Ag Fibrous Electrode by the

18686

J. Phys. Chem. C 2008, 112, 18686–18689

Fabrication of High Conductivity TiO2/Ag Fibrous Electrode by the Electrophoretic Deposition Method Z. Hosseini,† N. Taghavinia,*,†,‡ N. Sharifi,‡ M. Chavoshi,† and M. Rahman‡ Physics Department, Sharif UniVersity of Technology, Tehran 14588, Iran, and Institute for Nanoscience and Nanotechnology, Sharif UniVersity of Technology, Tehran 14588, Iran ReceiVed: May 25, 2008; ReVised Manuscript ReceiVed: September 7, 2008

TiO2 deposited on a membrane of Ag fibers was prepared as a photoelectrochemical cell electrode. Ag fibers were made by reduction of Ag complexes on cellulose fibers, followed by burning out the template. TiO2 photocatalyst layers were grown on the fibers by electrophoretic deposition of TiO2 nanoparticles. Ag fibers could be uniformly deposited. Photocatalytic tests by dye decomposition and electrochemical impedance spectroscopy (EIS) under UV illumination demonstrate that Ag fibers act as a good substrate that provides both high surface area and good separation of photogenerated electron-hole pairs and causes the enhancement of photocatalytic activity in comparison with a thin film of TiO2. 1. Introduction Since the breakthrough of wet photoelectrochemical (PEC) solar cells by Gra¨tzel and co-workers,1-4 a great deal of attention has been paid to nanoporous-nanocrystalline semiconductor electrodes. Among many semiconductors, TiO2 is widely used because it is nonphotocorrosive, nontoxic, and capable of photooxidative destruction of most organic pollutants.5-9 The photoelectrocatalytic efficiency of the nanocrystalline TiO2 system is strongly influenced by two key factors. First is the specific surface area of the electrode, which influences the efficiency of the heterogeneous catalytic reaction through more photoelectrons and photoholes reaching onto the surface.10-13 The use of fibrous structures is considered as one of the appropriate ways to achieve high surface area. The second key factor is the electron transport efficiency. The photogenerated electron-holes (e-h) in the semiconductor are likely to recombine, before they can cause photocatalytic reactions on the surface. To increase electron transport efficiency, the photogenerated electrons should be scavenged and guided into the circuit.14 This is related to electrode conductivity, which plays an important role in the efficiency of photoelectrochemical cells. Noble metal nanoparticles, such as Pt, Au, and Ag, loaded on the surface of TiO2 photocatalysts cause e-h pair separation through electron scavenging,14-17 as their fermi level is below the conduction band of TiO2. e-h separation through metal nanoparticles loading enhances the photo-oxidation efficiency on the TiO2 surface; however, to be useful in photoelectrochemical cells, the scavenged electrons should find a way out into the circuit to cause a reduction reaction on the counter electrode. This has led us to a structure consisting of Ag fibers deposited with TiO2 nanoparticles. The photogenerated electrons are expected to be easily scavenged into the neighboring Ag and be transferred into the circuit through low-conductivity Ag fibers. * Corresponding author. E-mail: [email protected]; +982166164532, +982166164570. † Physics Department. ‡ Institute for Nanoscience and Nanotechnology.

Tel:

We have employed the electrophoretic deposition (EPD) method to load TiO2 nanoparticles on Ag fibers. EPD is a costeffective technique for fabrication of coatings from particulate materials18 and is applicable to complex component shapes and large dimensions.18-20 Therefore, it is well suited for deposition on complex-shaped Ag fibers. The deposition can be seen as two combined processes: first the migration of charged colloidal nanoparticles under the applied electric field (electrophoresis) and second the coagulation process of the nanoparticles at the electrode. To obtain a homogeneous deposition, employing a well-dispersed and stable system is essential.18,21 2. Experimental Ag fibers were prepared as follows: 2 mL of ammonium hydroxide (Guangdong Guanghua, 25%) was slowly dropped into 10 mL of a 0.5 M AgNO3 (Wako) solution. The solution initially turned turbid and then became transparent. An amount of 10 mL of a 1 M KOH solution was added to the previous solution, and the solution became dark brown. It was made transparent again by addition of 1 mL of ammonium hydroxide. Filter paper (36 cm2) (Whatman 1001) as a template was put into the resulting solution. An amount of 10 mL of an aqueous solution of 0.7 M sucrose (C12H22O11, Merck) at pH 2 (by HNO3) was then added. Silver complexes were reduced on the surface of fibers, and a thin layer of silver formed on the template. The coated template was then removed and washed with deionized (DI) water. Finally, the template was burned out at 400 °C in air, resulting in the formation of conductive Ag fibers. The formed membrane of Ag fibers was fixed on a stainless steel sheet using conductive carbon tape. This structure was used as the substrate for electrophoretic deposition of TiO2 nanoparticles. Colloidal suspension of TiO2 nanoparticles was synthesized by mixing titaniumtetraisopropoxide (Merck), H2O2 (Merck), and H2O, with volume proportions of 12:90:200, respectively. The resulting solution was refluxed for 10 h to promote the crystallinity. The counter electrode (cathode) was stainless steel. The distance between two electrodes and the surface of the electrodes was 1 and 1 cm2, respectively. The electrodes were placed horizontally, and the substrate was the upper electrode. Electrophoretic deposition was carried out

10.1021/jp8046054 CCC: $40.75  2008 American Chemical Society Published on Web 11/04/2008

High Conductivity TiO2/Ag Fibrous Electrode

Figure 1. TEM image of TiO2 nanoparticles. The particles are elongated with average long and short axis sizes of 54 and 15 nm.

at a current density of 500 µA/cm2 and deposition time of 90 s, unless clearly quoted. Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) measurements were performed on a Philips XL30 machine. The surface composition of the silver fibers was characterized by X-ray photoelectron spectroscopy (XPS) with Specs EA 10Plus analyzer using Al KR (E ) 1486.6 eV) radiation. Binding energy (BE) scales were referenced at 285.0 eV, as determined by the location of the C 1s peak for a hydrocarbon (CHn) associated with adventitious contamination. The observation of TiO2 nanoparticles was made by transmission electron microscopy (TEM, Philips CM200). The zeta potential was measured by Malvern Zetasizer ZS. Photocatalytic activity tests were carried out by photocatalytic degradation of methylene blue (MB) in the presence of a TiO2 layer or TiO2/Ag fibers under UV illumination. The sample was put horizontally in 5 mL of a 5 µM MB solution with constant stirring for one hour in the dark, and then UV irradiation was started. After each 10 min illumination, the concentration of MB was determined by monitoring its 664 nm absorption peak (AvaSpec 2048-TEC UV-vis spectrometer). Electrochemical characterization of EIS Nyquist plots were performed in a three electrode system (Zahner elektrik IM6ex) using Ag/AgCl as the reference electrode, Pt as the counter electrode, and TiO2 thin film or TiO2/Ag membrane as the working electrode. Na2SO4 (0.5 M) aqueous solution was used as the electrolyte. Analysis of the EIS data was performed using ZView2 software. For illumination of samples in photoelectrochemical tests, an 8 W low pressure mercury lamp was employed. The intensity at the sample was 11 mW/cm2. The frequencies for EIS measurement were scanned from 104 Hz down to 0.01 Hz. 3. Results and Discussion The ion conductivity of the synthesized TiO2 sol was 200 µS/cm. Zeta potential of the particles was measured to be -50 mV, which is sufficiently large to keep the suspension stable. Figure 1 shows typical TEM images of nanoparticles. The particles are elongated with average long and short axis sizes of 54 and 15 nm. The crystalline phase has been verified as pure anatase using X-ray diffraction. Anatase is the photocatalytically most active crystalline phase of TiO2. The filter paper used as the template consists of fibers of predominantly cellulose composition, with an average diameter of 14 µm. The surface of template fibers after being coated by silver is shown in Figure 2a. The film has a granular morphology with grains of about 200 nm average size. After 30 min heat treatment in 400 °C, the organic template burns out and Ag fibers form. SEM images of Ag fibers are shown in Figures 2b and 2c. The structure follows, to some extent, the structure of the template, where tubular shapes with an average diameter of 10 µm can be observed. The Ag grains are about 1 µm in size,

J. Phys. Chem. C, Vol. 112, No. 47, 2008 18687 which demonstrates the sintering of the original Ag grains during heating. This provided good conductivity of the fibers which is needed for a successful EPD process, as well as good transport of photogenerated electrons of TiO2. The XPS spectra of Ag fibers obtained after heat treatment (Figure 2d) show that Ag 3d5/2 and Ag 3d3/2 peaks are located at 368.32 and 374.27 eV, respectively. These demonstrate no chemical shift and confirm the presence of nonoxidized Ag metal on the fiber surface.15,22 TiO2 nanoparticles were deposited on the Ag fibers using the EPD process. Penetration and deposition of TiO2 nanoparticles onto Ag fibers is shown in SEM images of Figure 3. Comparing Figures 2c and 3a, one notes that the Ag grains are covered by TiO2 nanoparticles, demonstrating a good surface coverage. Identical morphology in these two images demonstrates that the thickness of TiO2 coating is a few hundred nanometers. Backscattered electron image in Figure 3b at a larger scale also reveals the presence of a uniform coating of TiO2 on the fibers, as no Ag-TiO2 contrast can be seen. It was observed using EDS analysis that TiO2 loading on fibers linearly increases with increasing deposition time and current. However, at high deposition times and currents, it was difficult to avoid growing of TiO2 particle clusters (TiO2 film) on the surface of the fiber membrane. This effect is shown in Figure 4a for a typical condition. The field-induced flux of nanoparticles at high deposition currents is sufficiently large to jam pores near the surface of the fiber membrane,23 and this results in the deposition of a layer with a finite thickness covered on the surface of the fiber membrane. For 90 s deposition time, this effect was observed for current densities greater than 1000 µA/cm2. Figure 4b shows the results of EDS analysis for the amount of TiO2 loading on the inner parts of the fiber membrane for different currents. The decline of TiO2 deposition at higher currents is caused by the clogging of the membrane surface due to film formation. Current densities below 1000 µA/cm2 and deposition times shorter than 90 s were found ideal for the formation of uniform deposition of fibers. The degradation of MB was used as a test to verify the photocatalytic activity of TiO2/Ag fibers in comparison with TiO2 thin films. The TiO2 thin films were prepared using EPD on a flat stainless steel substrate, with conditions similar to TiO2/ Ag samples. The photocatalytic decomposition rate was calculated as ln(C0/C), where C0 is the initial concentration of MB solution and C is the concentration after each period of UV illumination. As shown in Figure 5, all the TiO2 thin films prepared at different conditions show similar photocatalytic activity which indicates that the deposition condition (deposition current and deposition time) in this range does not significantly change the photocatalytic activity. The same is true for TiO2/ Ag fibers, where all samples show similar photocatalytic activity. Also from Figure 5, it can be concluded that TiO2/Ag fibers show about a 1.7 times higher decomposition rate constant in comparison with TiO2 thin films. It is well-established that EIS Nyquist plots are associated with the charge transfer resistance and the separation efficiency of the photogenerated e-h pairs.24 Figure 6 shows the results of EIS Nyquist plots of TiO2/Ag fibers as well as the TiO2 thin film on a stainless steel substrate in the dark and under UV illumination. EIS data were analyzed in terms of an equivalent circuit model. By fitting the impedance spectrum to a model of an equivalent electrical circuit, useful information on the electrical response of the working electrode can be obtained. The equivalent circuit shown in Figure 7 has been proved effective in simulating this structure.24 Here, Rs is the solution

18688 J. Phys. Chem. C, Vol. 112, No. 47, 2008

Hosseini et al.

Figure 2. (a) SEM micrograph of Ag coating formed on the surface of the template, (b,c) SEM micrographs of Ag fibers, and (d) Ag 3d XPS spectra of the Ag fiber.

Figure 3. SEM micrographs of TiO2-deposited Ag fibers. Image (b) is a backscattered electron image.

resistance; Rf and CPEf are resistance and capacitance of the film; Rct is electron charge transfer resistance; and CPEdl is the double-layer capacitance. Both Cf and Cdl were replaced with constant phase element (CPE) in the fitting procedure due to the nonideal capacitance response of these interfaces. The impedance of CPE is given by ZCPE ) 1/C0(jω)m, where C0 is the admittance magnitude of CPE and m is an exponential term. Pure capacitance behavior is represented by m ) 1.25 Table 1 shows the results obtained by fitting the Nyquist plots with the equivalent circuit shown in Figure 7 in the absence and presence of UV light. As a simple comparison, the values of Rf, for TiO2/Ag fibers, are significantly smaller than those of TiO2 thin films. This shows that the electrode resistivity for TiO2/Ag fibers is lower, which can be attributed to thinner TiO2 layer thickness. Also, the values of Rf under UV illumination, in both cases, are smaller than those in the dark. The difference in Rf in dark and under UV is a 5-fold difference for TiO2 thin film samples and 20-fold for TiO2/Ag fibers. If the 5-fold difference is attributed to higher concentration of photogenerated electrons and holes in TiO2, the 20-fold difference in TiO2/Ag fibers indicates that more efficient separation of electrons and holes has occurred.24,25 This might be due to smaller TiO2 thickness and effective capture of electrons in the neighboring

Figure 4. (a) SEM micrograph of TiO2/Ag fibers prepared at 1000 µA/cm2 current density for 90 s. (b) Relative weight percent of TiO2 penetrated inside the Ag fiber membrane for different current densities. The data are obtained by EDS.

Figure 5. Comparison of photocatalytic MB decomposition rate for TiO2/Ag fiber samples and TiO2 thin film samples, prepared for 30-150 s deposition times. The deposition time has a slight effect on the photocatalytic activity of each type of sample; however, in general, TiO2/Ag samples show a higher degradation rate.

Ag. The efficient transfer of photogenerated electrons to Ag is expected to cause nonideality in the film capacitance, in the presence of UV illumination. This is manifested by the considerable decrease of mf for TiO2/Ag fibers to about 0.6, under UV illumination. This value is about 0.8 for TiO2 thin

High Conductivity TiO2/Ag Fibrous Electrode

J. Phys. Chem. C, Vol. 112, No. 47, 2008 18689 density less than 1000 µA/cm2. The effectiveness of EPD as a powerful and cost-effective technique for the deposition of TiO2 on complex-shaped fibers was demonstrated. The results show that noble metal fibers can be considered as effective substrates for the deposition of TiO2 photocatalysts, as they possess high surface area, allow easy diffusion of ions, promote the separation of photogenerated e-h pairs, and act as conductive junctions for low-resistivity electrodes. References and Notes

Figure 6. EIS Nyquist plots of the TiO2/Ag fibers and TiO2 thin film prepared in a similar condition. The low impedance part is magnified in the inset.

Figure 7. Equivalent circuit of TiO2 thin film and TiO2/Ag fibers. Rs is electrolyte resistance; Rf and CPEf are resistance and capacitance of film; Rct is charge transfer resistance; and Cdl is double layer capacitance.

TABLE 1: Fitted Results of EIS Spectra for TiO2 Thin Film and TiO2/Ag Fibers TiO2 thin film

TiO2/Ag fiber

parameters

dark

UV

dark

UV

Rs (Ω) Rf (Ω) Cf (F) mf Rct (Ω) Cdl (F) mdl

4.24 530 K 3.45e-5 0.9 2500 0.001 0.93

4.19 101 K 8.16e-5 0.81 1700 0.0004 0.91

4.22 90 1.9e-2 0.94 1260 0.002 0.94

4.29 4.4 1.9e-2 0.6 975 0.015 0.97

film samples. This also evidences better charge transfer efficiency for TiO2/Ag fibers. 4. Conclusion This work has demonstrated the successful application of EPD to manufacture TiO2-coated Ag fibers with good photoelectrocatalytic property. Ag fibers play the role of conductive junctions, as well as enhance the separation of photogenerated e-h pairs. The optimized EPD parameters to manufacture wellcoated fibers were deposition times shorter than 90 s and current

(1) Nazeeruddin, M. K.; Kay, A.; Rodicio, I.; Humphry, B. R.; Mueller, E.; Liska, P.; Vlachopoulos, N.; Gra¨tzel, M. J. Am. Chem. Soc. 1993, 115, 6382–6390. (2) Gra¨tzel, M. Curr. Opin. Colloid Interface Sci. 1999, 4, 314–321. (3) O’Regan, B.; Gra¨tzel, M. Nature 1991, 353, 737–40. (4) Gra¨tzel, M. Nature 2001, 414, 338–344. (5) Fujishima, A.; Hashimoto, K.; Watanabe, T. TiO2 Photocatalysis: Fundamentals and Applications; BKC: Tokyo, 1999. (6) Navio, J. A.; Cerrillos, C.; Pradera, M. A.; Morales, E.; GomezAriza, J. L Langmuir 1998, 14, 388–395. (7) Ranjit, K. T.; Willner, I.; Bossmann, S.; Braun, A. J. Phys. Chem. B 1998, 102, 9397–9403. (8) Hepel, M.; Kumarihamy, I.; Zhong, C. J. Electrochem. Commun. 2006, 8, 1439–1444. (9) Zou, Z. G.; Ye, J. H.; Sayama, K.; Arakawa, H. Nature 2001, 414, 625–627. (10) Yang, S. G.; Liu, Y. Z.; Sun, C. Appl. Catal., A 2006, 301, 284– 291. (11) Mor, G. K.; Shankar, K.; Paulose, M.; Varghese, O. K.; Grimes, C. A. Nano Lett. 2006, 6, 215–218. (12) Zhan, S. H.; Chen, D. R.; Jiao, X. L.; Tao, C. H. J. Phys. Chem. B 2006, 110, 11199–11204. (13) Peng, T. Y.; Hasegawa, A.; Qiu, J. R.; Hirao, K. Chem. Mater. 2003, 15, 2011–2016. (14) He, C.; Xiong, Y.; Chen, J.; Zha, C.; Zhu, X. J. Photochem. Photobiol., A 2003, 157, 71–79. (15) Jin, M.; Zhang, X.; Nishimoto, S.; Liu, Z.; Tryk, D. A.; Emeline, A. V.; Murakami, T.; Fujishima, A. J. Phys. Chem. C 2007, 111, 658–665. (16) Reddy, E. P.; Sun, B.; Smirniotis, P. G. J. Phys. Chem. B 2004, 108, 17198–17205. (17) Zheng, J.; Yu, H.; Li, X.; Zhang, S. Appl. Surf. Sci. 2008, 254, 1630–1635. (18) Besra, L.; Liu, M. Prog. Mater. Sci. 2007, 52, 1–61. (19) Kaya, C.; Boccaccini, A. R. J. Mater. Sci. Lett. 2001, 20, 1465– 1467. (20) Boccaccini, A. R.; Karapappas, P.; Marijuan, J. M.; Kaya, C. J. Mater. Sci. 2004, 39, 851–859. (21) Sarkar, P.; Nicholson, P. S. J. Am. Ceram. Soc. 1996, 79, 1978– 2002. (22) Chiua, Y.; Rambabu, U.; Hsu, M. H.; Shieh, H. P. D.; Chen, C. Y.; Lin, H. H. J. Appl. Phys. 2003, 94, 1996–2001. (23) Chung, Y. W.; Leu, I. C.; Lee, J. H.; Hon, M. H. J. Cryst. Growth 2005, 275, e2389-e2394. (24) Liu, H.; Cheng, Sh.; Wu, M.; Wu, H.; Zhang, J.; Li, W.; Cao, Ch. J. Phys. Chem. A 2000, 104, 7016–7020. (25) Li, M. C.; Shen, J. N. J. Solid State Electrochem. 2006, 10, 980–986.

JP8046054