**3. Results and discussion**

#### **3.1 Physical properties of tin oxide thin films obtained by sol-gel technique (the method of spreading)**

Fig. 1 shows the transmission spectra of SnOx films deposited on a glass substrate by spreading method. Transparency of films lies in the range of 80-90% and increases over the entire range of wavelengths with the increase of temperature up to 400°C.

X-ray diffraction studies of the SnOx films, obtained at various concentrations of tin ions in the colloidal solutions, led to the following results.

a. In the case of high tin concentration 0.83 mol/L (~ 0.04 ml aliquot) immediately after deposition and drying at 80°С, the film structure is similar to the amorphous (Fig. 2a)

micropipettes were deposited on the cleaned surface of the microscope glass slides. It was assumed that the number of tin atoms in the films will be identical (~ 3.25 × 10-5 mole), and the film thickness is ~ 350 nm. After the deposition, the films were dried for 1 hour at 80ºC. Then the samples were annealed at 100, 200 or 400ºC for 1 hour. The thickness of the films,

The SnO2 films with thickness of 300 nm were also fabricated by method of centrifugation. A solution was obtained by dissolving of anhydrous SnCl4 in 97% ethanol. The solution was deposited onto a glass substrate located on a table of centrifuge rotor, the rotation speed of which was ~ 3800 rpm. Centrifugation time was 3–5 s. Substrate with a layer was dried using an infrared radiation at 80°C for 3−5 minutes, and then in a muffle furnace at 400°C for 15 minutes. After cooling, the cycle repeats. Number of deposited SnO2 layers ranged

The SnO2 films with thickness of 300 nm were also deposited on cleaned microscopy glass slides by magnetron sputtering. The magnetron sputtering mode parameters were: cathode voltage Uc = 470 V, the ion beam current Iion = 35 mА, the argon-oxygen mixture pressure inside the chamber ~ 1−2.7 Pa, the oxygen concentration in the Ar-O2 mixture ~10%,

The SnO2 film's structure was investigated by X-ray diffraction using a narrow collimated (0.05×1.5 mm2) monochromatic (CuKα) X-ray beam directed at an angle of 5º to the sample surface. The average crystallite size was estimated from the width of X-rays lines by Jones method. The surface of the layers was analyzed by Atomic force microscopy (JSPM 5200, Jeol, Japan) using AFM AC technique. The optical transmittance spectra of SnOx films were measured in the wavelength range from 190 to 1100 nm by means of the SF-256 UVI and from 1100 to 2500 nm by means of the SF-256 NIR spectrophotometers (LOMO, Russia). Electrical resistance of the films was measured by a four probe technique at room temperature. For the measurements of electrical characteristics and parameters of gas sensitivity of the thin films in a wide temperature range was used a specialized

The glow discharge hydrogen plasma was generated at a pressure of 6.5 Pa with a capacitive coupled radio frequency (r.f.) power (27.12 MHz) of about 12.5 W. The temperature of

Fig. 1 shows the transmission spectra of SnOx films deposited on a glass substrate by spreading method. Transparency of films lies in the range of 80-90% and increases over the

X-ray diffraction studies of the SnOx films, obtained at various concentrations of tin ions in

a. In the case of high tin concentration 0.83 mol/L (~ 0.04 ml aliquot) immediately after deposition and drying at 80°С, the film structure is similar to the amorphous (Fig. 2a)

**3.1 Physical properties of tin oxide thin films obtained by sol-gel technique** 

entire range of wavelengths with the increase of temperature up to 400°C.

the colloidal solutions, led to the following results.

processing did not exceed 100°С. The processing time was 5 min.

deposition rate of films ~ 0.05 nm/c, the temperature of the substrate ~ 200°С.

estimated from the change in mass of the sample, was 360±40 nm.

from 12 to 15.

experimental setup.

**3. Results and discussion** 

**(the method of spreading)** 

and remains practically unchanged after annealing for 1 hour at 100°С (Fig. 2b). Separation of the broad band into two SnO2 lines, indicating the formation of crystallites, visually observed only after annealing at 200°С (Fig. 2c). Crystallite size in the planes (110) and (101) of SnO2 is ~ 1.5 nm, i.e. the crystallites are small and have an imperfect structure.

Fig. 1. The transmission spectra of tin oxide film synthesized by sol-gel technique: 1 − glass substrate, 2 − SnO2 film after annealing at 100°C, 3 − SnO2 film after annealing at 400°C.

b. A similar pattern after the deposition and annealing at 100°С is observed for films of SnO2, produced at lower tin concentrations of 0.41 mol/L (~ 0.08 ml aliquot) and 0.30 mol/L (~ 0.11 ml aliquot). After annealing at 200°С the differences are appeared, which manifest themselves in increasing intensities of SnO2 lines with decreasing of tin concentration in the colloidal solution.

Fig. 2. X-ray diffraction patterns and intensity curves for the SnOx films on glass substrates obtained by the sol-gel technique (0.04 ml solution with tin concentration of 0.83 mol/L, d = 320 nm) after: a) deposition and drying at a temperature of 80°С, and b) annealing at 100°С; c) annealing at 200°С.

Influence of Crystallization on the Properties of SnO2 Thin Films 229

necessary to determine the temperature at which a maximum sensitivity of the film to the

50 150 250 350 450 Т, о С

Fig. 4. Temperature dependence of the sensitivity of the SnO2 films (obtained from a

concentration: 1 − 15 mg/L of ethanol in the air; 2 − 0.7 mg/L of ethanol in the air.

colloidal solution with concentration of tin ions 0.14 mol/L) to the ethanol vapors of various

The greatest sensitivity of films to ethanol is observed at a temperature of 230°С. Adsorbed oxygen ions create a space charge region near the surface of SnO2 grains by extracting electrons from the material. Ethanol, being by nature a reducing gas, reacts with adsorbed

 ions and removes them from the surface of the grains, re-injecting electrons back into the material and thus lowering the resistance of the film. Peak sensitivity at 230°С shows that at this temperature, the amount of chemisorbed oxygen ions, which react with the molecules of

In Fig. 5 the sensitivity of the SnO2 films at 230°С as a function of the concentration of ethanol vapor is presented. Figs.4 and 5 show that sensitivity increases significantly with increasing of ethanol concentration. At low concentrations of ethanol vapor (0.1−1 mg/L), the dependence of the film sensitivity on the concentration is linear (Fig. 5), as is likely, there is a sufficient number of oxygen ions that react with molecules of ethanol. Linearity of sensitivity at low concentrations of ethanol can be used to create sensors basing on this film, which are sensitive to ethanol vapor in exhaled human. A visible sensitivity of the film is evident even when the concentration of ethanol vapor is about 0.05 mg/L (25

The films obtained by deposition of a colloidal solution with a concentration of tin atoms 0.83 mol/L (Fig. 5, curve 1), have higher gas sensitivity in comparison with films deposited from a solution with a concentration of tin atoms 0.14 mol/L (Fig. 5, curve 4 ). This may be due to the cluster structure of the film synthesized from a solution with a concentration of tin atoms 0.83 mol/L. Indeed, the average size of the crystallites in these films is much

1

2

∆R/R, %

test gas is observed.

О-

ppm).

smaller (1.5 nm).

ethanol, is maximal.

c. In case of the minimum concentration of 0.14 mol/L (~ 0.23 ml aliquot), the formation of a SnO2 polycrystalline film (Fig. 3a) with a crystallite size of 2.5−3.5 nm (Table 1) was observed immediately after deposition and drying. Annealing at 400°С led to the improvement of the crystallite perfection (4 lines) and an increase in their size up to ~ 5 nm (Fig. 3b).

Fig. 3. X-ray diffraction patterns and intensity curves for thin SnOx films on glass substrates obtained by the sol-gel technique (0.23 ml solution with tin concentration of 0.14 mol/L, d = 400 nm): a) after deposition and drying at a temperature of 100°С, b) after annealing at 400°С.


Table 1. Results of measurements of SnO2 crystallite sizes by X-ray diffraction

Thus, the SnO2 polycrystalline film immediately after deposition and drying without additional annealing at elevated temperatures by the sol-gel technique was derived. This may be due to better conditions for the processes of SnO2 crystallization and evaporation of HCl in the case of a lower concentration of tin atoms (0.14 mol/L) in solution. Additional annealing at 400°С leads to the formation of SnO2 crystallite with size of about 5 nm, ie close to optimal (6 nm) for high gas-sensitive films (Xu et al., 1991).

$$\text{Gas sensitivity was determined from the expression } \gamma = \frac{R\_0 - R\_{\frac{\pi}{2}}}{R\_0} \cdot 100\% = \frac{\Delta R}{R\_0} \cdot 100\% \text{, where } R\_0 \text{ is the total time of the system.}$$

R0 is the resistance of gas-sensitive layer in the clean air, Rg is the resistance of the layer in the mixture of air with detectable reducing gas. Fig.4 shows the temperature dependence of the sensitivity of SnO2 film for a given concentration of ethanol vapor in the atmosphere. Since the temperature range of sensitivity to different gases is different, it is

c. In case of the minimum concentration of 0.14 mol/L (~ 0.23 ml aliquot), the formation of a SnO2 polycrystalline film (Fig. 3a) with a crystallite size of 2.5−3.5 nm (Table 1) was observed immediately after deposition and drying. Annealing at 400°С led to the improvement of the crystallite perfection (4 lines) and an increase in their size up to ~ 5

**0,10**

**10 15 20 25 30 35 40** θ**, градус**

100°С 400°С

0 0

*R Rg R R R*

100% 100%

<sup>−</sup> <sup>Δ</sup> = ⋅ =⋅ , where

**SnO2(211)**

**0,20**

**0,30**

**0,40**

**I, относит.единицы**

Ι, arb. units

θ, degree θ, degree

Fig. 3. X-ray diffraction patterns and intensity curves for thin SnOx films on glass substrates obtained by the sol-gel technique (0.23 ml solution with tin concentration of 0.14 mol/L, d = 400 nm): a) after deposition and drying at a temperature of 100°С, b) after annealing at

Thus, the SnO2 polycrystalline film immediately after deposition and drying without additional annealing at elevated temperatures by the sol-gel technique was derived. This may be due to better conditions for the processes of SnO2 crystallization and evaporation of HCl in the case of a lower concentration of tin atoms (0.14 mol/L) in solution. Additional annealing at 400°С leads to the formation of SnO2 crystallite with size of about 5 nm, ie close

R0 is the resistance of gas-sensitive layer in the clean air, Rg is the resistance of the layer in the mixture of air with detectable reducing gas. Fig.4 shows the temperature dependence of the sensitivity of SnO2 film for a given concentration of ethanol vapor in the atmosphere. Since the temperature range of sensitivity to different gases is different, it is

γ

glass glass

SnO2 plane Crystallite size after annealing, nm

SnO2 (110) 2.5 5.5 SnO2 (101) 3.0 5.0 SnO2 (211) 3.5 3.0 Table 1. Results of measurements of SnO2 crystallite sizes by X-ray diffraction

**0,50**

**SnO2(110)**

**SnO2(101)**

**Стекло**

**SnO2(200)**

**0,60**

nm (Fig. 3b).

**0.40**

**10 15 20 25 30 35 40** θ**, градус**

to optimal (6 nm) for high gas-sensitive films (Xu et al., 1991).

Gas sensitivity was determined from the expression <sup>0</sup>

**SnO2(211)**

a) b)

**0.50**

**0.60**

**0.70**

**I, относит. единицы**

400°С.

Ι, arb. units

**0.80**

**SnO2(110)**

**Стекло**

**SnO2(101)**

**0.90**

necessary to determine the temperature at which a maximum sensitivity of the film to the test gas is observed.

Fig. 4. Temperature dependence of the sensitivity of the SnO2 films (obtained from a colloidal solution with concentration of tin ions 0.14 mol/L) to the ethanol vapors of various concentration: 1 − 15 mg/L of ethanol in the air; 2 − 0.7 mg/L of ethanol in the air.

The greatest sensitivity of films to ethanol is observed at a temperature of 230°С. Adsorbed oxygen ions create a space charge region near the surface of SnO2 grains by extracting electrons from the material. Ethanol, being by nature a reducing gas, reacts with adsorbed О ions and removes them from the surface of the grains, re-injecting electrons back into the material and thus lowering the resistance of the film. Peak sensitivity at 230°С shows that at this temperature, the amount of chemisorbed oxygen ions, which react with the molecules of ethanol, is maximal.

In Fig. 5 the sensitivity of the SnO2 films at 230°С as a function of the concentration of ethanol vapor is presented. Figs.4 and 5 show that sensitivity increases significantly with increasing of ethanol concentration. At low concentrations of ethanol vapor (0.1−1 mg/L), the dependence of the film sensitivity on the concentration is linear (Fig. 5), as is likely, there is a sufficient number of oxygen ions that react with molecules of ethanol. Linearity of sensitivity at low concentrations of ethanol can be used to create sensors basing on this film, which are sensitive to ethanol vapor in exhaled human. A visible sensitivity of the film is evident even when the concentration of ethanol vapor is about 0.05 mg/L (25 ppm).

The films obtained by deposition of a colloidal solution with a concentration of tin atoms 0.83 mol/L (Fig. 5, curve 1), have higher gas sensitivity in comparison with films deposited from a solution with a concentration of tin atoms 0.14 mol/L (Fig. 5, curve 4 ). This may be due to the cluster structure of the film synthesized from a solution with a concentration of tin atoms 0.83 mol/L. Indeed, the average size of the crystallites in these films is much smaller (1.5 nm).

Influence of Crystallization on the Properties of SnO2 Thin Films 231

the method of spreading, is transformed into a structure with an undulating surface, the wave amplitude of which is comparable to the thickness of the film. The heterogeneity of the thickness of the films may be the main reason for the lack of interference in the transmission spectra. Consequently, in order to obtain uniform thickness of the films with relatively smooth surface the centrifugation method was used, when under the centrifugal force

**500 nm** 

Fig. 7. Surface topography of tin dioxide film synthesized by sol-gel technique (a method of

**3.2 Optical, electrical, structural and sensory properties of the SnOx films, prepared** 

The films were made by centrifugation in order to reduce their electrical resistance. Scheme of the synthesis and study of the SnO2 films is shown in Figure 8. A solution of the desired tin concentration for producing of SnO2 film with thickness ~ 300 nm was obtained by dissolving of anhydrous SnCl4 in 97% ethanol. Kinematic viscosity of the solution was ~ 1.9 mm2/s. The radius of sol particles in solution, determined by the turbidimetric method

located on a table of centrifuge rotor. Rotational speed of the centrifuge was ~ 3800 rpm. Centrifugation time was 3−5 seconds. The deposited films were dried by an infrared emitter

As shown above, the films obtained by spreading, in the case of high concentration of tin atoms (0.83 mol/L) had higher gas sensitivity and lower response time and recovery time compared to films with a low concentration of tin atoms (0.14 mol/L). However, the films obtained by centrifugation from solution with a high concentration of tin atoms (0.83 mol/L) were susceptible to detachment due to poor adhesion of the film and had a greater thickness. Therefore, preference was given to films obtained from solution with low concentration of tin atoms (0.14 mol/L). Thin layers were obtained. By subsequent deposition of additional layers the multilayered films were obtained with good adhesion to

Characteristics of tin oxide films deposited by magnetron sputtering (pressure Ar-O2 mixture in the chamber −1 Pa) and sol-gel technique were compared. It was found that the relative resistance of films prepared by sol-gel technique, much faster decrease with increasing

at 80°C for 3−5 minutes and also in a muffle furnace at 400°C for 15 minutes.

glass substrate, acceptable optical properties and resistance of about 200 ohms.

spreading, a colloidal solution with a tin concentration 0.83 mol/L): a) atomic force

b)

, was 5.4 nm. The solution was deposited on a glass substrate,

10 mkm

during the sample rotation the alignment of the surface is reached.

a)

microscopy; b) scanning electron microscopy

according to formula 3 <sup>3</sup>

**by the sol-gel technique (centrifugation method)** 

4 *V <sup>r</sup>* <sup>=</sup> <sup>π</sup>

Fig. 5. The dependence of the sensitivity of the SnO2 films on the concentration of ethanol vapor at 230°С: 1 − film synthesized from a solution with a concentration of tin ions 0.83 mol/L, 2 − 0.41 mol/L, 3 − 0.30 mol/L; 4 − 0.14 mol/L.

To measure the response time, the changes of resistance are recorded as a function of time beginning from the moment when ethanol vapors are introduced into the chamber. The concentration of ethanol wondered equal to 1 mg/L (500 ppm) at T = 230°С. As a response time of the sensor was taken the time duration for which the sensor resistance dropped by 90%. Response time and recovery time of resistance to the initial value for the film deposited from a colloidal solution with a concentration of tin atoms 0.83 mol/L (Fig. 6, curve 1) are ~ 3 and 30 seconds, respectively, in contrast to the second film (0.14 mol/L), where response time and recovery time were twice as much − ~ 6 and 65 seconds, respectively.

Fig. 6. The dependence of the response time of SnO2 layer at 230°С (concentration of ethanol vapors 1 mg/L) on concentration of tin atoms in colloidal solution: 1 - 0.83 mol/L, 2 - 0.14 mol/L.

Figs 7a and 7b show the atomic force microscopy and scanning electron microscopy images, respectively, of the surface of tin oxide film, obtained by deposition of a colloidal solution with a tin concentration 0.83 mol/L on a glass substrate. Areas, protruding above the surface, have a light color, areas below the surface have a dark color (Fig.7a), and counting the height begins from the bottom point. In the process of drying, the SnO2 film, obtained by

0.0 0.2 0.4 0.6 0.8 1.0 С, мг/л

Fig. 5. The dependence of the sensitivity of the SnO2 films on the concentration of ethanol vapor at 230°С: 1 − film synthesized from a solution with a concentration of tin ions 0.83

To measure the response time, the changes of resistance are recorded as a function of time beginning from the moment when ethanol vapors are introduced into the chamber. The concentration of ethanol wondered equal to 1 mg/L (500 ppm) at T = 230°С. As a response time of the sensor was taken the time duration for which the sensor resistance dropped by 90%. Response time and recovery time of resistance to the initial value for the film deposited from a colloidal solution with a concentration of tin atoms 0.83 mol/L (Fig. 6, curve 1) are ~ 3 and 30 seconds, respectively, in contrast to the second film (0.14 mol/L), where response

t, s

Figs 7a and 7b show the atomic force microscopy and scanning electron microscopy images, respectively, of the surface of tin oxide film, obtained by deposition of a colloidal solution with a tin concentration 0.83 mol/L on a glass substrate. Areas, protruding above the surface, have a light color, areas below the surface have a dark color (Fig.7a), and counting the height begins from the bottom point. In the process of drying, the SnO2 film, obtained by

(concentration of ethanol vapors 1 mg/L) on concentration of tin atoms in colloidal

time and recovery time were twice as much − ~ 6 and 65 seconds, respectively.

C, mg/L

0

mol/L, 2 − 0.41 mol/L, 3 − 0.30 mol/L; 4 − 0.14 mol/L.

R, Mohm

solution: 1 - 0.83 mol/L, 2 - 0.14 mol/L.

Fig. 6. The dependence of the response time of SnO2 layer at 230°С

20

40

60

∆R/R, %

80

100

the method of spreading, is transformed into a structure with an undulating surface, the wave amplitude of which is comparable to the thickness of the film. The heterogeneity of the thickness of the films may be the main reason for the lack of interference in the transmission spectra. Consequently, in order to obtain uniform thickness of the films with relatively smooth surface the centrifugation method was used, when under the centrifugal force during the sample rotation the alignment of the surface is reached.

Fig. 7. Surface topography of tin dioxide film synthesized by sol-gel technique (a method of spreading, a colloidal solution with a tin concentration 0.83 mol/L): a) atomic force microscopy; b) scanning electron microscopy

#### **3.2 Optical, electrical, structural and sensory properties of the SnOx films, prepared by the sol-gel technique (centrifugation method)**

The films were made by centrifugation in order to reduce their electrical resistance. Scheme of the synthesis and study of the SnO2 films is shown in Figure 8. A solution of the desired tin concentration for producing of SnO2 film with thickness ~ 300 nm was obtained by dissolving of anhydrous SnCl4 in 97% ethanol. Kinematic viscosity of the solution was ~ 1.9 mm2/s. The radius of sol particles in solution, determined by the turbidimetric method according to formula 3 <sup>3</sup> 4 *V <sup>r</sup>* <sup>=</sup> <sup>π</sup> , was 5.4 nm. The solution was deposited on a glass substrate,

located on a table of centrifuge rotor. Rotational speed of the centrifuge was ~ 3800 rpm. Centrifugation time was 3−5 seconds. The deposited films were dried by an infrared emitter at 80°C for 3−5 minutes and also in a muffle furnace at 400°C for 15 minutes.

As shown above, the films obtained by spreading, in the case of high concentration of tin atoms (0.83 mol/L) had higher gas sensitivity and lower response time and recovery time compared to films with a low concentration of tin atoms (0.14 mol/L). However, the films obtained by centrifugation from solution with a high concentration of tin atoms (0.83 mol/L) were susceptible to detachment due to poor adhesion of the film and had a greater thickness. Therefore, preference was given to films obtained from solution with low concentration of tin atoms (0.14 mol/L). Thin layers were obtained. By subsequent deposition of additional layers the multilayered films were obtained with good adhesion to glass substrate, acceptable optical properties and resistance of about 200 ohms.

Characteristics of tin oxide films deposited by magnetron sputtering (pressure Ar-O2 mixture in the chamber −1 Pa) and sol-gel technique were compared. It was found that the relative resistance of films prepared by sol-gel technique, much faster decrease with increasing

Influence of Crystallization on the Properties of SnO2 Thin Films 233

The absolute values of the resistance of films, prepared by magnetron sputtering, changed in the temperature range 20-230°C from 69 to 11 kohm, while for the sol-gel films from 192 to 2.3 kohm. Low resistance of films obtained by magnetron sputtering at room temperature may be due to the presence of particles of tin, whose resistance increases with increasing temperature. On the other hand, the alleged occurrence of dielectric inclusions of tin monoxide may be the cause of the increased resistance of these films at 230°C in comparison

Fig. 10 shows the optical transmission spectra of films obtained in two ways. Transparency of films was about 90%. At long wavelengths there is a reduction of the transmission coefficient for the film SnOx obtained by magnetron sputtering. The absence of a similar reduction for the films prepared by sol-gel technique indicates a lack of tin particles in films

Table 2 shows the optical parameters determined from the transmission spectra by ways described in papers (Mishra et al., 2002; Ryzhikov et al., 2002). Films prepared by the sol-gel technique have high porosity, which is considered a positive factor contributing to increase

technique 1.73 318 4.10 5.35 2.49×<sup>103</sup> <sup>23</sup>

SnOx film prepared by magnetron sputtering (pressure of Ar-O2 mixture in the chamber 1 Pa), after the deposition had an amorphous structure. Annealing in the temperature range 400−550°C led to the formation of polycrystalline phases of SnO2, SnO, Sn2O3 (Fig. 11a). The results help to explain the slow decrease in the resistance of these films with increasing annealing temperature (Fig. 9) by the dielectric properties of the SnO crystallites. The SnOx film obtained by sol-gel technology had a polycrystalline structure after deposition and short-term annealing at 400°C for 15 minutes, included in the process of drying the film. Xray diffraction patterns and intensity curves of these films indicate the presence of only polycrystalline SnO2 phase (Fig. 11b). The crystallites sizes were of the order of 3−5 nm

deposition SnO2 plane Crystallite sizes, nm

SnO2 (101) 11.4 Sn2O3 (021) 8.5 SnO (101) 10.0

SnO2 (110) 5.2 SnO2 (101) 5.1 SnO2 (211) 3.0

Eg, band gap, eV

ρ, density, g/cm3

1.81 282 4.05 5.76 1.67×103 17

k, absorption coefficient, 1/cm

V, porosity, %

and, consequently, a better stoichiometry of the films of tin dioxide.

D, film thickness, nm

n, refractive index

(Table 3), that is more optimal to obtain the gas-sensitive films.

Table 2. Optical parameters of SnOх film

The method of film

Magnetron sputtering

The sol-gel technique

Table 3. Average crystallite sizes of SnO2 films

with the sol-gel films.

of their gas sensitivity.

The method of film deposition

Magnetron sputtering

The sol-gel

temperature (Fig. 9), and the increase of their resistance associated with a decrease in carrier mobility due to scattering by impurities occurs at higher temperatures (> 270°C).

Fig. 8. Scheme of the synthesis and study of nanostructured SnO2 films (sol-gel technology)

temperature (Fig. 9), and the increase of their resistance associated with a decrease in carrier

Dissolution оf SnCl4 in С2Н5ОН

temperature Control of

maturation of the solution

mobility due to scattering by impurities occurs at higher temperatures (> 270°C).

Determination of sol particle sizes by turbidimetric method

Deposition of the film on a substrate by centrifugation

Infrared drying for 2 minutes

Compaction of film by annealing at 400ºC for 15 minutes

Determination of film thickness by the change in sample mass

> The complex research 1.The method of Van der

Fig. 8. Scheme of the synthesis and study of nanostructured SnO2 films (sol-gel technology)

2.Measurement of transmission spectra 3.Temperature dependence

of resistance 4.Gas-sensitive characteristics 5.Investigation of the surface by AFM 6.X-ray analysis

Pauw

Heat treatment

+3 hours

+3 hours

+6 hours

Annealing 400ºC, 3 h

Annealing 400ºC, 6 h

Annealing 400ºC, 12h

Measurement of viscosity on

15 cycles

acidity of the solution

Plasma treatment for 5 minutes

oxygen

Research Film production Treatment techniques

hydrogen

The absolute values of the resistance of films, prepared by magnetron sputtering, changed in the temperature range 20-230°C from 69 to 11 kohm, while for the sol-gel films from 192 to 2.3 kohm. Low resistance of films obtained by magnetron sputtering at room temperature may be due to the presence of particles of tin, whose resistance increases with increasing temperature. On the other hand, the alleged occurrence of dielectric inclusions of tin monoxide may be the cause of the increased resistance of these films at 230°C in comparison with the sol-gel films.

Fig. 10 shows the optical transmission spectra of films obtained in two ways. Transparency of films was about 90%. At long wavelengths there is a reduction of the transmission coefficient for the film SnOx obtained by magnetron sputtering. The absence of a similar reduction for the films prepared by sol-gel technique indicates a lack of tin particles in films and, consequently, a better stoichiometry of the films of tin dioxide.

Table 2 shows the optical parameters determined from the transmission spectra by ways described in papers (Mishra et al., 2002; Ryzhikov et al., 2002). Films prepared by the sol-gel technique have high porosity, which is considered a positive factor contributing to increase of their gas sensitivity.


Table 2. Optical parameters of SnOх film

SnOx film prepared by magnetron sputtering (pressure of Ar-O2 mixture in the chamber 1 Pa), after the deposition had an amorphous structure. Annealing in the temperature range 400−550°C led to the formation of polycrystalline phases of SnO2, SnO, Sn2O3 (Fig. 11a). The results help to explain the slow decrease in the resistance of these films with increasing annealing temperature (Fig. 9) by the dielectric properties of the SnO crystallites. The SnOx film obtained by sol-gel technology had a polycrystalline structure after deposition and short-term annealing at 400°C for 15 minutes, included in the process of drying the film. Xray diffraction patterns and intensity curves of these films indicate the presence of only polycrystalline SnO2 phase (Fig. 11b). The crystallites sizes were of the order of 3−5 nm (Table 3), that is more optimal to obtain the gas-sensitive films.



Influence of Crystallization on the Properties of SnO2 Thin Films 235

(112)

(a)

technique.

was 70 and 90 s, respectively.

sensitivity to small concentrations of ethanol.

SnO2

Sn

Sn

O2

3 (011)

O2

3 (001)

(101)

(110)

SnO(101)

(200)

(210)

(211)

Sn

Fig. 12. Temperature dependence of the sensitivity of tin dioxide films to ethanol (concentration 1 mg/L): 1 − film deposited by magnetron sputtering, 2 − by sol-gel

O2

SnO(200)

Sn

Sn

by magnetron sputtering, and b) by the sol-gel technique.

ΔR/R, %

O2

3 (021)

O2

3 (130)

SnO

Fig. 11. X-ray diffraction patterns for thin SnOx films on glass substrates: a) films obtained

190 230 270 310 T, o C

The listed above factors also affect the dynamic performance. To measure the response time of gas-sensitive film, in the volume of the reactor a certain number of identified gas for the time of the order of tenths of a second was introduced. Fig. 13 shows the response time for the presence of ethanol vapor with a concentration of 1 mg/L (500 ppm), for which the film resistance dropped by 90%. For the films obtained by sol-gel method and magnetron sputtering, the response time was ~ 3 and 16 s, recovery time of resistance to the initial value

Fig. 14 shows the dependence of sensitivity of the SnO2 films synthesized by both sol-gel technique and magnetron sputtering, on the concentration of ethanol vapor at a temperature corresponding to the maximum of sensitivity. There is a sensitivity to trace amounts of ethanol vapor of films synthesized by sol-gel technique. In addition, the films have a linear

**1 2**

(103)

3 (041)

(b)

SnO2

(110)

(101)

(200)

(211)

(220)

(112)

Fig. 9. Temperature dependences of resistivity of the SnOx films: 1 − film deposited by magnetron sputtering, 2 − by sol-gel method.

Research of gas sensitivity of the deposited films was conducted in two stages. In the first stage, the temperature dependences of film resistance in both pure air, R0(T), and at a given concentration of the test gas, Rg(T), were measured. It was determined the temperature at which a maximum sensitivity of film to test gas observed. In the second stage, the effects of different concentrations of the gas at this temperature are measured.

Fig. 12 shows curves of temperature dependences of the sensitivity for films synthesized by both sol-gel method and magnetron sputtering, at an ethanol concentration of 1 mg/L. Maximum sensitivity of the sol-gel films to the ethanol vapors is shifted to lower temperatures (230°C), relative to films prepared by magnetron sputtering (270°C). This may be due to smaller crystallites, higher porosity and homogeneity of the phase composition of the sol-gel SnO2 films.

Fig. 10. Optical transmission spectra of SnOx films: 1 − glass, 2 − film, deposited by the sol-gel technique, 3 − film deposited by magnetron sputtering.

0 100 200 300 T, o C

Fig. 9. Temperature dependences of resistivity of the SnOx films: 1 − film deposited by

Research of gas sensitivity of the deposited films was conducted in two stages. In the first stage, the temperature dependences of film resistance in both pure air, R0(T), and at a given concentration of the test gas, Rg(T), were measured. It was determined the temperature at which a maximum sensitivity of film to test gas observed. In the second stage, the effects of

Fig. 12 shows curves of temperature dependences of the sensitivity for films synthesized by both sol-gel method and magnetron sputtering, at an ethanol concentration of 1 mg/L. Maximum sensitivity of the sol-gel films to the ethanol vapors is shifted to lower temperatures (230°C), relative to films prepared by magnetron sputtering (270°C). This may be due to smaller crystallites, higher porosity and homogeneity of the phase composition of

> 0 500 1000 1500 2000 2500 λ, nm

Fig. 10. Optical transmission spectra of SnOx films: 1 − glass, 2 − film, deposited by the

sol-gel technique, 3 − film deposited by magnetron sputtering.

**<sup>3</sup> <sup>2</sup>**

1

2

0,0

different concentrations of the gas at this temperature are measured.

0,2

0,4

0,6

R/Ro, arb.units

magnetron sputtering, 2 − by sol-gel method.

the sol-gel SnO2 films.

0

20

40

60

Т, %

80

**1**

100

0,8

1,0

Fig. 11. X-ray diffraction patterns for thin SnOx films on glass substrates: a) films obtained by magnetron sputtering, and b) by the sol-gel technique.

Fig. 12. Temperature dependence of the sensitivity of tin dioxide films to ethanol (concentration 1 mg/L): 1 − film deposited by magnetron sputtering, 2 − by sol-gel technique.

The listed above factors also affect the dynamic performance. To measure the response time of gas-sensitive film, in the volume of the reactor a certain number of identified gas for the time of the order of tenths of a second was introduced. Fig. 13 shows the response time for the presence of ethanol vapor with a concentration of 1 mg/L (500 ppm), for which the film resistance dropped by 90%. For the films obtained by sol-gel method and magnetron sputtering, the response time was ~ 3 and 16 s, recovery time of resistance to the initial value was 70 and 90 s, respectively.

Fig. 14 shows the dependence of sensitivity of the SnO2 films synthesized by both sol-gel technique and magnetron sputtering, on the concentration of ethanol vapor at a temperature corresponding to the maximum of sensitivity. There is a sensitivity to trace amounts of ethanol vapor of films synthesized by sol-gel technique. In addition, the films have a linear sensitivity to small concentrations of ethanol.

Influence of Crystallization on the Properties of SnO2 Thin Films 237

Fig. 15. Surface topography of the SnOx films on glass substrates: a) glass, b) SnOx film,

Ra, nm 2.31 3.58 0.66 Rzjis, nm 17.7 27.1 5.67 S, nm 2 264632.0 259867.1 251484.2 Rq, nm 2.92 4.69 0.842 Rz, nm 19.0 35.6 7.02 Sratio 1,06 1.04 1.01

Table 4 presents data on the film surface structure obtained using the processing software to an atomic force microscope, where: Ra is average roughness; Rzjis − average roughness of 10 points; S − area of the image; Rq − RMS roughness; Rz − the difference between maximum and minimum height of an image; Sratio − the ratio of the image area S to the area of flat

> SnОх films Magnetron sputtering The sol-gel method

**а) b) c)** 

deposited by magnetron sputtering, c) by sol-gel technique.

(glass)

Table 4. Parameters of the surface structure of SnO2 films on glass

**3.3 Influences of thermal and plasma treatments on phase composition, microstructure and physical properties of the SnOx films deposited by sol-gel** 

and structural properties of the SnOx films, deposited by the sol-gel techniques.

properties of the films is of a separate scientific and practical interest.

To modify the properties of SnO2 films, different ways of their treatment are widely used. Of particular interest is the study on the influence of thermal and plasma treatments on optical

As is known, the crystallization of SnO2 films on glass substrate takes place most intense at temperatures of 400−600°C. Heat treatment for 1 hour at temperatures of 550−600°C on energy costs can be equivalent to several hours of treatment at 400°C. However, the changes of the structural and sensory properties of the films may be non-equivalent in these conditions. Analysis of the influence of the duration of isothermal annealing on the

In recent years, the plasma treatment for modifying the properties of SnO2 films is used (Karapatnitski et al., 2000). It has been reported about the influence of plasma on the sensitivity of the sensors. An analysis of changes in the optical parameters determined from

surface S0.

Parameters Microscope slides

**technique and magnetron sputtering** 

Fig. 13. The dependence of the response time of SnO2 layer on ethanol vapors (concentration 1 mg/L): 1 − film deposited by magnetron sputtering, 2 − by sol-gel technique.

Fig. 14. The dependence of sensitivity of SnO2 films to the concentration of ethanol vapors: 1 − film deposited by magnetron sputtering, 2 − by sol-gel technique.

Fig. 15 shows the atomic-force microscope images of the surface microstructure of the glass substrate and SnOx film deposited by magnetron sputtering and sol-gel technique. The film deposited by magnetron sputtering, consists of large agglomerates, and the film synthesized by sol-gel technique has a fine-grained structure of the surface.

0 20 40 60 80 100 t, s

Fig. 13. The dependence of the response time of SnO2 layer on ethanol vapors (concentration

0 0,5 1 1,5 С, mg/L

Fig. 14. The dependence of sensitivity of SnO2 films to the concentration of ethanol vapors:

Fig. 15 shows the atomic-force microscope images of the surface microstructure of the glass substrate and SnOx film deposited by magnetron sputtering and sol-gel technique. The film deposited by magnetron sputtering, consists of large agglomerates, and the film synthesized

1 − film deposited by magnetron sputtering, 2 − by sol-gel technique.

by sol-gel technique has a fine-grained structure of the surface.

2 1

1

2

1 mg/L): 1 − film deposited by magnetron sputtering, 2 − by sol-gel technique.

10

0

20

40

ΔR/R0, %

60

80

15

20 25

R, kOh

м

30

35

Fig. 15. Surface topography of the SnOx films on glass substrates: a) glass, b) SnOx film, deposited by magnetron sputtering, c) by sol-gel technique.

Table 4 presents data on the film surface structure obtained using the processing software to an atomic force microscope, where: Ra is average roughness; Rzjis − average roughness of 10 points; S − area of the image; Rq − RMS roughness; Rz − the difference between maximum and minimum height of an image; Sratio − the ratio of the image area S to the area of flat surface S0.


Table 4. Parameters of the surface structure of SnO2 films on glass

#### **3.3 Influences of thermal and plasma treatments on phase composition, microstructure and physical properties of the SnOx films deposited by sol-gel technique and magnetron sputtering**

To modify the properties of SnO2 films, different ways of their treatment are widely used. Of particular interest is the study on the influence of thermal and plasma treatments on optical and structural properties of the SnOx films, deposited by the sol-gel techniques.

As is known, the crystallization of SnO2 films on glass substrate takes place most intense at temperatures of 400−600°C. Heat treatment for 1 hour at temperatures of 550−600°C on energy costs can be equivalent to several hours of treatment at 400°C. However, the changes of the structural and sensory properties of the films may be non-equivalent in these conditions. Analysis of the influence of the duration of isothermal annealing on the properties of the films is of a separate scientific and practical interest.

In recent years, the plasma treatment for modifying the properties of SnO2 films is used (Karapatnitski et al., 2000). It has been reported about the influence of plasma on the sensitivity of the sensors. An analysis of changes in the optical parameters determined from

Influence of Crystallization on the Properties of SnO2 Thin Films 239

minutes (Fig. 17b) and 20 minutes (Fig. 17c) the intensities of SnO2 lines are greatly reduced. This may occur due to destruction of crystallites of tin oxide under the influence of oxygen plasma to form a cluster structure of the film. Clustering is not accompanied by appreciable sputtering of the film, since after annealing at 550ºС for 1 h the structure of SnO2 crystallites is restored and the intensity of lines on X-ray Debye patterns almost restored (Fig. 17d). Table 5 shows the optical parameters of the films, calculated by standard methods (Song, 1999). It is seen that the thickness of the film after oxygen plasma treatment decreases

If the deposition of SnOx film is carried out in condition of a lack of oxygen at reduced pressure ~ 1 Pa in a Ar-O2 mixture in the chamber, an increase of the concentration of excess tin in the film results the formation of tin (β-Sn) crystallites with an average size of ~ 30 nm (Fig. 18a) during treatment by hydrogen plasma. So, that demonstrates a segregating effect by hydrogen plasma on the film structure. After annealing of these films in air at 550°C for 1 h, both as-treated and not treated by H-plasma, the appearance of X-ray lines of SnO2 (6 lines), Sn2O3 (5) and SnO (1) phases (Fig. 18b), is observed. Crystallite sizes are in the range of 8–18 nm. Processing of annealed samples by hydrogen plasma led to a blurring of line sections in the angle range of 15º < θ < 20º (in the region between arrows 1 in Figs. 18c and 19), corresponding to reflection from the plane systems of SnO(101), SnO2 (101) and (200), Sn2O3 (021) and (030) . Since the X-ray lines on the Debye photograph are a consequence of a number of specular reflections from the set of crystallites which are in reflecting position in accordance with the equation of Bragg 2 sin *d* ⋅ θ=λ , the blurring of line sections can occur when selective destruction (amorphization and clusterization) only those crystallites which are oriented to reflect in the area of blurring (between arrows 1). This assumes such orientation of SnOx film towards the movement of the plasma particles, that the symmetry and arrangement of atoms in the systems of planes of the crystallites SnO(101), SnO2(101) and (200), Sn2O3 (021) and (030) (plane A in Fig. 19) are disordered. The estimation of the angles α between the projection onto a plane perpendicular to the incident X-ray beam, of normals to the sample surface and to the planes A of crystallites which were in the reflecting position before clustering by hydrogen plasma treatment. Angle α for the plane systems SnO(101), SnO2(101) and (200), Sn2O3 (021) and (030) lies in the range of 15–45º, and the

slightly.

angle θ in the range of 15–20º.

n, refractive index

D, film thickness, nm

sputtering 1.830 280 4.05 1.65.

O-plasma 1.750 276 4.05 1.68.

H- plasma 1.805 294 4.05 2.25.

Eg., band gap, eV

Table 5. The parameters of SnOх films deposited by magnetron sputtering at a pressure of Ar-O2 mixture inside the camera − 2.7 Pa, after deposition and treatment by a glow

k, absorption coefficient, 1/cm

V, porosity, %

103 15.5 5.89

103 21.0 5.45

103 17.0 5.75

ρ, density, g/cm3

Sample

Magnetron

discharge plasma

transmission spectra in combination with changes of structural characteristics on the basis of X-ray diffraction and atomic force microscopy, may help to better understand the dynamics of the physical and structural properties of thin films of tin dioxide.

This section presents the results of the study of the effect of isothermal annealing (15 min, 3 hours, 6 hours and 12 hours at T = 400ºC) and processing by hydrogen and oxygen glow discharge plasma on the microstructure, optical and electrical properties, thickness, porosity and gas sensitivity of SnO2 films, deposited by the sol-gel technique (the method of centrifugation) on a glass substrate. The results are interpreted by comparing with recent data on the effect of plasma treatment on the properties of films obtained by magnetron sputtering or ion implantation.

#### **3.3.1 Influence of thermal and plasma treatment on phase composition, microstructure and physical properties of the SnOx films deposited by magnetron sputtering**

Fig. 16 shows the dependence of sensitivity on the concentration of ethanol vapor for thin SnO2 film, obtained by magnetron sputtering, after treatment by oxygen plasma (27.12 MHz, 12.5 W, 6.5 Pa, 100ºС, 5 or 20 min). The SnO2 film acquires a high sensitivity (> 50%) to ethanol vapor with concentrations below 0.2 mg/L after treatment by oxygen plasma for 5 and 20 min. The short-term treatment for 5 min was more effective for increasing the sensitivity of the SnO2 films (~70%). This significant effect of plasma treatment on the gassensitive properties of the SnO2 films, probably, is caused by changes in the structure of the films during processing.

Fig. 16. Dependence of the sensitivity on the concentration of ethanol vapor for SnO2 film: 1 − after deposition, 2 − after treatment by oxygen plasma for 5 minutes, 3 − for 20 minutes.

As shown in Fig. 17, immediately after the deposition by magnetron sputtering (cathode voltage is 470 V, discharge current − 35 mA, the pressure in a Ar-O2 mixture in the chamber − 2.7 Pa, oxygen concentration 10%, the rate of deposition of films ~ 0.05 nm/s, substrate temperature 200ºС) the films have a nanocrystalline structure and contain SnO2 crystallites with sizes about 4 nm. In this case there are clear reflections from the systems of planes of SnO2 with Miller indices (110), (101), (211) and (112). After oxygen plasma treatment for 5

transmission spectra in combination with changes of structural characteristics on the basis of X-ray diffraction and atomic force microscopy, may help to better understand the dynamics

This section presents the results of the study of the effect of isothermal annealing (15 min, 3 hours, 6 hours and 12 hours at T = 400ºC) and processing by hydrogen and oxygen glow discharge plasma on the microstructure, optical and electrical properties, thickness, porosity and gas sensitivity of SnO2 films, deposited by the sol-gel technique (the method of centrifugation) on a glass substrate. The results are interpreted by comparing with recent data on the effect of plasma treatment on the properties of films obtained by magnetron

of the physical and structural properties of thin films of tin dioxide.

**3.3.1 Influence of thermal and plasma treatment on phase composition,** 

**microstructure and physical properties of the SnOx films deposited by magnetron** 

Fig. 16 shows the dependence of sensitivity on the concentration of ethanol vapor for thin SnO2 film, obtained by magnetron sputtering, after treatment by oxygen plasma (27.12 MHz, 12.5 W, 6.5 Pa, 100ºС, 5 or 20 min). The SnO2 film acquires a high sensitivity (> 50%) to ethanol vapor with concentrations below 0.2 mg/L after treatment by oxygen plasma for 5 and 20 min. The short-term treatment for 5 min was more effective for increasing the sensitivity of the SnO2 films (~70%). This significant effect of plasma treatment on the gassensitive properties of the SnO2 films, probably, is caused by changes in the structure of the

> 0 0,5 1 1,5 C, mg/L

Fig. 16. Dependence of the sensitivity on the concentration of ethanol vapor for SnO2 film: 1 − after deposition, 2 − after treatment by oxygen plasma for 5 minutes, 3 − for 20 minutes.

As shown in Fig. 17, immediately after the deposition by magnetron sputtering (cathode voltage is 470 V, discharge current − 35 mA, the pressure in a Ar-O2 mixture in the chamber − 2.7 Pa, oxygen concentration 10%, the rate of deposition of films ~ 0.05 nm/s, substrate temperature 200ºС) the films have a nanocrystalline structure and contain SnO2 crystallites with sizes about 4 nm. In this case there are clear reflections from the systems of planes of SnO2 with Miller indices (110), (101), (211) and (112). After oxygen plasma treatment for 5

sputtering or ion implantation.

films during processing.

1

2

3

R/R, %

ΔR/R, %

**sputtering** 

minutes (Fig. 17b) and 20 minutes (Fig. 17c) the intensities of SnO2 lines are greatly reduced. This may occur due to destruction of crystallites of tin oxide under the influence of oxygen plasma to form a cluster structure of the film. Clustering is not accompanied by appreciable sputtering of the film, since after annealing at 550ºС for 1 h the structure of SnO2 crystallites is restored and the intensity of lines on X-ray Debye patterns almost restored (Fig. 17d). Table 5 shows the optical parameters of the films, calculated by standard methods (Song, 1999). It is seen that the thickness of the film after oxygen plasma treatment decreases slightly.

If the deposition of SnOx film is carried out in condition of a lack of oxygen at reduced pressure ~ 1 Pa in a Ar-O2 mixture in the chamber, an increase of the concentration of excess tin in the film results the formation of tin (β-Sn) crystallites with an average size of ~ 30 nm (Fig. 18a) during treatment by hydrogen plasma. So, that demonstrates a segregating effect by hydrogen plasma on the film structure. After annealing of these films in air at 550°C for 1 h, both as-treated and not treated by H-plasma, the appearance of X-ray lines of SnO2 (6 lines), Sn2O3 (5) and SnO (1) phases (Fig. 18b), is observed. Crystallite sizes are in the range of 8–18 nm. Processing of annealed samples by hydrogen plasma led to a blurring of line sections in the angle range of 15º < θ < 20º (in the region between arrows 1 in Figs. 18c and 19), corresponding to reflection from the plane systems of SnO(101), SnO2 (101) and (200), Sn2O3 (021) and (030) . Since the X-ray lines on the Debye photograph are a consequence of a number of specular reflections from the set of crystallites which are in reflecting position in accordance with the equation of Bragg 2 sin *d* ⋅ θ=λ , the blurring of line sections can occur when selective destruction (amorphization and clusterization) only those crystallites which are oriented to reflect in the area of blurring (between arrows 1). This assumes such orientation of SnOx film towards the movement of the plasma particles, that the symmetry and arrangement of atoms in the systems of planes of the crystallites SnO(101), SnO2(101) and (200), Sn2O3 (021) and (030) (plane A in Fig. 19) are disordered. The estimation of the angles α between the projection onto a plane perpendicular to the incident X-ray beam, of normals to the sample surface and to the planes A of crystallites which were in the reflecting position before clustering by hydrogen plasma treatment. Angle α for the plane systems SnO(101), SnO2(101) and (200), Sn2O3 (021) and (030) lies in the range of 15–45º, and the angle θ in the range of 15–20º.


Table 5. The parameters of SnOх films deposited by magnetron sputtering at a pressure of Ar-O2 mixture inside the camera − 2.7 Pa, after deposition and treatment by a glow discharge plasma

Influence of Crystallization on the Properties of SnO2 Thin Films 241

to more intense oxidation processes in the plasma-treated films in the annealing process due to their higher porosity (11.8 instead of 7.7 according to Table 6). As seen from Fig. 20 (curve 6), the processing by H-plasma of the annealed polycrystalline films, leads to a marked deterioration in transmittance in the wavelength range 300−1100 nm. This may be due to disordering of the structure and the formation of crystal-amorphous structure containing

> +Н2 +Н<sup>2</sup> SnO2 SnO Sn -Н2О -Н2О

Fig. 18. X-ray diffraction patterns and intensity curves for the SnOx films after deposition by

Further annealing at 550°C (1 h) permits to maximize the transparency of the films produced under low pressure in Ar-O2 mixture in the chamber (~ 1 Pa). This is due to the restoration of the structure of microregions, disordered during plasma treatment. It is assumed that the transformation of these areas during the oxidation into SnO2 crystallites with a higher density than Sn3O4, increases the porosity of the films (Table 6), which is important for the

The obtained data allows us to interpret the transmission spectra of SnOx films, obtained by magnetron sputtering under higher pressure in Ar-O2 mixture in the chamber ~ 2.7 Pa (Fig. 21). X-ray data show the presence of polycrystalline SnO2 phase immediately after the deposition and the absence of β-Sn crystallites (Fig. 17a). As is seen in the figure 21 (curves 3 and 4), a significant drop in *Т*(λ) in range of 1200−2500 nm after processing of the SnOx films

magnetron sputtering (pressure of 1 Pa in Ar-O2 mixture) and H-plasma treatment (a), deposition and annealing at 550°С for 1 h (b), the subsequent processing of H-plasma

(1)

opaque inclusions of SnO, in accordance with (1):

(c) and re-annealing at 550°С (d).

gas sensitivity of films.

Fig. 17. X-ray diffraction patterns and intensity curves for the thin SnO2 film after deposition by magnetron sputtering (pressure 2.7 Pa in Ar-O2 mixture) on a glass substrate (a), after treatment by a glow discharge oxygen plasma for 5 min (b), 20 min (c) and annealing at 550ºС for 1 h (d).

In other directions, areas of X-ray lines not subjected to blur (in the region between the arrows 2 in Figs. 18c and 19), the crystallites were solid and even increased their size, demonstrating segregate effects of hydrogen plasma. The order of arrangement of atoms in the systems of planes of SnO(101), SnO2 (101) and (200), Sn2O3 in these crystallites (plane B in Fig. 19) is intact. Angle α for these planes lies in the range of 15–(-45)º, and the angle θ in the same range of 15–20º. Therefore, the possibility of obtaining by treatment in glow discharge hydrogen plasma of crystal-amorphous nanostructures in which high-quality nanocrystals of tin oxide alternate with nanosized clusters of tin oxides, is shown.

Repeated annealing for 1 h at 550°С (Fig. 20, curve 7) causes the restoration of the integrity of the X-ray lines of SnO(101), SnO2(101) and (200), Sn2O3 (021) and (130) (Fig. 18d).

It follows from the optical transmission spectra (Fig. 20, curve 2), low pressure of Ar-O2 mixture in the chamber (~ 1 Pa), which leads to an increase in the concentration of excess tin in the film, leads to a significant deterioration of its transparency. The presence of significant amounts of tin clusters after deposition is probably the reason for the lack of transparency of the films at low wavelengths (<500 nm) and low transparency in the region above 500 nm, as β-Sn crystallites yet not observed on X-ray patterns. The formation of tin (β-Sn) crystallites with an average size of ~ 30 nm (Fig. 18a) after treatment by hydrogen plasma leads to a further reduction in the transparency of the film (Fig. 20, curve 3) over the entire considered range of wavelengths (500−1100 nm). The decomposition and oxidation of tin crystallites at the subsequent annealing in air at 550°C for 1 h (Fig. 18b) leads to significant increase of film transparency (Fig. 20, curve 4) both in range of > 500 nm, and at small length waves ~ 350−500 nm (*T*(λ) from 0 to ~ 80%), for which the film was not transparent. This may be caused not only by oxidation of the β-Sn crystallites, but also tin clusters. At the same time the transparency of these films is higher than the transparency of untreated by Hplasma films after annealing under the same conditions (Fig. 20, curve 5). This may be due

Fig. 17. X-ray diffraction patterns and intensity curves for the thin SnO2 film after deposition by magnetron sputtering (pressure 2.7 Pa in Ar-O2 mixture) on a glass substrate (a), after treatment by a glow discharge oxygen plasma for 5 min (b), 20 min (c) and annealing at

In other directions, areas of X-ray lines not subjected to blur (in the region between the arrows 2 in Figs. 18c and 19), the crystallites were solid and even increased their size, demonstrating segregate effects of hydrogen plasma. The order of arrangement of atoms in the systems of planes of SnO(101), SnO2 (101) and (200), Sn2O3 in these crystallites (plane B in Fig. 19) is intact. Angle α for these planes lies in the range of 15–(-45)º, and the angle θ in the same range of 15–20º. Therefore, the possibility of obtaining by treatment in glow discharge hydrogen plasma of crystal-amorphous nanostructures in which high-quality

Repeated annealing for 1 h at 550°С (Fig. 20, curve 7) causes the restoration of the integrity

It follows from the optical transmission spectra (Fig. 20, curve 2), low pressure of Ar-O2 mixture in the chamber (~ 1 Pa), which leads to an increase in the concentration of excess tin in the film, leads to a significant deterioration of its transparency. The presence of significant amounts of tin clusters after deposition is probably the reason for the lack of transparency of the films at low wavelengths (<500 nm) and low transparency in the region above 500 nm, as β-Sn crystallites yet not observed on X-ray patterns. The formation of tin (β-Sn) crystallites with an average size of ~ 30 nm (Fig. 18a) after treatment by hydrogen plasma leads to a further reduction in the transparency of the film (Fig. 20, curve 3) over the entire considered range of wavelengths (500−1100 nm). The decomposition and oxidation of tin crystallites at the subsequent annealing in air at 550°C for 1 h (Fig. 18b) leads to significant increase of film transparency (Fig. 20, curve 4) both in range of > 500 nm, and at small length waves ~ 350−500 nm (*T*(λ) from 0 to ~ 80%), for which the film was not transparent. This may be caused not only by oxidation of the β-Sn crystallites, but also tin clusters. At the same time the transparency of these films is higher than the transparency of untreated by Hplasma films after annealing under the same conditions (Fig. 20, curve 5). This may be due

nanocrystals of tin oxide alternate with nanosized clusters of tin oxides, is shown.

of the X-ray lines of SnO(101), SnO2(101) and (200), Sn2O3 (021) and (130) (Fig. 18d).

550ºС for 1 h (d).

to more intense oxidation processes in the plasma-treated films in the annealing process due to their higher porosity (11.8 instead of 7.7 according to Table 6). As seen from Fig. 20 (curve 6), the processing by H-plasma of the annealed polycrystalline films, leads to a marked deterioration in transmittance in the wavelength range 300−1100 nm. This may be due to disordering of the structure and the formation of crystal-amorphous structure containing opaque inclusions of SnO, in accordance with (1):

$$\begin{aligned} \text{SnO}\_2 \xrightarrow[\text{H}\_2\text{O}]{\text{+H}\_2} \text{SnO} \xrightarrow[\text{H}\_2\text{O}]{\text{+H}\_2} \text{Sn} \end{aligned} \tag{1}$$

Fig. 18. X-ray diffraction patterns and intensity curves for the SnOx films after deposition by magnetron sputtering (pressure of 1 Pa in Ar-O2 mixture) and H-plasma treatment (a), deposition and annealing at 550°С for 1 h (b), the subsequent processing of H-plasma (c) and re-annealing at 550°С (d).

Further annealing at 550°C (1 h) permits to maximize the transparency of the films produced under low pressure in Ar-O2 mixture in the chamber (~ 1 Pa). This is due to the restoration of the structure of microregions, disordered during plasma treatment. It is assumed that the transformation of these areas during the oxidation into SnO2 crystallites with a higher density than Sn3O4, increases the porosity of the films (Table 6), which is important for the gas sensitivity of films.

The obtained data allows us to interpret the transmission spectra of SnOx films, obtained by magnetron sputtering under higher pressure in Ar-O2 mixture in the chamber ~ 2.7 Pa (Fig. 21). X-ray data show the presence of polycrystalline SnO2 phase immediately after the deposition and the absence of β-Sn crystallites (Fig. 17a). As is seen in the figure 21 (curves 3 and 4), a significant drop in *Т*(λ) in range of 1200−2500 nm after processing of the SnOx films

Influence of Crystallization on the Properties of SnO2 Thin Films 243

Thus established that the short-term treatment (5 min) by O- or H-plasma on the deposited at a pressure of 2.7 Pa SnO2 film results in the formation of Sn clusters, which reduces the transparency of the film from 80% to 50% and 40%, respectively, in the near infrared region

**5** 

Fig. 20. Optical transmission spectra of glass substrate and thin SnOx films, deposited by magnetron sputtering (pressure 1 Pa of Ar-O2 mixture) and treated by H-plasma and annealing: 1 is a spectrum of substrate (glass), 2 – after deposition of SnOx film on glass, 3 – deposition + H-plasma; 4 – deposition + H-plasma + annealing at 550°С (1 h); 5 – deposition + annealing at 550°С; 6 – deposition + annealing at 550°C + H-plasma;

> D, film thickness, nm

Annealing 1.94 330 4.00 6.42 7.70

annealing 1.88 341 4.01 6.13 11.8

Table 6. Optical parameters of SnO*х* film deposited by magnetron sputtering at a pressure of

Eg., band gap, eV

1.82 296 4.05 5.82 16.2

ρ, density, g/cm3

V, porosity, %

7 – deposition + annealing at 550°C + H-plasma + annealing at 550°C.

n, refractive index

(1200-2500 nm).

Sequence of operations

Plasma -

Annealing plasma annealing

Ar-O2 mixture in the chamber − 1 Pa

for 5 min by O-plasma or H-plasma is observed. The decrease of the transmission of these films, consisting mainly of SnO2 crystallites (Fig. 17a), can occur as a result of increase of the concentration of free charge carriers due to segregation of excess atoms of tin and the formation of tin clusters with sizes of tenths of a nanometer. It should be noted that the phenomenon of reducing of the transparency due to the presence of nanoparticles in the films is less pronounced in the case of oxygen plasma treatment, because the process of segregation of tin nanoparticles is accompanied by a process of oxidation of some part of them up to SnO2.

Fig. 19. Illustration of the blurring of X-ray line sections (using SnO2(101) and SnO(101) as an example) in the angle ranges 15º < θ < 20º and 15º < α < 45º (in the region between arrows 1) after treatment of the annealed samples by H-plasma. The X-ray beam paths as follows: X1X2 is the X-ray beam incident at an angle of ψ = 5º to the sample surface, Х2Х3 is the direction of the transmitted beam, Х2Х4 is the X-ray beam reflected from the SnO2 (101) plane system of crystallite B (Plane B), Х2Х5 is the beam reflected from the plane system SnO2 (101) of crystallite A before its destruction by plasma treatment (Plane A). О1О2 is the straight line lying in the A plane and making an angle θ with Х2Х3 and Х2Х4. α is the angle between the projections onto the plane (perpendicular to the incident X-ray beam) of the normals to the sample surface and to the A plane.

for 5 min by O-plasma or H-plasma is observed. The decrease of the transmission of these films, consisting mainly of SnO2 crystallites (Fig. 17a), can occur as a result of increase of the concentration of free charge carriers due to segregation of excess atoms of tin and the formation of tin clusters with sizes of tenths of a nanometer. It should be noted that the phenomenon of reducing of the transparency due to the presence of nanoparticles in the films is less pronounced in the case of oxygen plasma treatment, because the process of segregation of tin nanoparticles is accompanied by a process of oxidation of some part of

Fig. 19. Illustration of the blurring of X-ray line sections (using SnO2(101) and SnO(101) as an example) in the angle ranges 15º < θ < 20º and 15º < α < 45º (in the region between arrows 1) after treatment of the annealed samples by H-plasma. The X-ray beam paths as follows: X1X2 is the X-ray beam incident at an angle of ψ = 5º to the sample surface, Х2Х3 is the direction of the transmitted beam, Х2Х4 is the X-ray beam reflected from the SnO2 (101) plane system of crystallite B (Plane B), Х2Х5 is the beam reflected from the plane system SnO2 (101) of crystallite A before its destruction by plasma treatment (Plane A). О1О2 is the straight line lying in the A plane and making an angle θ with Х2Х3 and Х2Х4. α is the angle between the projections onto the plane (perpendicular to the incident X-ray beam) of the normals to the

them up to SnO2.

sample surface and to the A plane.

Thus established that the short-term treatment (5 min) by O- or H-plasma on the deposited at a pressure of 2.7 Pa SnO2 film results in the formation of Sn clusters, which reduces the transparency of the film from 80% to 50% and 40%, respectively, in the near infrared region (1200-2500 nm).

Fig. 20. Optical transmission spectra of glass substrate and thin SnOx films, deposited by magnetron sputtering (pressure 1 Pa of Ar-O2 mixture) and treated by H-plasma and annealing: 1 is a spectrum of substrate (glass), 2 – after deposition of SnOx film on glass, 3 – deposition + H-plasma; 4 – deposition + H-plasma + annealing at 550°С (1 h); 5 – deposition + annealing at 550°С; 6 – deposition + annealing at 550°C + H-plasma; 7 – deposition + annealing at 550°C + H-plasma + annealing at 550°C.


Table 6. Optical parameters of SnO*х* film deposited by magnetron sputtering at a pressure of Ar-O2 mixture in the chamber − 1 Pa

Influence of Crystallization on the Properties of SnO2 Thin Films 245

treatment resulted in a slight decrease of 1−5% (Fig. 23, curve 3). Noticeable decrease in the transparency of the thin film after hydrogen plasma treatment can be attributed to the formation of opaque compounds such as SnO or Sn, formed by reducing properties of

However, X-ray diffraction data do not confirm the presence of crystallites of SnO and Sn in the films treated by hydrogen plasma (Table 7). Nevertheless, the presence of SnO inclusions in the amorphous state or in the form of clusters is possible that leads to an increase in the absorption coefficient from 2.5103 up to 5.91103 cm-1 (Table 7). The presence of metal clusters of tin in appreciable amounts is unlikely, since there is no reduction of the transmission coefficient Т(λ) in the range 1200-2500 nm, associated with an increase in the

The treatment by oxygen plasma does not result the formation of SnO, and the slight change in the transparency of the film is caused by the damaging effects of plasma on the structural perfection and crystallite size. Reducing the size of the SnO2 crystallites is caused by the destructive influence of the massive oxygen ions on the structure of the crystallites (Table 7). The increased density of the film in proportion to the reduction of its thickness is presumably caused by the filling of intergranular voids by SnO2 clusters. The increase of the SnO2 crystallites sizes during processing by hydrogen plasma is caused by the segregating effect of light hydrogen ions of plasma, under the influence of which the alteration, destruction, and merging of the crystallites are taken place. Reducing the thickness by 10% while maintaining the density of the film can be occur through the formation and desorption of H2O molecules and the formation of oxygen vacancies. The formation of SnO molecules on the surface of SnO2 crystallites is taken place; thereby the stoichiometry and transparency

> 0 500 1000 1500 2000 2500 λ, nm

Fig. 22. Optical transmission spectra of SnO2 films after isothermal annealing at 400ºC:

1 − glass substrate, 2 − the film after deposition and annealing for 15 minutes, 3 − 3 hours of annealing; 4 − 6 hours of annealing, 5 − 12 hours of annealing.

hydrogen in accordance with the reaction (1).

concentration of free charge carriers.

of the film are reduced.

70

75

80

85

Т, %

90

**1**

**5**

**2 3 4**

95

Fig. 21. Optical transmission spectra: the glass substrate (1); thin SnOx films on a glass substrate after deposition by magnetron sputtering (2) and treatment by oxygen (3) and hydrogen (4) plasma.

#### **3.3.2 Influence of isothermal annealing and plasma treatment on phase composition, microstructure and physical properties of the SnOx films deposited by sol-gel technique**

Colloidal solution for preparation of tin dioxide films was deposited on glass substrate. Rotational speed of the centrifuge was 3800 rpm. Centrifugation time was 3−5 s. The substrates with the deposited films were dried by an infrared emitter at 80°C for 3−5 min. Low temperature annealing was maintained to prevent the occurrence of cracks on the SnO2 films. Then the sample is placed in a muffle furnace and dried at 400°C for 15 min. The number of deposited layers of SnO2 was 15. The thickness of the deposited film was estimated from the weight of the film and was about 300 nm.

As shown earlier (Fig. 11b), X-ray diffraction study of SnO2 films after drying at 400ºC for 15 minutes showed that the films have a polycrystalline structure. The structure of SnO2 crystallites is sufficiently advanced, it contributed to recording X-ray reflections from 6 plane systems with Miller indices SnO2 (110), (101), (200), (211), (220), (112). The increase of the duration of isothermal annealing leads to an increase in the average size of SnO2 crystallites in the films: 15 minutes − 6 nm, 3 and 6 hours − little more than 6 nm, 12 hours − more than 10 nm.

Fig. 22 shows the spectra of optical transmission of tin dioxide films after annealing at 400ºС for 15 min, 3, 6 and 12 hours. Fig. 23 shows the spectra of optical transmission after treatment by oxygen and hydrogen plasma.

The prepared films have high transparency (~90%). With increasing of annealing time the transparency at short wavelengths is increased, which indicates the improvement of the stoichiometry of tin dioxide films and the removal of residual solvent (Fig. 22). As shown in Table 7, the increase of isothermal annealing time to 6 and 12 h leads to compaction of the film, reducing its thickness, porosity and absorption coefficient.

The treatment by glow discharge hydrogen plasma resulted in a decrease in the transparency on 3−15% in the short- wave range (Fig. 23, curve 4), while the oxygen plasma

1

4

0 500 1000 1500 2000 2500 λ, nm

Fig. 21. Optical transmission spectra: the glass substrate (1); thin SnOx films on a glass substrate after deposition by magnetron sputtering (2) and treatment by oxygen (3) and

**3.3.2 Influence of isothermal annealing and plasma treatment on phase composition, microstructure and physical properties of the SnOx films deposited by sol-gel** 

Colloidal solution for preparation of tin dioxide films was deposited on glass substrate. Rotational speed of the centrifuge was 3800 rpm. Centrifugation time was 3−5 s. The substrates with the deposited films were dried by an infrared emitter at 80°C for 3−5 min. Low temperature annealing was maintained to prevent the occurrence of cracks on the SnO2 films. Then the sample is placed in a muffle furnace and dried at 400°C for 15 min. The number of deposited layers of SnO2 was 15. The thickness of the deposited film was

As shown earlier (Fig. 11b), X-ray diffraction study of SnO2 films after drying at 400ºC for 15 minutes showed that the films have a polycrystalline structure. The structure of SnO2 crystallites is sufficiently advanced, it contributed to recording X-ray reflections from 6 plane systems with Miller indices SnO2 (110), (101), (200), (211), (220), (112). The increase of the duration of isothermal annealing leads to an increase in the average size of SnO2 crystallites in the films: 15 minutes − 6 nm, 3 and 6 hours − little more than 6 nm, 12 hours −

Fig. 22 shows the spectra of optical transmission of tin dioxide films after annealing at 400ºС for 15 min, 3, 6 and 12 hours. Fig. 23 shows the spectra of optical transmission after

The prepared films have high transparency (~90%). With increasing of annealing time the transparency at short wavelengths is increased, which indicates the improvement of the stoichiometry of tin dioxide films and the removal of residual solvent (Fig. 22). As shown in Table 7, the increase of isothermal annealing time to 6 and 12 h leads to compaction of the

The treatment by glow discharge hydrogen plasma resulted in a decrease in the transparency on 3−15% in the short- wave range (Fig. 23, curve 4), while the oxygen plasma

estimated from the weight of the film and was about 300 nm.

film, reducing its thickness, porosity and absorption coefficient.

3

2

0

20 40

60

T, %

hydrogen (4) plasma.

**technique** 

more than 10 nm.

treatment by oxygen and hydrogen plasma.

80

100

treatment resulted in a slight decrease of 1−5% (Fig. 23, curve 3). Noticeable decrease in the transparency of the thin film after hydrogen plasma treatment can be attributed to the formation of opaque compounds such as SnO or Sn, formed by reducing properties of hydrogen in accordance with the reaction (1).

However, X-ray diffraction data do not confirm the presence of crystallites of SnO and Sn in the films treated by hydrogen plasma (Table 7). Nevertheless, the presence of SnO inclusions in the amorphous state or in the form of clusters is possible that leads to an increase in the absorption coefficient from 2.5103 up to 5.91103 cm-1 (Table 7). The presence of metal clusters of tin in appreciable amounts is unlikely, since there is no reduction of the transmission coefficient Т(λ) in the range 1200-2500 nm, associated with an increase in the concentration of free charge carriers.

The treatment by oxygen plasma does not result the formation of SnO, and the slight change in the transparency of the film is caused by the damaging effects of plasma on the structural perfection and crystallite size. Reducing the size of the SnO2 crystallites is caused by the destructive influence of the massive oxygen ions on the structure of the crystallites (Table 7). The increased density of the film in proportion to the reduction of its thickness is presumably caused by the filling of intergranular voids by SnO2 clusters. The increase of the SnO2 crystallites sizes during processing by hydrogen plasma is caused by the segregating effect of light hydrogen ions of plasma, under the influence of which the alteration, destruction, and merging of the crystallites are taken place. Reducing the thickness by 10% while maintaining the density of the film can be occur through the formation and desorption of H2O molecules and the formation of oxygen vacancies. The formation of SnO molecules on the surface of SnO2 crystallites is taken place; thereby the stoichiometry and transparency of the film are reduced.

Fig. 22. Optical transmission spectra of SnO2 films after isothermal annealing at 400ºC: 1 − glass substrate, 2 − the film after deposition and annealing for 15 minutes, 3 − 3 hours of annealing; 4 − 6 hours of annealing, 5 − 12 hours of annealing.

Influence of Crystallization on the Properties of SnO2 Thin Films 247

0 3 6 9 12 Annealing time, hour

Fig. 24. The dependence of sheet resistance (22ºC) of the SnO2 film on the duration of

Fig. 25 shows the temperature dependence of the resistance of the films after isothermal annealing, and Fig. 26 - after treatment by hydrogen and oxygen plasmas. The substrate with the film placed on a heated table, located in a cylindrical chamber. The resistance of the deposited films decreases rapidly with increasing temperature (Figs. 25 and 26, curves 1). Increase of the duration of isothermal annealing has little influence on the temperature

The processing of the film by hydrogen plasma leads to an increase in the number of oxygen vacancies and, correspondingly, a decrease in resistance from 192 kohm to 3.1, resulting in a decrease to 1.6 kohm of resistance with increasing temperature is smooth and the relative resistance decreases slowly (Fig. 26, curve 3). Treatment by oxygen plasma leads to occurrence of excess oxygen in the film and, accordingly, to increase resistance from 192 to 640 kohm, resulting in a decrease to 1.9 kohm of resistance with increasing temperature is

> 0 50 100 150 200 250 300 T, <sup>o</sup> C

Fig. 25. Temperature dependences of the resistance of the SnO2 film after annealing at 400ºC

for 15 minutes (1 - **•**), 3 hours (2 - **□**), 6 hours (3 - **∆**) and 12 hours (4 - **○**).

1

2

0

100

200

ρ, kOhm·cm

dependence of resistance (Fig. 25, curves 2−4).

faster due to the desorption of excess oxygen (Fig. 26, curve 2).

isothermal annealing at 400ºC.

0,0

0,2

0,4

3

4

0,6

R/Ro, arb.units .

0,8

1,0

300

Fig. 23. Optical transmission spectra of SnO2 films after plasma treatment: 1 − substrate, 2 − after film deposition, 3 − after treatment by oxygen plasma, 4 − after treatment by hydrogen plasma.


Table 7. Optical and structural parameters of SnO2 films

The surface resistance of SnO2 films after deposition and annealing at 400ºС for 15 minutes by four-probe method is measured, and it was 16 ohm cm (Fig. 24). The surface resistance of the SnO2 film increases linearly with increasing duration of isothermal annealing (400ºC). Resistance measurements were performed at room temperature (22ºC). It is seen that each hour of annealing leads to an increase in surface resistance of ~ 24 kOhm cm (Fig. 24). This may be due to a decrease in the number of oxygen vacancies as a result of improving of the crystallite structure and stoichiometry during the long process of annealing in air.

After treatment in oxygen plasma, the film sheet resistance increased from 16 to 133 kohm·cm, apparently due to reducing the number of oxygen vacancies. In contrast, after treatment by hydrogen plasma, it decreased from 16 to 0.7 kohm·cm, presumably due to increase in the number of oxygen vacancies.

3

0 500 1000 1500 2000 2500 λ, nm

> 400ºС, 6 h

SnO2 (110) 5 7.5 6 8 5 5.5 6.5 SnO2 (101) 6.5 7 6.5 10 6.5 6 9.5 SnO2 (211) 6 6 6 14.5 6 4.5 10

Isothermal annealing Plasma treatment

400ºС, 15 min

Oplasma

Hplasma

400ºС, 12 h

Fig. 23. Optical transmission spectra of SnO2 films after plasma treatment: 1 − substrate, 2 − after film deposition, 3 − after treatment by oxygen plasma, 4 − after treatment by

> 400ºС, 3 h

n, refractive index 1.72 1.76 1.81 1.81 1.74 1.81 1.73 D, film thickness, nm 319 312 304 304 316 297 288 Eg, band gap, eV 4.1 4.1 4.1 4.1 4.1 4.1 4.0 Ρ, density, g/cm3 5.29 5.53 5.76 5.76 5.40 5.80 5.36 k, absorption coefficient, 1/cm 2.4103 2.3103 1.6103 0.6103 2.5103 3.05103 5.91103 V, porosity, % 23.9 20.5 17.2 17.1 22.3 16.5 22.8

The surface resistance of SnO2 films after deposition and annealing at 400ºС for 15 minutes by four-probe method is measured, and it was 16 ohm cm (Fig. 24). The surface resistance of the SnO2 film increases linearly with increasing duration of isothermal annealing (400ºC). Resistance measurements were performed at room temperature (22ºC). It is seen that each hour of annealing leads to an increase in surface resistance of ~ 24 kOhm cm (Fig. 24). This may be due to a decrease in the number of oxygen vacancies as a result of improving of the

After treatment in oxygen plasma, the film sheet resistance increased from 16 to 133 kohm·cm, apparently due to reducing the number of oxygen vacancies. In contrast, after treatment by hydrogen plasma, it decreased from 16 to 0.7 kohm·cm, presumably due to

crystallite structure and stoichiometry during the long process of annealing in air.

4

400ºС, 15 min

Table 7. Optical and structural parameters of SnO2 films

increase in the number of oxygen vacancies.

2

1

hydrogen plasma.

Average size ε of crystallites, nm

parameters

Optical and structural

Т, %

Fig. 24. The dependence of sheet resistance (22ºC) of the SnO2 film on the duration of isothermal annealing at 400ºC.

Fig. 25 shows the temperature dependence of the resistance of the films after isothermal annealing, and Fig. 26 - after treatment by hydrogen and oxygen plasmas. The substrate with the film placed on a heated table, located in a cylindrical chamber. The resistance of the deposited films decreases rapidly with increasing temperature (Figs. 25 and 26, curves 1). Increase of the duration of isothermal annealing has little influence on the temperature dependence of resistance (Fig. 25, curves 2−4).

The processing of the film by hydrogen plasma leads to an increase in the number of oxygen vacancies and, correspondingly, a decrease in resistance from 192 kohm to 3.1, resulting in a decrease to 1.6 kohm of resistance with increasing temperature is smooth and the relative resistance decreases slowly (Fig. 26, curve 3). Treatment by oxygen plasma leads to occurrence of excess oxygen in the film and, accordingly, to increase resistance from 192 to 640 kohm, resulting in a decrease to 1.9 kohm of resistance with increasing temperature is faster due to the desorption of excess oxygen (Fig. 26, curve 2).

Fig. 25. Temperature dependences of the resistance of the SnO2 film after annealing at 400ºC for 15 minutes (1 - **•**), 3 hours (2 - **□**), 6 hours (3 - **∆**) and 12 hours (4 - **○**).

Influence of Crystallization on the Properties of SnO2 Thin Films 249

b)

3

Ro/R, отн. ед.

Fig. 27. Temperature dependence of sensitivity to the vapors of ethanol (1 mg/L) of the SnO2 films: a) after annealing at 400ºC for 15 minutes (1), 3 hours (2), 6 hours (3) and 12 hours (4); b) after annealing at 400ºC for 15 minutes (1) and treatment by oxygen (2) and

Response time, s 5 3 2 2 5 2 3 Recovery time, s 90 140 130 75 90 110 100

2

3 4

1

hydrogen (3) plasmas for 5 minutes.

0

2

4

6

Ro/R, arb. un.

5 min.

8

а)

10

The parameters of sensitivity of SnO2 films

Ro/R, arb. un.

а)

210 230 250 270 T, <sup>o</sup> C

1

Table 8. Response time and recovery time of SnO2 films

0 0,5 1 1,5 С, mg/L

4

2

3

> 210 230 250 270 T, <sup>o</sup> C

> > min

0 0,5 1 1,5 С, mg/L

1

3

2

Oplasma

Hplasma

2

1

Isothermal annealing (400ºС) Plasma treatment

b)

15 min 3 h 6 h 12 h 400ºС, 15

0

2

4

6

Ro/R, отн. ед.

Fig. 28. The dependence of sensitivity on the concentration of ethanol vapor of SnO2 film: a) after isothermal annealing at 400ºC for 15 min (1), 3 h (2), 6 h (3) and 12 hours (4); b) after annealing at 400ºC for 15 min (1) and treatment by oxygen (2) and hydrogen (3) plasmas for

AFM images of the surface of tin dioxide multilayer film of size 500 × 500 nm (Fig. 29), annealed at 400ºC for 15 minutes, 3, 6 and 12 hours, suggest a fine-grained structure of the

8

10

Thin films of SnO2, subjected to isothermal annealing and plasma treatment, were investigated for sensitivity to ethanol vapor. Gas sensitivity in this case defined as the ratio

0 *g R R* γ = , (2)

where R0 is resistance of gas-sensitive layer in the clean air, Rg is the resistance of the layer in the mixture of air with detectable gas.

Fig. 27 shows the temperature dependence of the sensitivity for films synthesized by sol-gel technique at concentration of ethanol 1 mg/L after isothermal annealing (a) and treatment by plasma (b). Maximum sensitivity of the sol-gel films to the vapors of ethanol takes place at a temperature of 235ºС (Fig. 27, curves 1). There is a significant increase in sensitivity of the films both after annealing and after treatment by plasmas. The maximum sensitivity of the films is observed after annealing at 400ºC for 6 hours. A significant decrease in sensitivity after annealing for 12 hours in comparison with annealing for 3 and 6 hours may be due to a significant increase in the crystallite size from ~ 6.5 to ~ 11 nm. As shown by Xu et al. (1991), optimal crystallite size for good gas sensitivity should be ~ 6 nm.

Fig. 26. Temperature dependences of the resistance of the SnO2 film after deposition (1) and glow discharge oxygen (2) and hydrogen (3) plasma treatment for 5 min.

After processing by plasma, the sensitivity increases without increase of the operating temperature. The observed increase in sensitivity can be caused by an increase in the size of pores and cracks during the bombardment by ions of oxygen or hydrogen. The sensitivity of the films increases linearly (Fig. 28), if the concentration of ethanol vapor increases from 0.05 to 0.8 mg/L and above, after which saturation occurs. The growth of sensitivity is more intense after annealing or plasma treatment. The most intensive growth was observed after annealing for 3−6 hours. A marked sensitivity of the film is evident even when the concentration of ethanol vapor is less than 0.05 mg/L. Response time of sensor to ethanol with concentration of 1 mg/L is decreased from 5 to 2 sec both with the increase of annealing time and when exposed to plasma (Table 8). The recovery time of resistance to the initial value increases from 90 up to 140 s, except the 12-hour annealing (75 s).

Thin films of SnO2, subjected to isothermal annealing and plasma treatment, were investigated for sensitivity to ethanol vapor. Gas sensitivity in this case defined as the ratio

> 0 *g*

γ = , (2)

*R R*

where R0 is resistance of gas-sensitive layer in the clean air, Rg is the resistance of the layer

Fig. 27 shows the temperature dependence of the sensitivity for films synthesized by sol-gel technique at concentration of ethanol 1 mg/L after isothermal annealing (a) and treatment by plasma (b). Maximum sensitivity of the sol-gel films to the vapors of ethanol takes place at a temperature of 235ºС (Fig. 27, curves 1). There is a significant increase in sensitivity of the films both after annealing and after treatment by plasmas. The maximum sensitivity of the films is observed after annealing at 400ºC for 6 hours. A significant decrease in sensitivity after annealing for 12 hours in comparison with annealing for 3 and 6 hours may be due to a significant increase in the crystallite size from ~ 6.5 to ~ 11 nm. As shown by Xu

> 0 100 200 300 T, o C

Fig. 26. Temperature dependences of the resistance of the SnO2 film after deposition (1) and

After processing by plasma, the sensitivity increases without increase of the operating temperature. The observed increase in sensitivity can be caused by an increase in the size of pores and cracks during the bombardment by ions of oxygen or hydrogen. The sensitivity of the films increases linearly (Fig. 28), if the concentration of ethanol vapor increases from 0.05 to 0.8 mg/L and above, after which saturation occurs. The growth of sensitivity is more intense after annealing or plasma treatment. The most intensive growth was observed after annealing for 3−6 hours. A marked sensitivity of the film is evident even when the concentration of ethanol vapor is less than 0.05 mg/L. Response time of sensor to ethanol with concentration of 1 mg/L is decreased from 5 to 2 sec both with the increase of annealing time and when exposed to plasma (Table 8). The recovery time of resistance to the

3

1

et al. (1991), optimal crystallite size for good gas sensitivity should be ~ 6 nm.

2

0,0

glow discharge oxygen (2) and hydrogen (3) plasma treatment for 5 min.

initial value increases from 90 up to 140 s, except the 12-hour annealing (75 s).

0,2

0,4

0,6

R/Ro, arb. units .

0,8

1,0

in the mixture of air with detectable gas.

Fig. 27. Temperature dependence of sensitivity to the vapors of ethanol (1 mg/L) of the SnO2 films: a) after annealing at 400ºC for 15 minutes (1), 3 hours (2), 6 hours (3) and 12 hours (4); b) after annealing at 400ºC for 15 minutes (1) and treatment by oxygen (2) and hydrogen (3) plasmas for 5 minutes.


Table 8. Response time and recovery time of SnO2 films

Fig. 28. The dependence of sensitivity on the concentration of ethanol vapor of SnO2 film: a) after isothermal annealing at 400ºC for 15 min (1), 3 h (2), 6 h (3) and 12 hours (4); b) after annealing at 400ºC for 15 min (1) and treatment by oxygen (2) and hydrogen (3) plasmas for 5 min.

AFM images of the surface of tin dioxide multilayer film of size 500 × 500 nm (Fig. 29), annealed at 400ºC for 15 minutes, 3, 6 and 12 hours, suggest a fine-grained structure of the

Influence of Crystallization on the Properties of SnO2 Thin Films 251

Table 9 presents data on the effect of isothermal annealing and treatment by plasma on the roughness of the SnO2 films, where Ra is average roughness; Rzjis - average roughness of 10 points; Rq - RMS roughness; Rz - the difference between the maximum and minimum height of an image. The roughness of the SnO2 films is increased both after isothermal annealing and after the plasma treatment. This increases the gas sensitivity of film (Figs. 27b and 28b, curves 2 and 3) because the growth of roughness increases the surface area of the film. However, after isothermal annealing roughness varies slightly, but sensitivity is increased

Sample Rа, nm Rzjis, nm Rz, nm Rq, nm Glass substrate 2.31 17.7 19.0 2.92 400°С, 15 min 0.66 5.67 7.02 0.84 400°С, 3 h 0.79 6.32 6.93 0.98 400°С, 6 h 0.74 6.49 7.14 0.93 400°С, 12 h 1.02 9.06 10.0 1.27 400°С, 15 min 0.66 5.67 7.02 0.84 O- plasma 1.15 12.3 13.6 1.58 H- plasma 2.04 16.5 19.7 2.61

The essential differences in the films deposited by magnetron sputtering and sol-gel technique, include the composition, structure and stoichiometry influencing the properties of the films after deposition as well as after treatment. The comparison of the effect of treatment by hydrogen or oxygen plasma on the individual properties of the SnO2 films,

Sensitivity to ethanol vapor of films prepared by sol-gel technique, exceeds the sensitivity of the films obtained by magnetron sputtering (Fig. 31, curves 4 and 1). The maximum sensitivity of the sol-gel films at temperature ~ 235ºC reaches up to 53%. Treatment by both H-and O-plasma (Fig. 31, curves 5 and 6) increases the sensitivity of the film up to 66% at

> 190 230 270 310 T, <sup>o</sup> C

Fig. 31. Temperature dependence of sensitivity to the vapors of ethanol (1 mg/L) of the SnOx films after deposition by magnetron sputtering (1), H-plasma treatment (2) and O-plasma (3); deposition by sol-gel method (4), H-plasma (5) and O-plasma (6) treatment.

1

5

6

4

2

Table 9. Analysis of the surface topography of SnO2 films (500 × 500 nm)

obtained by these methods, is of particular scientific and practical interest.

3

ΔR/R, %

more significantly.

film surface. It turned out that after annealing at 400ºС for 15 minutes about 80% of the areas of film surface lies at elevations of 1.8−3.7 nm, after annealing for 3 hours – in the range 2.4−4.9 nm, for 6 hours – in the range of 2.7−4.9 nm and 12 hours – in the range of 3.2–6.4 nm. After annealing for 3 and 6 hours on the film surface are observed pronounced granules (Fig. 29b, c). After annealing for 12 hours the surface contains a large number of projections.

Fig. 29. Surface topography of the films of tin dioxide of 500×500 nm after annealing at 400ºC for 15 min (a), 3 h (b), 6 h (c) and 12 h (d).

Treatment by oxygen plasma (Fig. 30b) leads to the disintegration of the granular structure of the film, confirming the assumption about the clustering of structure. Treatment by hydrogen plasma (Fig. 30c) leads to the formation of agglomerates of sizes up to 200 nm. The dimensions of the agglomerates is much greater than the dimensions of the SnO2 crystallites (~ 9 nm), ie the agglomerates are composed of SnO2 crystallites.

Fig. 30. Surface topography of SnO2 film after after deposition and annealing (400ºC, 15 min) (a) and treatment by O-plasma (b) and H-plasma (c) for 5 min.

film surface. It turned out that after annealing at 400ºС for 15 minutes about 80% of the areas of film surface lies at elevations of 1.8−3.7 nm, after annealing for 3 hours – in the range 2.4−4.9 nm, for 6 hours – in the range of 2.7−4.9 nm and 12 hours – in the range of 3.2–6.4 nm. After annealing for 3 and 6 hours on the film surface are observed pronounced granules (Fig. 29b, c). After annealing for 12 hours the surface contains a large number of projections.

а) б)

a) b)

в) г)

Fig. 29. Surface topography of the films of tin dioxide of 500×500 nm after annealing at

crystallites (~ 9 nm), ie the agglomerates are composed of SnO2 crystallites.

(а) (b) (c)

(a) and treatment by O-plasma (b) and H-plasma (c) for 5 min.

Treatment by oxygen plasma (Fig. 30b) leads to the disintegration of the granular structure of the film, confirming the assumption about the clustering of structure. Treatment by hydrogen plasma (Fig. 30c) leads to the formation of agglomerates of sizes up to 200 nm. The dimensions of the agglomerates is much greater than the dimensions of the SnO2

Fig. 30. Surface topography of SnO2 film after after deposition and annealing (400ºC, 15 min)

c) d)

400ºC for 15 min (a), 3 h (b), 6 h (c) and 12 h (d).

Table 9 presents data on the effect of isothermal annealing and treatment by plasma on the roughness of the SnO2 films, where Ra is average roughness; Rzjis - average roughness of 10 points; Rq - RMS roughness; Rz - the difference between the maximum and minimum height of an image. The roughness of the SnO2 films is increased both after isothermal annealing and after the plasma treatment. This increases the gas sensitivity of film (Figs. 27b and 28b, curves 2 and 3) because the growth of roughness increases the surface area of the film. However, after isothermal annealing roughness varies slightly, but sensitivity is increased more significantly.


Table 9. Analysis of the surface topography of SnO2 films (500 × 500 nm)

The essential differences in the films deposited by magnetron sputtering and sol-gel technique, include the composition, structure and stoichiometry influencing the properties of the films after deposition as well as after treatment. The comparison of the effect of treatment by hydrogen or oxygen plasma on the individual properties of the SnO2 films, obtained by these methods, is of particular scientific and practical interest.

Sensitivity to ethanol vapor of films prepared by sol-gel technique, exceeds the sensitivity of the films obtained by magnetron sputtering (Fig. 31, curves 4 and 1). The maximum sensitivity of the sol-gel films at temperature ~ 235ºC reaches up to 53%. Treatment by both H-and O-plasma (Fig. 31, curves 5 and 6) increases the sensitivity of the film up to 66% at

Fig. 31. Temperature dependence of sensitivity to the vapors of ethanol (1 mg/L) of the SnOx films after deposition by magnetron sputtering (1), H-plasma treatment (2) and O-plasma (3); deposition by sol-gel method (4), H-plasma (5) and O-plasma (6) treatment.

Influence of Crystallization on the Properties of SnO2 Thin Films 253

3. The lower resistance at room temperature of films deposited by magnetron sputtering is caused by the presence of tin particles, and a smoother decrease of resistance with increasing temperature is caused by the inclusions of SnO. Films synthesized by the sol-gel technique are composed entirely of SnO2 phase, have optimal crystallite sizes (3-5 nm) and high values of porosity for higher gas sensitivity. There is no decrease in transmittance at long wavelengths due to absence of tin particles. After treatment by oxygen plasma the excess oxygen atoms increases the film resistance, and its rapid decrease is caused by the desorption of excess oxygen with increasing temperature and absence of SnO inclusions. The treatment by hydrogen plasma results in formation of oxygen vacancies and a

4. It is shown that tin dioxide films synthesized by sol-gel method, after isothermal annealing at 400ºС for 15 min, 3, 6 and 12 h have a high transparency (~ 80–90%). The optimum mode of thermal annealing (400ºC, 6 h) of SnO2 films is identified, which permits to achieve the maximum of gas sensitivity and minimum of response time (2 seconds) due to the better structure and optimal sizes (6 nm) of the crystallites. 5. Treatment by glow discharge hydrogen plasma of tin dioxide films synthesized by solgel technique resulted in a decrease in the transparency of 3–15% in the visible wavelength range due to the formation of opaque inclusions of SnO. Treatment by oxygen plasma resulted in a slight decrease in the transparency of 1−5% due to the

damaging effects of plasma on the structural perfection of the SnO2 crystallites. 6. For nonstoichiometric SnOx films (x<2), deposited by magnetron sputtering in condition of a lack of oxygen (pressure ArO2 mixture 1–2 Pa), the formation of polycrystalline β-Sn phase after treatment by H-plasma is shown. Both the segregation of Sn particles at temperatures of 150–200ºС and the formation of oxides SnO, Sn2O3 and SnO2 at temperatures of 50–550ºС, are taken place. The short-term treatment (5 min) by O- or H-plasma of deposited at a pressure of 2.7 Pa SnO2 film leads the formation of Sn clusters, which reduces the transparency in the near infrared region

7. The possibility of obtaining of the crystal-amorphous nanostructures in which crystallites of tin oxides alternate with clusters obtained by selective amorphization of SnO, SnO2 and Sn2O3 crystallites during treatment by glow discharge H-plasma, is shown. In this case, the order of the atoms in those (hkl) planes of crystallites is

8. The regularities of the influence of treatment by H- or O-plasma on growth of gas sensitivity of SnO2 films were studied. The plasma treatment of SnO2 films, obtained by the sol-gel technique, does not alter their phase composition, and gas sensitivity growth does not depend on the type of plasma. Treatment by O-plasma of films prepared by magnetron sputtering, leads to a more significant increase in gas sensitivity than by H-

9. The phenomena of self-organization of matter in SnOx layers were identified, consisting in the intensive formation of SnO2 crystallites with sizes of 4 nm in the process of film deposition by magnetron sputtering (pressure in the Ar-O2 mixture is 2.7 Pa), and SnO2 crystallites with sizes of 3 nm at the deposition of films by the sol-gel technique

(1200–2500 nm) from 80% to 50% and 40%, respectively.

plasma, due to the oxidation of SnOx film.

(concentration of Sn 0.14 mol/L in a solution).

violated, along which there are variations of the plasma particles.

decrease of film resistance.

the same operating temperature. For the films synthesized by magnetron sputtering, at an operating temperature of 270ºC sensitivity is about 45% (Fig. 31, curve 1). Treatment of H-plasma increases the sensitivity to 52% at the same operating temperature, the same treatment on oxygen plasma increases the sensitivity up to 78% with a decrease in operating temperature to 255ºC (Fig. 31, curves 2 and 3).

Thus, the regularities of the influence of treatment by H- or O-plasma on growth of the gas sensitivity of SnO2 films were established. The plasma treatment of SnO2 films, obtained by the sol-gel technique, does not alter their phase composition, resulting in gas sensitivity growth does not depend on the type of plasma. Treatment by O-plasma of films prepared by magnetron sputtering, leads to a more significant increase in gas sensitivity than by Hplasma, due to the oxidation of SnOx film.
