**3. Results and Discussion**

The TiO2 thin films produced by the surfactant-assisted sol-gel method described were transparent, covering the entire surface of the area of glass that was dipped and showed evidence of birefringence. All films produced exhibited good adherence to the substrate after annealing, and passed the scotch tape test. The average thickness of TiO2 thin films produced varied within the range of 42–220 nm, whereby the thinnest film at 42 nm was a one layer sample produced from a sol-gel solution that did not contain any added surfactant, and the thickest film, 220 nm was a three layer sample produced from a sol-gel solution with 0.003 mol·dm<sup>í</sup><sup>3</sup> of Tween® 20 surfactant added. The thickness of films produced was found to vary depending on the surfactant and concentrations used during processing, as well as the number of layers. The thin film thicknesses are shown in Table 2, where it can be seen that on average the thickness of the film increases with consecutive number of dips, as expected as this increases the number of layers of TiO2 on the surface. Single layer samples varied between 70–200 nm, two layer samples were within the thickness range of 110–210 nm, and three layer samples varied between 130–220 nm.

**Table 2.** Samples prepared via sol-gel using different types and concentrations of Tween® surfactant, and annealed at 500 °C for 15 min. A = no surfactant; B = Tween® 60 (6 × 10<sup>í</sup><sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ); C = Tween® 60 (4 × 10í<sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ); D = Tween® 60 (2 × 10<sup>í</sup><sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ); E = Tween® 40 (6 × 10í<sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ); F = Tween® 40 (4 × 10í<sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ); G = Tween® 20 (6 × 10í<sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ); H = Tween® 20 (4 × 10í<sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> ). Numbers in sample name represent number of layers. Particle sizes marked with asterisks (\*) denote agglomeration within the thin film. Contact angle measurements are given with standard deviation values.


Furthermore, the addition of surfactant was found on average to produce thicker thin films. For example, the three layer sample without surfactant, A3 had a thickness of 130 nm, whereas three layer samples produced with surfactant ranged in thickness depending on the type and concentration used; 180–220 nm, as shown in Table 2. The addition of surfactant increases the viscosity of the original sol-gel solution, and further increasing the surfactant concentration causes additional viscosity within the sol-gel, so the resulting thin films are thicker [12].

The thin film microstructure can be analysed using the SEM images, whereby it was found that all variants explored; the addition of different surfactants, variation in the number of dips, and the annealing temperature, all influenced the morphology of the TiO2 thin films produced as shown in Figures 1–3. These variations within the sol-gel processing used were also found to significantly affect the photocatalytic and wetting properties of the thin films. Surfactant addition was found to alter the morphology of the TiO2 thin films produced, as seen in the SEM images in Figures 1–3. In comparison to the thin films produced without surfactant, as in Figure 1, those films produced with the addition of Tween® surfactants (Figures 2 and 3) show less agglomeration, greater particle definition and exhibited a wider particle size range within samples.

**Figure 1.** SEM Images of samples prepared via sol-gel. A = no surfactant; numbers represent number of layers. Samples were annealed at 500 °C/15 min.

**Figure 2.** SEM Images of samples prepared by sol-gel with decreasing concentration of Tween® 40 surfactant. B = 0.0006 mol·dmí<sup>3</sup> ; C = 0.0003 mol·dm<sup>í</sup><sup>3</sup> ; numbers represent number of layers. Samples were annealed at 500 °C/15 min.

The addition of Tween® surfactants was found to exhibit a range of effects on the morphology of the thin films produced, depending on the concentration and type of surfactant used. Surfactant addition was found to decrease average particle size, from 130 nm for two layer samples produced without surfactant, to as low as 25 nm for two layer samples produced with Tween® 20 surfactant at the higher concentration of 6 × 10<sup>í</sup><sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> . The smaller particle sizes as listed in Table 2 are attributable to the role of the surfactant during the sol-gel growth phase, whereby the surfactant orients itself around growing titania particles restricting their growth to produce smaller particles. In addition, the samples produced with the addition of surfactant show greater particle definition and less agglomeration compared to the samples produced without surfactant (Figures 1–3) whereby the particles are also more angular due to the surfactant restricting their growth in a random way. This reduced particle size and enhanced particle definition increases the resulting surface area to volume ratio within the TiO2 thin film sample. This leads to improved functional properties, such as improved photocatalytic activity, which has been shown for Brij® type surfactants in a previous study [13]. For example, samples produced without surfactant showed a photocatalytic half-life for the degradation of resazurin ink ranging from 9.5 to 16.5 min for the three-layer and two-layer sample respectively. Those samples produced with Tween® surfactant exhibited half-lives ranging from 3 to 5 min for two layer samples. This decrease in half-life is attributable to the reduced average particle size, as well as the increased surface roughness of surfactant enhanced thin films as shown in Table 2, which both result in an overall increased surface area to volume ratio. This enables better adsorption of the resazurin dye to the thin film surface, and a greater surface area upon which the photocatalytic reaction can occur.

**Figure 3.** SEM Images of samples prepared by sol-gel with decreasing concentration of Tween® 20 surfactant. D = 0.0006 mol·dmí<sup>3</sup> ; E = 0.0003 mol·dm<sup>í</sup><sup>3</sup> ; numbers represent number of layers. Samples were annealed at 500 °C/15 min.

*3.2. Influence of Surfactant Addition on Average Surface Roughness of Thin Films* 

Surfactant addition has been found to increase the root mean square surface roughness of thin films by up to 180 nm, as sample A2 (produced without surfactant) has an average root mean square surface roughness of 11 nm, compared with sample B2 (produced with 0.006 mol·dm<sup>í</sup><sup>3</sup> of Tween® 40) which has an average surface roughness of 196 nm. This large increase in surface roughness is a result of the morphological changes within the thin film that have been described, whereby the particles produced with surfactant are smaller and also more angular in shape due to the surfactant obstruction during the sol-gel growth phase.

In addition, when the concentration of the surfactant is decreased from 6 × 10<sup>í</sup>4 mol·dmí<sup>3</sup> to 4 × 10<sup>í</sup><sup>4</sup> mol·dm<sup>í</sup><sup>3</sup> , as in samples D to E, the root mean square surface roughness decreases, whereby sample D2 exhibits the highest surface roughness of 350 nm, and E2 has a surface roughness of 191 nm. Further root mean square surface roughness values are given in Table 2. This increased surface roughness in the samples with increased concentration of surfactant can be explained by the effect of the surfactant as it surrounds the titania particles during the growth phase. When the surfactant concentration is reduced, as from sample D2 to E2, the growing particles are less restricted in their growth, meaning they can grow larger and more spherical, as can be seen in the SEM images D2 to C2 in Figure 3. This is reflected in the particle sizes, whereby sample D2 has an average particle size of 25 nm and E2 has an average particle size of 35 nm. However, it should be noted the role of agglomeration between particles present in sample D2, which also has had an effect to increase the surface roughness of the thin film. Those samples produced without surfactant have a reduced root mean square surface roughness in the range of 9–17 nm. There is no significant change in surface roughness depending on which surfactant type is used, however generally it is found that an increased concentration of surfactant increases the surface roughness. A 3D representation of the thin film surface roughness of samples produced with different surfactant types and concentrations are shown in Figure 4.

**Figure 4.** AFM 3D representation of thin film surface. (**a**) A2 (no surfactant); (**b**) B2 (Tween® 40, 0.006 mol·dmí<sup>3</sup> ); (**c**) C2 (Tween® 40, 0.003 mol·dmí<sup>3</sup> ); (**d**) D2 (Tween® 20, 0.006 mol·dm<sup>í</sup><sup>3</sup> ); (**e**) E2 (Tween® 20, 0.003 mol·dmí<sup>3</sup> ).

(**e**)
