Enhanced photosensitization of nanocrystalline TiO₂ film with a sulfonic sensitizer by surface protonation

Abstract

The interaction between an azo sensitizer containing an –SO₃⁻ anchoring group, 4-amino-5-hydroxy-6-azobenzene-1,7-naphthyl disulfonic sodium (K), and nanocrystalline TiO₂ film was studied using IR microscopy, and was attributed to physical adsorption due to the negligible peak shift of K. Photoelectrochemical experiments were carried out to characterize the interface charge transfer. Maximum incident photon-to-current conversion efficiency (IPCE) was greatly increased from 7.7 to 43.7% by HCl pretreatment of TiO₂ film. This strong photosensitization enhancement was assigned to surface protonation of the TiO₂ film, which enables K anions to be closely attracted to the TiO₂ nanocrystalline surface by electrostatic forces, thus facilitating interface charge transfer. An overall energy conversion efficiency of ∼2% was achieved for the sandwich-type photoelectrochemical cell constructed with K sensitized protonated TiO₂ film, while this value was 0.37% for the untreated TiO₂ film. Surface protonation provides a general solution to improving the photosensitization of TiO₂ film with dyes containing SO₃⁻ anchoring groups.

---

Surface modification of wide band gap semiconductors with sensitizers is a well-established procedure to alter their photophysical and photochemical properties.^1–3 In the sensitized systems, molecules attached to the semiconductor surface inject electrons into the conduction band of the semiconductor after being excited, thus extending the photoreponse of the semiconductor into the visible region. Research on such sensitized systems has greatly advanced since the development of nanocrystalline TiO₂ film and highly efficient Ru complex sensitizers by Grätzel and co-workers.^4,5 Dye-sensitized nanocrystalline TiO₂ photoelectrochemical cells have shown promising prospects in practical applications and also have stimulated enormous interest in fundamental research.^6–10 Nanocrystalline TiO₂ films provide huge interior surfaces for sensitizer attachment, which ensures high light harvesting efficiency. Interface charge transfer is another key step controlling overall energy conversion efficiency. The rate of this process is critically dependent on the degree of electronic coupling between the donor sensitizer orbital and the acceptor TiO₂ orbital,¹¹ which is obviously related to the way of sensitizer adsorption. Satisfactory performance has been achieved by attaching suitable anchoring groups on the sensitizer molecule.

The usual anchoring group employed is –COOH. Through a number of Raman and IR spectroscopic studies of sensitizer–TiO₂ systems, it was concluded that sensitizers functionalized with carboxylate groups anchor to the TiO₂ film through an ester-like bond with the surface –OH groups.^12–14 Recently, sensitizers with –SO₃⁻ as the anchoring group were also reported for photoelectric conversion.^15,16 A better understanding of the interaction between these sensitizers and the TiO₂ surface is obviously very important since it is closely related to sensitization efficiency, as mentioned above. However, to the best of our knowledge, no studies have been made in this regard.

Azo derivatives are cheap and stable dyes that have found wide applications in the textile industry.¹⁷ In the present study, an azo dye containing a –SO₃⁻ group, 4-amino-5-hydroxy-6-azobenzene-1,7-naphthyl disulfonic sodium (K), and having suitable absorption and redox properties was chosen to sensitize the nanocrystalline TiO₂ film. The nature of the interaction between sensitizer and TiO₂ was investigated using IR microscopy. The results show K is physisorbed onto the TiO₂ nanocrystalline surface. The surface of as-prepared TiO₂ is hydroxylated.¹⁸ Surface –OH groups may repel the –SO₃⁻ containing K from intimate contact, which is unfavorable for efficient interface charge transfer. If the TiO₂ surface is positively charged, will electrostatic forces attract K closer to the TiO₂ surface, and will interface charge-transfer be facilitated? We found that after surface protonation of a TiO₂ film by soaking in HCl solution, photoresponse was greatly enhanced as expected.† An overall energy conversion efficiency of about 2% was achieved in the improved case.

Experimental

Materials

ITO conducting glass (indium doped SnO₂, sheet resistance 20 Ω cm⁻², transmission >85% in the visible) was purchased from Shenzhen Weiguang Conducting Glass Company of China. K dye was synthesized and purified. Its structure was confirmed by ¹H NMR.¹⁷ LiI and carbowax were purchased from Aldrich. All the other solvents and chemicals were at least reagent grade (Beijing Chemical Company, China) and used without further purification. Propylene carbonate (PC) was dried with molecular sieves and then distilled before use.

Preparation of K coated nanocrystalline TiO₂ film

Preparation of TiO₂ nanocrystalline film has been described elsewhere.¹⁹ TiO₂ nanoparticles were synthesized by autoclaving the hydrolysis product of tetrabutyl titanate at 230 °C for 12 h. The TiO₂ particles were redispersed under ultrasonication with addition of carbowax (40% of TiO₂ weight). After evaporation, a viscous TiO₂ paste was obtained. ITO conducting glass was used as the substrate for depositing the TiO₂ film. To control the thickness of the film and to mask electric contact strips, two edges of the ITO glass plate were covered with adhesive tape. TiO₂ paste was spread uniformly on the substrate by sliding a glass rod along the tape spacer. After drying, the sample was heated in air for 30 min at 450 °C, then was soaked in 0.3 mM K solution in MeCN for 3 h. Excess dye was removed by rinsing with MeCN. For surface protonation, the TiO₂ film was soaked in an HCl solution of pH 2 for 3 h, then dried under a hot air flow before soaking in the K solution. The TiO₂ film thickness was measured to be 6 μm using an ALPHA-step 500 profilometer.

Micro IR and UV-Vis characterization

Pure K and trace amounts of sample scratched from K coated nanocrystalline TiO₂ film were subject to micro FTIR characterization using a MAGNA-IR 750 spectrometer series II combined with a Nic-Plan™ IR microscope (Nicolet). UV-Vis absorption spectra of K solution in MeCN and K coated TiO₂ film were measured using a Cary 1 E UV-Vis spectrometer (Varian); MeCN and bare ITO glass were used as references, respectively.

Electrochemical and photoelectrochemical experiments

Cyclic voltammetry (CV) of K was performed on a CHI650A electrochemical workstation (CH Instruments). A conventional three-electrode cell composed of Pt wire working electrode, Pt plate counter electrode and saturated calomel reference electrode (SCE) was used. NaClO₄ (0.5 M) in deionized water served as supporting electrolyte.

Photoelectrochemical experiments were carried out to characterize interface charge transfer between K and TiO₂. Sandwich-type cells were constructed with platinized ITO glass as counter electrode and 0.3 M LiI/0.03 M I₂ PC solution as electrolyte. A 150 W xenon lamp with a high intensity grating monochromator served as light source. Light intensities were measured with an EG&G light gauge radiometer/photometer (Gamma Scientific). In current–voltage measurements, a 10 cm water filter and a 350 nm cut-off filter were placed in the beam, and the 15% transmission loss in ITO glass was corrected. The current–voltage curves were obtained at a scan rate of 5 mV s⁻¹ in a two-electrode arrangement, which is identical to loading a variable resistor.¹⁹

Results and discussion

Absorption and redox properties

Fig. 1 shows the absorption spectra of K solution in MeCN and K coated TiO₂ film. K has a characteristic π–π^* electronic transition absorption band around 500 nm. This strong, broad absorption in the visible region is essential for high light harvesting efficiency. By comparing spectra a and b in Fig. 1, it can be seen that the absorption of K in solution and on TiO₂ film is similar and the two peaks are almost at the same wavelength, suggesting that K has been adsorbed onto the TiO₂ film without aggregation.

Fig. 1 Absorption spectra of (a) K solution in MeCN and (b) K coated nanocrystalline TiO₂ film.

Besides the absorption behavior, the availability of a suitable energy level is another key factor for efficient sensitization, which ensures that electron injection from the excited K molecule into the TiO₂ conduction band is thermodynamically possible. CV results on K give the redox potential of K as 0.58 V vs. SCE. Combining electrochemical and absorption spectroscopic data, the redox potential of excited state K is estimated to be −1.26 V vs. SCE (about −3.58 eV in vacuum level), about 0.36 eV above the conduction band edge (about −0.90 V vs. SCE, −3.94 eV in vacuum level) of a nanocrystalline TiO₂ film in aprotric solvent with no less than 0.1 M Li⁺ ion.²⁰ It can be seen that K has the proper absorption behavior and energy level to sensitize nanocrystalline TiO₂ film.

Nature of interaction between K and TiO₂

Microscope IR was measured between 4000 and 650 cm⁻¹. Strong absorption of TiO₂ below 1000 cm⁻¹ and broad absorption of –OH groups above 3400 cm⁻¹ were seen for uncoated and K coated TiO₂ film samples. Z. Wang et al.¹⁵ recently applied a series of hemicyanine derivatives containing the SO₃⁻ group to sensitize TiO₂ nanocrystalline films. They found that the presence of SO₃⁻ was essential for successful adsorption, as hemicyanines without SO₃⁻ were hardly adsorbed on TiO₂ films. Thus, only the –SO₃⁻ anchoring group is of concern for scrutinizing the nature of the adsorption.

A restricted region of the IR spectra of pure K and K adsorbed onto TiO₂ nanocrystalline film is shown in Fig. 2. The strong absorption peaks of spectrum a at 1214 and 1255 cm⁻¹ are attributed to the symmetric and asymmetric stretching modes of S[double bond, length half m-dash]O. For K adsorbed on TiO₂ film, the corresponding peaks are at 1217 and 1254 cm⁻¹. Upon adsorption of a sensitizer with a –COOH group, a 20 cm⁻¹ peak shift to higher frequency of the C[double bond, length half m-dash]O stretching mode was observed compared to free sensitizer, which is believed to arise from chemical adsorption of the sensitizers onto the oxide particles via formation of an ester-like bond between the carboxylic acid groups and the hydroxyl moieties on the TiO₂ surface.¹² In our case the minor S[double bond, length half m-dash]O absoption peak shifts imply that there is no chemical interaction between adsorbed K molecules and TiO₂. The relatively small electronegativity of S rules out the possibility of an H-bond between K and TiO₂, although this is found to be possible for –COOH and an oxide surface by molecular modelling calculations.¹³ K molecules may be physically adsorbed onto the TiO₂ nanocrystalline surface through a solvent layer²¹ or electrostatic forces.

Fig. 2 FTIR spectra of (a) pure K and (b) K coated nanocrystalline TiO₂ film.

K adsorbed onto a TiO₂ film is easily washed off with water, contrasting with sensitizers having –COOH as anchoring group, while K in aqueous solution was not adsorbed at all. These facts show that the interaction between K and TiO₂ is weak and support the physisorption nature of the interaction.

Photosensitization

TiO₂ is a wide band gap n-type semiconductor only responsive to UV light. The short-circuit photocurrent produced by photosensitization in the visible region of K coated, untreated and pretreated, TiO₂ film was measured. The action spectra are shown in Fig. 3, where the incident photon-to-current efficiency (IPCE) determined from the short-circuit photocurrent is plotted as a function of wavelength. In both cases the photocurrent extends to the visible region and follows the same profile as the absorption spectrum of K in solution. For the pretreated film the IPCE was increased more than fivefold over the whole region compared with the untreated film. The maximum IPCEs of the K sensitized, HCl pretreated and untreated, TiO₂ films are 43.7 and 7.7%, respectively.

Fig. 3 Action spectra of K sensitized nanocrystalline TiO₂ film pretreated with HCl (a) and without pretreatment (b).

To clarify this large enhancement of the photosensitization, the UV-Vis absorption character of the K coated TiO₂, with and without HCl pretreatment, was first investigated. As shown in Fig. 4, there is little difference between the two absorption spectra, indicating that the HCl pretreatment process does not affect the arrangement of K molecules on the TiO₂ nanocrystalline surface. This is understandable since K molecules do not tend to aggregate when adsorbed on TiO₂ nanoparticle, as concluded from Fig. 1. At the same time, TiO₂ is so stable and well crystallined after autoclaving that HCl of pH 2 will not change the surface morphology. So the enhancement of photosensitization cannot be rationalized in terms of the absorption itself.

Fig. 4 Absorption spectra of K coated TiO₂ film pretreated with HCl (a) and without pretreatment (b).

HCl pretreatment has another influence on the TiO₂ film. TiO₂ band edge positions or the flat band potential are known to be dependent on the pH value of the contact solution.^20,22,23 Adsorption of protons is associated with a downward energy shift of the band edges, and thus will increase the driving force for electron injection from excited K into the TiO₂ conduction band. But this cannot be the decisive factor for the great photosensitization enhancement. As discussed above, the energy difference between the excited K and the conduction band of TiO₂ already provides a driving force for interface charge transfer. Our experiments also show that HCl pretreatment of TiO₂ film does not induce any notable increase in IPCE when sensitized with the well-known Ru dye [cis-di(thiocyanato)bis(2,2′-bipyridyl-4,4′-dicarboxylate)ruthenium(II)] having –COOH as anchoring group.

Due to the physisorption nature of K on TiO₂ film, there is no interlocking bridge present between K and TiO₂ such as the ester-like bond present in –COOH containing sensitizer-TiO₂ systems. Surface –OH groups of TiO₂ nanocrystalline may prevent intimate contact between K and TiO₂, and thus retard efficient charge transfer. However, protons are easily adsorbed by TiO₂ nanocrystalline during soaking of the film in HCl aqueous solution. In the dye-coating process, that is dipping of the TiO₂ film in K solution, ionized K anions will be electrostatically bound to the positively charged TiO₂ nanocrystalline surface. The electrostatic interaction attracts K anions close to the protonated TiO₂ surface, consequently favoring efficient charge injection.

Fig. 5 shows current-voltage curves of photoelectrochemical cells constructed with K sensitized protonated and untreated nanocrystalline TiO₂ films under white light illumination of 27.2 mW cm⁻². Short-circuit current and open-circuit voltage for the cell constructed with the untreated TiO₂ nanocrystalline film are 0.39 mA cm⁻² and 0.44 V, respectively (see Fig. 5). For the cell constructed with the protonated TiO₂ nanocrystalline film, the corresponding values are 2.03 mA cm⁻² and 0.45 V. Both cells have the same fill factor of 0.6, giving overall energy conversion efficiencies of 0.37 and 1.96%. Both the short-circuit current and overall energy conversion efficiency are enhanced more than fivefold in magnitude by HCl pretreatment, which is attributed to the improved interface charge transfer resulting from close contact between sensitizer K molecules and the TiO₂ nanocrystalline film.

Fig. 5 Current–voltage curves of the photoelectrochemical cells constructed with K sensitized TiO₂ film pretreated with HCl (a) and without pretreatment (b).

Conclusion

Interaction between an azo sensitizer containing –SO₃⁻ as anchoring group, 4-amino-5-hydroxy-6-azobenzene-1,7-naphthyl disulfonic sodium (K) and nanocrystalline TiO₂ film is found to be physical adsorption through a micro IR investigation. A significant maximum IPCE increase from 7.7 to 43.7% was achieved by soaking the TiO₂ film in an HCl solution of pH 2 before dye-coating, which is attributed to the fact that surface protonation of the TiO₂ film enables K anions to be closely attracted to the TiO₂ surface by electrostatic forces, thus facilitating interface charge transfer. An overall energy conversion efficiency of ∼2% was achieved for the sandwich-type photoelectrochemical cell constructed with K sensitized protonated TiO₂ film, contrasting with a value of 0.37% for the untreated TiO₂ film. Surface protonation provides a general solution to improving sensitization efficiency of nanocrystalline TiO₂ films with dyes containing –SO₃⁻ anchoring groups.

Acknowledgements

Project 20073003 was supported by the National Natural Science Foundation of China.

References

1. M. Gleria and R. Memming, J. Phys. Chem., 1975, 98, 303.
2. P. K. Ghosh and T. G. Spiro, J. Am. Chem. Soc., 1980, 102, 5543.
3. M. Fujihara, N. Ohishi and T. Osa, Nature (London), 1977, 268, 226.
4. B. O'Regan and M. Grätzel, Nature (London), 1991, 353, 737.
5. M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Müller, P. Liska, N. Vlachopoulos and M. Grätzel, J. Am. Chem. Soc., 1993, 115, 6382.
6. Y. Ren, Z. Zhang, M. Yang, S. Fang and S. Cai, Sol. Energy Mater. Sol. Cells, 2001, 71, 253.
7. Y. Hao, M. Yang, C. Yu, S. Cai, M. Liu, L. Fan and Y. Li, Sol. Energy Mater. Sol. Cells, 1998, 56, 75.
8. A. Solbrand, A. Henningsson, S. Sodergren, H. Lindstrom, A. Hagfeldt and S. E. Lindquist, J. Phys. Chem. B, 1999, 103, 1078.
9. F. Willig, C. Zimmermann, S. Ramakrishna and W. Storck, Electrochim. Acta, 2000, 45, 4565.
10. S. A. Haque, Y. Tachibana, D. R. Klug and J. R. Durrant, J. Phys. Chem. B, 1998, 102, 1745.
11. J. M. Rehm, G. L. Mclendon, Y. Nagasawa, K. Yoshihara, J. Moser and M. Grätzel, J. Phys. Chem. B, 1996, 10, 9577.
12. P. Falaras, Sol. Energy Mater. Sol. Cells, 1998, 53, 163.
13. T. J. Meyer, G. J. Meyer, B. W. Pfennig, J. R. Schoonover, C. J. Timpson, J. F. Wall, C. Kobusch, X. Chen, B. M. Peek, C. G. Wall, W. Ou, B. W. Erickson and C. A. Bignozzi, Inorg. Chem., 1994, 33, 3952.
14. N. W. Duffy, K. D. Dobson, K. C. Gordon, B. H. Robinson and A. J. McOuillan, Chem. Phys. Lett., 1997, 266, 451.
15. Z. Wang, F. Li, C. Huang, L. Wang, M. Wei, L. Jin and N. Li, J. Phys. Chem. B, 2000, 104, 9676.
16. Z. Wang, F. Li and C. Huang, Chem. Commun., 2000, 2063.
17. H. Zhan and H. Tian, Dyes Pigm., 1998, 37, 231.
18. A. Hagfeldt and M. Grätzel, Chem. Rev., 1995, 95, 49.
19. Y. Ren, Z. Zhang, E. Gao, S. Fang and S. Cai, J. Appl. Electrochem., 2001, 31, 445.
20. G. Redmond and D. Fitzmaurice, J. Phys. Chem., 1993, 97, 1426.
21. S. Umapathy, A. M. Cartner, A. W. Parker and R. E. Hester, J. Phys. Chem., 1990, 94, 8880.
22. J. Morse and M. Grätzel, Helv. Chim. Acta, 1982, 65, 436.
23. G. Rothenberger, M. Grätzel and D. Fitzmaurice, J. Phys. Chem., 1992, 96, 5983.

---

Footnote

† During the publication process for this work, a paper was published (Z. S. Wang, Y. Liu, G. Yu, C. Xu, J. Zhang and D. Zhu, J. Phys. Chem. B, 2001, 105, 9422) reporting separate work on the photocurrent enhancement of a series of hemicyanine dyes containing the SO₃⁻ group with acid treatment.

---