Experimental Section

3.1 Materials

Materials for TiO₂ paste

1. Titanium dioxide Degussa P25 powder

2. Poly(ethylene glycol)-block- Poly(propylene glycol)-block- Poly(ethylene glycol) (P123) from Aldrich as a binder in TiO₂ paste and the chemical structure is :

3. Polyethylene glycol (PEG) with molecular weight 35,000 and 100,000 from Aldrich as a morphology controller in TiO2 paste and the chemical structure is:

4. N-butanol (C₄H₉OH) from ECHO as the solvent for TiO₂ paste:

5. Titanium dioxide 100nm nanoparticles from ISK for light-scattering layer

Materials for DSSCs

1. Titanium tetrachloride (TiCl₄) from SHOWA for post-treatment of TiO₂ film 2. Fluorine-doped tin oxide (FTO) conducting glass (8Ω/sq) from Hartford Glass 3. Ethanol (C2H5OH) from ECHO as a solvent for dye solution and the chemical

structure is:

34

4. Surlyn® (SX1170-60) from SOLARONIX as the spacer and sealing material 5. N719 dye from UniRegion Bio-Tech and the chemical structure is:

Materials for electrolyte

1. Lithium iodide (LiI) from MERCK 2. Iodine (I2) from SHOWA

3. 1-methylbenzimidazole from Alfa Aesar and the chemical structure is:

4. Guanidine thiocyanate from Alfa Aesar and the chemical structure is:

5. 1-Methoxypropionitrile from Alfa Aesar as a solvent for the electrolyte and the chemical structure is:

35

3.2 TiO

Paste Preparation

3.2-1 TiO₂ paste composition

The TiO₂ paste consist of commercial TiO₂ nanoparticles (P25 powder, Degussa), poly(ethylene glycol)-block-Poly(propylene glycol)-block-Poly(ethylene glycol) (P123) as binder, and polyethylene glycol (PEG) as porogen with butanol or water as solvent. In this study, we change the loading amount of PEG and different solution to control the pore morphology of TiO₂ film. The composition of varies TiO₂ pastes are shown below:

Table 3-1 TiO2 paste composition in butanol solution

0% PEG 5% PEG 10% PEG 15% PEG 20% PEG

TiO2 (P25) 0.3g

P123 0.1g

PEG (35k or 100k) 0g 0.1g 0.2g 0.4g 0.6g

H2O 0.1g

Butanol 1.2g 1.4g 1.4g 1.6g 1.6g

Table 3-1 shows TiO2 paste composition in butanol solution. The water addition is for PEG dissolution because PEG is hardly dissolved in butanol. The difference of butanol amount is for adjusting the paste viscosity to control the TiO2 film thickness at coating process. The PEG loading is set to 0%, 5%, 10%, 15% and 20% with two different molecular weights of PEG, 35,000 and 100,000. Table 3-2 shows TiO2 paste composition in water solution. TiO₂ pastes with both kinds of PEGs. And the paste of TiO2 scattering layer was composed of 0.3g 100nm TiO2 powder and 0.3g P123 in 2.4g n-butanol solvent.

36

Table 3-2 TiO2 paste composition in water solution

TiO2 (P25) P123 PEG (35k or 100k) H2O

15% PEG (35k) 0.3g 0.1g 0.4g 1.5g

15% PEG (100k) 0.3g 0.1g 0.4g 1.5g

3.2-2 TiO₂ paste mixing process

The uniform TiO2 paste is important. The TiO2 nanoparticles might aggregate to form large particles which would lose their high surface area and cause TiO2 film crack after film coating. The mixing procedures of the paste are:

1. Mixing PEG, P123 and the solution (butanol and water) and ultrasonic disperse to the mixture mixed well.

2. Mixing TiO2 and the mixture made at first step with stirring stick in the sample bottle.

3. Ultrasonic dispersing for 2 hours through room temperature water to prevent the solvent evaporation.

4. Stirring the sample by stirring or stirrer bar for 10 minutes.

5. Ultrasonic dispersing for 10 hours through room temperature water.

6. Settling for 1 hour but no longer than 2 hours before use.

37

3.3 TiO

film Preparation

FTO (fluorine-doped tin oxide) conducting glass (8Ω/sq) was cleaned by ultrasonic sieving for 20 min. The TiO₂ paste was coated to film on FTO glass by doctor-blade method. The TiO2 film thickness and active area (0.28 cm²) was controlled by adhesive tape with thickness 110 nm. After coating, the TiO₂ film was dried at room temperature in the air for one minute and removed the adhesive tape.

Then, the TiO₂ anode was sintered at 400℃ for 40 minutes and also for removing the organic loads. Here, we change the PEG burn-out rate which could also control the TiO₂ film morphology. The two different burn-out rates show below:

1. Burn out PEG loading in TiO2 film directly at 400℃ for 40minutes.

2. The TiO₂ film was baked at 100℃ for 30 minutes and the temperature was raised to 400℃ in 15 minutes then burned out PEG loading and sintered for 40min at 400℃.

Backing the TiO2 electrode at 100 ℃ for 30min could let PEG have time and energy to aggregate and form larger size pores after being burned out. After cooling down, the adhesive tape was applied again to define the same area for the scattering TiO₂ layer coating. The scattering TiO₂ layer was deposited by spin-coating with 600 rpm spin speed for 25 seconds and the TiO2 anode was also sintered at 400℃ for 30minutes after remove the adhesive tape. Figure 3-1 shows the schematic diagram of coating process.

38

FTO glass Adhesive tape

Adhesive tape Doctor-blade

TiO2 film Scattering layer

Scattering layer paste (c)

(b)

(a) (d)

Fig. 3-1 Coating process of the TiO2 film. (a) Doctor-blade coating TiO2 film. (b) TiO2 film sintered at 400℃. (c) Spin-coating scattering layer. (d) TiO2 electrode

sintered at 400℃.

The post-treatment was done by immersing the TiO2 anode into the 0.1 M TiCl4

water solution for 30 min in ice bath to form a very thin TiO₂ layer on TiO₂ particles which can improve the charge transfer between TiO2 particles and dye adsorption.[64 93] After TiCl4 treatment, the TiO2 anode was sintered at 400℃ for one hour to crystallize the TiO2 film to anatase phase.

The samples for BET analysis were also coated by doctor-blade method, and sintered by the same process. But the scattering layer and the post-treatment were not done for these samples. The TiO₂ films was scratched off from the FTO glass, and collected to certain amount for BET analysis.

39

3.4 DSSCs fabrication

The TiO2 photoanode was immersed in a 3×10^-4M N719 dye ethanol solution at room temperature for 24 hours for dye adsoption. After sensitized, the TiO₂ photoanode was dip into ethanol to remove extra dye which did not adsorb on the TiO₂ surface. DSSC was fabricated by sealing the dye-sensitized TiO₂ photoanode and Pt-sputtered counter electrode around 100℃ with hot melt sealing foil. The hot melt sealing foil was also a 60 μ m spacer. There are two little holes on the Pt- sputtered counter electrode for electrolyte injection. The electrolyte composition was 0.5M LiI, 0.05M I₂, 0.2M 1-methylbenzlmidazole and 0.5M guanidine thiocyanate in 1-Methoxypropionitrile solvent and was injected into the two little holes on the counter electrode. Extra electrolyte was removed and the two little holes were sealed also by the hot melt sealing foil with a normal glass. Figure 3-2 shows the schematic diagram of DSSC fabrication.

40

Fig. 3-2 The DSSC fabrication process. (a) Sealing the dye-sensitized photoanode and Pt-coated counter electrode. (b) Electrolyte injection. (c) Sealing the electrolyte

injection holes on counter electrode.

(a)

(c) (b)

FTO conduction glass

Pt-coated counter electrode Dye-sensitized TiO2 film

Hot melt sealing foil

Normal glass Hot melt sealing foil

Electrolyte

41

3.5 Characterization Techniques

3.5-1 Brunauer–Emmett–Teller (BET) Analysis

The pore size, pore size distribution, porosity and surface area of TiO2 film was measured by Brunauer–Emmett–Teller (BET) method, using Surface Area and Pore Size Analyzer (NOVA 1000e). The measurement point setting was 0-0.3 P/P0 17 points, 0.3-1 P/P₀ 7 points, 1-0 P/P₀ 15 points.

3.5-2 Scanning Electron Microscopy (SEM)

SEM is a powerful microscopy method for close viewing of nanoparticles. The microstructure of TiO2 films was studied by a scanning electron microscopy (SEM, HITACHI-S2500 JSM-6500F). The TiO₂ film was coated on normal glass for easier SEM sample preparation. The porous TiO2 film was observed by SEM including nanoparticle TiO₂ film and light-scattering layer. And the film thickness also defined by SEM observation.

3.5-3 Uv-Visible Light Spectrum

The amount of dye adsorption was determined by desorbing the dye from TiO2

film surface into 5mM NaOH aqueous solution and measuring its light absorbance by Uv- visible light spectrometer.The dye adsorption of TiO2 film could calculated by comparing the absorption intensity of dye desorbed NaOH solutions and a reference NaOH solution, and a reference dye solution with concentration 8×10^-5M was used for calculation.

42

3-5-4 Phtocurrent Examination

The AM1.5 solar simulator (Newport 3A) was used as the light source, and the incident light spectrum was AM1.5, 1 sun (100mW/cm²) calibrated with standard Si solar cell (ORIEL). The I-V curve was recorded with Keithley by scanning DSSC from -0.2V to 0.8V, and the photoelectrochemical characterizations of DSSCs were carried out by computer calculation with the active area 0.28cm².

3-5-5 Eectrochemical Ipedance Sectroscopy (EIS)

The impedances between the interfaces inside DSSCs were measured by electrochemical impedance spectroscopy (EIS) and presented in the form of the Nyquist plots. The EIS measurements were performed using AC impedance (PGSTAT100 AUTOLAB, Netherlands) over the frequency range from 0.01 to 10⁶ Hz with amplitudes 10mV. The results from EIS were then compared and disused with the efficiency results to provide a more specific explanation of the influence of the changes of TiO2 films.

43

3.6 Experimental Flow

The well mixed TiO2 paste was prepared with different solvents and then coating into films by doctor-blade method followed with sintering and PEG burn out process.

The SEM image was taken for thickness pore morphology observation of TiO2 film.

The TiO₂ films were scratched down from the FTO glass for the BET measurement.

And the pore size, porosity and surface area were analyzed. The TiO2 electrode was immersed into the dye solution for dye adsorption, then the dye desorbed into NaOH water solution for Uv-visible measurement. By Uv-visible spectrum, the dye adsorption amount could quantify after calculation. The TiO₂ electrodes with light-scattering layer after TiCl4 post-treatment were dye-sensitized and fabricated into DSSCs for efficiency measurement and EIS analysis. The relationship between pore morphology and DSSC performance would discuss.

Fig. 3-3 Experiment Design chart

44