Volume 4, number 2
OPTICAL PROPERTIES OF TIN OXIDE ON ANODIZED ALU~INIU~ A. ROOS and M. GEORGSON Department
of Technolog, 2 December
Box 534. S-751 ?I
The solar selective properties of tin oxide on anodized aluminium have been investigated. It has been shown that the pyrolytic spray method can be used to apply an IR-reflecting coating on black pigmented anodized aiuminium. This surface exhibits high solar selectivity and the excellent mechanical. chemical and thermal stability of this sandwich makes it an interesting alternative in solar energy applications.
The conversion of solar radiation into heat requires a selective absorber surface in order to be efficient at temperatures high enough for hot water production. The term “selective” in the case of photothermal conversion simply means that the surface in question should have high absorptivity in the solar wavelength range and low emissivity for thermal radiation [ 1,2]. Many different selective absorber surfaces have been suggested over the past 25 years 13-61 and the most common way to achieve the selectivity is to apply an oxide coating on top of a metal. The oxide thus absorbs solar radiation while the metal gives the desired infrared properties provided that the oxide coating is thin enough to be transparent for infrared radiation. This is the dilemma of this kind of selective absorber. The oxide has to be of the order of a few hundred nanometers only, and such a thin coating is usually more likely to deteriorate than a thick coating. This is unfortunate since an ~portant requ~ement is that the absorber surface is stable under every possible operating condition, Another possible way of achieving the selectivity is therefore of great interest, namely a thick absorbing layer covered with an equally thick transparent IR-reflector. In this case the transparent IR-reflector provides the low emissivity in the infrared and transmits the solar radiation to the next layer where it is absorbed. The base metal has nothing to do with the optical properties of this system. Not many transparent conductors exist that are suitable for this application because of the high re0 167-577x/86/$03.50 0 Elsevier Science Publishers B.V. (Nosh-HoUand Physics abashing Division)
quirements on durability. Doped SnO, is perhaps the most promising because of its very high thermal, chemical and mechanical stability. The problem is to find a suitable absorbing coating/metal combination, since highly conducting tin oxide can to date only be produced with the pyrolytic spray method. This means that the surface has to be heated to about 400°C and sprayed with a very corrosive SnCl, solution. The properties of Sn02 on enamelled steel [7,8] and black molybdenum ]9] have been reported earlier and in this Letter we report on the properties of tin oxide on anodized aluminium. This is a surface that offers many advantages: Aluminium is fairly inexpensive, light and has high thermal conductivity. The anodizing technique is well established and the resulting aluminium oxide can be pigmented to produce a highly absorbing coating. This surface is in fact used as a selective surface by itself, without the tin oxide, but then the oxide layer has to be thin as mentioned earlier. In our case the aluminium oxide is around 2 Mmthick and this surface is thermally and chemically stable enough to be coated with tin oxide using the pyrolytic method. The optical properties of tin oxide on glass and the importance of crystallization and doping have been reported earlier for films produced at our laboratory [ 10,l I]. It is straightforward to use the same technique to apply tin oxide onto the anodized aluminium surface. In this Letter we wish to point out that conductive tin oxide can be applied 49
Volume 4, number 2
not pigFig. 1. Reflectance versus wavelength for: mented AlsOs/Al, - - - pigmented AlsOs/Al, -.-.SnOs/not pigmented AlsOs/AI, . . . . SnOs/pigmented AlsOs/Al. (a) Visible,and near infrared wavelength range (total hemispherical reflectance). (b) Infrared wavelength range (specular reflectance).
aluminium, that the resulting surface shows reasonable selectivity and is extremely stable, In fig. 1 the optical reflectance of two anodized aluminium surfaces is shown with and without the tin oxide coating. Fig. la gives the total hemispherical reflectance in the solar wavelength range and fig, lb
the specular reflectance in the infrared. It is obvious that the reflectance in the infrared for the coated surfaces depends on the tin oxide only. The strong absorption lines present for the bare aluminium oxide are completely absent which, together with the high level of the reflectance, shows that the tin oxide films have the desired properties. The fact that a large part of the reflectance is diffuse, means that the total thermal emittance cannot be calculated with reasonable accuracy from the reflectance curves. Instead the total hemispherical emittance has been measured using a calorimetric method. This gave the values e = 0.25 for sample I in fig. 1 (not pigmented) and e = 0.34 for sample II (pigmented). The difference is to some extent due to surface roughness. In fig. 2 the surface profiles, as recorded by a mechanical stylus profilometer, are shown. It can be seen that sample II has a more pronounced surface roughness than sample I. It is clear that these emittance values are higher than can be achieved for tin oxide on glass. This, too, is partly due to surface roughness, but it must be pointed out that no efforts have been made at this stage to optimize the spray parameters for these surfaces. It is in no way obvious, however, that the same parameters (temperature, flow rate, concentration of the spray solution and doping) as for the glass case should be used to produce the best quality film on anozided aluminium. These considerations remain to be tested. We believe that it will be possible to considerably improve the emittance value of this type of surface. The total solar absorption as calculated from the curves in fig. 2a depends to a large extent on the refractive index of the tin oxide, which in this case is too thick to act as an antireflective layer. The high absorption in the near infrared range for both tin oxide coated samples is in accord with the Drude theory for
Fig, 2. Surface profiles for the SnOs/A120s/Al recorded by Sloan Tallysurf stylus instrument.
Volume 4, number 2
free-electron materials [ 121. This is advantageous in this case since a large part of the solar energy is found in this range. The total solar absorption (a) for the black sample with tin oxide is (Y= 0.9 1. This value can be expected to be fairly independent of preparation parameters and the thickness of the two oxide layers. The fact that makes this surface an attractive candidate for photothermal conversion is its superior stability. Preliminary tests have shown that this sur-
face can be immersed in a 10% sulphuric acid solution (pH= 10) and annealed at 230°C for three days without any detectable change in the optical properties. The samples have also been handled without any protection, washed, rinsed under tap water and dried without being affected. We have reasons to believe that further ageing tests will have no effects on this surface. In summary, we have shown that it is possible to apply a conductive tin oxide coating on anodized aluminium using the pyrolytic spray method. This yields a surface with high solar absorption (> 0.9) and low thermal emittance (< 0.35). It is based on thick layers of very stable compounds which results in a selective surface with extremely high thermal, mechanical and chemical stability.
We would like to express our sincere gratitude to Dr. Thommy Karlsson and Dr. Carl-G. Ribbing for their contributions to this work.
References [II 09. Agnihotri and B.K. Gupta, Solar selective surfaces (Wiley, New York, 1981).
PI B.O. Seraphin, in: Solar energy conversion, eds. A.E. Dixon and J.D. Leslie (Pergamon Press, London, 1979). 131 D.J. Close, CSIRO Report ED7 (1962). 141 J.C.C. Fan,Thin Solid Films 54 (1978) 139. [51 J.J. Cuomo, J.F. Ziegler and J.M. Woodall, Appl. Phys. Letters 26 (1975) 557.
[61 B. Karlsson,C.G. Ribbing, A. Roos, E. Valkonen and T. Karlsson, Physica Scripta 25 (1982) 826.
[71 F. Simonis, M. van der Leij and C.J. Hoogendoorn, Solar Energy Mat. l(l979)
181 C. Belle&, A. Bonanno, M. Conti, L. La Rotonda and R. Visentin, Nuovo Cimento 1C (1978) 6.
191 G.E. Carver, S. Karbal, A. Donnadieu, A. Chaoui and J.C. Manifacier, Mat. Res. Bull. 17 (1982) 527. [lOI T. Karlsson, A. Roos and C.G. Ribbing, Physica Scripta 25 (1982) 772. 1111 T. Karlsson, A. ROOSand C.G. Ribbing, Solar Energy Mat. 11 (1985) 469.
[I21 K.L. Chopra, F. Major and D.K. Pandya, Thin Solid Films 102 (1983) 1.