- Email: [email protected]
TiO2 nanocomposite films as solar selective absorber
Surface & Coatings Technology 302 (2016) 468–473
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Controllable fabrication and characterization of porous C/TiO2 nanocomposite films as solar selective absorber Zhizheng Wu a, Kangkai Wang a, Guangzhong Yuan a, Wen Li a, Chenlu Song a,b, Gaorong Han a,b, Yong Liu a,b,⁎ a b
State Key Laboratory of Silicon Materials and School of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, China Key Laboratory of Advanced Materials and Applications for Batteries of Zhejiang Province, China
a r t i c l e
i n f o
Article history: Received 25 July 2015 Revised 7 January 2016 Accepted in revised form 19 June 2016 Available online 21 June 2016 Keywords: Selective solar absorbers Porous C-TiO2 nanocomposite films PIPS
a b s t r a c t Porous C-TiO2 nanocomposite films prepared by photopolymerization-induced phase-separation method (PIPS) exhibited three dimensional interconnected pores structure. Scanning electron microscopy (SEM), atomic force microscope (AFM), thermo-gravimetric and differential thermal analysis (TG-DTA), high-resolution transmission electron microscopy (HRTEM), Raman spectroscopy, UV–Vis-NIR spectrophotometer are employed to characterize the structure and optical properties. The results illustrated that the photomonomer pentaerythritol triarylate (PETA) directly affects the relative molecular weight of polymer formed by photopolymerization to regulate the morphology and residual carbon of C-TiO2 nanocomposite films, and ultimately affects the solar selective absorption properties. The single layer of as-prepared porous C-TiO2 nanocomposite films exhibits high solar absorptance (α = 0.766–0.863) with low thermal emittance (ε = 0.06–0.12), yielding an optimized photothermal conversion efficiency η = α-ε of (0.706–0.762) corresponding to a wide redundancy PETA concentration 0.3– 0.9 (molar ratio to Ti precursor). The proportion of residual carbon contributed by PVP, PETA and poly-PETA, and other substances of the sol is 40.2%, 56.2% and 3.6%, respectively, which might be regulated by a suitable annealing process. Based on these results, porous C-TiO2 nanocomposite films are proved to be a novel candidate material for mid-temperature solar selective absorber coatings. © 2016 Elsevier B.V. All rights reserved.
1. Introduction With the aggravation of energy and environmental issues, the development of new energy technologies become a major breakthrough in solving the world's energy problems, and the preparation and study of solar selective absorption films of solar thermal technology has become the focus of attention. The search for spectrally selective solar absorbers conversion devices continue to be unabated that play the key role in solar photo-thermal [1–5]. The requirements of SSA are high absorption in the solar wavelength range and low thermal emittance in the far-infrared (FIR) wavelength ranges at the same time. These requirements translate to a material that has low reflectivity of less than 10% in the wavelength ranging from 0.3 to 2.5 μm and a high reflectivity of greater than 90% in the wavelength beyond 2.5 μm. As a matter of course, antireflection layers on the top of SSA are required to improve the absorbance as well [6,7]. More importantly, in order to realize industrialization, SSA should have the following essential properties: Firstly, the material has to be stable against the heat and water, and be exposed to ⁎ Corresponding author at: State Key Laboratory of Silicon Materials, School of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, China. E-mail address: [email protected] (Y. Liu).
http://dx.doi.org/10.1016/j.surfcoat.2016.06.050 0257-8972/© 2016 Elsevier B.V. All rights reserved.
the heat-transfer solvents and corrosive environments, especially for the middle and high temperature applications. Secondly, the components of these materials have to be nontoxic and sufficiently cheap, and must be easy to produce with low costs. So far, the well-known selective absorber surfaces are cermet materials [8–10], namely, the metal-dielectric composite coatings. Nickelpigmented Al2O3 (Ni-Al2O3) is a classic example for such a selective absorber system [11,12]. Other researchers designed a generation of solar selective absorbers which are SiO2, ZnO and NiO matrices dispersed with carbon nanoparticles [13,14]. Recently, we provided a newly design porous carbon-titania nanocomposite films with a well-defined interconnected macropores structure prepared via UV irradiation polymerization-induced phase-separation (PIPS) method [15,16]. The microstructure can be easily tuned and the exposure of the well-defined interconnected pores effectively enhances the scattering effect as a second phase resulting in an excellent solar selective absorption. The coatings also show a good stability at the high temperature (~ 500 °C vacuum). The pentaerythritol triacrylate (PETA), introduced as a photomonomer to form the porous structure and carbon source in the process, has not been discussed in detail yet. In the present work, we focus on PETA effects upon the microstructure and properties of nanocomposite films. Meanwhile, we also discuss
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473
469
Fig. 1. SEM images of the C/TiO2 films prepared with various PETA concentrations (molar ratio).
the contribution of the components in the precursor to the amount of residual carbon. Sols with various PETA concentrations were fabricated and the resultant as-prepared nanocomposite films were characterized and discussed in great detail.
2. Experimental details 2.1. Material All chemicals were obtained from Sinopharm Chemical Reagent Co. Ltd., except for polyvinylpyrrolidone k30 (PVP) and PETA purchased from Sigma-Aldrich Co. LLC, and used without further purification.
2.2. Nanocomposite preparation The porous C/TiO2 nanocomposite films were prepared on quartz and copper substrates (4 cm × 4 cm) by PIPS in the framework of sol-gel method for facilitating the destructive characterization and accomplishing performance characterization, respectively. The detail of the titania sol preparation have been described in Ref. [15–17]. Briefly, the mixtures with the C16H36O4Ti (TTB):C2H5OH:H2O:HNO3:DMF:PETA composition of 1:30:3:0.3:4:x (molar ratio) were used as the precursor solution, where x was varied from 0.3 to 1.2 to investigate the effects of PETA used as the photomonomer on the microstructure and optical properties. 1.5 g of PVP and the amount of 2, 2′-Azo bisisobatyronitrile (AIBN) used as radical initiator varied from 0.018 g to 0.073 g which keep PETA:AIBN =
Fig. 2. Thickness of the C/TiO2 films prepared with various PETA concentrations (molar ratio) measured by SEM.
470
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473
0.01 mol:0.035 g, were dissolved in the solution. After 30 min vigorous stirring under the ice-cooled condition, the sol was used to prepared films by spin coating method with a spin speed of 2000 rpm for 20 s on the cleaned substrates. Then the films were irradiated under UV light (λ = 365 nm, 7 mW/cm2) for 10 min. Afterwards, the irradiated films were annealed into a tube furnace with an inert atmosphere. These films were pre-heated at 200 °C to evaporate the solvent and then heated at 600 °C for 60 min (with 10 °C/min heating rate) to process the carbonization. 2.3. Characterization A scanning electron microscopy (SEM, Hitachi) was carried out to analyze the surface morphology and thickness of the prepared films with an accelerating voltage of 15 KV. The three dimensional surface topography was also probed using an atomic force microscope XE-100E from PSIA (AFM) operated in the tapping mode with a measuring area of 20 × 20 μm2. The surface roughness of the coatings was measured by Surface Imaging Systems equipped on XE-100E. The high-resolution transmission electron microscopy (HRTEM) was carried out by a Tecnai F20 from FEI using an acceleration voltage of 200 kV. A thermal analyzer (Thermo plus T8210, Rigaku) was used for the thermos gravimetry-differential thermal analysis (TG–DTA) with a heating rate of 5 °C/min from room temperature to 800 °C. The irradiated gel under the UV light for 10 min and dried at 70 °C for 48 h was prepared for the TGDTA analysis. Raman spectra were collected on a Renishaw InVia Raman microscope. The 532 nm line of a laser was used as excitation with the maximum laser power available of 50 mW. A spectrophotometer equipped with an integrating sphere and BaSO4 as a reference (UV3600, Shimadzu) was used to measure the reflectance in the range of 0.3–2.5 μm wavelength. And a Fourier Transform infrared spectroscopy (FTIR) with a gold mirror reference (Tensor 27, Bruker) was used to measure the reflectance in the range of 2.5–25 μm wavelength. The calculated hemispherical absorptance (α) and emissivity (ε) are respectively yielded by Eqs. (1) and (2) [13, 14],
Fig. 3. Three-dimensional AFM image of the sample prepared with the PETA concentration (molar ratio) x = 0.9.
like morphology (Fig. 1, x = 0.12). The cross-sectional SEM images (Fig. 2) show the film thickness increases from 281 nm to 659 nm with PETA concentration, because poly-PETA partially works as a carbon precursor [16]. The cross-sectional images are also indication of high PETA induces incompact structure, in agreement of the surface
2:5μm
α¼
∫ 0:3μm Ιsun ðλÞð1−RðλÞÞdλ 2:5μm
∫ 0:3μm Ιsun ðλÞdλ
ð1Þ
2:5μm
ε¼
∫ 2:5μm Ι B ðλÞð1−RðλÞÞdλ 2:5μm
∫ 2:5μm ΙB ðλÞdλ
ð2Þ
where R (λ), Isun(λ), and IB(λ) are the sample's spectral reflectance, the spectral power density of the solar radiation air mass (AM) 1.5, and the spectral radiance of a black body at 100 °C, respectively. In the present work, the photothermal conversion efficiency η is defined as: η ¼ α−ε
ð3Þ
3. Results and discussion The representative surface morphologies of the porous C/TiO2 nanocomposite films prepared with various PETA concentrations (x = 0.3– 1.2) have been characterized by SEM as shown in Fig. 1. All samples exhibit an interconnected macropores structure except that the sample prepared with x = 1.2 shows a collapsed morphology. The average diameter of the pores increases from around 50 nm to around 300 and 500 nm with the increase of the PETA concentration from 0.3 to 0.9. This can be explained that increase of PETA results in a larger average molecular weight of poly-PETA after UV-induced polymerizations, leading to a bigger pore size after the decomposition of poly-PETA. When PETA concentration exceeds the critical value, the average pore size is so large that the adjacent pores start to merge, resulting in a collapse-
Fig. 4. The reflectivity spectra of C/TiO2 films deposited on copper substrates with various PETA concentrations (x) in UV/Vis/Near-IR range (A) and IR range (B).
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473 Table 1 Photothermal conversion efficiency(η), solar absorption(α), thermal emittance(ε), and film thickness(h) of C/TiO2 nanocomposties prepared with various PETA concentrations. PETA (x)
h(nm)
α
ε
η
0.3 0.6 0.9 1.2
281 408 475 659
0.766 0.842 0.863 0.819
0.06 0.08 0.12 0.17
0.706 0.762 0.743 0.649
morphology. Open pores construct a rough surface, exhibiting a relatively high RMS roughness of about 54.78 nm (the sample with PETA x = 0.9) measured by AFM (Fig. 3), could lead to an enhancement of solar absorptance by multiple reflections. The measured UV/Vis/Near-IR and FTIR reflectance of the porous C/ TiO2 films deposited on copper substrate with various PETA concentrations are presented in Fig. 4. It is evident that the reflective spectra of films ranging from 0.3 μm to 1.5 μm is relatively low (around 20%) and the reflectance ranging from 5 μm to 25 μm is relatively high (around 80%), which demonstrate the typical selective absorption properties. The broad absorption band at about 2.9 μm (or 3450 cm−1) is assigned to Ti-OH stretching vibrations, and the other major band at about 6.13 μm (or 1630 cm−1) is possibly attributed to the stretching vibrations of sp2 coordinated C_C from the residual carbon in the composite films. According to Eqs.(1) and (2), the calculated α and ε of porous C/TiO2 nanocomposite films are listed in Table 1. The thermal emittance increases with PETA concentration, attributed to the film thickness effect. The solar absorption also increases with PETA concentration from x = 0.3 to x = 0.9, but after that a drop presents with PETA concentration x = 1.2 corresponding to the collapse-like structure. Therefore, one can see that the porous structure rather the thickness of the composite film dominates the solar absorption, consistent with the previous results [15]. Nevertheless, the thicker film does worsen the thermal emittance. On the basis of the competition relationship between the solar absorptance and thermal emittance, the samples exhibit the excellent photothermal conversion efficiency corresponding to a wide redundance of PETA concentration. Focusing on the microstructure of porous C/TiO2 nanocomposites, Raman spectroscopy has been used to characterize residual carbon containing in the films as shown in Fig. 5. Tow typical Raman shift peaks could be generally found from the Raman spectrum of nanocomposite materials containing carbon particles: a G peak at around 1580 cm−1 corresponding to the in-plane bond-stretching motion of pairs of C sp2 atoms with E2g symmetry at the zone center and a D peak at around 1355 cm−1 corresponding to the breathing motion of six fold rings
471
Table 2 Raman shifts (W), full width at half maximum (FWHM), peak intensity ratio (ID/IG) and dimension of the nanocrystalline (L) for the C/TiO2 nanocomposite prepared with various PETA concentrations (molar ratio). PETA (x)
WD(cm−1) WG(cm−1) DFWHM(cm−1) GFWHM(cm−1) ID/IG
L (nm)
0.3 0.6 0.9 1.2
1364.9 1373.5 1374.2 1380.0
1.44 1.20 1.19 1.18
1590.9 1591.3 1591.6 1605.5
245.95 243.59 243.27 252.93
102.72 100.81 105.00 97.35
0.790 0.787 0.785 0.770
with A1g symmetry near the zone boundary. The Raman spectra of the samples were fitted with a Breit-Wigner-Fano (BWF) line for the G peak and a Lorentzian for the D peak, and the resultant parameters of the fitting are summarized in Table 2. The mean sizes of graphitic nanocrystallites are also listed in Table 2 yielded by the following empirically calibrated equation [18]: IðDÞ ¼ kL IðGÞ
ð4Þ
where k is a constant depending on the incident laser wavelength in the Raman measurement and set to a value of 0.55 nm−2 in the present work. I(D) and I(G) are respectively the intensity of the D and G peak, and L is the dimension of cluster carbon nanocrystalline. From Table 2, the dimension of the nanocrystalline is much small with a value of around 1.18 nm–1.44 nm and decreases with increasing the PETA concentration. Since the samller graphite cluster will have higher modes corresponding to the D peak, the position of the D peak presents a blue shift with the increasing PETA concentration. In contrast, the G peak slightly shifts to a higher frequency with increasing the PETA concentration, suggesting that PETA might enhance the inner stress of the sp2 bonds to strengthen the phonon mode related to the G peak. The full width at half maximum(FWHM) of D peak increases with PETA concentration, indicating that the disorder degree of the residual carbon is reinforced. Considering the trends that nano-carbon particle dimension shrinking and the D peak FWHM narrowing with increasing PETA concentration, it might indicate that PETA disadvantages the order degree of the residual carbon in the nanocomposite films even though PETA worked as the carbon precursor. To further investigate the contributions of the precursor to the residual carbon in the composite films, three control groups (CGs) as shown in Table 3 were designed to determine the source of the residual carbon by TG experiments. The xerogel for TG measurements was made from the gel after irradiated by the UV light for 10 min and dried at 70 °C for 48 h. TG curves heated in air and N2 are shown in Fig. 6. For all samples, TG curves treated in air and nitrogen are stable above 550 °C, because the combustion (in air) and carbonization (in N2) processes almost accomplish very well at the temperature, indicating the annealing temperature in the present work set at 600 °C is reasonable. Sample CG1 mainly consists of TBT sol without PETA and PVP. TBT hydrolysis can produce small molecule butyl which is mainly volatilized as well as the solvent (EtOH and H2O) during the heat treatment process under 300 °C. Therefore, no carbonization reaction occurs in nitrogen, leading that TiO2 formation process is basically the same in both air and nitrogen. Considering the fact that no carbon left after the Table 3 Control Groups (CGs) for TG experiments. Component TBT
Fig. 5. Raman spectra of C/TiO2 nanocomposite thin films prepared with various PETA concentrations.
EtOH
H2O
HNO3
DMF
AIBN
PVP PETA (molar wt% ratio)
CGs
(molar ratio)
(molar ratio)
(molar ratio)
(molar ratio)
(molar ratio)
(g)
(1) (2) (3)
1 1 1
30 30 30
3 3 3
0.3 0.3 0.3
4 4 4
0.035 0 0.035 4% 0.035 4%
0 0 0.6
472
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473
Fig. 6. TG curves of xerogel from the control groups annealed in air and N2. A: CG (1); B: CG (2); C: CG (3).
combustion in air and ignoring the nonstoichiometric ratio of TiO2 in N2, the residual carbon is easy to yield from the TG curves by comparing the weight loss in air and N2. Therefore, for CG (1) the calculated weight ratio between the residual carbon and TiO2 is about 0.045:1, which is mainly contributed by the hydrolysate of TBT. For CG (2), the value is about 0.544:1, where the residual carbon mainly comes from the carbonization of PVP. And for CG (3), the value is about 1.240:1, where the residual carbon is mainly contributed by PVP, PETA and its photopolymerization production (poly-PETA). According to the proportion of the residual carbon and the TiO2, the contribution of the components in the precursors to the residual carbon can be yielded as: X:Y:Z = 40.2%:56.2%:3.6%, where X, Y, and Z are the contributions by PVP, PETA and poly-PETA, and hydrolysate of TBT, respectively. Based on the discussion above, PVP and PETA worked as the main carbon source of the composite film, and its combustion exothermic peaks located at 506.24 °C and 426.26 °C, respectively. Therefore, a suitable heat-treatment condition by adjusting the heating process might provide a practicable way to control residual carbon content in the composite film to optimize select absorption performance of the nanocomposite films. 4. Conclusions A newly proposed solar selective absorbers with composition of porous C/TiO2 nanocomposite films were successfully deposited on copper substrates by photopolymerization-induced phase-separation method. The coatings exhibited high solar absorptance (α = 0.766–0.863) and low thermal emittance (ε = 0.06–0.12) in a wide redundancy PETA concentration x = 0.3–0.9 (molar ratio to Ti precursor) for single layer. The SEM results show that all samples exhibit a fine interconnected macropores structure except that the sample prepared with PETA concentrations (x = 1.2) is collapsed. The average diameter of the pores increases from around 50 nm to 500 nm as well as the thickness increased from 281 nm to 659 nm with the increase of the PETA concentration from 0.3 to 0.9, and presents a collapse-like morphology with a continue increase of the PETA concentrations. The micro-Raman spectroscopy
data suggest a carbon cluster size of around 1.18–1.44 nm and an decrease in order degree of carbon with the increase of the PETA concentration. The results show that the main factor influencing the solar absorption is the three dimensional pore structure in the sample rather than the film thickness. Simultaneously, by the TG results, the proportion of residual carbon contributed by PVP, PETA and poly-PETA, other substances of the sol is calculated to 40.2%, 56.2% and 3.6% respectively. A suitable heat-treatment condition will be introduced to control the carbon content in the composite film and optimized the select absorption performance of the nanocomposite films which is under working. Acknowledgments The work was financially supported by National Natural Science Foundation of China (No. 51572236) and Zhejiang Provincial Natural Science Foundation of China (No. LY16E020002). References [1] J.A. Duddie, W.A. Beckman, Solar Engineering of Thermal Processes, second ed. Wiley, New York, 1991. [2] P. Konttinen, P.D. Lund, R.J. Kilpi, Mechanically manufactured selective solar absorber surfaces, Sol. Energy Mater. Sol. Cells 79 (2003) 273. [3] Z.C. Orel, Characterisation of high-temperature-resistant spectrally selective paints for solar absorbers, Sol. Energy Mater. Sol. Cells 57 (1999) 291. [4] S.S. Kanu, R. Binions, Thin films for solar control applications, Proc. R. Soc. A Math. Phys. 466 (2010) 19–44. [5] A. Avila-Garcia, U. Morales-Ortiz, Thermally and air–plasma-oxidized titanium and stainless steel plates as solar selective absorbers, Sol. Energy Mater. Sol. Cells 90 (2006) 2556. [6] Y. Matsai, S. Polarz, M. Antonietti, Silica-carbon nanocomposites-a new concept for the Design of Solar Absorbers, Adv. Funct. Mater. 12 (2002) 197–202. [7] D. Katzen, Y. Mastai, Thin films of silica-carbon nanocomposites for selective solar absorbers, Appl. Surf. Sci. 248 (2005) 514–517. [8] Q.C. Zhang, Y.G. Shen, High performance W-AlN cermet solar coatings designed by modelling calculations and deposited by DC magnetron sputtering, Sol. Energy Mater. Sol. Cells 81 (2004) 25–37. [9] G.B. Smith, A.V. Radchik, Columnar cermet structures in solar energy materials: can one model spectral response with simple effective medium theories, Sol. Energy Mater. Sol. Cells 54 (1998) 387–396.
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473 [10] Q.C. Zhang, Y.B. Yin, High efficiency Mo-Al2O3 cermet selective surfaces for hightemperature application, Sol. Energy Mater. Sol. Cells 40 (1996) 43–53. [11] T.S. Sathiaraj, R. Thangaraj, Optical properties of selectively absorbing r.f. sputtered Ni-Al2O3 composite films, Thin Solid Films 195 (1991) 33–42. [12] T. Tesfamichael, A. Roos, Treatment of antireflection on tin oxide coated anodized aluminum selective absorber surface, Sol. Energy Mater. Sol. Cells 54 (1998) 213–221. [13] G. Katumba, A. Forbes, Quantifying the efficacy of solar selective absorber materials: the case of carbon nanoparticles dispersed in SiO2, ZnO and NiO matrices, Proc. SPIE 7046 (2008) (70460B-1). [14] G. Katumba, L. Olumekor, Optical, thermal and structural characteristics of carbon nanoparticles embedded in ZnO and NiO as selective solar absorbers, Sol. Energy Mater. Sol. Cells 92 (2008) 1285–1292.
473
[15] B. Cheng, K.K. Wang, G.R. Han, Y. Liu, Porous carbon–titania nanocomposite films for spectrally solar selective absorbers, Sol. Energy Mater. Sol. Cells 133 (2015) 126–132. [16] B. Cheng, K.K. Wang, G.R. Han, Y. Liu, Preparation and characterization of porous carbon–titania nanocomposite films as solar selective absorbers, J. Alloys Compd. 635 (2015) 129–135. [17] J.X. Yao, M. Takahashi, T. Yoko, Controlled preparation of macroporous TiO2 films by photo polymerization-induced phase separation method and their photocatalytic performance, Thin Solid Films 517 (2009) 6479–6485. [18] A.C. Ferrari, J. Robertson, Interpretation of Raman spectra of disordered and amorphous carbon, Phys. Rev. B 61 (2000) 14095–14107.
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Controllable fabrication and characterization of porous C/TiO2 nanocomposite films as solar selective absorber Zhizheng Wu a, Kangkai Wang a, Guangzhong Yuan a, Wen Li a, Chenlu Song a,b, Gaorong Han a,b, Yong Liu a,b,⁎ a b
State Key Laboratory of Silicon Materials and School of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, China Key Laboratory of Advanced Materials and Applications for Batteries of Zhejiang Province, China
a r t i c l e
i n f o
Article history: Received 25 July 2015 Revised 7 January 2016 Accepted in revised form 19 June 2016 Available online 21 June 2016 Keywords: Selective solar absorbers Porous C-TiO2 nanocomposite films PIPS
a b s t r a c t Porous C-TiO2 nanocomposite films prepared by photopolymerization-induced phase-separation method (PIPS) exhibited three dimensional interconnected pores structure. Scanning electron microscopy (SEM), atomic force microscope (AFM), thermo-gravimetric and differential thermal analysis (TG-DTA), high-resolution transmission electron microscopy (HRTEM), Raman spectroscopy, UV–Vis-NIR spectrophotometer are employed to characterize the structure and optical properties. The results illustrated that the photomonomer pentaerythritol triarylate (PETA) directly affects the relative molecular weight of polymer formed by photopolymerization to regulate the morphology and residual carbon of C-TiO2 nanocomposite films, and ultimately affects the solar selective absorption properties. The single layer of as-prepared porous C-TiO2 nanocomposite films exhibits high solar absorptance (α = 0.766–0.863) with low thermal emittance (ε = 0.06–0.12), yielding an optimized photothermal conversion efficiency η = α-ε of (0.706–0.762) corresponding to a wide redundancy PETA concentration 0.3– 0.9 (molar ratio to Ti precursor). The proportion of residual carbon contributed by PVP, PETA and poly-PETA, and other substances of the sol is 40.2%, 56.2% and 3.6%, respectively, which might be regulated by a suitable annealing process. Based on these results, porous C-TiO2 nanocomposite films are proved to be a novel candidate material for mid-temperature solar selective absorber coatings. © 2016 Elsevier B.V. All rights reserved.
1. Introduction With the aggravation of energy and environmental issues, the development of new energy technologies become a major breakthrough in solving the world's energy problems, and the preparation and study of solar selective absorption films of solar thermal technology has become the focus of attention. The search for spectrally selective solar absorbers conversion devices continue to be unabated that play the key role in solar photo-thermal [1–5]. The requirements of SSA are high absorption in the solar wavelength range and low thermal emittance in the far-infrared (FIR) wavelength ranges at the same time. These requirements translate to a material that has low reflectivity of less than 10% in the wavelength ranging from 0.3 to 2.5 μm and a high reflectivity of greater than 90% in the wavelength beyond 2.5 μm. As a matter of course, antireflection layers on the top of SSA are required to improve the absorbance as well [6,7]. More importantly, in order to realize industrialization, SSA should have the following essential properties: Firstly, the material has to be stable against the heat and water, and be exposed to ⁎ Corresponding author at: State Key Laboratory of Silicon Materials, School of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, China. E-mail address: [email protected] (Y. Liu).
http://dx.doi.org/10.1016/j.surfcoat.2016.06.050 0257-8972/© 2016 Elsevier B.V. All rights reserved.
the heat-transfer solvents and corrosive environments, especially for the middle and high temperature applications. Secondly, the components of these materials have to be nontoxic and sufficiently cheap, and must be easy to produce with low costs. So far, the well-known selective absorber surfaces are cermet materials [8–10], namely, the metal-dielectric composite coatings. Nickelpigmented Al2O3 (Ni-Al2O3) is a classic example for such a selective absorber system [11,12]. Other researchers designed a generation of solar selective absorbers which are SiO2, ZnO and NiO matrices dispersed with carbon nanoparticles [13,14]. Recently, we provided a newly design porous carbon-titania nanocomposite films with a well-defined interconnected macropores structure prepared via UV irradiation polymerization-induced phase-separation (PIPS) method [15,16]. The microstructure can be easily tuned and the exposure of the well-defined interconnected pores effectively enhances the scattering effect as a second phase resulting in an excellent solar selective absorption. The coatings also show a good stability at the high temperature (~ 500 °C vacuum). The pentaerythritol triacrylate (PETA), introduced as a photomonomer to form the porous structure and carbon source in the process, has not been discussed in detail yet. In the present work, we focus on PETA effects upon the microstructure and properties of nanocomposite films. Meanwhile, we also discuss
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473
469
Fig. 1. SEM images of the C/TiO2 films prepared with various PETA concentrations (molar ratio).
the contribution of the components in the precursor to the amount of residual carbon. Sols with various PETA concentrations were fabricated and the resultant as-prepared nanocomposite films were characterized and discussed in great detail.
2. Experimental details 2.1. Material All chemicals were obtained from Sinopharm Chemical Reagent Co. Ltd., except for polyvinylpyrrolidone k30 (PVP) and PETA purchased from Sigma-Aldrich Co. LLC, and used without further purification.
2.2. Nanocomposite preparation The porous C/TiO2 nanocomposite films were prepared on quartz and copper substrates (4 cm × 4 cm) by PIPS in the framework of sol-gel method for facilitating the destructive characterization and accomplishing performance characterization, respectively. The detail of the titania sol preparation have been described in Ref. [15–17]. Briefly, the mixtures with the C16H36O4Ti (TTB):C2H5OH:H2O:HNO3:DMF:PETA composition of 1:30:3:0.3:4:x (molar ratio) were used as the precursor solution, where x was varied from 0.3 to 1.2 to investigate the effects of PETA used as the photomonomer on the microstructure and optical properties. 1.5 g of PVP and the amount of 2, 2′-Azo bisisobatyronitrile (AIBN) used as radical initiator varied from 0.018 g to 0.073 g which keep PETA:AIBN =
Fig. 2. Thickness of the C/TiO2 films prepared with various PETA concentrations (molar ratio) measured by SEM.
470
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473
0.01 mol:0.035 g, were dissolved in the solution. After 30 min vigorous stirring under the ice-cooled condition, the sol was used to prepared films by spin coating method with a spin speed of 2000 rpm for 20 s on the cleaned substrates. Then the films were irradiated under UV light (λ = 365 nm, 7 mW/cm2) for 10 min. Afterwards, the irradiated films were annealed into a tube furnace with an inert atmosphere. These films were pre-heated at 200 °C to evaporate the solvent and then heated at 600 °C for 60 min (with 10 °C/min heating rate) to process the carbonization. 2.3. Characterization A scanning electron microscopy (SEM, Hitachi) was carried out to analyze the surface morphology and thickness of the prepared films with an accelerating voltage of 15 KV. The three dimensional surface topography was also probed using an atomic force microscope XE-100E from PSIA (AFM) operated in the tapping mode with a measuring area of 20 × 20 μm2. The surface roughness of the coatings was measured by Surface Imaging Systems equipped on XE-100E. The high-resolution transmission electron microscopy (HRTEM) was carried out by a Tecnai F20 from FEI using an acceleration voltage of 200 kV. A thermal analyzer (Thermo plus T8210, Rigaku) was used for the thermos gravimetry-differential thermal analysis (TG–DTA) with a heating rate of 5 °C/min from room temperature to 800 °C. The irradiated gel under the UV light for 10 min and dried at 70 °C for 48 h was prepared for the TGDTA analysis. Raman spectra were collected on a Renishaw InVia Raman microscope. The 532 nm line of a laser was used as excitation with the maximum laser power available of 50 mW. A spectrophotometer equipped with an integrating sphere and BaSO4 as a reference (UV3600, Shimadzu) was used to measure the reflectance in the range of 0.3–2.5 μm wavelength. And a Fourier Transform infrared spectroscopy (FTIR) with a gold mirror reference (Tensor 27, Bruker) was used to measure the reflectance in the range of 2.5–25 μm wavelength. The calculated hemispherical absorptance (α) and emissivity (ε) are respectively yielded by Eqs. (1) and (2) [13, 14],
Fig. 3. Three-dimensional AFM image of the sample prepared with the PETA concentration (molar ratio) x = 0.9.
like morphology (Fig. 1, x = 0.12). The cross-sectional SEM images (Fig. 2) show the film thickness increases from 281 nm to 659 nm with PETA concentration, because poly-PETA partially works as a carbon precursor [16]. The cross-sectional images are also indication of high PETA induces incompact structure, in agreement of the surface
2:5μm
α¼
∫ 0:3μm Ιsun ðλÞð1−RðλÞÞdλ 2:5μm
∫ 0:3μm Ιsun ðλÞdλ
ð1Þ
2:5μm
ε¼
∫ 2:5μm Ι B ðλÞð1−RðλÞÞdλ 2:5μm
∫ 2:5μm ΙB ðλÞdλ
ð2Þ
where R (λ), Isun(λ), and IB(λ) are the sample's spectral reflectance, the spectral power density of the solar radiation air mass (AM) 1.5, and the spectral radiance of a black body at 100 °C, respectively. In the present work, the photothermal conversion efficiency η is defined as: η ¼ α−ε
ð3Þ
3. Results and discussion The representative surface morphologies of the porous C/TiO2 nanocomposite films prepared with various PETA concentrations (x = 0.3– 1.2) have been characterized by SEM as shown in Fig. 1. All samples exhibit an interconnected macropores structure except that the sample prepared with x = 1.2 shows a collapsed morphology. The average diameter of the pores increases from around 50 nm to around 300 and 500 nm with the increase of the PETA concentration from 0.3 to 0.9. This can be explained that increase of PETA results in a larger average molecular weight of poly-PETA after UV-induced polymerizations, leading to a bigger pore size after the decomposition of poly-PETA. When PETA concentration exceeds the critical value, the average pore size is so large that the adjacent pores start to merge, resulting in a collapse-
Fig. 4. The reflectivity spectra of C/TiO2 films deposited on copper substrates with various PETA concentrations (x) in UV/Vis/Near-IR range (A) and IR range (B).
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473 Table 1 Photothermal conversion efficiency(η), solar absorption(α), thermal emittance(ε), and film thickness(h) of C/TiO2 nanocomposties prepared with various PETA concentrations. PETA (x)
h(nm)
α
ε
η
0.3 0.6 0.9 1.2
281 408 475 659
0.766 0.842 0.863 0.819
0.06 0.08 0.12 0.17
0.706 0.762 0.743 0.649
morphology. Open pores construct a rough surface, exhibiting a relatively high RMS roughness of about 54.78 nm (the sample with PETA x = 0.9) measured by AFM (Fig. 3), could lead to an enhancement of solar absorptance by multiple reflections. The measured UV/Vis/Near-IR and FTIR reflectance of the porous C/ TiO2 films deposited on copper substrate with various PETA concentrations are presented in Fig. 4. It is evident that the reflective spectra of films ranging from 0.3 μm to 1.5 μm is relatively low (around 20%) and the reflectance ranging from 5 μm to 25 μm is relatively high (around 80%), which demonstrate the typical selective absorption properties. The broad absorption band at about 2.9 μm (or 3450 cm−1) is assigned to Ti-OH stretching vibrations, and the other major band at about 6.13 μm (or 1630 cm−1) is possibly attributed to the stretching vibrations of sp2 coordinated C_C from the residual carbon in the composite films. According to Eqs.(1) and (2), the calculated α and ε of porous C/TiO2 nanocomposite films are listed in Table 1. The thermal emittance increases with PETA concentration, attributed to the film thickness effect. The solar absorption also increases with PETA concentration from x = 0.3 to x = 0.9, but after that a drop presents with PETA concentration x = 1.2 corresponding to the collapse-like structure. Therefore, one can see that the porous structure rather the thickness of the composite film dominates the solar absorption, consistent with the previous results [15]. Nevertheless, the thicker film does worsen the thermal emittance. On the basis of the competition relationship between the solar absorptance and thermal emittance, the samples exhibit the excellent photothermal conversion efficiency corresponding to a wide redundance of PETA concentration. Focusing on the microstructure of porous C/TiO2 nanocomposites, Raman spectroscopy has been used to characterize residual carbon containing in the films as shown in Fig. 5. Tow typical Raman shift peaks could be generally found from the Raman spectrum of nanocomposite materials containing carbon particles: a G peak at around 1580 cm−1 corresponding to the in-plane bond-stretching motion of pairs of C sp2 atoms with E2g symmetry at the zone center and a D peak at around 1355 cm−1 corresponding to the breathing motion of six fold rings
471
Table 2 Raman shifts (W), full width at half maximum (FWHM), peak intensity ratio (ID/IG) and dimension of the nanocrystalline (L) for the C/TiO2 nanocomposite prepared with various PETA concentrations (molar ratio). PETA (x)
WD(cm−1) WG(cm−1) DFWHM(cm−1) GFWHM(cm−1) ID/IG
L (nm)
0.3 0.6 0.9 1.2
1364.9 1373.5 1374.2 1380.0
1.44 1.20 1.19 1.18
1590.9 1591.3 1591.6 1605.5
245.95 243.59 243.27 252.93
102.72 100.81 105.00 97.35
0.790 0.787 0.785 0.770
with A1g symmetry near the zone boundary. The Raman spectra of the samples were fitted with a Breit-Wigner-Fano (BWF) line for the G peak and a Lorentzian for the D peak, and the resultant parameters of the fitting are summarized in Table 2. The mean sizes of graphitic nanocrystallites are also listed in Table 2 yielded by the following empirically calibrated equation [18]: IðDÞ ¼ kL IðGÞ
ð4Þ
where k is a constant depending on the incident laser wavelength in the Raman measurement and set to a value of 0.55 nm−2 in the present work. I(D) and I(G) are respectively the intensity of the D and G peak, and L is the dimension of cluster carbon nanocrystalline. From Table 2, the dimension of the nanocrystalline is much small with a value of around 1.18 nm–1.44 nm and decreases with increasing the PETA concentration. Since the samller graphite cluster will have higher modes corresponding to the D peak, the position of the D peak presents a blue shift with the increasing PETA concentration. In contrast, the G peak slightly shifts to a higher frequency with increasing the PETA concentration, suggesting that PETA might enhance the inner stress of the sp2 bonds to strengthen the phonon mode related to the G peak. The full width at half maximum(FWHM) of D peak increases with PETA concentration, indicating that the disorder degree of the residual carbon is reinforced. Considering the trends that nano-carbon particle dimension shrinking and the D peak FWHM narrowing with increasing PETA concentration, it might indicate that PETA disadvantages the order degree of the residual carbon in the nanocomposite films even though PETA worked as the carbon precursor. To further investigate the contributions of the precursor to the residual carbon in the composite films, three control groups (CGs) as shown in Table 3 were designed to determine the source of the residual carbon by TG experiments. The xerogel for TG measurements was made from the gel after irradiated by the UV light for 10 min and dried at 70 °C for 48 h. TG curves heated in air and N2 are shown in Fig. 6. For all samples, TG curves treated in air and nitrogen are stable above 550 °C, because the combustion (in air) and carbonization (in N2) processes almost accomplish very well at the temperature, indicating the annealing temperature in the present work set at 600 °C is reasonable. Sample CG1 mainly consists of TBT sol without PETA and PVP. TBT hydrolysis can produce small molecule butyl which is mainly volatilized as well as the solvent (EtOH and H2O) during the heat treatment process under 300 °C. Therefore, no carbonization reaction occurs in nitrogen, leading that TiO2 formation process is basically the same in both air and nitrogen. Considering the fact that no carbon left after the Table 3 Control Groups (CGs) for TG experiments. Component TBT
Fig. 5. Raman spectra of C/TiO2 nanocomposite thin films prepared with various PETA concentrations.
EtOH
H2O
HNO3
DMF
AIBN
PVP PETA (molar wt% ratio)
CGs
(molar ratio)
(molar ratio)
(molar ratio)
(molar ratio)
(molar ratio)
(g)
(1) (2) (3)
1 1 1
30 30 30
3 3 3
0.3 0.3 0.3
4 4 4
0.035 0 0.035 4% 0.035 4%
0 0 0.6
472
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473
Fig. 6. TG curves of xerogel from the control groups annealed in air and N2. A: CG (1); B: CG (2); C: CG (3).
combustion in air and ignoring the nonstoichiometric ratio of TiO2 in N2, the residual carbon is easy to yield from the TG curves by comparing the weight loss in air and N2. Therefore, for CG (1) the calculated weight ratio between the residual carbon and TiO2 is about 0.045:1, which is mainly contributed by the hydrolysate of TBT. For CG (2), the value is about 0.544:1, where the residual carbon mainly comes from the carbonization of PVP. And for CG (3), the value is about 1.240:1, where the residual carbon is mainly contributed by PVP, PETA and its photopolymerization production (poly-PETA). According to the proportion of the residual carbon and the TiO2, the contribution of the components in the precursors to the residual carbon can be yielded as: X:Y:Z = 40.2%:56.2%:3.6%, where X, Y, and Z are the contributions by PVP, PETA and poly-PETA, and hydrolysate of TBT, respectively. Based on the discussion above, PVP and PETA worked as the main carbon source of the composite film, and its combustion exothermic peaks located at 506.24 °C and 426.26 °C, respectively. Therefore, a suitable heat-treatment condition by adjusting the heating process might provide a practicable way to control residual carbon content in the composite film to optimize select absorption performance of the nanocomposite films. 4. Conclusions A newly proposed solar selective absorbers with composition of porous C/TiO2 nanocomposite films were successfully deposited on copper substrates by photopolymerization-induced phase-separation method. The coatings exhibited high solar absorptance (α = 0.766–0.863) and low thermal emittance (ε = 0.06–0.12) in a wide redundancy PETA concentration x = 0.3–0.9 (molar ratio to Ti precursor) for single layer. The SEM results show that all samples exhibit a fine interconnected macropores structure except that the sample prepared with PETA concentrations (x = 1.2) is collapsed. The average diameter of the pores increases from around 50 nm to 500 nm as well as the thickness increased from 281 nm to 659 nm with the increase of the PETA concentration from 0.3 to 0.9, and presents a collapse-like morphology with a continue increase of the PETA concentrations. The micro-Raman spectroscopy
data suggest a carbon cluster size of around 1.18–1.44 nm and an decrease in order degree of carbon with the increase of the PETA concentration. The results show that the main factor influencing the solar absorption is the three dimensional pore structure in the sample rather than the film thickness. Simultaneously, by the TG results, the proportion of residual carbon contributed by PVP, PETA and poly-PETA, other substances of the sol is calculated to 40.2%, 56.2% and 3.6% respectively. A suitable heat-treatment condition will be introduced to control the carbon content in the composite film and optimized the select absorption performance of the nanocomposite films which is under working. Acknowledgments The work was financially supported by National Natural Science Foundation of China (No. 51572236) and Zhejiang Provincial Natural Science Foundation of China (No. LY16E020002). References [1] J.A. Duddie, W.A. Beckman, Solar Engineering of Thermal Processes, second ed. Wiley, New York, 1991. [2] P. Konttinen, P.D. Lund, R.J. Kilpi, Mechanically manufactured selective solar absorber surfaces, Sol. Energy Mater. Sol. Cells 79 (2003) 273. [3] Z.C. Orel, Characterisation of high-temperature-resistant spectrally selective paints for solar absorbers, Sol. Energy Mater. Sol. Cells 57 (1999) 291. [4] S.S. Kanu, R. Binions, Thin films for solar control applications, Proc. R. Soc. A Math. Phys. 466 (2010) 19–44. [5] A. Avila-Garcia, U. Morales-Ortiz, Thermally and air–plasma-oxidized titanium and stainless steel plates as solar selective absorbers, Sol. Energy Mater. Sol. Cells 90 (2006) 2556. [6] Y. Matsai, S. Polarz, M. Antonietti, Silica-carbon nanocomposites-a new concept for the Design of Solar Absorbers, Adv. Funct. Mater. 12 (2002) 197–202. [7] D. Katzen, Y. Mastai, Thin films of silica-carbon nanocomposites for selective solar absorbers, Appl. Surf. Sci. 248 (2005) 514–517. [8] Q.C. Zhang, Y.G. Shen, High performance W-AlN cermet solar coatings designed by modelling calculations and deposited by DC magnetron sputtering, Sol. Energy Mater. Sol. Cells 81 (2004) 25–37. [9] G.B. Smith, A.V. Radchik, Columnar cermet structures in solar energy materials: can one model spectral response with simple effective medium theories, Sol. Energy Mater. Sol. Cells 54 (1998) 387–396.
Z. Wu et al. / Surface & Coatings Technology 302 (2016) 468–473 [10] Q.C. Zhang, Y.B. Yin, High efficiency Mo-Al2O3 cermet selective surfaces for hightemperature application, Sol. Energy Mater. Sol. Cells 40 (1996) 43–53. [11] T.S. Sathiaraj, R. Thangaraj, Optical properties of selectively absorbing r.f. sputtered Ni-Al2O3 composite films, Thin Solid Films 195 (1991) 33–42. [12] T. Tesfamichael, A. Roos, Treatment of antireflection on tin oxide coated anodized aluminum selective absorber surface, Sol. Energy Mater. Sol. Cells 54 (1998) 213–221. [13] G. Katumba, A. Forbes, Quantifying the efficacy of solar selective absorber materials: the case of carbon nanoparticles dispersed in SiO2, ZnO and NiO matrices, Proc. SPIE 7046 (2008) (70460B-1). [14] G. Katumba, L. Olumekor, Optical, thermal and structural characteristics of carbon nanoparticles embedded in ZnO and NiO as selective solar absorbers, Sol. Energy Mater. Sol. Cells 92 (2008) 1285–1292.
473
[15] B. Cheng, K.K. Wang, G.R. Han, Y. Liu, Porous carbon–titania nanocomposite films for spectrally solar selective absorbers, Sol. Energy Mater. Sol. Cells 133 (2015) 126–132. [16] B. Cheng, K.K. Wang, G.R. Han, Y. Liu, Preparation and characterization of porous carbon–titania nanocomposite films as solar selective absorbers, J. Alloys Compd. 635 (2015) 129–135. [17] J.X. Yao, M. Takahashi, T. Yoko, Controlled preparation of macroporous TiO2 films by photo polymerization-induced phase separation method and their photocatalytic performance, Thin Solid Films 517 (2009) 6479–6485. [18] A.C. Ferrari, J. Robertson, Interpretation of Raman spectra of disordered and amorphous carbon, Phys. Rev. B 61 (2000) 14095–14107.