Photocatalytic mechanism of high-activity anatase TiO₂ with exposed (001) facets from molecular-atomic scale: HRTEM and Raman studies

Abstract

Anatase TiO₂ with a variant percentage of exposed (001) facets was prepared under hydrothermal processes by adjusting the volume of HF, and the photocatalytic mechanism was studied from atomic-molecular scale by HRTEM and Raman spectroscopy. It was revealed that: 1) From HRTEM observations, the surface of original TiO₂ with exposed (001) facets was clean without impurity, and the crystal lattice was clear and completed; however, when mixed with methylene blue (MB) solution, there were many 1 nm molecular absorbed at the surface of TiO₂; after the photocatalytic experiment, MB molecules disappeared and the TiO₂ lattice image became fuzzy. 2) The broken path of the MB chemical bond was obtained by Raman spectroscopy, i.e., after the irradiation of the light, the vibrational mode of C−N−C disappeared due to the chemical bond breakage, and the groups containing C−N bond and carbon rings were gradually decomposed. Accordingly, we propose that the driving force for breaking the chemical bond and the disappearance of groups is from the surface lattice distortion of TiO₂ during photocatalyzation.

Keywords

TiO₂ / exposed (001) facets / HRTEM / Raman spectroscopy / photocatalytic degradation mechanism

1 1 Introduction

With the development of society and industry, human beings have a rising requirement for new energies [1]. Solar energy, as we all know, has become one of the most abundant and promising renewable energy sources. In order to achieve the lofty goals of carbon peaking and carbon neutrality, solar energy must be made full use by new energy storage technologies to support the growth of new energy sectors [2–3].

Shaped phase change materials (PCMs) for solar energy storage have always been focused on keeping the shape of raw materials before and after phase shift [4]. Solid‒liquid and solid‒solid shaped phase change energy storage solutions are common among the thermal energy storage materials [5–8]. Typically, solid‒liquid PCMs are involved in salt hydrates [9], molten salts [10], etc. Because of their abundant reserves, high energy storage density, and optimum phase transition temperature, mirabilite (Na₂SO₄·10H₂O) phase change materials (MPCMs) have been widely used in greenhouses and other disciplines [11–12]. However, their thermal storage performance is affected by the problems of supercooling, serious phase stratification, and susceptible to leakage and corrosion, resulting in a short service life and limiting their applications [13–17]. In particular, a lot of attempts have been taken to remedy these problems through adding thickeners, nucleating agents, and other methods [18–20]. Usually, borax [12,21–22] was commonly used as a nucleating agent to minimize supercooling to less than 2 °C, and sodium carboxymethyl cellulose (CMC) [23] was used to successfully suppress phase stratification. More recently, nanomaterials such as nano-graphite [24], nanocellulose [25], nano-graphene [26], metal nanoparticles [27] and others have been successfully applied to hydrated salt phase change materials. But the common limitations of these methods are that nano-materials cannot achieve anti-leakage and anti-corrosion. As a result, specialized packaging methods, like the microcapsule method [28], greatly improved leak-proof performance and thermal conductivity. For instance, Zhang et al. [29] effectively developed a silica-coated phase change material of Na₂SO₄ for thermal energy storage. The melting and solidification temperatures of the microcapsule material were 33.6 and 6.0 °C, and the corresponding latent heats were 125.6 and 74.0 kJ·kg⁻¹, respectively; Xi et al. [30] used konjac glucomannan (KGM) as the supporting material to prepare super hydrophilic reduced graphene oxide (rGO) aerogel by hydrothermal reaction and freeze-drying, and then prepared sodium acetate trihydrate/KGM modified graphene oxide aerogel (SAT/KrGO) composite PCMs, which had stable thermal cycling and thermal storage capacity. Obviously, the carbon support with porous, high specific surface area and stable structure can not only prevent the leakage of phase change materials, but also serve as a heat conduction enhancer to improve the heat conduction performance, which has practical applications in solar thermal energy conversion and storage [31–35].

Cellulose carbon aerogels were employed to encapsulate the MPCMs because of their low density, large specific surface area, and high porosity [30,36–39]. Sodium CMC is a kind of cellulose ether derivative, which is widely used as stabilizer, thickener, and food additive in food applications [40]. Particularly, the −COONa in CH₂COONa of the chair molecular structure is a hydrophilic group and has a larger space volume than the original −CH₂, which is more hydrophilic than cellulose. However, the aerogel made by CMC is too weak to be utilized in practice [41].

Accordingly, carbon materials are combined into CMC-derived aerogels to boost their mechanical strength. For example, Huang et al. [42] developed CS-GO/CMC composite aerogel spheres with multi-scale hierarchically porous structure that had high performance in removing organic dyes and heavy metal ions from water and could be utilized for wastewater treatment; Xu et al. [43] successfully prepared carboxymethyl cellulose/graphene composite aerogel (CMC/GA) with acceptable adsorption capacity using a one-step hydrothermal process based on graphite oxide (GO) and a suitable proportion of inexpensive CMC. One elegant example comes from Jiang et al. [44], who created a lamellar structure by combining directed freezing, pre-oxidation, and carbonization with water-soluble CMC and hydroxypropyl methyl cellulose (HPMC) as carbon sources and GO as structure-oriented material, the carbon aerogels with excellent mechanical and pressure sensing properties, which have promising applications in wearable electrical systems.

With inspiration from the above researches, a new challenge for mirabilite PCMs is to develop stable shaped-structure carbon support with high adsorption capacity to package phase change materials while balancing quick heat transmission, high thermal energy storage stability, and extended service life.

Compared to metal framework, the GO surface possesses more OH, COOH, and CO groups [45], as well as superior interaction force and high adsorption capacity for PCMs, which further resists leakage. At the same time, GO has a high thermal conductivity (5000 W·m⁻¹·K⁻¹) and specific surface area (2600 m²·g⁻¹) as a conventional two-dimensional (2D) sheet structure and is cross-linked with CMC to produce a variety of porous structure. Furthermore, in addition to create hydrophilic carbon aerogels with increasing specific surface area based on GO, a tiny quantity of carbon nanofibers (CNFs) served as structural support and guidance were used to regulate the porous structure and decrease the volume shrinkage rate of the scaffolds. As a support of MPCMs, carbon skeleton will enhance the leak-proof performance of liquid mirabilite and make it have a longer cycle life, which has a driving effect on the further development of the application of MPCMs.

Herein, the CMC/rGO/CNFs carbon aerogels with lightweight, hydrophilic and stable shape were prepared by directional freeze-drying and low-temperature carbonization with CMC as raw material, using GO as layered structure guide material and CNFs as structural support material. The shaped mirabilite PCMs composites were prepared by vacuum impregnating molten MPCMs. And the microstructure, chemical composition, stability, thermal conductivity, solid‒liquid cycle thermal stability, and anti-leakage feature of the composite PCMs were then investigated. The stable composites based on ordered porous lamellae carbon aerogels have high heat storage, heat release properties and cycle life, and it is hoped that this study will contribute to a deeper understanding of new ideas for high-efficiency solar phase change energy storage.

2 2 Experimental

2.1 2.1 Materials

Graphene (purity > 99.5 wt.%), sodium CMC ([C₈H₁₁O₅Na]_n, purity: CP, viscosity: 800‒1200 mPa), carbon nanofibers (purity > 99.5 wt.%, 50‒200 nm in diameter, 1‒10 µm in longitudinal length), mirabilite (Na₂SO₄·10H₂O, purity: AR), crystalline sodium carbonate (Na₂CO₃·10H₂O, purity: AR), sodium chloride (NaCl, purity: AR), and borax (Na₂B₄O₇·10H₂O, purity: AR) were used as raw materials. All reagents were purchased from Aladdin Industrial Co., Ltd. Deionized water was self-made in our laboratory.

2.2 2.2 Preparation of CMC/rGO/CNFs carbon aerogels

The preparation procedure for CMC/rGO/CNFs carbon aerogels (CGCA) is shown in Fig.1(a). The GO was first prepared by Hummers’ method [45]. 0.025 g GO, 0.00875 g CNF, and 0.2 g CMC were dissolved in 10 mL deionized water, and ultrasonically oscillated for 30 min to obtain CMC/GO/CNFs hydrogels. Then, freezing-dry was taken for more than 36 h to remove moisture. Finally, the obtained CMC/GO/CNFs hydrogels were carbonized in the tube furnace, and the carbonization process was divided into three stages in argon atmosphere. In the first stage, the samples were heated from 20 to 300 °C with heating rate of 5 °C·min⁻¹; in the second stage, the temperature was raised from 300 to 400 °C with heating rate of 0.5 °C·min⁻¹ and held for 1 h; in the third stage, the temperature was raised from 400 to 550 °C with heating rate of 5 °C·min⁻¹ and held for 1 h. Using the same procedure, other samples of 0, 0.0025, and 0.0125 g CNFs were prepared as the control groups, particularly recorded as CGCA-0, CGCA-1, and CGCA-2, respectively.

Fig.1 Schematic diagrams of (a) CMC/rGO/CNFs carbon aerogels preparation and (b) CGCA-PCMs.

Full size|PPT slide

2.3 2.3 Preparation of CMC/rGO/CNFs carbon aerogels mirabilite PCMs

The preparation procedure for CMC/rGO/CNFs carbon aerogels mirabilite composite phase change materials (CGCA-PCMs) is shown in Fig.1(b). The mirabilite composite phase change materials were prepared by the matrix of Na₂SO₄·10H₂O:Na₂CO₃·10H₂O = 9:1 (mass fraction) [22]. 0.4 g sodium chloride and 0.4 g borax were added into 10 g mirabilite composite phase change materials with heating at 45 °C and stirred to melt into the liquid mirabilite composites. Whereafter, the above-prepared CMC/rGO/CNFs carbon aerogels were immersed in the PCMs at 45 °C and subjected to vacuum immersion treatment for 10 h to prepare CGCA-PCMs, and the other control samples are CGCA-PCMs-0, CGCA-PCMs-1, and CGCA-PCMs-2.

2.4 2.4 Evaluation of pore performance evaluation

First of all, the volume shrinkage rate of the samples was estimated by comparison with the samples of CGA without carbonization. Then the average diameter and the average height of the samples were measured with vernier caliper to calculate the volume of the samples, and the average mass of the samples at 25 °C was measured by the analytical balance. Subsequently, the true density meter (AccuPyc II 1340) was used to measure the density of the samples. Finally, the porosity P of the samples with different mass fractions of CNFs added was calculated according to Eq. (1):

(1)$P=\left(1-\frac{m}{v{\rho }_{\text{s}}}\right)\times 100\mathrm{\%}$

where, m is the mass of the sample, v is the apparent volume of the sample, and ρ_s is the true density of the sample. The specific surface area and porosity were tested by the nitrogen adsorption–desorption technique at 0.1 MPa (ASAP 2020C+HD88).

2.5 2.5 Characterization

A contact angle meter (Data physics OCA 20) was used to test the contact angle of the CGCA aerogels to prove their hydrophilicity. The morphology and detailed microstructure were characterized by scanning electron microscopy (SEM; JSM-6610LV). The chemical structure was investigated by the Fourier transform infrared spectroscopy (FTIR; Spectrum BXII). Furthermore, the elemental analysis was done by X-ray photoelectron spectroscopy (XPS; Escalabx). The micro confocal Raman spectrometer (LabRAM Odyssey) was used to describe the interior structure. The thermal property was evaluated by the differential scanning calorimetry (DSC; STA449F3-DSC200F3) in nitrogen atmosphere with heating/cooling rate of 3 °C·min⁻¹. The thermal stability was characterized by thermal gravimetric analysis (TGA; STA449 F3-STA449 F3) with heating rate of 2 °C·min⁻¹ in air atmosphere. The thermal conductivity was measured by the Hot Disk thermal constant analyzer (TPS 2200) three times. Thermal cycling stability was tested by putting the samples in a water bath equipped with a thermostat. Set the automatic cycle program temperature as: heating to 45 °C (above the melting temperature), heating time of 30 min, heating power of 1500 W, and then cooling to 5 °C (below the crystallization temperature), cooling time of 30 min, cooling power of 1500 W, and the automatic cycle program was set to 100, 500, and 1500 cycles.

3 3 Results and discussion

3.1 3.1 Nitrogen sorption, pore analysis and leak-proof performance of CGCA

The porosity evaluations of different samples are shown in Tab.1. It can be seen that the structure of Sample CGCA-2 with CNFs exceeding 35 wt.% GO is unstable and easy to break, and the porosity of the material cannot be determined. A possible explanation for this might be that the stable chemical bonds in this sample cannot be formed among carbons due to the increase of carbon content, so that it is easy to break up at room temperature and consequently the pore structure fails, and therefore it is difficult to support the phase change cycle process of the phase change energy storage materials. Samples CGCA-0 and CGCA-1 with CNFs 0 wt.% and CNFs 10 wt.% GO have large volume shrinkage and low porosity due to the lack or insufficient of CNFs, and obviously cannot be used for the encapsulation of phase change energy storage materials. Specially, sample CGCA with CNFs 35 wt.% GO has small volume shrinkage and with a porosity of 96.9%. So we strongly feel that the addition of CNFs effectively improves the toughness and enhances the structural stability of the CGCA, making the materials present a layered pore structure, which can effectively support the phase change process of the mirabilite PCMs in the pore structure. Through the comparison of porosity data, we can clearly see that CGCA has a higher porosity and is suitable for supporting phase change energy storage materials. Therefore, this paper further analyzes pore size characteristics of the CGCA with 35 wt.% CNFs of GO as a structural support material.

Tab.1 Average porosities of CGCA samples with different mass fractions of CNFs

Sample	Average diameter/cm	Average height/cm	Average volume/cm3	Average mass/g	ρs/(g·cm−3)	Volume shrinkage rate/%	Porosity/%
CGCA-0	0.1214	1.5678	0.0182	0.0170	1.37	> 80	31.8
CGCA-1	0.1656	2.7314	0.0588	0.0196	1.65	> 60	79.8
CGCA	0.3333	3.3000	0.2879	0.0415	4.62	10–40	96.9

Note: CGCA-2 is unstable, and easy to break.

Furthermore, N₂ adsorption was carried out for the activated CGCA sample in order to evaluate the pore properties. As shown in Fig.2(a), CGCA exhibits reversible type IV and H₃, respectively, showing a mesoporous structure [46], which is one of the main characteristics of mesoporous materials. The specific surface area (S_BET) of CGCA was calculated to be 213.3 m²·g⁻¹ using the Brunauer–Emmett–Teller (BET) model. The pore size of CGCA (inset of Fig.2(a)) is mainly concentrated around 5, 8, and 10 nm, further confirming its mesoporous structure. The above results indicate that the mesoporous CGCA scaffolds with higher specific surface area were obtained by the fiber structure of CMC and the guidance of CNFs and GO. It is the key to improve the loading rate and heat release performance of mirabilite.

Fig.2 (a) Nitrogen adsorption–desorption isotherms of CGCA (inset: pore width distribution curve). (b) Leak-proof photos of samples at 45 °C for 1 and 1500 cycles.

Full size|PPT slide

The leakage-proof property is very important for the PCMs thermal energy storage. The leakage results of PCMs, CGCA-PCMs-0 (the average volume is 0.0182 cm²) and CGCA-PCMs (the average volume is 0.2879 cm²) before and after cycling at 45 °C are shown in Fig.2(b). It is obvious that PCMs, GCCA-PCMs-0, and CGCA-PCMs maintain stable shapes in the crystalline state before cycling. When heated to 45 °C, the PCMs become liquid, and the CGCA-PCMs-0 has serious leakage, while no leakage can be found for the CGCA-PCMs after 1500 cycles. There is evidence to support the hypothesis that the mechanical properties of the carbon aerogel are enhanced by proper orientation enhancement of CNFs, making the carbon aerogel scaffold more stable. Therefore, the aerogel framework can be used as support material for mirabilite PCMs.

3.2 3.2 Morphology and microstructure of CGCA and CGCA-PCMs

The cross-section and longitudinal sections of CGA, CGCA-0, and CGCA, as well as the microstructure of CGCA impregnated with mirabilite crystals, were all characterized and presented in Fig.3. As shown in Fig.3(a), Fig.3(b) and Fig.3(c), it can be found that the CGCA forms many pores on the surface, and more regular and hierarchical structure is formed by carbonizing, which indicates that the carbonization process eliminates the hydrogen bonds between CMC nanofibers to form a porous structure. CGCA-0 without CNFs has no through-holes and displays a disordered pore structure, as seen in Fig.3(d). In contrast, according to the cross-sectional longitudinal sections in Fig.3(e) and Fig.3(f), CGCA has a regular sheet-like structure and abundant pore channels because it is guided by GO and a small addition of CNFs, and the three-dimensional (3D) pore structure formed after adding CNFs, which is often strongly beneficial for the adsorption of PCMs. Meanwhile, it has been demonstrated that CGCA has a high density of through holes, making it more favorable to the adsorption of phase change energy storage materials. Furthermore, it is commonly known that CGCA aerogels must be hydrophilic based on the strong water solubility of hydrated salts, in particular, it can be seen clearly that the mirabilite phase change material is completely immersed into the CGCA carbon framework in Fig.3(g) and Fig.3(h) of CGCA-PCMs, meaning that CMC/rGO/CNFs are of great absorbability to MPCMs. It is crucial to note that the contact angle between CGCA and water is 43.5° as shown in Fig.3(h), indicating that it has outstanding hydrophilicity.

Fig.3 SEM images of (a) CGA, (b)(c) CGCA, (d) CGCA-0 for its surface structure, (e) CGCA for its cross section structures, (f) CGCA for its longitudinal section structures, and (g)(h) CGCA-PCMs. (i) Contact angle between CGCA and water.

Full size|PPT slide

3.3 3.3 Chemical structure analysis of GO/CGCA/CGCA-PCMs

The FTIR spectra of GO, CGCA, and CMC are shown in Fig.4(a). The peaks at 3434 and 1439 cm⁻¹ are corresponding to the stretching vibrations of O−H and the in-plane bending vibration of −OH, respectively. The peaks at 2922, 1630, and 1150 cm⁻¹ are ascribed to the stretching vibrations of −CH₃, C=O, and C−O−C, respectively. Comparing the infrared peaks of the above three samples, it is found that there are hydrophilic groups such as O−H in CGCA, which determine their hydrophilicity and lay the foundation for the adsorption of PCMs [47].

Fig.4 (a) FTIR spectra of GO, CMC, and CGCA samples. (b) Raman spectra of GO and CGCA samples. (c) Raman spectra of PCMs and CGCA-PCMs samples. (d) EDS line spectra of CGCA samples before and after carbonization. (e) XPS wide-scan spectra of GO, CMC, and CGCA samples. (f)(g)(h) Spectra of C 1s of CMC, GO, and CGCA samples.

Full size|PPT slide

Particularly, Raman spectra are conducted to confirm the defects of the supporting materials in Fig.4(b). Two notable peaks are observed in CGCA, a typical D peak at about 1346 cm⁻¹ related to defects of the graphitic lattice of the sp³ hybridized carbon, and a G peak center at 1577 cm⁻¹ attributed to the presence of sp² bonded carbon. In addition, the intensity ratio of the D and G peak (I_D/I_G) was established to quantify the defects in the carbon of GO and the CGCA. The I_D/I_G values of GO and CGCA are 0.971 and 0.860, respectively, indicating that CGCA has lower density of defects than that of GO, because GO and CNFs successfully induce CMC to form a more ordered and less defective framework structure.

What is even more important is that the Raman spectra of PCMs and CGCA-PCMs also stand out in Fig.4(c). Obviously, the typical characteristic peaks at 985 and 3430 cm⁻¹ are attributed to Na₂SO₄·10H₂O. Meanwhile, there are S−O symmetry bending vibration peaks and S−O deformation bending vibration peaks around 450 and 622 cm⁻¹ in both PCMs and CGCA-PCMs, respectively, and it is also obvious that there is the S−O antisymmetric stretching vibration peak in [SO₄²⁻] with wave numbers of 1120 cm⁻¹. All the typical Raman peaks of PCMs are observed in CGCA-PCMs, indicating that PCMs are successfully introduced into the solid support. Besides, distinct D and G peaks appear in CGCA-PCMs, which also indicates that the CGCA can be successfully used as a support for PCMs [48].

The energy dispersive spectroscopy (EDS) results of CGCA before and after heat treatment in Fig.4(d) show that the elements contained in the aerogels before and after heat treatment are C, O, and Na, but the oxygen content is reduced by nearly half after heat treatment, which may be explained by the fact that due to the decrease of oxygen content on the surface of GO and CMC, GO undergoes a weak reduction reaction under 550 °C. On the other hand, the elements contained in CGCA were further analyzed by finer XPS spectra. As shown in Fig.4(e), the XPS spectra of CGCA show that C, O, and Na elements exist on the surface of CMC, GO, and CGCA, and the corresponding binding energies of C, O, Na 1s, and Na KLL are around 281, 526, 493, and 1073 eV, respectively, which proves that CGCA has good hydrophilicity. The corresponding high resolution spectra of C−C, C−O, O−C−O, and C=O groups located respectively at 284.7, 285.3, 287.6, and 288.8 eV in the C 1s spectrum are analyzed by XPS. The CMC sample shows at least three peaks at the binding energies of 284.50, 285.85, and 287.55 eV, which are mainly attributed to the C−C, C=C, C−O and O−C−O groups. The GO sample shows three peaks at the binding energies of 284.10, 286.55, and 288.30 eV, which can be attributed to C−C, C=C, C−O, and C=O. However, in the CGCA sample, four peaks are observed at the binding energies of 284.50 eV (C−C), 285.55 eV (C−O), 287.55 eV (O−C−O), and 288.30 eV (C=O) [49]. The variations of the binding energies further confirm the strong interaction between GO and CMC in Fig.4(f)–Fig.4(h). As expected, our data confirm the expected correlations among the CGCA carbon framework, GO, and CMC, which indicate that the CGCA aerogel has a stable chemical structure.

3.4 3.4 Supercooling of mirabilite composite PCMs and CGCA-PCMs

The degree of subcooling is commonly considered to be directly connected to the performance of the MPCMs. Because there is generally a meta-stable condition around the melting point that impacts the crystallization process, supercooling is a common phenomenon at the time of crystallization.

As illustrated in Fig.5(a), both PCMs and CGCA-PCMs have supercooling during solidification. However, the supercooling degree of CGCA-PCMs (0.7 °C) is much lower than that of PCMs (1.2 °C). We hypothesize that there is a major factor that promotes the drop of the supercooling degree. In general, this can be explained, in part, by the thermal conductivity differences between GO and other carbon materials, which is helpful for improving the thermal conductivity of CGCA aerogels. As a result, the thermal conductivity of CGCA prepared in this paper is 1.549 W·m⁻¹·K⁻¹ (Fig.5(b)). Simultaneously, there is satisfactory agreement that the carbonization at 550 °C is used in this paper, as slight carbonization ensures that GO does not lose its hydrophilicity due to being completely reduced, and it also retains the original high thermal conductivity of GO, so that the thermal conductivity of the CGCA prepared material remains relatively high.

Fig.5 (a) T-history pattern of PCMs and CGCA-PCMs. (b) Thermal conductivities of CGCA-PCMs. (c) Partial enlarged view of the green marked map in panel (a). (d) Schematic diagram of heterogeneous nucleation of CGCA-PCMs.

Full size|PPT slide

Furthermore, we know that the PCM has a thermal conductivity of 0.544 W·m⁻¹·K⁻¹, while CGCA-PCM has the thermal conductivity of 1.756 W·m⁻¹·K⁻¹ in Fig.5(b). In comparison, the conductivity of CGCA-PCM is three times that of PCMs. This demonstrates that the CGCA framework can convey the heat generated during mirabilite crystallization to the outside for heat exchange in time, hence, eliminating the heat transfer lag of the mirabilite phase transition material system. However, we all know that mirabilite requires a certain degree of subcooling to give energies in the process of reaching the critical point of crystallization, so we hypothesize that the magnitude of this part of the driving force can be reduced in a way that the heat is released and absorbed fast. This can be further confirmed by the compositional analysis, the reducing of time-space lag in crystallizing makes the phase transition temperature of the mirabilite PCMs actually controllable, which has a good theoretical guiding value for the use of the latent heat of the MPCMs.

Indeed, this may also be explained from the standpoint of thermodynamics present in the formation of mirabilite nuclei. The basic thermodynamic condition governing crystal growth is the principle of minimum free energy. The free energy of the mirabilite crystal is composed of its volume free energy and interface energy [50].

As shown in Fig.5(d), the radius of the mirabilite crystal nucleus formed on the inner wall of CGCA is r, then the change of the interface energy when the crystal nucleus formed is:

(2)$\mathrm{\Delta }{G}_S=S_{\text{nl}}{\sigma }_{\text{nl}}+S_{\text{nw}}{\sigma }_{\text{nw}}-S_{\text{lw}}{\sigma }_{\text{lw}}$

where, S_nl, S_nw, and S_lw are the interface area between the mirabilite crystal nucleus and the mirabilite liquid phase, the interface area between the mirabilite crystal nucleus and the inner wall of CGCA, and the interface area between the mirabilite liquid phase and the inner wall of CGCA, respectively; σ_nl, σ_nw, and σ_lw are the specific surface energy between the mirabilite crystal nucleus and the mirabilite liquid phase, the specific surface energy between the mirabilite crystal nucleus and the inner wall of CGCA, and the specific surface energy between the mirabilite liquid phase and the inner wall of CGCA, respectively.

At the intersection of three phases:

(3)$\mathrm{\Delta }{G}_S=\left(S_{\text{nl}}-\text{π}{r}^2{\sin }^2\theta \cos \theta \right){\sigma }_{\text{nl}}$

Additionally, the volume free energy caused by mirabilite nuclei is:

(4)$\mathrm{\Delta }{G}_t=V_n\mathrm{\Delta }{G}_V=\text{π}{r}^3\left(\frac{2-3\cos \theta +{\cos }^3\theta }{3}\right)\mathrm{\Delta }{G}_V$

Then, the total free energy caused by mirabilite nuclei is as follows:

(5)$\mathrm{\Delta }{G}_t+\mathrm{\Delta }{G}_S=\left(\frac{4r^3\mathrm{\Delta }{G}_V}{3}+4\text{π}{r}^2{\sigma }_{\text{nl}}\right)\times f(\theta )$

where, θ formed between the mirabilite nucleus and the inner wall of CGCA is 0° ≤ θ ≤ 180°. Obviously, in this way, while the nucleus of CGCA-PCMs grows, f (θ) is between 0 and 1, and the total free energy caused by mirabilite nuclei will be less than 1.

From this point of view, because the contact angle of CGCA is less than 90°, the total free energy between the mirabilite nucleus and the inner wall of CGCA could be less than 1, and CGCA-PCMs can undergo heterogeneous nucleation at lower supercooling degrees. Summarily, the surface of CGCA aerogels offers many non-uniform nucleation sites for mirabilite nuclei, which greatly intensifies the crystallization behavior of the hydrate salt and nucleates rapidly at lower temperature, thereby reducing the degree of supercooling of the CGCA-PCMs. Observably, the CGCA could be responsible for the remarkable crystallization of mirabilite phase change energy storage materials.

3.5 3.5 Thermal properties of mirabilite composite PCMs and CGCA-PCMs

It is obvious that the latent heat of CGCA-PCMs influences the performance of materials from this study, so the DSC curves of the original MPCMs and the CGCA-PCMs are measured in Fig.6. Firstly, we can observe from Fig.6(a) and 6(b) that CGCA-PCMs have phase transition features comparable to mirabilite PCMs. The phase transition behavior of the carbon aerogels generated by CMC and GO reveals that they have strong adsorption to the mirabilite PCMs. Secondly, in Fig.6(c), CGCA-PCMs-0 and CGCA-PCMs-1 have low loading rates due to their low porosities, which ultimately lead to very poor thermal performance of mirabilite composite phase change energy storage materials. However, the melting and solidification enthalpies of PCMs are 242.8 and 134.2 J·g⁻¹, respectively, while those of CGCA-PCMs are 157.1 and 114.8 J·g⁻¹, respectively. The heat storage density of CGCA-PCMs can exceed about 85.5%, which indicates that the CGCA-PCMs have significant heat storage and heat release capabilities, suitable phase change temperature.

Fig.6 (a)(b) DSC curves of PCMs and CGCA-PCMs. (c) Latent heat histograms of PCMs and CGCA-PCMs. (d)(e) DSC curves of CGCA-PCMs. (f) Latent heat histograms of CGCA-PCMs at 100, 500, and 1500 cycles.

Full size|PPT slide

To further explore the thermal cycling stability, the DSC curves of CGCA-PCMs for 100‒1500 solid‒liquid cycles are shown in Fig.6(d)–Fig.6(f). After the first cycle, the energy storage density is 157.1 J·g⁻¹, the crystallization exotherm is 114.8 J·g⁻¹, and the phase transition temperature is 16.2 °C. Interestingly, after 500 cycles, the energy storage density is 90.37 J·g⁻¹, the crystallization exotherm is 70.14 J·g⁻¹, and the phase transition temperature is 15.5 °C. Continuously, after 1500 cycles, the heat storage density of CGCA-PCMs drops to 75.37 J·g⁻¹, the crystallization exotherm is 52.03 J·g⁻¹, and the phase transition temperature is 13.9 °C, retaining about 47.97% of the energy storage density in the first cycle. Most importantly, the latent heat retention rates of crystallization (η) for samples are shown in Tab.2, which are calculated by Eq. (6) as follows:

Tab.2 DSC data for CGCA-PCMs of 100‒1500 cycles

Number of cycling	Melting				Solidifying			η/%
	to/°C	tp/°C	∆Hm/(J·g−1)		to/°C	tp/°C	∆Hc/(J·g−1)
1	31.4	34.7	157.1		14.3	16.2	114.80	‒
100	29.1	31.6	118.5		13.8	14.9	93.58	81.52
500	24.0	32.1	90.37		11.5	15.5	70.14	61.10
1500	12.3	23.7	75.37		12.4	13.9	52.03	45.32

Notes: to, onset temperature; tp, peak temperature.

(6)$\eta /\phantom{\eta \mathrm{\%}}\mathrm{\%}=\frac{\mathrm{\Delta }{H}_n}{\mathrm{\Delta }{H}_{\text{c}1}}\times 100$

where ∆H_c1 is the latent heat of crystallization of CGCA-PCMs before cycling, and n represents the number of cycling.

As shown in the Fig.6(f), the retention rates of the latent heat of crystallization of the CGCA-PCMs maintain approximately 61.10% and 45.32% after 500 and 1500 solid‒liquid cycles, respectively. It can be concluded that our experiments demonstrate the expected encapsulation dependence on the prepared porous aerogels that CGCA has strong mechanical properties and excellent shape stability. It is also one of the available ways for better applications of mirabilite phase change energy storage materials.

3.6 3.6 Thermal reliability of CGCA-PCMs

The absence of crystal water is crucial for judging phase change cycles in hydrated salts such as Na₂SO₄·10H₂O. In this paper, TGA analysis was applied to test the heat resistance and thermal stability of the samples. Fig.7(a) shows the whole TGA analysis curves of PCMs. It can be seen that the weight loss rates of PCMs reach 56.30% at 85.8 °C, and the exotherm is 1494 J·g⁻¹, subsequently, while the weight loss rates of CGCA-PCMs reach 50.21% at 89.5 °C in Fig.7(b), and the exotherm is 1058.7 J·g⁻¹, owing primarily to the loss of Na₂SO₄·10H₂O crystal water and the heat released by crystal water vaporization at 100 °C [51]. Its thermal degradation behavior is similar to that of the PCMs, indicating the energy storage materials have good chemical compatibilities with CGCA. Differently, the PCMs start to lose water at 38.6 °C while the CGCA-PCMs fail at 59.7 °C, which is summarized that the mirabilite composite PCMs employing CGCA aerogels as the support have improved thermal stability. Although there are still some challenges and limitations in its greater flexibility, we believe that the CGCA mirabilite composite phase change materials with high stability and cycle life still can potentially enable many future applications.

Fig.7 TGA curves of (a) PCMs and (b) CGCA-PCMs.

Full size|PPT slide

4 4 Conclusions

Using GO as the guiding material and CNFs as the structural support materials, in this work we successfully designed and fabricated the lightweight CGCA, which could be used as ideal support for mirabilite phase change energy storage materials. CGCA exhibits the following three characteristics. (1) The 3D lamellar structure of the CGCA has higher specific surface area and stable shape. After 1500 solid‒liquid cycles, there is no obvious leakage. (2) Good adsorption capacity and anti-leakage properties. The porosity of CGCA reaches 96.9% and it can adsorb mirabilite PCMs nearly 50 times than its own weight, ensuring higher energy storage density. (3) Ideal heat transfer path and excellent PCMs package. Based on the excellent properties of CGCA, the carbon aerogel mirabilite composite PCMs have desired phase change enthalpy, enhanced thermal conductivity and thermal cycling stability. After 1500 solid‒liquid cycles, the supercooling degree of the composites does not change significantly, and the latent heat rate decreases by 54.67%, indicating that the prepared CMC/rGO/CNFs mirabilite composite PCMs have excellent thermal cycle stability. In addition, the wide availability and low-cost of CMC, long-term cycle stability, and conversion from low-grade energy to high-grade energy (thermal storage) endow the composite phase change energy storage materials enormous economic advantages. Therefore, our work provides a new approach for efficient solar energy storage and thermal energy utilization in porous support hydrated salt composite phase change materials.

References

[1] Fujishima A, Honda  K. Electrochemical photolysis of water at a semiconductor electrode. Nature, 1972, 238(5358): 37–38

[2] Nishiyama N, Fujiwara  Y, Adachi K , . Preparation of porous metal-ion-doped titanium dioxide and the photocatalytic degradation of 4-chlorophenol under visible light irradiation. Applied Catalysis B: Environmental, 2015, 176–177: 347–353

[3] Obata K, Kishishita  K, Okemoto A , . Photocatalytic decomposition of NH₃ over TiO₂ catalysts doped with Fe. Applied Catalysis B: Environmental, 2014, 160–161: 200–203

[4] Choi J, Park  H, Hoffmann M R . Effects of single metal-ion doping on the visible-light photoreactivity of TiO₂. The Journal of Physical Chemistry C, 2010, 114(2): 783–792

[5] Qi D, Xing  M, Zhang J . Hydrophobic carbon-doped TiO₂/MCF-F composite as a high performance photocatalyst. The Journal of Physical Chemistry C, 2014, 118(14): 7329–7336

[6] Chen X, Lou  Y, Samia A C S , . Formation of oxynitride as the photocatalytic enhancing site in nitrogen-doped titania nanocatalysts: Comparison to a commercial nanopowder. Advanced Functional Materials, 2005, 15(1): 41–49

[7] Li D, Jiang  X, Zhang Y , . A novel route to ZnO/TiO₂ heterojunction composite fibers. Journal of Materials Research, 2013, 28(3): 507–512

[8] Luo C, Li  D, Wu W , . Preparation of porous micro–nano-structure NiO/ZnO heterojunction and its photocatalytic property. RSC Advances, 2014, 4(6): 3090–3095

[9] Luo C, Li  D, Wu W , . Preparation of 3D reticulated ZnO/CNF/NiO heteroarchitecture for high-performance photocatalysis. Applied Catalysis B: Environmental, 2015, 166–167: 217–223

[10] Lu Q, Lu  Z, Lu Y , . Photocatalytic synthesis and photovoltaic application of Ag–TiO₂ nanorod composites. Nano Letters, 2013, 13(11): 5698–5702

[11] Wu J, Luo  C, Li D , . Preparation of Au nanoparticle-decorated ZnO/NiO heterostructure via nonsolvent method for high-performance photocatalysis. Journal of Materials Science, 2017, 52(3): 1285–1295

[12] Pan X, Xu  Y J. Defect-mediated growth of noble-metal (Ag, Pt, and Pd) nanoparticles on TiO₂ with oxygen vacancies for photocatalytic redox reactions under visible light. The Journal of Physical Chemistry C, 2013, 117(35): 17996–18005

[13] Yarahmadi A, Sharifnia  S. Dye photosensitization of ZnO with metallophthalocyanines (Co, Ni and Cu) in photocatalytic conversion of greenhouse gases. Dyes and Pigments, 2014, 107(13): 140–145

[14] He Y, Tilocca  A, Dulub O , . Local ordering and electronic signatures of submonolayer water on anatase TiO₂(101). Nature Materials, 2009, 8(7): 585–589

[15] Bikondoa O, Pang  C L, Ithnin  R, . Direct visualization of defect-mediated dissociation of water on TiO₂ (110). Nature Materials, 2006, 5(3): 189–192

[16] Du Y, Deskins  N A, Zhang  Z, . Two pathways for water interaction with oxygen adatoms on TiO₂(110). Physical Review Letters, 2009, 102(9): 096102

[17] Ekström G N ,  McQuillan A J . In situ infrared spectroscopy of glyoxylic acid adsorption and photocatalysis on TiO₂ in aqueous solution. The Journal of Physical Chemistry B, 1999, 103(48): 10562–10565

[18] El-Maazawi M, Finken  A N, Nair  A B, . Adsorption and photocatalytic oxidation of acetone on TiO₂: An in situ transmission FT-IR study. Journal of Catalysis, 2000, 191(1): 138–146

[19] Sato S, Ueda  K, Kawasaki Y , . In situ IR observation of surface species during the photocatalytic decomposition of acetic acid over TiO₂ films. The Journal of Physical Chemistry B, 2002, 106(35): 9054–9058

[20] Kang M, Lee  J H, Lee  S H, . Preparation of TiO₂ film by the MOCVD method and analysis for decomposition of trichloro-ethylene using in situ FT-IR spectroscopy. Journal of Molecular Catalysis A: Chemical, 2003, 193(1–2): 273–283

[21] Yu Z, Chuang  S. In situ IR study of adsorbed species and photogenerated electrons during photocatalytic oxidation of ethanol on TiO₂. Journal of Catalysis, 2007, 246(1): 118–126

[22] Almeida A R, Moulijn  J A, Mul  G. In situ ATR-FTIR study on the selective photo-oxidation of cyclohexane over anatase TiO₂. The Journal of Physical Chemistry C, 2008, 112(5): 1552–1561

[23] Ho C, Shieh  C, Tseng C , . Decomposition pathways of glycolic acid on titanium dioxide. Journal of Catalysis, 2009, 261(2): 150–157

[24] Tseng C L, Chen  Y K, Wang  S H, . 2-Ethanolamine on TiO₂ investigated by in situ infrared spectroscopy. Adsorption, photochemistry, and its interaction with CO₂. The Journal of Physical Chemistry C, 2010, 114(27): 11835–11843

[25] Imanishi A, Okamura  T, Ohashi N , . Mechanism of water photooxidation reaction at atomically flat TiO₂ (rutile) (110) and (100) surfaces: dependence on solution pH. Journal of the American Chemical Society, 2007, 129(37): 11569–11578

[26] Fleming G J, Idriss  H. Probing the reaction pathways of DL-proline on TiO₂ (001) single crystal surfaces. Langmuir, 2004, 20(18): 7540–7546

[27] Cao Y, Yi  L, Huang L , . Mechanism and pathways of chlorfenapyr photocatalytic degradation in aqueous suspension of TiO₂. Environmental Science & Technology, 2006, 40(10): 3373–3377

[28] Zhang J, Zhang  Y, Lei Y , . Photocatalytic and degradation mechanisms of anatase TiO₂: A HRTEM study. Catalysis Science & Technology, 2011, 1(2): 273–278

[29] Tian F, Zhang  Y, Zhang J , . Raman spectroscopy: A new approach to measure the percentage of anatase TiO₂ exposed (001) facets. The Journal of Physical Chemistry C, 2012, 116(13): 7515–7519

[30] Yang H G, Sun  C H, Qiao  S Z, . Anatase TiO₂ single crystals with a large percentage of reactive facets. Nature, 2008, 453(7195): 638–641

[31] Liu P, Wang  Y, Zhang H , . Vapor-phase hydrothermal transformation of HTiOF₃ intermediates into {001} faceted anatase single-crystalline nanosheets. Small, 2012, 8(23): 3664–3673

[32] Bersani D, Lottici  P P, Ding  X Z. Phonon confinement effects in the Raman scattering by TiO₂ nanocrystals. Applied Physics Letters, 1998, 72(1): 73–75

Acknowledgements

This work was supported by the National Basic Research Program of China (973 Program, Grant No. 2009CB939705), the National Natural Science Foundation of China (Grant No. 11174227), and the Chinese Universities Scientific Fund.

RIGHTS & PERMISSIONS

2017 Higher Education Press and Springer-Verlag GmbH Germany