energies Article
Thermal and Stability Investigation of Phase Change Material Dispersions for Thermal Energy Storage by T-History and Optical Methods Maria Gabriela De Paola 1 , Natale Arcuri 1 , Vincenza Calabrò 2, * and Marilena De Simone 1 1
2
*
Department of Mechanical, Energy and Management Engineering (DIMEG), University of Calabria, Via P. Bucci 46/C, 87036 Rende, Italy;
[email protected] (M.G.D.P.);
[email protected] (N.A.);
[email protected] (M.D.S.) Department of Informatics, Modelling, Electronics and System Engineering (DIMES), University of Calabria, Via P. Bucci 39/C, 87036 Rende, Italy Correspondence:
[email protected]; Tel.: +39-0984496703
Academic Editor: Francesco Asdrubali Received: 29 September 2016; Accepted: 2 March 2017; Published: 13 March 2017
Abstract: Glauber’s salt (sodium sulphate decahydrate) is a promising phase change material (PCM) for use in the building sector, thanks to its high enthalpy of fusion associated with a proper phase transition temperature. It also offers economic and environmental advantages because it can be obtained as a byproduct from the disposal process of lead batteries. However, due to phenomena of phase segregation and supercooling, Glauber’s salt cannot be used in its pure state and requires the addition of rheological modifiers and nucleating agents. In this work, the initial thermal performances of mixtures based on Glauber’s salt with different compositions are compared by using the T-history method and adopting sonication for mixing, and following the same preparation procedure for all the samples. With fixed composition, the effects of the addition sequence of the reagents are also examined. The analysis carried out by optical methods based on light scattering (Turbiscan equipment) allowed us to identify the kinetics of destabilization for each sample and revealed the need to specify in detail the preparation stages of PCMs, in order to make the composition reproducible in the laboratory and on a wider scale. Keywords: thermal energy storage (TES); phase change material (PCM); Glauber’s salt; T-history; stability
1. Introduction Phase change materials (PCMs) are suitable products for thermal management solutions, because they store and release thermal energy during the process of melting and freezing. They are also very effective in providing thermal barriers or insulation, for example in temperature controlled transport. PCM dispersions based on salts can be very interesting for applications in the building sector, thanks to the high latent heat of fusion and the temperature at which this transition occurs. In such applications, they are enclosed in boards for the thermoregulation of internal environments. Panels can cover walls, floors, or ceilings, therefore large quantities of such materials are required. This need imposes limits in the choice of materials with regards to economic aspects, making it useful to address the recycling of waste materials, for example the Glauber’s salt that is obtained as a byproduct from the disposal process of lead batteries. The recovery is carried out by the crystallization of sulphate, which also allows for the elimination of traces of heavy metals, toxic to the environment and damaging to health. Energies 2017, 10, 354; doi:10.3390/en10030354
www.mdpi.com/journal/energies
Energies 2017, 10, 354
2 of 15
In particular, in the phase of the pastel desulfurization in which the lead sulphate is treated by using the carbonate or hydroxide of sodium, the following reactions take place [1,2]: PbSO4 + Na2 CO3 → PbCO3 + Na2 SO4
(1)
PbSO4 + 4NaOH → Pb(OH)2 + Na2 SO4
(2)
Due to its thermal properties, the Glauber’s salt has been one of the first salt hydrates to be considered suitable for applications in the building sector [3]. Glauber salt’s transition is a peritectic phase change where the solid Glauber salt produces a liquid phase consisting of a sodium sulphate solution and a solid phase consisting of Na2 SO4 [4]. In reference [5], data for the pure Glauber’s salt are available: the enthalpy of fusion is equal to 251.2 kJ/kg, the phase transition temperature is 32.4 ◦ C, and the specific heat of the solid and liquid phase are equal to 2.09 kJ/(kg·K) and 3.35 kJ/(kg·K), respectively. However, during initial investigations, problems of stability have been highlighted and, as a consequence, difficulties in long term performances. These problems were mainly due to two phenomena common to many other hydrated salts: supercooling and phase segregation. The first prevents the release of heat at a defined temperature, while the second, due to the incongruent melting, limits the stability of the structure which is manifested by a progressive reduction of the enthalpy of fusion. The solutions proposed in the literature involve adding a nucleating substance [6] and a rheological modifier [7], in addition to an excess of water whose function is to counter the low solubility of the salt [8]. When additives are mixed with pure PCMs, a heterogeneous mixture occurs. The chemical, physical, and thermal properties are completely different with respect to the pure compounds, and they have to be evaluated experimentally for such heterogeneous mixtures. The study presented in this paper is based on the characterization and the thermal analysis of two compositions with different percentages of Glauber’s salt, obtained by using bentonite as a thickener and borax as a nucleating agent. This paper presents results related to the compositional weight percentage of sodium sulphate of 25% and 31%, and of 67% and 60% of water chosen after a selection of different compositions, previously tested with the aim of achieving interesting PCM thermal properties. The objective is to analyze their effect on the enlargement of the solidification interval, on the lowering of the solidification temperature, and on the fusion enthalpy and stability in comparison to the pure salt. Since the samples are dispersions, it was necessary to identify the most appropriate type of mixing, opting for sonication as the more effective method as opposed to the magnetic stirring technique in reducing the size of the aggregates [9]. Two samples were prepared for each composition by varying the sequence of addition of the reagents. Each sample was analyzed successively by means of the T-history method for the determination of their thermal properties, and by using optical techniques to obtain information on their stability. The purpose of the paper was to introduce a methodology able to predict the stability of Glauber’s salt based PCMs in a relatively short time with a significant amount of samples, in order to guarantee that any inhomogeneity will be taken into account. By such a method, it is possible determine how both compositions and methods of preparation affect the properties of PCMs. The results have also been used in order to obtain further information about the thermal performance of the studied PCMs. 2. Materials and Methods 2.1. Reagents and Preparation of the Samples For the preparation of the samples the following reagents were utilized: distilled water, sodium sulfate 99.5% purity, borax 99.5% purity, potassium chloride 99% purity, purchased from Carlo Erba (Milan, Italy), and bentonite, provided by Sigma-Aldrich Italia (Milan, Italy). Table 1 shows the two analyzed compositional weight percentages.
Energies 2017, 10, 354
3 of 15
Table 1. Compositional weight percentage of the tested mixtures. Composition
Water (%)
Bentonite (%)
Sodium Sulphate (%)
Borax (%)
1 2
67 60
4 5
25 31
4 4
The compositional weight of bentonite and borax were selected thanks to preliminary experimental activity, carried out with the aim of investigating the optimal range of composition of the additives able to guarantee that the phase change occurs in the temperature range of interest for applications in building fields. As a consequence, the ratios between sodium sulphate and water are equal to 0.37 for sample 1 and 0.52 for sample 2. The samples have been made using a sonicator UP200S provided by Hielscher (Teltow, Germany); sonotrode S14 with a diameter of 14 mm), proceeding with 0.8 cycles of impulse followed by 0.2 s of pause at 100% of the length amplitude. Each composition (in total 100 g) has been obtained by the application of two different modalities: Modality (1): water and bentonite were sonicated for 10 min. After the addition of salts, the dispersion sonicated for a further 30 min. Modality (2): water and salts are sonicated for 30 min, and in a second step the bentonite was added and sonication was applied for 10 min. The samples will be identified in the paper following the nomenclature shown in Table 2. Table 2. Identification of the four samples in relation to the composition and modality of preparation. Sample
Composition
Modality of Preparation
C1(1) C1(2) C2(1) C2(2)
1 1 2 2
1 2 1 2
During sonication an increase of temperature to 70 ◦ C was observed in all modalities. For this reason, all the samples were cooled to 40 ◦ C in the thermostable bath, in order to apply the T-history method and optical analysis. It was possible to observe that all the samples appeared solid at temperatures lower than 19 ◦ C and liquid for temperature higher than 31 ◦ C. Sample preparation, as well as thermal and stability analyses, were replicated more than two times in order to guarantee a good reproducibility of the results. The data scattering was calculated for the thermal analysis and the optical stability analysis in terms of mean percentage error between the sets of data. 2.2. Thermal Analysis: T-History Method The thermal analysis was carried out by using the T-history method [10] in the temperature range from 10 ◦ C to 40 ◦ C. The original setup has been improved by using a refrigerated thermostat FOC 215 l from Velp Scientifica, (Usmate Velate (Monza Brianza), Italy)), as also suggested by Stankovic et al. [11], coupled in our case with the Software TEMPSoft™ supplied from Velp Scientifica, (Usmate Velate (Monza Brianza), Italy)) to program the heating and cooling ramps continuously, in order to obtain an adequate control of the environmental conditions, following the assigned function. Duran® test tubes (height of 16 cm and diameter of 1 cm) were used to store the PCM and reference material during the T-history analysis. Water was used as reference substance, because it does not show phase change in the investigated temperature range. The distance between test tubes was large enough to avoid any interference during heat transport.
Energies 2017, 10, 354
4 of 15
The T-history method was applied by decreasing the temperature from 40 ◦ C to 10 ◦ C with a step of then the samples stayed at 10 ◦ C for two hours, and after they were heated to 40 ◦ C (step of 1◦ /min). The cycle was repeated four times. Temperatures inside the PCM test tube as well as inside the reference (water) test tubes were continuously measured with resistance temperature detectors (RTDs) PT100, a miniature sensor RTD with high accuracy, with a diameter of 0.5 mm and length of 100 mm (able to guarantee fast and reproducible results). Two tests were performed for each sample. PCM mixture samples in the liquid phase, after sonication and cooled to 40 ◦ C, were immediately placed in the refrigerated thermostat and RTD probes were inserted in the test tubes. Simultaneously, test tubes with reference substances were heated at 40 ◦ C and the RTD probes were lodged in the refrigerated thermostat that will represent the environment at which the temperature is called the ambient temperature. Temperature profiles during the heating and cooling cycles were continuously measured, following the method proposed from Zhang and Jiang [10] and T-history curves were prepared. As reported in the literatures [10–12], cooling curves were used for the evaluation of ∆H. The results are therefore presented with reference to the method of Marin et al. [12], in which in addition to the enthalpy of fusion, the representation of the enthalpy-temperature trend is shown in the examined range. Such representation takes into account that PCMs do not have homogeneous compositions and, therefore, they have a rather wide solidification range. The enthalpy–temperature curve has been obtained by considering the changes in enthalpy in small temperature ranges according to Equation (3): 1◦ /min,
∆h P ( Ti ) =
mw ·c pw ( Ti ) + mt ·c pt ( Ti ) I mt · i ·∆Ti − ·c pt ( Ti )·∆Ti mP Ii ’ mP
(3)
where mw , mt , and mp are the mass of water, test tube, and PCM, respectively (kg); cpw and cpt are the specific heat of water and the test material (kJ/(kg·K)); for the ith step, ∆Ti is the temperature range (◦ C) with average value Ti , to which the enthalpy variation ∆hp (Ti ) (kJ/kg) corresponds; Ii is the integral with respect to time of the difference in temperature between the PCM and environment; Ii ’ is the integral with respect to time of the difference in temperature between the water and environment. The latent heat of solidification was obtained as [∆hp (TiS ) − ∆hp (TfS )] where TiS is the initial temperature of solidification calculated as the maximum value reached after sub-cooling and TfS is the final temperature, calculated according to as a point of inflection on the cooling curve of PCM [13]. The specific heat of PCM in the liquid (cpl ) and solid state (cps ) was calculated in reference to the article of Marinas, where the gradients of the two straight lines are observable in the enthalpy versus temperature trend [12,13]. 2.3. Kinetic of Destabilization PCMs made of Glauber’s salt and additives are dispersions. For this reason, they are subject to destabilization phenomena during the time that could affect their thermal properties. Generally, these phenomena concern particle migration that determines sedimentation and creaming, or variation of their dimensions (flocculation and coalescence). Destabilization could affect the thermal performance of PCMs, and consequently it is important to predict these phenomena from the beginning. When sedimentation, clarification, or flocculation occur, the texture of the materials is changed, and as a consequence the thermal conductivity, thermal capacity, heat of fusion or solidification, and density change: the system could be compared to a side or series resistance system, in terms of heat exchange. Fusion or solidification phenomena occur in different ways in the material and that could affect the effectiveness of the thermal behaviour of the PCM. Destabilization phenomena are not visible to the human eye when they are characterized by low kinetics, and only some sophisticated optical methods and instruments, such as the Turbiscan, are able to detect them from the initial stages [14–16]. The operating principle is based on multiple light scattering. The instrument scans samples with a volume of about 25 mL and draws transmission (T) and backscattering (BS) profiles along the height
Energies 2017,2017, 10, 354 5 of 14 Energies 10, 354 5 of 14 Energies 2017, 10, 354 5 of 514of 14 Energies 2017, 10, 354 Energies Energies 2017,2017, 10, 354 10, 354 5 of 514of 14 The operating principle is based on multiple lightlight scattering. The The instrument scans samples The principle is based on multiple scattering. instrument scans Energies 2017, 10, operating 354 5 samples of 15
The operating principle isand based ontransmission multiple light scattering. TheThe instrument scans samples The operating principle is and based on multiple light scattering. instrument scans samples withwith a volume of about 25 mL (T) and backscattering (BS)(BS) profiles along the the aoperating volume ofprinciple about 25is mL draws transmission (T) and backscattering profiles along The The operating principle based isdraws based on on multiple multiple light light scattering. scattering. TheThe instrument instrument scans scans samples samples with a volume of about 25 mL and draws transmission (T) and backscattering (BS) profiles along thethe with a volume of about 25 mL and draws transmission (T) and backscattering (BS) profiles along height of the test cell. The scan can be repeated for a long time, and at fixed intervals of time. The height of the test cell. scan can betransmission repeated for(T) a (T) long time, and at fixed intervals of along time. The with with a volume a volume of about of about 25 The mL 25 mL and and draws draws transmission andand backscattering backscattering (BS)(BS) profiles profiles along the the height of profiles the test cell. The scan cancan be be repeated fortime, a long time, and at fixed intervals of time. The height of the test cell. The repeated for athe long and atthe fixed intervals of time. of the test cell. The scan can bescan repeated for ainitial long and attime, fixed intervals of time. The more theThe more the are distinguished from the one, more unstable dispersion. more the profiles are distinguished from the initial one, the more unstable the dispersion. height height of the of the testtest cell.cell. TheThe scan scan cancan be repeated be repeated for for a long a long time, time, andand at fixed at fixed intervals intervals of time. of time. TheThe more the profiles distinguished from initial one, thethe more unstable thethe dispersion. more the profiles are distinguished from the initial one, unstable dispersion. profiles are distinguished from the initial one, the more unstable the dispersion. In the case ofare opaque dispersions, asthe those considered in more this investigation, the backscattering In the case of opaque dispersions, as those considered inunstable this investigation, the backscattering more more the the profiles profiles are are distinguished distinguished from from the the initial initial one, one, the the more more unstable the the dispersion. dispersion. In the case of opaque opaque dispersions, as those those considered in this this investigation, thethe backscattering Inmore the case of opaque dispersions, as those considered innull. this investigation, backscattering In is the case of dispersions, as considered investigation, the backscattering profile effective than the transmission profile that that is in null. profile is more effective than the transmission profile In the In the case case of opaque of opaque dispersions, dispersions, as those as those considered considered inisthis in this investigation, investigation, the the backscattering backscattering profile is more effective than the transmission profile that is null. profile is more effective than the transmission profile that is null. profile is more effective than the transmission profile that is null. The BS percentage along the test cell provides information regarding the homogeneity of the The BS percentage along the test cell provides information regarding the homogeneity of the profile profile is more is more effective effective than than the the transmission transmission profile profile thatthat is null. is null. The BS BS percentage along thethe testtest cellcell provides information regarding thethe homogeneity of the percentage along provides information regarding homogeneity of the TheThe percentage along the test cell provides information regarding the homogeneity of the sample, but in order to easily identify the phenomena in the samples, it is recommended to use the sample, but in order to easily identify the phenomena in the samples, it is recommended to use the TheThe BS BS percentage percentage along along the the testtest cellcell provides provides information information regarding regarding the the homogeneity homogeneity of the of the sample, but in order to easily identify the phenomena in the samples, it is recommended to use the sample, but in order to easily identify the phenomena in the samples, is recommended to use the sample, but in order to easily identify the phenomena in the samples, it recommended to use Delta mode. One of the profiles (by default, the first one) is as a as “blank” corresponding touse the Delta mode. ofeasily the profiles (bythe default, the first is used corresponding tothe the sample, sample, but but in order inOne order to to easily identify identify the phenomena phenomena in one) the in used the samples, samples, it aisit“blank” recommended is recommended to use to the Delta mode. One of the profiles (by(by default, thethe first one) isone) used aas“blank” corresponding tothus the Delta mode. One ofstate the profiles default, first one) is used aas“blank” corresponding to thus the the Delta mode. One of the profiles (by default, the first is as used afrom “blank” corresponding initial dispersion state of the sample. This profile is then subtracted from the other profiles, initial dispersion of the sample. This profile is then subtracted the other profiles, Delta Delta mode. mode. OneOne of the of the profiles profiles (by(by default, default, the the firstfirst one)one) is used is used as aas“blank” a “blank” corresponding corresponding to the to the initial dispersion state ofstate the sample. This profile is then from the other profiles, thus initial dispersion state of[17]. the sample. This profile is then subtracted from the other profiles, thus to the initial dispersion of the sample. This profile issubtracted then subtracted from the other profiles, emphasizing the variations emphasizing the variations [17]. initial initial dispersion dispersion state state of the of the sample. sample. ThisThis profile profile is then is then subtracted subtracted from from the the other other profiles, profiles, thusthus emphasizing the variations [17]. emphasizing the variations [17]. thus emphasizing the variations [17]. In Table 3, the examples of [17]. the results in terms of delta backscattering (∆BS) obtained withwith In Table 3,variations examples of the results in terms of delta backscattering (∆BS) obtained emphasizing emphasizing the variations [17]. In Table 3,examples examples ofreference thethe results in most terms ofbackscattering delta backscattering (∆BS) obtained with In Table 3,presented examples of results in terms of delta backscattering (∆BS) obtained with In Table 3,presented of the results in terms of delta (∆BS) obtained with Turbiscan Turbiscan are with to the important destabilization phenomena that occur Turbiscan are with reference to the most important destabilization phenomena that occur In In Table Table 3, examples 3, examples of the of the results results in terms in terms of delta of delta backscattering backscattering (∆BS) (∆BS) obtained obtained with with Turbiscan are presented with reference to the most important destabilization phenomena that occur Turbiscan are presented with reference to the most important destabilization phenomena that occur are presented with reference to the most important destabilization phenomena that occur with with inhomogeneous PCMs. with inhomogeneous PCMs. Turbiscan Turbiscan are are presented presented with with reference reference to the to the most most important important destabilization destabilization phenomena phenomena thatthat occur occur with inhomogeneous PCMs. with inhomogeneous PCMs. inhomogeneous PCMs. with with inhomogeneous inhomogeneous PCMs. PCMs. Table 3. Destabilization phenomena occurring in inhomogeneous dispersions and the analysis Table 3. Destabilization phenomena occurring in inhomogeneous dispersions and∆BS the ∆BS analysis Table 3.out Destabilization phenomena inhomogeneous ∆BS∆BS Table 3. Destabilization phenomena occurring in backscattering. inhomogeneous dispersions andand the the ∆BS analysis Table 3.by Destabilization phenomena occurring in inhomogeneous dispersions analysis carried Turbiscan instrument [17]. BS: backscattering. carried out by Turbiscan instrument [17]. BS: Table Table 3. Destabilization 3. Destabilization phenomena phenomena occurring occurring in inhomogeneous in inhomogeneous dispersions dispersions andand the the ∆BS∆BS analysis analysis carried outout by Turbiscan instrument [17].[17]. BS: BS: backscattering. Turbiscan instrument [17]. backscattering. carried by Turbiscan instrument backscattering. carried carried out out by Turbiscan by Turbiscan instrument instrument [17]. [17]. BS: BS: backscattering. backscattering. Particle Migration Particle Migration Destabilization Phenomena ΔBS ΔBS Destabilization Phenomena Particle Migration Particle Migration or Change in Size or Change in Size Destabilization Phenomena ΔBSΔBS Particle Migrationor Destabilization Phenomena Particle Particle Migration Migration or Change in Size ∆BS Destabilization Phenomena orChange Change inSize Size Destabilization Destabilization Phenomena Phenomena ΔBSΔBS or Change or Change in in Size in Size The density of theofdispersed The density the dispersed The density the dispersed The of the dispersed phase isdensity greater than the phase isof greater than the TheThe density density of the of the dispersed dispersed phase is greater than the phase is greater than the phase density of the continuous density of the continuous The density of the dispersed phase phase is greater is greater thanthan the the density the continuous of the continuous phase soofthe volume fraction isdensity greater than the density of the phase so the volume fraction density density of the of the continuous continuous phase so the volume fraction phase sodecrease the volume fraction phase so the volume Φcontinuous will decrease at the top of Φ will at the top of phase phase so the so the volume fraction fraction Φ will decrease atvolume the top of Φ will decrease at at the top of top fraction Φ will decrease at the the cell, and increase the the cell, and increase at the Φ will Φ will decrease decrease at the atatthe top top of of the cell, and increase the ofthe the cell, and increase cell, and increase atatthe bottom. bottom. the the cell, cell, and and increase increase at the at the the bottom. bottom. bottom. bottom. bottom. The dispersed phase has ahas a The dispersed phase The dispersed phase hasthe a a The dispersed phase has lower density than the lower density than The The dispersed dispersed phase phase has has a a lower a The dispersed phase has lower density than the lower density than the continuous phase so the continuous phase so the lower lower density density than than the the density than the continuous continuous phase the continuous phase sowill the phase volume fraction Φso will volume fraction Φ continuous continuous phase phase so the so Φ thewill sovolume the fraction volume fraction volume Φ will fraction Φthe will increase at the top of cell, increase at the top of the increase at the top of the cell,cell, volume volume fraction fraction Φ of will Φthe will increase at the top cell, increase at the top of the cell, and decrease at the bottom. and decrease at the bottom. and decrease at the bottom. increase increase at the atatthe top top of the of the cell,cell, and decrease the bottom. and decrease at the bottom. andand decrease decrease at the at the bottom. bottom. This This phenomenon will affect phenomenon will affect This phenomenon willof affect This phenomenon will affect the average diameter the the average diameter of the This phenomenon will affect the This This phenomenon phenomenon will will affect affect the average diameter of the the average diameter of the particles indiameter theinsample. It will particles the sample. It will average of the particles the the average average diameter diameter of the of the particles in the sample. It will inover the sample. It will increase over the increase the whole inparticles the sample. Itwhole will increase over particles particles in the inthe the sample. sample. It will It will increase over whole increase over the whole height of theof cell as the whole height of the cell as height the cell as increase increase over theas the whole whole height ofover the cell height ofor the cell as coalescence orflocculation flocculation are coalescence coalescence or flocculation height height of the of the cell cell as as coalescence or flocculation coalescence or flocculation global phenomena, taking place are global phenomena, are global phenomena, coalescence coalescence orinflocculation or flocculation are global phenomena, everywhere aeverywhere sample. are global phenomena, taking place everywhere in a in a taking place are are global global phenomena, phenomena, taking place everywhere in ain a taking place everywhere sample. sample. taking taking place place everywhere everywhere in ain a sample. sample. sample. sample.
By means ofofcomparing thethe profiles in time, the instrument determines the destabilization kinetics means comparing profiles in time, the the instrument determines the the destabilization By means of comparing the profiles in time, instrument determines destabilization By means of comparing the profiles in time, the instrument determines the destabilization By means of comparing the profiles in time, the instrument determines the destabilization based themeans computation of the Stability Index (TSI). The TSI is a The computation directly kinetics based on of the Stability Index (TSI). The TSIthe is adestabilization computation kinetics based on the computation of Turbiscan the Turbiscan Stability Index (TSI). TSI is a computation ByonBy means of the comparing of computation comparing theTurbiscan the profiles profiles in time, in time, the the instrument instrument determines determines the destabilization kinetics based on on the computation of transmission the Turbiscan Stability Index (TSI). The TSI issuperposition aiscomputation kinetics based the computation of the Turbiscan Stability Index (TSI). The TSI a computation based on the raw data of the transmission and backscattering signals. The superposition of the BS directly based on the raw data of the and backscattering signals. The of of directly based on the raw data of the transmission and backscattering signals. The superposition kinetics kinetics based based on on the the computation computation of the of the Turbiscan Turbiscan Stability Stability Index Index (TSI). (TSI). TheThe TSITSI is aiscomputation a computation directly based on thethe raw data ofwhile the transmission andand backscattering signals. The superposition of directly based on raw data of the transmission backscattering signals. The superposition of profiles is indicative of stability their deviation across time increases with the instability of the BS profiles is indicative of stability while their deviation across time increases with the instability the BS profiles is indicative of stability while their deviation across time increases with the instability directly directly based based on the on the rawraw data data of the of the transmission transmission and and backscattering backscattering signals. signals. TheThe superposition superposition of of the BS profiles is indicative of stability while their deviation across time increases with the instability the BS profiles is indicative of stability while their deviation across time increases with the instability the dispersions. TSI determined by using the of at given time. of TSIcan can be determined bytheir using thecomparison comparison of the the a at given time. of dispersions. the dispersions. The TSI can be determined bydeviation using theacross comparison ofscans thewith scans a given time. thethe the BS profiles BS profiles is The indicative isThe indicative of be stability of stability while while their deviation across time time increases increases with the the instability instability of the dispersions. The TSI cancan be determined by by using thethe comparison of the scans at aatgiven time. ofInthe dispersions. The TSI be determined using comparison of the scans a given time. In particular, the TSI is determined for the ith step [17]: particular, the TSI is determined for the ith step [17]: particular, the TSI is determined for the ith step [17]: of the of the dispersions. dispersions. TheThe TSITSI cancan be determined be determined by using by using the the comparison comparison of the of the scans scans at aatgiven a given time. time. In particular, thethe TSITSI is determined for for thethe ith ith stepstep [17]: In particular, is determined [17]: In particular, In particular, the the TSI TSI is determined is determined for for the the ith step ith step [17]: [17]: (1) (1) by comparing every scaniscan (h) of measurement to the one one scani−1 (h),i−1on by comparing every i(h)a of a measurement to previous the previous scan (h), the on selected the selected (1) (1)by comparing every scanscan i(h)i(h) of aofmeasurement to the previous oneone scanscan i−1(h), on on thethe selected by comparing every a measurement to the previous i−1(h), selected height h; height h; every (1) (1)by by comparing comparing every scanscan i(h) i(h) of aofmeasurement a measurement to the to the previous previous oneone scanscan i−1(h), i−1(h), on on the the selected selected height h; h; height height height h; h;
Energies 2017, 10, 354
6 of 15
by comparing every scani (h) of a measurement to the previous one scani −1 (h), on the selected height h; Energies 2017, 10, 354 6 of 14 (2) by dividing the result by the total selected height (H) in order to obtain a result which does not depend on the quantity of product in the measured cell: (2) by dividing the result by the total selected height (H) in order to obtain a result which does not depend on the quantity of product in the∑measured − scani−1 (h)] [scani (h) cell: (4) TSI = ∑ h H ∑ i (ℎ) − (ℎ) (4) = (1)
The TSI computation sums up all the variations in the sample, to give as a result a unique number The TSI computation sums up all variations in thethe sample, tostronger give as the a result a unique reflecting the destabilization occurring in athe sample. The higher TSI, the destabilization number reflecting the destabilization occurring sample. The higher the TSI, the stronger the in the sample. Consequently, an asymptotic valuein of aTSI could mean two things: destabilization in the sample. Consequently, an asymptotic value of TSI could mean two things: (1) The material material is is stable: stable: in in this this case, case, TSI TSI has has aa constant constant value value from from the (1) The the initial initial time time and and during during the the entire analysis. Usually, this happens for homogeneous materials. entire analysis. Usually, this happens for homogeneous materials. (2) The material material is is unstable unstable at at the the beginning beginning and and becomes becomes stable stable after after aa long long time time (more (more than than 24 24 h): h): (2) The in this case, the asymptotic value could be observed after a long time because the original sample in this case, the asymptotic value could be observed after a long time because the original changedchanged its texture. As a consequence, the material probably became stablestable but with verya sample its texture. As a consequence, the material probably became but awith different texture, with different overall overall thermalthermal properties, like conductivity, thermal capacity, very different texture, with different properties, like conductivity, thermal density, or other physical properties: It became a different material with respect to the initial capacity, density, or other physical properties: It became a different material with respect toone the and itsone thermal behavior as a PCM) needs to beneeds recalculated. initial and its thermal(i.e., behavior (i.e., as a PCM) to be recalculated.
the investigation investigationproposed proposedininthis thispaper, paper,the theanalyses analyses were conducted considering a period In the were conducted considering a period of ◦ of h 24and h and temperature the maximum the observation range °C) in 24 the the temperature was was fixedfixed at theatmaximum valuevalue in thein observation range (40 C)(40 in which which the sample in thephase. liquid phase. the sample remainsremains in the liquid 3. 3. Results Results and and Discussion Discussion By following modalities 1 or12,orit 2, is possible to observe significant and visible By preparing preparingthe thesample sample following modalities it is possible to observe significant and differences in their textures. In particular, the first sample (both for C1 and C2) has an appearance visible differences in their textures. In particular, the first sample (both for C1 and C2) has an more viscous thanviscous the second. Figure 1 shows the1comparison between the compositions C1(1) and appearance more than the second. Figure shows the comparison between the compositions C1(2). C1(1) and C1(2).
Figure 1. Sample with composition C1(1) in comparison to the sample with composition C1(2). Figure 1. Sample with composition C1(1) in comparison to the sample with composition C1(2).
Sample C1(1) appears like a dough, with a rheological behavior intermediate between a liquid Sample C1(1) appears like a dough, with a rheological a liquid and a solid system. On the contrary, sample C1(2) appears asbehavior a liquid. intermediate To the naked between eye, it appears to and a solid system. On the contrary, sample C1(2) appears as a liquid. To the naked eye, it appears to have a more homogeneous texture than C1(1), where the inhomogeneity is more evident. have a more homogeneous texture than C1(1), where the inhomogeneity is more evident. 3.1. Thermal Analysis of the Samples In Figure 2, the results of the T-history method applied at each sample Ci(j), are reported. They depict the heating and cooling cycles obtained in the refrigerated thermostat (the environment) where the temperature was changed from 40° to 10 °C following the imposed function, as described in Section 2.2. In the figure, the temperature evolution of the refrigerated thermostat where the samples are lodged (ambient temperature) as well as the temperature evolution in the PCM and water (reference
Energies 2017, 10, 354
7 of 15
3.1. Thermal Analysis of the Samples In Figure 2, the results of the T-history method applied at each sample Ci(j), are reported. They depict the heating and cooling cycles obtained in the refrigerated thermostat (the environment) where the temperature was changed from 40◦ to 10 ◦ C following the imposed function, as described in Section 2.2. In the figure, the temperature evolution of the refrigerated thermostat where the samples are lodged (ambient temperature) as well as the temperature evolution in the PCM and water (reference are reported. Energies 2017,substance) 10, 354 7 of 14
(a)
(b)
(c)
(d)
Figure 2. 2. Heating Heatingand andcooling cooling cycles all the samples, during the T-history method. (a): Sample Figure cycles forfor all the samples, during the T-history method. (a): Sample C1(1); C1(1); (b): sample C1(2); (c): sample C2(1); and (d): sample C2(2). (b): sample C1(2); (c): sample C2(1); and (d): sample C2(2).
It is possible to see the supercooling limited to 2–3 °C and the repeatability of the cycles. As It is possible to see the supercooling limited to 2–3 ◦ C and the repeatability of the cycles. reported in the literatures [10–13], cooling curves have been used for the second part of the method, As reported in the literatures [10–13], cooling curves have been used for the second part of the then by considering the cooling curves and Equation (3), the enthalpy versus temperature curves method, then by considering the cooling curves and Equation (3), the enthalpy versus temperature were determined for all the samples. The results obtained for sample C1 are shown in Figure 3. curves were determined for all the samples. The results obtained for sample C1 are shown in Figure 3. The thermal properties of the first sample, estimated by considering the method described in Section 2.2, are summarized in Table 4. The first composition has a percentage of sodium sulphate equal to 25%. In both modalities of preparation, a decrement of the temperature range of solidification with respect to the value of the pure salt is observed. This phenomenon could be correlated to the presence of additives that also determines an extension of the solidification interval. Such range of temperature is greater for the first composition and marks a major degree of inhomogeneity. Furthermore, it need to be taken into account that for sample C1(1), the water and bentonite were first mixed. Thus, it can be predicted that part of the water will be trapped by the bentonite silicate layers, thus reducing the water available to mix with the sodium sulphate, affecting melting temperature range.
Figure 3. Sample C1. Enthalpy vs. temperature curves with variation of the modality of preparation.
The thermal properties of the first sample, estimated by considering the method described in Section 2.2, are summarized in Table 4.
C1(1); (b): sample C1(2); (c): sample C2(1); and (d): sample C2(2).
It is possible to see the supercooling limited to 2–3 °C and the repeatability of the cycles. As reported in the literatures [10–13], cooling curves have been used for the second part of the method, then by considering the cooling curves and Equation (3), the enthalpy versus temperature curves Energies 2017, 10, 354 8 of 15 were determined for all the samples. The results obtained for sample C1 are shown in Figure 3.
Energies 2017, 10, 354
8 of 14
Figureof3. 3. the Sample C1. Enthalpy vs. with of of Figure Sample C1. will Enthalpy vs. temperature temperature curves with variation variation of the the modality modality of preparation. preparation. that part water be trapped by thecurves bentonite silicate layers, thus reducing the water available to mix with the sodium sulphate, affecting melting temperature range. The thermal properties of of the estimated considering the method described in Table 4. Thermal properties thefirst firstsample, sample C1 prepared by by means of the modalities (1) and (2). Section 2.2, 4. are summarized in Table 4. sample C1 prepared by means of the modalities (1) and (2). Table Thermal properties of the first ◦ The first composition has a percentage sodium equal In both cpl (kJ/kg ·K) modalities of Sample TiS ( C) TfS (◦ C) of∆H (kJ/kg)sulphate cps (kJ/kg ·K)to 25%. Sample TiS (°C) TfS (°C) ∆H (kJ/kg) cps (kJ/kg·K) cpl (kJ/kg·K) preparation, a decrement of the temperature range of solidification to the value of the C1(1) 19.26 122.28 4.10 with respect C1(1) 29.75 29.75 19.26 122.28 4.10 2.70 2.70 pure salt is observed. phenomenon correlated to3.24 the presence 3.63 of additives that also C1(2) This25.63 21.07 could be 87.99 C1(2) 25.63 21.07 87.99 3.24 3.63 determines an extension of the solidification interval. Such range of temperature is greater for the first composition and marks a major degree of inhomogeneity. Furthermore, it need to be taken into specific heats, heats,for forthe thesolid solidand andliquid liquid phase, larger than those Glauber’s The specific phase, areare larger than those for for the the purepure Glauber’s salt account that for sample C1(1), the water and bentonite were first mixed. Thus, it can be predicted salt reported in introduction the introduction [5].fact In the factexcess the excess of water is known to elevate the specific reported in the [5]. In of water is known to elevate the specific heat.heat. In Figure 4, the the enthalpies enthalpies of of fusion fusionof ofthe thetwo twosamples sampleswith withcomposition composition2 2are arecompared comparedand anda asignificant significantdifference differenceininthe thevalues valuesisisnoticed. noticed.
Figure 4. 4. Sample Figure Sample C2. C2. Enthalpy Enthalpy vs. vs. temperature temperature curves curves with with variation variation of of the the modality modality of of preparation. preparation.
The thermal properties of the second sample are presented in Table 5. The thermal properties of the second sample are presented in Table 5. Table 5. Thermal properties of the second sample C2 prepared by means of the modalities (1) and (2). Table 5. Thermal properties of the second sample C2 prepared by means of the modalities (1) and (2). Sample TiS (°C) TfS (°C) ∆H (kJ/kg) cps (kJ/kg·K) cpl (kJ/kg·K) (◦ C) c2.43 Sample (◦ C) TfS21.15 ∆H (kJ/kg) cps4.64 (kJ/kg·K) C2(1) TiS 28.06 72.65 pl (kJ/kg·K) C2(2) 30.97 24.45 171.9 3.62 3.80 C2(1) 28.06 21.15 72.65 4.64 2.43 C2(2)
30.97
24.45
171.9
3.62
3.80
Both the C2 compositions contain a percentage by weight of sodium sulphate equal to 31%, however the sample obtained by modality 1, C2(1), presents a thermal performance less effective Both the C2 compositions contain a percentage by weight of sodium sulphate equal to 31%, than that of sample C2(2) and also than that of the C1 samples that contain a minor percentage of however the sample obtained by modality 1, C2(1), presents a thermal performance less effective than sodium sulphate (25%). This result is probably due to the high viscosity of C2(1) that reduces the that of sample C2(2) and also than that of the C1 samples that contain a minor percentage of sodium efficacy of the sonication process: the same sonication power is not able to effectively break up the anhydrous sodium sulphate, reducing the formation of the decahydrate form (Glauber’s salt) and therefore the final thermal performance of the sample. In all the analyzed compositions, the size of supercooling is 2–3 °C and it can be considered very reduced in comparison to the values higher than 15 °C indicated in the literature for the pure salt [18], revealing the efficacy of the adopted nucleant.
Energies 2017, 10, 354
9 of 15
sulphate (25%). This result is probably due to the high viscosity of C2(1) that reduces the efficacy of the sonication process: the same sonication power is not able to effectively break up the anhydrous sodium sulphate, reducing the formation of the decahydrate form (Glauber’s salt) and therefore the final thermal performance of the sample. In all the analyzed compositions, the size of supercooling is 2–3 ◦ C and it can be considered very reduced in comparison to the values higher than 15 ◦ C indicated in the literature for the pure salt [18], revealing the efficacy of the adopted nucleant. Further consideration could be performed taking into account the phase diagram of the sodium sulphate–water system, as proposed by Biswas [19]. In order to apply such a diagram, the amount of Na2 SO4 has been calculated and the corresponding theoretical melting temperature values have been estimated for each sample. It is important to notice that the phase diagram is relative to a system without additives, therefore the amount of sodium sulphate has been recalculated assuming a system without bentonite and borax. For sample 1 where the amount of Sodium sulphate is 25%, the corresponding value for the phase diagram has been estimated at 27.17%. That corresponds to a condition in which the anhydrous sodium sulphate crystals are not expected in the studied range of temperature (10–40 ◦ C). Furthermore, the theoretical transition temperature is 25 ◦ C, and the value is within the experimental range reported in Table 4 for both C1(1) and C1(2). For sample 2, the amount of charged sodium sulphate is 31% and, assuming a system without additives, the corresponding value is 34.07%. In the phase diagram, this value corresponds to a limit condition and, beyond this value, the formation of anhydrous sodium sulphate is expected. With this value in the phase diagram, the theoretical transition temperature could be expected to be 32 ◦ C. This value is more similar to that obtained in Table 5, because the additives play a more significant role during the activity of Glauber’s salt as a PCM. With the amount of sodium sulphate and water reported in Table 1, it is possible to expect that the availability of water is greater for sample 1 than for sample 2. Bentonite and sodium sulphate compete for water. Sample 1 has enough water for both, and consequently, the theoretical melting point is in the observed range of temperature Tis and Tfs for both samples C1(1) and C1(2). On the other hand, for sample 2 where less water is available, the modality plays a more significant role. Modality 1 gives an immediate possibility for bentonite to hydrate itself with water: less water remains for the Glauber’s salt formation. Consequently, the temperature range where melting occurs is closer to a system where a smaller amount of hydrated salt is present. With modality 2, Glauber’s salt could be easily obtained during the first sonication. Residual water becomes available for the hydration of bentonite. As a consequence, thermal performance closer to Glauber’s salt could be reached, as confirmed for the temperature range shown in Table 5. Finally, the reproducibility of the data was confirmed from the evaluation of the mean error percentage between the data sets of each sample. With reference to the data shown in Figures 3 and 4, mean errors range from 4.45% and 5.56%. 3.2. Stability Analysis of the Samples The BS profiles of the samples C1(1) and C1(2) are shown in Figure 5. The composition C1(1) shows in each scan a percentage of the BS variable along the height of the cell, revealing the inhomogeneity of the sample. The composition C1(2) is rather homogeneus until 36 mm height; from this point, sedimentation and creaming phenomena become evident. Thus the sonication is more effective on the second dispersion that is less viscous. As a consequence, its solidification range (shown in Figure 3) is more limited (the first one is 10.49 ◦ C, the second one is limited to 4.56 ◦ C) while the total ∆H of the process is comparable. ∆BS profiles are presented in Figure 6 in order to underline the nature of these phenomena.
Finally, the reproducibility of the data was confirmed from the evaluation of the mean error percentage between the data sets of each sample. With reference to the data shown in Figures 3 and 4, mean errors range from 4.45% and 5.56%. 3.2. Stability Analysis of the Samples
Energies 2017, 10, 354
10 of 15
The BS profiles of the samples C1(1) and C1(2) are shown in Figure 5.
(a)
Energies 2017, 10, 354
10 of 14
The composition C1(1) shows in each scan a percentage of the BS variable along the height of the cell, revealing the inhomogeneity of the sample. The composition C1(2) is rather homogeneus until 36 mm height; from this point, sedimentation and creaming phenomena become evident. Thus the sonication is more effective on the second dispersion that is less viscous. As a (b) 3) is more limited (the first one is 10.49 °C, the consequence, its solidification range (shown in Figure second one isFigure limited to Backscattering 4.56 °C) while(BS) thepercentage total ΔH of process comparable. 5. (a) of the C1(1) sample;isand (b) of C1(2) sample. Figure 5.are (a)presented Backscattering (BS) percentage of C1(1) sample; (b) of ∆BS profiles in Figure 6 in order to underline theand nature of C1(2) thesesample. phenomena.
(a)
(b) Figure 6. (a) ∆BS of C1(1) sample; and (b) ∆BS of C1(2) sample. Figure 6. (a) ∆BS of C1(1) sample; and (b) ∆BS of C1(2) sample.
In sample C1(1), the flocculation phenomena are dominant and the shift of the trends up and down during the time depends on the particle diameter, smaller or larger than the incident radiation [17]. In sample C1(2), sedimentation, creaming, and flocculation are observed. The presence of sedimentation and creaming is probably due to the lower viscosity of the dispersion. Also for the composition C2, the second method of preparation makes the dispersions more
Energies 2017, 10, 354
11 of 15
In sample C1(1), the flocculation phenomena are dominant and the shift of the trends up and down during the time depends on the particle diameter, smaller or larger than the incident radiation [17]. In sample C1(2), sedimentation, creaming, and flocculation are observed. The presence of sedimentation and creaming is probably due to the lower viscosity of the dispersion. Also for the composition C2, the second method of preparation makes the dispersions more Energies 2017, 10, 354 11 of 14 homogeneous, as shown in Figure 7.
(a)
(b) Figure Figure 7. 7. (a) (a) BS BS percentage percentage of of C2(1) C2(1) sample; sample; and and (b) (b) of of C2(2) C2(2) sample. sample.
Unlike the previous case, a minor effect on the solidification temperature interval is observed, as shown in Figures 3 and 4. However for this composition, in both sequences of reagent additions, only the flocculation is reported. The absence of sedimentation and creaming is probably correlated to a higher viscosity of composition C2(2) in comparison to C1(2) thanks to the higher contents of bentonite and the lower percentage of water. ∆BS percentages, obtained for composition C2, are presented in Figure 8 for both modalities of preparation. The stability analysis has been quantified by considering the TSI behavior and comparing the trends obtained in 24 h. As already stated, such an index could be used as a measure of the progressive destabilization of the system. It is important to notice (a) that the optical testing obtained by Turbiscan is disruptive and non-repeatable. Thus, it is possible to test each sample one time during the 24 h period, with a scan of 5 min. It seems interesting to do the comparison between the repeated samples in terms of the TSI data scattering: the mean values of the errors in terms of TSI behaviour has been found, ranging from 6.25% to 10.2%. In Figure 9, the TSI values are reported with reference to sample C1 for the two different preparation modalities. It appears evident that modality 2 is not adequate for
Energies 2017, 10, 354
12 of 15
the preparation of Glauber’s salt, as discussed by observing the features of the samples in Figure 2. TSI values increase progressively until 15, for sample C1(2). This trend indicates that the sonication of (b) water and bentonite is preferable in the first step of preparation of such composition.. That allows us Figure 7. (a) BS percentage of C2(1) sample; and (b) of C2(2) sample. to achieve more homogeneous samples.
(a) Energies 2017, 10, 354
12 of 14
The stability analysis has been quantified by considering the TSI behavior and comparing the trends obtained in 24 h. As already stated, such an index could be used as a measure of the progressive destabilization of the system. It is important to notice that the optical testing obtained by Turbiscan is disruptive and non-repeatable. Thus, it is possible to test each sample one time during the 24 h period, with a scan of 5 min. It seems interesting to do the comparison between the repeated samples in terms of the TSI data scattering: the mean values of the errors in terms of TSI behaviour has been found, ranging from 6.25% to 10.2%. In Figure 9, the TSI values are reported with reference to sample C1 for the two different preparation modalities. It appears evident that modality 2 is not adequate for the preparation of Glauber’s salt, as discussed by observing the features of the samples in Figure 2. TSI values increase progressively until (b)15, for sample C1(2). This trend indicates that the sonication of water and bentonite is preferable in the first step of preparation of such composition.. ∆BS of of C2(1) C2(1) sample; sample; and and (b) ∆BS of C2(2) sample. Figure 8. 8. (a) ∆BS Figure That allows us to achieve more(a)homogeneous samples. (b) ∆BS of C2(2) sample.
Figure 9. Sample C1. Comparison of the time evolution of the Turbiscan Stability Index (TSI) as a Figure 9. Sample C1. Comparison of the time evolution of the Turbiscan Stability Index (TSI) as function a functionofofthe themodality modalityofofpreparation. preparation.
In order to evaluate the influence of composition, TSI behavior has been reported in Figure 10 In order to evaluate the influence of composition, TSI behavior has been reported in Figure 10 for for samples C1(1), C2(1), and C2(2). That allows us to compare the influence of composition on samples C1(1), C2(1), and C2(2). That allows us to compare the influence of composition on stability stability (samples C1(1) and C2(1)) and the influence of modality in the cases where more (samples C1(1) and C2(1)) and the influence of modality in the cases where more concentrated solutions concentrated solutions were used (samples C2(1) and C2(2)). It is interesting to note that the TSI for were used (samples C2(1) and C2(2)). It is interesting to note that the TSI for C1(1) is almost double C1(1) is almost double the TSI of C2(1). No significant difference can be observed for C2(1) and C2(2). That could be due to the fact that the sample C2 is characterized by a higher thickener (as bentonite)/water ratio, able to guarantee a greater stability of the solution.
Figure 9. Sample C1. Comparison of the time evolution of the Turbiscan Stability Index (TSI) as a function of the modality of preparation.
In order to evaluate the influence of composition, TSI behavior has been reported in Figure 10 for samples C1(1), C2(1), and C2(2). That allows us to compare the influence of composition on Energies 2017, 10, 354 13 of 15 stability (samples C1(1) and C2(1)) and the influence of modality in the cases where more concentrated solutions were used (samples C2(1) and C2(2)). It is interesting to note that the TSI for C1(1) the TSIdifference of C2(1). can No be significant can C2(2). be observed for C2(1) the TSIisofalmost C2(1). double No significant observeddifference for C2(1) and That could be dueand to C2(2). be due fact that the is characterized by a higher thickener the factThat that could the sample C2to is the characterized by sample a higherC2 thickener (as bentonite)/water ratio, able(as to bentonite)/water ratio, able to a greater stability of the solution. guarantee a greater stability of guarantee the solution.
Figure 10. 10. Comparison composition and modality of Figure Comparison of of the the time timeevolution evolutionofofTSI TSIasasa afunction functionofofthe the composition and modality preparation. of preparation.
In any case, samples C2(1) and C2(2) are more stable than C1(1). However, sample C2(2) offers better thermal performances in comparison to C2(1), and consequently it should be preferred for further studies. 4. Conclusions In this paper, samples of PCMs based on Glauber’s salt dispersions have been studied, by changing their composition and modality of preparation with the aim of estimating, by the T-history method, their thermal properties, and by means of optical methods, their stability. It has been observed that both compositions and modalities of preparation significantly affect the phase change enthalpy, melting temperature range, and structural stability. This is due to how the components are distributed in the PCM solutions at different thermal conditions, affecting both dispersion and viscosity as also verified for other families of heterogeneous materials [20,21]. This is more relevant when PCMs are subject to numerous thermal cycles or when crystallization or supercooling are involved. In particular, the nucleant selected in order to limit the supercooling has produced satisfactory results in all the analyzed cases, as it is able to reduce this phenomenon to a few degrees celcius. With reference to the thickener, it was observed that the stability of the dispersion improves for the samples that present a greater bentonite/water ratio, thanks to the higher viscosity. In defining the optimal composition of Glauber’s salt and additives, the results underlined that the thermal performances and the stability of the dispersion cannot be defined by considering only the percentage of the composition. The final properties depend on the preparation method and in particular, on the sequence used for adding the reagents and on the viscosity of the mixture. In any case, the aim of this paper is to provide an example of a procedure useful for testing the quality of inhomogeneous samples by considering both thermal performance and stability. The different compositions and modalities of preparation have been tested in order to underline the different behaviors of PCMs in relation to these variables. We suggest the use of this method as a way to characterize PCMs and, consequently, as a method of preparation and composition optimization. Of course, further studies should be carried out at different lifetimes, depending on the purpose of the investigation. However, it is important to verify how the preparation methodology can affect the initial thermal properties and the stability of samples.
Energies 2017, 10, 354
14 of 15
Acknowledgments: This publication is part of a project that has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 657466 (INPATH-TES). Author Contributions: Maria Gabriela De Paola carried out the experimental activity. She realized the T-history system and she introduced the use of Turbiscan as method for the characterization of PCM dispersions. Furthermore she carried out, with other authors for the analysis and elaboration of data. Finally, she was responsible of the reference selection. Natale Arcuri cooperated with the acquisition of experimental instrument and during the preliminary part of the research. He was also involved in the study of literature. Vincenza Calabrò, corresponding author, followed and coordinated the experimental activity. She was mainly involved in the analysis and elaboration of experimental data in order to describe the phenomena that occurred. Furthermore she has worked as in the research setup phases, as for writing and revision of the manuscript. Marilena De Simone, responsible of the European Union’s Horizon 2020 research and innovation program under grant agreement No. 657466 (INPATH-TES), was involved in the analysis and elaboration of experimental data. She plays a significant role for manuscript preparation and for writing and revising it until the publication. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5.
6. 7.
8. 9.
10. 11. 12.
13. 14.
15.
16.
Atanasova, D.; Lyakov, N.; Vassilev, V.; Angelov, G.; Haralampiev, G. Desulphurization of lead cake by sodium carbonate and sodium hydroxide. J. Univ. Chem. Technol. Metall. 2008, 43, 267–272. Atanasova, D.A. Hydrometallurgical processing of dumped lead paste for lead acid batteries. Bulg. Chem. Commun. 2009, 41, 285–296. Telkes, M. Thixotropic Mixture and Method of Making Same. U.S. Patent 3,986,969 A, 19 October 1976. Derluyn, H.; Saidov, T.A.; Espinosa-Marzal, R.M.; Pel, L.; Scherer, G.W. Sodium sulfate heptahydrate I: The growth of single crystals. J. Cryst. Growth 2011, 329, 44–51. [CrossRef] Eckoff, S.; Okos, M. Solar Energy Heat Storage for Home, Farm and Small Business: Suggestions on Selecting and Using Thermal Storage Materials and Facilities. Available online: https://www.extension.purdue.edu/ extmedia/ae/ae-89.html (accessed on 22 September 2016). Mehling, H.; Cabeza, L.F. Heat and Cold Storage with PCM: An up to date Introduction into Basics and Applications; Springer: Berlin/Heidelberg, Germany, 2008. Feng, G.; He, N.; Li, G.; Chen, Q.; Li, H.; Huang, K. Experimental research of anhydrous Na2 SO4 +H2 O phase change energy storage system. In Proceedings of the APEC Conference on Low-carbon Towns and Physical Energy Storage, Changsha, China, 25–26 May 2013. Prabhu, P.A.; Shinde, N.N.; Patil, P.S. Review of phase change materials for thermal energy storage applications. Int. J. Eng. Appl. 2012, 2, 871–875. Ultrasonic Production of Nano-Size Dispersions and Emulsions. Available online: http://www. euronanoforum2011.eu/wp-content/uploads/2011/10/enf2011_production-nanostructures_hielscher_fin.pdf (accessed on 22 September 2016). Zhang, Y.; Jiang, Y. A simple method, the T–history method, of determining the heat of fusion, specific heat and thermal conductivity of phase-change materials. Meas. Sci. Technol. 1999, 10, 201–205. Stankovic, S.B.; Kyriacou, P.A. Improved measurement technique for the characterization of organic and inorganic phase change materials using the T-history method. Appl. Energy 2013, 109, 433–440. [CrossRef] Marín, J.M.; Zalba, B.; Cabeza, L.F.; Mehling, H. Determination of enthalpy–temperature curves of phase change materials with the temperature-history method: Improvement to temperature dependent properties. Meas. Sci. Technol. 2013, 14, 184–189. [CrossRef] Hong, H.; Kim, S.K.; Kim, Y.S. Accuracy improvement of T-history method for measuring heat of fusion of various materials. Int. J. Refrig. 2004, 27, 360–366. [CrossRef] Mengual, O.; Meunier, G.; Cayre, I.; Puech, K.; Snabre, P. Characterisation of instability of concentrated dispersions by a new optical analyser: The TURBISCAN MA 1000. Colloids Surf. A 1999, 152, 111–123. [CrossRef] Mengual, O.; Meunier, G.; Cayre, I.; Puech, K.; Snabre, P. Turbiscan MA 2000: Multiple light scattering measurement for concentrated emulsion and suspension instability analysis. Talanta 1999, 50, 445–456. [CrossRef] De Paola, M.G.; De Simone, M.; Arcuri, N.; Calabrò, V. Crossed analysis by T-history and Turbiscan for the characterization of PCM with Glauber salt. In Proceedings of the INNOSTORAGE Conference, Beer-Sheva, Israel, 16–18 February 2016.
Energies 2017, 10, 354
17. 18.
19. 20. 21.
15 of 15
Data Interpretation for Turbiscan Analysis. Available online: http://files.instrument.com.cn/FilesCenter/ 20061226/SH101155-36019.pdf (accessed on 2 March 2017). García-Romero, A.; Diarce, G.; Ibarretxe, J.; Urresti, A.; Sala, J.M. Influence of the experimental conditions on the subcooling of Glauber’s salt when used as PCM. Sol. Energy Mater. Sol. Cells 2012, 102, 189–195. [CrossRef] Biswas, D.R. Thermal Energy Storage using sodium sulphate decahydrate and water. Sol. Energy 1977, 19, 99–100. [CrossRef] Lupi, F.R.; Gabriele, D.; Baldino, N.; Seta, L.; de Cindio, B.; De Rose, C. Stabilization of meat suspensions by organogelation: A rheological approach. Eur. J. Lipid Sci. Technol. 2012, 114, 1381–1389. [CrossRef] Seta, L.; Baldino, N.; Gabriele, D.; Lupi, F.R.; de Cindio, B. Rheology and adsorption behaviour of β-casein and β-lactoglobulin mixed layers at the sunflower oil/water interface. Colloids Surf. A Physicochem. Eng. Asp. 2014, 441, 669–677. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).