Applied Thermal Engineering 112 (2017) 871–880

Contents lists available at ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Research Paper

Renewable syngas production from thermal cracking of glycerol over

praseodymium-promoted Ni/Al2O3 catalyst
Mohd Nasir Nor Shahirah a,b, Bamidele V. Ayodele a,b, Jolius Gimbun a,b, Su Shiung Lam c, Chin Kui Cheng a,b,⇑
a
Faculty of Chemical & Natural Resources Engineering, Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia
b
Centre of Excellence for Advanced Research in Fluid Flow (CARIFF), Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan, Pahang, Malaysia
c
Eastern Corridor Renewable Energy Group (ECRE), School of Ocean Engineering, Universiti Malaysia Terengganu, 21030 Kuala Terengganu, Terengganu, Malaysia

h i g h l i g h t s

Glycerol was catalytically pyrolyze into syngas over 3wt%Pr-20wt%Ni/77wt%a-Al2O3.

1

The highest H2 rate and H2 yield were at 0.02593 mol g cat s1 and 29.04% respectively.
1

Langmuir-Hinshelwood (LH) model showed activation energy of 37.36 kJ mol .

a r t i c l e i n f o a b s t r a c t

Article history: In this study, the kinetics of glycerol pyrolysis over a 3wt%Pr-20wt%Ni/77wt%a-Al2O3 catalyst
Received 16 July 2016 was investigated. The catalyst was synthesized via wet-impregnation method and was characterized
Revised 22 September 2016 using temperature-programmed calcination (TPC), temperature-programmed reduction (TPR),
Accepted 20 October 2016
N2-physisorption, FESEM imaging, X-ray diffraction and CO2-/NH3-temperature-programmed desorption
Available online 21 October 2016
(TPD). The catalytic activity of the as-synthesized 3 wt% Pr-Ni/a-Al2O3 catalyst was evaluated in a stainless
steel fixed bed reactor at temperatures that ranged from 973 K to 1073 K and a weight-hourly-space-
Keywords:
velocity (WHSV) of 4.5  104 ml g1 h1 under the atmospheric condition. The main gaseous products from
Glycerol
Kinetics
catalytic glycerol pyrolysis were H2, CO, CO2 and CH4 (descending ranking) with the highest H2 formation
Nickel rate and H2 yield of 0.02593 mol g cat1 s1 and 29.04%, respectively. The analysis of the kinetic data
Praseodymium obtained from the glycerol pyrolysis showed activation energy of 37.36 kJ mol1. Based on the mechanistic
Pyrolysis modeling, it can be deduced that the rate determining step of the glycerol pyrolysis was via a single site
Syngas associative adsorption with molecular surface reaction as the rate-determining step.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction tion, it is expedient for an increase in the industrial utilization of

the glycerol [6]. There has been a growing interest in the produc-
Glycerol is a polyol compound commonly produced as a by- tion of synthesis gas (syngas) from glycerol. Syngas, which mainly
product from the oleo-chemical and biodiesel plants [1]. It is consists of a mixture of H2 and CO, is a chemical intermediate for
widely used as a chemical building block in pharmaceutical, cos- the production of various value-added chemicals and oxygenated
metic, and food industries [2]. In recent time, there has been a fuels [7]. Presently, syngas is commercially sourced from the steam
tremendous increase in glycerol production as a by-product from reforming of natural gas [8]. Nevertheless, due to a fast depletion in
the biodiesel synthesis [3] due to growing demand for biodiesel finite fossil fuel such as natural gas, the problem of sustainability
as a liquid transportation fuel [4]. Indeed, it has been projected that might arise in the near future. Therefore, the utilization of glycerol
by the year 2020, about 3 megatons of crude glycerol will be pro- as a feedstock for the production of syngas has been touted as a
duced, which may create a surfeit of glycerol in the market [5]. more sustainable route for syngas production.
Hence, in order to keep pace with the increasing glycerol produc- An extensive review by Lin [9] showed that several methods, viz.
steam reforming, partial oxidation and auto-thermal reforming, as
well as CO2 reforming have been investigated for the valorization
⇑ Corresponding author at: Faculty of Chemical & Natural Resources Engineering,
of glycerol. All these processes showed promising routes for the syn-
Universiti Malaysia Pahang, Lebuhraya Tun Razak, 26300 Gambang, Kuantan,
Pahang, Malaysia.
gas production. However, the major challenge still lies in the tran-
E-mail address: [email protected] (C.K. Cheng). sient formation of carbon, which was directly dependent on the

http://dx.doi.org/10.1016/j.applthermaleng.2016.10.126
1359-4311/Ó 2016 Elsevier Ltd. All rights reserved.
872 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880

concentration of glycerol feed [10–12]. On the other hand, there is 2. Experimental

also a growing interest in the valorization of glycerol via catalytic
pyrolysis process. The catalytic pyrolysis process is a thermal crack- 2.1. Catalyst synthesis
ing process in an oxygen-free environment employing a suitable
catalyst. Similar to the aforementioned processes, catalytic pyroly- The 3wt%Pr-20wt%Ni/77wt%a-Al2O3 catalyst was prepared via
sis is also a thermal-based process. Consequently, it is highly wet-impregnation method following the procedures outlined by
endothermic and often requires considerable amount of heat input. previous researchers [7,11]. The as-received a-alumina support
Unfortunately, this has often resulted in carbon deposition through (Sigma-Aldrich) was firstly calcined in a muffle Carbolite furnace
coking and sintering routes [13,14]. In view of this, several catalytic at 1073 K for 6 h. To bring the temperature of furnace from room
systems have been employed for the glycerol pyrolysis to syngas. temperature to the calcination temperature of 1073 K, a heating
Notably, noble metals showed high activity for syngas production rate of 10 K min1 was employed. Subsequently, the a-alumina
via catalytic pyrolysis [15]. However, these noble metals are expen- (15.4 g) was mixed with 45.0 ml of aqueous solution containing
sive and therefore not economically-feasible for industrial applica- 19.81 g of Ni/(NO3)26H2O (99.99% trace metal basis, Sigma-
tion. Preferably, a non-noble metal such as Ni has been reported to Aldrich) and 1.85 g of Pr(NO3)36H2O (99.99% trace metal basis,
exhibit comparable activity with the noble metals for glycerol Sigma-Aldrich) to produce a final formulation of 3wt%Pr-20wt%
pyrolysis [16]. In addition, it is readily available and inexpensive. Ni/77wt%a-Al2O3. The resulting mixture was then magnetically-
Nonetheless, it is more sensitive to carbon deposition which often stirred for 3 h under ambient condition, followed by oven-drying
leads to catalyst deactivation [12]. For mitigation, an introduction at 393 K for 24 h. The dried solid was then air-calcined at 1073 K
of alien metals such as alkaline earth oxides and elements from for 5 h employing a heating rate of 10 K min1. It was finally
the lanthanide series could assist. Furthermore, employing a suit- crushed and sieved to 140–250 lm particle range for reaction
able support such as a-Al2O3 can also help in enhancing the stability studies.
of the catalyst [11].
Indeed, most of the previous works were on hydrogen produc- 2.2. Fresh catalyst characterization
tion from glycerol steam reforming [17–19], which represent dif-
ferent catalytic behavior and effects compared to the current The temperature programmed calcination (TPC) of the fresh cat-
work. In particular, the steam reforming process required H2O as alyst was performed on a TGA Q-500 system. The TPC was carried
the feedstock and released greenhouse gas, CO2 as a by-product out from room temperature to 1173 K employing heating rates of
[20]. In the catalytic partial oxidation, high conversions were 10, 15 and 20 K min1, respectively, under flow of 60 ml min1 of
achieved and equilibrium compositions were obtained using dif- high purity air. Subsequently, kinetics pertaining to the gas-solid
ferent metallic catalysts such as Ni, Pt and Rh based catalysts, interaction was evaluated using the Kissinger-Akahira-Sunose
which were the most commonly used catalysts in the said reaction (KAS) model-free expression [33,34] as in Eq. (1):
[14,21]. However, exothermic behavior from partial oxidation pro-

cess has caused local hotspot in the catalyst bed and led to catalyst b AR Ea
ln ¼ ln  ð1Þ
deactivation [22]. In addition, previous studies also analyzed the T 2a Ea RT a
behavior of the auto-thermal reforming of glycerol in the presence
of Pt and Ni based catalysts, respectively, with conversion closed to where b is the heating rate (K min1), A represents the pre-
equilibrium [23,24]. However, auto-thermal reforming produced exponential factor (s1), Ea denotes for apparent activation energy
lower H2 yield, thermodynamically, compared to steam reforming (kJ mol1), T a is the apparent temperature (K) and R is the ideal
[25,26]. The hydrogen production from glycerol dry reforming gas constant (8.314 J mol1 K1).
were also reported from previous works over different catalyst The textural properties (i.e. BET specific surface area and pore-
types [27–29]. In contrast to the steam reforming system, the main size distribution) of the freshly calcined catalyst were obtained
advantage of dry reforming is this process is more environmental- from the multipoint Brunauer-Emmett-Teller (BET) method and
friendly as it consumed CO2 while released H2O as the by-product Barett-Joyner-Halenda (BJH) method, respectively [35,36]. The
[30]. However, dry reforming requires gasifying agent such as CO2 raw isotherm data were sourced from N2 physisorption carried
which add on to the feedstock costing. out at 77 K in a Thermo Scientific Surfer Gas Adsorption Porosime-
In contrast, there are not many published works concerning the ter unit. Prior to the analysis, the fresh catalyst was pre-treated
catalytic thermal cracking of glycerol. A previous study [31] overnight under vacuum at 573 K. The density of the catalyst
showed that pyrolysis at 1073 K over an activated carbon produced was measured using Gas Pycnometer 1340 AccuPyc. The surface
H2 yield ranged 22.6–30.1%. Another independent research [32] structure and elemental composition of the solid samples were
has found that glycerol pyrolysis at 973 K produced lower amount analyzed using the FESEM (JOEL/JSM-7800F) at 5–15 kV of voltage
of H2 gas which was circa 22.0%. In lieu of the knowledge gap sta- acceleration, with up to 30 kX magnifications.
ted, the current work focuses on the synthesis and characterization The XRD measurement was carried out on a Shimadzu XRD-
of praseodymium (Pr) promoted Ni/a-Al2O3 catalyst for glycerol 6000 model employing X-ray radiation generated from a nickel-
pyrolysis to syngas. The 3 wt% Pr-Ni/a-Al2O3 catalyst was synthe- filtered Cu Ka. Wavelength (k) of 1.5418 Å at 40 mA and 45 kV with
sized using wet impregnation method and subsequently character- scanning that ranged from 10° to 80° were employed. Subse-
ized for its physicochemical properties. The catalytic activity of quently, the crystallite size was estimated using the Scherrer equa-
glycerol pyrolysis was then investigated at temperatures that ran- tion as in Eq. (2):
ged 973–1073 K and also at different partial pressures of glycerol. kSch k
To surmise, the novelties of this work that have not been reported Dp ¼ ð2Þ
b cos h
elsewhere were: (1) Application of thermally-induced glycerol
cracking/pyrolysis that was assisted by 3 wt% Pr-Ni/a-Al2O3 cata- where Dp is the mean crystallite size (Å), kSch is the Scherrer con-
lyst, (2) estimations of kinetics parameters, as well as elucidation stant (0.90), k is the wavelength of X-ray, b is the angular width
of mechanisms that can represent the thermally-driven glycerol for half-maximum intensity (degree, °) and h is half of the Bragg’s
cracking and (3) glycerol conversion, syngas product as a function angle (degree, °).
of reacting conditions, viz. temperature, glycerol partial pressure In order to determine the reducibility and acidity/basicity prop-
and 3 wt% Pr-Ni/a-Al2O3 catalyst. erty of the catalyst, both temperature programmed reduction (TPR)
 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880 873

and temperature programmed desorption (TPD) analyses were can be assigned to the metal oxides formation resulting from the
performed in a Thermo Finnigan TPDRO 1100 instrument that decomposition of nitrate species (cf. Eqs. (6) and (7)) [38,39].
was equipped with a thermal conductivity detector (TCD). The
TPR analysis was performed from room temperature to 1173 K at PrðNO3 Þ3  6H2 O ! PrðNO3 Þ3 þ 6H2 O ð4Þ
a 10 K min1 ramping rate and employing 20 ml min1 of 5% NiðNO3 Þ3  6H2 O ! NiðNO3 Þ3 þ 6H2 O ð5Þ
H2/95% N2 gas mixture with a holding time of 1 h. The acidity of
2PrðNO3 Þ3 ! Pr2 O3 þ 6NO2 þ 3=2O2 ð6Þ
the catalyst was examined via TPD using NH3 as a probe molecule.
Approximately 0.5 g of the catalyst was placed in a reactor, then 2NiðNO3 Þ3 ! 2NiO þ 6NO2 þ 2O2 ð7Þ
treated in flow of 30 ml min1 of N2 for 30 min, followed by satu- It is to be reiterated that from this thermal profile, the as-
ration with NH3 in He (30 ml min1) for 1 h, up to 393 K (ramping synthesized catalyst was thermally stable at calcination tempera-
rate of 40 K min1) for the adsorption of NH3 onto the catalyst. The ture greater than 600 K. Hence, the calcination temperature for
reactor was then purged with He (30 ml min1) at room tempera- the fresh Pr-Ni/a-Al2O3 catalyst was set at 1073 K to ensure that
ture for 30 min to remove the excess NH3 in the gas phase. The NH3 the nitrate precursor was fully decomposed whilst at the same
desorption was then analyzed from room temperature to 1173 K time, to pre-condition the catalyst for the ensuing glycerol pyroly-
under He flow (20 K min1, 30 ml min1), and the desorbed NH3 sis study. Furthermore, there is no significant variation in the ther-
was detected by the TCD. For CO2-TPD, approximately 0.5 g of mal characteristic of the uncalcined catalyst with changes in the
the catalyst was pretreated with N2 gas at 30 ml min1. Then, ramping rates. Notably, higher ramping rate has shifted the peak
CO2 was adsorbed at 30 ml min1 from 323 to 423 K, followed by center to the higher temperature region. This is because when
holding at 423 K for 60 min. The desorption was subsequently per- the ramping rate was increased, the nitrate decomposition would
formed in 30 ml min1 of He with ramping up to 1173 K. The heat require more time to achieve the same extent of decomposition.
of desorption was determined from Eq. (3): Consequently, this has resulted in a shift to higher temperature
b ðDHdesorption Þ ðDHdesorption ÞAsat (longer reaction time translated into higher temperature). This
ln ¼ þ ln ð3Þ observation is typical of a non-isothermal gas-solid interaction
T 2p RT p RC
between the solid catalyst and air. This trend can be modeled using
where b is the temperature ramping rate (K min1), T p is the peak a ‘‘model-free” non-isothermal kinetics expression. Fig. 1b shows
temperature (K), R is the ideal gas constant, Asat is the quantity the modeling of the non-isothermal kinetics data representing
adsorbed at saturation and C is a constant related to desorption the sharpest peaks (peak IV that represents metal nitrate decom-
rate. position) to the Kissinger-Akahira-Sunose (KAS) model-free
expression (cf. Eq. (1)). Activation energy of 123.7 kJ mol1 was
obtained from the data-fitting with R2-value of 0.99. In the previ-
3. Results and discussion
ous works, activation energy obtained from the KAS model has
been reported. This includes the works by Ayodele et al. [40] with
3.1. Fresh catalyst characterization
Ea in the range of 104.61–230.26 kJ mol1 and Lee et al. [41] with
Ea in the range of 122.73–250.26 kJ mol1. The activation energy
Fig. 1a shows the weight profile (%) and derivative weight pro-
obtained in the current work is within the range of values reported
files obtained from the temperature-programmed calcination
by these authors. This may be ascribed to a similar catalyst prepa-
(TPC) of the uncalcined (dried only) catalyst. The TPC profiles of
ration protocol.
the catalyst consistently demonstrate four derivative weight peaks
Fig. 2 shows the N2 adsorption/desorption isotherms of the cal-
(I–IV) associated with weight loss from the onset of calcination
cined as-synthesized Pr-Ni/a-Al2O3 catalyst. At P/Po less than 0.60,
until 600 K. Thereafter, the thermal profiles stabilized with no sign
the isotherms show a linear trend, symptomatic of mono-layer
of weight reduction. The derivative weight loss represented by
absorption of N2 on the pores of the solid specimen. Beyond the
peaks I–III at 373–513 K can be attributed to the sequential
P/Po of 0.60, isotherms show an upward trend due to the formation
removal of physical water and also hydration water from the Pr
of multiple layers of N2 molecules. The exhibited trend in Fig. 2 is
and Ni-precursors, respectively, as represented in Eqs. (4) and (5)
consistent with a mesoporous solid classified under type IV iso-
[37]. In addition, the peaks coded IV that centered at circa 575 K
therm based on the IUPAC classification. From these isotherms,
the textural properties were obtained and are summarized in
Table 1. The same analysis was also performed on the individual
a-Al2O3 for hierarchical comparison. Although a-Al2O3 is a very
IV

I II
III

Physical and hydrated Metal nitrate Decomposition

water elimination decomposition completed

Fig. 1a. Temperature programmed calcination showing the weight loss and Fig. 1b. Kissinger-Akahira-Sunose (KAS) linear model data for kinetics parameter
derivative weight profiles for Pr-Ni/a-Al2O3 catalyst. determinations of the as-synthesized Pr-Ni/a-Al2O3.
874 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880

stable support for high temperature reactions such as glycerol

pyrolysis, it is often characterized by low BET specific surface area
as also obtained in the this study and in the previous studies
[42,43]. The dispersion of Pr and Ni on the a-Al2O3 has resulted
in BET surface area increment from 2.59 m2 g1 for the pristine
a-Al2O3 to 5.18 m2 g1. This can be attributed to the deposition
of bulkier metals on the support surface. The larger pore diameter
of 18.20 nm that was obtained for the 3 wt% Pr-Ni/a-Al2O3 catalyst
compared to just 3.08 nm for the a-alumina has lend credence to
the supposition, as formation of macropores could be more notable
in the former.
The FESEM’s surface morphology of the a-Al2O3 and the 3 wt%
Pr-Ni/a-Al2O3 catalyst is shown in Fig. 3. It can be observed that
the a-Al2O3 support has irregular shape and possesses smooth sur-
face which explains its small BET specific surface area. Upon metal
co-impregnation, catalyst surface has become coarser and bulkier
due to the crystallite formation. It can be further observed from
Fig. 2. N2 physisorption analysis for the as-synthesized Pr-Ni/a-Al2O3. the image in Fig. 3 that these particles were well-distributed over
the a-Al2O3 support. The morphology transformation was consis-
tent with the textural properties obtained from the aforemen-
tioned N2 adsorption-desorption study.
The XRD pattern of freshly calcined 3 wt% Pr-Ni/a-Al2O3 cata-
Table 1 lyst is shown in Fig. 4. The peaks at 2h = 37.21°, 43.31°, 62.78°
Textural properties of the as-synthesized catalyst and pure alumina. and 75.35° can be assigned to the cubic structure of NiO phase,
Catalysts BET specific surface area Pore diameter Pore volume while the peaks at 2h = 25.84°, 35.41°, 52.79°, 57.73°, 66.72°,
(m2 g1) (nm) (cm3 g1) 68.86° and 77.65° are due to the NiAl2O4 [44]. The crystalline phase
a-Al2O3 5.18 3.09 0.0063 associated with the Pr element cannot be detected from the XRD
3 wt% Pr-Ni/a-Al2O3 2.59 18.20 0.0064 patterns; this is most likely due to their well-dispersion nature.
The crystallite diameter of the sharpest peak from the XRD pattern

Fig. 3. FESEM images of (a) a-Al2O3 at 30 kX magnification, (b) 3 wt% Pr-Ni/a-Al2O3 at 20 kX magnification and (c) 3 wt% Pr-Ni/a-Al2O3 at 30 kX magnification.
 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880 875

Fig. 4. X-ray diffraction patterns of the 3 wt% Pr-Ni/a-Al2O3 catalyst.

Fig. 6. (a) CO2-TPD profiles and (b) NH3-TPD profiles of 3 wt% Pr-Ni/a-Al2O3
catalyst.
Fig. 5. TPR analysis of freshly calcined 3 wt% Pr-Ni/a-Al2O3 catalyst.

temperatures ranged 519–848 K (refers to Fig. 6(a)) could be

was computed from the Scherrer equation (refers to Eq. (2)) as
attributed to a weak Lewis basic since it is generally believed that
43.77 nm, indicating the nano-range of the as-synthesized catalyst.
an adsorbed CO2 on a weaker basic site would desorb at lower tem-
The TPR profile of the 3 wt% Pr-Ni/a-Al2O3 catalyst is depicted
perature region [49]. The second CO2-desorption peak at 886–
in Fig. 5. Interestingly, broad and intense H2-consumption peaks
1084 K can be assigned to a strong Lewis basic site [50]. The des-
at 460 and 697 K are visible, an indication of the H2-reduction of
orption energy (DHd ) representing the strength of basic site was
both NiO and NiAl2O4 to Nio [45] as shown in Eqs. (8) and (9).
estimated as 43.1 kJ mol1 (weak Lewis) and 69.4 kJ mol1 (strong
The H2-reduction of NiO has been reported to occur in three differ-
Lewis). Furthermore, the acidic property of the catalyst was char-
ent regions denoted as a (683–773 K), b (853–1013 K) and ƴ
acterized by three discernible NH3 desorption peaks which are
(1063–1113 K) based on the reduction temperature [46]. Hence,
recorded at 837 K, 895 K and 968 K, respectively (cf. Fig. 6(b)).
the NiO reduction in this study can be assigned to ‘‘type-a” since
These three peaks signify strong Lewis acid sites. Similarly, DHd
the reduction temperature fall within the 683–773 K domain. This
was estimated from the NH3-desorption as 48.7, 57.8, and
‘‘a-type” NiO reduction signifies weak interaction between the free
69.3 kJ mol1. Overall, the 3 wt% Pr-Ni/a-Al2O3 catalyst was more
NiO species and the a-Al2O3 support.
acidic because the acid to basic ratio was at 2.14. The presence of
more acidic site would ensure an interaction with the OH-group
NiO þ H2 ! Ni þ H2 O ð8Þ of the glycerol backbone during the surface adsorption (mechanis-
NiAl2 O4 þ H2 ! Ni þ Al2 O3 þ H2 O ð9Þ tic studies).
The CO2-TPD and NH3-TPD profiles and the acid/basic proper-
ties are shown in Fig. 6 and summarized in Table 2, respectively. 3.2. Catalytic pyrolysis reaction studies
The CO2-TPD profile reveals two different peaks. A Lewis site was
represented by DHd value lower than 125.0 kJ mol1 whilst a 3.2.1. Effects of partial pressure of glycerol and reaction temperature
Brönsted site was synonym with DHd in the range of 125.0– The effects of glycerol partial pressure on the rate of formation
145.0 kJ mol1 [47]. In addition, the strength of acid/basic sites of the gaseous products are depicted in Fig. 7. It can be observed
can be classified into three categories based on the desorption tem- that the formation rates of all gaseous products (H2, CO, CO2 and
peratures viz., weak sites (<523 K), medium sites (523–673 K) and CH4) increased with glycerol partial pressure. This confirms that
strong sites (>673 K) [48]. The first CO2-desorption peak at all the products (main products or byproducts) were originating
876 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880

Table 2
Summary of acid-base properties.

Analysis Peak Type of site DHd (kJ mol1) Acid/basic site concentration (lmol g1)
CO2-TPD I (519–848 K) Lewis 43.1 63.03
II (886–1084 K) Lewis 69.4 65.24
Total: 128.27
NH3-TPD I (785–860 K) Lewis 48.7 84.71
II (869–913 K) Lewis 57.8 92.73
III (917–1014 K) Lewis 69.3 97.21
Total: 274.65
Acid: basic sites ratio 2.14

Fig. 7. Effects of temperature and glycerol partial pressure on the rates of formation for (a) H2, (b) CH4, (c) CO2 and (d) CO over the 3 wt% Pr-Ni/a-Al2O3.

Table 3
The product yield as function of partial pressure of glycerol and pyrolysis temperature over 3 wt% Pr-Ni/a-Al2O3 catalyst.

Pgly YH2 (%) YCH4 (%) YCO2 (%) YCO (%)

(kPa) 973 K 1023 K 1073 K 973 K 1023 K 1073 K 973 K 1023 K 1073 K 973 K 1023 K 1073 K
4.5 24.86 20.64 26.27 1.74 1.83 0.25 3.49 3.67 2.50 15.70 14.68 22.51
9.0 20.25 23.10 29.04 1.12 0.91 0.39 3.37 3.62 0.65 14.62 14.50 15.49
13.5 22.11 23.09 26.37 0.95 0.77 0.86 2.85 3.08 0.86 15.22 14.62 18.01
18.0 25.27 19.37 23.55 1.98 0.62 0.71 3.96 2.46 0.71 13.88 11.69 14.98
22.5 27.42 19.95 24.38 1.28 1.21 1.20 4.49 2.42 1.20 14.11 13.30 16.25

from the glycerol as the sole reactant. In addition, the H2 (0.005– mol g1 s1) profiles as shown in Fig. 7(b) and (c). This indicates
0.025 mol g1 s1) and CO (0.002–0.012 mol g1 s1) profiles (cf. that the H2 and CO were the major products from the glycerol
Fig. 7(a) and (d)) show the highest formation rates, as opposed to pyrolysis as in Eq. (10).
the CH4 (105 to 103 mol g1 s1) and CO2 (105 to 3  103 -
C3 H8 O3 ! 3CO þ 4H2 ð10Þ
 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880 877

Table 4
Glycerol consumption rates and product formation rates over the 3 wt% Pr-Ni/a-Al2O3 catalyst.

Temperature (K) Glycerol partial pressure (kPa) Glycerol consumption rate (mol g cat1 s1) Product reaction rates (mol g cat1 s1)
H2 CH4 CO2 CO
973 0 0 0 0 0 0
973 4.51 0.00104 0.00382 0.00015 0.00040 0.00215
973 9.02 0.00268 0.00875 0.00030 0.00084 0.00401
973 13.53 0.00511 0.01295 0.00047 0.00119 0.00618
973 18.04 0.00631 0.01453 0.00101 0.00225 0.00744
973 22.55 0.00957 0.01869 0.00102 0.00309 0.00923
1023 0 0 0 0 0 0
1023 4.51 0.00115 0.00438 0.00018 0.00055 0.00207
1023 9.02 0.00227 0.00732 0.00033 0.00100 0.00421
1023 13.53 0.00434 0.01188 0.00035 0.00115 0.00603
1023 18.04 0.00617 0.01819 0.00048 0.00142 0.00667
1023 22.55 0.00948 0.02438 0.00089 0.00150 0.00935
1073 0 0 0 0 0 0
1073 4.51 0.00165 0.00532 0.00004 0.00017 0.00286
1073 9.02 0.00340 0.01147 0.00012 0.00020 0.00332
1073 13.53 0.00508 0.01542 0.00031 0.00045 0.00782
1073 18.04 0.00751 0.01872 0.00038 0.00050 0.00869
1073 22.55 0.00990 0.02593 0.00068 0.00107 0.01198

Table 5
Kinetics data for reactant and products from the glycerol pyrolysis.

Power law Species A (mol m2 s1 kPab) EA (kJ mol1) b

Reactant Glycerol 0.000339 7.14 1.34
Products H2 0.000295 12.81 1.47
CH4 0.003891 20.77 1.24
CO2 0.001678 16.18 1.28
CO 0.000508 14.94 1.52

3H2 þ CO ! CH4 þ H2 O ð11Þ

2CO ! CO2 þ C ð12Þ
Furthermore, it can be seen that generally, H2 and CO formation
rates increased with reaction temperature, which was consistent
with the Arrhenius trend. In contrast, both CH4 and CO2 exhibited
the opposite trend with rising temperature. As both methanation
reaction and Boudouard reaction are exothermic, from thermody-
namic standpoint, the formation of CH4 and CO2 was not favorable
at higher temperature, thus explained the downward trend.
In terms of product yield, the yields of H2, CO, CH4, and CO2 as a
function of temperature and partial pressure are summarized in
Table 3. The highest average yield of H2 and CO can be obtained
at the temperature of 1073 K, with an average value of 25.92%
and 17.45%, respectively. However, the highest average of CH4
and CO2 (1.41% and 3.63%) was observed at the lowest pyrolysis
temperature of 973 K. These trends were in sync with the afore-
Fig. 8. Power law of predicted and observed pyrolysis rate for 3 wt% Pr-Ni/a-Al2O3 mentioned formation rates.
catalyst.

3.2.2. Power law modeling

The reaction rates obtained from the experiments were mod-
Table 6
eled to the power-law kinetic equation as in Eq. (13).
Langmuir-Hinshelwood rate models.
EA
1
No. Model Description r G ðmol g s1 Þ ¼ Ae RT PbG ð13Þ
1 krxn P G Single site associative adsorption of glycerol (G) with surface
ð1þK G PG Þ

pffiffiffiffi
reaction as a rate-determining step where ðr G Þ is the glycerol pyrolysis rate, A is the pre-exponential
2  krxnpffiffiffiffiffiffiffiffi
PG
ffi
Single site dissociative adsorption of glycerol (G) with surface factor (s1), EA is the activation energy (kJ mol1), R is the universal
1þ K G PG reaction as a rate-determining step
gas constant (8.314 J mol1 K1), T is the reaction temperature (K),
PG is the reactant’s partial pressure (kPa) and b is the order of reac-
For the species CO2 and CH4, its several folds lower magnitude tion. The kinetic data employed for the power-law modeling are
in formation rates indicate that these two products were byprod- provided in Table 4. These data were analyzed using Levenberg-
ucts of competing reaction pathways, with CH4 likely from the Marquardt non-linear regression optimization method. The result-
methanation reaction (cf. Eq. (11)) whilst CO2 could be from the ing kinetic parameters from the analysis are summarized in Table 5.
Boudouard reaction as in Eq. (12). Parity plot shown in Fig. 8 was employed to check the adequacy of
878 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880

Table 7
LH models parameter estimation based on glycerol consumption rate.

Model no. Temperature (K) krxn (mol m2 s1 kPab) KG (kPa1) R2
04
1 973 2.15  10 0.0216 0.99
1023 2.61  1004 0.0168 0.99
1073 3.31  1004 0.0109 0.99
2 973 N/A N/A N/A
1023 N/A N/A N/A
1073 N/A N/A N/A

Table 8
BMW guidelines for Model 1 estimation.

Parameter Value R2 1st guideline 2nd guideline

1 1
DSG (J mol K ) 92.30 0.97 92.30 P 10 92.30 6 98.49
DHG (kJ mol1) 79.06 0.97
1/T (K1) ln krxn EA (kJ mol1) R2
0.001028 8.445 37.36 0.9925
0.000978 8.251
0.000932 8.013

the power-law model. The activation energy for the glycerol con- C3 H8 O3 þ 2X $ HCOOX  X þ CH3 CHOH  X þ H2
sumption rate was 7.14 kJ mol1, which suggests a superior perfor- HCOOX  X ! CO2 þ H  X
mance. In comparison, Hemings and co-workers undertook detailed
CH3 CHOH  X þ H  X ! CH3 O  X þ CH3  X
kinetics study of pyrolysis and oxidation of glycerol, with reported
EA value of 10 kJ mol1 [51], while EA of 6.4–12.4 kJ mol1 were CH3 O  X þ X ! CH2  X þ OH  X
obtained from another previous research [52]. In addition, Hou CH2  X þ H  X ! CH3  X þ X
and Hughes [43] have also reported activation energy of CH3  X þ H  X ! CH4 þ 2X
11.8 kJ mol1. Furthermore, the main gaseous products (H2 and CH3 O  X þ X ! CH2 O  X þ H  X
CO) showed activation energy values of 12.0 and 15 kJ mol1,
CH2 O  X þ X ! HCO  X þ H  X
respectively, whilst the CH4 and CO2 exhibited slightly higher acti-
vation energy values of 16.0 and 21.0 kJ mol1, respectively. This HCO  X þ X ! CO  X þ H  X
may be due to the slower formation rates of CH4 and CO2 since CO  X $ CO þ X
these species were the byproducts from glycerol pyrolysis. In addi- H  X þ H  X $ H2 þ 2X
tion, the reaction orders were estimated in the range of 1.2–1.6
which deviates from 0.1 to 0.6 as reported by Cheng et al. [11]. This where X is the active site.
discrepancy might be due to variation in the catalytic system as Using quasi-steady-state approximation and incorporating the
well as the reaction conditions. concept of most abundant reactive intermediate (MARI), two
models corresponding to the reaction mechanisms described ear-
3.2.3. Langmuir-Hinshelwood (LH) single site mechanism lier were developed and are summarized in Table 6. The LH models
Based on the NH3-TPD results, the Langmuir-Hinshelwood (LH) were regressed using Levenberg-Marquardt non-linear regression
single site mechanism was proposed. It assumes that the glycerol in POLYMATHTM. The kinetics parameters estimated from the LH
molecule has either molecularly- or dissociatively-adsorbed on model using Polymath software are shown in Table 7. It can be
the identical site of the catalyst. Based on single site mechanism, seen that only Model 1 gave good fitting with regression coefficient
the pyrolysis reaction entails the chemical bonding of the adsorbed R2 of 0.99 whereas Model 2 failed to converge. In order to verify
glycerol vapor onto the active site as follow: the thermodynamic consistency of Model 1, the kinetic parameters
Molecular adsorption of glycerol in Table 7 were subjected to a further validation using Eq. (14)
[53]:
C3 H8 O3 þ X $ C3 H8 O3  X
C3 H8 O3  X þ X ! HCOO  X þ CH3 CHOH  X þ H2  DH DS
ln K ¼ þ ð14Þ
HCOOX  X ! CO2 þ H  X ðRTÞ R
CH3 CHOH  X þ H  X ! CH3 O  X þ CH3  X
where K is the adsorption constant (kPa1), DH is the enthalpy
CH3 O  X þ X ! CH2  X þ OH  X (kJ mol1), DS is the entropy (J mol1 K1), R is the universal gas
CH2  X þ H  X ! CH3  X þ X constant (8.314 J mol1 K1), T is the reaction temperature (K).
CH3  X þ H  X ! CH4 þ 2X The criteria based on Eqs. (15) and (16) must be satisfied to ensure
that the Model 1 is thermodynamically valid.
CH3 O  X þ X ! CH2 O  X þ H  X
CH2 O  X þ X ! HCO  X þ H  X
10 6 DS ð15Þ
HCO  X þ X ! CO  X þ H  X
 DS 6 12:2  0:0014DH ð16Þ
CO  X $ CO þ X
The results of the thermodynamic consistency of the model are
H  X þ H  X $ H2 þ 2X
summarized in Table 8. Significantly, Model 1 has fulfilled the cri-
Dissociative adsorption of glycerol teria in Eqs. (15) and (16), suggesting that the said model was ther-
 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880 879

modynamically consistent. Furthermore, the activation energy for [18] M.L. Dieuzeide, V. Iannibeli, M. Jobbagi, N. Amadeo, Steam reforming of
glycerol over Ni/Mg/c-Al2O3 catalysts. Effect of calcination temperatures, Int. J.
the catalytic glycerol pyrolysis was evaluated from the LH model
Hydrogen Energy 37 (2012) 14926–14930.
as 37.36 kJ mol1. [19] L.F. Bobadilla, A. Álvarez, M.I. Domínguez, F. Romero-Sarria, M.A. Centeno, M.
Montes, J.A. Odriozola, Influence of the shape of Ni catalyst in the glycerol
steam reforming, Appl. Catal. B: Environ. 123 (2012) 379–390.
[20] L. O’Sullivan, Hydrogen as an Alternative Energy Source. Environmentalism,
4. Conclusion
2008. Retrieved from <http://suite101.com/article/hydrogen-as-an-
alternative-energy-source-a59685#>.
We have reported here, the catalytic syngas production from [21] S. Carrettin, P. McMorn, P. Johnston, K. Griffin, C.J. Kiely, G.J. Hutchings,
glycerol pyrolysis over 3 wt% Pr-Ni/a-Al2O3 catalyst. The 3 wt% Oxidation of glycerol using supported Pt, Pd and Au catalysts, Phys. Chem.
Chem. Phys. 5 (2003) 1329–1336.
Pr-Ni/a-Al2O3 catalyst was synthesized via wet impregnation [22] M. Nilsson, L.J. Pettersson, B. Lindstro, Hydrogen generation from Dimethyl
method. It was found that the order of product formation rates Ether for fuel cell auxiliary power units, Energy Fuels 20 (5) (2006) 2164–
with respect to H2, CO, CO2 and CH4 were 1.47, 1.24, 1.28 and 2169.
[23] P.J. Dauenhauer, J.R. Salge, L.D. Schmidt, Renewable hydrogen by autothermal
1.52, respectively, with corresponding activation energy values of steam reforming of volatile carbohydrates, J. Catal. 244 (2006) 238–247.
12.81, 20.77, 16.18 and 14.94 kJ mol1, respectively. Moreover, [24] A.M.D. Douette, S.Q. Turn, W. Wang, V.I. Keffer, Experimental investigation of
Langmuir-Hinshelwood model was also employed for the model- hydrogen production from glycerin reforming, Energy Fuels 21 (2007) 3499–
3504.
ing of the kinetics data. Statistical and thermodynamic discrimina- [25] S. Adhikari, S. Fernando, A. Haryanto, Hydrogen production from glycerol: an
tion of the associated kinetic parameters indicate that the most update, Energy Conv. Manage. 50 (2009) 2600–2604.
adequate representation of the mechanisms of the glycerol pyroly- [26] F.J. Gutiérrez Ortiz, P. Ollero, A. Serrera, A. Sanz, Thermodynamic analysis of
the autothermal reforming of glycerol using supercritical water, Int. J.
sis involved a molecular single site adsorption of the glycerol with
Hydrogen Energy 36 (19) (2011) 12186–12199.
surface reaction as the rate-determining step. [27] H.C. Lee, K.W. Siew, M.R. Khan, S.Y. Chin, J. Gimbun, C.K. Cheng, Catalytic
performance of cement clinker supported nickel catalyst in glycerol dry
reforming, J. Energy Chem. 23 (5) (2014) 645–656.
Acknowledgements [28] Y. Fernandez, A. Arenillas, J.M. Bermudez, J.A. Menendez, Comparative study of
conventional and microwave-assisted pyrolysis, steam and dry reforming of
glycerol for syngas production, using a carbonaceous catalyst, J. Anal. Appl.
Authors would like to acknowledge the financial support (RACE Pyrol. 88 (2010) 155–159.
fund with vot. no. of RDU151302) from the Ministry of Higher Edu- [29] K.W. Siew, H.C. Lee, J. Gimbun, C.K. Cheng, Characterization of La-promoted Ni/
Al2O3 catalysts for hydrogen production from glycerol dry reforming, J. Energy
cation, Malaysia. Nor Shahirah Mohd Nasir acknowledges the stu-
Chem. 23 (1) (2014) 15–21.
dentship (MyBrain15) awarded by the same ministry. [30] J. Gao, Z.Y. Hou, H. Lou, X.M. Zheng, Dry (CO2) reforming, Fuel Cells 7 (2011)
192–213.
[31] Y. Fernández, A. Arenillas, M.A. Díez, J.J. Pis, J.A. Menéndez, Pyrolysis of glycerol
References over activated carbons for syngas production, J. Anal. Appl. Pyrol. 84 (2009)
145–150.
[32] V. Skoulou, P. Manara, A. Zabaniotou, H2 enriched fuels from co-pyrolysis of
[1] H.W. Tan, A.R. Abdul Aziz, M.K. Aroua, Glycerol production and its applications
crude glycerol with biomass, J. Anal. Appl. Pyrol. 97 (2012) 198–204.
as a raw material: a review, Renew. Sustain. Energy Rev. 27 (2013) 118–127.
[33] K.W. Siew, H.C. Lee, J. Gimbun, C.K. Cheng, Production of CO-rich hydrogen gas
[2] M.S. Ardi, M.K. Aroua, N.A. Hashim, Progress, prospect and challenges in
from glycerol dry reforming over La-promoted Ni/Al2O3 catalyst, Int. J.
glycerol purification process: a review, Renew. Sustain. Energy Rev. 42 (2015)
Hydrogen Energy 39 (13) (2014) 6927–6936.
1164–1173.
[34] K.W. Siew, H.C. Lee, J. Gimbun, S.Y. Chin, M.R. Khan, Y.H. Taufiq-Yap, C.K.
[3] T. Ito, Y. Nakashimada, K. Senba, T. Matsui, N. Nishio, Hydrogen and ethanol
Cheng, Syngas production from glycerol-dry(CO2) reforming over La-promoted
production from glycerol-containing wastes discharged after biodiesel
Ni/Al2O3 catalyst, Renew. Energy 74 (3) (2015) 441–447.
manufacturing process, J. Biosci. Bioeng. 100 (2005) 260–265.
[35] E.P. Barrett, L.G. Joyner, P.P. Halenda, The determination of pore volume and
[4] H. Ong, H.H. Masjuki, T.M.I. Mahlia, Engine performance and emissions using
area distributions in porous substances. I. Computations from nitrogen
Jatropha curcas, Ceiba pentandra and Calophyllum inophyllum biodiesel in a CI
isotherms, J. Am. Chem. Soc. 73 (1951) 373–380.
diesel engine, Energy 69 (2014) 427–445.
[36] S. Brunauer, P.H. Emmett, E. Teller, Adsorption of gases in multimolecular
[5] C.A.G. Quispe, C.J.R. Coronado, J.A. Carvalho Jr., Glycerol: production,
layers, J. Am. Chem. Soc. 60 (1938) 309–319.
consumption, prices, characterization and new trends in combustion, Renew.
[37] S.Y. Foo, Oxidative Dry Reforming of Methane over Alumina-Supported Co-Ni
Sustain. Energy Rev. 27 (2013) 475–493.
Catalyst Systems Doctor of Philosophy, 2012.
[6] F. Bauer, C. Hulteberg, Is there a future in glycerol as a feedstock in the
[38] W. Brockner, C. Ehrhardt, M. Gjikaj, Thermal decomposition of nickel nitrate
production of biofuels and biochemicals?, Biofuels Bioprod Biorefining 7
hexahydrate, Ni(NO3)26H2O, in comparison to Co(NO3)26H2O and Ca
(2013) 43–51.
(NO3)24H2O, Thermochim. Acta 456 (2007) 64–68.
[7] B.V. Ayodele, M.R. Khan, C.K. Cheng, Syngas production from CO2 reforming of
[39] B.A.A. Balboul, Synthesis course and surface properties of praseodymium oxide
methane over ceria supported cobalt catalyst: effects of reactants partial
obtained via thermal decomposition of praseodymium acetate: impacts of the
pressure, J. Nat. Gas Sci. Eng. 27 (2015) 1016–1023.
decomposition atmosphere, J. Anal. Appl. Pyrol. 88 (2010) 192–198.
[8] K. Aasberg-Petersen, I. Dybkjær, C.V. Ovesen, N.C. Schjødt, J. Sehested, S.G.
[40] B.V. Ayodele, M.A. Hossain, S.L. Chong, J.C. Soh, S. Abdullah, M.R. Khan, C.K.
Thomsen, Natural gas to synthesis gas - catalysts and catalytic processes, J. Nat.
Cheng, Non-isothermal kinetics and mechanistic study of thermal
Gas Sci. Eng. 3 (2011) 423–459.
decomposition of light rare earth metal nitrate hydrates using
[9] Y.C. Lin, Catalytic valorization of glycerol to hydrogen and syngas, Int. J.
thermogravimetric analysis, J. Therm. Anal. Calorim. 1–13 (2016).
Hydrogen Energy 38 (2013) 2678–2700.
[41] H.C. Lee, K.W. Siew, J. Gimbun, C.K. Cheng, Synthesis and characterisation of
[10] B. Zhang, X. Tang, Y. Li, Y. Xu, W. Shen, Hydrogen production from steam
cement clinker-supported nickel catalyst for glycerol dry reforming, Chem.
reforming of ethanol and glycerol over ceria-supported metal catalysts, Int. J.
Eng. J. 255 (2014) 245–256.
Hydrogen Energy 32 (2007) 2367–2373.
[42] Z. Hou, Characterization of Ca-promoted Ni/a-Al2O3 catalyst for CH4 reforming
[11] C.K. Cheng, S.Y. Foo, A.A. Adesina, Glycerol steam reforming over bimetallic Co-
with CO2, Appl. Catal. A Gen. 253 (2003) 381–387.
Ni/Al2O3, Ind. Eng. Chem. Res. 49 (2010) 10804–10817.
[43] K. Hou, R. Hughes, The kinetics of methane steam reforming over a Ni/a-Al2O
[12] C.K. Cheng, S.Y. Foo, A.A. Adesina, Carbon deposition on bimetallic Co–Ni/Al2O3
catalyst, Chem. Eng. J. 82 (2001) 311–328.
catalyst during steam reforming of glycerol, Catal. Today 164 (2011) 268–274.
[44] S. Tada, M. Yokoyama, R. Kikuchi, T. Haneda, H. Kameyama, N2O pulse titration
[13] J. Sehested, Four challenges for nickel steam-reforming catalysts, Catal. Today
of Ni/a-Al2O3 catalysts: a new technique applicable to nickel surface-area
111 (2006) 103–110.
determination of nickel-based catalysts, J. Phys. Chem. C 117 (2013) 14652–
[14] D.C. Rennard, J.S. Kruger, L.D. Schmidt, Autothermal catalytic partial oxidation
14658.
of glycerol to syngas and to non-equilibrium products, ChemSusChem 2 (2009)
[45] M. Koike, C. Ishikawa, D. Li, L. Wang, Y. Nakagawa, K. Tomishige,
89–98.
Catalytic performance of manganese-promoted nickel catalysts for the steam
[15] E.E. Iojoiu, M.E. Domine, T. Davidian, N. Guilhaume, C. Mirodatos, Hydrogen
reforming of tar from biomass pyrolysis to synthesis gas, Fuel 103 (2013) 122–
production by sequential cracking of biomass-derived pyrolysis oil over noble
129.
metal catalysts supported on ceria-zirconia, Appl. Catal. A Gen. 323 (2007)
[46] D. Hu, J. Gao, Y. Ping, L. Jia, P. Gunawan, Z. Zhong, G. Xu, F. Gu, F. Su, Enhanced
147–161.
investigation of CO methanation over Ni/Al2O3 catalysts for synthetic natural
[16] T. Valliyappan, D. Ferdous, N.N. Bakhshi, A.K. Dalai, Production of hydrogen
gas production, Ind. Eng. Chem. Res. 51 (2012) 4875–4886.
and syngas via steam gasification of glycerol in a fixed-bed reactor, Top. Catal.
[47] G. Yaluris, R.B. Larson, J.M. Kobe, M.R. Gonzalez, K.B. Fogash, J.A. Dumestic,
49 (2008) 59–67.
Selective poisoning and deactivation of acid sites on sulfated zirconia catalysts
[17] V.V. Thyssen, T.A. Maia, E.M. Assaf, Ni supported on La2O3eSiO2 used to
for n-butane isomerization, J. Catal. 158 (1996) 336–342.
catalyze glycerol steam reforming, Fuel 105 (2013) 358–363.
880 M.N.N. Shahirah et al. / Applied Thermal Engineering 112 (2017) 871–880

[48] A. Auroux, R. Monaci, E. Rombi, V. Solinas, A. Sorrentino, E. Santacesaria, Acid [51] E.B. Hemings, C. Cavallotti, A. Cuoci, T. Faravelli, E. Ranzi, A detailed kinetic
sites investigation of simple and mixed oxides by TPD and microcalorimetric study of pyrolysis and oxidation of glycerol (propane-1,2,3-triol), Combust. Sci.
techniques, Thermochim. Acta 379 (2001) 227–231. Technol. 184 (7–8) (2012) 1164–1178.
[49] L. Pino, A. Vita, F. Cipitì, M. Laganà, V. Recupero, Hydrogen production by [52] Ch. Pichas, P. Pomonis, D. Petrakis, A. Ladavos, Kinetic study of the catalytic dry
methane tri-reforming process over Ni–ceria catalysts: effect of La-doping, reforming of CH4 with CO2 over La2xSrxNiO4 perovskite-type oxides, Appl.
Appl. Catal. B Environ. 104 (2011) 64–73. Catal. A: Gen. 386 (2010) 116–123.
[50] V.R. Choudhary, V.H. Rane, A novel method for measuring base strength [53] M. Vannice, Entropies of adsorption in heterogeneous catalytic reactions, J.
distribution on basic solid catalysts under operating conditions, Catal. Lett. 4 Catal. 56 (1979) 358–362.
(1990) 101–106.