Post on 04-Mar-2023
nline at www.sciencedirect.com
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8
Available o
www. i ifi i r .org
ScienceDirect
journal homepage: www.elsevier .com/locate/ i j refr ig
Experimental and theoretical study of adsorptionkinetics of Difluoromethane onto activated carbons
Ahmed A. Askalany a,*, Bidyut Baran Saha b,c
a Mechanical Engineering Department, Faculty of Industrial Education, Sohag University, Sohag 82524, Egyptb Interdisciplinary Graduate School of Engineering Sciences, Kyushu University, Kasuga-koen 6-1, Kasuga-shi,
Fukuoka 816-8580, Japanc International Institute for Carbon-Neutral Energy Research (WPI-I2CNER), Kyushu University, 744 Motooka,
Nishi-ku, Fukuoka 819-0395, Japan
a r t i c l e i n f o
Article history:
Received 19 June 2014
Received in revised form
28 September 2014
Accepted 7 October 2014
Available online 16 October 2014
Keywords:
Adsorption kinetics
Activated carbon
Difluoromethane
LDF
FD
* Corresponding author. Tel.: þ20862730572,E-mail address: ahmed_askalany3@yahoo
http://dx.doi.org/10.1016/j.ijrefrig.2014.10.0090140-7007/© 2014 Elsevier Ltd and IIR. All rig
a b s t r a c t
This paper presents theoretical and experimental adsorption kinetics of Difluoromethane
(HFC-32) onto activated carbon powder of type Maxsorb III and activated carbon fiber of
type A-20. The experimental runs have been conducted on an apparatus that has been
designed and built specially for this purpose. The adsorption kinetics have been deter-
mined at different adsorption temperatures ranging from 25 �C to 65 �C. The experimental
data are reduced and fitted with linear driving force (LDF) and Fickian diffusion (FD)
models. It has been found that both LDF and FD models are able to simulate the adsorption
kinetics of HFC-32 onto activated carbon powder and fiber with an error of ±5%.
© 2014 Elsevier Ltd and IIR. All rights reserved.
Etude exp�erimentale et th�eorique de la cin�etique d’adsorptionde de difluorom�ethane sur du charbon activ�e
Mots cl�es : Cin�etique d’adsorption ; Charbon activ�e ; Difluorom�ethane ; Lin�eaire force motrice ; Diffusion de Fickian
1. Introduction
According to the dramatic increase of the global energy de-
mand that has been occupied by a shortage in energy supply,
renewable energy applications becomeone of the solutions for
theenergy crisis.Oneof these applicationswhichmaybeaid in
mobile: þ201028721274..com (A.A. Askalany).
hts reserved.
solving this problem is the thermally powered adsorption
cooling system. Unremitting researches have been carried out
to bring the adsorption cooling systems to the existence (Alam
et al. 2013; Askalany et al., 2012, 2013a; Chan et al., 2012; Dabler
and Mittelbach, 2012; Gordeeva and Aristov, 2014; Khattab
et al., 2012; Sharafian and Bahrami, 2014; Tso and Chao, 2012).
Nomenclature
A surface area, m2
C adsorption uptake, kg kg�1
C0 equilibrium uptake, kg kg�1
Ds surface diffusion, m2 s�1
Ds0 pre-exponential coefficient, m2 s�1
Ea activation energy, kJ kg�1
F constant
m mass of adsorbate, kg
n constant
Rp particle radius, m
t time, s�1
V volume, m3
Subscripts
ads adsorber
eva evaporator
exp experimental
f final status
i initial
pore pore volume
sol solid
theo theoretical
tube connecting tubes
void void spaces between adsorbent particles
z measurement number
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8 161
In order to build an adsorption cooling system, there are
some indispensable steps which should be followed. The
first step is studying the adsorption isotherms of the
adsorbent/refrigerant pair. The second step is studying the
adsorption kinetics of the working pair. Adsorption kinetics
could be defined simply as time-variation of the rate of
adsorption capacity for an adsorbent/adsorbate pair. For
better design of adsorption chillers, it is essential to deter-
mine accurately the kinetics of adsorption pair. Many re-
searches have been conducted to determine the adsorption
kinetics of different adsorption pairs (Atakan et al., 2013;
Dawoud, 2013; Glaznev and Aristov, 2008). These results
have been conducted to detect the most accurate theoretical
model which fits the experimental data of the studied
adsorption pair.
Difluoromethane (HFC-32) is a hydofluoroocarbon refrig-
erant that has been used either as a pure fluid for low tem-
perature refrigeration or as an ingredient of 400 series
nonazeotropic refrigerants 407a,b,c and 410a,b for packaged
air conditioning applications. HFC-32 is also present in R-504
that was used as a refrigerant but now phased out under
Montreal Protocol because of the presence of chloropenta-
fluoroethane (CFC-115) which has large ozone depletion po-
tential. Adsorption data on HFC-32 are expected to be
beneficial for two distinct applications, namely, for possible
thermal compression when used as a pure component and
separation of the mixtures at the end of the life cycle
(Askalany et al., 2013b).
Table 1 e Specifications of the adsorbents.
ACP ACF
Pore volume (cm3 g�1) 1.7 1.0
Surface area/m2 g�1 3200 2200
Mean pore diameter/nm 2 2
Amount in adsorption cell (g) 10.10 2.75
2. Working pairs and experimentalapparatus
Activated carbon fiber (ACF) of type A-20 and activated carbon
powder (ACP) of type Maxsorb III, were developed by AD0ALLCo. Ltd., Japan, and Kansai Coke & Chemicals Co. Ltd., Japan,
respectively. Table 1 summarizes the characteristics of the
studied adsorbents. A detailed description and chemical
analysis of the used adsorbents have been presented by
Askalany et al., 2013b. Samples of HFC-32 were supplied by
Daikin Co. Ltd., Japan and helium by Asahi Sanso Shokai Ltd.,
Japan, respectively with a stated purity of 99.995% for both
HFC-32 and helium.
An experimental apparatus has been designed and built to
investigate the adsorption kinetics of HFC-32 onto activated
carbonbased adsorbents. It consistsmainly of an adsorber and
an evaporator as shown in Fig. 1. The Adsorber and the evap-
orator are connected through 1/4 inch steel tube. In order to
control the flow of refrigerant, seven ball valves are installed.
The adsorber is 2500 cm3 bottle made of a stainless steel (SUS
304) and withstands a maximum loading pressure of 3 MPa.
A copper tube coil of a diameter of 1/8 inch has been
fabricated inside the adsorber to improve heating and
cooling the adsorbent as shown in Fig. 2. The copper coil is
connected to a water circulator. The circulator has a heating
and cooling capacity of 1.7 kW which used to control the
temperature of the water bath. The adsorber has a mesh
which is used to prevent the migration of the adsorbent
particles. The evaporator is made of stainless steel (SUS 304)
bottle of a volume of 3 L (3000 cm3). It has been designed to
manage a maximum pressure of 3 MPa. Each of the adsorber
and the evaporator is installed inside a water bath with
0.2 m3 volume in order to manage and regulate its temper-
ature. Each water bath is connected to a water circulator in
order to heat or cool the water and to achieve a preselected
temperature.
The apparatus has been connected to a vacuum pump
which has been used for evacuation purposes. A number of
type K thermocouples are attached at the locations in the test
rig where the temperature has to be measured.
Pressure transducers are used to measure the pressure of
the adsorber and the evaporator. A high precisionCoriolis flow
meter is installed between the evaporator and the adsorber to
measure the flow rate of the refrigerant during the adsorption
Fig. 1 e Schematic diagram of the experimental apparatus adsorption-kinetics evaluation.
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8162
processes. The accuracy and sensitivity of the used in-
struments have been furnished in Table 2.
3. Test procedure
Before inserting the adsorbent inside the adsorber, it has been
heated up to 150 �C for duration of 3 h to extract its moisture
then it has been weighed. After inserting the adsorbent into
the adsorber, the whole apparatus is evacuated using a vac-
uum pump with a pumping speed of 500 L min�1 and 9.3 Pa
ultimate pressure. The apparatus has been evacuated to a
pressure of 10 Pa in about 3 h in order to ensure that there are
no residual gases inside the apparatus.
Temperatureof theadsorberandevaporator isadjusted tobe
at the same value using the water circulators and water paths.
The evaporator is then charged by refrigerant to a certain
pressure. After reaching to almost steady state conditions of
pressure and temperatureof the evaporator, thevalvesbetween
the evaporator and the adsorber are suddenly opened. Hence
the refrigerant vapor flows from the evaporator to the adsorber
through the flowmeter. The valves are closed after 100 s where
most of the adsorption process occurs at that time as shown in
Fig. 3. The refrigerant is then charged again to the evaporator.
The filling and transient period of the evaporator has been
shown in Fig. 4 for a representative temperature of 35 �C. Zone Ishows thefillingperiod andzone II shows the transientuntil the
stable conditions.Thisprocesshasbeenrepeated six timesuntil
reached to the maximum uptake of the refrigerant by the
adsorbent. The maximum pressure of the evaporator is not
allowed to exceed the saturation pressure in order to avoid
liquid transfer to the adsorber. Thepressure and temperature of
the evaporator and the adsorber has been recorded every 1 s. At
the end of measurements for an assigned adsorption-
temperature the whole system has to be evacuated. The evac-
uation process using vacuum pump continue for 24 h, during
which the system was heated up to 100 �C to eliminate any re-
sidual gases from the system.
In order to investigate the effect of the temperature on the
adsorption kinetics, the adsorption process has been repeated
at different temperatures in the range of 25 �Ce65 �C. Theadsorbed amount of the refrigerant has been measured by
high sensitive flow meter and also determined by using the
volumetric method. An example for the raw data of the flow
meter has been shown in Fig. 5 at adsorption temperature of
35 �C. In order to determine the adsorbed amount of the
refrigerant in grams at a certain adsorption temperature,
equations from 1 to 4 have been used. All densities are
determined by REFPROP9 software at the equilibrium
conditions.
Fig. 2 e Adsorber construction.
Table 2 e Characteristics of the utilized instruments.
Instrument Type Accuracy Sensitivity
Thermocouple K ±0.05 �C 0.01 �CPressure
transducer
Digital ±0.25% of reading 1 Pa
Flow meter Coriolis ±0.5% of reading 4 g min�1
Balance Digital ±0.05 g 0.1 g
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8 163
mz ¼ mz�1 þ Dmload;z � Dmvoid;z (1)
Dmeva;z ¼�rads;i � rads;f
�eva
ðVeva þ VtubeÞ (2)
Dmvoid;z ¼�rads;z � rads;z�1
�ads cell
ðVvoidÞ (3)
Vvoid ¼ Vads ��madsorbent
rsolþ Vpore
�(4)
Fig. 3 e Pressure and temperature records at 35 �C of ACP/
HFC-32.
Fig. 5 e An example for flow meter readings at adsorption
temperature of 35 �C of ACP/HFC-32.
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8164
Here adsorbent refers to the used activated carbon. The
experimental data of ACP/HFC-32 and ACF/HFC-32 have been
presented in Table 3 and Table 4, respectively.
Table 3 e Experimental data of ACP/HFC-32.
t, sec Tevaa, �C Tads
b, �C P, MPa C, kg kg�1
130 25.2 26.1 0.44 0.48
250 23.2 23.9 0.66 0.80
370 25.2 26.3 1.05 1.02
490 25.1 26.1 1.36 1.25
600 24.8 24.9 1.36 1.27
130 34.2 35.5 0.44 0.36
250 34.1 35.3 0.75 0.66
370 34.4 36.0 0.97 0.92
490 34.5 35.5 1.18 1.05
610 34.5 34.8 1.40 1.18
710 34.7 34.9 1.39 1.20
130 45.7 46.7 0.23 0.33
250 43.9 44.1 0.41 0.60
370 44.2 44.6 0.64 0.80
490 44.2 45.4 0.96 0.99
600 44.6 44.9 0.94 1.10
4. Theoretical models
4.1. Semi-infinite model
Ruthven (1984) reported that during the initial phase of
adsorption, surface diffusion coefficient follows the uptake
behavior of a semi-infinite medium of any particle shape and
can be expressed as:
CC0
¼ 2AV
ffiffiffiffiffiffiffiDstp
r(5)
A plot of the relative uptake and square root of time for
each isotherm yields with a slope of 2A=VffiffiffiffiffiffiffiffiffiffiffiDs=p
p. So that the
values of Ds could be extracted which are expressed as:
Ds ¼ Ds0 exp
��Ea
RT
�(6)
Which can be rearranged as.
Fig. 4 e An example of filling and transient period of the
evaporator.
lnðDsÞ ¼ lnðDs0Þ � Ea
RT(7)
By plotting ln(Ds) against T�1, one can get the numerical
values of activation energy and the pre-exponential coeffi-
cient. This plot is known as the Arrhenius plot. The slope
yields -Ea/R and the intercept provides the constant, Ds0.
130 54.8 54.1 0.55 0.33
250 54.6 55.6 0.76 0.55
370 54.3 55.5 0.97 0.74
490 54.1 54.5 1.37 0.92
600 54.4 54.8 1.35 0.97
100 63.7 63.9 0.42 0.23
210 64.1 64.6 0.74 0.42
320 64.2 65.1 0.71 0.60
450 64.2 65.0 1.18 0.77
570 64.2 64.8 1.37 0.88
680 64.3 64.8 1.37 0.94
130 74.0 74.0 0.47 0.20
250 74.0 75.1 0.77 0.42
370 74.2 75.2 0.96 0.56
490 73.9 74.0 1.37 0.72
600 74.2 75.2 1.36 0.76
a ads: adsorber.b eva: evaporator.
Table 4 e Experimental data of ACF/HFC-32.
t, sec Teva, �C Tads, �C P, MPa C, kg kg�1
130 24.9 25.2 0.31 0.33
250 24.9 25.1 0.53 0.50
370 24.9 24.7 0.76 0.69
490 24.9 24.9 1.00 0.81
567 24.3 25.1 0.99 0.88
630 24.3 25.2 1.30 0.93
688 24.8 24.4 1.30 0.94
128 33.9 33.9 0.10 0.30
246 34.3 34.8 0.26 0.44
366 34.0 35.0 0.45 0.60
485 34.6 34.8 0.66 0.75
606 34.3 34.4 0.95 0.83
722 34.7 35.3 0.94 0.86
124 44.0 44.4 0.25 0.22
243 44.1 44.3 0.45 0.39
364 43.7 44.2 0.66 0.56
484 43.4 44.0 0.88 0.67
602 44.1 44.6 1.10 0.75
719 44.8 45.5 1.08 0.80
130 52.5 53.4 0.40 0.25
250 52.8 53.8 0.64 0.40
370 53.2 54.6 0.87 0.51
490 53.6 54.5 1.10 0.60
610 53.8 54.4 1.31 0.67
710 54.7 55.3 1.30 0.69
130 62.5 63.1 0.40 0.23
250 62.7 63.1 0.65 0.36
370 63.1 64.0 0.88 0.46
490 63.4 63.3 1.10 0.54
610 63.5 63.9 1.31 0.61
710 64.5 65.5 1.30 0.63
130 72.3 73.4 0.42 0.15
250 72.6 73.8 0.65 0.23
370 72.7 73.3 0.89 0.30
490 73.1 74.0 1.11 0.35
610 73.3 74.3 1.32 0.37
710 74.1 75.3 1.31 0.39
Fig. 6 e Time variation of rate of change of adsorbate
concentration for ACP/HFC-32.
Fig. 7 e Arrhenius plot of ACP/HFC-32.
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8 165
4.2. Linear Driving Force model (LDF)
The Linear Driving Force (LDF) model for gas adsorption ki-
netics is frequently and successfully used for analysis of
adsorption column dynamic data and for adsorptive process
designs because it is simple, analytic, and physically consis-
tent (Sircar and Hufton, 2000). The LDF model is shown in Eq.
(8) (Ng et al., 2013).
vcvt
¼ FDs
R2p
ðc0 � cÞ (8)
Where the values of the constant F is 15 for spherical shape
and 8 for cylinderical shape (El-Sharkawy et al., 2006).
Fig. 8 e Fitting of LDF (solid lines) and FD (dashed lines)
models with experimental adsorption kinetics for ACP/
HFC-32.
Fig. 11 e Arrhenius plot of ACF/HFC-32.
Fig. 9 e Difference between calculated and measured
adsorption uptakes for LDF and FD models against time for
ACP/HFC-32.
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8166
4.3. Fickian diffusion (FD)
The Fickian diffusion model is used to evaluate adsorption
kinetics of various types of adsorbate pairs. The Fickianmodel
is valid when the diffusivity is independent of adsorbate
concentration (Habib et al., 2010). The diffusion equation
which expresses the adsorption rate considering a constant
diffusivity is given by Eq. (9) (Sircar and Hufton, 2000).
cc0
¼ 1� 6p2
X∞n¼1
1n2
exp
� n2p2Dst
R2p
!(9)
where n is an integer number.
Fig. 10 e Time variation of rate of change of adsorbate
concentration for ACF/HFC-32.
5. Results
5.1. ACP/HFC-32
Fig. 6 shows the rate of change of adsorption capacity for ACP/
HFC-32 adsorption pair. The adsorption capacity rate de-
creases with the increase in time of adsorption. Fig. 7 shows
the Arrhenius plot of ACP/R32 which is essential to determine
the numerical values of Ds0 and Ea. Fig. 8 shows fitting of LDF
and FD models with the experimental data of adsorption ki-
netics for ACP/R HFC-32. This figure indicates that the both
models have a good fitting with the experimental data of
adsorption kinetics of ACP/HFC-32 adsorption pair. Fig. 9
shows the difference between the calculated and experi-
mental adsorption kinetics of ACP/HFC-32 and the difference
Fig. 12 e Fitting of LDF (solid lines) and FD (dashed lines)
models with experimental adsorption kinetics for ACF/
HFC-32.
Fig. 13 e Difference between calculated and measured
adsorption uptakes for LDF and FD models against time for
ACF/HFC-32.
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8 167
has been calculated employing the following equation, Eq.
(10).
DC ¼ Ctheo � Cexp
Cexp*100 (10)
According to the statistical error analysis the maximum
error percentage is 9.2% and the mean error is 4.6% for FD and
6.1% for LDF. It is clear from this figure that most of the
calculated values have error within ±5% which is acceptable
from the engineering point of view.
5.2. ACF/HFC-32
Fig. 10 shows rate of change of adsorption capacity of ACF/
HFC-32 with time at different adsorption temperatures.
Fig. 11 shows the Arrhenius plot of ACF/HFC-32.
Fig. 12 shows the fitting of LDF and FD models with the
experimental data of ACF/HFC-32. These figures indicate
that the two different theoretical models fit well with the
experimental data of the adsorption kinetics of ACF/HFC-32
pair.
Fig. 13 shows the difference between theoretical and
experimental adsorption kinetics for both LDF and FD
models. According to the statistical error analysis the
maximum error is �10% and the mean error is 6.7% for FD
and 4.2% for LDF.
Table 5 shows the activation energy and pre-exponential
coefficient for ACP/HFC-32 and ACF/HFC-32.
Table 5 e Activation energy and pre-exponentialcoefficient for ACP/HFC-32 and ACF/HFC-32.
Parameter ACP/HFC-32 ACF/HFC-32
Ds0, m2 s�1 1.2 � 10�13 1.3 � 10�13
Ea, kJ kg�1 23.8 84
6. Conclusions
The adsorption kinetics of ACP/HFC-32 and ACF/HFC-32
adsorbent/adsorbate pairs have been investigated experi-
mentally and theoretically. An experimental apparatus has
been designed and built to conduct the experimental runs.
The experimental data have been fitted with two commonly
used theoretical models (LDF and FD). The fitting has been
carried out at various adsorption temperatures ranging from
25 �C to 65 �C. The numerical values of the activation energy
and pre-exponential coefficient for ACP/HFC-32 and ACF/HFC-
32 have been estimated.
The fitting shows that both LDF and FD models could be
effectively used to simulate the adsorption kinetics of ACP/
HFC-32 and ACF/HFC-32 where the differences between
calculated and measured values are within ±5%.
r e f e r e n c e s
Alam, K.C.A., Saha, B.B., Akisawa, A., 2013. Adsorption coolingdriven by solar collector: a case study for Tokyo solar data.Appl. Therm. Eng. 50 (2), 1603e1609.
Askalany, A.A., Salem, M., Ismael, I.M., Ali, A.H.H., Morsy, M.G.,2012. A review on adsorption cooling systems with adsorbentcarbon. Renew. Sustain. Energy Rev. 16, 493e500.
Askalany, A.A., Saha, B.B., Ahmed, M.S., Ismail, I.M., 2013a.Adsorption cooling system employing granular activatedcarboneR134a pair for renewable energy applications. Int. J.Refrigeration 36 (3), 1037e1044.
Askalany, A.A., Saha, B.B., Uddin, K., Miyzaki, T., Koyama, S.,Srinivasan, K., Ismail, I.M., 2013b. Adsorptionisotherms and heat of adsorption of difluoromethaneon activated carbons. J. Chem. Eng. Data 58 (10),2828e2834.
Atakan, A., Fueldner, G., Munz, G., Henninger, S., Tatlier, M., 2013.Adsorption kinetics and isotherms of zeolite coatings directlycrystallized on fibrous plates for heat pump applications.Appl. Therm. Eng. 58 (1e2), 273e280.
Chan, K.C., Chao, C.Y.H., Sze-To, G.N., Hui, K.S., 2012.Performance predictions for a new zeolite 13X/CaCl2composite adsorbent for adsorption cooling systems. Int. J.Heat. Mass Transf. 55 (11e12), 3214e3224.
Dabler, I., Mittelbach, W., 2012. Solar cooling with adsorptionchillers. Energy Procedia 30, 921e929.
Dawoud, B., 2013. Water vapor adsorption kinetics on small andfull scale zeolite coated adsorbers; A comparison. Appl.Therm. Eng. 50 (2), 1645e1651.
El-Sharkawy, I.I., Saha, B.B., Koyama, S., Ng, K.C., 2006. Astudy on the kinetics of ethanol-activated carbon fiber:theory and experiments. Int. J. Heat. Mass Transf. 49,3104e3110.
Glaznev, I.S., Aristov, Y.I., 2008. Kinetics of water adsorption onloose grains of SWS-1L under isobaric stages of adsorptionheat pumps: the effect of residual air. Int. J. Heat. Mass Transf.51 (25e26), 5823e5827.
Gordeeva, L., Aristov, Y., 2014. Dynamic study of methanoladsorption on activated carbon ACM-35.4 for enhancing thespecific cooling power of adsorptive chillers. Appl. Energy 117,127e133.
Habib, K., Saha, B.B., Rahman, K.A., Chakraborty, A., Koyama, S.,Ng, K.C., 2010. Experimental study on adsorption kinetics ofactivated carbon/R134a and activated carbon/R507A pairs. Int.J. Refrigeration 33, 706e713.
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 4 9 ( 2 0 1 5 ) 1 6 0e1 6 8168
Khattab, N.M., Sharawy, H., Helmy, M., 2012. Development ofnovel solar adsorption cooling tube. Energy Procedia 18,709e714.
Ng, K.C., Thu, K., Kim, Y., Chakraborty, A., Amy, G., 2013.Adsorption desalination: an emerging low-cost thermaldesalination method. Desalination 308, 161e179.
Ruthven, D.M., 1984. Principles of Adsorption and AdsorptionProcesses. John Willy & Sons, Inc, pp. 128e130.
Sharafian, A., Bahrami, M., 2014. Assessment of adsorber beddesigns in waste-heat driven adsorption cooling systems for
vehicle air conditioning and refrigeration. Renew. Sustain.Energy Rev. 30, 440e451.
Sircar, S., Hufton, J.R., 2000. Why does the linear driving forcemodel for adsorption kinetics work? Adsorption 6,137e147.
Tso, C.Y., Chao, C.Y.H., 2012. Activated carbon, silica-gel andcalcium chloride composite adsorbents for energy efficientsolar adsorption cooling and dehumidification systems. Int. J.Refrigeration 35 (6), 1626e1638.