Sepiolite supported stearic acid composites for thermal energy storage

Abstract

In this paper, novel composite phase change materials (PCMs) were prepared by absorbing stearic acid (SA) into sepiolite (α-sepiolite, β-sepiolite) via a vacuum impregnation method. The microstructures, thermal properties and thermal reliability of the composite PCMs were characterized by SEM, TG and DSC. The results indicate that the maximum SA absorption of α-sepiolite (α-SPL) and β-sepiolite (β-SPL) could reach as high as 60% and 49%. The latent heat of α-SPL/SA and β-SPL/SA were 118.7 J g⁻¹ and 95.8 J g⁻¹, respectively. The thermal cyclic test indicated α-SPL/SA and β-SPL/SA composites have excellent structural stability and thermal reliability. The thermal performance of the composite PCM gypsum boards was also evaluated. The composite PCMs could have a great potential application for thermal energy storage.

---

1. Introduction

With the development of our society, the consumption of energy is increasing. After the oil crisis happened in the 1970s, solar radiation has been considered as one of the most prospective energy sources. But solar energy is intermittent, and can be utilised on a large scale only if high efficiency storage technology is developed.¹ Thermal energy storage, which includes sensible heat storage and latent heat storage, provides a convenient and useful way to reduce the consumption of energy in the building sector and to improve the efficiency of the heat storage. However, the sensible heat storage has apparent defects, it requires large volumes and the temperature which the energy release is vary. Phase change materials (PCMs) for latent heat storage have been carried out due to their high thermal energy storage, heat storage density and release at a constant temperature range.^2–4 These characteristics make the latent heat thermal energy storage (LHTES) system widely used in the area of solar energy storage, industrial conservation and building. It's possible to store a large amount of thermal energy in building without large structural mass associated with sensible heat storage by using PCMs as LHTES materials.⁵

PCM-enhanced materials can allow the storage of large amounts of thermal energy without transforming the temperature. During the last two decades, various of candidate organic and inorganic PCMs, such as fatty acid esters, paraffin waxes, non-paraffin organic compounds, binary and ternary mixtures and salt hydrates, have been researched as PCMs for LHTES applications on account of their disparate phase change intervals.^6–10 As a linear chain fatty acid, stearic acid (SA) was found to possess many satisfactory characteristics in the investigated PCMs.^11–14 As a consequence, it can be use as thermal storage materials for factional fluid, passive solar space heating and to reduce indoor energy wave in buildings. However, the leakage of organic PCMs during the phase change process limits their use to some scopes. The composite PCMs have been studied to prevent the leakage by absorption of the suitable PCMs into porous materials.^15–22 In recent years, the composite PCMs have attracted the attention because of the main benefits of easily producibility by inexpensive technology and directly usage. The researches have been investigated on the thermal energy storage characteristics of various composite PCMs, such as expended perlite,²³ vermiculite,^24,25 kaolin,^26,27 pumice^28–34 and diatomite.^35,36 These studies showed that the composite PCMs have the great potential applications for thermal energy storage.

Sepiolite (SPL), a hydrated magnesium silicate (Si₁₂Mg₈O₃₀(OH)₄(OH₂)₄·8H₂O), is a kind of natural fibrous clay mineral with discontinuity interior tunnels and blocks along the fiber direction. According to the crystal shape, SPL is divided into α-sepiolite (α-SPL) and β-sepiolite (β-SPL), the former is large bundle of fibrous crystals and the latter is short and thin fibrous crystals. The block structure of SPL consists of a central magnesia sheet and two tetrahedral silica sheets, and the alternation of the silica sheets causes the tunnel structure.^37–39 On account of the special structure, easy availability, low cost and abundant, SPL has been widely used in agriculture, geology, industry and adsorptive support.^40–45 Features of pore structure, such as the pore diameter distribution, pore volume, and the BET surface area, are the main factors for the selection of supports for impregnating PCMs. The pore structure of SPL is far better than that of the supports such as kaolin, bentonite, etc. With all the mentioned above, SPL can be an appropriate candidate as support to prepare composite PCMs for LHTES.⁴⁶

In this study, α-SPL/SA and β-SPL/SA were prepared as the novel composite PCMs by the vacuum impregnation method. Furthermore, the structure characteristics, thermal stability, phase change properties and thermal reliability of α-SPL/SA and β-SPL/SA composites were investigated.

2. Experimental

2.1 Materials

Stearic acid (C₁₈H₃₆O₂) was analytically pure and supplied by Shanghai Sinopharm Chemical Reagent Co., Ltd., China. The α-SPL and β-SPL were collected from mineral deposit located in Hebei and Hunan Province in China. They were ground in a grinder and sieved using a 74 μm sieve and the dried at 100 °C for 24 h before use. Gypsum powder was purchased from the Xiongshi Construction Material Plant, China.

2.2 Characterization

The microstructure and morphology were characterized by X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR) and scanning electron microscopy (SEM). The XRD experiments were performed on the specimen using a RigakuD/max 2550 (40 kV, 40 mA) with Cu Kα (λ = 0.15406 nm) irradiation at a scan rate of 2° min⁻¹ in the range of 5–80°. FTIR spectra of the samples were obtained in the range of 4000–500 cm⁻¹ on a Nicolet 5700 spectrophotometer. Specimens used for FTIR measurement were performed by mixing the KBr with the sample powder and the mixtures were pressed into pellets. The SEM images of the samples were investigated using a JEOL JSM-6360LV SEM. Before SEM image, the samples were ground and the surfaces were covered with gold. Thermal conductivities of the samples were investigated by hot-wire method (TC 3000 heat conduction modulus testing instrument). 4 g composites were pressed at 20 MPa for 1 min in the mould (the diameter of the mould is 20 mm). And then the made-up samples were identified in the TC 3000 heat conduction modulus testing instrument. Thermal properties and thermal stability of the composite PCMs were determined by thermo-gravimetric analysis (TGA), differential scanning calorimetry (DSC) analysis. TGA (TA Instruments, STA8000) was performed at a heating rate of 10 °C min⁻¹ up to 600 °C in a nitrogen atmosphere. DSC analysis (TA Instruments, Q10) was carried out in the range of 25–80 °C, at a rate of 5 °C min⁻¹ under argon flow at atmospheric pressure. To evaluate the pore structures of the supports, nitrogen gas adsorption–desorption isothermals were carried out at 77 K using an ASAP 2020 unit. The composite PCMs gypsum boards were prepared and its thermal energy storage performance was also evaluated.

2.3 Preparation of the α-SPL/SA and β-SPL/SA composites

The composite PCMs were prepared by vacuum impregnation method. First, 20 g SA was placed in a conical flask with 10 g α-SPL, a round flask was used to connect a vacuum pump to the conical flask for preventing backward suction. The conical flask was heated at 90 °C and the vacuum was evacuated to −0.1 MPa for 30 min. Then, the vacuum pump and air entry was closed to forced the SA to penetrate into the α-SPL, with ultrasonic heating at 90 °C for 5 min. To remove excess SA by thermal filtration, the composite PCMs were maintained at 90 °C for 48 h as the mass did not decrease longer. After cooling, the mixture was ground to obtain the α-SPL/SA composite. Furthermore, the β-SPL/SA composite was also prepared by the method mentioned above.

2.4 Preparation of the composite PCMs gypsum boards

The composite PCMs gypsum boards were prepared by mixing the composite PCMs at mass fraction of 20% with the gypsum slurry (water/gypsum ratio of 0.5). To achieve the homogeneous distribution of components, the mixtures were uniformly stirred and placed into a plastic mold (150 mm × 150 mm × 10 mm). The mold retained at room temperature for 24 h after being vibrated for 5 minutes. Finally, the mold was removed and the molded composite PCMs gypsum boards were dried in an oven at 50 °C for 24 h. Ordinary gypsum board without composite PCMs was also prepared by the way mentioned above. The thermal energy performance of the gypsum boards was determined by a self-designed heating setup. The setup consisted of a computer-recording system linked with a data-logger, a 275 W infrared lamp (the heating source) situated in 250 mm above the top pan, and a small test room (150 mm × 150 mm × 150 mm). The temperature recorder was fixed on the inner surface of the gypsum boards and the centre of the test room to monitor the change of temperature with time.

3. Results and discussion

XRD patterns and FTIR spectra of α-SPL, β-SPL, SA and α-SPL/SA, β-SPL/SA composites are shown in Fig. 1. The characteristic peaks of α-SPL appear at 7.3°, 27.8° and 35.0°, those of β-SPL appear at 7.2°, 26.5° and 36.2°, and that of calcite (the main impurity in the natural SPL) appears at 29.4° in the XRD pattern of supports (Fig. 1a), it indicates that α-SPL and β-SPL have the similar crystalline structure. As porous supporting materials, the crystal structure of α-SPL and β-SPL had not been affected during the preparation of the composite PCMs. After SA was permeated into the supports, the characteristic peaks of supports and SA appear in the XRD pattern of the composite PCMs, showing that SA has been successfully impregnated into the α-SPL and β-SPL. Therefore, the crystallization states of α-SPL/SA and β-SPL/SA composites are well preserved and stable.

Fig. 1 (a) XRD patterns and (b) FTIR spectra of α-SPL, β-SPL, SA, α-SPL/SA and β-SPL/SA composites.

The chemical structure and the specific interactions of α-SPL/SA and β-SPL/SA composites were investigated by FTIR spectroscope. As shown in Fig. 1b, the FTIR spectrum of α-SPL has a stretching vibration of MgOH (dioctahedral) at 3615 cm⁻¹ and a stretching vibration of –OH from coordinated water at 3563 cm⁻¹. The peak at 1662 cm⁻¹ represents the bending vibration of –OH. The bands at 877 cm⁻¹ and 666 cm⁻¹ represent the bending vibration of carbonate and Mg₃OH. The FTIR spectrum of β-SPL is similar to that of the α-SPL, which indicates that α-SPL and β-SPL have the similar chemical structure. In the pure SA spectrum, the absorption peaks at wave number of 2919 cm⁻¹ and 2849 cm⁻¹ represent the symmetrical stretching vibration of –CH₃ and –CH₂, respectively. The sharp absorption peak at 1898 cm⁻¹ corresponds to the C=O stretching vibration. The FTIR spectra of the composite PCMs contain the characteristic peaks of the supports and SA, no significant new peaks were observed in the FTIR spectrum of the composite PCMs, indicating no chemical reactions between the supports and SA, the bending features between supports and SA were mainly hydrogen bonding and physical interaction, which had no influence on the structure of α-SPL and β-SPL during the melting and freezing process. Thus, α-SPL/SA and β-SPL/SA composites have stable crystal structure and chemical structure for thermal energy storage.

The morphologies of α-SPL, β-SPL, α-SPL/SA and β-SPL/SA composites at microscale, and the corresponding EDS patterns of the composite PCMs are shown in Fig. 2. The α-SPL and β-SPL have a fibrous morphology and the surface of fibers is full of trench (Fig. 2a & e). α-SPL fiber is longer and β-SPL fiber bundles are agglutinated as a bed, which leads to the difference specific surface area and the pore volume of two supports. When considered the composite PCMs images, the surfaces of α-SPL and β-SPL were occupied entirely by SA chain, which can account for the adsorption of SA into the supports successfully, and a sharp rise was clearly observed in the number of carbon atom because a large amount of SA was impregnated into α-SPL and β-SPL (Fig. 2d & h).

Fig. 2 Microstructures of: (a) α-SPL, (b) α-SPL/SA, (c) α-SPL/SA after 200 thermal circles, (e) β-SPL, (f) β-SPL/SA, (g) β-SPL/SA after 200 thermal circles, and the corresponding EDS patterns of: (d) α-SPL/SA, (h) β-SPL/SA.

The thermal conductivity is an important indicator for the applicability of phase change materials. Low thermal conductivity severely reduces the rate of heat storage and release during the melting and freezing cycles. Thermal conductivity measurements were conducted at room temperature and the results were the average of three measurements. The thermal conductivity of pure SA was 0.26 W m⁻¹ K⁻¹. The thermal conductivity of α-SPL/SA was 0.57 W m⁻¹ K⁻¹, which increased by 119% of the pure SA. The thermal conductivity of β-SPL/SA was 0.76 W m⁻¹ K⁻¹, which increased by 192% of the pure SA (Fig. 3). The thermal conductivities of the α-SPL/SA and β-SPL/SA composites were obviously improved when SA was impregnated in supports pores.

Fig. 3 Thermal conductivities of the samples.

The thermo-gravimetric analysis curves and DSC curves of SA, α-SPL/SA and β-SPL/SA composites are shown in Fig. 4. There are no obvious weight loss and decomposition reaction in the range of 25 °C and 180 °C (Fig. 4a). It reveals the α-SPL/SA and β-SPL/SA can be used repeatedly below 180 °C. The sharp weight loss from 180 °C to 300 °C is due to SA evaporation. The weight losses for the α-SPL/SA and β-SPL/SA below 600 °C were 70% and 53%, respectively, which included the water evaporation of supports and SA evaporation. According to the TG analysis, the weight losses for the α-SPL and β-SPL below 600 °C were 11% and 5%, and the weight loss for the SA below 600 °C was 98%, which can be concluded that the maximum mass fractions of SA absorbed into the α-SPL and β-SPL were 60% ((70% − 11%)/98%) and 49% ((53% − 5%)/98%), respectively. Therefore, α-SPL/SA and β-SPL/SA composites have great thermal stability in the operating temperature range.

Fig. 4 (a) TG curves and (b) DSC curves of SA, α-SPL/SA and β-SPL/SA composites.

Phase change properties including melting and freezing temperature and the heat storage capacity were measured by DSC thermal analysis. The DSC curves are shown in Fig. 4b. The latent heat was derived from DSC curves by calculating the peak area. The latent heats of melting and freezing are calculated from the area of endothermic peak and exothermic peaks. The data from the DSC thermograms were presented in Table 1. Combining Fig. 4b with Table 1, it can be seen that the melting temperatures of SA absorbed into pore of supports decreased slightly, which could be attributed to the weak interaction. The interactions in composite PCMs include van der Waals forces and hydrogen bonding, which is in keeping with the FTIR results that there were no chemical reactions between the supports and SA. Therefore, the lower melting temperatures are observed in α-SPL/SA and β-SPL/SA composites. The freezing temperatures of SA and β-SPL/SA are 69.1 °C and 68.2 °C, but the freezing temperature of α-SPL/SA is 60.1 °C. Actually, the interaction between SA and porous supports of capillary and surface tension forces could affect the rate of crystallization, the larger capillary and surface tension forces between SA and α-SPL is contributed to the larger BET surface area of α-SPL (Table 2), which then decreases the rate of crystallization in the α-SPL/SA composites, so, the α-SPL/SA has the lower freezing temperature than SA and β-SPL/SA. The latent heat values are less than the theoretic values (for α-SPL/SA, ΔH_m = 206.1 J g⁻¹ × 60% = 123.7 J g⁻¹; for β-SPL/SA, ΔH_m = 206.1 J g⁻¹ × 49% = 101.0 J g⁻¹). The slight decreases of the latent heat values were caused by the crystallizing of SA, which was hindered for interactions between SA and supports and cut down the latent heats of the composite PCMs.⁴⁷

Table 1 DSC data of the SA, α-SPL/SA and β-SPL/SA composites

Sample	Melting temperature (Tm, °C)	Latent heat of melting (J g^−1)	Freezing temperature (Tf, °C)	Latent heat of freezing (J g^−1)
SA	70.8	206.1	69.1	201.2
α-SPL/SA	68.0	118.7	60.1	114.9
β-SPL/SA	67.1	95.8	68.2	94.4

---

Table 2 Physical properties of α-SPL and β-SPL

Samples	Average pore diameter (nm)	Total pore volumes (cm^3 g^−1)	Surface area (m^2 g^−1)
α-SPL	7.66	0.21	108.68
β-SPL	9.46	0.12	51.58

---

In order to further explore the reason of the large latent heat of α-SPL/SA and β-SPL/SA, the pore structures of α-SPL and β-SPL were measured. The nitrogen adsorption–desorption isotherms of the α-SPL and β-SPL are shown in Fig. 5, the II type isotherms indicated that the mesopores conform to IUPAC classification. Insert of Fig. 5 shows the pore size distributions of the α-SPL and β-SPL in the mesoporous range were between 2 and 20 nm. The average pore diameter, total pore volume, and BET surface area of the α-SPL and β-SPL are summarized in Table 2. Combining Fig. 5 with Table 2, the pore volume of α-SPL is larger than that of β-SPL, indicating that α-SPL has more adsorption site. Therefore, the SA segments in the α-SPL/SA are easier to move and crystallizable, showing that SA in the α-SPL/SA composite has relatively higher crystallinity than that in the β-SPL/SA composite, which results in the larger latent heats.⁴⁸

Fig. 5 Nitrogen adsorption–desorption isotherms and pore size distribution of α-SPL and β-SPL.

In addition, α-SPL and β-SPL have the excellent physical properties and have advantages for adsorption of SA on account of capillary and surface tension forces. The comparison of thermal properties of α-SPL/SA and β-SPL/SA composites with that of the composite PCMs reported in literatures are presented in Table 3. The latent heat capacities of α-SPL/SA and β-SPL/SA composites are larger than those of other composite PCMs, which is due to the larger pore volume and BET surface area of supports and the latent heat of SA.

Table 3 Comparison of thermal properties of α-SPL/SA and β-SPL/SA composites with that of the composite PCMs reported in literature

Composite PCM	Melting temperature (°C)	Latent heat (J g^−1)	Incorporated rate (%)	References
Stearic acid/bentonite	53.2	48.4	37	49
Capric–stearic acid/activated-attapulgiate	21.8	72.6	50	50
Stearic acid/activated montmorillonite	59.9	84.4	48	51
Capric–stearic acid/gypsum	23.8	49.0	25	52
Stearic acid/kaolin	53.9	59.3	37	53
Paraffin/diatomite	41.1	70.5	47	54
β-SPL/SA	67.1	94.4	49	This work
α-SPL/SA	68.0	118.7	60	This work

---

According to the above results, an overall schematic presentation for preparing α-SPL/SA and β-SPL/SA composites is shown in Fig. 6. The α-SPL has the longer fibrous bundle than β-SPL, and the block structure of α-SPL and β-SPL consists of a central magnesia sheet and two tetrahedral silica sheets and the alternation of the silica sheets cause the tunnel structure. With the vacuum impregnation, SA was fully dispersed in the α-SPL and β-SPL interior tunnel structure by capillary and surface tension forces. It indicates that the SPL-base composite PCMs can be considered as efficient energy storage materials.

Fig. 6 Schematic representation for preparing α-SPL/SA and β-SPL/SA composites.

The composite PCMs must be chemically and thermally stable, no or less transform in its chemical structure and thermal properties after long-term use. The thermal cyclic test (200 melting/freezing cycles) was performed to investigate the chemical and thermal reliability of the composite PCMs. Fig. 7 shows the FTIR spectra and DSC curves of α-SPL/SA and β-SPL/SA composites before and after 200 thermal cycles. FTIR was performed to investigate the chemical reliability. None of the shapes or frequencies of characteristic peaks changed obviously of α-SPL/SA and β-SPL/SA composites after 200 thermal cycles (Fig. 7a). It showed that no chemical structure was influenced and no reaction happened during thermal cycles. SEM was performed to investigate the thermal reliability. The SEM photographs of α-SPL/SA and β-SPL/SA after the thermal cycling experiments were shown in Fig. 2c & g. The morphology of the fiber structure was intact, SA was fully dispersed in the α-SPL and β-SPL pores and nearly no empty pores were observed, and α-SPL/SA and β-SPL/SA composites showed nearly no change after 200 thermal circles. The DSC was also performed to investigate the thermal reliability. After thermal cyclic test, the phase change temperatures of α-SPL/SA and β-SPL/SA composites changed by 0.1% and 1.8%, respectively, and the latent heats of α-SPL/SA and β-SPL/SA composites changed by 0.2% and 1.7%, respectively (Fig. 7b). The results approximate the instrumental uncertainty (±1%). Therefore, the α-SPL/SA and β-SPL/SA composites have excellent thermal reliability and can be used repeatedly in the operating temperature range.

Fig. 7 (a) FTIR spectra and (b) DSC curves of α-SPL/SA and β-SPL/SA composites before and after 200 thermal cycles.

The thermal performance of the composite PCMs gypsum boards was further studied. The thermal performance comparisons of α-SPL/SA and β-SPL/SA composite gypsum boards with ordinary gypsum board are demonstrated in Fig. 8. Compared with the ordinary gypsum board, the inner surface temperature of the composite PCMs gypsum boards is obviously lower during the heating process, but higher during the cooling process (Fig. 8a). It indicates that the composite PCMs gypsum boards have the energy absorption and release characteristics during the heating and cooling process. The comparisons of ordinary gypsum board shows that the maximum inner surface temperature with the α-SPL/SA and β-SPL/SA composite gypsum boards are 5.8 °C and 5.0 °C, which reveals that the α-SPL/SA and β-SPL/SA composite gypsum boards have obvious thermal energy storage properties at the higher heating temperature than melting temperature. Fig. 8b shows the test room central air temperature comparisons of ordinary gypsum board with the composite PCMs gypsum boards. The obtained temperature variation trends of the composite PCMs gypsum boards and ordinary gypsum board (Fig. 8b) are similar to that in Fig. 8a. Therefore, the composite PCMs gypsum boards have better endothermal and exothermal properties, preferable thermal storage performance. It indicates that the α-SPL/SA and β-SPL/SA composite gypsum boards have advantages for thermal energy storage applications.

Fig. 8 Thermal performance comparisons of α-SPL/SA and β-SPL/SA composite gypsum boards with ordinary gypsum board: (a) inner surface temperature and (b) test rooms central temperature.

The novel composite PCMs could be suitable for the application in the roof and outer wall in the summer of hot regions with large temperature difference between day and night, such as inland areas close to desert. The highest direct sunlight temperatures in these regions were generally higher than the phase change temperature in summer, while the nighttime temperatures dropped. The composite PCMs could store thermal energy during the day, and release thermal energy against the cold at night. Therefore, the composite PCMs could reduce the indoor temperature swing and save the building energy consumption. But the novel composite PCMs still involve some drawbacks, the relatively low thermal conductivity may limit its application in some areas.

Additionally, the economics of the novel composite PCMs were further investigated. As the natural mineral, sepiolite only costs 100 USD per ton, and SA costs 800 USD per ton. Generally, the composite PCMs, prepared by using n-octadecane as organic PCM within porous supports, nearly cost 500 000 USD per ton. Microencapsulated different PCM with TiO₂ shell prepared by a sol–gel method also costs 200 000 USD per ton. So, the novel composite PCM shows cheaper than current composite PCMs, showing obvious advantages for thermal energy storage applications.

4. Conclusions

The composite PCMs were prepared by vacuum impregnation of SA into α-SPL and β-SPL. The maximum mass fractions of SA absorbed into the α-SPL and β-SPL reached 60% and 49%, respectively, without the leakage of melted SA from the composite PCMs. α-SPL/SA and β-SPL/SA composites have large latent heat capacity and excellent thermal reliability, and the composite PCMs gypsum boards show better endothermal and exothermal properties and energy storage ability, indicating that α-SPL/SA and β-SPL/SA composites could have a great potential application for thermal energy storage.

Acknowledgements

This work was supported by the National Science Fund for Distinguished Young Scholars (51225403), the Hunan Provincial Science and Technology Project (2016RS2004, 2015TP1006), the Innovation Driven Plan of Central South University (2016CX015) and the ShengHua Scholar Project of CSU.

Notes and references

1. M. K. Rathod and J. Banerjee, Renewable Sustainable Energy Rev., 2013, 18, 246–258.
2. M. Li and Z. Wu, Renewable Sustainable Energy Rev., 2012, 16, 2094–2101.
3. Y. Yuan, N. Zhang, W. Tao, X. Cao and Y. He, Renewable Sustainable Energy Rev., 2014, 29, 482–496.
4. K. Pielichowska and K. Pielichowski, Prog. Mater. Sci., 2014, 65, 67–123.
5. A. Sharma, A. Shukla, C. R. Chen and S. Dwivedi, Energy and Buildings, 2013, 64, 403–407.
6. Y. Yuan, N. Zhang, T. Li, X. Cao and W. Long, Energy, 2016, 97, 488–497.
7. F. Kuznik, D. David, K. Johannes and J. J. Roux, Renewable Sustainable Energy Rev., 2011, 15, 379–391.
8. K. Peng, L. Fu, J. Ouyang and H. Yang, Adv. Funct. Mater., 2016, 26, 2666–2675.
9. X. Li, Q. Yang, J. Ouyang, H. Yang and S. Chang, Appl. Clay Sci., 2016, 126, 306–312.
10. Y. Zhang, M. Long, P. Huang, H. Yang, S. Chang, Y. Hu, A. Tang and L. Mao, Sci. Rep., 2016, 6, 33335.
11. K. Peng, L. Fu, H. Yang, J. Ouyang and A. Tang, Nano Res., 2016 DOI:10.1007/s12274-016-1315-3.
12. K. Peng, L. Fu, H. Yang and J. Ouyang, Sci. Rep., 2016, 6, 19723.
13. B. Zalba, J. M. a. Marín, L. F. Cabeza and H. Mehling, Appl. Therm. Eng., 2003, 23, 251–283.
14. A. Sari and K. Kaygusuz, Sol. Energy, 2001, 71, 365–376.
15. C. Zhi, S. Di, M. Qin and G. Fang, Energy and Buildings, 2015, 86, 1–6.
16. F. Tang, D. Su, Y. Tang and G. Fang, Sol. Energy Mater. Sol. Cells, 2015, 141, 218–224.
17. Z. Lu, B. Xu, J. Zhang, Y. Zhu, G. Sun and Z. Li, Sol. Energy, 2014, 108, 460–466.
18. Y. Zhang, A. Tang, H. Yang and J. Ouyang, Appl. Clay Sci., 2016, 119, 8–17.
19. W. Ding, J. Ouyang and H. Yang, Powder Technol., 2016, 292, 169–175.
20. M. Niu, X. Li, J. Ouyang and H. Yang, RSC. Adv., 2016, 6, 44106–44112.
21. M. Niu, H. Yang, X. Zhang, Y. Wang and A. Tang, ACS Appl. Mater. Interfaces, 2016, 8, 17312–17320.
22. J. Ouyang, H. Yang and A. Tang, Mater. Des., 2016, 92, 261–267.
23. K. Peng, J. Zhang, H. Yang and J. Ouyang, RSC Adv., 2015, 5, 66134–66140.
24. A. Karaipekli and A. Sari, Solar Energy, 2009, 83, 323–332.
25. O. Chung, S. G. Jeong and S. Kim, Sol. Energy Mater. Sol. Cells, 2015, 137, 107–112.
26. S. Liu and H. Yang, Energy Technol., 2015, 3, 77–83.
27. S. A. Memon, T. Y. Lo, X. Shi, S. Barbhuiya and H. Cui, Appl. Therm. Eng., 2013, 59, 336–347.
28. Z. Zhou, J. Ouyang, H. Yang and A. Tang, Appl. Clay Sci., 2016, 121, 63–70.
29. L. Fu, C. Huo, X. He and H. Yang, RSC Adv., 2015, 5, 20414–20423.
30. X. He and H. Yang, Dalton Trans., 2015, 44, 1673–1679.
31. X. He, Q. Yang, L. Fu and H. Yang, Funct. Mater. Lett., 2015, 8, 1550056.
32. J. Jin, L. Fu, J. Ouyang and H. Yang, Sci. Rep., 2015, 5, 12429.
33. A. Sari, C. Alkan, A. Biçer and C. Bilgin, Int. J. Energy Res., 2014, 38, 1478–1491.
34. A. Karaipekli and A. Sari, Sol. Energy Mater. Sol. Cells, 2016, 149, 19–28.
35. Z. Sun, Y. Zhang, S. Zheng, Y. Park and R. L. Frost, Thermochim. Acta, 2013, 558, 16–21.
36. A. Biçer and A. Sari, Energy Convers. Manage., 2013, 69, 148–156.
37. M. Ying and G. Zhang, Chem. Eng. J., 2016, 288, 70–78.
38. W. Ding, H. Yang and J. Ouyang, RSC Adv., 2015, 5, 67184–67194.
39. X. Li, J. Ouyang, Y. Zhou and H. Yang, Sci. Rep., 2015, 5, 13763.
40. Y. Zhang, J. Tan, M. Long, H. Yang, S. Yuan, A. Tang, S. Chang and Y. Hu, J. Mater. Chem. B, 2016, 4, 7406–7414.
41. K. Peng, H. Yan and J. Ouyang, Powder Technol., 2015, 286, 678–683.
42. C. Li, L. Fu, J. Ouyang, A. Tang and H. Yang, Appl. Clay Sci., 2015, 115, 212–220.
43. C. Huo, J. Ouyang and H. Yang, Sci. Rep., 2014, 4, 1–9.
44. H. Yang, A. Tang, J. Ouyang, M. Li and S. Mann, J. Mater. Chem. B, 2010, 114, 2390–2398.
45. X. Li, J. Luo, Q. Gao and J. Li, RSC Adv., 2016, 6, 45158–45165.
46. D. Yang, S. Shi, L. Xiong, H. Guo, H. Zhang, X. Chen, C. Wang and X. Chen, Sol. Energy Mater. Sol. Cells, 2016, 144, 228–234.
47. C. Wang, L. Feng, H. Yang, G. Xin, W. Li, J. Zheng, W. Tian and X. Li, Phys. Chem. Chem. Phys., 2012, 14, 13233–13238.
48. C. Li and H. Yang, J. Am. Ceram. Soc., 2013, 96, 2793–2798.
49. C. Li, L. Fu, J. Ouyang and H. Yang, Sci. Rep., 2012, 3, 1908.
50. S. Song, L. Dong, S. Chen, H. Xie and C. Xiong, Energy Convers. Manage., 2014, 81, 306–311.
51. Y. Wang, H. Zheng, H. X. Feng and D. Y. Zhang, Energy and Buildings, 2012, 47, 467–473.
52. A. Sari, A. Karaipekli and K. Kaygusuz, Int. J. Energy Res., 2008, 32, 154–160.
53. S. Liu and H. Yang, Appl. Clay Sci., 2014, 101, 277–281.
54. B. Xu and Z. Li, Appl. Energy, 2013, 105, 229–237.

---