Introduction

Solar selective coatings on the solar receiver are selected according to the specific characteristics that achieve high solar absorptance (αsol) in the visible region and low thermal emittance (εtherm) in the infrared region [1]. There are many techniques for selective coating preparation such as evaporation, sputtering, electro-chemical deposition (anodization), chemical vapour deposition, spray pyrolysis, sol–gel/dip-coating, electro plating, painting and other methods [2]. Anodization is a common technique used to protect the aluminium (Al) from corrosion by creating a thin oxide film on the Al surface. The solar selectivity of the solar absorber materials can be optimized by varying the metal and ceramic constituent elements, thickness, concentration, shape, size, and orientation [1]. Several metal pigments can be used to deposit into the pores of the film as a solar selective absorber. The thickness of Al2O3 based cermet coatings was considered in previous reports [3,4,5], which correlated with spectral selective properties. In addition, variation of the content of pigment or composite coating on the surface has an influence on the spectral selectivity (αsol and εtherm) of the solar selective absorbers [2, 3]. To improve the spectral selectivity of the solar selective absorber and high corrosion resistance, several pigments deposited in the Al2O3 film have been studied with different techniques, e.g. coatings of black Cu–Ni, black Ni–Co by electro-deposition [6, 7], Co–Al2O3 by anodization [8], Cu–CuAl2O4 hybrids deposited in anodic aluminium oxide (AAO) by electrochemical processes [9], carbon nano-particles embedded in SiO2, ZnO and NiO matrices by a sol–gel technique [10], W–Al2O3 cermet, Mo–Al2O3, Pt–Al2O3, Ag–Al2O3 nanocermet, and CrNxOy/SiO2 coated with a sputtering technique [11,12,13,14]. Tin (Sn) has been considered a good candidate pigment embedded into the pores of Al2O3 film on Al substrate which has been successfully prepared by the anodization process as Sn–Al2O3 [15, 16]. The corrosion resistance of Sn–Al2O3 is higher than that of the Ni–Al2O3 [16]. Presently, there are a few prior researches on the influence of Sn content on phase, morphology, thermal conductivity, and solar selectivity of Al2O3-based cermet coatings.

Therefore, the purposes of the present study are preparation of the Sn–Al2O3 based solar selective absorber by anodic anodization with three different contents of Sn pigment. Phase, morphology and chemical distributions of the Sn–Al2O3 coatings were characterized. The influence of Sn contents on the αsol, εtherm and thermal conductivity of the coatings were also investigated, including the discussion of solar radiation and heat transfer mechanism through the Sn–Al2O3 coatings.

Experimental section

Preparation of anodized aluminium

Aluminium (Al) sheet of 9 cm wide × 160 cm long with a thickness of 0.02 cm was used as a substrate. It was followed by the anodization process, under suitable conditions calculated by the surface area of the samples. First, the aluminium sheet was degreased with a commercial cleaning solution, and then etched by NaOH solution, and immersed in 165 g/L sulphuric acid solution. The Al substrate was connected with an anode (+) and a lead plate was used as the cathode (−), and then a direct current of electrical density with 1000 A at the constant potential of 15–17 V at 20 °C was applied for 45 min to achieve the Al2O3 film thickness of 9.1–9.5 µm. Subsequently, the formed Al2O3 film was directly transferred to the colouring bath containing tin (II) sulphate (SnSO4). The Sn pigment was deposited into the pores of Al2O3 film for 5, 7 and 10 min, which coded as Sn–Al2O3-5, Sn–Al2O3-7 and Sn–Al2O3-10, respectively. Then, the porous Al2O3 was sealed in a hot water tank at 50–60 °C for 30 min.

Instrumentation

The phase and morphology of the samples were characterized by X-ray diffraction technique (XRD, PHILIPS X’Pert-MPD) with Cu–K α radiation in the 2θ range of 20–90°, a low vacuum scanning electron microscope equipped with an energy dispersive X-ray analyser (LV-SEM–EDX, JSM-5910 JEOL). The reflectance (R) was measured using the ultraviolet visible near infrared (UV–Vis–NIR) spectrophotometer (Shimadzu UV-3101PC) in the wavelength of 300–2500 nm, and the Fourier transform infrared spectrophotometer (FTIR, Bruker Tensor 27) in the wavelength of infrared region (2500–25,000 nm). The αsol and the εtherm were calculated from the measured R spectrum by the following equations [2]:

$$ \alpha _{\text{sol}} = \frac{{\int\limits_{{0.3\upmu{\text {m}}}}^{{2.5\upmu{\text {m}}}} {I_{s} \left( \lambda \right)\left( {1 - R\left( \lambda \right)} \right)d\lambda} }}{{\int\limits_{{0.3\upmu{\text {m}}}}^{{2.5\upmu{\text {m}}}} {I_{s} \left( \lambda \right)d\lambda } }} $$
(1)
$$ \varepsilon _{\text{therm}} = \frac{{\int\limits_{{2.5\upmu{\text{m}}}}^{{25\upmu{\text{m}}}} {I_{b} \left( \lambda \right)\left( {1 - R\left( \lambda \right)} \right)d\lambda } }}{{\int\limits_{{2.5\upmu{\text{m}}}}^{{25\upmu{\text{m}}}} {I_{b} \left( \lambda \right)d\lambda } }} $$
(2)

where I s (λ) is the solar spectral irradiance at AM 1.5, and R(λ) is the measured reflectance spectrum of the coating. Furthermore, the thermal diffusivity (κ d ) was measured by laser flash analyser (LFA 447 NanoFlash-A) over the temperature of 25–275 °C according to ASTM 1461-13 standard. The thermal conductivity (k) was then calculated with density (ρ) and specific heat (C p ) as follows:

$$ k = k_{d} \rho C_{p} $$
(3)

Results and discussion

Figure 1 shows the XRD patterns of Sn–Al2O3 with various Sn contents, compared with the JCPDS database No. 01-1176, 01-1243 and 01-0926 for Al, Al2O3 and Sn phases, respectively. The phase of Al and Sn were indexed for all samples, and the phase intensity of Sn was increased by increasing the Sn contents. The crystallite size was also calculated by Scherrer’s equation [17] using the Sn (200) peak in the XRD spectrum.

Fig. 1
figure 1

The XRD patterns of the Sn–Al2O3 with different Sn contents

Full size image
$$ D \, = \, \frac{K\lambda }{\beta \cos \theta } $$
(4)

K is the crystallite shape factor which usually takes a value of 0.9, λ is the X-ray wavelength of Cu–K α radiation (0.15406 nm), β is the full width at half maximum (FWHM) in radians and θ is the Bragg’s angle of the (200) peak of Sn. The calculated results for the Sn–Al2O3-5, Sn–Al2O3-7 and Sn–Al2O3-10 were 23.24, 23.24 and 46.48 nm, respectively. It is observed that the crystallite size of the Sn–Al2O3-10 sample is double the size of the others, indicating that the crystal growth of Sn inside the Al2O3 pores continuously occurred in the colouring bath beyond 7 min. According to the double intensity for (200) plane of Sn, the crystallinity of the highest content of Sn is also improved due to the sufficient time in the colouring process. However, the Al2O3 phase was not found in the XRD patterns because its structure is the amorphous phase, which is consistent with previous studies of W–Al2O3 [11], Pt–Al2O3 [18] and Mo–Al2O3 [13].

Figure 2 presents the SEM images for surface morphology of the Sn–Al2O3 coatings prepared by the anodization process. In Fig. 2a–c, the surface of the samples was quite homogeneous similar to the ordinary anodized Al substrate prepared to be a solar absorber black coatings [3, 19, 20]. The shallow marks of the surfaces occurred during the phenomenon of sealing the Al2O3 pores. Figure 2d–f indicates that the Al2O3 films were formed with a thickness of 9.1–9.5 µm, with a perpendicular pore structure and compact Al2O3 barrier on Al substrate, corresponding with Ni–Al2O3 selective absorber [21]. Fundamentally, the elements with higher atomic numbers exhibit the lighter region in that area compared to lower atomic numbers. It is observed that there are different contrast regions on the Sn–Al2O3 films, due to the different atomic numbers of the elements. The Al2O3 phase exhibit dark grey regions in all the samples. While, Sn has its highest atomic number in the Al2O3 phase, the Sn rich reveal a spike shape and light region close to the interface of Al substrate, according to a previous study [16]. The increasing and sharpness of the lighter regions of each sample depend on an increase in the Sn content of the Al2O3 pores.

Fig. 2
figure 2

LV-SEM images of (a–c) surface and (d–f) sectional image of Sn–Al2O3-5, Sn–Al2O3-7 and Sn–Al2O3-10

Full size image

To ascertain the increase of Sn content in the porous Al2O3 films, SEM–EDX with line scan analysis techniques were used (as shown in Fig. 3). All the samples consist of detected Al, O and Sn elements, corresponding to the XRD pattern. The Sn content exhibited a high count rate close to the interface of the Al substrate, and the accumulation of Sn according to the incremental deposition time.

Fig. 3
figure 3

SEM–EDX line scanning profiles of a Sn–Al2O3-5, b Sn–Al2O3-7 and c Sn–Al2O3-10 coatings

Full size image

In addition, the chemical distribution over a cross section, the surface of the Sn–Al2O3 coatings indicated the different chemical distributions in each layer, as presented by the EDX mapping images in Fig. 4. The fundamental elements of these samples were dispersed in different densities in each region. Al displayed the highest density in the Al substrate, while the O revealed the highest density in the Al2O3 layer, and the Sn also appeared in the Al2O3 layer but the highest density is at the interface of the Al substrate, consistent with the EDX line scanning images (Fig. 3). Furthermore, the highest concentration of Sn in all the samples was found at 1 µm distance from the interface. The dispersion distance of Sn, from the interface, in Sn–Al2O3-5, is 1 µm, increasing to 4 µm and 7 µm in the Sn–Al2O3-7 and the Sn–Al2O3-10, respectively.

Fig. 4
figure 4

SEM image and EDX mapping analysis of the a Sn–Al2O3-5, b Sn–Al2O3-7 and c Sn–Al2O3-10 coatings

Full size image

Figure 5a presents the measured spectral reflectance (R) of Sn–Al2O3 with different Sn contents in the whole wavelength range of UV–Vis-NIR regions. All the Sn–Al2O3 samples exhibit the low R in the wavelength range of 300–2000 nm. However, the R of all samples gradually increased with the increasing wavelength due to the maximum of the scattering and reflectance efficiencies at the high wavelength [1]. With increasing Sn content in the pores of Al2O3, the spectral R of the samples relatively decreased in the wavelength range of UV–Vis-NIR regions due to the thickness of Al2O3 layer behaved as anti-reflection (AR) layer reducing the reflection between the air and the Sn pigment layer [18, 22]. The αsol of the samples are calculated and summarized in Table 1, together with the previous reports. The αsol of the Sn–Al2O3-5, Sn–Al2O3-7 and Sn–Al2O3-10 were 0.89, 0.93 and 0.94, respectively. All are close to commercial solar absorbers (0.85–0.95) [1], and consistent with solar absorbers with related metal-Al2O3 film such as Co–Al2O3, W–Al2O3, Ag–Al2O3, Mo–Al2O3, Pt–Al2O3 and Ni–Al2O3 [8, 11,12,13, 18, 21]. In addition, αsol of the Sn–Al2O3 samples increased according to the increase of Sn content as a result of the solar radiation trapped inside the Al2O3 pores by the light resonant scattering with Sn particles [1, 21]. It corresponds to a previous study which found that the Al2O3 layer plays the role of antireflection and the amount of Sn pigment filling in the Al2O3 pores affect the αsol [16]. Furthermore, the measured spectral R in the wavelength range of the IR region is presented in Fig. 5b. The spectral R was above 0.95 for all the samples in the wavelengths of 2.5–8.0 μm, and there was no display of the spectral R beyond 17.5 μm due to extreme noise. It is observed that the spectral R has little effect on Sn content in the wavelength range of the IR region. The εtherm was calculated by Eq. 2, on the Planck blackbody distribution at 100 °C, as shown in Table 1. The εtherm was 0.23, 0.24 and 0.21 for Sn–Al2O3-5, Sn–Al2O3-7 and Sn–Al2O3-10, respectively. It seems that the εtherm is insignificantly changed with the increasing Sn content. When considering previous studies, the εtherm of this study was lower than the Pt–Al2O3 coating (εtherm = 0.36) [23] and close to the Ni–Al2O3 coatings (εtherm = 0.14–0.23) [21] and Mo–Al2O3 coating (εtherm = 0.19–0.27) [24]. As the results shown in Fig. 5, all the Sn–Al2O3 samples exhibit low R in the wavelength range of 300–2000 nm, and high R in the wavelength range of 2.5–17.5 μm, according to the theoretical property of solar absorber materials for achieving a high solar selectivity. Therefore, increasing Sn content in the pore of Al2O3 leads to a significant increase in the αsol, but has little effect on the εtherm.

Fig. 5
figure 5

Spectral reflectance (R) of Sn–Al2O3 solar absorber with different Sn contents in the wavelength range of a the UV–Vis-NIR regions compared with the solar spectrum at AM 1.5 and b the IR region

Full size image
Table 1 The solar absorptance (αsol) and thermal emittance (εtherm) of the Sn–Al2O3 samples with different Sn contents compared with the previous studies
Full size table

Figure 6 presents the temperature dependences of thermal conductivity for non-anodized Al and Sn–Al2O3 solar absorber with various Sn contents to propose the quantity of heat transmitted through a unit thickness of the samples at interval temperatures from 25 to 275 °C. The thermal conductivity of non-anodized Al was the highest related to the thermal conductivity of pure Al which was reported by National Standard Reference Data Series-National Bureau of Standards [25]. While, the thermal conductivity of the Sn–Al2O3 samples was lower than that of the non-anodized Al because Al2O3 film on the surface presents the same as an insulator and imperfections in the atomic structure as ceramic behavior. It is observed that the thermal conductivity of all the samples slightly decreased with increasing temperature, indicating that the Sn–Al2O3 coatings are able to be used in special thermal applications by maintaining the thermal conductivity of materials over the operational temperature range (50–275 °C). Furthermore, the thermal conductivity in each sample decreased with increasing Sn content as a result of the increasing thickness of Sn layer at the interface leading to obstruct free electrons and phonon contributions.

Fig. 6
figure 6

Temperature dependences of thermal conductivity of non-anodized Al and Sn–Al2O3 with different Sn contents

Full size image

Figure 7 represents a diagram of the heat flow in the sample for thermal conductivity measurement indicating the thermal direction from Al substrate to the Al2O3 layer. In the Sn–Al2O3 samples, thermal conduction in the Al substrate occurs from free electrons, where heat is transferred through kinetic energy in the elements but is easily scattered [26]. The heat transferred through the Al2O3 layer by lattice vibration (phonons), which is directly transmitted from high to low temperature regions. It is well known that free electrons are predominant in the thermal conductivity of metals and non-anodized Al, causing high thermal conductivity. However, the thermal conductivity of all Sn–Al2O3 samples was higher than commercial hard anodic coatings (0.7 W/m K) and also greater than previous studies with Al anodic coatings using several electrolytes (no more than 1.33 W/m K) [27]. The solar spectrum is converted to thermal energy at the Sn–Al2O3 solar absorber, then, useful heat is easily transferred through the Al substrate to the exchanger. It implies that the heat loss from the Sn–Al2O3 solar absorber to its surroundings is blocked and reduced by the Al2O3 layer, as illustrated in Fig. 8. Therefore, the low thermal conductivity of the Sn–Al2O3 solar absorber (Al2O3 layer) has advantages in solar thermal applications. However, the melting point of Sn is not too high (232 °C), compared to the melting point of other pigments (Ni, W, Mo, Ag, Co, Pt and Ti). Based on the operational temperature ranges, solar absorber materials for solar thermal applications can be classified as: (1) low temperature (T < 100 °C), (2) mid-temperature (100 °C < T < 400 °C) and (3) high temperature (T > 400 °C) [1]. In the previous studies, metal-Al2O3 solar absorbers are stable in a mid-temperature range between 250 and 350 °C of solar thermal applications [1, 28, 29]. Therefore, the Sn–Al2O3 solar absorber is considered to be thermally stable for solar collectors operated at low and mid-temperatures such as flat plate collectors and evacuated tube solar collectors, which are generally applied for water heating systems (T < 150 °C). In the present study, the thermal stability of the Sn–Al2O3 solar absorber is still unknown and further experiments need to be conducted, including a long-term (e.g. year) thermal stability test in a practical solar collector.

Fig. 7
figure 7

Heat flow in the thermal conductivity measurement on the Sn–Al2O3 specimen

Full size image
Fig. 8
figure 8

Solar radiation and heat transfer through the Sn–Al2O3 specimen

Full size image

Conclusions

The Sn-pigment on aluminium oxide (Sn–Al2O3) with selective absorber successfully prepared by anodization using dilute sulphuric acid followed by electrolysis in SnSO4 solution (coloration) for various three contents of Sn in the Al2O3 film. The Sn–Al2O3 samples were composed of Al and Sn phases while the Al2O3 phase was not detected due to the amorphous phase. The Sn rich was a spike shape expanding into the oxide film layer and had a high concentration at the interface of the Al substrate. The R of the Sn–Al2O3 solar absorbers decreased as the Sn contents increased over the whole solar spectrum region (300–2500 nm) and the Al2O3 layer plays the role of anti-reflection. The αsol increased according to the increase of Sn contents (0.89–0.94). While the εtherm changes were insignificant when compared to the increase of Sn content. The thermal conductivity of anodized Al samples was relatively lower than non-anodized Al, especially with Sn filling in the pores leading to the decrease in heat loss from the Al substrate in the surroundings. It can be concluded that the increased Sn contents in Sn–Al2O3 is able to enhance the solar selective properties and hence considered to be a candidate for good solar absorber materials for solar collectors operated at low and mid-temperature.