Reusable gypsum based catalyst for synthesis of glycerol carbonate from glycerol and urea

Article abstract of DOI:10.1016/j.apcata.2015.06.024

In this study, the catalytic carbonylation of glycerol with urea in the presence of gypsum based heterogeneous catalyst is reported for the first time. Gypsum (CaSO₄·2H₂O) is one of the two calcium sulphate minerals found in nature a

Full text of DOI:10.1016/j.apcata.2015.06.024

Accepted Manuscript
Title: Reusable Gypsum Based Catalyst for Synthesis of
Glycerol Carbonate from Glycerol and Urea
Author: Nor Ain Syuhada Zuhaimi Vidhyaa Paroo Indran
Mohd Asyrak Deraman Nor Farihan Mudrikah Gaanty Pragas
Maniam Yun Hin Taufiq-Yap Mohd Hasbi Ab. Rahim
PII:
S0926-860X(15)30040-5
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2015.06.024
APCATA 15436
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
28-11-2014
2-6-2015
19-6-2015
Please cite this article as: Nor Ain Syuhada Zuhaimi, Vidhyaa Paroo Indran,
Mohd Asyrak Deraman, Nor Farihan Mudrikah, Gaanty Pragas Maniam, Yun Hin
Taufiq-Yap, Mohd Hasbi Ab.Rahim, Reusable Gypsum Based Catalyst for Synthesis
of Glycerol Carbonate from Glycerol and Urea, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2015.06.024
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. 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.

Reusable Gypsum Based Catalyst for Synthesis of Glycerol Carbonate
from Glycerol and Urea
a
a
a
a
Nor Ain Syuhada Zuhaimi , Vidhyaa Paroo Indran , Mohd Asyrak Deraman , Nor Farihan Mudrikah , Gaanty
a,b
c
a,d*
Pragas Maniam ,Yun Hin Taufiq-Yap and Mohd Hasbi Ab. Rahim
a
Faculty of Industrial Sciences & Technology, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300
Kuantan, Pahang, MALAYSIA.
Central Laboratory, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Kuantan, Pahang,
b
MALAYSIA.
c
Catalysis Science and Technology Research Centre, Faculty of Science, Universiti Putra Malaysia,
4
3400 Serdang, Selangor, MALAYSIA.
d
Centre for Earth Resources Research & Management, Universiti Malaysia Pahang, Lebuhraya Tun
Razak, 26300 Kuantan, Pahang, MALAYSIA.
*
E-mail: mohdhasbi@ump.edu.my
Highlights
•
•
•
•
•
First report on utilizing gypsum as catalyst for synthesis of organic carbonate.
Different pre-treatment produced different crystalline structure of catalyst.
The yield of glycerol carbonate depends on the specific phase of gypsum.
Accelerated formation of glycerol carbonate via glycerol carbamate intermediate.
The catalyst is reusable.
Graphical abstractReusable Gypsum Based Catalyst for Synthesis of Glycerol
Carbonate from Glycerol and Urea
a
a
a
Nor Ain Syuhada Zuhaimi , Vidhyaa Paroo Indran , Mohd Asyrak Deraman , Nor Farihan
a
a,b
c
a,d*
Mudrikah , Gaanty Pragas Maniam ,Yun Hin Taufiq-Yap and Mohd Hasbi Ab. Rahim
a
Faculty of Industrial Sciences & Technology, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Kuantan,
Pahang, MALAYSIA.
b
Central Laboratory, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Kuantan, Pahang, MALAYSIA.
c
Catalysis Science and Technology Research Centre, Faculty of Science, Universiti Putra Malaysia, 43400 Serdang,
Selangor, MALAYSIA.
d
Centre for Earth Resources Research & Management, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300
Kuantan, Pahang, MALAYSIA.
*
E-mail: mohdhasbi@ump.edu.my

Abstract
In this study, the catalytic carbonylation of glycerol with urea in the presence of gypsum based
heterogeneous catalyst is reported for the first time. Gypsum (CaSO ·2H O) is one of the two
4 2
calcium sulphate minerals found in nature and also one of the waste materials produced from
advanced material industrial processing plant. The effect of physical and chemical pre-treatment
procedures on gypsum was investigated. To obtain the catalyst structure-activity relationship, the
treated catalysts were characterized by means of several characterization techniques (i.e. XRD,
2 3
TGA, BET surface area, SEM, FTIR, CO -TPD, NH -TPD and Hammett test). Tuneable
physico-chemical properties of gypsum based catalysts were successfully prepared by varying
the pre-treatment techniques, which later on contributed to the variation of catalytic activity
toward glycerol carbonate formation from glycerol. The highest catalytic activity obtained was
4
for catalyst consisting ß-CaSO phase where it produced 92.8% conversion of glycerol, 90.1%
selectivity and 83.6% yield of glycerol carbonate, respectively. The gypsum catalyst is easily
recoverable and reusable for subsequent cycles of reaction. Similar physico-chemical properties
of fresh and used catalyst were confirmed through XRD, FTIR and Hammett test analysis.
Besides, the mechanistic pathway of glycerol carbonate was confirmed through the formation of
glycerol carbamate as intermediate compound which was further established through time online
1
3
analysis study using C NMR and ATR-FTIR, respectively. The study also clearly supports
conversion of waste into wealth while promising proper disposal of waste to produce value
added product.
Keywords: Glycerol; Glycerol Carbonate; Gypsum; Heterogeneous catalyst; Pre-treatment
1
. Introduction
The development of new routes of chemical products from renewable feedstock is one of the
critical importance for a sustainable future. Nowadays, the expanding production of biodiesel has
generated glycerol in quantities exceeding the current demand. This leads a drop of the glycerol
price and the risk of seeing a surplus of glycerol. By the year 2016, the world biodiesel market is
estimated to be at 308 billion kg, which means more than 33 billion kg of crude glycerol will be
produced every year [1]. The production cost of biodiesel increases by U.S. $0.0022/L for every
U.S. $0.022/kg reduction in glycerol selling price [2]. The potential sale of this product could
make biodiesel cheaper. Thus, it is important to find alternative uses for glycerol. Recently, the
new opportunities for conversion of glycerol waste into high value-added product have increased
due to the special structure, properties and the renewable feature of glycerol [3]. The glycerol
waste can make a comfortable place in the global market by using it as a source of feedstock for
value-added product such as glycerol carbonate [4].
Glycerol carbonate is one of the important building blocks for chemical industry and
characterized by a low toxicity, good biodegradability and high boiling point, which are useful
for solvents and chemical intermediate [5]. Due to the combination of wide reactivity and bio
based origin, carbonates become a versatile and renewable building block for organic chemistry

as well as possibility to become a major chemical intermediate, for instance, in the
manufacturing of polymer, adhesives, foams, coatings and lubricants [6-8].
Glycerol carbonate is produced from glycerol in many different routes. Previously, it was
synthesized in conventional method via phosgenation of methanol [9]. However, due to high
toxicity, chemical and corrosive nature of phosgene, the alternative routes were suggested such
as from ethylene carbonate and propylene [10]. Glycerol carbonate also synthesized from
glycerol with dimethyl carbonate. This route has its disadvantages because the reaction required
expensive catalyst and higher ratio of dimethyl carbonate to glycerol as well as a shift in
chemical equilibrium [11]. Alternatively, an equally attractive route to prepare glycerol
carbonate is the reaction of glycerol with urea. This phosgene-free process utilised low cost raw
material with low toxicity [12]. Besides, there is a possibility to convert ammonia formed during
reaction back into urea using carbon dioxide captured technology [9].
Gypsum is the most common naturally occurring sulphate minerals with the chemical formula
of CaSO
4
·2H
2
O. Gypsum is also produced as a waste product from advanced material
) processing plant [13]. Titanium (IV)
manufacturing industry, such as titanium dioxide (TiO
2
oxide is extracted from ilmenite ores by sulphuric acid digestion and the spent acid is neutralized
by calcium carbonate thus produces the waste gypsum [14]. Hunstman Tioxide is one of the
largest companies that produce TiO in the world. The capacity of their plant in Malaysia alone
2
is about 56 000 metric tons per year. This industry is estimated to produce ~ 400 000 tonnes of
waste gypsum annually [15-16]. Therefore, for the first time this present research utilizes
commercial gypsum material as a model catalyst for the synthesis of glycerol carbonate from
glycerol and urea.
2
. Experimental
2
.1. Chemicals
Glycerol (99.5%) and urea (AR grade) were purchased from Friendemann Schmidt
Chemical. Standard glycerol carbonate, hydrogen peroxide (30% w/w H
tetraethylene glycol (TEG) were purchased from Sigma-Aldrich while commercial gypsum
CaSO ·H O) and calcium oxide (CaO) were purchased from R&M chemical. All other
2 2
O in water) and
(
4
2
chemicals except gypsum and calcium oxide were used directly without further purification and
pre-treatment.
2
.2. Catalyst preparation
The commercial gypsum (denoted as GypF) is calcined in static air using a vertical furnace
for 3h at 150°C and 800°C respectively. The catalysts obtained were denoted as Gyp150 and
Gyp800. The gypsum treated with H was prepared by treating the fresh commercial gypsum
material with 30% H in H O. In specific, 11.5g of GypF was added into 20ml of H . Then
the solution was stirred for 3h with the heating at 40°C. The solution was then filtered and
2
O
2
2
O
2
2
2 2
O

o
washed with hot water for 3 times. Next, the sample was dried in the oven at 110 C for 16h.
Then, the catalyst obtained was denoted as GypH For calcium oxide, the fresh commercial
.
sample was calcined in the furnace at 800°C for 3h and the catalyst obtained was denoted as
CaO800.
2
.3. Catalyst characterization
Thermo-gravimetric Analyser Mettler Toledo TGA/DSC-HT/1600 was used to monitor the
o
o
weight loss and phase transformation in the range of temperature between 25 C to 1000 C under
static air environment.
X-ray diffraction (XRD) patterns of the catalysts were recorded on a Rigaku Miniflex II
diffractometer and were set to be analysed at the range of 15 degree to 80 degree at 2 Theta (θ)
for crystalline phase determination with Cu Kα radiation.
The BET (Brunauer-Emmett-Teller) method was used to measure the total specific areas of
catalysts using Micromeritics TriStar II 3020.
Scanning Electron Microscopy (SEM) brand Zeiss Evo 50 was used to study the surface
morphology of the catalyst.
The acido-basic property of catalyst was determined by three different methods (Hammett
test, CO -TPD, and NH -TPD). Firstly, by utilizing Hammett test, the indicator 2,4-
2 3
Dinitroaniline, Phenolphthalein, Methyl red and Methyl orange were used to determine the
qualitative acidic as well as basic properties of the catalyst. In specific, 25mg of catalyst was
weighed and prepared in four batches and 5ml of methanol was added to the catalyst. Then, 1ml
of the indicator was added to 4ml of methanol. Final volume of 5ml indicators were added
separately to the catalysts weighed in batches. The mixture was then left to equilibrate for 2h.
The colour changes were observed and noted.
The quantitative basicity strength was analysed using Temperature Programmed Desorption
of CO
conductivity detector (TCD). 50mg of the catalyst sample was treated at 200°C for 1h by passing
helium at a flow rate of 20ml/min. Then, the sample was saturated with CO at a flow rate of
0ml/min and at a temperature ramping from 10°C/min up to 200°C for 1h. After that, the
2 2
(CO -TPD), (Thermo Finnigan TPD/R/O 1100). The TPD was equipped with a thermal
2
3
system was flushed with helium (20ml/min) at 200°C for 1h to eliminate physisorbed CO
Desorption was carried out from ambient temperature of 900°C at a heating ramp of 5°C/min.
2
.
3 3
Temperature Programmed Desorption of NH (NH -TPD) was used to determine the
quantitative acidity strength. Approximately, 50mg of sample was placed in a quartz tube and
treated for 1h at 200°C under a nitrogen flow of 20ml/min. Then, the sample was dosed with
o
NH
3
o
, again at 20ml/min. After dosing, the samples were cooled down to 50 C and then heated to
o
9
00 C at 10 C/min, under helium flow of 20 ml/min.
2
.4. Catalytic Testing

The reaction was carried out using a three neck round bottom flask. Firstly, 13.8g of glycerol
was added into the 100ml three necks round bottom flask as the solution was heated until 150°C
and stirred at 340 rpm under the flow of nitrogen gas. The temperature and stirring rate were
controlled by using a temperature controller (IKA@ ETS-D5). When the temperature of the
glycerol is reaching 150°C, 13.5g of urea and 0.25g of catalyst were added into the reaction
solution. The sample of liquid product produced was sampled at specific sampling time then
subjected for analysis.
2
.5. Catalyst Reusability Procedure
For catalyst reusability test, the tested catalyst was removed after completing the catalytic
reaction by using centrifugal, washed with excess of water then drying it in oven for 16h at
10°C. The catalyst obtained denoted as GypR. It then undergoes another catalytic reaction
1
directly without any further treatment.
In addition to that, the reaction solution after the reaction process was also subjected to ICP-
MS and CHNOS analysis, which in order to analyse any possibility of metal components leach-
out into the reaction medium. The amounts of metal leach-out can be calculated and compared to
the actual amount of metal available in catalyst prior to the reaction. For ICP-MS analysis
procedure, 0.5ml of sample diluted to 50ml of 2% HNO
standards of 2% HNO were prepared before the analysis, which are (0, 0.5. 1.0, 2.0, 5.0 and
0.0 ppm) respectively. Then, the analysis was completed using Agilent 7700x ICP-MS.
3
(Fischer Chemical). Five different
3
5
CHNOS analysis was carried out by combustion method using Vario Macro Elemental Analyser
CHNSO.
2
.6. Analysis of product
Gas Chromatography GC-FID Agilent Technologies 7890A equipped with a Varian
Capillary Column, CP-PoraBOND Q (25m, 0.53mm, 10µm) was used to analyse the product of
reaction. 50µl of sample was taken at each time and diluted into 1450µl of deionised (DI) water.
Helium gas was used as the carrier gas with a flow rate of 35ml/min. The temperature of the
o
o
injector and the detector were 225 C and 250 C, respectively. The temperature of the column
o
o
o
o
was programmed to be at 2min with initial hold at 80 C, a 10 C/min ramp from 80 C to 250 C
o
o
o
and 15 C/min ramp from 250 C to 300 C with 3min holding time. The split ratio was 1:10 and
injection volume was 1µl. Tetraethylene glycol (TEG) was used as an external standard.
Standard glycerol and glycerol carbonate with different range of concentration were used for
calibration purpose.
Time online analysis of products was also analysed using Nuclear Magnetic Resonance
(NMR) BRUKER Ultra Shield Plus 500MHz and Attenuated Total Reflectance-Fourier
Transform Infrared spectroscopy (ATR-FTIR) Perkin Elmer, USA. For the case of analysis of
liquid product using ATR-FTIR, 50µl of sample was immediately sampled and one drop was

placed onto the sample holder. The sample was then immediately analysed. In case of NMR, the
sample for 13C NMR analysis was prepared by sampling 150µl of liquid sample into the NMR
tubes and then diluting with 500µl of deuterated methanol-d .
4
3
3
3
. Results and discussion
.1. Characterization of catalyst
.1.1. Thermo-gravimetric analysis (TGA)
In order to determine the calcination temperature as well as the effect of temperature on
GypF (CaSO ·2H O), the GypF was subjected to thermogravimetric analysis. Fig. 1 illustrates
4
2
the TGA curve for GypF catalyst. Single stage decomposition was clearly observed from the
TGA thermogram. The decomposition of the precursor started at 80°C and completed at 200°C.
The total weight loss was about 0.74mg from 3.63mg of sample placed at the beginning of the
test equal to 20.5% of weight loss. The magnitude of weight loss (about 20% of the original
sample weight) complies with stoichiometrically expected for complete dehydration of the
gypsum to the anhydrite (CaSO ) form [17]. The weight losses are more easily determined by
4
integrating the peaks at 130°C in the first derivative (DTG) curve, which also indicates
apparently only one weight loss.
Fig. 1. Thermo-gravimetric analysis (TGA) of GypF (a) weight percentage (%) loss and (b) derivative
thermal gravimetric (DTG)

3
.1.2. X-ray diffraction (XRD)
Fig. 2 illustrates the X-ray diffractogram of GypF, Gyp150, Gyp800 and GypH labelled as a,
b, c and d, respectively. XRD pattern of GypF showed characteristics of CaSO
at 2θ = 20.90º, 23.58º, 29.31º and 48.02º (JCPDS File: 33-0311). Gyp150 showed the
characteristic of ß-CaSO diffraction peaks presence at 2θ = 25.88°, 29.94° and 32.02°. Lopez-
Delgado and co-workers in their study stated that CaSO ·2H O will transform into a different
phase of anhydrite (CaSO ) after subjected to different temperature treatment [18]. ß-CaSO will
be produced in the temperature range between 163 to 360 C, while at higher temperature up to
4 2
·2H O presence
4
4
2
4
4
o
o
3
70 C, the beta (ß) phase will be transformed into gamma (γ) phase. Their observation is in-line
with the XRD pattern obtained for Gyp800 catalyst where the diffraction pattern showed
characteristic of γ-CaSO . The gamma phase of anhydrite was reported to be insoluble compared
to beta phase [19]. In addition to heat treatment, chemical treatment of fresh gypsum with
hydrogen peroxide (H 30% in water) GypH showed the characteristic of CaSO ·0.5H O [20].
4
2
O
2
4
2
Fig. 2. X-ray Diffraction pattern of a) GypF b) GypH c) Gyp150 d) Gyp800
The crystallographic data were summarised in Table S1 using Debye-Scherrer Equation [21]
FWHM cos θ (1)

D is the average crystallite size in mm, λ is the wavelength of the X-ray radiation (Cu Kα, 0.1548
nm), FWHM is the full width at half-maximum, and θ is the diffraction angle.
The crystallite sizes were such calculated using Debye–Scherrer Equation shows that GypF
has smaller particles compared to the treated samples. The difference in crystallite size between
treated and untreated commercial gypsum is expected since heat treatment will induce the
agglomeration of particle, which subsequently increase the crystallite size. The variation of
crystallite size is also in-line with the surface area data analysed using BET method whereby
smaller crystallite size material showing bigger specific surface area.
3
.1.3. BET Surface Area (SBET
)
2
The specific surface area for GypF (10.8m /g) is five times higher compared to the treated
samples. As previously mentioned, the heat treatment was responsible in inducing the
agglomeration of the sample. The effect is more pronounce at higher temperature and causing
sample to bind closely to each other thus decreasing the surface area. Indran and co-workers also
reported this phenomenon where higher temperature will decrease the surface area of catalyst
o
[
22]. It was clearly true since gypsum calcined under static air at 800 C for similar calcination
2
duration shows almost half BET surface area value (i.e. 1.4 m /g) than the one observed with
o
2
gypsum calcined at 150 C (i.e. 2.6 m /g).
3
.1.4. Scanning Electron Microscope (SEM)
Surface morphology using SEM analysis in Fig. 3 showed that the different pre-treatment did
not significantly alter the morphology of the gypsum based catalyst. Gyp150 and Gyp800
showed a slight difference in morphology from GypF due to the deformation of water during
calcinations thus altered the composition from CaSO
4
·2H
2
O
to CaSO . GypH showed the
4
morphology of gypsum hemihydrate (CaSO
similar material reported in literature [23].
4 2
·0.5H O) that is similar to the SEM morphology of
(a)
(b)

(c)
(d)
Fig. 3. SEM of a) GypF b) GypH c) Gyp150 d) Gyp800
3
3
.1.5 Acidity and Basicity Analysis
.1.5.1. Hammett Test
Hammett test has found a wide application in the examination of acid-base properties of
heterogeneous catalysts. Visual or spectrophotometric observation of the colour change of an
absorbed indicator makes it possible for the assessment of the basic or acid strength of the
surface of catalyst [24]. GypF, GypH, Gyp150 and Gyp800 showed negative results for both
basic (2,4-dinitroaniline, H_=15 and phenolphthalein, H_=8.2) and acidic (methyl red, H_=5.1
and methyl orange, H_=3.7) indicator. Thus, it is suggested that gypsum are neutral in nature due
2
-
to the counter effect of both strong acid from (SO
4
) and strong base from Ca. It is important to
note that gypsum was known to be neutral material as reported by Azimi et al [25]
3
2 3
.1.5.2. Temperature Programmed Desorption (CO -TPD and NH -TPD)
It can be seen from CO
characteristic of basic material where there is no peak corresponds to CO
observed for the range of tested temperature. However, with NH as a chemical probe instead of
CO , one desorption peak was observed at the maximum desorption peak at 211 C accounted to
amount desorbed NH of 119.7μmol/g (Fig. S2). Nevertheless, the value was relatively low
which suggests a weak acidic site. The NH -TPD data obtained for anhydrite gypsum (CaSO ) is
in agreement with similar analysis reported by Hu et al [26]. In their study, NH -TPD of sample
containing CaSO :CaO with the ratio of 9:1 produced significantly weak NH desorption signal
compare to the lower ratio of CaSO (1:9) presence in the sample. Both TPD results clearly
confirmed the Hammett test analysis, which indicated Gyp800 is relatively neutral in nature.
2
-TPD profile of Gyp800 (Fig. S1) that the catalyst does not have the
2
desorption as being
3
o
2
3
3
4
3
4
3
4
3
.2. Catalytic activity

3
.2.1. The effect of catalyst pre-treatment on gypsum towards glycerol conversion, glycerol
carbonate selectivity and yield.
Table 1 shows the catalytic activity of gypsum based catalysts towards glycerol conversion,
glycerol carbonate selectivity and yield respectively. The yield of glycerol carbonate obtained
through catalysed reactions is significantly higher compared to uncatalysed reaction. The
reaction without catalyst only produced 25.8% yield of glycerol carbonate. It is therefore
apparent that the important role of the catalyst is to direct the selectivity towards the required
product, glycerol carbonate. It was reported in previous literatures that the first step of the
reaction involving conversion of glycerol into glycerol carbamate which proceeds at higher rate
via uncatalysed homogeneous reaction, whereas the catalyst played a crucial role to selectively
converted carbamate intermediate into glycerol carbonate [9,27]. The statement was further
supported by proposed mechanistic pathways carried out in this study. The highest catalytic
activity obtained from Gyp150 was 92.8% conversion of glycerol, 90.1% selectivity and 83.6%
yield of glycerol carbonate respectively. The turnover frequency (TOF) for Gyp150 also showed
the highest value compared to other gypsum samples. It is worth to note that the TOF value
calculated for Gyp150 is also higher and comparable to the TOF value reported in recently
published works [28-29]. It is important to note that different phase of CaSO
significant role in controlling the catalytic activity and selectivity since only 73.8% of glycerol
carbonate yield observed with γ-CaSO (Gyp800) as a catalyst. However, it is essential to point
out here that γ-CaSO is insoluble anhydrite material compared to ß-CaSO identified in Gyp150.
Thus, it is suggested that higher yield of glycerol carbonate observed with ß-CaSO might have
4
also plays a
4
4
4
4
been contributed by some of the metal component leach-out during reaction and involved as
homogeneous reaction instead of fully surface reaction. This assumption was supported by the
analysis of solution reaction after reaction for both Gyp150 and Gyp800 using ICP-MS and
CHNOS analyser, respectively. The ICP-MS results show that 0.54% of Ca is detected in the
liquid solution for the reaction in the presence of Gyp150 compared to 0.21% quantified for
Gyp800. Besides that, analysis of possible sulphur leach-out during the catalytic reaction was
determined using CHNOS analyser. The data analysis showed 0.47% of sulphur leach-out for
Gyp150, which is almost 3 times higher compared to the similar analysis for Gyp800.
Table 1
o
Reaction conditions: Temperature, 150 C; Gas, N ; Glycerol: Urea, 1: 1.5 (Molar ratio); Catalyst mass 0.25g; Time:
2
4
h; Standard stirring rate, 340 rpm, TOF: calculated based on the glycerol carbonate yield per gram catalyst per total
reaction time (h). RSD: <5%. Note: Gly. is Glycerol; GC : Glycerol carbonate; (3) is 2,3-dihydroxypropyl
carbamate; (5) is 4-(hydroxymethyl)oxazolin-2-one; (6) is (2-oxo-1,3-dioxolan-4-yl) methyl carbamate.
Compared to calcined samples, the uncalcined gypsum (GypF) shows inferior catalytic
performance whereby the observed glycerol carbonate yield is only at 59.9%. The inferior
catalytic activity of this particular catalyst may partially attribute to the presence of water

molecule inside the catalyst lattice that consequently affects the catalytic process. This argument
further proved by the data obtained from H treated catalyst (GypH) where it showed the
2
O
2
lowest yield and selectivity of glycerol carbonate compared to others. The effect of water content
in glycerol was also studied by Lu and co-workers [30]. The yield of glycerol carbonate will
rapidly decrease in the presence of water more than 2.0wt%.
Since acido-basic properties were shown to have significant effect on overall catalytic
activity for glycerol carbonate via carbonylation of glycerol with urea [9], detailed
characterizations have been carried out. Though, the qualitative and quantitative analysis
performed using TPD technique with NH
3
and CO
2
as chemical probe, respectively and
phase
4
) where these ions present at the catalyst surface could act
Hammett test showed that gypsum based catalyst is neutral in nature. However, γ-CaSO
4
2
+
2-
of Gyp800 consists of Ca and (SO
as Lewis acid and conjugate base respectively. It is reported by Climent et al. that Lewis acid
sites activates the carbonyl group of urea while the conjugated base site activates hydroxyl group
on glycerol to form glycerol carbonate [9]. On top of that, it has been reported that neutral
catalyst could catalyse glycerol carbonylation to yield glycerol carbonate [31]. Thus, it is
believed that gypsum offers ions at its surface that could particularly act as the active site for the
synthesis of glycerol carbonate using glycerol and urea.
In order to identify whether Ca plays a major role in superior activity of gypsum based
catalyst, CaO has been subjected to similar reaction conditions. For the record, CaO as catalyst
for glycerol carbonate formation via carbonylation of glycerol using urea was not reported in any
previous literature. Though, CaO has been used by Simantujak and co-workers for converting
glycerol into glycerol carbonate through transesterification of dimethyl carbonate with glycerol
and the reaction produced 90.2% yield of glycerol carbonate [32]. However, glycerol carbonate
synthesis via carbonylation approach involving urea and CaO as catalyst clearly produced lower
yield of glycerol carbonate (51.0%) compared to 73.8% attained with γ-CaSO
important to note here that the level of glycerol conversion between CaO and γ-CaSO
similar which is 89.5% for CaO and 89.1% for γ-CaSO , respectively. Conversely, the selectivity
of glycerol carbonate which is believed to originate from selective transformation of glycerol
carbamate is observed to be higher with the presence of γ-CaSO (82.8%) compared to CaO
57.0%). The observation is clearly in-line with the role of Lewis acid and conjugate based as
4
catalyst. It is
4
is almost
4
4
(
2
-
4
mentioned previously. Thus, it is postulated that (SO ) presence in anhydrite phase of gypsum
2
-
may act as more efficient conjugate base compare to O available with CaO catalyst.
Theoretically, for the catalytic reaction involving heterogeneous catalyst, the specific surface
area and particle size of catalyst plays a crucial role in assisting the catalytic activity and
selectivity [33]. However, both factors are clearly not contributing for this study as GypF with
higher specific surface area and smaller crystallite size produced lower yield of glycerol
carbonate.
3
.2.2. Time Online Analysis (TOL)

4
Since the γ-CaSO is an insoluble phase of anhydrite, Gyp800 was then used as catalyst for
time-online study. Fig. 4 illustrates the time online analysis (TOL) for 10h conducted in the
presence of Gyp800 catalyst. The conversion increased up to 80% for the first 2h with
maximization to 96% at 6h whereas the selectivity of glycerol carbonate gradually decreased
from 6h to 10h. The decreases in glycerol carbonate selectivity at prolong reaction tie is
attributed to the transformation of this target product into compound 6 and at the same time
formation of compound 5, which is originated from carbamate intermediate had also increased. It
also clear from the figure that glycerol carbamate (compound 3) was formed and quickly
decomposes into glycerol carbonate.
Fig. 4. Time online profile of glycerol carbonate synthesis in the presence of Gyp800. Reaction
conditions: glycerol/urea molar ratio: 1 : 1.5, temperature: 150°C, catalyst mass: 0.25g. Key: x glycerol
conversion (%),
ꢀ
selectivity (%) to glycerol carbonate,
selectivity (%) to 4-(hydroxymethyl) oxazolidin-2-one (5), ♦ selectivity (%) to (2-oxo-
,3-dioxolan-4-yl) methyl carbamate (6)
●
selectivity (%) to 2,3-dihydroxyproyl
carbamate (3),
p
1
3
.2.3. Mechanistic study
1
3
13
Fig. 5a showed the C NMR analysis for the time online analysis (TOL) of Gyp800. The C
NMR patterns prove that glycerol carbonate (4) was selectively formed through the
carbonylation of intermediate 2.3-dihydroxypropyl carbamate (3). At 6h of reaction time, 4-
(hydroxymethyl) oxazolin-2-one (5) and (2-oxo-1,3-dioxolan-4-yl) methyl carbamate (6) were
present. It is evident that selectivity towards glycerol carbonate formation decreased from 6h
onwards due to the formation of compound (5) and (6). Further reaction of glycerol carbonate
with excess of urea will form (6) and the intermediate 2,3-dihydroxyl carbamate (3) forms
compound (5) [34]. Besides that, Fig. 5b indicates the chemical shift at 165 ppm which

corresponds to urea (compound 2) is steadily reducing with prolonged reaction time. For the
record, Hasbi and co-workers in 2012 have reported the same mechanistic pathway with this
current study as well with Kim and co-workers in 2013 [34,35]. However, in this study the
glycerol carbamate was selectively decomposed in an accelerated way to glycerol carbonate and
this observation is similar to recently reported study by Indran and co-researchers [22].
13
Fig. 5. a) C NMR overlay of standard glycerol (STD GLY), standard glycerol carbonate (STD GC) and
TOL of Gyp800. Key: (1) glycerol; (2) urea; (3) 2,3-dihydroxypropyl carbamate; (4) glycerol carbonate;
13
(
5) 4-(hydroxymethyl)oxazolin-2- one. b) C NMR overlay of standard glycerol (STD GLY), standard
glycerol carbonate (STD GC) and TOL of Gyp800. Key: (2) urea and (4) glycerol carbonate.
1
3
An analysis using ATR-FTIR was carried out to validate the mechanistic study using C
NMR. Fig. 6 showed the ATR-FTIR spectra obtained from the reaction sampling at 0.5h to 10h.
-1
IR peak unique to glycerol was identified at 1450cm whereas the IR peaks for urea assigned at
-
1
-1
1
(
620cm and 1665cm were clearly observed at 0.5h. However, IR peak of glycerol carbamate
3) was not clearly observed during 0.5h due to the reaction occurred in an accelerated way to
directly decompose glycerol carbamate into glycerol carbonate (4) and this phenomena agreed
1
3
well with the analysis of C NMR. Glycerol carbamate (3) was observed at 2h and 4h with peak
at (1715cm ). The peak of glycerol carbamate was gradually shifted to the 4-(hydroxymethyl)
oxazolidin-2-one (5) at a longer reaction time. On the other hand, peak at 1790cm indicated
C=O stretching from glycerol carbonate. This indicated formation of glycerol carbonate from 2h
onwards. Isocyanic acid compound with N=C=O stretching that usually exist in range peak
-1
-1

-1
2
250cm that was previously reported by Aresta and co-workers was not observed in this
reaction [12]. Thus, it was confirmed that the formation of glycerol carbonate does not follow the
isocyanic acid pathway but follows the pathway of selective transformation of glycerol
carbamate.
Fig. 6. ATR-FTIR spectra for time online analysis of Gyp800 reaction from 0.5h to 10h. Key: (1)
glycerol; (2) urea; (3) 2,3-dihydroxypropyl carbamate; (4) glycerol carbonate; (5) 4-(hydroxymethyl)
oxazolin-2- one
3
.2.3. Stability and reusability of catalyst
One of the key factors that must be considered for heterogeneous catalysts operating in liquid
phase systems is the possibility that active components can leach into the reaction mixture,
thereby leading to catalyst deactivation or, to the formation of an active homogeneous catalyst.
Therefore, the reusability of catalyst was investigated by utilizing it in four consecutive batch
runs. Fig. 7 presents the results of glycerol conversion, selectivity and yield of glycerol carbonate
obtained up to fourth consecutive catalytic reaction in the presence of Gyp800. The results
confirm that this catalyst is reusable with no appreciable decrease in its activity and selectivity.
In addition to that and as previously mentioned, the reaction solution was subjected to ICP-MS
and CHNOS analysis in order to determine the possibility of metal leach-out. The insignificant
amount of leaching confirms the stability of the catalyst. The reusability of Gyp800 catalyst is
4
expected since the γ-CaSO structure was reported to be an insoluble phase by Rolnick [19].

Fig. 7. Reusability study of Gyp800 on glycerol conversion, glycerol carbonates selectivity and yield.
Reaction conditions: Temperature, 150˚C; Gas, N ; Glycerol: Urea, 1: 1.5 (Molar ratio); Catalyst mass
.25g; Time: 4h; Standard stirring rate, 340 rpm.
2
0
To further validate the stability of gypsum based catalyst, the used catalyst was subjected to
different characterization techniques (XRD, FTIR and Hammett test) in order to inspect the
changes in its physico-chemical properties. XRD diffractogram (Fig. 8) revealed that similar
diffraction peaks were detected for both fresh and used catalyst. Table S2 shows that the average
crystallite size for the used catalyst was very similar compared to the fresh catalyst, which shows
agglomeration did not occur.
The XRD data was further supported by FTIR analysis of both fresh and used catalyst. It is
clear from FTIR spectra (Fig. 9) that the characteristic peak was almost similarly presented for
-1
the both Gyp800 and GypR catalyst. However, additional IR peak was observed at 2919cm and
-1
peak at 3448cm , broader than Gyp800 which is attributed to the presence of OH- functional
-1
group. Additional peak also observed at 1717cm due to the presence of carbonyl group C=O
functional group. The presence of O-H and C=O can be explained by the excess of glycerol and
products molecule that were not totally removed during washing step.
Besides, Hammett test analysis for used catalyst also shows similar observation of fresh
catalyst counterpart. Therefore confirm the stability and reusability of gypsum (Gyp800) toward
producing glycerol carbonate from glycerol and urea.

Fig. 8. XRD pattern of a) Gyp800 b) GypR
Fig. 9. FTIR Spectra of a) Gyp800 b) GypR
4
. Conclusion

It is clear from the work details in this study that gypsum based catalyst is efficient and
reusable for the synthesis of glycerol carbonate from glycerol and urea. The reaction is
environmental friendly and also introduces a proper and an alternative way for disposal of
gypsum. Simple pre-treatment techniques were shown to successfully produce different physico-
chemical properties of gypsum based catalysts. The ß-CaSO
active phase where it produced 92.8% conversion of glycerol, 90.1% of selectivity and 83.6%
yield of glycerol carbonate, respectively. However, the slightly less productive γ-CaSO phase is
essentially important where it shows true heterogeneous catalyst characteristics with a similar
4
structure was identified as the most
4
2
+
catalytic performance at consecutive cycles of reaction. The presence of Ca as Lewis acid sites
2
-
4
with (SO ) as conjugate base sites in gypsum based catalyst is believed to responsible for the
catalytic activity and selectivity. As a final point, the stability of sulphur inside the catalyst
structure shall add to the viability of the glycerol carbonate to be utilised in different broad
applications without causing extra issues.
Acknowledgements
The authors acknowledge Universiti Malaysia Pahang for UMP Internal grants scheme
(RDU120363) and Graduate Research Scheme (GRS) as well as the Ministry of Education for
Research Acculturation Collaborative Effort (RACE, RDU121301). We also would like to thank
the Ministry of Education Malaysia for MyBrain15.
References
[1]
[2]
[3]
[4]
[5]
X. Fan, R. Burton, Y. Zhou, Open Fuels Energ. Sci. 3 (2010) 17-22
Y. Zheng, X. Chen, Y. Shen ,Chem. Rev. 108 (2008) 5253-5277
Z. Gholami, A. Z. Abdullah, K. Lee, Renew. Sust. Energ. Rev. 39 (2014) 327-341
M. Ayoub, A. Z. Abdullah, Renew. Sust. Energ. Rev. 16 (2012) 2671-2686
A. V. M. Nunes, G. V. S. M. Carrera, V. Najdanovic-Visak, M. Ponte, Fluid Phase
Equilib. 358 (2013) 105-107
[6]
M. O. Sonnati, S. Amigoni, E. P. Te de Givenchy, T. Darmanin, O. Choulet, F. Guittard,
Green Chem. 15 (2013) 283-306.
[
[
7]
8]
T. Mizuno, T. Nakai, M. Mihara, Heteroat. Chem. 21 (2010) 541–545
J. R. Ochoa-Gomez, O. Gomez-Jimenez-Aberasturi, C. Ramirez-Lopez, M. Belsue, Org.
Process Res. Dev. 16 (2012) 389-399
[
[
[
9]
M. J. Climent, A. Corma, P. D. Frutos, S. Iborra, M. Noy, A. Velty, Patricia Concepción,
J. Catal. 269 (2010) 140-149
10] Z. Mouloungui, J.-W. Yoo, C.-A. Gachen, A. Gaset, G. Vermeersch, European Patent,
EP0739888 A1 (1996) 1–13
11] S. C. Kim, Y. H. Kim, H. Lee, D. Y. Yoon, B. K. Song, J. Mol. Catal. B: Enzym. 49
(2007) 75–78.
[
[
12] M. Aresta, A. Dibenedetto, F. Nocito, C. Ferragina, J. Catal. 268 (2009)106-114
13] R. A. Kamarudin, M. S. Zakaria, MJAS 11 (2007) 57-64

[
[
14] I. Fauziah, S. Zauyah, T. Jamal, Sci. Total Environ. 188 (1996) 243-251
15] A. Azdarpour, M. Asadullah, R. Junin, M. Manan, H. Hamidi, E. Mohammadian, Fuel
Process. Technol. 126 (2014) 429-434
[16] M. J. Richards, Ensign 6 (1992) 3-16
[17] Y. Wang, Y. Kim, H. K. Christenson, F. C. Meldrum, Supp. Material for Chem.
Commun., R. Soc. Chem. (2011) 1-8
[18]
A. Lopez-Delgado, S. Lopez-Andres, I. Padilla, M. Alvarez, R. Galindo, A. J. Vazquez,
Geomaterials 4 (2014) 82-91
[
19] L. S. Rolnick, “The stability of gypsum and anhydrite in the geologic environment,” Ph.
D. dissertation, Massachusetts Ins. of Tech., The John Hopkins Univ., US, 1950
20] K. Lanjekar, V. K. Rathod, J. Environ. Chem. Eng. 1 (2013) 1231-1236
21] V. Uvarov, I. Popov. Mater. Charact. 85 (2013) 111-123
[
[
[
22] V. P. Indran, N. A. S. Zuhaimi, M. A. Deraman, G. P. Maniam, M. M. Yusoff, T. Y. Y.
Hin, M. H. A. Rahim, RSC Adv. (2014) 25257–25267
[23] V. Leskericiene, Mater. Sci. 19 (2013) 197-202
[24] J. Kijenski, B. Zielinski, R. Zadrolny, S. Malinowski, J. Res Inst. Catal. Hokkaido Uni.
2
7 (1979) 145-1568
25] G. Azimi, V.G. Papangelakis and J.E. Dutrizac, Fluid Phase Equilibr., 260 (2007) 300–
15.
[
3
[
[
26] X. Hu, S. Ju, L. Wu, M. Zhao and C. Dong, Appl Mech Mater. 457-458 (2014) 88-92.
27] C. Hammond, J.A. Lopez-Sanchez, M. H. A. Rahim, N. Dimitratos, R. L. Jenkins, A. F.
Carley, Q. He, C. J. Keily, D. W. Knight, G. J. Hutchings, Dalton Trans. 40 (2012) 3927-
3
937.
28] T. W. Turney, A. Patti, W. Gates, U. Shaheen, S. Kulasegaram, Green Chem., 15 (2013)
925-1931
29] K. Jagadeeswaraiah, C. R. Kumar, P. S. S. Prasad, S. Loridant, N. Lingaiah, Appl. Catal.,
A 469 (2014) 165– 172
[
[
1
[
[
30] P. Lu, H. Wang, K. Hu, Chem. Eng. J. 228 (2013) 147–154
31] J. Chen, C. Wang, B. Dong, W. Leng, J. Huang, R. Ge, Y. Gao, Chin. J. Catal., 36 (2015)
3
36-343.
[32] F. S. H. Simanjuntak, T. K. Kim, S. D. Lee, B. S. Ahn, H. S. Kim, H. Lee, Appl. Catal.,
A 401 (2011) 220– 225
[33] Y. Zhang, Z. Hong, Y. Ren, D. Wang, Y. Yokogawa, J. Non-Cryst. Solids 402 (2014)
1
3
49-152D. Kim, K. Park, M. Kim, D. Kang, J. Yang, D. Park, Appl. Catal., A 473 (2014)
1-40
[34] M. H. A. Rahim, Q. He, J. A. Lopez-Sanchez, C. Hammond, N. Dimitros, M. Sanker, A.
F. Carley, C. J. Kiely, D. W. Knight, G. J. Hutchings, Catal. Sci. Technol. 2 (2012) 1914–
1
924
[35] D. Kim, K. Park, M. Kim, D. Kang, J. Yang, D. Park, Appl. Catal., A 473 (2014) 31-40

LIST OF TABLE
Table 1
The effect of gypsum pre-treatment on glycerol conversion, glycerol carbonate selectivity and
yield.
Entry Catalyst Temp. heat
treatment
Gly.
Conv.
%
Selectivity %
GC yield
%
TOF
-1
GC
(3)
(5)
(6)
(mmol g
(
°C)
-1
cat. h )
-
1
2
3
4
Blank
CaO800
GypF
-
78.7
89.5
92.5
92.8
32.8
57.0
64.8
90.1
24.0
4.5
9.0
0.6
43.1
26.7
18.1
8.4
-
25.8
51.0
59.9
83.6
800
-
150
7.3
5.4
0.9
76.3
89.7
110.5
Gyp150
5
6
Gyp800
GypH
800
110
89.1
91.0
82.8
43.8
5.1
10.3
29.6
1.7
73.8
39.8
125.4
59.6
14.8
11.8
o
Reaction conditions: Temperature, 150 C; Gas, N ; Glycerol: Urea, 1: 1.5 (Molar ratio); Catalyst mass 0.25g; Time:
2
4
h; Standard stirring rate, 340 rpm, TOF: calculated based on the glycerol carbonate yield per gram catalyst per total
reaction time (h). RSD: <5%. Note: Gly. is Glycerol; GC : Glycerol carbonate; (3) is 2,3-dihydroxypropyl
carbamate; (5) is 4-(hydroxymethyl)oxazolin-2-one; (6) is (2-oxo-1,3-dioxolan-4-yl) methyl carbamate.
The effect of gypsum pre-treatment on glycerol conversion, glycerol carbonate selectivity and
yield.
Entry Catalyst Temp. heat
treatment
Gly.
Conv.
%
Selectivity %
GC yield
%
TOF
-1
(
mmol g
(
°C)
-1
GC
(3)
(5)
(6)
cat. h )

1
2
3
4
Blank
CaO800
GypF
-
78.7
89.5
92.5
92.8
32.8
57.0
64.8
90.1
24.0
4.5
9.0
0.6
43.1
26.7
18.1
8.4
-
25.8
51.0
59.9
83.6
-
800
-
150
7.3
5.4
0.9
76.3
89.7
110.5
Gyp150
5
6
Gyp800
GypH
800
110
89.1
91.0
82.8
43.8
5.1
10.3
29.6
1.7
73.8
39.8
125.4
59.6
14.8
11.8