Microporous and Mesoporous Materials 296 (2020) 109999

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

journal homepage: http://www.elsevier.com/locate/micromeso

Hydration mechanism in Ce-exchanged zeolites and heat release

performances upon adsorption of water vapour in support of their potential
use in thermochemical storage of energy under mild conditions of
adsorbent regeneration and saturation
Hao Wu, Philippe Trens, Bernard Fraisse, Fabrice Salles *, Jerzy Zajac **
ICGM, Universit�e Montpellier, CNRS, ENSCM, Montpellier, France

A R T I C L E I N F O A B S T R A C T

Keywords: The potential use of commercially available 13X zeolite, modified by ion-exchange with cerium compensating
13X zeolite cations possessing high charges and high hydration energies, has been tested in view of low-temperature storage
Cation exchange of solar energy performed under mild operating conditions of low regeneration temperatures and low pressures
Trivalent and tetravalent cerium
of water vapour during the adsorption step. Structural and textural properties, sorption behaviour towards water
Water vapour adsorption
Cation distribution
vapour of three selected samples containing various proportions of Ce3þ and Ce4þ compensating cations and the
Adsorption isotherms pristine Naþ-13X zeolite were studied by a variety of experimental techniques including Wavelength Dispersive
Heats of adsorption X-Ray Fluorescence, Energy Dispersive X-ray Spectroscopy, X-ray diffraction, Thermogravimetric analysis, as
Gas flow calorimetry well as measurements of the adsorption of gaseous nitrogen at 77 K and water vapour at 313 K. Based on the
Monte Carlo simulations structure refinement procedure applied to the experimental XRD patterns, it was demonstrated that extra-
Thermochemical energy storage framework cerium cations were preferentially located on sites I’ and II in dry and hydrated zeolites, showing
Mild adsorbent regeneration and saturation relatively little dependence on the hydration level. Monte Carlo simulations were used to determine the limit
values of the amount adsorbed and differential heat of adsorption, which could be obtained experimentally if the
zeolite samples were completely dried. The potential of Ce-containing zeolites as adsorbents for the thermo­
chemical energy storage was finally determined under flow conditions by firstly dehydrating samples at 353 K or
423 K and then saturating them at 296 K with water vapour at a mole fraction of 0.03. The choice of the
operating conditions was decided so as to maintain the stability of the zeolite structure while taking the risk of
reduced thermal performance of zeolite adsorbents undergoing incomplete regeneration-dehydration. Under
such mild conditions, the modified 13X zeolites exhibited enhanced thermal performance in comparison with
that of the pristine 13X, by releasing between 700 and 1100 kJ per kg of the adsorbent during a period of 6–8 h.
Through a complementary study based on calorimetry measurements and molecular simulations, the under­
standing of the hydration-dehydration steps in Ce-exchanged zeolites and cation displacement upon hydration
has allowed to establish the best compromise for the conditions of zeolite regeneration and saturation in the case
of heat long-term storage applications.

1. Introduction key issue for sustainable development and climate change, especially in
the building sector which is responsible for approximately 40% of the
The rapidly increasing demand for energy, combined with the energy consumed in the EU countries resulting in significant greenhouse
decrease of fossil fuel resources and growing greenhouse emissions, calls gases emissions [1]. In attempts to circumvent issues due to the inter­
for the continued development of ambitious alternatives to move to­ mittent nature of the renewable energy, many studies have focused on
wards a sustainable, renewable and environment-friendly way to pro­ converting it into more useful forms such as thermal energy or electricity
vide energy. Nowadays, renewable energy sources are considered as a [2–6]. Amongst the numerous possibilities for energy conversion, solar

* Corresponding author.
** Corresponding author.
E-mail addresses: fabrice.salles@umontpellier.fr (F. Salles), jerzy.zajac@umontpellier.fr (J. Zajac).

https://doi.org/10.1016/j.micromeso.2020.109999
Received 23 October 2019; Received in revised form 17 December 2019; Accepted 2 January 2020
Available online 3 January 2020
1387-1811/© 2020 Elsevier Inc. All rights reserved.
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

energy harvested in abundance during summer may be exploited to increase in the heat release when adsorbing water vapour at low hu­
increase the internal energy of an appropriate storage material (i.e., midity contents was predicted on this basis. Despite a decrease in the
charging step) and thus be stored until cold days in winter when the number of multivalent cations necessary to compensate the negative
heating supply system is under pressure. Converted subsequently into charge of the zeolite framework, the net energy gain was justified by a
heat during the discharging step, it may be used as a supplemental source non-linear increase in the molar enthalpy of hydration when compared
of space heating [6]. with the initial Naþ counter-ion; e.g., 6325 kJ mol 1, Ce4þ; 3365 kJ
In view of future uses in low-temperature thermochemical storage of mol 1, Ce3þ; 415 kJ mol 1, Naþ, according to Ref. [30]. Furthermore,
solar energy, solid materials capable of releasing much heat during the capacity of cerium cations to adapt their electronic configuration to
adsorption of vapours have the potential to yield high storage densities their local environment underlying their particular redox properties was
[5,6]. With the open sorption systems operating in the moist-air flow also taken into account [32]. Multi-scale experimental and modelling
mode, the heat released upon the adsorption of water vapour may serve studies, including adsorption and batch calorimetry measurements,
to increase the temperature of the dry air leaving the reactor. The warm X-ray diffraction, and Monte Carlo simulations, were performed to
gas is subsequently directed to the space heating loop of the heated elucidate the hydration mechanism occurring in the Ce-exchanged
building. Various kinds of solid materials with a propensity for being zeolite samples, as well as to optimise the regeneration conditions.
reversibly dried and hydrated, and additionally characterised by good The motion of the extra-framework cations in the structure was moni­
hydrothermal stability, have been already tested as adsorbents for the tored by coupling X-Ray diffraction patterns with structural resolution
thermochemical storage (TCS) uses [6–16]. Porous zeolites and and Monte Carlo simulations. Moreover, the heat release performances
zeolite-based structures commonly stand out from the others as of the selected samples, in terms of the integral heat of adsorption and
combining high specific surface areas, pronounced hydrophilic char­ the kinetics of the heat evolvement, were evaluated based on the results
acter of their surface, and good thermal/hydrothermal stability [17,18]. of calorimetry measurements carried out with the aid of a gas flow
By the way, zeolite 13X was the first adsorbent practically applied in a microcalorimeter operating in the saturation mode. In the light of these
short-term low-temperature thermochemical storage for daily heating of results, the use of cerium-exchanged zeolites as adsorbents for the
a school building in Germany [19,20]. Nevertheless, the handling of low-temperature thermochemical storage of energy by sorption was
zeolites as effective adsorbents of water vapour may pose some practical discussed.
problems [6].
Firstly, when the zeolite-water affinity is strong even at low humidity 2. Experimental section
levels, the adsorption isotherm is similar to the Type I curve possessing a
very steep initial portion. In order to ensure the complete regeneration 2.1. Materials
of the adsorbent after each discharging-hydration stage, harsh drying
conditions are necessary (e.g., deep vacuum treatment at high temper­ High purity (ACS reagent grade, assay � 99.0%) cerium(III) nitrate
atures). Furthermore, the use of an almost perfectly tight tank is imposed hexahydrate, Ce(NO3)3.6H2O, and ammonium sulfate, (NH4)2SO4, were
on the storage of the dried adsorbent between the successive charging purchased from Sigma-Aldrich and employed without further purifica­
and discharging steps so as to avoid the presence of water traces inside tion. The 13X zeolite powder with a primary particle size of around 2 μm
it. Otherwise, the adsorbent will be deactivated in an uncontrolled used in the present study was also a Sigma-Aldrich product. According to
manner before the next discharging. Secondly, 13X-type zeolites subject the technical specification, it contained only Naþ as a compensating
to a great number of repeated drying-hydration cycles even at somewhat cation. The 18.2 MΩ cm ultrapure water (PURELAB® Chorus 1, ELGA
lower temperatures appear to undergo a progressive amorphisation and Veolia) was used in the synthesis and ion exchange procedures, as well
the concomitant evolution of the pore structure [21]. as in the adsorption and calorimetry measurements. The gas adsorption
One possible solution to this issue could be to modulate the hydro­ and calorimetry experiments were carried out by making use of
philic character of zeolite surface either by cation exchange or to modify ALPHAGAZ 2 grade helium and nitrogen (Air Liquide, France).
the substitution ratio [22]. For example, it was proven that the nature of
extra-framework cations had a strong influence on the hydration energy 2.2. Preparation of ion-exchanged zeolites
in porous solids [23–27]. Going further in this direction, commercially
available 13X samples, containing mostly Naþ as extra-framework Three ion-exchanged 13X zeolite samples were prepared in the
compensating cations and subsequently modified by a partial ion ex­ following way.
change with divalent Ca and Mg cations, were demonstrated to yield A stock 0.05 mol L 1 aqueous solution of Ce3þ was prepared by
interestingly enough amounts of heat during the adsorption of water dissolving the Ce(NO3)3.6H2O powder in ultrapure water. After the
vapour if the activation procedure had been performed at 473 K [27]. addition of 1 g of 13X zeolite to 50 mL of this solution, the suspension
However, the migration of extra-framework cations among various was firstly stirred at 338 K for 3 h and then centrifuged for 10 min at
crystallographic sites, which accompanies hydration and dehydration 11000 rpm. The solid sample was subsequently separated from the su­
sequences [28,29], may be undesirable for long-term control of the TCS pernatant liquid. This procedure was repeated twice more, with the
system. suspension stirring step being extended to 24 h upon the last cycle. Then,
The main objective of the present work was to design a cation- the resulting solid sample was shaken again with 50 mL of ultrapure
exchanged 13X zeolite with enhanced heat release characteristics, thus water at room temperature for 2 h and centrifuged in the same way as
allowing avoiding the regeneration cycle to be performed under harsh before. The whole procedure was repeated 5 times, with the final stirring
conditions. Therefore, the main challenge was to increase the heat being prolonged to 24 h. The solid was finally dehydrated by freeze-
evolved upon the adsorption of water vapour onto a zeolite structure drying (lyophilisation) under vacuum in a Cryonext freeze-drier and
that had not been completely dehydrated in the preceding desorption stored in unsealed vials. The sample achieved in this way was denoted as
step. The migration of compensating cations upon hydration-drying Ce1-13X.
stages to be reduced as much as possible was also targeted in the pre­ In order to maximise the exchange ratio, the pristine 13X sample was
sent work. By referring to the research strategy developed previously first washed with ultrapure water and a preliminary exchange step was
[27], similar commercially available 13X sample was modified by then performed for 2 h by using 50 mL of aqueous (NH4)2SO4 solution.
substituting the pristine extra-framework ions with various cerium cat­ The aim here was to exchange the primary Naþ cations for NHþ 4 ones and
ions. The choice of cerium was dictated by the fact that this rare earth thus facilitate further exchange with cerium. Similar exchange proced­
metal could form both Ce3þ and Ce4þ cations possessing high energies of ure was carried out twice making use of 50 mL of solution containing
hydration in aqueous solutions [30,31]. Therefore, the expected Ce3þ at a concentration of 0.005 mol L 1. The purification and drying

2
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

procedures were the same as before. The resulting sample was named as the aim of determining the extra-framework positions of charge-
Ce2-13X. compensating cations. The recorded XRD patterns were refined by
An additional zeolite sample containing a great quantity of Ce4þ as a using Le Bail method and the SUPERFLIP software to solve the crystal
compensating cation was also prepared. According to Hunter [33], structure of Ce-exchanged zeolites at different hydration states [35].
oxidation of extra-framework Ce3þ to Ce4þ was possible upon calcina­ Furthermore, the JANA2006 software was employed to refine the crystal
tion of the zeolite structure at high temperatures in air. For this purpose, structure [36] and specify the location of cerium cations.
a part of the Ce2-13X sample was heated at 723 K in air during 12 h. As a
result, a change of colour was observed from white to pale yellow thus 2.4. Measurements of the adsorption of water vapour in various
indicating the oxidation of Ce(III) to Ce(IV). This sample was referred to experimental regimes
as Ce3-13X.
The adsorption of water vapour under static equilibrium conditions
2.3. Zeolite sample characterization was studied at 313 K by means of a home-made apparatus previously
described in the literature [37]. Prior to the adsorption measurements,
The chemical formulas of the initial 13X and Ce3þ-exchanged 13X the zeolite samples underwent an initial thermal treatment at 423 K
samples were determined with the aid of Wavelength Dispersive X-Ray under a vacuum pressure of 10 5 Torr. For the purpose of the experi­
Fluorescence (WDXRF) spectrometer (Axios Max, PANalytical) and En­ ment, the temperature of the thermoregulated chamber was set to 313 K
ergy Dispersive X-ray Spectroscopy (EDX) on the scanning electron and an adsorption curve was determined based on the manome­
microscope (SEM) (FEI Quanta 200 FEG). Since it was not possible to tric/volumetric measuring principles. More precisely, the vapour pres­
find the proportion between Ce3þ and Ce4þ in Ce3-13X, the chemical sure was measured before and after the sorption step. Assuming the
formula of this sample remained undetermined. vapour as a perfect gas, the difference in pressure induced by sorption
Textural properties of the solids were characterized based on the could be related to the adsorbed amount, after a careful determination of
gaseous nitrogen adsorption isotherms measured at 77 K using a the sample cell volume. One particular interest of the present sorption
micromeritics ASAP2010 sorption equipment. Samples were degassed at device was the possibility to set the initial pressures rather than the
473 K for 12 h under secondary vacuum prior to adsorption measure­ equilibrium ones. This possibility allowed the sample itself to define the
ments. The apparent specific surface areas were then calculated on the number of data in the sorption isotherm. Indeed, in the next sorption
basis of the BET model by taking 0.162 nm2 as a cross sectional area for step, the initial pressure was calculated by adding the equilibrium
one adsorbed N2 molecule [34]. pressure measured in the former step to a defined vapour pressure [38].
Some characteristics of the pristine 13X zeolite and cerium- The uncertainties of the measurements could be evaluated by taking
exchanged samples tested in the present study are given in Table 1. into account the slight scattering of the data in the sorption traces.
Thermogravimetric analysis (TGA) was performed on a NETZSCH However, the adsorbed amounts at saturation (see further sections) were
Jupiter STA 449, under an argon flow of 50 mL min 1. Samples were consistent with the previous results reported on similar materials [39].
heated from 298 K to 823 K in an alumina crucible by using a heating The use of a Calvet-type differential and isothermal microcalorimeter
ramp of 5 � C min 1. (Setaram instrumentation C80) connected to the static volumetric rig
The resistance of the zeolite structure to cation exchange was allowed the differential molar heat of adsorption to be additionally
monitored by X-ray diffraction (XRD) at room temperature when measured [38].
exposing powder samples to Cu-Kα X-ray radiation with the aid of Philips The heat release performances of the zeolite samples were also
X’Pert PRO MPD equipped with an X’Celerator Scientific detector using evaluated under dynamic conditions by applying a 4 V ms Microscal
0.04 rad Soller slits. The same radiation source was used in the in situ flow microcalorimeter operating in a saturation mode [40]. Helium was
XRD analysis at high temperatures on a Panalytical Empyrean diffrac­ used as a carrier gas at a flow rate of 2 mL min 1. Each adsorbent was
tometer with Pixcel3D detector, using 0.02 rad Soller slits to prevent pre-activated in a flow of argon at 423 K during 24 h in a separate
axial divergence and thus improve peak shape and symmetry. This XRD degassing rig outside the calorimeter. A weighted solid sample (about
equipment included an Anton Paar HTK16 chamber with platinum 80 mg) was introduced to fill a 50 μL calorimetric cell. The sample was
heating filament, which offered the possibility of working in static air subsequently flushed with helium to reach the initial thermal equilib­
atmosphere at various temperatures. The zeolite particles were sus­ rium. An in-line mounted saturation vessel filled with ultrapure water
pended in a few drops of water and deposited on a 10% rhodium- and thermostated by a water bath was used to continuously saturate the
platinum slide to form a thin and uniform layer at the end of water flowing helium with water vapour at 296 K. After having attained a
evaporation process. The subsequent desorption of water from the sur­ steady thermal baseline, the flow of helium containing water vapour at a
face was achieved by starting with a completely hydrated sample at partial pressure of 2.8 kPa was directed to the calorimetric cell still at a
room temperature, firstly, and then by increasing the temperature step constant flow rate of 2 mL min 1. The complete saturation of the zeolite
by step. This facility was used to obtain the XRD patterns for Ce3þ- sample with water vapour was decided on the basis of the return of the
containing samples successively at 303, 353, 423, 573, and 723 K, thus thermal signal to baseline. Calibration of the areas under the thermal
simulating different hydration states. The diffraction data were quanti­ profile was carried out by dissipating a known amount of energy in the
tatively analysed using the Rietveld refinement approach mostly with cell, with the aid of a calibration probe placed directly within the
adsorbent bed, under the same flowing conditions and integrating the
Table 1 related exothermic peaks. The total amount of heat released during each
Chemical formula, atomic Si:Al ratio, degree of cation exchange, and apparent saturation run corresponded to the integral heat of adsorption. After the
specific surface area for the zeolite samples studied. saturation stage, the adsorbent was degassed under a constant flow of
helium. This regeneration step lasted 1 day and night, with the adsor­
Sample Chemical formula Si: Exchange Sap (m2
Al ratio (%) g 1) bent being heated at 353 K during 24 h and then cooled down to room
temperature overnight. Altogether, three degassing-saturation cycles
13X Na87,8(SiO2)108,9(AlO2)83,1, xH2O 1.27 – 519
Ce1- Na10,6Ce24,2(SiO2)109,6(AlO2)82,4, 1.33 88 426 were performed for each zeolite sample.
13X xH2O
Ce2- Na2,6Ce27,4(SiO2)109(AlO2)83, xH2O 1.31 97 661 2.5. Computational section
13X

N.B. An apparent specific surface area of 578 m2 g 1

was determined for Ce3- Grand Canonical Monte Carlo (GCMC) simulations were performed
13X using a home-made code to localize the position of extra-framework

3
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

cations and evaluate the sorption properties of the investigated zeolite

structures. To reproduce the 13X zeolite structure, the space group Fd3m
with a unit cell parameter of 24.85 � 0.03 Å was considered [41]. In the
unit cell, 16 hexagonal prisms, 8 sodalite cages, and 8 supercages
composed the structure, thus resulting in a general chemical formula of
Mx/nAlxSi192 xO384⋅yH2O, with Mnþ being a given compensating cation:
x varied from 0 to 96, y corresponded to the number of water molecules
in the unit cell. In the present study, the atomic Si/Al ratio ranged be­
tween 1.0 and 1.5 and a rigid framework was considered; i.e., 88 sub­
stitutions of Si4þ by Al3þ distributed randomly following the
Loewenstein’s rule.
In order to predict the adsorption enthalpies and isotherms, a stan­
dard methodology was used based on the equilibrium of chemical po­
tentials between the adsorbed phase within the pores and the gaseous
phase at a given temperature [42]. In the present case, temperature,
chemical potentials for water vapour and unit cell volume were kept as
fixed. The interactions between adsorbate molecules and the solid were
then calculated by using 12-6 Lennard Jones and electrostatic potentials.
Fig. 1. Room temperature X-ray diffraction patterns recorded for the pristine
During the GCMC simulations, the 3D periodic boundary conditions
13X, Ce1-13X, Ce2-13X, and Ce3-13X, as defined in the Experimental section.
were applied, while the motion of water molecules included translation,
rotation, deletion and insertion randomly considered during the equil­
cerium nitrate particles are irreversibly trapped within some pores of the
ibration steps.
zeolite. Certainly, the chemical formula and the ion exchange ratio re­
Computations were performed at 298 K using 107 Monte Carlo pro­
ported in Table 1 are no longer valid. This sample with less defined
duction steps following 107 equilibration steps. The partial atomic
characteristics is still of interest in view of the present application,
charges (see Table S1 in Supplementary Material) and the 12-6 Lennard-
especially when comparing with Ce2-13X characterised by an exchange
Jones (LJ) atomic parameters for the framework were taken from the
ratio close to 100%.
literature [27] and the UFF respectively, while formal charges and UFF
were used for the extra-framework cations. Water molecules were
considered as rigid by using the four-site charged LJ TIP4P/2005 model 3.1. Effects of temperature and hydration state
[27].
The parameters of the adsorbate/adsorbent LJ interatomic potential In the second stage, the effect of temperature and hydration state on
were then calculated using the Lorentz-Berthelot combining rule. A cut- the crystallographic positions of the atoms within the framework, as
off radius of 12 Å was applied to all LJ interactions, and the long-range well as on the positions of the extra-framework cations was investigated.
electrostatic interactions were handled by applying the summation Several characteristic temperatures were selected in relation with the
Ewald technique. In these calculations, extra-framework cations were operating conditions under which zeolites 13X had been previously
allowed to move within the porous structure. tested in the context of thermochemical storage of energy [3], as well as
under which their dehydration or oxidation of Ce3þ to Ce4þ had been
3. Results and discussion studied [33,44]. An example of in situ XRD patterns recorded at these
different temperatures and hydration states is given in Fig. 2.
Given the higher valence of cerium species and some changes of the The X-ray diffractograms recorded on Ce1-13X have been chosen for
textural parameters reported in Table 1, the question arises as to illustrative purposes since this sample contains additional peaks
whether the zeolite structure has been preserved after ion exchange with compared to the others. For example, it is worth noting that the addi­
Ce3þ cations and also after oxidation from Ce3þ to Ce4þ. For this pur­ tional peak of bulk cerium nitrate is to a very great extent reduced after
pose, the three ion-exchanged zeolite samples and the pristine 13X heating near and above 423 K, which supports well the previously for­
material were studied by XRD. Fig. 1 shows the experimental XRD warded hypothesis of its location outside the zeolite framework.
patterns recorded for all samples at room temperature. Changes in the intensity of other diffraction peaks recorded at different
Despite some changes in the relative intensity of the peaks observed temperatures indicate that the electronic density has been modified
in the XRD patterns in Fig. 1, the number and the positions of these peaks when raising the temperature. This modification of the electronic den­
remain essentially the same. It can be thus concluded that ion exchange sity can be attributed to changing positions of the extra-framework
in the 13X faujasite does not induce great modification of the zeolite cations and progressive removal of the adsorbed water molecules.
structure. This result supports the hypothesis that the structure is close Indeed, an increase in the temperature induces the passage of the zeolite
to be rigid and behaves in this manner upon ion exchange, as has been structure to a less hydrated state because of the enhanced evaporation of
already reported in the literature [43]. Furthermore, the number and water. Such a decreasing loading of water inside the pores leads to
valence of compensating cations have no impact on the structure since observe better crystallization and thus resulting in modification of both
the peak positions in the XRD patterns of samples containing Ce3þ and the shape and the intensity of the peaks (which also modifies the relative
Ce4þ are always the same. In all cases, the observed modification of the intensities in the whole diffractogram). This in turn may affect the
peak intensity can be ascribed to the modification of the electronic mobility of the extra-framework cations [28,45]. It is important to note
density inside the pores. Nevertheless, the XRD pattern of Ce1-13X that the diffractograms recorded at 303 K in the beginning and at the end
presents an exception to the trends described above. Indeed, a new of the experiment are not strictly identical to each other. A net change in
peak can be seen at a 2θ value of 13.2� . The comparison with the XRD colour of the recovered sample additionally argues in favour of the
pattern of the bulk Ce(NO3)3.4H2O (c.f., Fig. S1 in Supplementary Ma­ irreversible oxidation of extra-framework Ce3þ cations, which has
terial) provides a plausible argument that some quantity of non-reacted occurred after complete dehydration of Ce1-13X and its heating likely at
cerium salt has remained after ion exchange and purification proced­ 723 K in the static air atmosphere inside the XRD chamber.
ures. Since a marked decrease in the apparent specific surface area has Based on the XRD patterns recorded at various temperatures for ze­
been inferred from the measurements of the gaseous N2 adsorption olites exchanged with Ce3þ, the Rietveld method was used for the
isotherm for this sample, it seems reasonable to conclude that such refinement of the zeolite structural model. In consequence, it was

4
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

Fig. 2. In situ XRD patterns recorded at different temperatures corresponding to a variety of hydration states for Ce1-13X. At the final state, the sample is cooled back
to 303 K, after having been heated to 723 K.

possible to analyse the location of extra-framework cations and to According to the results reported in Fig. S5 and in Table 2, nearly
compare them with the theoretical positions inferred from Monte Carlo complete regeneration of adsorbents is possible only when the temper­
simulations. For computational purposes, two model zeolite structures ature is at least equal to 573 K. In consequence, the results of structure
saturated with Ce3þ and Ce4þ were considered, namely Ce3þ-13X and refinement at this temperature will be confronted with those obtained
Ce4þ-13X. Some detailed results are reported in Supplementary Mate­ for the model Ce3þ-13X sample at the dry state. By analogy, the sample
rial: Fig. S2 shows various crystallographic sites occupied by extra- heated at 723 K will be regarded as simulating the dry state of 13X
framework cations in 13X zeolites; typical snapshots of the unit cell in saturated with Ce4þ cations.
the Ce-exchanged zeolites at different temperatures are given in Fig. S3 The positions of the extra-framework cations as inferred from
(structure refinement procedure), whereas those of the model Ce3þ-13X structure refinement and Monte Carlo simulations are reported in
and Ce4þ-13X in dry and hydrated states are reproduced in Fig. S4 Table 3.
(Monte Carlo simulations). Taking into account the results obtained from the structure refine­
For reasons of comparison between the experimental and theoretical ment, the type I0 and II sites represent the preferential positions for
results and in order to ensure the equivalence of various hydrated states, extra-framework cations in dry and hydrated zeolites. Furthermore, this
mass loss profiles of the zeolite samples studied as a function of the method does not reveal much difference between the hydrated and dry
temperature, as obtained by means of thermogravimetric analysis states: a systematic shift from sites II towards sites I0 is observed when
(TGA), are given in Fig. S5 in Supplementary Material. The use of air has
been avoided during the analysis period so as not to oxidize the
Table 3
compensating Ce3þ ions. The mass loss stops somewhat before 600 K in
Positions of the extra-framework cations in the Ce-exchanged zeolite samples as
the case of 13X, whereas it still continues, although to a much lesser
inferred from structure refinement based on the experimental XRD patterns and
extent, above this temperature for Ce-containing samples. Much higher Monte Carlo simulations. Various crystallographic sites occupied by extra-
energies of cation hydration in aqueous solutions for Ce3þ and Ce4þ in framework cations in 13X zeolites are specified in Fig. S2 in Supplementary
comparison with that of Naþ can be taken into account to explain this Material. The occupation rate on a given crystallographic site or within
trend, since higher temperatures will be necessary to remove last water supercages (SC) is given by a number before the symbol of the site. Real ad­
molecules from the hydration shells surrounding the cerium compen­ sorbents and models appearing on the same line are regarded as corresponding
sating cations in the zeolite structures. Table 2 shows the percentage of to similar hydration states.
mass loss measured until 3 temperatures envisaged for adsorbent Structure refinement Monte Carlo simulation
regeneration.
Temp. Dominant Sites Model Hydration Sites
(K) cation state

303 Ce3þ 4 I’ þ Ce3þ- hydrated 6 I’ þ 9 II þ 9

Table 2
22 II 13X III þ 5 SC
Percentage of mass loss monitored up to a given temperature for the zeolite 353 Ce3þ 4 I’ þ
samples studied. These values were calculated on the basis of thermogravimetric 22 II
analysis (the corresponding TGA curves are reported in Fig. S5 in Supplementary 423 Ce3þ 13 I’ þ
Material). 11 II
573 Ce3þ 20 I’ þ Ce3þ- dry 7 (I þ I0 ) þ 16
Sample 13X Ce1-13X Ce2-13X Ce3-13X
5 II 13X II þ 6 III0
Temperature (K) Mass loss (%) 723 Ce4þ 31 I0 Ce4þ- dry 8 (I þ I0 ) þ 9
353 5 5 4 5 13X II þ 5 III0
423 15 15 13 15 303 Ce4þ 31 I0 Ce4þ- hydrated 6 I’ þ 3 II þ 9
573 25 23 22 22 13X III’ þ 4 SC

5
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

the hydration degree decreases. When the sample is heated up to 353 K, the temperatures applied during the charging and discharging stages.
the sample retains much water and the positions of Ce3þ cations do not For the purpose of the present study, the choice of these temperatures
change at all. Increasing further the temperature to 423 K already puts was dictated by the need to be close to the operating conditions
the sample in a state approaching the dry one. As a first important frequently tested in a thermochemical storage with the use of 13X zeo­
conclusion drawn from these observations, the hydration phenomenon lites [3]. Therefore, the experimental adsorption isotherms were
has rather little effect on the location of extra-framework cations. This measured at 313 K after having submitted the samples to a vacuum
can be ascribed to the high charges of the Ce3þ and Ce4þ compensating treatment at 423 K. The quantity of water adsorption was plotted as a
cations, thus making their electrostatic interactions with the zeolite function of the relative vapour pressure (the saturation vapour pressure
framework much stronger. The lack of pronounced hydration effect, for water at 313 K is 7.4 kPa). In order to gain an idea about the
being particularly pronounced in the case of 13X zeolite containing Ce4þ maximum adsorption capacity, the theoretical adsorption curves were
cations, is of high importance in the context of the use of these adsor­ calculated for the model Naþ-13X, Ce3þ-13X and Ce4þ-13X samples by
bents in the thermochemical energy storage. These results are in fairly referring to Monte Carlo simulations performed at 298 K. The experi­
good agreement with the preferential crystallographic positions inferred mental and simulated adsorption isotherms have been collected in
from Monte Carlo simulations performed on 13X zeolite structures Fig. 3.
saturated with Ce3þ and Ce4þ compensating ions. Here the type I’, II, The comparison between the experimental and simulated adsorption
and III sites have been identified as the main adsorption centres. It is isotherms reveals two opposite trends with the type and proportion of
worth noting that similar cation distributions have been already re­ the extra-framework compensating cation. Two portions of the adsorp­
ported for zeolites containing Naþ, Csþ, and Co2þ cations in the dry state tion curves particularly deserve to be analysed: the initial segment
[41,46–49]. Some cerium cations appear to be located close to the (almost linear) at very low relative pressures, which gives an idea about
supercages in the hydrated samples. It may be argued that the high the affinity of the adsorbate for the adsorbent surface, and the one at
energy of supercage hydration is responsible for the migration of the higher relative pressures, providing an estimate of the total adsorption
extra-framework cations initially positioned close to the supercages to­ capacity of a given sample. On the experimental side, Ce2-13X (con­
wards the positions localised inside these supercages. Note that Naþ taining mostly Ce3þ as a compensating cation) and Ce3-13X (containing
compensating cations were demonstrated to occupy crystallographic some proportion of Ce4þ) adsorb less water vapour than the pristine 13X
sites depending to a small extent on the hydration or dry state of the sample (sodium form) at various relative pressures. Since the initial
zeolite samples [47]. It should be also remembered that the repartition slopes are very similar to one another being almost vertical, it is difficult
of the aluminium moieties within the framework can strongly influence to precisely identify which material would generate stronger in­
the distribution of the extra-framework cations. teractions with water molecules. This is confirmed by the enthalpy
curves shown in the inset of Fig. 3. The curve extrapolation to zero
surface coverage provides quite similar heat values within experimental
3.2. Sorption properties of the Ce-exchanged zeolite sample error.
Furthermore, the two cerium-exchanged samples are characterised
Altogether, the molecular simulations have proven useful for eluci­ by similar adsorption capacities. Surprisingly, there is little difference in
dating the mechanism of water adsorption onto cerium-exchanged ze­ the quantity of adsorption between 13X and Ce1-13X, with the excep­
olites and the resulting redistribution of the compensating cations tion of the high pressure range where the latter adsorbs more water
within the pores, in line with the earlier arguments advanced in the vapour. To better differentiate between these two samples with respect
literature [50]. Once the pertinent structural properties of the to their affinity towards water, the variations of the differential molar
Ce-exchanged zeolite samples had been determined, further studies heat of adsorption in the range of small quantities of water adsorption
were necessary to estimate the extent and energetics of adsorption onto (up to about 7 wt%) are additionally illustrated in Fig. 3. The differential
these adsorbents. heat, being of the order of 100 kJ per mole of water in the very
The adsorption capacity towards water vapour obviously depends on

Fig. 3. The experimental adsorption isotherms for

water vapour onto various zeolite samples measured
at 313 K after having submitted the samples to a
relatively mild vacuum treatment at 423 K (left panel)
and the theoretical adsorption curves for model
zeolite samples obtained at 298 K by Monte Carlo
simulations (right panel). The experimental and
theoretical quantities of adsorption are presented here
using two different scales. The inset within the left
panel shows the variations of the differential molar
heat of adsorption as a function of the water uptake
for 13X and Ce1-13X in the initial adsorption range;
the horizontal line represents the molar heat of water
evaporation.

6
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

beginning, decreases quickly with increasing adsorption values, in line (adsorption) and increasing the regeneration temperature from 393 to
with the energetic heterogeneity of the adsorption sites which are 423 K.
occupied by water molecules under the experimental conditions applied Gas flow calorimeter was used here as a measuring tool to determine
here. the integral heat of adsorption when the solid sample placed in the
On the basis of simulated adsorption isotherms, the adsorption ca­ calorimetric cell was saturated with water vapour at a partial pressure of
pacity of the zeolite samples should move in the opposite direction: Naþ- 2.8 kPa (i.e., mole fraction of 0.03). Three dehydration-saturation cycles
13X < Ce3þ-13X < Ce4þ-13X. For very low relative pressures, the were performed for each zeolite sample under a constant flow of helium,
theoretical adsorption curves are superimposed on one another. The either pure or containing water vapour, at a flow rate of 2 mL min 1.
molar enthalpy of adsorption for the first water molecule was thus Prior to the first saturation-degassing cycle, the adsorbent was activated
calculated with the aid of Monte Carlo simulations to gain an idea about by a vacuum treatment at 423 K. The resulting heat values are presented
the surface affinity towards water vapour. The following enthalpy values in the form of histogram bars in Fig. 4, whereas the corresponding heat
were obtained: 90 kJ mol 1, Naþ-13X; 320 kJ mol 1, Ce3þ-13X; release profiles are given in Fig. S6 in Supplementary Material.
400 kJ mol 1, Ce4þ-13X. It is clear from Fig. 4 that the cerium-exchanged samples are char­
What is most important here is the acknowledgement that the acterised by the enhanced heat release in comparison with the pristine
quantities of water adsorption predicted by Monte Carlo simulations are 13X sample. The difference is particularly pronounced for the first
much higher than those indicated by the adsorption experiments. To saturation cycle. Here the saturation of Ce3-13X by water vapour under
rationalise this finding, it is necessary to recall that the three model the experimental conditions applied yields more than 1100 kJ per kg of
samples simulate the behaviour of the perfectly dried zeolite frame­ the sample, whereas integral heats of 950 and 800 kJ kg 1 are reported
works neutralised by one type of extra-framework cation. According to for Ce2-13X and 13X, respectively. When the vacuum treatment at 423 K
the results reported in the previous section, the temperature of thermal is replaced by a flow of helium at 353 K, less energy is recovered upon
treatment applied (i.e., 423 K) is insufficient for complete adsorbent the second and the third cycles: about 700 kJ kg 1 for the three cerium-
dehydration and some water molecules are likely present in the hydra­ exchanged zeolites and 550–600 kJ kg 1 for 13X. This result emphasises
tion shells which surround the compensating cations in the zeolite again the importance of the regeneration temperature in the heat release
structures. In turn, the adsorption of such molecules would be more cycle.
exothermic. Nonetheless, since they finally remain, in different quanti­ In the case of cerium-exchanged samples, there is little difference
ties, in the samples treated at 423 K, the heat subsequently released between the integral heat values obtained in the second and third
during the adsorption of water vapour is markedly reduced compared to saturation stage, which means that the saturation-dehydration cycles
the theoretical heat amount that would have been obtained if the sam­ become more regular at the end. For 13X, a marked decrease continues
ples had been completely dehydrated. Another conclusion to be drawn during the three cycles. In comparison with the thermal performance of
from the comparison between the experimental and theoretical iso­ the adsorbents achieved in the first adsorption stage after a more energy-
therms with respect to Ce3-13X may be that this sample rather contains consuming regeneration, between 64 and 74% of the integral heat is still
a mixture of Ce4þ and Ce3þ compensating cations as a result of incom­ obtained in further dehydration-saturation cycles. This performance of
plete oxidation of Ce(III) to Ce(IV). There should be numerous Naþ Ce-exchanged zeolites is a good compromise when compared with the
cations still remaining in Ce1-13X after ion exchange, with much cerium case of Mg- or Ca-exchanged zeolites for which, respectively, 810 and
being involved in the Ce(NO3)3.4H2O phase located within the pores. 310 kJ per kg of each sample were obtained during the first hydration
Finally, lowering the adsorption temperature from 313 K (experiment) after an initial vacuum pre-treatment at 473 K [27]. For the latter zeo­
to 298 K (simulations) increases the amount of water vapour retained by lites, only 30–50% of the initial energy was recovered during next
the zeolite sample in line with the exothermic character of the adsorp­ saturation cycles preceded by a regeneration at 298 K.
tion phenomenon. To underline again the importance of the regeneration and adsorp­
tion conditions, the limit values of the integral heat of adsorption were
3.3. Thermal performances of cerium-exchanged zeolites under flow
conditions

The potential of Ce-exchanged zeolites as adsorbents for the ther­

mochemical energy storage was finally considered by taking advantage
of the results and conclusions presented in the previous sections. The
operating conditions during the laboratory testing were defined mostly
in line with the primary objective of the present study to reduce as much
as possible the temperature of the adsorbent regeneration after a hy­
dration stage performed at low pressures of water vapour. To minimize
the mobility of the compensating cations, while avoiding the uncon­
trolled oxidation of Ce3þ ions, and thus to strengthen the stability of the
zeolite structure submitted, in the industrial practice, to a large number
of repeated regeneration-saturation cycles, it was decided to decrease
the regeneration temperature down to 353 K (contrary to the tempera­
ture of 423 K applied previously in measurements under static condi­
tions of equilibrium); the adsorption of water vapour was carried out at
296 K. Studies of the effect of the regeneration temperature on the extent
and kinetics of water adsorption onto zeolite 13X from the gas phase
reported in the literature were also the motivation underlying this
Fig. 4. Integral heat of adsorption released during 3 consecutive saturation
change of the operating conditions. For example [44], it was shown that
cycles for the zeolite samples studied after a regeneration stage performed by a
changes in the regeneration temperature between 353 and 393 K had no flow of helium at 353 K. It is important to note that, prior to the first saturation-
significant impact neither on the quantity of water adsorption nor on the degassing cycle, the adsorbent was activated by a vacuum treatment at 423 K.
effective diffusivity, the latter being rather decreased with an increase in The experimental errors are marked as green vertical bars. (For interpretation
the adsorbate concentration; the upward trend in the adsorption ca­ of the references to colour in this figure legend, the reader is referred to the
pacity was observed when decreasing the temperature of discharging Web version of this article.)

7
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

assessed on the basis of the theoretical chemical compositions for two exchanged zeolite structure was hindered when oscillating between
model solids saturated with Ce3þ and Ce4þ compensating cations on one room temperature and 353 K. The oxidation of Ce3þ ions to Ce4þ ones by
side, and the hydration energies of these cations in aqueous solutions, on heating at 723 K in air was found to be irreversible.
the other side. The chemical formulas used in Monte Carlo simulations The potential of Ce-exchanged zeolites as adsorbents for the ther­
were as follows: Ce29NaSi104Al88O384 (molar mass ¼ 15515 g mol 1), mochemical energy storage was finally determined under flow condi­
Ce3þ-13X; Ce22Si104Al88O384 (molar mass ¼ 14512 g mol 1), Ce4þ-13X. tions by firstly dehydrating samples at 353 K and then saturating them at
By identifying the adsorption heat per 1 mol of compensating cation 296 K with water vapour at a mole fraction of 0.03. The choice of the
with its molar hydration heat and by dividing the final result by the operating conditions was decided so as to maintain the stability of the
molar mass of the model zeolite sample, it was possible to calculate the zeolite structure while taking the risk of reduced thermal performance of
limit value of the integral heat of adsorption per kg of the solid. The zeolite adsorbents undergoing incomplete regeneration-dehydration.
following limit values were obtained: 6290 kJ kg 1, Ce3þ-13X; 9590 kJ Three degassing-saturation cycles were carried out. Prior to the first
kg 1, Ce4þ-13X. It is clear that the heat values determined for the Ce- saturation stage, the adsorbent was activated by a vacuum treatment at
containing samples in the present section are far from these limit 423 K. Under such operating conditions, the cerium-containing samples
values which could be reached experimentally if the completely dried were found to release between 700 and 1100 kJ per kg of the sample
samples were saturated with water vapour at saturated vapour pressure. over a period of 6–8 h, with the highest heat being reported for the
The kinetics of the heat evolvement also changes from one sample to zeolite sample containing Ce4þ compensating cations during the first
another (c.f., Fig. S6 in Supplementary Material). At first sight, the heat saturation cycle. Higher heat values compared to those attained for the
release period lasts for 6–8 h. In the case of pristine 13X sample, a pristine 13X zeolite under the same conditions argued in favour of the
gradual increase in the heat of adsorption is observed on either side of enhanced thermal performance of cerium-containing zeolites suffi­
the central maximum attaining about 2 kJ kg 1 somewhere in the ciently enough to envisage the periodic heat release during the dis­
middle of the discharging period. A steeper increase in the heat release is charging step of the thermochemical storage performed under mild
observed in the beginning of the saturation cycle for Ce1-13X and the operating conditions.
maximum is between 2.5 and 3.5 kJ kg 1, depending on the cycle
number. In consequence, this Ce-contained sample is characterised by a Author-contribution
larger heat release window. As far as Ce2-13X and Ce3-13X are con­
cerned, the plots of the instantaneous value of the heat released upon H.W, P.T., B.F. carried out the experiments for calorimetry, isotherm
adsorption as a function of the discharging time represent left-skewed determination and XRD measurements. F.S. performed molecular sim­
distribution curves; the maximum heat values are about 3.2–3.5 and ulations, P.T., J.Z. and F.S. wrote the manuscript. J.Z. supervised the
3.2–3.8 kJ kg 1, respectively. Altogether, the Ce-exchanged samples project.
provide relatively much heat over a rather short period of a few hours.

4. Conclusion Declaration of competing interests

An alternative solution to harvest solar energy and store this energy The authors declare that they have no known competing financial
as chemical potential without much energy loss can be proposed based interests or personal relationships that could have appeared to influence
on the adsorption of water vapour occurring in natural porous adsor­ the work reported in this paper.
bents, thus allowing the investment costs to be greatly reduced.
Nevertheless, the main challenge is still to optimise the operating con­ Acknowledgement
ditions corresponding to the charging (adsorbent drying-regeneration)
and discharging (vapour adsorption) stages, which may have negative One of the authors, H. Wu, greatly acknowledges the financial sup­
effects on the hydrothermal stability of the adsorbent submitted to a port of his work by French Ministry of National Education, Higher Ed­
large number of repeated regeneration-saturation cycles. Within the ucation and Research. The authors also want to thank J. Fullenwarth and
framework of the present study, the potential use of commercially Dr. N. Donzel of the Platform of Analysis and Characterisation (PAC),
available zeolites, modified by ion-exchange with cerium compensating Po
^le Chimie Balard, for their assistance, respectively, in X-ray Fluores­
cations possessing high charges and high hydration energies, was tested cence and N2 adsorption measurements.
under mild operating conditions of low regeneration temperatures and
low pressures of water vapour during the adsorption step. Appendix A. Supplementary data
A research approach combining experimental characterization
techniques and theoretical methods was employed to monitor changes Supplementary data to this article can be found online at https://doi.
in the structural and textural properties of the zeolite samples as a org/10.1016/j.micromeso.2020.109999.
function of the hydration state and oxidation number of extra-
framework cerium cations. Based on the structure refinement proced­ References
ure applied to the experimental XRD patterns, it was demonstrated that
extra-framework cerium cations were preferentially located on sites I’ [1] L. P�erez-Lombard, J. Ortiz, C. Pout, A review on buildings energy consumption
information, Energy Build. 40 (2008) 394–398, https://doi.org/10.1016/j.
and II in dry and hydrated zeolites, showing relatively little dependence enbuild.2007.03.007.
on the hydration level. The experimental adsorption isotherms for water [2] P. Tatsidjodoung, N. Le Pierr�es, L. Luo, A review of potential materials for thermal
vapour onto cerium-containing zeolites and the pristine 13X sample energy storage in building applications, Renew. Sustain. Energy Rev. 18 (2013)
327–349, https://doi.org/10.1016/j.rser.2012.10.025.
were measured at 313 K after having submitted the samples to a vacuum [3] N. Yu, R.Z. Wang, L.W. Wang, Sorption thermal storage for solar energy, Prog.
treatment at 423 K. Monte Carlo simulations were used to determine the Energy Combust. Sci. 39 (2013) 489–514, https://doi.org/10.1016/j.
limit values of the amount adsorbed and differential heat of adsorption, pecs.2013.05.004.
[4] K.E. N’Tsoukpoe, H. Liu, N. Le Pierr� es, L. Luo, A review on long-term sorption solar
which could be obtained experimentally if the zeolite samples were
energy storage, Renew. Sustain. Energy Rev. 13 (2009) 2385–2396, https://doi.
completely dried. The comparison between the experimental and theo­ org/10.1016/j.rser.2009.05.008.
retical values, together with the results of thermogravimetric analysis, [5] S. Kiyabu, J.S. Lowe, A. Ahmed, D.J. Siegel, Computational screening of hydration
provided strong evidence for the presence of water molecules in the reactions for thermal energy storage: new materials and design rules, Chem. Mater.
30 (2018), https://doi.org/10.1021/acs.chemmater.7b05230, 2006–2017.
hydration shells surrounding Naþ, Ce3þ, and Ce4þ compensating cations [6] H. Wu, F. Salles, J. Zajac, A critical review of solid materials for low-temperature
in the zeolite structures. The mobility of cerium cations in the Ce- thermochemical storage of solar energy based on solid-vapour adsorption in view

8
H. Wu et al. Microporous and Mesoporous Materials 296 (2020) 109999

of space heating uses, Molecules 24 (2019) 945, https://doi.org/10.3390/ the mechanism of water vapor adsorption, Sol. Energy Mater. Sol. Cells 179 (2018)
molecules24050945. 223–230, https://doi.org/10.1016/j.solmat.2017.11.020.
[7] C. Bales, P. Gantenbein, A. Hauer, H.-M. Henning, D. Jaenig, H. Kerskes, et al., [28] A. Di Lella, N. Desbiens, A. Boutin, I. Demachy, P. Ungerer, J.-P. Bellat, et al.,
Report B2 of Subtask B: Thermal Properties of Materials for Thermo-Chemical Molecular simulation studies of water physisorption in zeolites, Phys. Chem. Chem.
Storage of Solar Heat : a Technical Report of Subtask B, 2005. Consulted on 11th Phys. 8 (2006) 5396–5406, https://doi.org/10.1039/b610621h.
September 2018. [29] C. Abrioux, B. Coasne, G. Maurin, F. Henn, M. Jeffroy, A. Boutin, Cation behavior
[8] L. Navarro, A. de Gracia, S. Colclough, M. Browne, S.J. McCormack, P. Griffiths, et in faujasite zeolites upon water adsorption: a combination of Monte Carlo and
al., Thermal energy storage in building integrated thermal systems: a review. Part molecular dynamics simulations, J. Phys. Chem. C 113 (2009) 10696–10705,
1. active storage systems, Renew. Energy 88 (2015) 526–547, https://doi.org/ https://doi.org/10.1021/jp902274t.
10.1016/j.renene.2015.11.040. [30] Y. Marcus, A simple empirical model describing the thermodynamics of hydration
[9] H. Deshmukh, M.P. Maiya, S. Srinivasa Murthy, Study of sorption based energy of ions of widely varying charges, sizes, and shapes, Biophys. Chem. 51 (1994)
storage system with silica gel for heating application, Appl. Therm. Eng. 111 111–127, https://doi.org/10.1016/0301-4622(94)00051-4.
(2017) 1640–1646, https://doi.org/10.1016/j.applthermaleng.2016.07.069. [31] D.W. Smith, Ionic hydration enthalpies, J. Chem. Educ. 54 (1977) 540, https://doi.
[10] S. Mitra, N. Aswin, P. Dutta, Scaling analysis and numerical studies on water org/10.1021/ed054p540.
vapour adsorption in a columnar porous silica gel bed, Int. J. Heat Mass Transf. 95 [32] J. Paier, C. Penschke, J. Sauer, Oxygen defects and surface chemistry of Ceria:
(2016) 853–864, https://doi.org/10.1016/j.ijheatmasstransfer.2015.12.011. quantum chemical studies compared to experiment, Chem. Rev. 113 (2013)
[11] A. Jabbari-Hichri, S. Bennici, A. Auroux, Effect of aluminum sulfate addition on the 3949–3985, https://doi.org/10.1021/cr3004949.
thermal storage performance of mesoporous SBA-15 and MCM-41 materials, Sol. [33] F.D. Hunter, Cation positions in Cerium X zeolites, J. Catal. 246259 (1971)
Energy Mater. Sol. Cells 149 (2016) 232–241, https://doi.org/10.1016/j. 246–259, https://doi.org/10.1016/0021-9517(71)90086-8.
solmat.2016.01.033. [34] F. Rouquerol, L. Luciani, P. Llewellyn, R. Denoyel, J. Rouquerol, Texture des
[12] A. Khutia, H.U. Rammelberg, T. Schmidt, S. Henninger, C. Janiak, Water sorption mat� eriaux pulv�
erulents ou poreux, Tech. l’Ing� enieur. P1050 (2003) 1–24.
cycle measurements on functionalized MIL-101Cr for heat transformation [35] L. Palatinus, Superflip — A Computer Program for Solution of Crystal Structures
application, Chem. Mater. 25 (2013) 790–798, https://doi.org/10.1021/ from X-Ray Diffraction Data in Arbitrary Dimension, User manual, Lausanne, 2007.
cm304055k. [36] V. Petrícek, M. Du�sek, L. Palatinus, Crystallographic computing system JANA2006:
[13] J. Canivet, A. Fateeva, Y. Guo, B. Coasne, D. Farrusseng, Water adsorption in general features, Zeitschrift Fur Krist 229 (2014) 345–352, https://doi.org/
MOFs: fundamentals and applications, Chem. Soc. Rev. (2014) 5594–5617, 10.1515/zkri-2014-1737.
https://doi.org/10.1039/c4cs00078a. [37] N. Tanchoux, P. Trens, D. Maldonado, F. Di Renzo, F. Fajula, The adsorption of
[14] M.F. De Lange, K.J.F.M. Verouden, T.J.H. Vlugt, J. Gascon, F. Kapteijn, Adsorption- hexane over MCM-41 type materials, Colloids Surfaces A Physicochem. Eng. Asp.
driven heat pumps: the potential of metal-organic frameworks, Chem. Rev. 115 246 (2004) 1–8, https://doi.org/10.1016/j.colsurfa.2004.06.033.
(2015) 12205–12250, https://doi.org/10.1021/acs.chemrev.5b00059. [38] P. Trens, H. Belarbi, C. Shepherd, P. Gonzalez, N.A. Ramsahye, U.H. Lee, Y.K. Seo,
[15] S.K. Henninger, H.A. Habib, C. Janiak, MOFs as adsorbents for low temperature J.S. Chang, Adsorption and separation of xylene isomers vapors onto the chromium
heating and cooling applications, J. Am. Chem. Soc. 131 (2009) 2776–2777, terephthalate-based porous material MIL-101(Cr): An experimental and
https://doi.org/10.1021/ja808444z. computational studyAuthor links open overlay panel, Micro. Meso. Mater 183
[16] B. Han, A. Chakraborty, Water adsorption studies on synthesized alkali-ions doped (2014) 17–22, https://doi.org/10.1016/j.micromeso.2013.08.040.
Al-fumarate MOFs and Al-fumarate zeolite composites for higher water uptakes [39] S. Yamanaka, P.B. Malla, S. Komarneni, Water sorption and desorption isotherms
and faster kinetics, Microporous Mesoporous Mater. 288 (2019) 109590, https:// of some naturally-occurring zeolites, Zeolites 9 (1989) 18–22, https://doi.org/
doi.org/10.1016/j.micromeso.2019.109590. 10.1016/0144-2449(89)90004-3.
[17] B. Mette, H. Kerskes, H. Drück, H. Müller-Steinhagen, Experimental and numerical [40] J. Zajac, R. Dutartre, D.J. Jones, J. Roziere, Determination of surface acidity of
investigations on the water vapor adsorption isotherms and kinetics of binderless powdered porous materials based on ammonia chemisorption: comparison of flow-
zeolite 13X, Int. J. Heat Mass Transf. 71 (2014) 555–561, https://doi.org/10.1016/ microcalorimetry with batch volumetric method and temperature- programmed
j.ijheatmasstransfer.2013.12.061. desorption, Thermochim. Acta 379 (2001) 123–130, https://doi.org/10.1016/
[18] D. Lefebvre, F.H. Tezel, A review of energy storage technologies with a focus on S0040-6031(01)00611-6.
adsorption thermal energy storage processes for heating applications, Renew. [41] G. Vitale, C.F. Mellot, L.M. Bull, A.K. Cheetham, Neutron diffraction and
Sustain. Energy Rev. 67 (2017) 116–125, https://doi.org/10.1016/j. computational study of zeolite NaX: influence of SIII‘ cations on its complex with
rser.2016.08.019. benzene, J. Phys. Chem. B 101 (1997) 4559–4564, https://doi.org/10.1021/
[19] R.A. Shigeishi, C.H. Langford, B.R. Hollebone, Solar energy storage using chemical jp970393x.
potential changes associated with drying of zeolites, Sol. Energy 23 (1979) [42] D. Frenkel, B. Smit, Understanding Molecular Simulation, second ed., From
489–495, https://doi.org/10.1016/0038-092X(79)90072-0. Algorithms to Applications, Academic Press, Hardcover ISBN: 9780122673511,
[20] T. Schmidt, D. Mangold, H. Müller-Steinhagen, Central solar heating plants with Published Date: 19th October 2001.
seasonal storage in Germany, Sol. Energy 76 (2004) 165–174, https://doi.org/ [43] A. Sartbaeva, S.A. Wells, M.M.J. Treacy, M.F. Thorpe, The flexibility window in
10.1016/j.solener.2003.07.025. zeolites, Nat. Mater. 5 (2006) 962–965, https://doi.org/10.1038/nmat1784.
[21] G. Storch, G. Reichenauer, F. Scheffler, A. Hauer, Hydrothermal stability of [44] Ş.Ç. Sayilgan, M. Mobedi, S. Ülkü, Effect of regeneration temperature on
pelletized zeolite 13X for energy storage applications, Adsorption 14 (2008) adsorption equilibria and mass diffusivity of zeolite 13X-water pair, Microporous
275–281, https://doi.org/10.1007/s10450-007-9092-7. Mesoporous Mater. 224 (2016) 9–16, https://doi.org/10.1016/j.
[22] K.M. Kim, H.T. Oh, S.J. Lim, K. Ho, Y. Park, C.H. Lee, Adsorption equilibria of micromeso.2015.10.041.
water vapor on zeolite 3A, zeolite 13X, and dealuminated Y zeolite, J. Chem. Eng. [45] J. Hunger, I.A. Beta, H. B€ohlig, C. Ling, H. Jobic, B. Hunger, Adsorption structures
Data 61 (2016) 1547–1554, https://doi.org/10.1021/acs.jced.5b00927. of water in NaX studied by DRIFT spectroscopy and neutron powder diffraction,
[23] J.C. Moïse, J.P. Bellat, A. M�
ethivier, Adsorption of water vapor on X and Y zeolites J. Phys. Chem. B 110 (2006) 342–353, https://doi.org/10.1021/jp054636u.
exchanged with barium, Microporous Mesoporous Mater. 43 (2001) 91–101, [46] D.H. Olson, The crystal-structure of dehydrated NaX, Zeolites 15 (1995) 439–443,
https://doi.org/10.1016/S1387-1811(00)00352-8. https://doi.org/10.1016/0144-2449(95)00029-6.
[24] P. Aprea, B. de Gennaro, N. Gargiulo, A. Peluso, B. Liguori, F. Iucolano, et al., Sr-, [47] C. Abrioux, B. Coasne, G. Maurin, F. Henn, A. Boutin, A. Di Lella, et al.,
Zn- and Cd-exchanged zeolitic materials as water vapor adsorbents for thermal A molecular simulation study of the distribution of cation in zeolites, Adsorption
energy storage applications, Appl. Therm. Eng. 106 (2016) 1217–1224, https:// 14 (2008) 743–754, https://doi.org/10.1007/s10450-008-9123-z.
doi.org/10.1016/j.applthermaleng.2016.06.066. [48] M. Jeffroy, A. Boutin, A.H. Fuchs, Understanding the equilibrium ion exchange
[25] F. Salles, J.M. Douillard, O. Bildstein, C. Gaudin, B. Prelot, J. Zajac, et al., Driving properties in faujasite zeolite from Monte Carlo simulations, J. Phys. Chem. B 115
force for the hydration of the swelling clays: case of montmorillonites saturated (2011) 15059–15066, https://doi.org/10.1021/jp209067n.
with alkaline-earth cations, J. Colloid Interface Sci. 395 (2013) 269–276, https:// [49] M. Jeffroy, E. Borissenko, A. Boutin, A. Di Lella, F. Porcher, M. Souhassou, et al.,
doi.org/10.1016/j.jcis.2012.12.050. Evidence of a framework induced cation redistribution upon water adsorption in
[26] F. Salles, J.-M. Douillard, O. Bildstein, S. El Ghazi, B. Pr�elot, J. Zajac, et al., cobalt exchanged X faujasite zeolite: a joint experimental and simulation study,
Diffusion of interlayer cations in swelling clays as a function of water content: case Microporous Mesoporous Mater. 138 (2011) 45–50, https://doi.org/10.1016/j.
of montmorillonites saturated with alkali cations, J. Phys. Chem. C 119 (2015) micromeso.2010.09.031.
10370–10378, https://doi.org/10.1021/jp512986d. [50] M. Jeffroy, C.N. Draghi, A. Boutin, A new molecular simulation method to
[27] D. Alby, F. Salles, J. Fullenwarth, J. Zajac, On the use of metal cation-exchanged determine both aluminum and cation location in cationic zeolites, Chem. Mater. 29
zeolites in sorption thermochemical storage: some practical aspects in reference to (2017) 513–523, https://doi.org/10.1021/acs.chemmater.6b03011.