Electrolyte Concentration Effect of a Photoelectrochemical Cell Consisting of TiO_𝟐 Nanotube Anode

Abstract

The photoelectrochemical responses of a TiO₂ nanotube anode in ethylene glycol (EG), glycerol, ammonia, ethanol, urea, and Na₂S electrolytes with different concentrations were investigated. The TiO₂ nanotube anode was highly efficient in photoelectrocatalysis in these solutions under UV light illumination. The photocurrent density is obviously affected by the concentration change. Na₂S generated the highest photocurrent density at 0, 1, and 2 V bias voltages, but its concentration does not significantly affect the photocurrent density. Urea shows high open circuit voltage at proper concentration and low photocurrent at different concentrations. Externally applied bias voltage is also an important factor that changes the photoelectrochemical reaction process. In view of the open circuit voltage, EG, ammonia, and ethanol fuel cells show the trend that the open circuit voltage (OCV) increases with the increase of the concentration of the solutions. Glycerol has the highest OCV compared with others, and it deceases with the increase in the concentration because of the high viscosity. The OCV of the urea and Na₂S solutions did not show obvious concentration effect.

1. Introduction

Titanium Na₂S dioxide (TiO₂) has been widely studied because it has good photovoltaic property. Photoelectrochemical cell (PEC) is a device that could degrade pollutants, splitting water by utilizing photon energy while electric energy is generated. Clean energy generation by environmentally friend method is a very important issue. Hydrogen is used in fuel cells, but nature gas is still a main source for hydrogen production. It contains contaminative byproducts. Only a few part of hydrogen is produced by water splitting in the world [1]. Pollutant degradation is another very important application of PEC. TiO₂ nanostructured anode has the high potential for clean energy production and hazard material degradation. Many hazardous organic materials can be converted into clean substances by TiO₂ anode PEC, which include methylene blue [2, 3], glucose [4], organic compounds [5, 6], waste water [7], dye pollutant [8], and even CO₂ [9] of the greenhouse gas.

By using various electrolytes, different levels of open circuit voltage, current density, filling factor can be reached [10–12]. The bias potential and concentration of electrolytes can also influence the performance of a PEC cell. Normally, the additionally applied external bias voltage can further enhance the PEC reaction process because the conduction band edge of anode material cannot lie above the energy level. Quan et al. [13] selected 0.0, 0.2, 0.4, and 0.6 V bias and 0, 0.005 M, and 0.01 M Na₂SO₄ electrolyte used in PEC and study on the effect. Normally, the higher bias and concentration enhanced the PEC process. But not all of the high concentration can benefit the fuel cell. There is a watershed for concentration effect on PEC. When the concentration reached a level, performance of PEC will be stabilized [14]. Some papers mentioned the concentration effect on the performance of PEC, but the concentration range is all within a small range from 0.001 M to 0.5 M [4, 15, 16]. Such a small range concentration has limitation. A wide range (0 to 100% or statured) of concentration would allow us to know this effect better. Milczarek et al. [17] studied the concentration effect of sulfuric acid within the range from 0.5 to 5 M. No significant effect due to the change in the sulfuric acid concentration was found. Most of people would like to choose 1 M H₂SO₄ as an optimal concentration for PEC. The iodine concentration effect was studied by Hao et al. [18]. When the iodine concentration increased from 0.025 to 0.1 M, OCV and photocurrent density decreased, while the fill factor increased. The best concentration of electrolyte can enhance the performance of PEC. But different solution shows different concentration effect.

In this paper, we selected three main types of solutions including alcohols, polyols, and some pollutant materials. Different with others’ research, the concentration range is broader, which is from 0 to saturated. The purpose of this study is to find the best concentration for each solution and compare the different processes of these electrolytes in PEC.

2. Experimental

2.1. Chemicals

All chemicals were used as received without further purification. Ti foil (99.9% purity, 0.127 mm), ethanol (99% purity), ethylene glycol (99% purity), glycerol (99% purity), urea (99.9% purity), Na₂S·9H₂O (98% purity), NH₄OH (30% wt. NH₃), H₃PO₄ (85% wt in water), and NaF (98% purity) were purchased from Alfa Aesar.

2.2. TiO₂ Nanotube Fabrication

Electrochemical anodization method was applied to synthesize self-organized TiO₂ nanotube by using titanium foil (3 mm × 40 mm). Before electrochemical anodization, the titanium foil was cleaned using acetone, ethanol, and distilled water, and dried in air stream. TiO₂ nanotube was fabricated in 1 M H₃PO₄ and 0.2 M NaF solutions where the Ti foil was used as anode and a platinum wire was used as cathode. The voltage and time for anodizing Ti are 20 V and 19 hours. The nanotube was cleaned by using distilled water and the samples were dried in air. Anodization was performed at 25°C. The TiO₂ nanotube was annealed in the air at 450°C for 4 hours to convert the amorphous structure of TiO₂ into anatase crystalline structure.

2.3. Fuel Concentration

We used ethanol, and ethylene glycol, glycerol, urea, Na₂S, ammonia as the fuels. The concentrations of ethylene glycol, glycerol, and NH₄OH were from 0 to 100% wt. The concentrations of urea were from 0 to 108 g/mL water (saturated), and those of Na₂S were from 0 to 20 g/mL water (saturated) at 25°C.

2.4. Photoelectrocatalytic Response

A photoelectrochemical fuel cell was made using the TiO₂ NTs anode, Pt wire cathode. Ethanol, ethylene glycol, glycerol, urea, Na₂S, and ammonia were used as the fuels, and the ultraviolet lamp was the UVL-21 (365 nm UV, 4 W, 0.16 Amps) with the illumination intensity of 40 mW/cm². The water bath is used to keep the temperature as constant. A CHI 400A electrochemical workstation was used to measure the open circuit voltage and current density of the fuel cell and supply the bias potential (Figure 1).

Figure 1 

Schematic diagram of the PEC system.

3. Results and Discussion

3.1. TEM Image Analysis

TiO₂ nanotube specimen was examined by transmission electron microscopy (TEM) using the JEOL 2100F TEM. Figure 2 is the TEM image of a typical TiO₂ nanotube. The TiO₂ nanotube was scraped off from the Ti substrate before examination on the TEM. The length of nanotube is 770 nm and the diameter is 200 nm.

Figure 2 

TEM image of a TiO2 nanotube.

3.2. Photocurrent Density

For testing the open circuit voltage, we used the two-electrode method (Figure 1). Compared with the method using the Ag/AgCl reference, the two electrode approach is more stable because there is no unwanted chemical reaction during testing. In order to study fuel concentration effect on the PEC process, we selected five solutions which contained three major substances, alcohols, polyols, and pollutant materials. These materials show different PEC properties. For different fuels, the reaction mechanisms are different [19]. Ethanol is an alcohol substance, and the reaction mechanisms are as follows (also shown in Figure 3): C₂H₅OH+2h⁺⟶CH₃CHO+2H⁺CH₃CHO+H₂O+2h⁺⟶CH₃COOH+2H⁺CH₃COOH⟶CH₄+CO₂ (1) The total reaction is C₂H₅OH+H₂O+4h⁺⟶2CO₂+CH₄+4H⁺ (2) At the anode, TiO₂ is excited under light energy to form holes and electrons on the surface. Holes will react with the fuels such as ethanol (Figure 3), and water and the reactions can generate CO₂, CH₄ and H⁺. H⁺ flows to cathode through the electrolyte. The higher transfer rate of H⁺ and e⁻ will result in higher current density. The external bias can increase the potential energy of system and the flow rate of electrons and H⁺. So the reaction rate increases with the bias potential level.

Figure 3 

Mechanism of PEC using ethanol as the fuel.

For the polyols substance such as glycerol, the reaction is C₃H₈O₃+3H₂O+14h⁺⟶3CO₂+14H⁺ (3) For same amount of molecules, glycerol generates more CO₂ and H⁺ than ethanol does.

For the ammonia, the chemical reaction at photo anode is as follows: NH₃+3h⁺⟶12N₂+3H⁺ (4) The Na₂S oxidative process is following the reaction 2S²⁻+2h⁺⟶S₂²⁻ (5) where h⁺ is the hole formed on TiO₂.

Bias potential draws the induced electrons to the counter electrode. The higher value of the bias, the higher the driver force of the system. A larger photocurrent density should be obtained. We have the same conclusion from the experimental studies as verified by the results from Figure 4 to Figure 9; the highest photocurrent is obtained at 2 V bias and the lowest is obtained at 0 V bias.

Figure 4 

Photocurrent density versus concentration of ethanol at different bias potentials.

The positive and negative ion transfer velocities (𝑣₊ and 𝑣₋) increase with the electric field intensity 𝐸⋅𝑣₊ = 𝜇₊𝐸 and 𝑣₋=𝜇₋𝐸⋅𝜇 is the ion transfer rate. A high bias voltage can supply a high value of 𝐸. The conductivity of the solution is 𝜎=𝑛𝑒𝜇₊+𝜇₋, (6) where 𝜎 is the conductivity, 𝑛 is the ion concentration, and 𝑒 is the absolute value of electronic charge.

Current density increases with the increase in the ion concentration and mobility. But the mobility of ions increases with decreasing in the solution viscosity. Since glycerol (Figure 4) is a high viscosity substance, with the increase in concentration of glycerol, the high viscosity will impede the ion transfer process.

The alcohols, polyols fuels have the similar performance (Figures 4–7). But the glucose has an opposite trend [4]. The glucose and alcohols are similar organic materials which cannot conduct electrons. But why they show different properties? The reason is due to the fact that the glucose concentration range is small. They only picked the range from 0 to 6 mM/L. If they select the range from 0 to statured, the result may be changed. Alcohols and polyols are organic substances and the polarity of them is weak. The pure alcohols do not have conductivity. With the water content increase, the water is ionized and the electron conductivity is improved. The photocurrent density of ethanol is the lowest. Compared with ethylene glycol and glycerol, the value of photocurrent density by using ethanol fuel is about 50% lower than that of others. Viscosity is another important factor affecting the microscale diffusion. Current is the density of ions in motion. It is very similar with a concentration flux of mass; that is, Fick’s law is satisfied as follows: 𝐽(𝑥,𝑡)=𝑐(𝑥,𝑡)𝑣(𝑥,𝑡), (7) where 𝐽 is the current density, 𝑐 the concentration, and 𝑣 the ion transfer rate.

Figure 5 

Photocurrent density versus concentration of ethylene glycol at different bias potentials.

Figure 6 

Photocurrent density versus concentration of glycerol at different bias potentials.

Figure 7 

Photocurrent density versus concentration of urea solution at different bias potentials.

Some solutions with high concentrations have low ion transfer rates, such as alcohols, polyols solutions. These substances are needed to be in electrolyte solutions. The pure alcohols and polyols are not ideal as fuels for PEC. If the concentration increases, the attractive force between the ions will constrain the ion transfer. The micromass transfer and diffusion of alcohol and polyol molecules limit the reactions on the surface of TiO₂ nanotubes. There is a diffusion layer between the TiO₂ nanotube surface and electrolyte. With the concentration being increased, the ionization of alcohols and polyols is decreased. The transfer rate in the diffusion layer is low and causes the photocurrent density to be low. Figure 4 shows 𝐼 versus 𝑐 curves of ethanol with different bias. The highest photocurrent is at 1 wt.% which is even lower than the case of pure water as the fuel. The photocurrent density of ethylene glycol is higher than that of pure water (Figure 5). The maximum photocurrent is obtained at 5% wt. of ethylene glycol. The glycerol shows different 𝐼 versus 𝑐 characteristic from ethanol and ethylene glycol. The photocurrent density of glycerol is constant and high 𝑡 0–25 wt%. When the concentration exceeds 25%, the photocurrent decreases sharply.

Urea, Na₂S, and ammonia are electrolytes that contain lager amounts of free ions. The conductivity of urea, Na₂S, and ammonia is higher than that of alcohols and polyols. In the experiments, the concentration of urea, Na₂S, and ammonia is selected from zero to saturation levels. The current density almost has no change with the concentration variations (Figures 7–9).

Figure 8 

Photocurrent density versus concentration of Na2S at different bias potentials.

Figure 9 

Photocurrent density versus concentration of NH4OH at different bias potentials.

Different solutions showed different photocurrent densities. Na₂S solution has the highest photocurrent density which is about 100, 300, or 450 mA/m² at 0, 1, or 2 V bias. Generally, ionized solutions have higher photocurrent density than alcohols and polyols. Ethanol has the lowest photocurrent density which is about 10, 75, or 150 mA/m² at 0, 1, or 2 V bias. This photocurrent density generated by ethanol is even lower than pure water. The trends of 𝐼 versus 𝑐 curves of Na₂S and ammonia are dissimilar to others. The photocurrent almost showed no change with the increase of the concentration (Figures 8 and 9). The ions mobility is already high, and the high concentration cannot improve it further in Na₂S and ammonia solutions.

3.3. Kinetics Study of Influences on Photocurrent Responding

Figures 4–6 show the photocurrent density obtained from different electrolytes and at various bias potential levels using alcohols, polyols as fuels. The photocurrent density decreases with the concentration increasing. Computer simulation reveals that the 𝐼 versus 𝑐 curve of ethylene glycol fits. 𝐼=−𝑅₁×𝑐+𝑅₂, (8) where 𝑅₁ and 𝑅₂ are constants and 𝑐 is the concentration. For ethylene glycol at 2 V bias (Figure 10), the process can be divided into two parts (0–5 wt.% and 5–100 wt.%). For 0–5 wt.%, the 𝑅₁, 𝑅₂ are −11.5 and 218.25. For 5–100% wt., the values of 𝑅₁, 𝑅₂ are 2.85 and 290, respectively. 𝑅₁ and 𝑅₂ are related to the bias. With the increase of the bias, the values of 𝑅₁ and 𝑅₂ increased. Higher values of 𝑅₁ and 𝑅₂ also measure present higher kinetic energies of these systems.

Figure 10 

Computing results versus experimental data (𝐼 versus 𝑐) in ethylene glycol at 2 V.

The 𝐼 versus 𝑐 curve of urea is not linear (Figure 7). We found that the 𝐼 versus 𝑐 curve follows 𝐼=−𝑃₁×(𝑐−25)¹⁶+𝑃₂, (9) where 𝑃₁ and 𝑃₂ are two parameters. For urea at 2 V bias (Figure 11), the process may be divided into two parts (0–32.61 g/mL and 32.61–50 g/mL). For concentration of 0–32.61 g/mL, 𝑃₁, 𝑃₂ are 4.5𝐸−21 and 315, respectively. For concentration of 32.61–50 g/mL, 𝑃₁, 𝑃₂ are 1.2𝐸−21 and 315, respectively. 𝑃₁ determined the shape of the plot and 𝑃₂ is related to the maximum value of the photocurrent density. We can use these equations to find any concentration versus photocurrent of urea electrolyte.

Figure 11 

Computing results versus experimental data (𝐼 versus 𝑐) in urea at 2 V.

The behavior of Na₂S and ammonia can simply be expressed by 𝐼=𝑄, (10) where 𝑄 is a constant. There is no obvious concentration effect on the photocurrent density for these solutions.

3.4. Photo Voltage

For the open circuit voltage (OCV) test, we also used the two-electrode method. The trend of OCV is that with the increase in concentration of solution, the OCV is increased except for glycerol solution (Figure 12). Glycerol is a high viscosity solution with low ion mobility. When the concentration of glycerol reaches 90%, the OCV is shapely decreased. So, 90% glycerol is a maximum concentration to obtain a reasonable OCV from the system. Glycerol has the maximum photovoltage compared with others and ethanol always has the minimum photo voltage (Figure 12). EG, glycerol, ammonia, and ethanol show a constant voltage with the increase in concentration. But the photo voltage of urea and Na₂S show more fluctuations than others (Figure 13). There are larger concentration effects for urea and Na₂S solutions than for EG, glycerol, ammonia, and ethanol.

Figure 12 

Photovoltage versus concentration of ethylene glycol, glycerol, and ammonia.

(a)

(b)

Figure 13 

(a) Photo voltage versus concentration of urea. (b) Photo voltage versus concentration of Na2S.

4. Conclusions

The kinetics and mechanism of TiO₂ nanotube photoelectrochemical cells with different concentrations of ethanol, EG, glycerol, ammonia, urea, and Na₂S were studied. These solutions can be divided into alcohols, polyols, and electrolyte pollutants three major types. The TiO₂ nanotube anode was highly efficient in photoelectrocatalysis in these solutions under UV light illumination. The concentration effect on kinetic of the PEC with alcohols and polyols is obvious. With the concentration being increased, the photocurrent density was decreased. The reason for this is that the viscosity and mobility of solutions changed at different concentrations. For the concentration effect on electrolyte pollutant materials such as urea, Na₂S, and ammonia, urea has large photocurrent density at middle range (5–45 g/100 mL) and has low photocurrent at low or high concentration regime. Na₂S showed the maximum photocurrent density at 0, 1, 2 V bias compared with others, and the concentration of Na₂S does not affect the value of photocurrent density. External bias voltage is another significant factor affecting the PEC. Obviously, the high bias leads to a large photocurrent. The photocurrent increasing rate was decreased with the increase in the bias potential. The relationship of 𝐼 versus 𝑐 can be defined by 𝐼=−𝑅₁×𝑐+𝑅₂ for alcohols, polyols, and 𝐼=−𝑃₁×(𝑐−25)¹⁶+𝑃₂ for urea by selecting different constant 𝑅₁, 𝑅₂, 𝑃₁, and 𝑃₂. Since the behavior of Na₂S and ammonia shows a horizontal line, it can be expressed as 𝐼=𝑄.

In view of the open circuit voltage, EG, ammonia, and ethanol show OCV increasing with the increase of the solution concentration. Glycerol has the maximum OCV and it decreased at the high concentration because of the high viscosity. Urea and Na₂S solutions did not show obviously concentration effect.

Acknowledgments

The support from the United States Environmental Protection Agency (EPA) under Grant no. SU83529701 is gratefully acknowledged. The transmission electron microscopic (TEM) research carried out at the Center for Functional Nanomaterials, Brookhaven National Laboratory, was supported by the US Department of Energy, Office of Basic Energy Sciences, under Contract no. DE-AC02-98CH10886.