Water

Available online at www.sciencedirect.com

ScienceDirect Procedia Chemistry 12 (2014) 34 – 40

New Processes and Materials Based on Electrochemical Concepts at the Microscopic Level Symposium, MicroEchem 2013

Relation between morphology and photoelectrochemical performance of TiO2 nanotubes arrays grown in ethylene glycol/water Próspero Acevedo-Peña*, Ignacio González Universidad Autónoma Metropolitana-Iztapalapa, Departamento de Química. Av. San Rafael Atlixco No 186. C.P. 09340. Mexico City, Mexico.

Abstract The TiO2 nanotube films were prepared by anodizing Ti plates in 0.2 M NH4F ethylene glycol/10% H2O under different formation voltages keeping fixed the time length, or by keeping fixed the formation voltage and varying the time length. The morphology of the TiO2 film obtained was observed by SEM images and different morphological parameters were derived from them. Furthermore, the optical and semiconducting properties of TiO2 films were also measured. The photoelectrochemical performance toward water oxidation of the TiO2 only showed to be dependent with the inner diameter of the nanotubes, that could be related to the interaction of the film with the light and the transport of species in the electrolyte inward or outward the film. © Published by Elsevier B.V. B.V. This is an open access article under the CC BY-NC-ND license ©2014 2014The TheAuthors. Authors. Published by Elsevier (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of the Sociedad Mexicana de Electroquimica. Peer-review under responsibility of the Sociedad Mexicana de Electroquimica "Keywords: Titanium dioxide nanotubes; Surface morphology; Photoelectrochemistry; Species transport in electrolyte"

1. Introduction Growth of TiO2 nanotubes by anodization of Ti substrates in fluoride containing media is a process that has been increasingly gaining importance as a result of its flexibility and simplicity of formation, highly ordered structure, directional pathway for electron transport and the paramount performance showed by these films in different applications such as: photocatalysis,1 photoelectrochemical solar cells,2 water splitting,3 among others.4,5

* Corresponding author. Tel/ fax: +52 55 5804 4671. E-mail address: [email protected]

1876-6196 © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Sociedad Mexicana de Electroquimica doi:10.1016/j.proche.2014.12.038

Próspero Acevedo-Peña and Ignacio González / Procedia Chemistry 12 (2014) 34 – 40

Particularly, the water oxidation has drawn attention of different researchers since it is the half-reaction occurring at the anode (TiO2 film) during photoelectrochemical water splitting.3,6-9 TiO2 nanotubes can be obtained by anodization in aqueous electrolytes as well as in organic media containing fluoride ions.1-9 The latter one is the most commonly used to fabricate films for photoelectrochemical applications due to its simplicity for controlling the morphology (tube diameter and length) through operational variables. To design these structures and provide more efficient anodes, it is necessary to understand different parameters that govern the photoelectrochemical performance of TiO2 nanotube films. The purpose of this study is to evaluate the relation between the processing variables during the anodizing process (formation voltage and time length), over the morphology, electronic properties and photoelectrochemical performance for water oxidation, of TiO2 nanotubes film growth in 0.2 M NH4F Ethylene glycol/10% H2O. Nomenclature VF tF di l w dt P R Efb Nd Eg OCPon

Voltage employed during anodization process Time length of anodization process Inner diameter Tube length Wall thickness Tube to tube distance Porosity Roughness factor Flat band potential Donor density Band gap energy Open circuit potential measurements under illumination

2. Experimental details The TiO2 nanotube films were obtained by potentiostatic anodization in a two electrode cell at different voltages (VF = 15, 20, 25, 30 and 35 V) for 60 min or at 30 V for different time lengths (tF = 15, 30, 60, 120 and 180 min) in a 0.2 MNH4F ethylene glycol/10 % H2O electrolyte. Pt (99.99 % Alfa Aesar) was employed as counter electrode placed at 2 cm from the Ti foil (99.95 % Alfa Aesar). The electrolyte was stirred with a magnetic bar during anodization. Finally, the films were thoroughly cleaned with ethanol and Millipore water (18.2 MŸcm), left to air dry, and heat treated in ambient air at 450 °C (10 °Cmin-1) for 30 min to obtain anatase polymorph.1,3 SEM images were acquired using a JSM 7600F high field emission microscope, with an accelerating voltage of 10.0 kV. Morphological characteristics of the formed films were estimated using the iTEM software from Olympus soft imaging solutions. The UV-Vis diffuse reflectance spectra were measured using a Varian Cary 100 spectrometer equipped with an integration sphere. The (photo)electrochemical tests were carried out in a conventional three electrode cell equipped with a quartz window allowing the UV light illumination of the entire portion (1.23 cm2) of the TiO2 nanotube film exposed to the electrolyte. An Ag/AgCl (3.0 M KCl) electrode was employed as reference electrode and a graphite bar (99.999 % Alfa Aesar) as counter electrode. The 0.1 M HClO4 aqueous electrolyte used for film characterization was prepared employing Millipore water (18.2 MŸcm) and HClO4 (JT Baker, 69 %). Before each test, the electrolyte was bubbled with N2 gas for 30 min and N2 atmosphere was preserved during the experiments. The illumination was made using a Newport Q Housing (Model 60025) equipped with a 100 W Hg arc lamp. The semiconducting properties of the films were estimated from Mott-Schottky plots. The space charge capacitance of the film was extracted from the high frequency time constant of the potentiostatic EIS spectra obtained at 50 mV intervals between 0.75 V vs. Ag/AgCl and 0.0 V vs. Ag/AgCl.3 Prior to each measurement, the measuring potential was imposed for 10 minutes in order to stabilize the interfaces, then the EIS spectra were

35

36

Próspero Acevedo-Peña and Ignacio González / Procedia Chemistry 12 (2014) 34 – 40

collected in a frequency interval between 10 kHz and 10 mHz with AC perturbation of ±10 mV (peak to peak). Finally, the experimental EIS spectra were fitted to an equivalent circuit, R(Q(R(QR))), using Boukamp software. The (photo)electrochemical characterization was carried out using a BAS Epsillon potentiostat and EIS measurements were performed in an EG&G PAR model 283 potentiostat/galvanostat, coupled to a SI model 1260 Solartron frequency response analyzer.

Fig. 1. SEM images obtained for the as-anodized TiO2 nanotube films grown at different VF for 60 min and at 30 V for different tF.

3. Results and discussion The SEM images of the as-anodized films obtained at different VF for 60 min, and at 30 V for different tF, are shown in Figure 1. For all the VF and tF employed TiO2 nanotubes arrays were obtained, however the tube length and diameter were considerably altered by changing VF or tF. Increasing VF as well as tF causes an improvement in the nanotube morphology, avoiding the formation of precipitates present on top of the films. The morphological characteristics of the obtained TiO2 nanotubes were estimated from SEM images similar to those shown in Figure 1, and are summarized in the Table 1. Increasing VF and tF carried to an increase in the inner tube diameter, di, as well as their length, l; nonetheless, neither of these two variables caused a change in the wall thickness of the nanotubes, w. Besides, the film porosity, P, and the roughness factor, R, gets continuously higher values as VF and tF were increased; being R more sensible to changes in tF than VF, during the film growth.

Próspero Acevedo-Peña and Ignacio González / Procedia Chemistry 12 (2014) 34 – 40 Table 1. Morphological parameters of the TiO2 nanotube films obtained by anodization at different conditions, estimated from SEM images similar to those shown in Figure 1. Inner diameter, di; tube length, l; wall thickness, w; tube to tube distance, dt; porosity, P; roughness factor, R. tF (min)

VF (V)

di* (nm)

l* (nm)

w* (nm)

60

15

32.5

927.3

10.8

60

20

45.8

1181.9

60

25

59.2

1517.9

60

30

71.7

60

35

15

30

30 60 120

30

76.8

2342.6

180

30

82.9

3780.5

dt* (nm)

P

R

51.3

0.71

101.0

12.0

71.7

0.70

102.0

11.1

75.5

0.71

117.8

1836.5

10.6

88.5

0.73

128.0

84.3

2140.4

10.9

97.7

0.73

127.9

56.9

757.1

11.3

75.0

0.70

60.2

30

68.5

1244.2

11.9

80.6

0.71

86.2

30

71.7

1836.5

10.6

88.5

0.73

128.0

10.5

99.5

0.74

156.3

10.4

105.9

0.75

239.2

*The values reported in the table are an average of 500 measurements for di, dt and w and 100 measurements for l.

The semiconducting properties of the films were estimated in the dark using Mott-Schottcky plots.1,3 For this purpose the space charge capacitance was evaluated from EIS spectra obtained at different potentials in which no faradaic processes are taking place over the TiO2 surface in a 0.1 M HClO4 aqueous electrolyte.1,3,8 The flat band potential, Efb, and the donor density, Nd, are tabulated in the Table 2. Increasing the formation voltage of the films caused a slight displacement of Nd toward higher values, and changes the Efb to continuously less positive values. In contrast, increasing the anodizing time length carried to a detriment in the Nd values measured and displaces the Efb toward less positive values. Table 2. Effect of TiO2 nanotubes growth conditions (VF and tF) over their semiconducting properties (donor density, Nd, and flat band potential, Efb), and the band gap energy, Eg. tF

VF

(min)

(V)

Nd × 1021 (cm-3)

Efb (V vs Ag/AgCl)

Eg (eV)

60

15

0.22

0.49

2.9

60

20

1.73

0.45

2.9

60

25

3.58

0.45

3.0

60

30

2.28

0.38

3.0

60

35

1.55

0.39

3.2

15

30

6.06

0.49

3.1

30

30

7.05

0.45

3.1

60

30

2.28

0.38

3.0

120

30

0.84

0.37

2.9

180

30

0.51

0.28

2.8

VF and tF differently influence the Nd values, showing that the defects formed in the TiO2 during its growing might be of different nature. When VF increases, it causes an increment in the diffusion coefficient for the F- ions inside the compact layer formed in the metal/oxide interface,9 so a greater insertion of these ions is expected when the TiO2 nanotube films are grown at continuously higher VF. Modifying TiO2 by doping with F- ions induce energetic states inside the gap, provoking a displacement of the Efb towards the conduction band of the TiO2, increasing the Nd.10,11 However, even when the same trend in Efb is observed, the estimated Nd decreased as tF got increasingly higher, evidencing a further modification of the film probably related to the insertion of other species present in the electrolyte, which is compensating the charge, inducing energetic states in the gap over the valence

37

38

Próspero Acevedo-Peña and Ignacio González / Procedia Chemistry 12 (2014) 34 – 40

band of n-type semiconductors.12,13 In the case of TiO2, and due to the composition of the anodizing bath, it might be C or N.14 The variation of the band gap energy (Eg) with the processing variables is also summarized in Table 2. The increase in VF led to a shift towards increasing Eg values; while the increase in tF had the opposite effect, diminishing the estimated values for Eg. This behavior is in accordance with that observed in the variation of Nd in Table 2; because the insertion of C or N induces energetic states over the TiO2 valence band, decreasing Eg.10,11,14 Additionally, according to previous studies, doping TiO2 with these elements leads to a shift of Efb towards more distant values from the TiO2 conduction band,12,13 disagreeing with the trend observed in Table 2. However, it is worth mentioning that the measurement of semiconductor properties using EIS is limited to compact oxide film,1,3 so in order to obtain information about the flat band potential of the nanotube film, open circuit potential measurements were performed under illumination (OCPon), see Figure 2a. The increase in VF and tF during the growth of the TiO2 nanotubes modified the OCPon, showing that defects induced by the insertion of N and/or C during film growth, primarily impacts the properties of the porous outer film, and no the internal compact layer. This behavior is consistent with that observed by other researchers, that have shown than fluoride ions migrates faster than oxygen, or C and N species, due to their size and mobility in the TiO2 lattice; so fluoride ions are mainly present in the compact film, while species of C and N, remain TiO2 nanotubes.9

Fig. 2. Effect of the processing variables during TiO2 nanotubes arrays growth over: (a) the open circuit potential measured under illumination (OCPon), and (b) the photocurrent potentiostatically measured at 1.25 V vs Ag/AgCl. Both measurements were carried in a 0.1 M HClO4 electrolyte.

Próspero Acevedo-Peña and Ignacio González / Procedia Chemistry 12 (2014) 34 – 40

The effect of the VF and tF over the photocurrent potentiostatically measured at 1.25 V vs Ag/AgCl in a 0.1 M HClO4 electrolyte is shown in Figure 2 b. In both cases, the photocurrent increase as VF and tF takes increasingly higher values, reaching a maximum for 30 V when the time length is fixed at 60 min (filled points) and at 120 min when the formation voltage is fixed at 30 V (filled squares). The formation of a maximum in the measured photocurrent with VF and tF does not correlate directly with the electronic properties of the film (Efb, Nd or Eg), see Table 2. From the different morphological parameters reported in Table 1 the inner diameter of the nanotubes, di, is the only one that seems to show a monotonic dependence on the photoelectrochemical performance, exhibiting a maximum value around 71-77 nm, Figure 3. This dependence may be related to the fact that the diameter of the nanotubes determines the interaction of light with the film.15 Although various studies have linked the formation of the maximum to the morphology of the film, there is controversy about the optimal values of inner diameter and length of the nanotubes for photoelectrochemical applications,15 showing that these values can be related to other variables such as: the system used for photoelectrochemical characterization, the electrolyte used for anodizing, or to the subsequent heat treatment to which the film is subjected; that they are consequently altering the electronic properties of the films.

Figure 3. Dependence of the photocurrent potentiostatically measured at 1.25 V vs Ag/AgCl in a 0.1 M HClO4 electrolyte with the TiO2 nanotube inner diameter (di). 4. Conclusions

By varying the formation voltage and time length during the anodizing process, the morphology of the TiO2 film obtained was considerably modified varying their inner diameter from around 38 nm to 82 nm, and their length from around 0.9 μm to 3.8 μm. Furthermore, the optical and semiconducting properties of TiO2 films were also modified with the processing variables; this behavior was attributed to the insertion of species in the electrolyte in the TiO2 lattice during their growth. However, the photoelectrochemical performance toward water oxidation of the TiO2 only showed to be dependent with the inner diameter of the nanotubes, that could be related to the interaction of the film with the light and the transport of species in the electrolyte inward or outward the film. Acknowledgements The authors are indebted to the CONACyT for their financial support to carry out this work (Projects CB 2008/105655 and INFR 2011 1 163250). P. Acevedo-Peña gratefuls to CONACyT for the PhD fellowship granted. The authors thank to Laboratorio Central de Microscopía Electrónica (UAM-I) for SEM images.

39

40

Próspero Acevedo-Peña and Ignacio González / Procedia Chemistry 12 (2014) 34 – 40

References 1. Acevedo-Peña P, Carrera-Crespo JE, González F, González G. Effect of heat treatent on the cristal phase composition, semiconducting properties and photoelectrocatalytic color removal efficiency of TiO2 nanotube arrays. Electrochim Acta 2014; 140: 564-571. 2. Wang J, Lin Z. Dye-Sensitized TiO2 Nanotube Solar Cells: Rational Structural and Surface Engineering on TiO2 Nanotubes. Chem-Asian J 2012; 7: 2754-2762. 3. Acevedo-Peña P, González I. TiO2 nanotubes formed in aqueous media: relationship between morphology, electrochemical properties and photoelectrochemical performance for water oxidation. J Electrochem Soc 2013; 160: H452-H458. 4. Li J, Yu H, Lin C-J. A Photoelectrochemical Study of n-Doped TiO2 Nanotube Arrays as the Photoanodes for Cathodic Protection of SS. J Electrochem Soc. 2007; 154: C631-C636. 5. Wu X, Ling Y, Liu L, Huang Z. Enhanced photoelectrocatalytic degradation of methylene blue on smooth TiO2 nanotube array and its impedance analysis. J Electrochem Soc 2009; 156: K65-K71. 6. Kong D-S, Zhang X-D, Wang J, Wang C, Zhao X, Feng Y-Y, Li W-J. A photoelectrochemical study on the features of carbonate-catalyzed water oxidation at illuminated TiO2/Solution interface. J Solid State Electrochem 2013; 17: 69-77. 7. Zhang Z, Wang P. Optimization of photoelectrochemical water splitting performance on hierarchical TiO2 nanotube arrays. Energy Environ Sci 2013; 5: 6506-6512. 8. Acevedo-Peña P, González I. TiO2 photoanodes prepared by cathodic electrophoretic deposition in 2-propanol: effect of the electric field and deposition time. J Solid State Electrochem 2013; 17: 519-526. 9. Roy P, Berger S, Schmuki P. TiO2 Nanotubes: Synthesis and Applications. Angew Chem Int Ed 2011; 50: 2904-2939. 10. Dozzi MV, Selli E. Doping TiO2 with p-block elements: Effects on photocatalytic activity. J Photochem Photobiol C 2013; 14: 13-28. 11. Di Valentin C, Pacchioni G. Trends in non-metal doping of anatase TiO2: B, C, N and F. Catal Today 2013; 206: 12-18. 12. Cheng XF, Leng WH, Liu DP, Xu YM, Zhang JQ, Cao CN. Electrochemical Preparation and Characterization of Surface-Fluorinated TiO2 Nanoporous Film and Its Enhanced Photoelectrochemical and Photocatalytic Properties. J Phys Chem C 2008; 112: 8725-8734. 13.Di Valentin C, Pacchioni G, Selloni A. Reduced and n-Type Doped TiO2: Nature of Ti3+ Species. J Phys Chem C. 2009; 113: 20543-20552. 14. Acevedo-Peña P, González F, González G, González I. Effect of Anatase crystal orientation on the photoelectrochemical performance of anodic TiO2 nanotubes. Phys Chem Chem Phys 2014; Accepted for publication. 15. Paramasivam I, Jha H, Liu N, Schmuki P. A Review of Photocatalysis using Self-organized TiO2 Nanotubes and Other Ordered Oxide Nanostructures. Small 2012; 8: 3073-3103.