**3.1. Solution properties**

## *3.1.1. Viscosity*

This part is devoted to study the effect of solution aging and its viscosity on the films thickness. To do this study, five samples are developed, successively on the day, the next day, after seven days, after 10 days, and after fourteen days of solution synthesis. The conservation of the sol during the 14 days is made at room temperature.

Table 1 shows an important change of layers thickness with the increasing of the solution viscosity. After a repose period of 14 days of synthesis solution doped with ZrO2, the thickness of layer changes from 32 nm, with a viscosity 10 mPa.s, to 81 nm when the viscosity is 180 mPa.s. We notice that the ZrO2-doped TiO2 solution becomes more viscous over time (Figure 1). This reflects the rate of polycondensation reaction progress.


**Table 1.** Variation of lm thicknesses d (nm) with solution viscosity.

**Figure 1.** Variation of lm thicknesses d (nm) with solution viscosity η(mPas.s)

## **3.2. Powder Properties (Xerogel)**

### *3.2.1. X-Ray Diffraction (XRD)*

210 Heat Treatment – Conventional and Novel Applications

0.1° 2θ

output of 20 mw.

**3.1. Solution properties** 

**3. Results** 

*3.1.1. Viscosity* 

and films diffraction, Fourier transforms infrared (FTIR), Scanning electron microscopy (SEM), Raman spectroscopy, differential scanning calorimetric (DSC), Scanning electron microscopy (SEM), the energy dispersive X-ray spectrometry (EDX) and UV spectroscopy.

To determine the transformation points, the obtained powdered xerogels were analyzed by Differential Scanning Calorimetry (DSC) using a SETARAM DSC–92 analyzer equipped with a processor and a measuring cell. The thermal cycle applied consists of heating from room temperature to 520°C, holding for 5min at this temperature and finally cooling back to room temperature with the same rate (5°C/min). X-ray powder diffraction was performed by Siemens D5005 diffractometer using a Cu Kα1 radiation. The patterns

and counting time of 1 s.step-1. The UV absorption studies were carried out using

UV-VIS double–beam spectrophotometer SHIMADZU (UV3101PC). Its useful range is between 190 and 3200 nm. The treatment of the spectra was performed using the UVPC software. A surface profiler DEKTAK 3ST AUTO1 (VEECO) was used to determine film thicknesses. Raman spectra were recorded in a back scattering configuration with a Jobbin Yvon micro Raman spectrometer coupled to a DX40 Olympus microscope. The samples of doped and undoped TiO2 thin films were excited with a 632.8 nm wavelength with an

This part is devoted to study the effect of solution aging and its viscosity on the films thickness. To do this study, five samples are developed, successively on the day, the next day, after seven days, after 10 days, and after fourteen days of solution synthesis. The

Table 1 shows an important change of layers thickness with the increasing of the solution viscosity. After a repose period of 14 days of synthesis solution doped with ZrO2, the thickness of layer changes from 32 nm, with a viscosity 10 mPa.s, to 81 nm when the viscosity is 180 mPa.s. We notice that the ZrO2-doped TiO2 solution becomes more viscous

**solution age η (mPa.s) Measured thicknesses d (nm)** 

conservation of the sol during the 14 days is made at room temperature.

over time (Figure 1). This reflects the rate of polycondensation reaction progress.

**0 day** 10 32 **1 day** 20 33 **7 days** 60 39 **10 days** 110 61 **14 days** 180 81

**Table 1.** Variation of lm thicknesses d (nm) with solution viscosity.

θ

, with a step length of

were scanned at room temperature, over the angular range 10-70° 2

Figure 2a and b shows the X-ray diffraction (XRD) patterns of TiO2 xerogels of undoped (Figure 2a) and 5% ZrO2-doped TiO2 (Figure 2b). The XRD pattern evolution of titanium xerogel obtained after the evaporation of the organic compounds during 3 months of aging at ambient temperature shows that it is an amorphous phase as reported in [54].

It has been reported that the used acid catalyst, during sol–gel preparation, plays a crucial role for determining the TiO2 phase, in literature [54, 55], they found that powder is amorphous when they use acetic acid as catalyst. However, when using formic acid they found that, in addition to amorphous phase, there is an amount of the anatase nanoparticles. This analysis of the doped TiO2 xerogel exhibits that the addition of 5% ZrO2 (Figure 2b) would be largely sufficient to form nanoparticles of anatase which crystallizes with (101) plane. It is interesting to note that the addition of a minor amount of ZrO2 starts crystallization of anatase. Whereas, A. Kitiyanan et al. [56] B. Neppolian et al. [57] reported that addition of ZrO2 has no effect on TiO2 oxide morphology..

#### 212 Heat Treatment – Conventional and Novel Applications

### *3.2.2. Thermal analysis*

The differential scanning calorimetric (DSC) curves of undoped TiO2 and 5% ZrO2-doped TiO2 xerogels are shown in Figure 3 It is interesting to note that both doped and undoped xerogels showed a similar thermal behavior in the temperature range 20–250 °C. Generally, weight loss corresponds to the evaporation of water, thermal decomposition of butanol as well as carbonization or combustion of acetic acid and other organic compounds [58-60] which constitute metal alkoxides. Hence, the above thermal events were represented by an endothermic peak spreading from 50 to 250 °C.

Synthesis, Characterization and Properties of

*TiO2 Xerogel* 

*386*

*Xerogel*

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 213

0 100 200 300 400 500

*5% ZrO2*

 *doped TiO2*

*T(°C)*

**Figure 3.** Differential scanning calorimetric curves of xerogels: (a) undoped TiO2 and (b) 5 % ZrO2-

The measured values of thin films thickness, given in Figure 4, were determined with a surface proler for various layers and at different annealing temperatures. It is clearly observed that the film thickness increases with the number of dipping and annealing

*286*

0


**3.3. Thin films properties** 

*3.3.1. Structural Properties* 

*3.3.1.1. Thin films thickness* 

doped TiO2.

temperatures.

0

40

80

120

160 *dQ/dt(mw) Exo*

40

80

120

**Figure 2.** Evolution of XRD patterns of xerogels (a) undoped TiO2, and (b) 5% ZrO2-doped TiO2.

A double exothermic peak in the 260–450 °C temperature range of TiO2 xerogel can be attributed to the crystallization of titanium oxide [61].

The addition of 5% of zirconium oxide led to a shift of exothermic peak phase towards lower temperatures. This may be due to the speeding up of the crystallization of titanium oxide compared to the undoped one.

**Figure 3.** Differential scanning calorimetric curves of xerogels: (a) undoped TiO2 and (b) 5 % ZrO2 doped TiO2.

### **3.3. Thin films properties**

212 Heat Treatment – Conventional and Novel Applications

endothermic peak spreading from 50 to 250 °C.

The differential scanning calorimetric (DSC) curves of undoped TiO2 and 5% ZrO2-doped TiO2 xerogels are shown in Figure 3 It is interesting to note that both doped and undoped xerogels showed a similar thermal behavior in the temperature range 20–250 °C. Generally, weight loss corresponds to the evaporation of water, thermal decomposition of butanol as well as carbonization or combustion of acetic acid and other organic compounds [58-60] which constitute metal alkoxides. Hence, the above thermal events were represented by an

**Figure 2.** Evolution of XRD patterns of xerogels (a) undoped TiO2, and (b) 5% ZrO2-doped TiO2.

attributed to the crystallization of titanium oxide [61].

compared to the undoped one.

A double exothermic peak in the 260–450 °C temperature range of TiO2 xerogel can be

The addition of 5% of zirconium oxide led to a shift of exothermic peak phase towards lower temperatures. This may be due to the speeding up of the crystallization of titanium oxide

*3.2.2. Thermal analysis* 

#### *3.3.1. Structural Properties*

#### *3.3.1.1. Thin films thickness*

The measured values of thin films thickness, given in Figure 4, were determined with a surface proler for various layers and at different annealing temperatures. It is clearly observed that the film thickness increases with the number of dipping and annealing temperatures.

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 215

**Deposition rate (cm.s-1)** 0.6 0.8 1 1.4 1.8 2 **Thickness (**± **0.1 nm)** 38 56 74 93 116 143

**Table 2.** Variation of lm thickness d (nm) for different deposition rate.

**Figure 5.** Variation of thickness logarithm with deposition rate logarithm.

examined at different annealing temperatures (350 to 450 °C).

Our samples were analyzed using X-ray diffraction (XRD) to investigate the transformation behaviors. We have studied the structural properties of undoped TiO2, and doped with 5% ZrO2 thin films and deposited by the sol–gel method. The dip-coated thin films have been

Figure 6 shows XRD patterns of 5% ZrO2-doped TiO2 thin film obtained after different dipping and treated at 450 °C. However, Figure 7 shows XRD patterns of 5% ZrO2-doped TiO2 thin film obtained after 2 dipping and treated at various annealing temperatures at 350 °C, 400 °C and

This analysis of the doped TiO2 thin films exhibits that the addition of a minor amount of ZrO2 would be largely sufficient to form 4 sharp diffraction peaks at 25.35°, 35.51°, 50.90°

450 °C. Clearly, titanium oxide starts to crystallize starting from annealing at 350 °C.

*3.3.1.3. X-Ray Diffraction (XRD)* 

**a. Influence of annealing temperature** 

3.3.1.3.1. Crystallization

**Figure 4.** Variation of lm thicknesses d (nm) for different annealing temperatures and different dipping

#### *3.3.1.2. Study of deposition rate*

A thin film deposited by dip-coating method has a thickness which can be controlled by the deposition rate. To simplify, d thickness depends on the speed according to the following relationship:

$$d = kV\_d^a$$

Where:

Vd is the deposition rate, k is the empirical factor depending on the viscosity, surface tension and density of the solution used and a is the exponent with 2/3 value according to Landau and Levich [62], and 1/2 according to Michels et al. [63] or even proportionately at the speed of dipping Hewak et al. [64].

We then set the objective of determining this factor to validate one of these different models. So, we have prepared six samples with different deposition rate from 0.6 cm.s-1 to 2 cm.s-1 in the same conditions (deposited in solution with a viscosity 40 mPa.s at 21°C and treated at 400°C); we measured their thicknesses by profilometer and the results are grouped in table 2 and represented in figure 5.

This curve shows a linear increase in ln(d) versus ln(Vd). For the exponent a, values of 0,965 have been obtained, which is in good agreement with the exponent obtained by Hewak et al. [64]. This simple comparison shows that the increase in the speed of dipping, results an increase in the thickness of doped thin films.


**Table 2.** Variation of lm thickness d (nm) for different deposition rate.

**Figure 5.** Variation of thickness logarithm with deposition rate logarithm.

### *3.3.1.3. X-Ray Diffraction (XRD)*

3.3.1.3.1. Crystallization

214 Heat Treatment – Conventional and Novel Applications

dipping

relationship:

Where:

*3.3.1.2. Study of deposition rate* 

of dipping Hewak et al. [64].

and represented in figure 5.

increase in the thickness of doped thin films.

**Figure 4.** Variation of lm thicknesses d (nm) for different annealing temperatures and different

A thin film deposited by dip-coating method has a thickness which can be controlled by the deposition rate. To simplify, d thickness depends on the speed according to the following

> . *<sup>a</sup> <sup>d</sup> d kV* =

Vd is the deposition rate, k is the empirical factor depending on the viscosity, surface tension and density of the solution used and a is the exponent with 2/3 value according to Landau and Levich [62], and 1/2 according to Michels et al. [63] or even proportionately at the speed

We then set the objective of determining this factor to validate one of these different models. So, we have prepared six samples with different deposition rate from 0.6 cm.s-1 to 2 cm.s-1 in the same conditions (deposited in solution with a viscosity 40 mPa.s at 21°C and treated at 400°C); we measured their thicknesses by profilometer and the results are grouped in table 2

This curve shows a linear increase in ln(d) versus ln(Vd). For the exponent a, values of 0,965 have been obtained, which is in good agreement with the exponent obtained by Hewak et al. [64]. This simple comparison shows that the increase in the speed of dipping, results an

#### **a. Influence of annealing temperature**

Our samples were analyzed using X-ray diffraction (XRD) to investigate the transformation behaviors. We have studied the structural properties of undoped TiO2, and doped with 5% ZrO2 thin films and deposited by the sol–gel method. The dip-coated thin films have been examined at different annealing temperatures (350 to 450 °C).

Figure 6 shows XRD patterns of 5% ZrO2-doped TiO2 thin film obtained after different dipping and treated at 450 °C. However, Figure 7 shows XRD patterns of 5% ZrO2-doped TiO2 thin film obtained after 2 dipping and treated at various annealing temperatures at 350 °C, 400 °C and 450 °C. Clearly, titanium oxide starts to crystallize starting from annealing at 350 °C.

This analysis of the doped TiO2 thin films exhibits that the addition of a minor amount of ZrO2 would be largely sufficient to form 4 sharp diffraction peaks at 25.35°, 35.51°, 50.90° and 60.76°, these are assigned to (101), (112), (200), and (105) planes which are attributed to anatase nanoparticles (crystalline) phase of TiO2.

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 217

**Figure 7.** Evolution of diffraction patterns of 5% ZrO2-doped TiO2 thin lms; obtained at various

**Figure 8.** Comparison between undoped and 5% ZrO2-doped TiO2 diffraction patterns.

The crystallite size D of TiO2 doped with ZrO2 thin lms can be deduced from XRD line

3.3.1.3.2. Surface morphology and grain size

broadening using Scherrer equation [65]:

annealing temperatures (350, 400, 450 °C) for the same thickness.

Furthermore, all XRD patterns show a peak at 30.41° corresponding to (111) plane which is attributed to the brookite formation whatever the annealing temperature.

Peak intensities corresponding to characteristic planes of anatase (101) and brookite (111) phases are obviously increased with the increase of annealing temperature and number of dipping. This latter is interpreted as due not only to increase in proportion of titanium oxide but also to the improvement of the crystalline quality.

However, in the same conditions Kitiyanan et *al.* [56] B. Neppolian et *al.* [57] showed that titanium oxide, crystallizes in anatase phases and they found that addition of minor amount of ZrO2 has not contributed to any change in the TiO2 morphology.

## **b. Influence of doping with ZrO2**

Comparison between XRD patterns (Figure 8) of both undoped and doped with 5% ZrO2 thin lms obtained after 2 dippings and annealed at 400 °C temperature showed a similar behavior, so characteristic peaks correspond to the crystallization of anatase and brookite phases of the doped state is shifted to larger angles compared to the undoped one.

**Figure 6.** Evolution of diffraction patterns of 5% ZrO2-doped TiO2 thin lms; obtained after various dipping (2, 4, 6, 8) annealed at 450°C. : substrat, : anatase, : brookite.

Synthesis, Characterization and Properties of Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 217

**Figure 7.** Evolution of diffraction patterns of 5% ZrO2-doped TiO2 thin lms; obtained at various annealing temperatures (350, 400, 450 °C) for the same thickness.

**Figure 8.** Comparison between undoped and 5% ZrO2-doped TiO2 diffraction patterns.

3.3.1.3.2. Surface morphology and grain size

216 Heat Treatment – Conventional and Novel Applications

**b. Influence of doping with ZrO2**

anatase nanoparticles (crystalline) phase of TiO2.

but also to the improvement of the crystalline quality.

and 60.76°, these are assigned to (101), (112), (200), and (105) planes which are attributed to

Furthermore, all XRD patterns show a peak at 30.41° corresponding to (111) plane which is

Peak intensities corresponding to characteristic planes of anatase (101) and brookite (111) phases are obviously increased with the increase of annealing temperature and number of dipping. This latter is interpreted as due not only to increase in proportion of titanium oxide

However, in the same conditions Kitiyanan et *al.* [56] B. Neppolian et *al.* [57] showed that titanium oxide, crystallizes in anatase phases and they found that addition of minor amount

Comparison between XRD patterns (Figure 8) of both undoped and doped with 5% ZrO2 thin lms obtained after 2 dippings and annealed at 400 °C temperature showed a similar behavior, so characteristic peaks correspond to the crystallization of anatase and brookite

phases of the doped state is shifted to larger angles compared to the undoped one.

**Figure 6.** Evolution of diffraction patterns of 5% ZrO2-doped TiO2 thin lms; obtained after various

dipping (2, 4, 6, 8) annealed at 450°C. : substrat, : anatase, : brookite.

attributed to the brookite formation whatever the annealing temperature.

of ZrO2 has not contributed to any change in the TiO2 morphology.

The crystallite size D of TiO2 doped with ZrO2 thin lms can be deduced from XRD line broadening using Scherrer equation [65]:

$$\mathbf{D} = \frac{0.94 \times \mathcal{A}}{\sqrt{\left(\Delta\_{hkl}^2 - \Delta\_{instr}^2\right)}} \frac{1}{\cos\theta}$$

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 219

The computed values of grain sizes, given in Table 3, were calculated for different annealing temperatures with the same thickness. Thus, the obtained grain sizes of anatase and brookite increase from 8.58 nm to 20.56 nm and from 17.50 nm to 18.06 nm, respectively. In

It is interesting to note that the grain size improves and the defects like dislocation density and strain in the lms decrease with lm thickness. This may be due to the improvement in crystallinity in the lms with lm thickness. As we also note that the variation of the strain is perfectly correlated with that of the dislocation density δ. When these increase, they cause the decrease in grain size and leads to recrystallization of the nanoparticles. Furthermore, the stages of nucleation, growth and coalescence become stable, which causing the reduction

The evolution of the grain size D according the annealing temperature can be interpreted by

( ) 0 /, *D D exp E k T a b* = −

The values of activation energies of crystallization corresponding respectively to anatase and brookite phases are calculated from the curve showed in figure 9, we note that the activation energy of the anatase Ea=0,096 eV crystallization is lower than that of the brookite Ea=0,012 eV. This means that the formation of anatase requires more energy than that of

The Raman spectra of undoped and 5% ZrO2-doped TiO2 thin films annealed at 450°C for different dipping (figure 10) display various peaks related to titanium oxide as anatase and brookite phases. These spectra exhibit bands at around 138 (strong), 235 (weak), 514 (weak) and 632 cm-1 (medium)) for the thin layers of ZrO2-doped TiO2 corresponding to the Eg modes of vibration. The above bands can be assigned to anatase phase except the band 235 cm-1 corresponding to the B1g modes of vibration, which is due to the crystallization of brookite phase. While bands of 144, 188 and 651 cm-1 can be assigned to both anatase and

A slight shift of the most intense peak, Eg, to smaller wavenumber is observed for all thin films doped with ZrO2 by comparison with anatase of undoped phase (figure 11). Similar displacements have been previously reported in XRD patterns and they can be correlated

fact, as annealing temperature increases grain sizes also increases.

of constraints in the film formed.

• Ea is the energy activation of crystallization;

• The size D tends towards the infinite for high temperatures [67].

with the confinement effects in nano-structured anatase crystallites.

the Arrhenius law (figure 9):

• KB the Boltzmann constant; • D0 pre-exponential factor.

Where:

brookite.

*3.3.1.4. RAMAN* 

brookite phases [68,69].

λ is the wavelength of X-ray beam (Cu Kα=1.5406 Å), Δhkl is the full width at half maximum (FWHM) of (hkl) diffraction peak, Δinstr is the FWHM corresponding to the instrumental limit, and θ is the Bragg angle.

Using the size of the crystallites, the dislocation density (δ) [66], the number of crystallites per unit surface area (N) and strain in the lms (ε), which are newly introduced by Ray et al. [66], has been determined:

$$
\delta = \frac{1}{D^2}
$$

$$
N = \frac{d}{D^3}
$$

$$
\mathcal{E} = \frac{\Delta(2\theta) \cdot \cos\theta}{4}
$$

Where d is the lm thickness.

The calculated structural parameters are presented in Table 3.


**Table 3.** Structural parameters of xerogels and thin lms, for different annealing temperatures and same thickness.

The computed values of grain sizes, given in Table 3, were calculated for different annealing temperatures with the same thickness. Thus, the obtained grain sizes of anatase and brookite increase from 8.58 nm to 20.56 nm and from 17.50 nm to 18.06 nm, respectively. In fact, as annealing temperature increases grain sizes also increases.

It is interesting to note that the grain size improves and the defects like dislocation density and strain in the lms decrease with lm thickness. This may be due to the improvement in crystallinity in the lms with lm thickness. As we also note that the variation of the strain is perfectly correlated with that of the dislocation density δ. When these increase, they cause the decrease in grain size and leads to recrystallization of the nanoparticles. Furthermore, the stages of nucleation, growth and coalescence become stable, which causing the reduction of constraints in the film formed.

The evolution of the grain size D according the annealing temperature can be interpreted by the Arrhenius law (figure 9):

$$D = D\_0 \exp\left(-E\_a / k\_b T\right).$$

Where:

218 Heat Treatment – Conventional and Novel Applications

limit, and θ is the Bragg angle.

[66], has been determined:

Where d is the lm thickness.

Same thickness

Xerogel 3 months at

Xerogel 3 months at

350 °C

400 °C

450 °C

Undoped TiO2

> 5

same thickness.

**%**ZrO2-doped TiO**2**

( ) 2 2 0.94 1 <sup>D</sup>

<sup>×</sup> <sup>=</sup> Δ −Δ

*hkl instr*

λ is the wavelength of X-ray beam (Cu Kα=1.5406 Å), Δhkl is the full width at half maximum (FWHM) of (hkl) diffraction peak, Δinstr is the FWHM corresponding to the instrumental

Using the size of the crystallites, the dislocation density (δ) [66], the number of crystallites per unit surface area (N) and strain in the lms (ε), which are newly introduced by Ray et al.

> 2 1 *D* δ=

3 *<sup>d</sup> <sup>N</sup> D* =

> θ

T ambient Amorphous - - - - -

T ambient Anatase (101) 14,80 - - -

traits/nm2

Anatase (101) 8,58 135,84 99,74 3,11 Brookite (111) 17,50 32,65 11,76 3,89 Anatase (112) 16,66 36,03 13,62 4,52 Anatase (200) 14,74 46,03 19,67 6,69 Anatase (105) 16,33 37,50 14,47 7,53

Anatase (101) 10,09 98,22 73,01 3,07 Brookite (111) 17,61 32,25 13,73 3,82 Anatase (112) 17,27 33,53 14,56 4,44 Anatase (200) 15,57 41,25 19,87 6,37 Anatase (105) 18,71 28,57 11,45 7,58

Anatase (101) 13,92 51,61 29,29 2,98 Brookite (111) 18,06 30,66 13,41 3,74 Anatase (112) 19,09 27,44 11,36 4,38 Anatase (200) 18,63 28,81 12,22 6,34 Anatase (105) 20,56 23,66 9,09 7,47

) N (10-3 nm-2) ε (10-4)

(2 ) cos 4 θ

Δ ⋅ <sup>=</sup>

**Table 3.** Structural parameters of xerogels and thin lms, for different annealing temperatures and

ε

Annealed at Phase (hkl) L (nm) <sup>δ</sup> (10-4

The calculated structural parameters are presented in Table 3.

λ

cos

θ


The values of activation energies of crystallization corresponding respectively to anatase and brookite phases are calculated from the curve showed in figure 9, we note that the activation energy of the anatase Ea=0,096 eV crystallization is lower than that of the brookite Ea=0,012 eV. This means that the formation of anatase requires more energy than that of brookite.

## *3.3.1.4. RAMAN*

The Raman spectra of undoped and 5% ZrO2-doped TiO2 thin films annealed at 450°C for different dipping (figure 10) display various peaks related to titanium oxide as anatase and brookite phases. These spectra exhibit bands at around 138 (strong), 235 (weak), 514 (weak) and 632 cm-1 (medium)) for the thin layers of ZrO2-doped TiO2 corresponding to the Eg modes of vibration. The above bands can be assigned to anatase phase except the band 235 cm-1 corresponding to the B1g modes of vibration, which is due to the crystallization of brookite phase. While bands of 144, 188 and 651 cm-1 can be assigned to both anatase and brookite phases [68,69].

A slight shift of the most intense peak, Eg, to smaller wavenumber is observed for all thin films doped with ZrO2 by comparison with anatase of undoped phase (figure 11). Similar displacements have been previously reported in XRD patterns and they can be correlated with the confinement effects in nano-structured anatase crystallites.

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 221

**Figure 10.** Raman spectrum of 5% ZrO2-doped TiO2 thin films annealed at 450°C for different dipping;

**Figure 11.** Comparison between Raman spectrum of undoped and 5% ZrO2-doped TiO2 thin films;

=anatase, =brookite.

=anatase, =brookite.

**Figure 9.** Plot of log (D) versus (1000/T) for determination of activation energies of anatase and brookite

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 221

220 Heat Treatment – Conventional and Novel Applications

**Figure 9.** Plot of log (D) versus (1000/T) for determination of activation energies of anatase and brookite

**Figure 10.** Raman spectrum of 5% ZrO2-doped TiO2 thin films annealed at 450°C for different dipping; =anatase, =brookite.

**Figure 11.** Comparison between Raman spectrum of undoped and 5% ZrO2-doped TiO2 thin films; =anatase, =brookite.

*3.3.1.5. FTIR* 

Figure 12 shows the infrared absorption spectrum of the 5% ZrO2 -doped TiO2 lms annealed at different temperature. The peak at 2360 cm-1 resulted from the adsorbed H2O molecules, which were not removed completely after sol–gel coating. The peaks at 1242 cm-1, 1111 cm-1, 1035 cm-1 and 860 cm-1 correspond to the vibration mode of Ti-OH [70, 71].

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 223

5% ZrO2-doped TiO2 thin films deposited on ITO substrates and obtained after various annealing temperature at 350°C and 450°C were coated and examined in a scanning electron microscope (SEM) to investigate their structure and surface characteristics. We observed that the coating was homogeneous without any visual cracking over a wide area. The

The surface composition of films is further identied by EDX measurement. EDX result shown in Figure 13 demonstrates that the peaks of Ti, O and Zr can be clearly seen in the survey spectrum. While the other elements as Si, In, Ca, Na and Mg are the components of

 **O Na Mg Si Ca In Zr Ti Total at.%** 40,13 2,03 3,15 37,83 3,07 4,39 1,23 8,17 100

**Table 4.** Elemental composition (at. %) of 5% ZrO2 -doped TiO2 thin films, treated at 450 ° C.

increase in the treatment temperature, did not affect the uniformity of the film.

The chemical compositions of thin film analyzed are given in table 4.

**Figure 13.** EDX spectra of 5% ZrO2-doped TiO2 thin films annealed at 450°C.

Figure 14 display diffused scattering UV-VIS transmittance spectra of TiO2 thin lms undoped and 5% ZrO2-doped TiO2, for different annealing temperatures from 350°C to 450°C and different numbers of dipping (3, 4, 6, 8 dipping) in the wavelength range 300–800 nm, where the film due to interference phenomena between the wave fronts generated at

*3.3.1.6. Scanning electron microscopy (SEM) and EDX*

ITO substrate.

*3.3.2. Optical properties* 

*3.3.2.1. UV absorption analysis* 

The band around 665 cm-1 was attributed to the vibration mode of Ti-O-Ti bond [72] and another bond appears around 455 cm-1, this is the O-Ti-O band corresponding to the crystalline titania in the anatase form [73-76].

We find that the vibration bands intensity located in the vicinity of 665 cm-1 and 455 cm-1 increase when annealing temperature increases. This indicates that the number of Ti-O-Ti and O-Ti-O links of titanium dioxide crystallization is also growing.

**Figure 12.** FTIR spectra of 5% ZrO2-doped TiO2 thin lms, obtained at various annealing temperatures (350, 400, 450 °C).

#### *3.3.1.6. Scanning electron microscopy (SEM) and EDX*

222 Heat Treatment – Conventional and Novel Applications

crystalline titania in the anatase form [73-76].

Figure 12 shows the infrared absorption spectrum of the 5% ZrO2 -doped TiO2 lms annealed at different temperature. The peak at 2360 cm-1 resulted from the adsorbed H2O molecules, which were not removed completely after sol–gel coating. The peaks at 1242 cm-1,

The band around 665 cm-1 was attributed to the vibration mode of Ti-O-Ti bond [72] and another bond appears around 455 cm-1, this is the O-Ti-O band corresponding to the

We find that the vibration bands intensity located in the vicinity of 665 cm-1 and 455 cm-1 increase when annealing temperature increases. This indicates that the number of Ti-O-Ti

**Figure 12.** FTIR spectra of 5% ZrO2-doped TiO2 thin lms, obtained at various annealing temperatures

1111 cm-1, 1035 cm-1 and 860 cm-1 correspond to the vibration mode of Ti-OH [70, 71].

and O-Ti-O links of titanium dioxide crystallization is also growing.

*3.3.1.5. FTIR* 

(350, 400, 450 °C).

5% ZrO2-doped TiO2 thin films deposited on ITO substrates and obtained after various annealing temperature at 350°C and 450°C were coated and examined in a scanning electron microscope (SEM) to investigate their structure and surface characteristics. We observed that the coating was homogeneous without any visual cracking over a wide area. The increase in the treatment temperature, did not affect the uniformity of the film.

The surface composition of films is further identied by EDX measurement. EDX result shown in Figure 13 demonstrates that the peaks of Ti, O and Zr can be clearly seen in the survey spectrum. While the other elements as Si, In, Ca, Na and Mg are the components of ITO substrate.

The chemical compositions of thin film analyzed are given in table 4.


**Figure 13.** EDX spectra of 5% ZrO2-doped TiO2 thin films annealed at 450°C.

## *3.3.2. Optical properties*

#### *3.3.2.1. UV absorption analysis*

Figure 14 display diffused scattering UV-VIS transmittance spectra of TiO2 thin lms undoped and 5% ZrO2-doped TiO2, for different annealing temperatures from 350°C to 450°C and different numbers of dipping (3, 4, 6, 8 dipping) in the wavelength range 300–800 nm, where the film due to interference phenomena between the wave fronts generated at

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 225

If the film surface is rough, the radiation in film/air interface undergo scattering in all directions instead of a reflection. Oh et al. [28], Kim et al. [77] show that the interference fringes are due the increase in thin films thickness. The occurrence of such fringes means

Analysis of UV– VIS transmission spectra shows that the 5% ZrO2-doped TiO2 thin lms are transparent in the visible range and opaque in the UV region, whatever are the annealing

The amplitude of interference spectra increased with increasing calcination temperature. These results show that the refractive index of TiO2 thin films is increased while the film thickness is decreased. This can be due to the formation stage of anatase and with the

A slight shift of transmission curves to lower wavelengths is observed for curves of ZrO2 doped thin lms in comparison with those undoped (figure 14) This shift is ascribed to the

The refractive index of TiO2 thin lms was calculated from measured UV–VIS transmittance spectrum. The evaluation method used in this work is based on the analysis of UV–VIS transmittance spectrum of a weakly absorbing lm deposited on a non-absorbing substrate [78]. The refractive index n (λ) over the spectral range is calculated by using the envelopes

<sup>0</sup> ( ) () () *<sup>S</sup> n S Sn n*

( ) 2 2 max min

Where n0 is the refractive index of air, ns is the refractive index of the lm, Tmax is the maximum envelope, and Tmin is the minimum envelope. The thickness of the lms was adjusted to provide the best ts to the measured spectra. In this study, all the deposited

<sup>1</sup> () () () () 2 <sup>2</sup> () () *S S*

<sup>2</sup> <sup>1</sup>

− 

*n*

<sup>−</sup> =− ×

*nd*

Where nd is refractive index of pore-free anatase (nd = 2.52) [80], and n is refractive index of

<sup>1</sup> 100(%) <sup>2</sup> <sup>1</sup>

=+ −

*T T S n n nn*

<sup>−</sup> = ++

0 0

The porosity of the thin lms is calculated using the following equation [79]

*Porosité*

 λ

λ

λ

22 2

λ λ

max min

×

 λ

 λ

*T T* λ

λ

that our lms are sufciently thick

temperature and number of dipping

increase in the grain size [28].

increase in band gap energy.

*3.3.2.2. Refractive index, density, thickness and porosity* 

that are tted to the measured extreme:

lms are assumed to be homogeneous.

porous thin films.

**Figure 14.** UV–VIS spectra of undoped TiO2 and 5% ZrO2-doped TiO2 thin films, for various dipping and at different annealing temperatures.

the two interfaces (air and substrate) defines the sinusoidal behavior of the curves' transmittance versus wavelength of light. The curves showed a similar behavior in the temperature range:


As can be seen, all the spectrums exhibit interference fringes, which are due to the multiple reflections at the two film edges, i.e. at the film/air and the film/substrate interfaces. This indicates that the top film surface is smooth and uniform and exhibits a good transparency in the visible region. So that the excellent surface quality and homogeneity of the film were confirmed from the appearance of interference fringes in the transmission spectra. This occurs when the film surface is reflecting without much scattering/absorption in the bulk of the film.

If the film surface is rough, the radiation in film/air interface undergo scattering in all directions instead of a reflection. Oh et al. [28], Kim et al. [77] show that the interference fringes are due the increase in thin films thickness. The occurrence of such fringes means that our lms are sufciently thick

Analysis of UV– VIS transmission spectra shows that the 5% ZrO2-doped TiO2 thin lms are transparent in the visible range and opaque in the UV region, whatever are the annealing temperature and number of dipping

The amplitude of interference spectra increased with increasing calcination temperature. These results show that the refractive index of TiO2 thin films is increased while the film thickness is decreased. This can be due to the formation stage of anatase and with the increase in the grain size [28].

A slight shift of transmission curves to lower wavelengths is observed for curves of ZrO2 doped thin lms in comparison with those undoped (figure 14) This shift is ascribed to the increase in band gap energy.

## *3.3.2.2. Refractive index, density, thickness and porosity*

224 Heat Treatment – Conventional and Novel Applications

and at different annealing temperatures.

the transition electronic inter-band.

interest in undoped or doped TiO2 thin films.

temperature range:

**Figure 14.** UV–VIS spectra of undoped TiO2 and 5% ZrO2-doped TiO2 thin films, for various dipping

the two interfaces (air and substrate) defines the sinusoidal behavior of the curves' transmittance versus wavelength of light. The curves showed a similar behavior in the

• A **region** characterized by a strong absorption at λ< 380 nm, this absorption is due to

• High transmittance region, from 60 to 95% on a wide range of wavelength in the visible region (from 380 to 800 nm) has been observed which may be used in applications in solar cells. High transparency is one of the most important properties that explain the

As can be seen, all the spectrums exhibit interference fringes, which are due to the multiple reflections at the two film edges, i.e. at the film/air and the film/substrate interfaces. This indicates that the top film surface is smooth and uniform and exhibits a good transparency in the visible region. So that the excellent surface quality and homogeneity of the film were confirmed from the appearance of interference fringes in the transmission spectra. This occurs when the film surface is reflecting without much scattering/absorption in the bulk of the film.

The refractive index of TiO2 thin lms was calculated from measured UV–VIS transmittance spectrum. The evaluation method used in this work is based on the analysis of UV–VIS transmittance spectrum of a weakly absorbing lm deposited on a non-absorbing substrate [78]. The refractive index n (λ) over the spectral range is calculated by using the envelopes that are tted to the measured extreme:

$$n(\mathcal{A}) = \sqrt{S + \sqrt{S^2 - n\_0^2(\mathcal{A})n\_S^2(\mathcal{A})}}$$

$$S = \frac{1}{2} \left( n\_0^2(\mathcal{A}) + n\_S^2(\mathcal{A}) \right) + 2n\_0 n\_S \frac{T\_{\text{max}}(\mathcal{A}) - T\_{\text{min}}(\mathcal{A})}{T\_{\text{max}}(\mathcal{A}) \times T\_{\text{min}}(\mathcal{A})}$$

Where n0 is the refractive index of air, ns is the refractive index of the lm, Tmax is the maximum envelope, and Tmin is the minimum envelope. The thickness of the lms was adjusted to provide the best ts to the measured spectra. In this study, all the deposited lms are assumed to be homogeneous.

The porosity of the thin lms is calculated using the following equation [79]

$$Porosité} = \left(1 - \frac{n^2 - 1}{n\_d^2 - 1}\right) \times 100 (\%)$$

Where nd is refractive index of pore-free anatase (nd = 2.52) [80], and n is refractive index of porous thin films.

The relationship between density, and refractive index for the polymorphs was proposed by Gladstone-Dale [81]. This relationship is as follows:

Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 227

**Figure 15.** Variation of refractive index (n) of 5% ZrO2-doped TiO2 for different annealing temperatures

**Figure 16.** Variation of porosity (p) of 5% ZrO2-doped TiO2 for different annealing temperatures and

and different thickness

different thickness.

$$n = 1 + 0.4\rho$$

Where :

n: mean index of refraction, d: density 0,40: Gladstone-Dale constant for TiO2 The thickness of the films was calculated using the equation:

$$d = \lambda\_1 \lambda\_2 / 2 \text{ ( $\lambda\_1 n\_2 - \lambda\_2 n\_1$ )}$$

Where *n*1 and *n*2 are the refractive indices corresponding to the wavelengths λ1 and λ2, respectively [82].

The results of the computed refractive index (n) (figure 15), density (*ρ*) and porosity (p) (figure 16) are shown in Table 5. It is noted that the refractive index and the density of thin lms of doped titanium oxide increases with increasing annealing temperature and number of dipping; due to phase transition (anatase, anatase–brookite), which increases grain sizes and/or the density of layers.

This phenomenon is related to crystallization, pores destruction and densification of associated film, as well as the elimination of organic compounds.

However, the porosity decreases with increasing annealing temperature and film thickness.


**Table 5.** Variation of refractive index (n), density (ρ), porosity (p) of undoped and 5% ZrO2-doped TiO2 for different annealing temperatures and different thickness.

The calculated values of thin films thickness are given in table 6. It is clearly observed that the film thickness increases with the number of dipping and annealing temperatures, which is in good agreement with results obtained previously of thickness determined with a surface proler.

226 Heat Treatment – Conventional and Novel Applications

Where :

d: density

respectively [82].

**350°C** 

**400°C** 

**450°C** 

surface proler.

n: mean index of refraction,

and/or the density of layers.

0,40: Gladstone-Dale constant for TiO2

Gladstone-Dale [81]. This relationship is as follows:

The thickness of the films was calculated using the equation:

associated film, as well as the elimination of organic compounds.

for different annealing temperatures and different thickness.

The relationship between density, and refractive index for the polymorphs was proposed by

*n* = + 1 0,4

������ 2 ����� � ���� ⁄ � Where *n*1 and *n*2 are the refractive indices corresponding to the wavelengths λ1 and λ2,

The results of the computed refractive index (n) (figure 15), density (*ρ*) and porosity (p) (figure 16) are shown in Table 5. It is noted that the refractive index and the density of thin lms of doped titanium oxide increases with increasing annealing temperature and number of dipping; due to phase transition (anatase, anatase–brookite), which increases grain sizes

This phenomenon is related to crystallization, pores destruction and densification of

However, the porosity decreases with increasing annealing temperature and film thickness.

*T (°C) Films of n ρ P(%) n ρ P(%) n ρ P(%)* 

**Table 5.** Variation of refractive index (n), density (ρ), porosity (p) of undoped and 5% ZrO2-doped TiO2

The calculated values of thin films thickness are given in table 6. It is clearly observed that the film thickness increases with the number of dipping and annealing temperatures, which is in good agreement with results obtained previously of thickness determined with a

**TiO2** *1,92 2,30 49,8 2,15 2,88 36,3* 2,19 *2,98 29,1*  **TiO2 : ZrO2** *1,62 1,55 69,3 2,13 2,83 33,9 2,18 2,95 29,7* 

**TiO2** *2,11 2,78 35,5 2,21 3,03 27,4 2,25 3,13 24,1*  **TiO2 : ZrO2** *1,91 2,28 50,3 2,18 2,95 29,9 2,21 3,03 27,4* 

**TiO2** *2,15 2,88 36,3 2,29 3,23 20,7 2,37 3,43 13,7*  **TiO2 : ZrO2** *2,17 2,93 30,7 2,23 3,08 25,7 2,29 3,23 20,7* 

*4 Dipping 6 Dipping 8 Dipping* 

ρ

**Figure 15.** Variation of refractive index (n) of 5% ZrO2-doped TiO2 for different annealing temperatures and different thickness

**Figure 16.** Variation of porosity (p) of 5% ZrO2-doped TiO2 for different annealing temperatures and different thickness.


Synthesis, Characterization and Properties of

Zirconium Oxide (ZrO2)-Doped Titanium Oxide (TiO2) Thin Films Obtained via Sol-Gel Process 229

**Figure 17.** Plot of (αhν)2 versus (hν) for determination of band gap of undoped and 5% ZrO2-doped

In this study, we investigated the transformation behaviors and the effect; of a smaller ratio range of ZrO2; doping on the surface area of TiO2 thin films, band gap energy, variations of crystal granularity, phase composition and especially on the evolution of the crystallite size and defects concentration with annealing treatments and layers thickness of the samples produced. So that in this chapter, we report the study of structural, thermal and optical properties of ZrO2-doped TiO2 thin films deposited by the

Analyses of doped TiO2 xerogel show that addition of 5% ZrO2 would be largely sufficient to form nanoparticles of anatase (size of grain of 14.78 nm) by contrast to that of undoped TiO2. X-ray diffraction and Raman spectroscopy analyses exhibit that doped thin films obtained starting from annealing at 350°C crystallize in both anatase and brookite phases. Calculation of grain sizes by Scherrer's formula, gives sizes ranging from 8.58 to 20.56 nm and we note an increase in grain sizes by increasing the annealing temperature for all structures. Raman spectroscopy studies confirms the results found by XRD and reveal that

TiO2.

**4. Conclusion** 

sol–gel process

**Table 6.** Variation of calculated lm thicknesses d (nm) for different annealing temperatures and different dipping.

### *3.3.2.3. Optical band gap:*

The band gap is then found as the intercept of the linear portion of the plot. For a direct band gap semi-conductor, the absorption near the band edge can be estimated from the following equation known as the Tauc plot [83]:

$$(ah\nu)^{\ast} = C(h\nu^{\ast} - E\_{\mathfrak{g}})^{\ast}$$

Where C is a constant, Eg the optical band, α is the optical absorption coefcient, h*ν* is the photon energy gap, h the Plank's constant and the exponent n characterizes the nature of band transition; the values of n = 1/2 and 3/2 correspond to direct allowed and direct forbidden transitions, n = 2 and 3 are related to indirect allowed and indirect forbidden transitions [83] and in the cases for a direct band gap semi-conductor like TiO2 the relation become [84, 85]:

$$(ah\nu)^2 = \mathcal{C}\left(h\nu - E\_g\right)$$

The energy band gap (Eg) of the lms can be estimated by plotting (αh*ν*)2 versus h*ν* (Figure 17), then extrapolating the straight-line part of the plot to the photon energy axis. The energy band gap of 5% ZrO2-doped TiO2 lms, given in table 7, decrease owing to an increase in annealing temperatures and the number of dipping. The values are 3.65 and 3.54 eV at 350°C and 450°C respectively.

This decrease was correlated with grains size increases with temperature, when the latter increases the defects and impurities tend to disappear causing a reorganization of the structure. We find that doping with ZrO2 causes an increase in the band gap by contrast to that of undoped TiO2 (3.50 eV).


**Table 7.** Variation of band gap of 5% ZrO2-doped TiO2 thin films for different annealing temperatures and different thickness.

**Figure 17.** Plot of (αhν)2 versus (hν) for determination of band gap of undoped and 5% ZrO2-doped TiO2.
