C for 1h and then treated with a 5% NH
3
/He stream at 100 °C for 1h. Thereafter, the
. NH was gradually
sample was flushed with He at 100 °C for 6 h to remove the weakly adsorbed NH
3
3
liberated as the temperature was increased from 100 to 500 °C at 10 °C min1, and was detected by
on-line gas chromatography (GC) with a thermal conductivity detector. To determine the basicities of the
catalysts, they were subjected to CO
was injected instead of NH
.3 Testing the Catalyst and Analyzing the Product
All of the catalysts (particles 0.8–1.2 mm in diameter, 20 g) were flushed with N
2
-TPD in a manner similar to NH
3
-TPD, with the exception that CO
2
3
.
2
2
at 240 °C for 4 h
before methylation. A mixture of HFiP (Sinochem, Lantian, > 99%) and DMC (Aladdin, > 99%) in a 1:2
molar ratio was fed into a down-flow reactor using a syringe pump, and the mixture was allowed to react
at 240 °C and 101.1 kPa. The catalysts were tested at a contact time of 8 s by analyzing the outlet reaction
mixture by GC using a GC-2014C chromatograph (Shimadzu). The GC instrument was fitted with a
DB-VRX capillary column (30 m  0.45 mm  0.25 μm), whose temperature was first held at 50 °C for 5

1
min, and then heated to 200 °C at 10 °C min , and finally held at 200 °C for 8 min. The injection port
and the hydrogen flame ionization detector were maintained at a constant temperature of 200 °C. The
6


1
carrier gas (He) was introduced at 10 mL min . The conversion (X
H
) of HFiP and the selectivity (S ) of
H
HFE-356mmz were determined using a reported method [11].
3
. Results and discussion
3
.1 Activities of the Catalysts for the Methylation of HFiP
Despite the low catalytic activities of pure MgO and Al
mixed Mg-Al oxides promoted the reaction efficiently. Since no by-products were detected in the outlet
reaction mixture for any of these catalysts, the S for HFE-356mmz was considered to be 100% in all
cases. However, the X of HFiP differed among the various Mg-Al oxides (Table 1), with 3/2Mg-Al
exhibited the highest activity.
2
O in the HFiP methylation reaction, all
3
H
H
2
Table 1. X data for the vapor-phase methylation of HFiP with DMC over
H
a
various catalysts.
Catalyst b
X
c
(%)
d
1 2
H
Total activity (mmol h m )
MgO
3.6
0.09
0.73
0.75
0.94
0.70
0.56
0.33
0.20
0.04
3
2
3
1
2
1
1
/1Mg-Al
/1Mg-Al
/2Mg-Al
/1Mg-Al
/3Mg-Al
/2Mg-Al
/3Mg-Al
2
2
2
2
2
2
2
48.9
66.1
89.5
76.2
70.9
42.7
29.0
7.9
2 3
γ-Al O
a
Reaction conditions: temperature = 240 °C; contact time = 8 s; HFiP injection rate = 139.8
1
mmol h ; running time = 4 h.
b
2
The as-prepared Mg-Al mixed oxides are labeled as “a/b Mg-Al ”, where “a/b” represents the
molar ratio of MgO and Al
2
O
3
in the mixed oxides.
c
X
H
(HFiP conversion) was determined by GC against a calibrated internal standard.
d
1
2
Total activity (mmol h m ) was determined from the molar reaction rate of HFiP and the
specific surface area of the catalysts used in this reaction.
3
.2 Characterization and Evaluation the Catalyst
7

To determine the reason for the different catalytic activities of the mixed Mg-Al oxides, the
crystalline phases, surface areas, and surface acid-base properties of all catalysts were carefully
investigated. Then, the 500 h lifetime of the optimized catalyst was evaluated and compared with that of
MgF .
2
The crystalline phases of the catalysts were examined by XRD. As summarized in Table 2, Al
and MgO exhibited γ-Al and MgO (cubic) phases, respectively. The XRD patterns of the Mg-rich
mixed oxides (Entries 2–4, Table 2) exhibited typical diffraction peaks of MgO and γ-Al , with the
2
O
3
2
O
3
2
O
3
intensities of the MgO diffraction peaks observed to increase with increasing MgO content. On the
contrary, the Mg-Al mixed oxides with larger Al contents (Entries 6–9, Table 2) only exhibited
diffraction peaks typical of γ-Al
additional MgAl phase was observed for 1/1Mg-Al
transition from the Mg-rich to the Al-rich phase. Although the 1/3Mg-Al
mixed oxides exhibited catalytic activity (Table 1) that increased with increasing γ-Al
intensity, the 1/2Mg-Al oxide, which showed pure γ-Al diffractions, was much less active. Therefore,
catalytic activity does not correlate with the intensities of the γ-Al diffractions. In addition, the
/1Mg-Al oxide that exhibited diffraction peaks from both MgO and γ-Al , and the 1/1Mg-Al
that showed diffraction peaks corresponding to the MgAl phase and γ-Al , exhibited similar
catalytic activities, which indicates that the nature of the Mg phase (MgO or MgAl ) has a negligible
2
O
3
, suggesting that MgO is well dispersed in these γ-Al
(Entry 5, Table 2), which is probably due to the
, 1/2Mg-Al , and 3/2Mg-Al
-diffraction-peak
2
O
3
lattices. An
2
O
4
2
2
2
2
2
O
3
2
2
O
3
2
O
3
3
2
2
O
3
2
oxide
2
O
4
2
O
3
2
O
4
influence on activity. Therefore, we conclude that catalytic activity is not determined by the crystalline
phase of the catalyst.
8

Table 2. BET surface areas and XRD data for MgO, Al
2
3
O , and the Mg-Al mixed oxide samples.
Mg/Al
2
molar ratio
Surface area
XRD structural analysis
Entry
Catalyst a
Measuredb
2
1
Phases detected
Theoretical
（m g ）
56.0
1
2
3
4
5
6
7
8
9
MgO
-
-
MgO periclase
3/1Mg-Al
2/1Mg-Al
3/2Mg-Al
1/1Mg-Al
2/3Mg-Al
1/2Mg-Al
1/3Mg-Al
2
2
2
2
2
2
2
3.0
2.0
1.5
1.0
0.68
0.50
0.33
-
3.4
2.1
1.6
1.0
0.68
0.44
0.31
-
93.9
MgO periclase + γ-Al
MgO periclase + γ-Al
MgO periclase + γ-Al
2
2
2
O
O
O
3
3
3
122.4
133.2
152.2
177.2
179.7
205.4
264.5
2 4 2 3
MgAl O + γ-Al O
γ-Al
γ-Al
γ-Al
γ-Al
2
2
2
2
O
3
O
3
O
3
O
3
2 3
Al O
aThe as-prepared Mg-Al mixed oxides are labeled as “a/b Mg-Al
”, where “a/b” represents the molar ratio of MgO and Al
O
3
in the mixed
2
2
oxides.
bMg/Al
molar ratios in the final precipitates were determined by ICP-OES.
2
The BET surface areas of the catalysts are listed in Table 2. Among these catalysts, γ-Al
2
3
O and
2
1
2
1
MgO exhibit the largest (264.5 m g ) and smallest (56 m g ) surface areas respectively. The surface
area of the Mg-Al mixed oxide was observed to decrease with increasing Mg content. Therefore, the
surface properties of these catalysts can readily be tuned by adjusting the Mg content; however, the
surface areas and catalytic activities do not correlate. Therefore, we conclude that the activity of the
catalyst is not determined by its surface area.
The acidities of the catalysts were next studied by NH
20], NH -TPD peaks in the 80–150 °C, 150–300 °C, and 300–450 °C ranges correspond to weak (W),
medium (M), and strong (S) acidities, respectively. The NH -TPD profile for Al displays two
overlapping peaks at about 180 °C and 305 °C arising from NH adsorbed at medium and strongly acidic
sites; similar bands in the NH -TPD curves were observed for the Al-rich samples (Entries 5–9, Table 3).
The NH -desorption temperature increased with increasing Al content. The desorption of NH from the
3
-TPD (Fig. 1). According to previous studies
[
3
3
2
O
3
3
3
3
3
MgO and MgO-rich samples (Entries 1–4, Table 3) gave rise to weak bands below 170 °C, indicating the
existence of sites with weak and medium acidities. The acid-site densities and percentage contribution of
9

each peak were determined by integrating the TPD profiles, the results of which are presented in Table 3.
MgO exhibited the lowest total evolved NH , which increased with increasing Al content (Entries 1–9,
3
Table 3). Moreover, the S peak contributed less with decreasing Al content, while the W peak contributed
more, and the contribution of the M peak first increased and then decreased.
Al O3
2
1
/3Mg-Al2
1
/2Mg-Al2
2
1
/3Mg-Al2
/1Mg-Al2
3
/2Mg-Al2
2
3
/1Mg-Al2
/1Mg-Al2
MgO
0
100
200
300
400
500
Temperature (℃)
Fig. 1. NH -TPD profiles for the catalysts examined in this study.
3
The basicities of the catalysts were studied by CO
different base strengths (Fig. 2). The desorption peaks at 80–150 °C, 150–300 °C, and > 300 °C are
attributed to the desorption of CO from the sites of weak (W), medium (M), and strong (S) basicities,
respectively. The CO -TPD profile of MgO displays two broad peaks at about 320 °C and 220 °C,
indicating the presence of sites of strong and medium basicity; similar peaks were observed for the Mg-rich
samples (Entries 2–4, Table 3). The CO -desorption temperature was observed to decrease with decreasing
Mg content. On the other hand, the CO -desorption temperatures of the Al and Al-rich samples (Entries
–9, Table 3), decrease with increasing Al content, and only sites of medium and weak basicity were
observed for the Al , 1/3Mg-Al and 1/2Mg-Al samples. The total amount of desorbed CO and the
2
-TPD, which revealed the existence of sites with
2
2
2
2
2
O
3
6
2
O
3
2
2
2
1
0

percentage contribution of each peak determined by integrating the TPD profiles are presented in Table 3.
In general, the contribution of the S peak decreased, while that of the W peak increased, and that of the M
peak first increased and then decreased with increasing Al content (Entries 1–9, Table 3).
According to the acidity and basicity data for various catalysts, Al
2
O and MgO are dominated by
3
strongly acidic and basic centers, respectively, while the mixed oxides contain both weakly acidic and
basic centers as well as those of medium strength. When referring to the catalytic activities, we found that
the 3/2Mg-Al
87.8%, Table 3) and medium-strength basic centers (70.6%, Table 3), which may contribute to
this catalyst exhibiting the highest activity in the methylation reaction. Similarly, all other
2
oxide contains the highest percentage of both medium-strength acidic centers
(

1

2
catalysts that show high catalytic activities, such as 3/1Mg-Al oxide (0.73 mmol h m , Table
2

1

2

1

2
1
), 1/1Mg-Al
Table 1), exhibit relatively high percentages of both medium-strength acidic and basic centers
Table 3). These results are in good agreement with those obtained using metal fluorides as
catalysts [11].
2
oxide (0.70 mmol h m , Table 1), and 2/1Mg-Al
2
oxide (0.75 mmol h m ,
(
MgO
2
/1Mg-Al2
/1Mg-Al2
3
3
/2Mg-Al2
1/1Mg-Al2
2/3Mg-Al2
1/2Mg-Al2
1
/3Mg-Al2
Al O3
2
0
100
200
300
400
500
Temperature (℃)
Fig. 2. CO -TPD profiles for the catalysts examined in this study.
2
1
1

Table 3. Acid–base surface properties of MgO, Al
2
O , and Mg-Al mixed oxide samples
3
b
2
c
2
Densities of acidic sites (µmol m )
Densities of basic sites (µmol m )
Entry
Catalyst a
Wd,e
Md,e
Sd,e
Total evolved NH
0.22
Wd,e
Md,e
Sd,e
Total evolved CO
8.44
3
2
0
.15
0.07
0.0
0.77
3.89
3.78
1
2
3
4
5
6
7
8
9
MgO
(
68.9%) (31.1%) (0.0%)
.11 0.15 0.0
43.6%) (56.4%) (0.0%)
.15 0.36 0.0
28.5%) (71.5%) (0.0%)
.07 0.61 0.01
10.1%) (87.8%) (2.1%)
.06 0.58 0.07
8.5%) (81.7%) (9.8%)
.04 0.72 0.14
4.6%) (80.3%) (15.1%)
.0 0.89 0.24
0.0%) (78.8%) (21.2%)
.0 0.60 0.47
0.0%) (56.1%) (43.9%)
.00 0.31 1.09
0.0%) (22.4%) (77.6%)
(9.1%)
0.34
(46.1%) (44.8%)
1.64 0.77
(59.6%) (28.1%)
0.90 0.22
(68.1%) (17.3%)
1.40 0.18
(70.6%) (9.1%)
0.47 0.04
(66.2%) (5.2%)
0.49 0.03
(49.6%) (3.1%)
0.19 0.00
(36.4%) (0.0%)
0.26 0.0
(30.6%) (0.0%)
0
3/1Mg-Al
2/1Mg-Al
3/2Mg-Al
1/1Mg-Al
2/3Mg-Al
1/2Mg-Al
1/3Mg-Al
2
2
2
2
2
2
2
0.26
0.51
0.69
0.71
0.90
1.13
1.07
1.40
2.75
1.32
1.98
0.71
0.98
0.51
0.85
0.17
(
(12.3%)
0.20
0
(
(15.6%)
0.40
0
(
(20.3%)
0.20
0
(
(28.6%)
0.46
0
(
(47.3%)
0.32
0
(
(63.6%)
0.59
0
(
(69.4%)
0.15
0
0.02
0.0
2 3
Al O
(
(90.5%) (9.5%)
(0.0%)
aThe as-prepared Mg-Al mixed oxides are labeled as “a/b Mg-Al
2
2 3
”, where “a/b” represents the molar ratio of MgO and Al O in the mixed
oxides.
b
TPD of adsorbed CO
2
.
c
TPD of adsorbed NH
3
.
d
e
W, weak; M, medium; S, strong.
The percentage contribution of each peak is given in parentheses.
To compare the performance between 3/2Mg-Al
h lifetime of 3/2Mg-Al was evaluated preliminarily. As shown in Fig. 3, the conversion of HFiP
decreased slowly from 89.5% to 83.2% with HFE-356mmz selectivity unchanged, which indicated that
/2Mg-Al has a good stability in 500 h. Although the initial activity of 3/2Mg-Al (89.5%) is slightly lower
than that of MgF (94.3%), 3/2Mg-Al has the advantages of simple preparation and no use of corrosive HF,
so it still has good application value.
2
and MgF
2
(studied in our previous work), the 500
2
3
2
2
2
2
1
2

100
80
60
40
20
0
0
100
200
300
400
500
Time/h
Fig. 3. The selectivity of HFE-356mmz (red dots) and conversion of HFiP (black squares) over 500 h.
Reaction conditions: temperature = 240 °C; contact time = 8 s.
3
.3 Reaction mechanism
In our previous work, we proposed a reasonable reaction mechanism for the vapor-phase
methylation of HFiP using DMC over a metal fluoride [11]. Herein, the strong relationships between
catalytic activity and surface acid-base properties using mixed Mg-Al oxides correlate well with those
observed using metal fluorides, providing further support for the proposed acid-base synergistic
mechanism
.
In Mg-Al mixed oxides, the medium strength Lewis acid is attributed to the surface defect of
3
+
3+
2

2+
Al in Al -O -Mg ion, while the medium strength Lewis base is assignable to the strong
2

electronegativity O [21,22]. Accordingly, a rational reaction mechanism of gas-phase
methylation of HFiP and DMC catalyzed by Mg-Al binary catalyst is proposed in Scheme 2.
Firstly, the carbonyl oxygen atom of DMC is adsorbed to the unsaturated coordination center of
3
+
2

2+
Al atom and HFiP is absorbed onto the O atom in Al -O -Mg through hydrogen bonding
.
1
3

Thereafter, the oxygen atom in HFiP attacks the methoxy carbon atom in DMC to produce
HFE-356mmz, with CO and methanol liberated as volatile by-products and the catalyst regenerated. In
2
addition, dimethyl ether was also formed through the condensation of methanol.
Scheme 2. Mechanism for the methylation of HFiP with DMC over mixed Mg-Al oxide.
This reaction mechanism well explains the strong activity of the catalyst with medium Lewis
acidities and basicities (e.g., 3/2Mg-Al
facilitate the adsorption of both DMC and HFiP, as well as product desorption. If the catalyst is of strong
Lewis acidity (e.g., Al ), the carbonyl oxygen of DMC is strongly adsorbed onto the catalyst [23]. On
the other hand, if strong basic (e.g., MgO) or acidic molecules such as CO or HFiP itself is strongly
2
). Only catalysts with medium Lewis acidities and basicities
2
O
3
2
absorbed onto the catalyst [24], the catalyst activity would diminish seriously. Still, other mechanistic
possibilities exist taking into account the strong affinity of the F-atoms in fluorocarbons with alumina
based Lewis acids [25,26]. More experiments to further clarify the reaction mechanism are underway.
1
4

4. Conclusion
HFE-356mmz was successfully synthesized through the vapor-phase methylation of HFiP with
DMC using Mg-Al mixed oxide catalysts instead of metal fluorides. Among various mixed Mg-Al
oxides, the 3/2Mg-Al oxide showed the highest catalytic activity due to its moderate Lewis acidity and
basicity. The HFiP conversion and the activity of 3/2Mg-Al
respectively, moreover, 3/2Mg-Al gives good activity and stability in 500 h. The relationship between
2

1
2
2
were 89.4% and 0.94 mmol h m ,
2
catalytic activity and the surface acid-base properties of the metal oxide catalyst were found to correlate
well with those observed for metal fluorides, which provides further evidence of the acid-base
synergistic mechanism involving activation of DMC at a Lewis acidic site through its carbonyl oxygen
and activation of HFiP at an adjacent Lewis basic site through H-bonding. Our study provides deep
insight into this reaction by unambiguously elucidating its reaction mechanism. Since the preparation of
metal fluorides requires the use of HF, proving the applicability of metal oxides and confirming their
comparable catalytic activity to that of metal fluoride make the reaction more universal and operable.
The comprehensive mechanistic details provided, and our exploration of applicable catalysts for this
reaction will promote the practical use of this chemistry for the pollution-free and large-scale production
of HFE-356mmz.
Declaration of Conflict Interest
The authors declare that there is no conflict interest.
Acknowledgments
The authors are grateful for financial support from Sinochem, Lantian (grant no. ZHLT2015AC20).
1
5

References
[
[
[
1] R.E. Benedick, Int. Negot. 1 (1996) 231–246.
2] A. Sekiya, S. Misaki, J. Fluorine Chem. 101 (2000) 215–221.
3] P. Blowers, D.M. Moline, K.F. Tetrault, R.R. Wheeler, S.L. Tuchawena, Environ. Sci. Technol. 42
(2008) 1301–1307.
[
4] W. Li, S. Qin, Q. Zhang, Z. Guo, X. Zhang, Chem. Phys. Lett. 714 (2019) 125–130.
5] A. Yoshinori, C. Liang, K. Naoki, et al, Heat source machine and operating method therefor, Int.
Patent Appl. WO 2017138614 (2017).
[
[
6] C.C. Li, A method of cleaning using partially fluorinated ethers having a tertiary structure, Int. Patent
Appl. WO 9309272 (1993).
[
7] Q. An, Z. Lv, W. Shangguan, B. Qiao, W. Shangguan, Preparation method and application of
anti-fouling and anti-fingerprint fluorosilicone resin for touch screen glass, Chinese Patent CN
1
06866972 (2017).
[8] A.R. Massah, M. Mosharafian, A.R. Momeni, H. Aliyan, H.J. Naghash, M. Adibnejad, Synth.
Commun. 37 (2007) 1807–1815.
[
[
[
[
[
[
9] R.A.W. Johnstone, M.E. Rose, Tetrahedron 35 (1979) 2169–2173.
10] A. Leggio, A. Liguori, A. Napoli, C. Siciliano, G.J. Sindona, J. Org. Chem. 66 (2001) 2246–2250.
11] W. Li, G. Yang, F. Lu, X. Zhang, Appl. Catal., A 594 (2020) 117416.
12] T. Beutel, J. Chem. Soc. Faraday Trans. 94 (1998) 985–993.
13] T.M. Jyothi, T. Raja, M.B. Talawar, B.S. Rao, Appl. Catal. A 211 (2001) 41–46.
14] V.K. Díez, C.R. Apesteguía, J.I. Di Cosimo, J. Catal. 215 (2003) 220–233.
1
6

[15] M. Bolognini, F.Cavani, D. Scagliarini, C. Flego, C. Perego, M. Saba, Catal. Today 75 (2002)
1
03-111.
[
[
[
[
16] V. Crocellà, G. Cerrato, G. Magnacca, et al., Dalton Trans. 39 (2010) 8527-8537.
17] V.R.L. Constantino, T.J. Pinnavaia, Inorg. Chem. 34 (1995) 883–892.
18] F. Prinetto, G. Ghiotti, R. Durand, D. Tichit, J. Phys. Chem. B 104 (2000) 11117–11126.
19]M. León, E. Díaz, S. Bennici, A. Vega, S. Ordónez, A. Auroux, Ind. Eng. Chem. Res. 49 (2010)
3
663–3671.
[
[
[
[
[
20] S. Narayanan, K. Deshpande, Appl. Catal. A 135 (1996) 125–135.
21] V. Crocellà, G. Cerrato, G. Magnacca, Dalton Transactions 39(2010) 8527–8537.
22] F. Prinetto, G. Ghiotti, R. Durand, J. Phys. Chem. B 104(2000) 11117–11126.
23] A. Astruc, C. Cochon, S. Dessources, S. Célérier, S. Brunet, Appl. Catal. A 453 (2013) 20–27.
24] V. Crocellà, G. Cerrato, G. Magnacca, C. Morterra, F. Cavani, S. Cocchi, S. Passeri, D. Scagliarini,
C. Flego, C. Perego, J. Catal. 270 (2010) 125–135.
[
25] T. Stahl, H. F. T. Klare, M. Oestreich, ACS Catal. 3 (2010) 1578–1587.
26] A. Comas-Vives, M. Schwarzwälder, C. Copéret, P. Sautet, J. Phys. Chem. C 119 (2015) 7156–
[
7
163.
1
7