ultrafine powder in aqueous solutions - Springer Link

METALS AND MATERIALS International, Vol. 8, No. 1 (2002), pp. 103~109

Redox Abilities of Rutile TiO2 Ultrafine Powder in Aqueous Solutions Jae Sung Song, Dong Yoon Lee, Won Jae Lee and Sun Jae Kim* Electric and Magnetic Devices Group, Korea Electrotechnology Research Institute (KERI) 28-1 Sungju-dong, Changwon 641-120 Korea *Sejong Advanced Institute of Nano Technologies, Sejong University 98 Kunja-dong, Kwangjin-ku, Seoul 143-747, Korea Redox abilities of rutile TiO2 powder with an acicular primary particle, having about a 200 m2/g BET surface area obtained by a homogeneous precipitation process in aqueous TiOCl2 solution at 50oC for 4 hrs, were investigated using a photocatalytic reaction in aqueous 4-chlorophenol, Cu- and Pb-EDTA solutions. Its abilities were then compared with those of commercial P-25 powder, together with investigating those of anatase TiO2 powder obtained from the aqueous TiOCl2 solution, having a similar surface area and primary particle shape as those of the rutile powder. The powder was more effective than the anatase or P-25 powder, while the anatase powder unexpectedly showed the slowest decomposition rate and the smallest amount in the same experiments in spite of similar particle shapes and surface area. From the results, it is found that the excellent photo redox abilities of the powder is likely to be caused by the specific powder preparation method regardless of crystalline structures even when having similar particle shapes and values in the surface area. − Also, many OH attached to the surface of the TiO2 particle appeared to interfere with the adsorption of decomposing target materials to the TiO2 surface in the solution during the photocatalytic reaction, resulting in a delay in the reaction. Keywords : rutile TiO2 powder, redox ability, photocatalytic reaction, crystalline structures, decomposition

1. INTRODUCTION The photocatalytic process, applying a non-polluted energy source such as sunlight to the metal oxide semiconducting material, TiO2, has been actively studied in the fields of removing non-degradable organic chemicals such as phenol [1] and 4-chlorophenol [2] and recovering heavy metal ions [3] in wastewater because it especially induces the formation of stable by-products in the environment compared to conventional processes such as chlorination, ozonolysis, air stripping and so on and because it has relative simplicity and low operating cost for decomposing them completely. For a photocatalytic reaction to utilize the photonic energies having wavelengths shorter than about 400 nm reaching the surface of the ground effectively, modifications of TiO2 powder such as its minuteness up to nano size, metal ion doping for use in visible light and adaptation of dual semiconductor systems are needed to make it an efficient photocatalyst [4]. Generally, nanosized photocatalyst particles have the added advantage of providing a transparent solution for efficient photoredox processes. On the other hand, Wang et al. [5] demonstrated that there exists an optimal particle size of about 10 nm in nanocrystalline TiO2 systems for maximum

photocatalytic efficiency. Because of Somorjais suggestion [6] for the reasons of showing a lower efficiency in the rutile TiO2 phase in the two crystal structures of TiO2 being because the recombination of the electron-hole pair produced by UV light irradiation occurring more rapidly on the surface of the rutile phase and the amounts of reactants and hydroxides attached to the surface of the rutile phase being smaller than those of the anatase TiO2 phase, which has higher efficiency, the object material used for the photocatalytic reaction has undoubtedly been an anatase TiO2 phase in the two crystal structures of TiO2 for several decades. However, although nanostructured TiO 2 ultrafine powder with more than a 150 m2/g BET surface area by a homogeneous precipitation process at ambient or low temperatures (HPPLT) consisted of a pure rutile phase alone, it displayed a superior photocatalytic ability in recovering Cu and Pb ions as lumps from the aqueous solution by photo-reduction [7]. It could be thought that the superior photo-reductive ability of TiO2 powder by HPPLT is ascribed to its having a chestnut bur shape as a nanocrystalline structure with a large BET surface area and acicular-type primary particle coagulated with radializing in all directions, probably supporting more rapid photocatalytic reaction between metal

Jae Sung Song et al.

ions and photo excited conduction-band electrons without their recombination with valence-band holes for neutralization. Therefore, there is ample interest in accurately characterizing the fact that the photocatalytic abilities for the ultrafine TiO2 powder depends on the preparation method rather than the crystalline phase or surface area itself after obtaining the rutile TiO2 powder for a superior photocatalytic reaction which was crystallized together with the precipitation in the aqueous TiOCl2 solution during HPPLT. In this report, the photocatalytic abilities of the rutile TiO2 powder by HPPLT were compared with those of commercial P-25 powder with mainly the anatase phase as a benchmarking material when removing 4-chlorophenol and heavy metal ions such as Cu and Pb ions from aqueous solutions, together with investigating those of anatase TiO2 powder consisting of similarly shaped primary particles and having almost the same surface area as those of the rutile TiO2 powder. The reason the rutile TiO2 powder by HPPLT shows superior efficiency in the photocatalysis was also investigated together with the effect of OH- attached to the surface of the TiO2 particle from the wet processing for powder preparation.

2. EXPERIMENTAL PROCEDURE For the quantitative evaluation of photocatalytic properties for ultrafine TiO2 powders, two kinds of the powders were prepared with almost the same surface area and similar primary particle shape in different crystalline structures using crude TiCl4. First of all, as a starting material for obtaining photocatalytic TiO2 powders with rutile and anatase phases having high surface areas, an aqueous TiOCl2 solution having 4.7 M of Ti4+ concentration was made. Adding distilled water to the aqueous TiOCl2 solution to make a 0.4~0.7 M of Ti4+ concentration, crystalline TiO2 ultrafine powder with a pure rutile phase was prepared by the homogeneous precipitation process by only heating the solution at 50oC for 4 hrs (HPPLT). The TiO2 powder (HPPLT TiO2), with the pure rutile phase which has a BET surface area of − 180~200 m2/g and Cl concentration of less than 1 ppm, was obtained with about a 0.3 μm sized spherical secondary particle, which consists of a primary acicular particle with a thickness of 3~7 nm and length of 70~150 nm. Secondly, ultrafine TiO2 powder with a pure anatase phase, using the techniques as reported by Seo et al. [9], was obtained after preparing Ti-hydroxide by slow reaction between NH 4OH

and aqueous TiOCl2 solution.. And then, after maintaining it for 12 hrs in a 5M NaOH solution at 150oC followed by drying at 100oC for 12 hrs in air, the obtained anatase TiO2 powder, which has a BET surface area of about 200 m2/g, consists of about a 10 μm sized irregular secondary particle having spherical primary particles with sizes less than 10 nm in it, especially containing acicular typed primary particles (length 150 nm x diameter 10~20 nm) consisting mainly of a nanotube particle with a wall thickness of 2 nm at its surface, similar to that of the rutile powder by HPPLT. In addition, for obtaining accurate reference data for the reaction, the photocatalytic properties of two ultrafine TiO2 powders were compared with that of the commercial P-25 powder (Degussa Co.) as a benchmarking material. The crystalline structures and BET surface areas of ultrafine TiO2 powders used in the experiments are summarized in Table 1. Here, slurry typed TiO2 means dispersed HPPLT TiO2 suspension not dried after washing completely up to about pH 6. Photocatalytic reactions using four kinds of ultrafine TiO2 powders under ultraviolet irradiation were carried out in a 100 mL batch typed quartz reactor having an exterior irradiation source with an 800-watt power high pressure mercury arc lamp. In the oxidative experiments for an organic material, 1 mmol 4-chlorophenol (Sigma Co.) was used. In the reductive experiments for metal ions, aqueous Cu-EDTA and Pb-EDTA solutions with 50 ppm Cu and 150 wppm Pb ions, respectively, were used. The amount of total organic carbon (TOC) that varied during the photocatalytic reaction was measured using a TOC analyzer, and the degree of decomposition of 4-chlorophenol itself was confirmed by comparing the spectra obtained before and after the reaction using a luminescence − spectrometer. Also, the Cl concentration released from 4chlorophenol by the reaction or naturally existing in the aqueous TiO2 solution was analyzed using ion chromatography. Here, the remaining metal ion concentrations in the solution with the reaction were detected by an AA spectrophotometer. All the analyses were performed for the aqueous solution used, which had already been filtered to 0.2 and separated for the removal of the ultrafine TiO2 powders. Also, a TEM equipped with EDS was used for analyzing the composition and shape of the TiO2-metal complexes containing metal or metal oxides, which had been removed by photo-oxidation or -reduction reaction from the aqueous solutions.

Table 1. Physical properties of various TiO2 powders used in the experiment

Crystalline Phase 2 Specific Surface Area (m /g, BET)

Rutile TiO2 Powder (HPPLT TiO2 powder)

Slurry typed TiO2 (HPPLT TiO 2 suspension)

P-25 TiO2 Powder

Anatase TiO2 Powder

100% Rutile 180~200

100% Rutile Not determined

70% Anatase+30% Rutile ~ 55

100% Anatase ~ 200

Redox Abilities of Rutile TiO2 Ultrafine Powder in Aqueous Solutions

3. RESULTS AND DISCUSSION Recently, S.J. Kim et al. [7] developed a new synthetic process (HPPLT) to produce a rutile TiO2 ultrafine powder having more than a 150 m2/g BET surface area from aqueous TiOCl2 solution made from TiCl4 by simply heating it to below 70oC and drying it at 100~150oC. They have suggested that the spontaneous homogeneous precipitation reaction in a highly acidic TiOCl2 aqueous solution would be maintained through the growth of nuclei such as yellow hydroxide TiO(OH)2 together with its crystallization. The crystalline structures for HPPLT TiO2 precipitates in the acidic solution just after the reaction and in the dried state after the complete washing were analyzed using Raman spectroscopy and XRD apparatus. Fig. 1 shows the comparison of the Raman spectrum of commercial TiO2 powder (Aldrich Co.), as a standard sample powder, consisting of

Fig. 1. Raman spectra for crude TiCl4, TiO2 standard powder from Aldrich Co, 5 M TiOCl2, 0.5 M TiOCl2 and glass bottle as a reference, where the spectra for 0.5 M TiOCl2 were obtained before and after the o precipitation at 50 C for 4 hrs.

mainly a rutile phase together with a small anatase phase, with those of the solutions with and without TiO2 precipitates before and after HPPLT at 50oC for 4 hrs, respectively crude TiCl4 and the aqueous TiOCl2 solution. The spectrum characteristics of TiCl4 are quite different from those of water added TiOCl2. The characteristics of aqueous TiOCl2 solution with 0.5M of Ti4+ concentration even nearly become the same as that of water, showing no peaks. On the other hand, the E1g(R) and A1g(R) peaks in Fig. 1(a), corresponding to 438 and 604 cm−1 in Fig. 1(b), which are the characteristic peaks of the rutile phase, only appearing in the acidic solution without an anatase characteristic peak, A1g(A). −1 Here, the peak at the 241 cm is coming from phonon scattering, hence it has nothing to do with the crystal structure of TiO2 itself. Thus, from the spectrum of the TiO2 precipitates in the acidic solution with 0.5M of Ti4+ after the reaction at 50oC for 4hrs (“After” marked in this figure), the precipitates appear to be formed as a rutile crystalline phase in the precipitation reaction before the drying process. In general, stable hydroxide such as Ti(OH)4 is transformed into amorphous, anatase and rutile phases, in that order, as the heat-treatment temperature is increased. However, in the XRD results shown in Fig. 2, the crystallinities of as-prepared and HPPLT TiO2 powders, heat-treated after being obtained at 50oC for 4 hrs, are quite different from that of the P-25 powder, which consists of mixed rutile and anatase phases, because they show a pure rutile phase regardless of the heat treatment temperatures. Because the anatase

Fig. 2. XRD patterns for the TiO2 powder by HPPLT before and after the calcinations at various temperatures for 1 hr in air.


Fig. 3. Fluorescence spectrum profiles during degradation of 4-chlorophenol in the rutile TiO2 suspension (experimental condition: λex= 250 nm, λemi=310 nm).

phase in the HPPLT TiO2 powders heat-treated below 650oC is not detected at all, the TiO2 powder is thought to have already had a pure rutile phase alone when it was precipitated in the solution. That is to say, it can be thought that the rutile TiO2 ultrafine powder produced from the aqueous TiOCl2 solution by HPPLT has been crystallized not by the heat treatment in the drying process but directly during the reaction in the precipitation process. In order to evaluate photo-oxidative characteristics for 4chlorophenol using HPPLT TiO2 powder, first of all the aqueous solution containing 1.0 mmol 4-chlorophenol and 1.0 g/L TiO2 powder was irradiated together with magnetic stirring for 15 min under UV light, and then the degree of decomposition of 4-chlorophenol by the photocatalytic reaction, as shown in Fig. 3, was detected by its fluorescent characteristics peak at 310 nm to see whether it was destroyed or not. As the irradiation time of UV light for the photocatalytic reaction increases, the characteristic peak intensity at 310 nm decreases gradually, showing the decomposition of most of the 4-chlorophenol at 15 min. At the same time the peak intensity at about 360 nm increases gradually with the irradiation time. It can be said, thus, that the formation of intermediate organic products having fluorescent characteristics at the peak of about 360 nm has occurred with the decomposition of most of the 4-chlorophenol itself by the photocatalytic reaction. Fig. 4 shows the concentrations of the 4-chlorophenol, − intermediates and Cl -ion, and the pH value in the aqueous solution irradiated under the UV light for the quantitative evaluation from the result of Fig. 3. Here, the concentration

Fig. 4. The concentrations of 4-chlorophenol (4CP), intermediates − and Cl -ion, the changes in the pH value with degradation of 4CP in the rutile TiO2 powder suspensions. For comparison, the concentrations of 4CP with the degradation of 4CP in the P-25 and anatase TiO2 powder suspensions were also added.

of the 4-chlorophenol was calculated from the comparison of fluorescent intensities at 310 nm with irradiation time when using the HPPLT TiO2 powder as a photocatalyst. This method was also applied for both of P-25 and anatase TiO2 powders. − Because the concentrations of the Cl -ion and intermediates, and the pH value in the solution for these powders changed all together with the same decreasing or increasing tendencies, the experimental results for the HPPLT TiO2 powder alone are shown in this figure. Also, the concentration of the intermediates produced during the photocatalytic reaction was calculated from the difference between the amounts of total organic carbon (TOC), as shown later in Fig. 7, and the concentration of 4-chlorophenol calculated in this figure, assuming it has six carbons. It can be seen that there are no adsorptions of 4-chlorophenol to the HPPLT and P-25 TiO2 particles except for the anatase powder with a large adsorption, after stirring in the dark room for 30 min. However, as the UV light irradiation time increases, the anatase TiO2 powder, regardless of its large adsorption of 4-chlorophenol, hardly decomposes 4-chlorophenol from the solution, while the other two powders decompose 4-chlorophenol greatly − while showing the increases in the Cl and intermediate concentrations in the solution. In that, the HPPLT TiO2 powder as a rutile phase finally shows a larger decrease in the decomposition of the 4-chlorophenol than the P-25 with the irradiation time. Therefore, it can be thought that the HPPLTed TiO2 powder was more effective than the P-25 powder for the large decomposition of 4-chlorophenol in the aqueous solution by the photocatalytic reaction although there appeared no adsorption for 4-chlorophenol in the initial reac-


tion, due to no photocatalytic effect of the anatase TiO2 powder on the decomposition of 4-chlorophenol. By the way, Al-Ekabi et al. [10] and other researchers have reported that 4-chlorophenol became finally mineralized to CO2 and HCl after going through the formation of many intermediates during the photocatalytic reaction for 4chlorophenol using TiO2. But it is thought that by considering the formation of many intermediates by the decomposition of 4-chlorophenol using the photocatalytic reaction, the investigation for changes in the TOC rather than the concentration of 4-chlorophenol itself should be carried out together with a longer irradiation time of the UV light for an appropriate photocatalytic evaluation for various TiO2 powders. Then, the photocatalytic reaction was carried out using various TiO2 powders, like the rutile and anatase TiO2 powders with similar values in the surface area and particle shape in the different crystalline structures, the P-25 TiO2 powder with a smaller surface area consisting of mainly the anatase phase, and slurry typed TiO2 powder, not to be put through the drying process in the HPPLT. Then the amounts of TOC in the aqueous solution with the UV light irradiation time were measured for 300 min, as shown in Fig. 7. Here, the slurry typed TiO2 powder is expected to have many OH- ions attached to its surface. The amount of each of these TiO2 powders used in the reaction was fixed to 1.0 g/L in our reactor conditions because the photocatalytic reaction was not active due to smaller reaction sites and higher light scattering when it was greater or less than 1.0 g/L, respectively. Regardless of various preparation methods for the photocatalysts, there is no adsorption of 4-chlorophenol to all the TiO2 particles contrary to the previous investigations for the fluorescent characteristics. Thus, it appears that the large adsorption of 4-chlorophenol to the nanotube TiO2 powder obtained in the fluorescent characteristics was caused due to the transfer to unknown organic materials because it was confirmed that the characteristics of 4-chlorophenol was easily recovered by putting a HNO3 droplet into the aqueous 4-chlorophenol solution containing the anatase TiO2 powder, which had complex and various fluorescent peaks. In Fig. 5, as the irradiation time of the UV light increases, the mineralization of organic carbon from the aqueous solution to CO2 and HCl was underway varying with the preparation methods for the powders rather than with the difference in the crystal structures. The slurry typed TiO2 powder with a rutile phase, confirmed from the Raman spectrum of Fig. 1, shows weak photocatalytic ability for 100 min of the irradiation time, but after 100 min the amount of TOC decreases somewhat more rapidly than in the case of UV light alone without TiO2 powder. It can be seen that the rutile TiO2 powder showed almost the same decomposition rate for 4-chlorophenol as that of the P-25 powder at more than 60 min. However, the rutile TiO2 powder shows a 10% faster decomposition rate than does the P-25 powder in the end because

Fig. 5. Variations in the amount of total organic carbon (TOC) with the degradation of 4CP by photocatalytic process with/without TiO2 powder under UV light, where four kinds of powders - rutile, slurry typed, P-25 and anatase TiO2 - powders were used in the process with the TiO 2 powder.

its decomposition rate is 2 times faster than that of the latter in a time shorter than 30 min. On the other hand, the anatase TiO2 powder with almost the same values in the surface area and thickness of the primary particle as that of rutile TiO2 powder has a similar decomposition for 4-chlorophenol to that of the P-25 powder in the early stage, but with the increasing of the irradiation time its decomposition rate decreases gradually, and then finally showed the smallest amount of decomposition, although the result is not shown in the figure. Therefore, at the starting of the reaction in Fig. 3, it is once more apparent that the large decrease in the concentration of 4-chlorophenol by the fluorescent characteristics was not by adsorption to the TiO2 particles but by the transfer of 4-chlorophenol to another unknown organic during stirring at room temperature. Also, it can be thought that the superior photocatalytic effect of the rutile TiO2 powder by HPPLT compared to the other powders would be due to the preparation method to give the acicular typed shape of the primary particle likely without defects at the TiO2 surface, supporting many reaction sites as well as efficient utilization of the valence band-holes produced by the UV light irradiation without the electron-hole pair recombination. Fig. 6 shows the photocatalytic results of the rutile and slurry typed, P-25, and anatase TiO2 powders in the 100 mL batch type reactor experiment, when 2.0 g/L of each of these powders were added to aqueous Cu-EDTA solutions having about a 50 ppm Cu ion concentration for the photo-reductive reaction. In this figure, the concentration remaining in the solution at 0 min of irradiation time indicates adsorbed Cu ion concentration to the TiO2 particle in the dark room for 30 min. It is shown that both the initial Cu ion adsorption and


because the photocatalytic reactivity of the TiO2 powder had been enhanced for 24 hrs by increasing the drying time at 50oC. From the above results and SEM/TEM analysis, the acicular shape of the primary particles with a 3~7 nm very thin thickness in the agglomerated shape of the HPPLT powder, rather than the round one in the easily dispersed P-25 powder, seems to have greatly affected the photocatalytic reaction to have little electron-hole recombination, in addition to a higher value in the surface area. In the meantime, it was confirmed that the Cu and Pb lumps obtained by the photocatalytic reaction were easily oxidized or converted into metal oxide or metal-carbonate forms in the aqueous solution even in deaeroated conditions. Thus, our next research for the purification treatment of the wastewater will be carried out to accurately control the oxidizing or reducing atmospheres. Fig. 6. The Cu ion concentrations remaining in the aqueous solutions with the degradation of Cu-EDTA by the photocatalytic process with TiO2 powder under the UV light, where four kinds of powders - the rutile, slurry typed, P-25 and anatase TiO2 - were used in the process with the TiO2 powder.

removal rate of the rutile TiO2 powder is larger and faster than those of the slurry typed, P-25 and anatase powders. The slurry typed TiO2 powder shows the largest Cu ion adsorption probably due to the existence of OH- at the TiO2 surface, but shows very slow or delayed photo-reductive results because it had almost the same amount in the reduction as in those of the other powders after the reaction for at least 60 min. This may be due to the existence of OH- acting as an obstacle to the role of EDTA, the sacrificial donor agent, which should react first with the valence band-holes rather than the conduction band-electrons photo-excited under the UV light irradiation. Then, it is likely that the transfer of the photo-excited electrons by the UV light irradiated on the TiO2 surface into the reduced Cu ion is terribly inhibited. In the HPPLT and P-25 TiO2 powders having a full or partial rutile phase, respectively, the higher the value of the surface area is the larger the adsorption is, and then the HPPLT TiO2 powder as a rutile phase has superior photo-reductive properties to that of the P-25 powder in the Cu-EDTA solution. However, the anatase TiO2 powder with a high surface area as well as similar particle shape shows no adsorption of the Cu-EDTA at the initial reaction. Furthermore, the photocatalytic ability for the anatase TiO2 powder is very low. This strongly means that the photocatalytic reaction of the ultrafine TiO2 powder depends on the powder preparation method rather than the value of the surface area or the difference in the crystalline structures. It was confirmed even in the preparation method of the HPPLT that the removal amount of the OH- ions attached to the TiO2 surface was very critical in enhancing the photocatalytic ability of the TiO2 powder

4. CONCLUSIONS Redox abilities of rutile TiO2 powder, consisting of 3~7 nm thick and 70~150 nm long acicular particles and having about a 200 m2/g BET surface area obtained by a homogeneous precipitation process in aqueous TiOCl2 solution at 50oC for 4 hrs, were investigated using photocatalytic reaction in aqueous 4-chlorophenol, Cu- and Pb-EDTA solutions. Its abilities were then compared with those of commercial P-25 powder, and those of ultrafine TiO2 powder with a pure anatase phase obtained from aqueous TiOCl2 solution, consisting of similar acicular particles and surface area as those of the rutile powder. The conclusions from the results are as follows: Firstly, the rutile TiO2 powder was more effective, showing the fastest decomposition rate and the largest amount in the photo redox, than the anatase or P-25 powder when 4chlorophenol and metal-EDTAs were completely decomposed by the photocatalytic reaction. Secondly, the anatase TiO2 powder unexpectedly showed the slowest decomposition rate and the smallest amount in the same experiments regardless of their similar shapes and surface areas. Thirdly, the excellent photo redox abilities in the rutile TiO2 powder with the anatase phase and a large surface area, was likely to give no defects at the TiO2 surface, regardless of the crystalline structures when having similar particle shapes and values in the surface area. Fourthly, many OH- attached to the surface of a TiO2 particle appear to interfere with the adsorption of decomposing target materials to the TiO2 surface in the solution during the photocatalytic reaction, resulting in a delay in the reaction. Therefore, in the wet processing for preparing ultrafine TiO2 powder, the appropriate removal of OH- from the TiO2 surface in the drying process would be essential to superior photocatalytic abilities.


REFERENCES 1. I. Ilisz, Z. Laszlo and A. Dombi, Appl. Catal. A 180, 25 (1999). 2. X. Li, J. W. Cubbage, T. A. Tetzlaff and W. S. Jenks, J. Org. Chem. 64, 8509 (1999). 3. J. Chen, D. F. Ollis, W. H. Rulkens and H. Bruning, Colloids & Surf. A 151, 393 (1999). 4. Z. Zhang, C. C. Wang, R. Zakaria and J. Y. Ying, J. Phys. Chem. B 102, 10871 (1998). 5. C. C. Wang, Z. Zhang and J. Y. Ying, Nanostruct. Mater. 9,

583 (1997). 6. G. A. Somorjai, Chemistry in Two Dimensions: Surface, p. 551, Cornel University Press, Ithaca, U.S.A. (1981). 7. S. J. Kim, S. D. Park, C. K. Rhee, W. W. Kim and S. Park, Scripta mater. 44, 1229 (2001). 8. R. R. Bacsa and J. Kiwi, Appl. Catal. B 16, 19 (1998). 9. D. S. Seo, J. K. Lee and H. Kim, J. Kor. Ceram. Soc. 37, 700 (2000). 10. H. A. Ekabi, N. Serpone, E. Pelizzetti, C. Minero, M. A. Fox and R. B. Draper, Langmuir 5, 20 (1989).