Study of water in Ca-montmorillonite by thermal analysis and positron annihilation lifetime spectroscopy
- 65 Downloads
The aim of this study is to characterize adsorbed liquid in montmorillonite structure for different levels of adsorption by both thermoanalytical and microstructural methods. Montmorillonite of Ca type is used for the analysis. Water desorption process occurring typically between 50 and 180 °C is analysed in details by thermogravimetric analysis. Thermal response of adsorbed water for the selected steps of desorption process is studied by differential scanning calorimetry. Corresponding characterization of free volume is performed by positron annihilation lifetime spectroscopy. An attempt to determine a correlation of characterization method results is provided.
KeywordsMontmorillonite Thermogravimetric analysis DSC Positron annihilation lifetime spectroscopy
Montmorillonite is a microcrystalline mineral from the group of phyllosilicates and belongs in nature among the most widespread minerals. It has an extremely wide use in various industries, for example as catalytic processes agent, desiccant to remove moisture, inorganic filler in the polymer industry [1, 2]. Most of its applications use its unique sorption and colloidal properties. Chemically, it is hydrated sodium calcium aluminium magnesium silicate hydroxide (Na,Ca)0.33(Al,Mg)2(Si4O10)(OH)2·nH2O; exact ratio of cations varies with source.
Particular feature of montmorillonite is its rather complex porous structure. It is formed by layered sheet particles of average diameter ~ 500 nm and thickness of ~ 1.0 nm with specific inner and outer surface which is characterized by enhanced cation exchange capacity. Such a structure contains rather high amount of interlayer space. Water and other polar liquids are highly adsorbed by the surface of montmorillonite particles. As the liquid can penetrate to interlayer space, montmorillonite structure swells. The amount of expansion is due largely to the type of exchangeable cation contained in the structure. In certain cases, the material can increase its volume several times.
The liquid adsorbed by porous structure finds itself in so-called confined state. Various studies of the systems porous structure-adsorbed liquid have revealed that numerous physical properties of such confined liquids and probably also solids might be modified or even principally changed [3, 4, 5, 6, 7]. The variability of thermal and mechanical properties of such systems still needs to be investigated. Moreover, the water itself represents a liquid with relatively complex thermal behaviour and merits further analysis of the influence of confined state on its physical properties.
The theme of water interaction and montmorillonite is a subject of long-term wide research , but studies using combination of thermal and microstructural characterization, namely by PALS, are rather rare. This work deals with properties of water in Ca-montmorillonite. The choice of calcium montmorillonite was influenced by the fact that calcium ions are most often represented in the variable positions of the montmorillonite. However, natural montmorillonites typically do not contain only one ion type in these positions. For better characterization of our montmorillonite, the original form of montmorillonite was converted to a monionic calcium form. Further an attempt to correlate the analysis with microstructural data obtained via free-volume approach is presented. The study is a part of complex investigation of the structure and properties of confined nanosystems [6, 7, 8, 9, 10, 11].
Materials and methods
Montmorillonite was isolated by decanting method from 4% water suspension of bentonite from deposit of Stará Kremnička–Jelšový potok (Slovak Republic) . Ca2+-doped montmorillonite was prepared by ion-exchange reaction in the water solution of CaCl2 (concentration csol = 0.2 mol dm−3) using five repeated cycles. Montmorillonite doped in this way was washed by distilled water and dried in vacuum. More detailed information on bentonite and montmorillonite from this site is given elsewhere [19, 20].
Microstructural and chemical analysis was carried out using JEOL JSM 6610 scanning electron microscope (SEM) operated at 15 kV and equipped with energy-dispersive spectroscopy detector (EDX) (Oxford Instruments).
Anhydrous montmorillonite was prepared by drying for 24 h at 393 K (120 °C) in a vacuum of 10−6 mbar. Samples from this material for desorption experiments were closed for at least 24 h in a water vapour container to prepare samples in fully saturated state. The pure water LC–MS Ultra CHROMASOLV from Honeywell/Riedel–de Haen was used.
The TG analysis measurements were taken using TGA7 (PerkinElmer) in both linear heating and isothermal regimes using dynamic argon atmosphere. For isothermal analysis, the sample was kept at temperature Tan = 323 K (30 °C) for 600 min. Linear heating measurements were taken in temperature region 303–453 K (30.0–180.0 °C) with the rates w = 1.0, 2.0, 5.0 and 10.0 K min−1. Mass loss coefficient ml, which determines the mass loss of the sample during measurements, can be directly determined from relative mass coefficient k measured by TG unit as ml = 1 − k.
DSC measurements were taken using power-compensation DSC8500 (PerkinElmer) with automatic intracooler (200 K). Samples of 15–30 mg encapsulated in air and measured in dynamic nitrogen atmosphere were used. Both hermetic and non-hermetic encapsulation was used for specific way of study of thermal properties. For certain analyses, the samples of material after specific pre-treatment by TG were analysed.
Continuous cooling and heating regimes with the rates of w = − 2.0 or + 2.0 K min−1, respectively, were applied. Each DSC measuring cycle started at 303.0 K, and the cool/heat procedure was used. It means that firstly each sample was slowly cooled to 213.0 K; a temperature which is sufficiently below the equilibrium crystallization temperature of water and the solidification was measured; then, it was continuously heated up to 303.0 K and the corresponding melt transition was detected.
Positron annihilation lifetime spectra were measured by the conventional fast–fast coincidence time spectrometer with the time resolution about 320 ps FWHM determined using Al defect-free sample. LT programme  was used for the analysis of spectra. Three lifetime components were fitted during the analysis. Shorter lifetimes τ1 and τ2 (in subnanosecond range) are connected with the annihilation of para-positronium and free positrons. Larger lifetime τ3 (in nanosecond range) originated from the annihilation of ortho-positronium in free volumes of different sizes is important for this study and was discussed later. The correction on the positron annihilation in the source as well as in the sample container kapton foil window was taken into account.
Results and discussion
The occurrence of tiny peak at ~ − 41.0 °C (~ 232.0 K) is typically observed during DSC analysis of confined water-containing systems. This peak is ascribed to the crystallization of confined water in the spherical cavities by homogenous nucleation [23, 24]. Its presence shows that some part of confined water in montmorillonite structure finds itself in regions where it is isolated from ice front percolation during solidification. The consequence is that it can be supercooled down to the temperature, which is considered as a lowest attainable temperature for pure water when crystallization by homogeneous nucleation occurs .
This effect is more complex for montmorillonite than for previously studied confined water-containing systems . It manifests itself by the presence of two peaks of slightly different intensity with onset temperatures ~ − 32.0 °C (~ 241.0 K) and ~ − 39.0 °C (~ 234.0 K), respectively. This is probably related to specificity of montmorillonite pores structure; however, the interpretation is still open. This effect is relatively more pronounced as well, the area of corresponding peaks represents more than 8% total area of solidification peaks, while, e.g. for controlled pore glass, CPG126 was that almost negligible (< 0.1%) . Besides good reproducibility, these two peaks exhibit one more specific feature which will be explained later.
Typical temperatures for individual steps (corresponding to maximum of derivative peaks) were estimated as T1 = 35 °C (308 K), T2 = 60 °C (333 K) and T3 = 110 °C (383 K); the values were determined for the heating rate of w = 2.0 K min−1 (Fig. 3a). Corresponding mass losses were determined as ml1 = 5.4%, ml2 = 17.9% and ml3 = 22.7%, respectively. Three samples were prepared in these states by initial heating in TG unit to corresponding temperature after which the sample was immediately hermetically encapsulated and analysed by DSC. The results of their analyses were compared with that of as-prepared sample (Fig. 3b). From the results, it is evident that a significant part of the water giving a thermal response during DSC analysis is released during the first step of dehydration. Its complex nature seems to be conserved during releasing. This part of adsorbed water represents only its small fraction and is rapidly released from montmorillonite pores. Heating of sample to temperature T1 causes the loss of approximately only one tenth of its total amount (entire loss after heating to 180 °C is ~ 25%), but already almost two-thirds of this part are released as can be estimated from the corresponding DSC peak areas.
Analogous characterization was done by the TG analysis of sample by annealing at 30 °C (Fig. 4b). Total area of DSC solidification peaks is compared with corresponding TG curve. Derivative of TG curve shows several steps of its evolution. Similarity with solidification peaks area evolution allows assume that first two steps correspond to the release of corresponding forms of adsorbed water. Using this assumption, one can estimate that loosening of corresponding part of adsorbed water (giving thermal response in DSC) was finished roughly after 45–50 min.; the sample lost ~ 14–15.0% of its mass during this time (which is ~ 60% of total amount adsorbed water). Using this coefficient for the estimation of corresponding mass of water in hermetically encapsulated sample analysed by DSC (Fig. 2), one obtain total enthalpy of solidification for 1.cooling run 260.2–278.8 J g−1 and total enthalpy of melting for 1.heating run 286.8–307.3 J g−1. An analogous discrepancy between enthalpy values for cooling and heating runs as well as deficit to table data of that for pure water (333.6 J g−1) had been obtained for previous studies of confined water-containing systems . The discrepancy seems to be a specificity of the used method of thermal analysis of water (it was observed even for pure water). For the observed deficit to table data, it had been supposed that it had confirmed an existence of certain part of water that does not crystallize during solidification. However, in this case such an interpretation is problematic. It has to be noted that the used estimation is loaded with a certain error as the evaporation rates are not necessarily the same in the DSC and TG units. Moreover, there is obviously an overlapping with the second step of dehydration (corresponding to the release of water molecules bound to the + 2 exchanged cations) as the material was continuously losing its mass in the course of annealing, e.g. for 600 min., it lost 17.8% of its mass.
The formation of Ps in montmorillonite is small, as shown by other measurements . Intensity I3 is less than 2% for all measured values. It increases from 0.8 to 2% as the water content increases.
According to TG study dehydration of water-adsorbed Ca-montmorillonite (elaborated by vapour adsorption process) takes place in three apparent steps. The material loses > 25% of its original mass. This result is in principal agreement with the usually observed thermal behaviour of analogous form of montmorillonite.
A significant part of water giving a thermal response during cooling–heating treatment of material by DSC is released during the first step of dehydration. Its thermal behaviour is rather similar to that of confined water in other fine porous structures. The solidification of this part of water consists mainly of heterogeneous nucleation, the onset temperature of which depends on water content. Relatively important fraction of this part exhibits homogeneous nucleation with constant onset temperature ~ − 32.0 °C (~ 241.0 K). The release of this part is finished after the release of ~ 60% of total amount adsorbed water.
The following two steps of dehydration correspond well to usually accepted model of three forms of water molecules bonding in montmorillonite, e.g. water molecules bonded to exchangeable cations and water molecules bonded to layers of the structure.
PALS measurements showed that for the water content in the range of concentration values ~ 0.15–0.36, which corresponds to the concentration range when water giving thermal response in DSC (released during first step of dehydration) is present in the structure, the longest lifetime spectrum component τ3 linearly decreases from 3.0 to 1.8 ns which is the value close to that of liquid water. Interlayer spacing of Ca-montmorillonite without water (c = 0) determined from corresponding lifetime spectrum component τ3 value (using parallel plate model) is in good agreement with the value determined by other structural methods.
Obviously further research is essential to understand more aspect of really complex thermal behaviour of this type confined water system.
Work was partially supported by projects VEGA 2/0127/17, VEGA 2/0157/17, APVV-16-0369 and APVV-15-0621.
- 14.Zaleski, R. EELViS. http://eelvis.sourceforge.net (2009). Accessed 15 Jun 2011.
- 16.Goworek T. Positronium as a probe of small free volumes in crystals, polymers and porous media. Ann Univ Maria Curie-Sklodowska, Lublin-Polonia. 2014;69:1–110.Google Scholar
- 18.Jesenák K. Laboratory device for sedimentation separation of powders. Ceramics. 1994;38:35–6.Google Scholar
- 19.Jesenák K, Kuchta Ľ, Guller L, Fúsková J. Physico-chemical properties of Bentonite “Stará Kremnička – Jelšový potok” I: particle size distribution. Mineralia Slovaca. 1997;29:439–42.Google Scholar
- 20.Jesenák K, Fajnor V. Distribution of trace elements in bentonite Stará Kremnička – Jelšový potok. Mineralia Slovaca. 1995;27:221–4.Google Scholar