Effect of Annealing Temperature on Structural and Optical Properties of SnO₂ Nanoparticles

INTRODUCTION
For some time now, Habte et al., showed SnO₂ nanostructures have been getting a lot of attention in optics, gas sensors, and solar cells because of their unique properties [1]. They seem well suited for optoelectronic uses due to a broad energy band gap and also a very good conductivity, caused by oxygen vacancies [2-5]. As noted by Stadler, Even if a tremendous progress has been made in the preparation of SnO₂ nanoparticles, the part about being able to precisely engineer their structural features and optical characteristics, so that we can optimize the overall performance, is still quite a major challenge [6]. Thermal annealing, based on Sun et al., is one of the best approaches used to tune crystal defects, crystallite size, and the outer surface of the nanoparticles [7].
The need for a systematic investigation of the annealing temperature is mostly dictated by the fact that many properties of SnO₂, like its optical gap, photoluminescence intensity, and the peaks positions, plus its surface resistance, really do depend strongly on the heat treatment history [8-10]. So, mentioned by Sefardjella et al., if the temperature goes up too much, it can trigger excessive grain growth and a noticeable fall in the specific surface area, and also it may induce phase transitions, which are not really welcome when we think about sensor use or photocatalytic tasks [11]. On the other hand, if the temperature is too low, then organic groups are not fully removed, and the crystallization degree cannot increase properly. Therefore, finding the most appropriate temperature range where crystal ordering improves while optical defects are still preserved is very crucial for this research field [12-14].
Another requirement for the research according to Kolesnikov et al., is related to how naturally complex the interrelation is between annealing temperature and oxygen vacancies within SnO₂ nanoparticles [15]. Those oxygen vacancies can influence material properties in two rather different ways, first they boost conductance as well as optical responses by forming energy states inside the gap, and second if there is too much of these defects then non-radiative recombination can happen and the quantum efficiency may go down [16-18]. The way mentioned by Bharti et al., annealing temperature acts on both the favorable side and the unfavorable side of this whole situation might become a good pathway for designing oxide nanostructures [19]. Older studies stated by [20], mostly talked about variations in particle size or changes in the energy gap, but they did not really take into account defect formation or how that links to optical properties. In other words, the correlation between defect development and the optical behavior was somehow missing, especially when temperature is the control parameter.
Also, because thin films, plus nanostructured inks are getting developed more and more, there appears this need for powder materials that have particle size distributions as noted by Ponte et al., that are quite well controlled, and also they need specific optical features [21]. The annealing process, which described by Kong et al., is kind of a key player in the industrial fabrication stage, becomes a step that can either confirm the making of high-quality material, or on the contrary it can ruin everything and the synthesis turns into something basically null and void [22-25]. Therefore, in [18] discussed it’s really important to understand how structural properties like lattice strain, lattice parameter, and crystallite size along with optical properties, such as energy bandgap and emission spectra, change when the annealing temperature changes. This interest comes both from theoretical angles and practical needs, and honestly it matters either way.
With all these aspects in mind and previous works review [27-32], in the current work we try to investigate how the annealing temperature actually affects the structural as well as optical characteristics of tin oxide nanoparticles, which are prepared via sol gel method. It should be said that the temperature range we picked is expected to be not just one thing, but rather move from relatively low values, to avoid surface defects in the particles, to higher ones for better crystallinity. So, in this article, we have attempted to run a rather broad exploration of the concurrent variations in the structural and optical parameters when the samples are annealed. Overall, we believe the obtained findings will help us understand better, how to design more efficient SnO2 optical and sensing elements, maybe even with improved performance in real use.

MATERIALS AND METHODS
Synthesis of Tin Dioxide Nanoparticles by Sol-Gel Method
For making SnO₂ nanoparticles, tin(II) chloride dihydrate (SnCl₂•2H₂O) was used as the initial precursor and absolute ethanol acted like the solvent. First the calculated amount of tin salt was dissolved in the ethanol,then stirred with a magnetic stirrer around room temperature for about 45 minutes so that a more uniform, homogeneous mix was obtained. After that, deionized water was added slowly into the mixture to trigger hydrolysis, and during the whole period the pH was kept somewhere in the range of 2–3,through drops of hydrochloric acid. The resulting white gel was allowed to sit, and age at room temperature for 12 hours. Then it was dried in an oven at 80°C,for 24 hours. In the end, the yellow solid that formed was crushed into powder.

Heat treatment and annealing program
The base powder that came out of the synthesis process was sort of divided into four little samples, and each one was later annealed at a different temperature. Those four samples were all kept inside alumina crucibles inside a tubular electric furnace, and then exposed to a regular flow of air. For the heating, the temperature increment rate was set to 5°C/min, pretty constant, more or less. After reaching the target temperature, every sample was held at that fixed temperature for 2 h, no changing conditions in between. The temperatures chosen for this step were 300, 500, 700 and 900°C. Then, once the required time was over, the furnace was simply let to cool back down to room temperature and the resulting powders were removed from the furnace, without any abrupt thermal shift.

Characterization tools and methods
Like, the XRD pattern of the samples was gotten using a diffractometer with Cu Kα radiation (λ = 1.5406 Å) at angular steps of 0.02 deg, spanning from 20 up to 80 degrees. The average crystallite size was then estimated using the Scherrer relation and basically the width of the most prominent peaks. For the optical behavior, a diffuse reflectance spectrum was recorded in the visible–ultraviolet range with a spectrophotometer equipped with an integrating sphere, and the bandgap values were computed from the Tauc plots. Also, the PL spectra of the materials were collected at room temperature using an excitation wavelength of 325 nm, and the laser power was kept low. Everything was done under identical experimental settings, without any extra handling or further preparation of the powder samples.

RESULTS AND DISCUSSION
For this study, tin oxide nanoparticles that were produced by a sol-gel synthesis process have been heated at four different temperatures, in order to show kind of their structures and optical behavior. XRD, UV-vis, and photoluminescence measurements were done on each sample, so that quantitative data could be gathered and is shown in the Table 1.
As the temperatures during the annealing process rise, a small shift toward higher angles shows up in all the peaks. It indicates a decrease in the compressive strain inside the lattice, moving closer to bulk parameterization, or, in other words, settling nearer to the lattice bulk state. This behavior is seen most clearly in the more closely packed reflections like (110) and (211), where the lattice seems to relax a bit faster as the annealing goes on.
The distance between the interphase in all samples looks very close to the standard card readings, and the relative deviation comes to be almost zero at a temperature of 900 °C. From this it can be said that, at high temperatures, there are no more stresses left in the crystal lattice structure, or at least that is what the results hint at.
The lattice parameters show a small reduction in value, on account of an increase in temperature due to the elimination of imperfections in the crystal, more or less. Meanwhile the c/a ratio stays pretty much constant, because the tetragonal symmetry is kept intact by the crystal during annealing.
The fact that there is an increase in crystallite size from 6.8 nm at 300°C to 38.6 nm at 900°C suggests that there has been considerable growth of crystals, due to the intrusion of grain boundaries. Also, the difference in size in both directions is kind of minimal, which indicates isotropic growth.
When temperature goes up there’s a noticeable drop in both strain levels and dislocation density. This decrease in the strain numbers, going from about 0.59% down to 0.11%, feels like clear proof that something atomic is reorganizing, and that the stress inside the lattice is reduced. 
This behavior is described in Fig. 1 below where there is this increase in intensity, and a decrease in peak width as the temperature rises. I mean it seems to match what one would expect, with a kind of crystalline growth going on, plus less of a crowd of defects in the lattice structure of tetragonal cassiterite.
As the annealing temperature goes up, the energy gap value drops, starting at 3.68 and then going down to 3.59 eV, it is getting closer to the bulk energy gap value. At the same time the Auerbach energy value, falls from about 420 to 110 meV, which implies a sort of decay of those irregularities in the absorption tail.
There is a very strong negative correlation between crystal size and the energy band gap ( r = -0.986) and, kind of obviously, a very strong positive correlation between the strain and the energy band gap (r = 0.994). 
Taking the linear portion of the curves and extrapolating it out to the energy scale, it turns out that the pattern of gap shrinkage still keeps going. In other words, the drop in the gap value of about 0.09 eV for samples ranging from 300 to 900°, can be mostly put down to a fading quantum size effect and also to the way levels form near the band edge, with less “regular” structuring than before.
The PL spectrum has three main emission peaks, like the ultraviolet glow very near the band edge about 360 nm, plus blue and green emissions which show up due to oxygen vacancies. As the temperature rises, the UV emission goes up a lot, while the defect related emission keeps dropping continuously.
The reduction in the value of the FWHM of the UV emission peak, from 4.21 to 8.13 nm, kinda suggests that the size distribution became more uniform, and this leads to lower energy dispersion within the band. Also, the ratio of the UV to visible emission intensity grows about 25 times when the temperature is raised.
This move of the color coordinates into the bluer areas, with the rise of temperature is mostly linked to a reduced contribution from the green luminescence and to the dominance of ultraviolet luminescence. It kind of shows that you can tune the emission color just by changing the annealing temperature.
Reflectance go down and absorption go up, because crystallization happens and the scattering losses reduction occurs right at the grain boundary region. The 900°C sample, that one had the lowest reflectance and the highest absorption overall, it was treated as the best option, for photocatalytic applications.
The sample with the 900° annealing temperature is in the boundary region between the nano and bulk states regarding the crystalline state, while in terms of energy gap and Auerbach energy, it represents an almost ideal crystal. The exceptionally high UV to visible ratio in the spectrum of this sample makes it quite different from any previously studied nanostructures. To summarize, it can be stated that this work clearly demonstrates how the precise control over the annealing temperature can help us obtain a controlled combination of the nanoparticle crystal state and its optical characteristics. The observed linear dependencies between the reduction in crystal strain and decrease in energy gap and appearance of exciton luminescence instead of defects luminescence give us a continuous image of how the material evolves.

CONCLUSION
It was found in an organized manner how annealing temperatures affect the structure and optical properties of tin dioxide nanoparticles produced using the sol-gel technique. Based on the results of the analysis of the diffraction pattern by X-rays, it can be concluded that as temperature rises from 300°C to 900°C, the crystallite size increases from 6.8 nm to 38.6 nm, demonstrating the onset of solid-state sintering and grain boundary movement. In addition, the increase in temperature caused a reduction in lattice strain from 0.59% to 0.11%, as well as a two-fold decrease in dislocation density, suggesting the rearrangement of atoms, as well as depletion of structural defects in the process of annealing. This is confirmed by the minor shifts of the diffraction peaks towards larger angles. On the basis of optical characteristics, a decrease in the direct energy gap was observed when increasing the temperature from 3.68 eV at 300°C to 3.59 eV at 900°C. This observation was substantiated by the correlation analysis performed according to the strong negative correlation between energy gap and crystal growth (r = -0.986) and a positive one between energy gap and strain (r = 0.994). This effect can be explained by decreasing the quantum confinement effects influence and joining oxygen vacancy levels to the valence band edge. Moreover, Auerbach energy was found to have significantly lowered from 420 to 110 mEv, showing an absence of absorption tails and a high level of local ordering of oxygen vacancies. Thus, taking into account the reflectance and absorption coefficient results, the optimal sample is the one obtained at 900°C since it has the lowest value of reflectance (550 nm) and the maximum absorption edge. The analysis of photoluminescence spectrum showed that annealing resulted in a qualitatively and quantitatively significant change with increasing temperature, the intensity of the emission in ultraviolet region (360 nm) grew significantly, and at the same time the blue and green emissions arising as a result of oxygen vacancies were continually decreased. Integral ratio between UV visible intensities changed from less than one for 300° C sample to 6.78 for 900°C one. This is also associated with a reduction in the full-width-half-maximum of the UV peak from 4.21 to 8.13 nm, which indicated homogeneity of the intraband energy distribution and predominance of excitonic recombination processes. It should be mentioned that, as a result, CIE chromaticity coordinates moved from greenish white region to blue one. In conclusion, this paper has shown how the annealing temperature acts as an effective means of controlling the crystallinity, energy bandgap, and emission wavelength of the SnO₂ nanoparticles. Statistical correlations can be established between the various parameters to make predictions of performance at intermediate temperatures. Among all the samples prepared, the sample annealed at 900°C performed better in terms of nanometer particle size, bulk-like energy gap, least disorder, and high edge/defect emission. This information can prove valuable in the development of methods of production of tin oxide sensors and LEDs.

CONFLICT OF INTEREST
The authors declare that there is no conflict of interests regarding the publication of this manuscript.