Photobleaching and photomineralization of azobenzene and substituted azobenzenes in aqueous solution by photocatalytic membranes immobilizing titanium dioxide

The kinetics of photobleaching (by spectrophotometric analysis) and integral photomineralization (by total organic carbon (TOC) analysis) of azobenzene (I) and substituted azobenzenes in aqueous solution were followed in laboratory-scale runs on photocatalytic membranes immobilizing 30 +/- 3 wt.% of semiconductor TiO2. Experiments were carried out by the technique described in preceding papers of this series, employing stoichiometric hydrogen peroxide as the oxygen donor. The following azobenzenes were examined: (4-diethylamino)-phenylazobenzene (II), 4'-(((4-diethylamino)phenyl)azo) benzoic acid (III), 4'-(((2-amino-5-diethylamino)phenyl)azo) benzoic acid (IV), 4'-(((2-acetamido-4-diethylamino)phenyl)azo) benzoic acid (V), 4'-(((4-dimethylamino)phenyl)azo) benzenesulphonic acid, sodium salt (VI) and 4'-(((2-acetamido-4-diethylamino)phenyl)azo) benzenesulphonic acid, sodium salt (VII). From the Langmuir-Hinshelwood treatment of the initial rate data as a function of the initial concentration ((0.10-1.0) x 10(-3) M), the kinetic parameter k and the pseudo-thermodynamic parameter K for photobleaching were obtained. With regard to photobleaching, I and II were certainly the most reactive, followed by IV. The remaining molecules showed a photooxidation rate of one-third to one-quarter of that of I chosen as reference structure. Consequently, the presence of an amino group in the 4-position (II) does not stabilize the azobenzene structure against photo-oxidation leading to bleaching, whereas the same group in the 2-position (IV) decreases the photobleaching rate by about 40% when a carboxylic group is also present in the 1'-position. Acetylation of this amino group, such as in V, decreases the photobleaching rate more markedly. With regard to photomineralization, it was observed that, when photobleaching was virtually complete, a certain amount of TOC was already mineralized. The maximum amount of TOC remaining at the end of photobleaching ranged from about 90% to about 30%, varying with the dye structure and initial concentration as well as with the power and type of irradiation source. By examining the TOC concentration profiles as a function of the substituted azobenzene structure, the following hypotheses were proposed: 1. during the photobleaching period, the ring containing the diethylamino group breaks down (more markedly if further amino or acetamido groups are present in the structure), and photomineralization of the other ring occurs more slowly; 2. both rings break down, within certain limits, during photobleaching; however, the aliphatic fragments containing carboxyl or sulphonic groups are mineralized more slowly. The fact that a small initial plateau in the TOC profile is followed by another more evident plateau at the end of the photobleaching period, for both VI and VII, suggests that hypothesis (1) is more probable in these cases. When the second plateau is reduced to a sigmoidal curve or an inflection point, hypothesis (2) also needs to be considered. (C) 1997 Elsevier Science S.A.