1. Introduction
Semiconductor oxide thin films are materials with numerous applications in electronic and optoelectronic devices as well as some other applications such as protective coatings, heat mirrors, and catalysis [1-3]. In the context of world energy demand, particularly energy conversion, transparent conductive oxides (TCOs) based on semiconductor oxides, play a crucial role in the development of new-fashioned thin film solar cells [4]. As a matter of fact, in the organic solar cells, such as in dye-sensitized nanocrystalline TiO2 solar cells (DSSCs), TCOs are used in tandem structures developed on ordinary glass and flexible substrates [5].
In this respect, TCOs for DSSCs manufacturing should present a good chemical stability due to the possible chemical reactions occurring in the manufacturing process, as in the annealing steps, and also the chemical interactions promoted by electrolytes in the line of duty, that define the lifetime of the device. In our developed work on organic thin film solar cells, we have noticed that tin oxide electrodes offer the chemical stability required for device processing, contrary to results obtained from indium oxide films that, despite the higher electrical conductivity, is presented a degradation to metallic indium during the process, which in turn affects adversely the performance of the device.
The manufacturing of thin film devices based on nonsophisticated techniques, with a competitive performance has encouraged many researchers in the challenge of reaching higher efficiencies through the improvement of the material properties. Chemical deposition techniques provide a simple and economical way of processing good quality films demanded in the manufacturing of different devices. Some of the relevant chemical techniques for the deposition of quality films are sol-gel [6], homogeneous precipitation [7], chemical vapor deposition [8], and chemical spray techniques [9], among others. Results about deposition of quality SnO2 films by chemical spray technique either in pneumatic or ultrasonic atomization process have been continuously reported [10,11]. In the pneumatic process, the deposition is developed at atmospheric pressure and the system is very simple; whereas in the ultrasonic, it is necessary a closed reaction chamber, as a diluted fog instead of a jet of droplets is now directed on the hot substrate. In this deposition technique any slight disturbance generated by the exhaust system can modify the fog pattern, causing a non uniform growth, and hence, films with a like-rainbow finish are obtained. Nevertheless, according to our experience an advantage of the ultrasonic route is the significantly saving of reactants during the processing films.
As a consequence of the economic adaptation of the atomization equipment, the number of reports on SnO2 thin films by ultrasonic spray is increasing now, and a better understanding of the process involved is being reached. Studies on the effect of post annealing treatments [12], substrate temperature [13], texture [14], microstructure [15], and doping [16] have been published. E. Elangovan, et al. [11] reported SnO2:F thin films with a resistivity in the order of 2 × 10–4 Ωcm, starting from stannous chloride (SnCl2) and ammonium fluoride (NH4F). Also, the role of solution preparation in the deposition of tin oxide films by chemical spray technique has been raised [17]. Tin oxide thin films deposited by ultrasonic spray pyrolysis are usually reported starting from SnCl2 and ethanol. In our case due to limitations of equipment, methanol has been adopted as the main solvent, since this can be atomized easily.
In this work, the effect of both water content in the starting solutions and substrate temperature on the physical characteristics of SnO2 films, deposited by ultrasonic spray pyrolysis, starting from tin chloride and hydrofluoric acid, is reported.
2. Experimental
SnO2:F thin films were deposited by ultrasonic spray pyrolysis from a 0.2 M starting solution of tin chloride (SnCl4·5H2O), dissolved in a mix of water and methanol. The water content was varied in five different volumes, namely, 2.5, 5.0, 10.0, 20.0 and 30.0 ml for a total solution volume of 100 ml. Hydrofluoric acid at a fixed [F]/[Sn] ratio of 30 at % was used as F precursor. The selection of this ratio is based in previous experimental work, where we found that the optimum fluorine doping concentration oscillate between 20 and 40 at %.
SnO2:F samples were deposited on 2.5 cm × 5.0 cm clean glass substrates at five different substrate temperatures, namely, 375˚C, 400˚C, 425˚C, and 450˚C, at a fixed deposition time of 12 min. The cleaning process was as follows: 1) a five minutes ultrasonic bath in a trichloroethylene for degreasing the substrates; followed by 2) a five minutes bath in methyl alcohol; 3) a five minutes ultrasonic bath in acetone [CH3COCH3]; and finally, 4) a drying process by a jet of gas nitrogen [N2].
The electrical sheet resistance of all as-deposited samples was measured by the four point probe technique by using a Veeco equipment, with the appropriate geometric correction, p/ln2 = 4.53. The structure of the as-deposited films was characterized by means of X-ray diffraction in a Pan-Analytical XPert Pro system, by using the θ - 2θ technique, based on the Cu-Kα radiation (λ = 1.5405 Å). Scanning electron micrographs were obtained from a Jeol JSM 5400 LV microscope. Chemical composition of all the SnO2 films was determined by energy dispersive spectroscopy (EDS) with a detector MORAN (Quest) having a 136 eV resolution. The optical transmittance at normal incidence was measured with a double-beam UVVis Shimadzu spectrophotometer, in the UV-visible region (300 - 1000 nm) without glass substrate correction. The film thickness was estimated according to the Manifacier’s formula [18], and corroborated by direct measurements with a KLA Tencor P15 profilometer. The values estimated were around 600 nm.
3. Results and Discussion
3.1. Electrical Properties
Electrical resistivities for all as-grown SnO2:F films deposited were calculated from the mathematical product of the respective sheet resistances and thickness values. In Table 1 are listed the results obtained. From these, it can be observed that an increase in the water content leads to a decreasing in the electrical resistivity, reaching a minimum magnitude in samples deposited at 450˚C from a starting solution with
. Further increase in the substrate temperature increases the electrical resistance of the films. These results show that both substrate temperature and water content influence the growth kinetic and consequently the physical properties of the films. This behavior is associated with the variation of the F incorporation into the SnO2 lattice as the solution conditions and deposition temperature are changed.
Table 1. Electrical resistivity values of SnO2:F films deposited at different substrate temperatures, from starting solutions with different 
ratios.
3.2. Structure
Figure 1 shows the X ray diffraction spectra of SnO2:F thin films deposited at 450˚C from starting solutions with different water contents, for 2θ values from 20˚ to 70˚. All samples were polycrystalline, and the peaks fit well with the different reflections of the SnO2 cubic rutile structure, ASTM standard card JCPDS No. 41-1445 [19]. All spectra show the (200) direction as preferred growth orientation, whereas the (110) and (310) directions increase with the water content in the starting solution. These results demonstrate that, as was expected, the preferred growth orientation is sensible to the water content in the starting solutions. Absence of the Sn3O4 phase was also confirmed, as substrate temperatures higher than 400˚C guarantee only SnO2 formation. Additionally, no fluorine phases were detected despite the fact that a high [F/Zn] = 30 at % ratio in starting solution, was added. This result confirms the highly volatile character of fluorine compounds during the growth process, and consequently its low efficiency of incorporation into the SnO2 lattice.