An efficient and sustainable production of triacetin from the acetylation of glycerol using magnetic solid acid catalysts under mild conditions

Article abstract of DOI:10.1016/j.cattod.2015.07.011

The efficient and selective acetylation of glycerol to produce triacetin is achieved using magnetic solid acids as catalysts. The Fe-based materials including Fe-Sn-Ti (OH)_x, Fe-Sn-Ti(SO₄^2-) and Fe-Sn-Ti(SO₄^2-)-t (t represents the temperature for the heating treatment) were successfully prepared and employed in the acetylation of glycerol, respectively. As a result, 100% conversion and 99.0% selectivity for triacetin was obtained in the presence of a catalytic amount of Fe-Sn-Ti(SO₄^2-)-400 at 80 °C for 30 min, which exhibits higher catalytic activity than those of some molecular sieves. The magnetic catalytic materials were respectively characterized by XRD, IR, TG-DTG, BET and NH₃-TPD techniques. Moreover, the effects of reaction temperature and reaction time in the glycerol acetylation are investigated in detail. Finally, based on the experimental results and reaction phenomena, a possible mechanism for the catalytic reaction is proposed.

Full text of DOI:10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
An efficient and sustainable production of triacetin
from the acetylation of glycerol using magnetic
solid acid catalysts under mild conditions
Jinyan Sun, Xinli Tong∗, Linhao Yu, Jun Wan∗
Tianjin Key Laboratory of Organic Solar Cells and Photochemical Conversion, School of Chemistry and Chemical Engineering,
Tianjin University of Technology, Tianjin 300384, P. R. China
a r t i c l e i n f o
a b s t r a c t
Article history:
The efficient and selective acetylation of glycerol to produce triacetin is achieved using magnetic solid
acids as catalysts. The Fe-based materials including Fe-Sn-Ti (OH)x, Fe-Sn-Ti(SO42−) and Fe-Sn-Ti(SO42−)-t
(t represents the temperature for the heating treatment) were successfully prepared and employed in the
acetylation of glycerol, respectively. As a result, 100% conversion and 99.0% selectivity for triacetin was
obtained in the presence of a catalytic amount of Fe-Sn-Ti(SO42−)-400 at 80 ◦C for 30 min, which exhibits
higher catalytic activity than those of some molecular sieves. The magnetic catalytic materials were
respectively characterized by XRD, IR, TG-DTG, BET and NH3-TPD techniques. Moreover, the effects of
reaction temperature and reaction time in the glycerol acetylation are investigated in detail. Finally, based
on the experimental results and reaction phenomena, a possible mechanism for the catalytic reaction is
proposed.
Received 4 May 2015
Received in revised form 4 July 2015
Accepted 11 July 2015
Available online xxx
Keywords:
Glycerol
Acetylation
Triacetin
Magnetic material
Solid acids
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
environmental and economic issues. For avoiding these problems,
the acetylation of glycerol with acetic acid was extensively studied
The rapid increase of biodiesel production has led to a wider
availability of crude glycerol as a by-product in chemical industry.
Thus, the efficient utilization of excessive and cheap glycerol are
becoming a new research topic at present, in which the etherifi-
cation [1,2] and esterification processes [3,4] to fuel additive are
considered as the promising route to conversion of glycerol.
The glycerol derivatives such as glycerol esters, acetyl glycerol
and glycerol acetal have good potential for the usage as fuel
components that assist to a decreasing in particles, hydrocarbons,
carbon monoxide and unregulated aldehydes emission. As shown
in Scheme 1, the acetylation of glycerol contains three main steps,
involving consecutive formations of monoacetin, diacetin, and
triacetin via esterification reaction in the presence of acetylation
reagents. In the previous study, the glycerol acetylation process was
often accelerated by the acid catalysts in homogeneous media, for
example, over sulfuric acid, hydrofluoric acid, p-toluenesulfonic
acid or acidic ionic liquids [5,6]. However, these mineral acid
catalysts have the obvious disadvantages such as the severe
in the presence of solid catalysts [3,7,8]. Generally, the used
heterogeneous catalysts included the acid exchange resins [9–11],
K-10 montmorillonite, HZSM-5 [3], HUSY [8], niobium-zirconium
mixed oxides [12], PMo-NaUSY [13], supported heteropolyacids
[14–16], mesoporous silica [17], and sulfonic acid functionalized
deoxycellulose [18], sulfated zirconia and niobic acid [19] etc.
Particularly, the cation-exchange resins showed the high catalytic
activity as well as the selectivity of higher esters where the triacetin
are lower compared to diacetin. Very recently, Huang et al. [20]
reported that the acetylation of glycerol to triacetin was achieved
with the heteropolyacid-based ionic liquids as catalysts in which
a 100% conversion of glycerol and a superior selectivity (86–99%)
was obtained. Khayoon et al. [21] found that the Y/SBA-3 catalyst
could promote the formation of diacetin and triacetin in the
glycerol acetylation where the selectivity of 34% and 55% toward
diacetin and triacetin was achieved under suitable conditions. Kale
et al. [22] investigated the esterification of glycerol with acetic
acid over acidic Amberlyst ion-exchange resins in the presence of
toluene as an entrainer. As a result, a 95% selectivity of triacetin at
complete glycerol conversion was obtained.
In this study, a series of magnetic solid catalysts and molecu-
lar sieves were successfully prepared and further employed in the
acetylation of glycerol under mild conditions. It is found that a 100%
∗
Corresponding authors. Tel./fax: +86 22 60214259.
E-mail addresses: tongxinli@tju.edu.cn, tongxli@hotmail.com (X. Tong),
wanjuntjlg@sina.com (J. Wan).
http://dx.doi.org/10.1016/j.cattod.2015.07.011
0920-5861/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
2
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
O
OH
O
O
OH
O
H3CCO
OCCH3
HO
HO
OCCH3
O
OH
HO
OCCH3
acetylation
acetylation
O
O
H3CCO
OCCH3
OH
OCCH3
OCCH3
OH
O
HO
OCCH3
O
Scheme 1. The reaction route for the acetylation of glycerol.
2.3. The synthesis of Fe-Sn-Ti (SO42−) solid acid
conversion and 99.0% selectivity for triacetin was obtained using
the Fe-Sn-Ti(SO42−)-400 as the catalyst at 80 ◦C for 30 min. More-
over, based on the reaction phenomena and experimental results,
a possible reaction mechanism is proposed to explain the catalytic
process.
Stannic chloride pentahydrate (17.5 g) is added in three-neck
flask. Under rapid stirring, 15 mL magnetic matrix (0.1 mol/L) and
10 mL tetrabutyl titanate are poured into the above flask, and then
NH3·H2O is slowly dropped into the mixture to form the pasty liq-
uids. The produced solid is filtered and washed numerous times
to be neutral. Then, the solid sample is dried at 100 ◦C in oven,
and the obtained material is referred as Fe-Sn-Ti (OH)x. In the fol-
lowing, the above solid product is further sulfated through being
soaked with 1 mol/L (NH4)2SO4 solution for 24 h, and then filtered,
dried to obtain the sulfated Fe-Sn-Ti material that is referred as
Fe-Sn-Ti (SO42−). In addition, the sulfated Fe-Sn-Ti materials are
calcined at different temperatures. Therein, the obtained samples
calcined at 400 ◦C, 500 ◦C, 600 ◦C and 700 ◦C are signified by the
Fe-Sn-Ti (SO42−)-400, Fe-Sn-Ti (SO42−)-500, Fe-Sn-Ti (SO42−)-600
and Fe-Sn-Ti (SO42−)-700 in the next discussion.
2. Experimental
2.1. Reagents and instruments
Glycerol, ferrous sulfate, ferric sulfate, stannic chloride pentahy-
drate, acetic acid, ammonia, tin(II) acetate, NaOH, acetic anhydride,
ammonium sulfate, HZSM-5, H-␤ zoelite, nitric acid, deionized
water and ethanol are analytic grade and got from commercial
sources.
The monoacetin, diacetin and triacetin as the standard samples
are purchased from Alfa Aesar.
The measurement of X-ray diffraction (XRD) was performed by
diffractometer with Cu K␣ radiation (0.02◦ resolution) and was col-
lected from 20 to 80◦ [2␪]. The spectra of Fourier transform infrared
spectroscopy (FT-IR) are recorded on a Nicolet Nexus spectrome-
ter in the 400–4000 cm−1 range. The thermal analysis (TG-DTG) is
performed with NETASCH TG 209F3 instrument, and the data are
shown from 0–1000 ◦C. BET surface areas, pore volumes, and aver-
age pore diameters of the prepared samples are obtained from N2
(77 K) adsorption measurement using a Micromeritics ASAP2020 M
system, in which the samples are pretreated under vacuum at
150 ◦C for 4 h before the measurement. The average pore diameters
are calculated according to Barrett–Joyner–Halenda (BJH) model in
absorption and desorption period. The acid properties of the mag-
netic solid catalysts was determined by temperature-programmed
desorption of ammonia (NH3-TPD). Before the adsorption of ammo-
nia the samples were treated under helium at 500 ◦C (from 25 to
500 ◦C in 40 min) for 1 h. The samples were then cooled to 25 ◦C
in He flow, then treated with a NH3 flow for 30 min at 100 ◦C.
The physisorbed ammonia was eliminated by flowing He for 1 h at
100 ◦C. NH3-TPD was run between 100 ◦C and 950 ◦C at 10 ◦C/min
and followed by an online gas chromatograph (GC) provided with
a thermal conductivity detector.
2.4. The preparation of deAl-ˇ and Sn-ˇ zoelites
The preparation of the deAl-␤ and Sn-␤ zoelites is similar with
the reference [23]. In brief, 1) deAl-␤ zeolite is synthesized via the
dealumination of H-␤. The commercial H-␤ zeolite was dealumi-
nated by treatment in HNO3 solution (13 M) at 100 ◦C for 20 h [20
mL·g-1(zeolite)]. 2) The Sn-␤ zeolite is synthesized by the solid-
state ion-exchange (SSIE) process. The appropriate amount of tin(II)
acetate was grinded with the required amount of dealuminated
zeolite for 15 min. In addition, these prepared zeolites were cal-
cined in an air flow at 550 ◦C before being used.
2.5. General procedure for the acetylation of glycerol
All the acetylations of glycerol with acetic anhydride or acetic
acid were performed in a 120 mL steel autoclave equipped with
the magnetic stirring and a temperature controller. Typical proce-
dure for catalytic process is given as follows: a 1.5 g glycerol, 8.39 g
acetic anhydride and the catalyst (2.5 wt.%) were charged into the
autoclave; Under stirring, the mixture was preheated to 80 ◦C and
kept for 30 min after the reactor was sealed. After the reaction, the
mixture was transferred to a volumetric flask and was diluted with
anhydrous ethanol. The conversion of glycerol and the selectivity
of product were attained using the gas chromatograph with the
internal standard method.
The quantitative analyses of the products are performed on
a GC apparatus with FID detector. The capillary column is HP-
5, 30 m × 0.25 mm × 1.0 ␮m. In addition, the qualitative analysis
for the product is carried out on the Agilent 6890/5973 Gas
Chromatograph-Mass Spectrometer (GC-MS) instrument.
3. The physical properties of solid acid catalysts
2.2. The preparation of magnetic matrix and its treatment method
3.1. The IR spectra of different catalysts
The typical procedure is given in the following: a mixture of fer-
rous sulfate/ferric sulfate [Fe2SO4/Fe2(SO4)3] was added in a vessel
and solved with deionized water. The solution was then heated
to 45 ◦C and NH3·H2O was added to adjust pH value to 10–11
using acidometer. The reaction was kept for 1 h. The produced
magnetic matrix was separated with magnetic separation method
and washed repeatedly to the neutrality. The magnetic matrix was
stored with suspension liquid to use.
Figure 1 shows the IR spectra of different catalytic materials
including Fe-Sn-Ti(OH)x, Fe-Sn-Ti (SO42−), Fe-Sn-Ti (SO42−)-400
and Fe-Sn-Ti (SO42−)-600. It was found that the peaks of
3100–3250 cm-1 and 3420–3550 cm-1 in the spectrum of Fe-Sn-
Ti(OH)x are attributed to the stretching vibration of combined
water and O H bond. Moreover, these peaks of water and O
H
bond become weak after the Fe-Sn-Ti(OH)x is sulfated. Moreover,
the characteristic peaks of the sulfated metal oxides often occur
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
3
Fig. 1. The IR spectra of different magnetic catalytic materials.
Fig. 2. The XRD patterns of different magnetic catalytic materials.
between 900 and 1400 cm−1, assigned to the stretching vibrations
of the S O bond or S O bond [24]. In the spectra of Fe-Sn-Ti (SO42−),
Fe-Sn-Ti (SO42−)-400 and Fe-Sn-Ti (SO42−)-600, the typical mew
peaks of 1095.5 cm−1, 1141.8 cm−1 and 1201.6 cm−1 represent the
formation of the superacid structures.
graph of Fe-Sn-Ti (SO42−)-400, while the loss of weight appears
in the region of 620-830 ◦C in the graph of Fe-Sn-Ti (SO42−)-600
2−
material. These indicates that the chelation of SO4 and metal
O
bond has different states. The structure of superacid was par-
tially destroyed after being calcined at 600 ◦C. The acid sites of
Fe-Sn-Ti (SO42−)-600 is more confined than Fe-Sn-Ti (SO42−)-400
material.
3.2. XRD patterns
In Fig. 2, the XRD patterns for the magnetic materials Fe-Sn-Ti
(OH)x, Fe-Sn-Ti (SO42−), Fe-Sn-Ti (SO42−)-400, Fe-Sn-Ti (SO42−)-
600 and Fe-Sn-Ti (SO42−)-700 are presented in which the crystal
structures of these materials is indicated. The peaks at 26.749◦,
34.061◦ and 52.228◦ are the characteristic diffraction of Titanium
3.4. BET measurement
Textural properties of the catalyst materials derived from the
nitrogen physisorption are shown in Table 1. The results indicate
that the surface area of solid catalyst is increased from 16.96 m2 g−1
to 18.88 m2 g−1 and the pore volume is also elevated to from
0.13 cm3 g−1 to 0.15 cm3 g−1 after being calcined at 400 ◦C. Then,
the surface area is decreased along with the increase of the cal-
cined temperature and the surface area is only 11.64 m2 g−1 for
Fe-Sn-Ti (SO42−)-700 catalyst. Moreover, there is a certain increase
Tin Oxide [(Ti0.3Sn0.7)O2]. It indicates that there is the weak rela-
2−
tionship of Ti-O-Sn bond in the Fe-Sn-Ti(OH)x and Fe-Sn-Ti (SO4
)
materials. Furthermore, the new peaks at 38.100◦, 54.935◦, 58.355◦,
62.258◦ and 65.185◦ begin to appear in the materials of Fe-Sn-
Ti (SO42−)-400, Fe-Sn-Ti (SO42−)-500 and Fe-Sn-Ti (SO42−)-600
which are attributed to the formation of cassiterite. This shows that
there exists a mutual function between Sn and Fe element for these
catalytic materials. Otherwise, more peaks at 39.491◦ and 72.757◦
is shown up in the Fe-Sn-Ti (SO42−)-700 material. It is probably
due to the stronger function appearing between Sn-O-Ti and Fe3O4
support after high temperature treatment.
in the average pore diameters with the elevation of the calcination
2−
temperature, in which the value of 2.8240 nm for Fe-Sn-Ti (SO4
)
is increased to the value of 8.0875 nm for Fe-Sn-Ti (SO42−)-700
material.
3.3. TG-DTG detection
3.5. NH3-TPD detection
The detection of TG-DTG of magnetic catalytic materials was
performed and the results are shown in Fig. 3. It was found that
there exists two obvious regions for the weight loss in the curve
of Fe-Sn-Ti (OH)x material in which the scope of temperature
is 40–100 ◦C and 130–600 ◦C domain. The first loss of weight is
attributed to the remove of water absorbed in the surfaces of the
material. The second weight loss should be due to the dehydra-
tion of Sn(OH)x and Ti(OH)4 to form the oxides. In the TG-DTG
graph of Fe-Sn-Ti (SO42−), the loss of weight appear in the region
of 50–140 ◦C, 160–305 ◦C, 315–385 ◦C, 400–490 ◦C and 520–670 ◦C,
respectively. The first and second loss of weight is contributed to
the remove of water absorbed and hydroxide groups in the sur-
faces of the catalysts. The next other three losses of weight are
contributed to the remove of SOx and the breakdown of acid struc-
ture. Besides, the TG-DTG detection of Fe-Sn-Ti (SO42−)-400 and
Fe-Sn-Ti (SO42−)-600 materials was also performed, and the loss of
weight appears in the region of 410-580 ◦C and 590-820 ◦C in the
In order to study acidic properties of magnetic solid catalysts,
the NH3-TPD was performed from 100 to 950 ◦C, and the experi-
mental results are presented in the Fig. 4. It can be seen that three
peaks (212 ◦C, 529 ◦C and 718 ◦C) occur on the NH3-TPD profile of
the Fe-Sn-Ti (SO42−)-400 catalyst. Herein, the first peak showed the
existence of weak acidic sites, and the second peak indicated the
co-existence of medium and strong acidic sites [25,26]. The third
peak should be contributed to the self-decomposition of the cata-
lyst. Similarly, there exist three kinds of peaks on the profile of the
Fe-Sn-Ti (SO42−)-500 catalyst, while the amount of acidic sites (cor-
responds to peak area) is less than that of the Fe-Sn-Ti (SO42−)-400
catalyst. In case of the Fe-Sn-Ti (SO42−)-600 catalyst, the amount of
acidic sites become much less than the above two catalysts. Espe-
cially, the weak acid nearly disappeared. Furthermore, no apparent
peak come out on the NH3-TPD profile of the Fe-Sn-Ti (SO42−)-700
catalyst, which showed that no acidic sites were generated on this
catalyst.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
4
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
Fig. 3. The TG-DTG result for different magnetic catalytic materials [(1) refers to Fe-Sn-Ti (OH)x; (2) refers to Fe-Sn-Ti (SO42−); (3) refers to Fe-Sn-Ti (SO42−)-400; (4) refers
to Fe-Sn-Ti (SO42−)-600].
4. Results and discussion
(entry 2). To our surprise, the conversion is increased to 100% and
selectivity of triacetin reaches 99% when the Fe-Sn-Ti(SO42−)-400
was employed as the catalyst (entry 3). Moreover, a 99% conver-
sion of glycerol in 97% or 96% selectivity of triacetin was obtained
in the presence of Fe-Sn-Ti (SO42−)-500 or Fe-Sn-Ti (SO42−)-600
catalyst (entries 4 and 5). This should be due to the promotion of
the formed structure of superacid in the reaction. In addition, based
on the results of NH3-TPD measurements, the amount of acid on the
Fe-Sn-Ti (SO42−)-600 catalyst is much less than that on the Fe-Sn-
Ti (SO42−)-400 catalyst; besides, there very little weak acidic sites
exist on the Fe-Sn-Ti (SO42−)-600 catalyst. Thus, it can be concluded
that the catalytic activity is closely related to the acid strength of
catalyst and the amount of acid is not a determining factor. How-
ever, the conversion of glycerol is reduced to 68% and monoacetin
become main product as much as 82% selectivity in the presence
of Fe-Sn-Ti (SO42−)-700 catalyst (entry 6). It is comparable to the
4.1. The atecylation of glycerol to triacetin by the magnetic
catalytic material
In the beginning, the catalytic performance of different solid
acids is first investigated in the acetylation of glycerol with acetic
anhydride (shown in Scheme 2), in which a series of magnetic
materials, zeolites and Amberlyst-15 was employed as the cata-
lysts. As comparison, the blank experiment was also performed.
The obtained experimental results are summarized in Table 2.
Seen from the experimental data, it is found that 49% conversion
of glycerol was obtained in the absence of any catalyst where the
selectivity of monoacetin, diacetin and triacetin is 48%, 46% and
6%, respectively (entry 1). With the Fe-Sn-Ti (SO42−) as the cata-
lyst, a 58% conversion and 83% selectivity of monoacetin is attained
Table 1
the BET data for different acid catalysts.
Catalysts
BET surface area (m2 g−1
)
Pore volume (cm3 g−1
BJH adsorption
)
Average pore diameter (nm)
BJH desorption
BJH adsorption
BJH
desorption
2−
Fe-Sn-Ti(SO4
)
16.9606
18.8818
16.5536
14.9134
11.6472
0.127524
0.145857
0.134225
0.123552
0.077765
0.128372
0.153853
0.141268
0.134966
0.110199
2.8240
4.2046
5.2298
6.1426
8.0875
2.9477
3.7972
4.6957
5.6663
9.2812
Fe-Sn-Ti(SO42−)-400
Fe-Sn-Ti(SO42−)-500
Fe-Sn-Ti(SO42−)-600
Fe-Sn-Ti(SO42−)-700
˚
˚
Volume of pores cumulated between 17.000 A and 3,000.000 A diameter.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
5
O
O
O
O
OH
O
OH
OH
HO
OCCH3
acetic anhydride
solid catalyst
+
+
H3CCO
OCCH3
HO
OCCH3
OH
+
H3CCO
OCCH3
O
Scheme 2. The catalytic acetylation of glycerol with acetic anhydride in the presence of acid catalyst.
Table 2
The acetylation of glycerol and acetic anhydride with different catalystsa
Entry
Catalyst
Conv. (%)b
Product distribution (%)b
Monoacetin
Diacetin
Triacetin
1
2
3
4
5
6
7
8
Blank
Fe-Sn-Ti (SO4
49
58
100
99
99
68
89
74
99
99
99
48
83
0
0
0
82
59
78
14
0
46
17
1
3
4
17
37
21
62
1
6
0
99
97
96
1
4
1
24
99
26
2−
)
Fe-Sn-Ti (SO42−)-400
Fe-Sn-Ti (SO42−)-500
Fe-Sn-Ti (SO42−)-600
Fe-Sn-Ti (SO42−)-700
Sn-␤
deAl-␤
HZSM-5
9
10
Amberlyst-15
11c
Fe-Sn-Ti (SO42−)-400
13
61
a
Reaction conditions: glycerol 1.5 g, acetic anhydride 8.39 g and 0.05 g catalyst is added, at 80 ◦C, for 30 min under magnetic stirring.
The data are obtained by GC with the internal standard method.
About 1 mL H2O is added into reaction mixture.
b
c
result of acetylation reaction using uncalcined Fe-Sn-Ti (SO42−) cat-
alyst which matches the disappear of superacid after being treated
at 700 ◦C high temperature. Furthermore, some common acid cat-
alysts were also studied in the acetylation of glycerol under similar
conditions. As a result, it was found that 89% or 74% conversions
was obtained using the molecular sieve Sn-␤ or deAl-␤ as the cata-
lyst, respectively; meanwhile, the major products are monoacetin
and diacetin (entries 7 and 8). In the presence of HZSM-5 cata-
lyst, the conversion of glycerol arrived at 99%, and the selectivity
of diacetin and triacetin was respectively 62% and 24% (entry 9),
which can be attributed to the occurrence of the stronger acidic
sites. Also, the Amberlyst-15 is employed as the solid catalyst and
a 99% conversion in 99% selectivity of triacetin was attained (entry
10). This is near to the data with magnetic solid catalytic mate-
rials. On the other hand, with the Fe-Sn-Ti (SO42−)-400 catalyst,
the selectivity of triacetin is decreased to 26% and the selectivity
of monoacetin and diacetin is elevated to 13% and 61% when 1 mL
pure H2O is added into reaction mixture; meanwhile, the conver-
sion of glycerol was still 99% (entry 11). It exhibits that more water
is disadvantageous to the production of triacetin in this catalytic
process.
4.2. The effect of temperature on the acetylation of glycerol
In the following, the influence of temperature on the acetylation
of glycerol with acetic anhydride in the presence of the Fe-Sn-Ti
(SO42−)-400 catalyst was investigated and the results are shown
in Fig. 5. It can be seen that the conversion of glycerol is gradually
elevated, and the selectivity of monoacetin decreases and the selec-
tivity of triacetin slowly increases along with the temperature being
increased from 50 ◦C to 80 ◦C. Furthermore, the conversion and the
product distribution is almost unchanged when the temperature
was increased to 90 ◦C in the reaction.
Fig. 5. Effect of temperature in the acetylation of glycerol with Fe-Sn-Ti (SO42−)-400
Fig. 4. The NH3-TPD profiles of different magnetic solid catalysts.
catalyst.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
6
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
Fig. 6. Effect of time in the acetylation of glycerol with Fe-Sn-Ti(SO42−)-400 catalyst.
4.3. Effect of reaction time on the acetylation of glycerol
4.4. The acetylation of glycerol with acetic acid under different
temperatures
Fig. 6 showed the effect of reaction time on the acetylation
of glycerol with acetic anhydride in the presence of the Fe-Sn-Ti
(SO42−)-400 catalyst. As a result, it was found that the conver-
sion of glycerol gradually increases along with the reaction time
being increased from 10 min to 30 min, in which the selectivity
of monoacetin is slowly decreased, and the diacetin selectivity
was firstly increased to top value and then decreased, and the
selectivity of triacetin was increased all the time. These charac-
ters are completely suitable for the typical consecutive reaction.
Moreover, the conversion of glycerol and product selectivity almost
keep unchanged after being further reacted for 40 min and more
period.
In order to further reveal the character of acetylation of glycerol
using magnetic catalytic materials, the acetic acid is employed as
the acetylating reagent at different temperatures. The experimen-
tal results are presented in Fig. 7. As a result, the main product
is monoacetin and the by-product is diacetin when acetic acid
replaced the acetic anhydride in the reaction. Therein, a 32.3% con-
version and 100% selectivity of monoacetin was obtained when
the reaction was performed at 80 ◦C. Along with the elevation of
temperature, the conversion was gradually increased and the selec-
tivity was slowly reduced. When the temperature is increased to
140 ◦C, the conversion and selectivity of monoacetin is respectively
75.7% and 90.3%.
Fig. 7. The acetylation of glycerol with acetic acid by Fe-Sn-Ti (SO42−)-400 catalyst
(Reaction conditions: 1.5 g glycerol, acetic acid 9.79 g and 0.05 g catalyst is employed,
for 30 min, under the magnetic stirring).
Fig. 8. The recycling of the Fe-Ti-Sn(SO42−)-400 catalyst in the acetylation of glyc-
erol.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
7
Fig. 9. Reaction mechanism on the acetylation of glycerol with magnetic catalyst.
4.5. The investigation on the recycling of catalyst
5. Conclusions
The recycling of the magnetic catalyst has been examined. It
should be mentioned that the catalysts is easily separated from
the reaction mixture owing to its own magnetic performance. The
investigations are performed at 80 ◦C for 30 min in the presence
of Fe-Ti-Sn (SO42−)-400 catalyst. After the reaction, the solid cata-
lyst is seperated from the reaction solution with the assistance of
the magnet. It is first washed with ethanol for 3 times, and then
is dried and calcined at 400 ◦C for 3 h in the oven. The obtained
magnetic catalyst is further used in the next run by adding the
glycerol and acetic anhydride as reactants. In Fig. 8, it can be
seen that the yield of triacetin almost keeps unchanged in the
former three recycled reactions which shows that the catalyst
retains a very high activity for the acetylation of glycerol. Fur-
thermore, the yield of triacetin decreases a little (about 9%) after
being recycled four times. All these results indicate that the mag-
netic catalytic material is efficient and recycled in the glycerol
acetylation.
The efficient conversion of glycerol to triacetin with acetin anhy-
dride has been successfully performed with magnetic solid acid
catalysts. A 100% conversion of glycerol and 99.0% selectivity of
triacetin is obtained using the Fe-Sn-Ti (SO42−)-400 catalyst at
80 ◦C for 30 min. This exhibits that the magnetic catalytic mate-
rials is very promising in the utilization of glycerol. Moreover,
there exist the special effects on the reaction temperature and
time in the acetylation of glycerol. Furthermore, a possible reac-
tion mechanism is proposed to explain the high conversion and
good selectivity of this catalytic process.
Acknowledgements
This research work was financially supported by the National
Natural Science Foundation of China (21003093) and the State
Key Program of the National Natural Science Foundation of China
(21336008).
References
4.6. The reaction mechanism for the acetylation of glycerol
[1] H.J. Lee, D. Seung, K.S. Jung, H. Kim, I.N. Filimon, Etherification of glycerol by
isobutylene: tuning the product composition, Appl. Catal. A Gen. 390 (2010)
235–244.
Based on the obtained results and reaction phenomena, a pos-
sible catalytic reaction mechanism has been proposed (see Fig. 9).
First, acetic anhydride can function with solid acid catalyst in which
carbonyl group is activated with hydrogen proton or Lewis acid site.
Then, the activated carbonyl of acetic anhydride is attacked by the
glycerol molecule to generate the monoacetin along with the lose
of a acetic acid molecule. Furthermore, the monoacetin reacts with
activated carbonyl group of another molecular acetic anhydride,
which leads to produce the diacetin (e.g. 2-hydroxypropane-1,3-
diyl diacetate and 3-hydroxypropane-1,2-diyl diacetate). Finally,
the activated carbonyl group of third acetic anhydride is further
attacked by the diacetin, and the triacetin is obtained with the lose
of a molecular acetic acid. According the principle of reversible
reaction, the diacetin and triacetin can probably be converted to
monoacetin and diacetin during the whole process if there exists
a certain water molecule (shown in dotted line). This also explains
the high selectivity of triacetin when acetic anhydride is used which
is contributed to very small amount of molecular water in that
situation.
[2] F. Frusteri, F. Arena, G. Bonura, C. Cannilla, L. Spadaro, O. Blasi, Catalytic
etherification of glycerol by tert-butyl alcohol to produce oxygenated
additives for diesel fuel, Appl. Catal. A Gen. 367 (2009) 77–83.
[3] V.L.C. Goncalves, B.P. Pinto, J.C. Silva, C.J.A. Mota, Acetylation of glycerol
catalyzed by different solid acids, Catal. Today 133 (2008) 673–677.
[4] P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Acetylation of
glycerol over heteropolyacids supported on activated carbon, Catal. Commun.
12 (2011) 573–576.
[5] J. Gui, X. Cong, D. Liu, X. Zhang, Z. Sun, Novel Brønsted acidic ionic liquid as
efficient and reusable catalyst system for esterification, Catal. Commun. 5
(2004) 473–477.
[6] D.C. Forbes, K.J. Weaver, Brønsted acidic ionic liquids: the dependence on
water of the Fischer esterification of acetic acid and ethanol, J. Mol. Catal. A
Chem. 214 (2004) 129–132.
[7] L. Zhou, E. Al-Zaini, A.A. Adesin, Catalytic characteristics and parameters
optimization of the glycerol acetylation over solid acid catalysts, Fuel 103
(2013) 617–625.
[8] I. Kim, J. Kim, D. Lee, A comparative study on catalytic properties of solid acid
catalysts for glycerol acetylation at low temperatures, Appl. Catal. B Environ.
148–149 (2014) 295–303.
[9] A. Heidekum, M.A. Harmer, Addition of carboxylic acids to cyclic olefins
catalyzed by strong acidic ion-exchange resins, J. Catal. 181 (1999) 217–222.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011

G Model
CATTOD-9682; No. of Pages8
ARTICLE IN PRESS
8
J. Sun et al. / Catalysis Today xxx (2015) xxx–xxx
[10] K. Klepacova, D. Mravec, M. Bajus, Brønsted acidic ionic liquids: the
dependence on water of the Fischer esterification of acetic acid and ethanol,
Appl. Catal. A Gen. 294 (2005) 141–147.
[11] I. Dosuna-Rodríguez, E.M. Gaigneaux, Glycerol acetylation catalysed by ion
exchange resins, Catal. Today 195 (2012) 14–21.
[18] I. Kim, J. Kim, D. Lee, Sulfonic acid functionalized deoxycellulose catalysts for
glycerol acetylation to fuel additives, Appl. Catal. A Gen. 482 (2014) 31–37.
[19] M.L. Testa, V.L. Parola, L.F. Liotta, A.M. Venezia, Screening of different solid
acid catalysts for glycerol acetylation, J. Mol. Catal. A Chem. 367 (2013)
69–76.
[12] P. Lauriol-Garbey, G. Postole, S. Loridant, A. Auroux, V. Belliere-Baca, P. Rey,
Acid-base properties of niobium-zirconium mixed oxide catalysts for glycerol
dehydration by calorimetric and catalytic investigation, Appl. Catal. B Environ.
106 (2011) 94–102.
[13] P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, Acid -base properties of
niobium-zirconium mixed oxide catalysts for glycerol dehydration by
calorimetric and catalytic investigation, Catal. Commun. 10 (2009) 481–484.
[14] S.K. Bhorodwaj, D.K. Dutta, Activated clay supported heteropoly acid catalysts
for esterification of acetic acid with butanol, Appl. Clay Sci. 53 (2011)
347–352.
[20] M. Huang, X. Han, C. Hung, J. Lin, P. Wu, J. Wu, S. Liu, Heteropolyacid-based
ionic liquids as efficient homogeneous catalysts for acetylation of glycerol, J.
Catal. 320 (2014) 42–51.
[21] M.S. Khayoon, S. Triwahyono, B.H. Hameed, A.A. Jalil, Improved production of
fuel oxygenates via glycerol acetylation with acetic acid, Chem. Eng. J. 243
(2014) 473–484.
[22] S. Kale, S.B. Umbarkar, M.K. Dongare, R. Eckelt, U. Armbruster, A. Martin,
Selective formation of triacetin by glycerol acetylation using
acidicion-exchange resins as catalyst and toluene as an entrainer, Appl. Catal.
A Gen. 490 (2015) 10–16.
[15] P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Glycerol
acetylation over dodecatungstophosphoric acid immobilized into a silica
matrix as catalyst, Appl. Catal. B Environ. 91 (2009) 416–422.
[16] P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Esterification of
glycerol with acetic acid over dodecamolybdophosphoric acid encaged in USY
zeolite, Catal. Commun. 10 (2009) 481–484.
[17] J.A. Melero, R. van Grieken, G. Morales, M. Paniagua, Acidic Mesoporous Silica
for the Acetylation of Glycerol: Synthesis of Bioadditives to Petrol Fuel, Energ.
Fuel 21 (2007) 1782–1791.
[23] C. Hammond, S. Conrad, I. Hermans, Simple and scalable preparation of highly
active Lewis acidic Sn-beta, Angew. Chem. Int. 51 (2012) 11736–11739.
[24] C. Guo, S. Yao, J. Cao, Z. Qian, Alkylation of isobutane with butenes over solid
superacids, SO42-/ZrO2 and SO42-/TiO2, Appl. Catal. A Gen. 107 (1994)
229–238.
[25] H. Atia, U. Armbruster, A. Martin, Dehydration of glycerol in gas phase using
heteropolyacid catalysts as active compounds, J. Catal. 258 (2008) 71–82.
[26] Y.T. Kim, K.-D. Jung, E.D. Park, Gas-phase dehydration of glycerol over
silica-alumina catalysts, Appl. Catal. B Environ. 107 (2011) 177–187.
Please cite this article in press as: J. Sun, et al., An efficient and sustainable production of triacetin from the acetylation of glycerol using
magnetic solid acid catalysts under mild conditions, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.07.011