Hydroxylated anatase surface

Rapid e / h recombination, the reverse of equation 3, necessitates that D andM be pre-adsorbed prior to light excitation of the Ti02 photocatalyst. In the case of a hydrated and hydroxylated Ti02 anatase surface, hole trapping by interfacial electron transfer occurs via equation 6 to give surface-bound OH radicals (43,44). The necessity for pre-adsorbed D andM for efficient charge carrier trapping calls attention to the importance of adsorption—desorption equihbria in... 

The authors assume that the activation of Mg(OH)2+Ti02 (anatase) mixture involves the transition of proton from OH groups on anatase surface to the hydroxyl group of Mg(OH)2 which results in the formation of water and Mg-O-Ti bond, the latter being a nucleus for the formation of MgTiOj phase during consequent calcination and agglomeration. 

The amphoteric character of the hydroxylated Ti02 surface has been demonstrated by Boehm and co-workers in a series of potentiometric titration experiments performed both with anatase and rutile samples. Approximately half of the total amount of hydroxyls present on the Ti02 surface, which underwent neutralization with a diluted (0.01 M) sodium hydroxide solution, has been described as relatively strongly acidic, with a pK value of 2.9, and the remainder as weakly acidic with a pK of 12.7. However, the use of a Langmuir-type adsorption equation as a basis for estimating these acid dissociation constants has raised questions about the real significance of the above values. ... 

The essential point arising from the above discussion is that the saturation coverage of titanium dioxide with the surface-bonded peroxo species, photogenerated in alkaline or neutral solutions, ranges most likely from 4 to 5 peroxo groups per nm . The latter value is close to the total number of active OH groups present initially on each nm of fully hydroxylated (and unilluminated) anatase surface. 

The results were different in one important respect from the earlier ones for photolysis of Fe(III) complexes in solution. The Fe(II) yield here showed very little dependence on t-butanol concentration and indicated a primary quantum yield of 0.170, approximately half the extrapolated intercept for methanol of 0.330. In ferric perchlorate photolysis studied earlier, both alcohols gave a common extrapolated yield equal to the independently determined primary yield for hydroxyl radical production. In other words, at the TiOa surface we appear to generate more "OH radical for reaction with CHaOH than "OH radical" to react with t-butanol. From these results, we infer that a hole reaching the anatase surface may produce one of two distinct oxidants in approximately equal quantity. These are a species capable of abstracting hydrogen (e.g. the OH radical) and a second less reactive oxidant. Preliminary results from parallel experiments... 

The most commonly employed titania modification in catalysis is anatase, whereas there are fewer applications of rutile. In many reports, mixed ana-tase-rutile samples were investigated. In sum, at least 12 types of OH groups on the anatase surface have been proposed by different authors (115). The most frequently reported hydroxyls are characterized by IR bands around 3720-3715 and 3675 cm (27,28,115,269,272-274,330,357,504,651, 678,679). These two bands are well discemable after evacuation at a temperature of 673 K. A band around 3640 cm is also often reported and characterizes less stable hydroxyls (230,273,356,504,651). For a Degussa P25 sample (containing rutile), an additional band at 3662 cm was reported (273). An increase in the evacuation temperature results in a complex spectrum containing a set of low-intensity bands. 

The concentration of the various types of surface hydroxyls, the surface acidity constants (Ki nt and K2 ), the pzc and the surface charge oo= F [SOH2+] - [SO ]) are determined using potentiometric titrations [44-54]. In the case of titania which is a mixture of simple oxides (anatase and rutile) only the values of pzc and oq may be determined [55]. 

Further chemical evidence for the existence of hydroxyl groups on anatase was obtained by Herrmann (305). Anatase made by flame hydrolysis of TiCl4 was used. Its surface area was 60 m /gm. The... 

Although the surface models for anatase and rutile, as proposed by different authors, are idealized and differ from each other in details, it can certainly be concluded that coordinatively unsaturated Ti4+cations, O2- ions, and OH groups in widely varying configurations should be exposed on partially hydrated and/or hydroxylated surfaces. Depending on the local environments of these sites, a wide spectrum of possible intermolecular interactions should be the consequence which may render specific adsorption processes possible. Finally, the ease of the surface reduction of titanium dioxides due to hydrocarbon contamination (19) leads to the formation of new types of surface sites and to drastic changes of the surface properties. 

Similarly, for the surface hydroxyl groups of the anatase, the increase of the electron density on the oxygen atoms may not only ease the dissociation of the hydrogen but also its adsorption. 

The substantial parameter at the modeling of the electric double layer at metal oxide-electrolyte solution interface is a number of the hydroxyl group per surface unit of the oxide. For the titanium dioxide, although different crystalline faces form the surface [rutile 60% of the surface is formed by the face (110) whereas for anatase by (001)] the same density 12.8 of —OH group/nm2 is assumed [28]. That results from the very similar intersection of the elementary cells of the mentioned face, which have the highest density of the atoms in both oxides. 

A separate problem, having an influence on the properties of the surface is the purity of the experimental sample. It means not only the presence of anionic or cationic impurities, but also the presence of the one crystalline form in another one. For example, the even small amounts of the anatase on the surface of the rutile may essentially change the properties of the latter. At the beginning, the edl was characterized by two values pHpzc and pHiep. The critical review of these data was made by Parfitt [164]. A comprehensive survey of pHpzc values, up to nineteenth was done by Lyklema [22]. After the site binding model was worked up, the edl is characterized by the surface hydroxyl group reaction constants. The values of these constants for TiC>2 in different solutions are presented in the papers by James and Parks and Schindler [11,16]. 

---