Spectrally selective Cu_1.5Mn_1.5O₄ spinel ceramic pigments for solar thermal applications

Abstract

In this work, Cu_1.5Mn_1.5O₄ spinel ceramic pigments have been successfully prepared by a facile and cost-effective sol–gel self-combustion method and annealed at a temperature ranging from 500 °C to 900 °C for 1 h. The reaction process was discussed in detail through thermal analysis, X-ray diffraction (XRD), and Fourier transformation infrared spectroscopy (FTIR) techniques. The optical properties of the ceramic pigments annealed at various temperatures were also determined from the corresponding diffuse reflectance spectra in the 2.5–20 μm range, which elucidated that optimum spectrally selective paint coatings could be fabricated according to the pre-selection ceramic pigments. Ceramic pigments were utilized to fabricate thickness sensitive spectrally selective (TSSS) paint coatings by means of the convenient and practical spray-coating technique, and TSSS paint coatings based on pigments annealed at 700 °C showed absorptance of α_s = 0.914–0.923 and emittance of ε_T = 0.244–0.357. Besides, thermal stability tests were carried out by a prolonged and extended accelerated thermal investigation at 227 °C. The TSSS paint coatings exhibited no observable visual changes and the performance criterion (PC) values reached the requirement needed.

---

1 Introduction

During the last few decades, transition metal oxides with a spinel structure have attracted significant attention due to their unique properties such as chemical inertness, high corrosion resistance, high mechanical strength, good thermal shock resistance, excellent optical and catalytic properties, causing the emergence of potential applications ranging from optics, electronics, magnetism, and catalysis to energy conversion and storage.^1–5 In particular, these features endow this material with promising application as solar selective absorbers in solar-thermal energy conversion systems.^6,7 From the viewpoint of practical application, the solar absorber materials employed in low-to-mid temperature domain such as plate solar collectors exhibits a high solar absorptance in the UV-Vis-NIR solar spectrum and a low thermal emittance in the IR wavelength region.⁸

Many research efforts have been devoted to the preparation of the ternary and the binary spinel compounds serving as the solar-absorbing materials in the last decades, for instance, CuFeMnO_x, CuCoMnO_x, CuMnO_x spinels and Cu_xCo_yO_z spinel-like.⁹ Kaluža and co-workers have succeeded in fabricating the black colored CuFeMnO₄ spinel powders and CuFeMn-oxide/SiO₂ film using sol–gel process at the heat-treatment temperature of 500 °C.¹⁰ The prepared film was composed of the upper Cu_1.4Mn_1.6O₄ spinel layer and the lower amorphous SiO₂ layer, which exhibited absorptance of around 0.6 and emittance of 0.29–0.39. Due to the segregation of Fe₂O₃ phase upon heat treatment, the film displayed a reddish brown hue which caused a low solar absorptance. To deal with this problem, researchers managed to substitute Fe with Co and ultimately synthesized CuCoMnO_x spinel film.^11,12 Some of CuCoMnO_x spinel films exhibited high spectral selectivity. However, the preparation cost of large scale solar absorber film of this type based on Co element is not low enough because the much higher price of the cobalt salt precursor than that of other transition metal salt. The simpler binary CuMnO_x spinel derived from the CuCoMnO_x spinel also showed the characteristic of the spectral selectivity. CuMn-spinel thin films were deposited on aluminum substrates by dip-coating method and subsequently sintered in air at 450–500 °C.¹³ The formation of the copper–manganese-oxide film had the absorptance of 0.86–0.87. Nevertheless, the excellent spectral selectivity of the film could be achieved only if the anti-reflection SiO₂ coating was added to this film.¹⁴ In addition, for this film, some problems that arose from using sol–gel dip-coating method for large scale deposition, such as the border effect and heterogeneity of the film, had a negative effect on its application.¹⁵ It has been reported that solar selective absorber films consisting of pure Cu_1.5Mn_1.5O₄ spinel phase have also been successfully deposited on aluminum and stainless steel substrates by sol–gel dip-coating method.¹⁶ However, Cu_1.5Mn_1.5O₄ thin films showed a poor solar absorptance. The material mentioned above fails to meet the low-cost, easy access and practice application requirements which need to be taken into consideration. Hence, efforts have been made to decrease the cost of raw material and to pursue a simple and practical large-scale fabrication technique. To date, it has not attracted any attention for the usage of Cu_1.5Mn_1.5O₄ powder as a promising ceramic pigment for preparing the TSSS paint coating for solar thermal application.

The aim of the present work is to fabricate Cu_1.5Mn_1.5O₄ ceramic pigments as solar-absorber material by the facile and cost-effective sol–gel self-combustion method, which has been intensively investigated owing to preparing ceramic pigments with high purity, better chemical homogeneity, small grain size, and relatively low crystallization temperature than some traditional methods.¹⁷ To the best of our knowledge, the formation of Cu_1.5Mn_1.5O₄ ceramic pigments depended on not only the sol–gel self-combustion process, but also the annealing process. Hence, the effects of the annealing temperature on the formation of ceramic pigments were systematically studies using the corresponding characterization technology. Additionally, ceramic pigments showing different optical properties were also discussed when annealed at various temperatures. Subsequently, ceramic pigments were employed to fabricate TSSS paint coatings by the convenient and practical spray-coating technique. The influences of ceramic pigments annealed at difference temperatures on the spectral selectivity of paint coatings were also investigated in detail through measuring optical properties. Furthermore, the accelerated thermal stability test was performed to evaluate the service lifetime of paint coatings in low-to-mid temperature region.

2 Experimental section

2.1 Synthesis of Cu_1.5Mn_1.5O₄ spinel ceramic pigments

Cu_1.5Mn_1.5O₄ spinel ceramic pigments were synthesized by sol–gel self-combustion technique. All of the analytical-grade reagents were purchased and used as received. Copper nitrate trihydrate (Cu(NO₃)₂·3H₂O) and manganese nitrate (Mn(NO₃)₂ (50%)) were firstly dissolved in an adequate amount of ultrapure water with the Cu/Mn molar ratio of 1 : 1. An appropriate amount of citric acid was then added into the prepared aqueous solution to chelate Cu²⁺ and Mn²⁺. After stirring for period of time, the polyethylene glycol 200 (PEG 200) was added to the solution as an esterifying agent, which took part in chelation reactions. The ideal reaction between citric acid and metal nitrates for synthesis of Cu_1.5Mn_1.5O₄ spinel is presented as the following equation:

The mixture solution was adjusted to pH = 7.0 by slowly dropping ammonia, and successively stirred for 1 h to obtain a homogeneous solution. The prepared solution was subsequently heated at 70 °C with continuous electromagnetic stirring for dehydration until a high-viscous gel formed. The gel was put into an oven preheated to 130 °C for the adequate period of time to form the Cu–Mn-citric xerogel with a light green color. Then, the xerogel was ignited in atmosphere using a few drops of absolute ethanol as initiating combustion agent and burnt in a self-combustion manner with rapid evolution of a large quantity of fume, yielding voluminous powders. Finally, the as-burnt powders were annealed at 500–900 °C for 1 h with a heating rate of 5 °C min⁻¹ to obtain Cu_1.5Mn_1.5O₄ spinel ceramic pigments.

2.2 Preparation of TSSS paint coatings

The aluminum substrates with a purity of 99.5% and a dimension of 50 mm × 60 mm × 0.5 mm were degreased by ultrasonic treatment in absolute alcohol and acetone for 15 min, respectively. Then the substrates were completely cleaned up by deionized water. Well cleaned substrates were dried off by blowing with N₂. TSSS paint coatings were prepared according to the literature report.¹⁸ Pigment dispersion was firstly prepared by mixing the pigments (as-burnt powder or Cu_1.5Mn_1.5O₄ spinel ceramic powder) with the commercially corresponding binders (silicone modified by epoxy resin) and solvent in specific proportions and ground in a ball mill for 12 h. The mass ratio of the pigment to binder was kept at appropriate value, which the paint showed excellent film-forming property and the ratio of the binders was not too high to influence the thermal emittance of the paint coating. Subsequently, appropriate amount of curing agent was injected into the mixture system to form paint. Ultimately, the paint was sprayed on aluminium substrates to obtain TSSS paint coatings which were dried at room temperature for 24 h. The thickness of the TSSS paint coating varied with the spraying time. A diagram for the sample preparation is shown in Fig. 1.

Fig. 1 A diagram of employing sol–gel self-combustion method to fabricate Cu_1.5Mn_1.5O₄ spinel ceramic pigment based TSSS paint coatings.

2.3 Characterization techniques

TG and DTA curves for the xerogel sample were recorded with a Netzsch STA449C simultaneous thermal analyzer (Netsch, Frankfurt, Germany) at the temperature ranging from room temperature to 900 °C in air (flow rate 20 ml min⁻¹) with a heating rate of 10 °C min⁻¹. FTIR spectra were recorded on Bruker TENSOR 27 FTIR spectrometer by employing the KBr disk technique in the wavelength range 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹ using 32 scans for each sample. The phase identification of the as-burnt powder and annealed ceramic pigments were performed using X-ray diffraction (XRD) on a Rigaku D/max 2400/PC diffractometer with Cu Kα (λ = 1.5406 Å) at 40 kV and 150 mA. Diffraction patterns in the 10–80° region were recorded at a rate of 5 °C min⁻¹. The morphologies of obtained powders were investigated by using field emission scanning electron microscope (FE-SEM, SU8020, Hitachi).

The average of twenty measurements for the thickness of paint coatings were performed by QuaNix 4500 coating thickness gauge (Automation Dr NixGmbH & Co. KG, Köln, Germany). To evaluate the optical properties of the TSSS coatings, the reflectance spectrum in the wavelength range 0.3–2.5 μm was measured in a Perkin Elmer Lambda 950UV/VIS/NIR spectrometer equipped with an integrating sphere (module 150 mm). The total reflectance was measured relative to a BaSO₄ reference. The infrared near normal specular reflectance was measured from 2.5 to 20 μm on a Bruker TENSOR 27 FTIR spectrometer, equipped with an integrating sphere (A562-G/Q) using a gold plate as reference. The average of five measurements for each samples was performed to create reflectance spectra and the normal α_s and ε_T values were calculated using eqn (2) and (3).¹⁹ The solar absorptance α_s is theoretically defined as a weighted fraction between absorbed radiation and incoming solar radiation. Where λ is wavelength, R(λ) reflectance and I_s(λ) direct normal solar irradiance. It is defined according to ISO standard 9845-1, normal radiance, AM1.5.

Normal thermal emittance ε_T is equally a weighted fraction but between emitted radiation and the Planck black body distribution, I_b(λ,T), at temperature T.

Accordingly, thermal emittance of sample is denoted as ε₁₀₀ in this work, when it is obtained at 100 °C.

3 Results and discussion

3.1 Characterizations of the Cu_1.5Mn_1.5O₄ ceramic pigments

3.1.1 TG-DTA characterization of the xerogel. Thermal analysis is performed to understand the decomposition behavior of the xerogel precursors and the formation of metal oxides. According to the TG-DTA trace, the phase evolution of the xerogel precursor is shown in Fig. 2. Four clear weight losses are plotted in the TG curve, and simultaneous exothermic peaks emerge in the DTA curve. The first broad weight loss of ∼17% from room temperature to 189 °C is assigned to the dehydration of residual water in the xerogel, which appears the exothermic peak at around 192 °C on DTA curve. The second weight loss of ∼23%, between 198 °C and 236 °C on TG curve, corresponds to the autocatalytic oxidation–reduction reaction between the nitrate and citrate acid.²⁰ The third weight loss of ∼18% occurring between 236 °C and 264 °C can be related to decomposition of PEG 200 in xerogel. The last weight loss of ∼8% at the temperature ranging from 265 °C to 298 °C, concurrent with the weaker exothermic peak at about 279 °C on DTA curve, is ascribed to the decomposition of residual organic compound after combustion. Thereafter, there is no evidence of the phase evolution taking place up to 900 °C and the mass of the sample almost maintains constant above 300 °C in the TG curve, indicating that the organics are burnt out and inorganic compounds are formed. Therefore, it is reasonable that calcination temperature for powders should be kept at 300 °C or above after sol–gel self-combustion. Phase evolution of the sample need to be supported from the spectroscopic and the XRD analysis.

Fig. 2 TG and DTA curves of xerogel precursor.

3.1.2 FTIR study. Chemical and structural changes that happen in self-combustion and calcinations process can be monitored by a spectroscopic analysis. This may be helpful to elucidate the combustion reaction mechanism. The FTIR spectra of the xerogel precursor and ceramic pigments annealed at different temperatures are plotted in Fig. 3. The xerogel spectrum illustrates that the broad band centered in the region of 3600–3000 cm⁻¹ can be assigned to symmetric and anti-symmetric stretching vibration modes of the hydroxyl.²¹ These hydroxyls are derived from citric acid and residual water. Only the anti-symmetric stretching vibration band of carboxylate ion related to citric acid are situated at 1627 cm⁻¹, while the lower strength of the symmetric band at about 1400 cm⁻¹ is covered up partially by characteristic band of N–O.²² The absorption bands at 1760 cm⁻¹ and 1730 cm⁻¹ respectively could not be found in despite of two types of free carboxyl groups in citric acid.²³ Therefore, it suggests that citric acid serves as tridentate ligand. The band around at 1384 cm⁻¹ and 826 cm⁻¹ are the typical vibration peak of nitrate species.²⁴ After self-combustion, the number and intensity of peaks decrease obviously, which reveals that most of the NO₃⁻ and COO⁻ take part in the combustion reaction.²⁵ Hence, the self-combustion for xerogel can be described as a thermally induced anionic redox reaction, in which the citrate ions act as a reductant and nitrate ions as an oxidant.²⁶ In terms of the as-burnt powders, very weak bands (1627 cm⁻¹ and 1384 cm⁻¹) can still be found, which shows the existence of some organic residues. For both of the as-burnt and the annealed powders, a series of new absorption bands at 616–500 cm⁻¹ indicate the presence of metal oxide.^27,28

Fig. 3 FTIR spectrophotometer for xerogel, as-burnt powder, and powders annealed at different temperatures.

3.1.3 XRD study. The XRD pattern analysis is performed to determine the crystal structure of the prepared powders, and the results are shown in Fig. 4. For the as-burnt powder, several intense diffraction peaks indicate that it is composited of mixed phase, such as Cu_1.5Mn_1.5O₄ (ICDD-PDF No. 35-1172), Mn₂O₃ (ICCD-PDF No. 41-1442), CuO (ICCD-PDF No. 41-254), and Mn₃O₄ (ICCD-PDF No. 18-803). Take the thermal analysis into account, it can be concluded that the temperature reached in sol–gel combustion process is sufficient for forming the Cu_1.5Mn_1.5O₄ spinel phase. For powders heat-treatment at 500 °C, the XRD patterns are well matched with the crystalline Cu_1.5Mn_1.5O₄ spinel structure, and Mn₃O₄ peak disappears and the intensity of the Mn₂O₃ main peak decreases, which could be attributed to the tetragonal Mn₃O₄ transformed into cubic Mn₂O₃ structure with elevating annealing temperature,²⁹ and then the solid-state redox reaction takes place between CuO and Mn₂O₃.³⁰ In our case, we could expect that Cu_1.5Mn_1.5O₄ spinel phase appears at lower temperatures due to the powders have been subjected to combustion process. In the case of powders annealed at 700 °C and 900 °C, the diffraction peaks mainly correspond to the Cu_1.5Mn_1.5O₄ spinel phase, but coexistence of the CuO phase. The copper-rich spinel (i.e. Cu_1.5Mn_1.5O₄) can be fabricated only at annealing temperature not much higher than 500 °C in literature.^13,31 Once the temperature exceeds 500 °C, the Cu–Mn–O phase diagram³² indicates that CuO will segregate. The diffraction intensity of the CuO based on powder annealed at 900 °C is stronger than that of powder annealed at 700 °C, which shows that the extent of the segregation gradually deepens with the annealing temperature. Furthermore, with elevating the annealing temperature, the diffraction peaks of Cu_1.5Mn_1.5O₄ spinel phase become sharper and more intense, and the full width at half-maximum (FWHM) decreases, implying better crystallinity and increasing of the particles size.

Fig. 4 XRD patterns of the as-burnt powder and powders annealed at different temperatures.

3.1.4 FE-SEM micrographs. Fig. 5 presents the FE-SEM images of as-burnt powders and Cu_1.5Mn_1.5O₄ spinel ceramic powders annealed at different temperatures for 1 h. As seen from Fig. 5a, seeing that a large amount of gases are liberated in the self-combustion process, the as-burnt powder exhibits the relatively spongy agglomeration and porosity. However, as shown in Fig. 5b, the formation of multigrain agglomeration are observed in ceramic powder consisting of very fine crystallites, as they tend to fuse together to form regular crystal structure. It can be clearly found that the average sizes of the Cu_1.5Mn_1.5O₄ spinel particles dramatically increase and the particles become interconnected with elevating the annealing temperature from 500 to 900 °C. It is noted that particles annealed at 900 °C possess a polyhedral shape. The formation of extraordinary morphologies can be conceivably explicated by means of the existence of the large microscopic air gaps between the particles leading to a reduction in diffusion rate which impedes the particle growth.³³ As seen in corresponding inset of the FE-SEM micrographs, the representation of the as-burnt powders and annealed ceramic powders in color is dark-gray. As a result, they can be taken full advantage of as the candidates for solar-absorbing materials. It has been reported that solar absorptance of those pigments in UV-Vis-NIR range (0.3–2.5 μm) show no clear different due to they are same in color.^33,34 Therefore, the effect of the thermal emittance of the powder on the ultimate TSSS paint coating achieving high spectral selectivity is of vital importance. To eliminate the artificial operation error, the same qualities of powders specimen are put into the same containers and are flush with the edge of the containers by the aid of a piece of glass. Thereafter, the preparation of the powder specimen is employed in IR reflectance spectrum measurement. Fig. 6 depicts the reflectance spectrum of each powder sample and the corresponding emittance calculated from the reflectance spectrum. Thermal emittance of each powder sample shows no durative descent with elevating the annealing temperature. Though the perfecting of crystals at higher temperature can cause the powder increasing in reflectance and obtaining the lower thermal emittance, the influence of other factors on the thermal emittance should be taken into account and investigated further. Considering Cu_1.5Mn_1.5O₄ spinel with two Jahn–Teller active ions, the presence of ingredient segregation is likely to cause the lattice distortion with increasing the annealing temperature, which probably changes the vibration and rotation modes of molecules to enhance the intrinsic absorption of the particles.³⁴ As for powders annealed at 700 to 900 °C, the results are well agreement with the reflectance decline in infrared region, leading to the thermal emittance raise. Hence, the discussion above is conducive to explain that the thermal emittance of powder annealed at 900 °C exceeds that of powder annealed at 700 °C.

Fig. 5 FE-SEM images and corresponding (inset) photographs of the powders: (a) as-burnt powder, (b) powder annealed at 500 °C, (c) powder annealed at 700 °C, and (d) powder annealed at 900 °C, respectively.

Fig. 6 The reflectance spectra in the range from 4000 to 500 cm⁻¹ of as-burnt powder and powders annealed at different temperatures with corresponding thermal emittance (ε₁₀₀).

3.2 Characterizations of the paint coatings

3.2.1 Spectral selectivity. Considering that spinel Cu_1.5Mn_1.5O₄ films deposited on aluminum substrates have been used as spectrally selective absorber coating, the Cu_1.5Mn_1.5O₄ spinel ceramic powders are employed as solar-absorbing pigments to prepare TSSS paint coatings. Moreover, transition metal oxides have been widely used as pigments for paint coatings in the literature³⁵ due to their semiconductor properties. Therefore, we compare the performance of paint coatings using the as-burnt powders and Cu_1.5Mn_1.5O₄ spinel ceramic powders annealed at various temperatures as pigments for fabricating TSSS paint coatings. Sample A, sample B, sample C, and sample D represent paint coatings using as-burnt powder, ceramic powders annealed at 500 °C, 700 °C, and 900 °C, respectively. The reflectance spectra of those paint coatings in the range from 0.3 to 20 μm are shown in Fig. 7a–d. The absorptance (α_s), thermal emittance (ε₁₀₀), the coating thicknesses, and corresponding standard deviation, are recorded in Table 1. As indicated in Table 1, both the values of α_s and ε₁₀₀ of each sample increase with the coating thickness and the increase of the ε₁₀₀ is remarkable than that of α_s. Besides, the thermal emittance of the paint coatings rising and falling trends are in accordance with that of pigment powders (as-burnt powder and ceramic powders annealed at various temperatures). Furthermore, the thermal emittance of paint coatings are much lower than that of as-synthesized powders due to the low thermal emittance of aluminum substrates (ε₁₀₀ = 0.04) employed to fabricate TSSS paint coatings. The α_s values of paint coating based on as-burnt powder ranges from 0.874 to 0.909 while ε₁₀₀ ranges from 0.223 to 0.390, which displays the lowest spectral selectivity. As can be seen from the Fig. 7b–d, for the paint coatings based on ceramic powders we observe that the absorption edge shifts towards long wavelengths (0.9–2.0 μm), which leads to the higher α_s for paint coatings based on ceramic powders than that of paint coating based on as-burnt powder. In terms of sample B, the values for these two parameters are α_s = 0.910–0.921 and ε₁₀₀ = 0.286–0.377, which indicates the spectral selectivity is much lower than that of sample C and sample D. The values of α_s and ε₁₀₀ for sample D are typically 0.920–0.929 and 0.266–0.382, which exhibits the highest α_s and ε₁₀₀ among all those samples. Compared with other three samples, sample C based on ceramic pigment annealed at 700 °C has the optimum spectral selectivity (α_s = 0.914–0.923 and ε₁₀₀ = 0.244–0.357). Therefore, the spectral selectivity for TSSS paint coatings based on ceramic pigment annealed at 700 °C can bear comparison with that of other paint coatings reported in literature.^36,37

Fig. 7 The reflectance spectra of paint coatings with different thickness based on (a) as-burnt powder, (b) powder annealed at 500 °C, (c) powder annealed at 700 °C, and (d) powder annealed at 900 °C.

Table 1 Optical parameters and the thickness of TSSS paint coatings based on the as-burnt and annealed pigments

Pigment calcining temp.	Coating sample	d^a (μm)	αs^b	ε100^c
a d (μm) the thickness of the paint coatings.b αs the solar absorptance of the paint coatings.c ε100 thermal emittance of the paint coatings at 100 °C.
As-burnt	A1	1.63 ± 0.07	0.874 ± 0.002	0.223 ± 0.001
	A2	2.19 ± 0.05	0.880 ± 0.002	0.233 ± 0.002
	A3	2.84 ± 0.06	0.890 ± 0.002	0.316 ± 0.001
	A4	3.77 ± 0.05	0.899 ± 0.001	0.328 ± 0.002
	A5	4.52 ± 0.05	0.909 ± 0.002	0.390 ± 0.002
500 °C	B1	1.96 ± 0.06	0.910 ± 0.002	0.286 ± 0.001
	B2	2.43 ± 0.05	0.913 ± 0.001	0.325 ± 0.001
	B3	3.65 ± 0.06	0.915 ± 0.001	0.330 ± 0.001
	B4	4.67 ± 0.05	0.917 ± 0.002	0.367 ± 0.002
	B5	4.87 ± 0.07	0.921 ± 0.001	0.377 ± 0.002
700 °C	C1	1.78 ± 0.05	0.914 ± 0.002	0.244 ± 0.002
	C2	2.16 ± 0.05	0.917 ± 0.002	0.290 ± 0.001
	C3	3.08 ± 0.06	0.918 ± 0.001	0.325 ± 0.001
	C4	4.04 ± 0.07	0.922 ± 0.001	0.348 ± 0.002
	C5	4.46 ± 0.05	0.923 ± 0.002	0.357 ± 0.001
900 °C	D1	1.85 ± 0.05	0.920 ± 0.002	0.266 ± 0.001
	D2	2.45 ± 0.07	0.921 ± 0.001	0.302 ± 0.002
	D3	3.96 ± 0.07	0.922 ± 0.001	0.345 ± 0.002
	D4	4.68 ± 0.05	0.926 ± 0.001	0.354 ± 0.001
	D5	5.12 ± 0.06	0.929 ± 0.002	0.382 ± 0.002

---

3.2.2 The accelerated thermal stability test. Apart from good optical properties, the other momentous issue related to solar absorber coatings cares about their long term stability during collector's service lifetime. The influence of three kinds of degradation processes, such as high temperature, high humidity, and aggressive airborne pollutant on service lifetime of the selective absorber coating have been reported.³⁸ Considering that a selective absorber coating is strictly insulated under a transparent glass cover, thermal stability of the absorber coating is very crucial and the absorber would definitely degrade with time at the operating temperatures, which shrinks the life time and eventually leads to failure.³⁹ Therefore, we evaluate the service lifetime of the TSSS paint coating based on its thermal behavior in low-to-mid temperature range. The accelerated thermal stability test is performed using a circulating air furnace based on the PC (performance criterion) value (PC = −Δα_s + 0.5Δε₁₀₀) of IEA SHC Task 27.⁴⁰ Acceptable PC values have to be lower than 0.05, which is derived from the definition that the influence of ageing on the efficiency of the collector must be less than 5% after 25 years.⁴¹ Specific method concerning the accelerated thermal stability study is discussed in the literature.^42,43 Seeing that the test temperature depended on the absorptance and thermal emittance of the TSSS paint coating, a testing temperature of 227 °C was selected for the accelerated thermal stability studies for sample C2 (α_s = 0.917, ε₁₀₀ = 0.290) and C4 (α_s = 0.922, ε₁₀₀ = 0.348), simplifying the test and acquiring comparable result. Testing times were 18, 36, 75, 150, 300 and 600 h, and after each interval time, samples were taken out from the furnace and measured.

To make sure an obvious visual distinction among the reflectance spectra of samples before and after the accelerated thermal test, only the reflectance spectra in the typical time interval (0, 36, 150, 600 h) are vividly plotted in Fig. 8a and b. Comparing the reflectance spectra, a remarkable distinguish in reflectance is observed although the shape of them is very similar. Consequently, it is concluded that the variation of organic compounds in paint coating brings about the disparity between the paint coating before and after the accelerated thermal test.

Fig. 8 The reflectance spectra of paint coatings before and after accelerated thermal stability aging test at 227 °C with the interval time of 36, 150, 600 h: (a) sample C2 and (b) sample C4.

Comparison of the average values of absorptance and emittance, together with the corresponding PC values for the samples, have been recorded. As shown in Table 2, the variation of absorptance before and after the thermal test (Δα_s) can be considered to be negligible, but the emittance has a slight change after the accelerated thermal test. Thermal emittance for both sample C2 and C4 decrease after 18 h at the first exposure at 227 °C. The decline of thermal emittance may be related to removing some of residual organic compounds and forming thin coatings with the heat treatment. The remarkable distinction emerging from thermal emittance for sample C2 and C4 can be attributed to the different thickness of the paint coatings. The relation between the thermal emittance and structural or chemical change of the paint coatings need to be studied more carefully in future. Once the accelerated thermal test is carried out, the change of emittance (Δε₁₀₀) increases with the testing time. Considering the marginal change of the absorptance, the PC values of the paint coatings are more influenced by the change of thermal emittance. The results in Table 2 show that the corresponding PC values of paint coatings also increase with the test time, but the final PC values are far below the critical value (PC < 0.01 after 600 h) for which an absorber can be considered qualified. Furthermore, none of the paint coatings shows observable visual changes and the aluminium substrate exposed. The adhesion property of the TSSS paint coating with a dimension of 50 mm × 85 mm × 0.5 mm is evaluated by the circle-drawing test after it is subjected to heat treatment at 227 °C for 600 h. Although the circle-drawing test is relatively simple and crude, it quite clearly reveals that the paint coating shows the good adhesion for aluminium substrate. It should point out that the dimension of paint coating is decided by the test scope of the instrument. The digital photos of the circle-drawing test for TSSS paint coatings are illustrated in Fig. 9. It is obvious that there is no damage for TSSS paint coatings after the circle-drawing test. Therefore, we can expect that those TSSS paint coatings exhibit good adhesion properties for aluminium substrate. On the basis of the mentioned discussion, it is demonstrated that our paint coatings are qualified for the accelerated thermal stability tests and can be safely used in solar collectors (low-to-mid temperature region) within an extended timeframe.

Table 2 Optical parameters and PC values for sample C2 and sample C4 after accelerated thermal stability tests

Testing times	Sample C2			Sample C4
	αs^a	ε100^b	PC^c	αs	ε100	PC
a αs the solar absorptance of the paint coatings.b ε100 thermal emittance of the paint coatings at 100 °C.c PC performance criterion values of the paint coatings after the accelerated thermal stability tests.
0 h	0.917 ± 0.002	0.290 ± 0.001	—	0.922 ± 0.001	0.348 ± 0.002	—
18 h	0.917 ± 0.001	0.274 ± 0.001	−0.008	0.921 ± 0.001	0.296 ± 0.001	−0.025
36 h	0.918 ± 0.001	0.277 ± 0.001	−0.008	0.922 ± 0.001	0.298 ± 0.002	−0.020
75 h	0.918 ± 0.001	0.278 ± 0.001	−0.007	0.923 ± 0.001	0.350 ± 0.001	0
150 h	0.917 ± 0.001	0.280 ± 0.001	−0.005	0.922 ± 0.001	0.357 ± 0.001	0.005
300 h	0.919 ± 0.001	0.280 ± 0.001	−0.007	0.923 ± 0.001	0.358 ± 0.002	0.004
600 h	0.919 ± 0.001	0.295 ± 0.002	0.001	0.925 ± 0.002	0.362 ± 0.002	0.004

---

Fig. 9 The digital photos of the circle-drawing test for sample C2 and sample C4.

4 Conclusions

The citric acid based sol–gel combustion method is shown to be an attractive method to prepare Cu_1.5Mn_1.5O₄ spinel ceramic pigments. The XRD analysis suggests that the heat generated is enough for the formation of spinel Cu_1.5Mn_1.5O₄ during the self-combustion process, but many impurity phases coexist. Pure Cu_1.5Mn_1.5O₄ spinel can be obtained after the as-burnt powder is annealed at temperature ranging from 500 to 900 °C for 1 h. FE-SEM morphology analysis shows that very fine crystallites tend to fuse together to form regular crystal structure when the as-burnt powder is subsequently subjected to calcinations process. Both of the as-burnt powders and ceramic powders are used as pigments to prepare TSSS paint coatings on aluminium substrate by convenient and cost-effective spray-coating technique. Comparing all TSSS paint coatings fabricated, it can be found that the paint coatings based on ceramic pigments annealed at 700 °C exhibit the optimum α_s values of 0.914–0.922 and ε₁₀₀ values of 0.244–0.358. These values are comparable with literature data for paint coatings. Except for good spectral selectivity, the accelerated thermal stability tests reveal our paint coatings can reach the qualified requirement and have great potential as a candidate for solar collector in low-to-mid temperature range.

Acknowledgements

This work was financially supported by, the Western Light Talents Training Program of CAS, the Science and Technology Support Program of Gansu Province (Grant 1304GKCA025), and the National Natural Science Foundation of China (Grant 51402315).

References

1. M. Rubat du Merac, I. E. Reimanis and H. J. Kleebe, J. Am. Ceram. Soc., 2015, 98, 2130–2138.
2. H. Kim, D. H. Seo, H. Kim, I. Park, J. Hong, K. Y. Park and K. Kang, Chem. Mater., 2012, 24, 720–725.
3. A. B. Rinkevich, A. V. Korolev, M. I. Samoylovich, S. M. Klescheva and D. V. Perov, J. Magn. Magn. Mater., 2016, 399, 216–220.
4. H. J. Zhao, M. G. Lin, K. G. Fang, J. Zhou, Z. Y. Liu, G. F. Zeng and Y. H. Sun, RSC Adv., 2015, 5, 67630–67637.
5. S. J. Peng, L. L. Li, Y. X. Hu, M. Srinivasan, F. Y. Cheng, J. Chen and S. Ramakrishna, ACS Nano, 2015, 9, 1945–1954.
6. S. Wijewardane and D. Y. Goswami, Renewable Sustainable Energy Rev., 2012, 16, 1863–1873.
7. Q. F. Geng, X. Zhao, X. H. Gao and G. Liu, J. Am. Ceram. Soc., 2011, 94, 827–832.
8. R. Elyasi, S. M. Nowee, H. A. Namaghi and N. Ramezanian, Sol. Energy Mater. Sol. Cells, 2015, 141, 57–70.
9. A. Amri, Z. T. Jiang, T. Pryor, C. Y. Yin and S. Djordjevic, Renewable Sustainable Energy Rev., 2014, 36, 316–328.
10. L. KalužA, A. Šurca-Vuk, B. Orel, G. Dražič and P. Pelicon, J. Sol-Gel Sci. Technol., 2001, 20, 61–83.
11. L. Kaluža, B. Orel, G. Dražič and M. Kohl, Sol. Energy Mater. Sol. Cells, 2001, 70, 187–201.
12. N. EI Mahallawy, M. Shoeib and Y. Ali, J. Coat. Technol. Res., 2014, 11, 979–991.
13. R. Bayón, G. S. Vicente, C. Maffiotte and Á. Morales, Sol. Energy Mater. Sol. Cells, 2008, 92, 1211–1216.
14. R. Bayón, G. S. Vicente, C. Maffiotte and Á. Morales, Renewable Energy, 2008, 33, 348–353.
15. R. Bayón, G. S. Vicente and Á. Morales, Sol. Energy Mater. Sol. Cells, 2010, 94, 998–1004.
16. S. Pal, D. Diso, S. Franza, A. Licciulli and L. Rizzo, J. Mater. Sci., 2013, 48, 8268–8276.
17. S. Alamolhoda, S. M. Mirkazemi, T. Shahjooyi and N. Benvidi, J. Alloys Compd., 2015, 638, 121–126.
18. I. Jerman, M. Mihelčič, D. Verhovšek, J. Kovač and B. Orel, Sol. Energy Mater. Sol. Cells, 2011, 95, 423–431.
19. B. Cheng, K. K. Wang, K. P. Wang, W. Jiang, B. J. Cong, C. L. Song, S. H. Jia, G. R. Han and Y. Liu, Sol. Energy Mater. Sol. Cells, 2015, 133, 126–132.
20. P. P. Hankare, K. R. Sanadi, K. M. Garadkar, D. R. Patil and I. S. Mulla, J. Alloys Compd., 2013, 553, 383–388.
21. J. Y. Meng, X. F. Tang, W. Li, H. F. Shi and X. X. Zhang, Thermochim. Acta, 2013, 558, 83–86.
22. A. I. Mitsionis, T. Vaimakis and C. Trapalis, Ceram. Int., 2010, 36, 623–634.
23. W. D. Li, J. Z. Li and J. K. Guo, J. Eur. Ceram. Soc., 2003, 23, 2289–2295.
24. J. Li, Y. B. Pan, F. G. Qiu, Y. S. Wu and J. K. Guo, Ceram. Int., 2008, 34, 141–149.
25. Y. Y. Meng, M. H. He, Q. Zeng, D. L. Jiao, S. Shukla, R. V. Ramanujan and Z. W. Liu, J. Alloys Compd., 2014, 583, 220–225.
26. L. Yang, G. X. Xi and J. J. Liu, Ceram. Int., 2015, 41, 3555–3560.
27. S. A. Hosseini, A. Niaei, D. Salari, M. C. Alvarez-Galvan and J. L. G. Fierro, Ceram. Int., 2014, 40, 6157–6163.
28. A. F. Costa, P. M. Pimentel, F. M. Aquino, D. M. A. Melo, M. A. F. Melo and I. M. G. Santos, Mater. Lett., 2013, 112, 58–61.
29. S. S. Falahatgar, F. E. Ghodsi, F. Z. Tepehan, G. G. Tepehan and İ. Turhan, Appl. Surf. Sci., 2014, 289, 289–299.
30. R. E. Vandenberghe, G. G. Robbrecht and V. A. M. Brabers, Mater. Res. Bull., 1973, 8, 571–579.
31. S. Behar, P. Gonzalez, P. Agulhon, F. Quignard and D. Świerczyński, Catal. Today, 2012, 189, 35–41.
32. P. Wei, M. Bieringer, L. M. D. Cranswick and A. Petric, J. Mater. Sci., 2010, 45, 1056–1064.
33. Q. F. Geng, X. Zhao, X. H. Gao, S. R. Yang and G. Liu, J. Sol-Gel Sci. Technol., 2012, 61, 281–288.
34. Q. F. Geng, X. Zhao, X. H. Gao, H. C. Yu, S. R. Yang and G. Liu, Sol. Energy Mater. Sol. Cells, 2012, 105, 293–301.
35. B. Orel, H. Spreizer, L. S. Perše, M. Fir, A. Š. Vuk, D. Merlini, M. Vodlan and M. Köhlc, Sol. Energy Mater. Sol. Cells, 2007, 91, 93–107.
36. L. S. Perše, M. Mihelčič and B. Orel, Mater. Chem. Phys., 2015, 149–150, 368–377.
37. B. Japelj, A. Š. Vuk, B. Orel, L. S. Peršea, I. Jerman and J. Kovač, Sol. Energy Mater. Sol. Cells, 2008, 92, 1149–1161.
38. R. Kunič, M. Koželj, B. Orel, A. Š. Vuk, A. Vilčnik, L. S. Perše, D. Merlini and S. Brunold, Sol. Energy Mater. Sol. Cells, 2009, 93, 630–640.
39. A. Amri, Z. T. Jiang, N. Wyatt, C. Y. Yin, N. Mondinos, T. Pryor and M. M. Rahman, Ceram. Int., 2014, 40, 16569–16575.
40. B. Carlsson, IEA SHC Task 27, Swedish, Borås, 9, 2004.
41. S. Brunold, U. Frei, B. Carlsson, K. Möller and M. Köhl, Sol. Energy, 2000, 68, 313–323.
42. B. Carlsson, K. Möller, M. Köhl, U. Frei and S. Brunold, Sol. Energy Mater. Sol. Cells, 2000, 61, 255–275.
43. M. Köhl, M. Heck, S. Brunold, U. Frei, B. Carlsson and K. Möller, Sol. Energy Mater. Sol. Cells, 2004, 84, 275–289.

---