Photovoltaic performances of heterostructures based on photoelectrochemically textured macroporous n type silicon in contact with a sprayed transparent, conductive oxide (TCO = SnO₂) are limited by three main factors : (i) the deposition process must be carried out at a temperature lower than 400 C to keep a high energy barrier ; (ii) the contacting SnO₂ layer must be undoped for the same reason ; (iii) as a metallic grid is unefficient on top of a macroporous structure, the current collection has to be achieved thanks to a TCO overlayer with a resistivity as low as possible [1]. From a more general point of view, there is an interest to realize highly conductive TCO layers for electrochemical or solid state applications, at relatively low temperatures and by a low cost technique.
Antimony doping produces badly conductive tin oxide layers with a resistivity of the order of 10^-2 W.cm at 400 C, increasing up to 1 W.cm around 350 C. From this point of view, fluorine doping is much more attractive. Using a standard spray solution (SnCl₄ 0.2M / CH₃OH + NH₄F ; F/Sn=70%) optimized for a deposition temperature of 500 C, resistivity of 5-8 x10^-3 W.cm can be obtained around 380 C. This is about one order of magnitude below the best resistivity values we obtain at 500 C (i.e. 4 x10^-4 W.cm). Physical properties of sprayed SnO₂ : F (FTO) films are sensitive to any change of the cross-linked process parameters, mainly the deposition temperature, the solution composition, the solution and gas fluxes, the geometry of the deposition chamber, the exhaust of reaction products...
Keeping in mind that the film growth is depending on a lot of coupled experimental factors, it is worthwhile to compare our present know-how to the existing literature data on low-temperature sprayed SnO₂ films. Figure 1 shows the results taken out from references [2] and [3]. It is clear that increasing the F/Sn ratio tends to decrease the film resistivity, but with a visible saturation effect depending on the deposition temperature. At our knowledge, the data issued from reference [3], corresponding to very large F/Sn ratios, exhibit the lowest resistivity values never obtained for FTO sprayed films and constitute a real challenge for people working in this field. It appears that particular experimental conditions were found to improve the incorporation of fluorine and chlorine atoms as donor centres, reaching carrier density as large as 3 x10²¹ cm^-3, to be compared with 4 x10²⁰ cm^-3 for F / Sn=70% in our standard conditions. Starting with SnCl₄ and NH₄F as precursors leads to a competitive doping by Cl and F, strongly enhanced when the deposition temperature is decreased below 400 C [4]. AES and SIMMS studies showed that the F/Sn and Cl/Sn ratios in the films increased correlatively with the increase of the F/Sn ratio in solution [2]. From our experience, the ratio F/Sn is an important factor to lower the FTO resistivity at low deposition temperature but not the only determining one.
For instance, with a TCO overlayer of resistivity 10^-3 W.cm deposited at 380 C, we currently obtain for a flat-mirror Si(100)/SnO₂ cell : I_sc = 22 mA.cm^-2 ; V_oc = 560 mV ; fill factor = 0.52 for a current collection by a dot of 3 mm² and a diode area of 64 mm².

Fig.1 : Resistivity of FTO layers deposited at various temperatures with respect to the F/Sn atomic ratio in the spray solution, for doping with NH₄F [2,3] and SnF₂ [5]. The arrows represent our best resistivity values presently.

Another way recently explored was to consider a single precursor for tin and fluorine. Thus, using SnF₂/CH₃OH solutions (F/Sn = 200%), FTO films with a resitivity of 6.10-4 W.cm were realized in the deposition temperature range 350-400 C [5]. For the moment, we are able to reproduce these values, but at higher temperatures and a further optimization work is needed. Interestingly, when starting from fluorine-rich precursors, as SnF₂ in CH₃OH solution, sprayed around 400 C, it is possible to obtain yellow-colored FTO layers. These layers present a high dc resistivity (0.01 to 1 W.cm). The F/Sn ratio inside the film was found high (10 to 20% as determined by Nuclear Reaction Analysis) with a constant profile. RHEED observations show that these FTO layers are constituted by very small grains (grain size < 10 nm). A strong near infrared absorption was observed, indicating that the SnO₂ grains are conductive by fluorine doping, with a carrier density of 6.1019 cm-3, as deduced from the plasma frequency. This class of FTO material can be considered as a dispersion of SnO₂ conductive nanocrystallites embedded in an intergranular, badly conductive, -Sn-F-Sn-F- matrix.

REFERENCES :

[1]	H. Cachet, J. Bruneaux, G. Folcher, C. Lévy-Clément, C. Vard and M. Neumann-Spallart, Solar Energy Mater. Solar Cells, 46 (1997) 101-114.
[2]	M. Fantini, I. Torriani, Thin Solid Films, 138 (1986) 255-265.
[3]	C. Agashe, M.G. Takwale, B.R. Marathe and V.G. Bhide, Solar Energy Mater., 17 (1988) 99-117.
[4]	I. Chambouleyron, C. Constantino, M. Fantini and M. Farias, Solar Energy Mater., 9 (1983) 127-138.
[5]	G.C. Morris and A.E. Mc Elnea, Appl. Surf. Sci., 92 (1996) 167-170.

[BACK] [MAIN] [FURTHER]