Mechanisms and Intermediates

As a result of the current interest and investigative activity into various aspects of the PCO process, some consistent mechanisms of oxidation are beginning to emerge, although many conflicting ideas are still ubiquitous. As noted, PCO demonstrates a Langmuir-Hinshelwood relationship with respect to organic substrate concentration [Equation (11)]. It has been assumed that this dependence was directly due to a Langmuir isotherm dependence of the substrate adsorption. For example, Matthews [56] investigated both the PCO and dark adsorption of methylene blue onto TiOa. Both were described using a Langmuir isotherm expression, suggesting that adsorption of the substrate is a critical step in the oxidation. Several descriptions of organic adsorption and competition from reactants and products such as 02 and CP have

Initial Toluene Concentration (niM)

Figure 5 Enhancement of PCO rates in the presence of dissolved copper (10-5 M) and iron (10~5 M) as a function of substrate concentration. (From Butler and Davis [88].)

Initial Toluene Concentration (niM)

Figure 5 Enhancement of PCO rates in the presence of dissolved copper (10-5 M) and iron (10~5 M) as a function of substrate concentration. (From Butler and Davis [88].)

been published. For example, Al-Ekabi and Serpone [21] studied the photocatalytic oxidation of chlorinated phenols. They attempted to establish the role of organic adsorption in the PCO process, although possible competitive effects of oxygen are ignored.

Kormann et al. [66] suggested that under proper pH conditions, when the organic substrate is favorably adsorbed, the reactions take place at the surface of the catalyst; a second-order kinetic expression was presented based on adsorbed substrate and oxygen concentration, both of which were quantified using Langmuir adsorption expressions. Conversely, CIR-FTIR examination of a photocatalyst suspension revealed no chemical bonding of 3,4-dichlorobiphenyl to the Ti02 [47].

Currently, the most widely accepted PCO mechanism employs hydroxyl radicals as intermediates in the oxidation process. Adsorption of the substrate is not required to produce Langmuir-Hinshelwood kinetics if electron-hole and radical recombination reactions are considered in the overall photocatalytic mechanism [7,89].

Evidence has consistently accrued implicating OH- radicals in the photocatalytic oxidation of organics. Studying the role of OH- radicals in the oxidation of salicylic acid, Matthews [20] found that the rate of degradation of the organic acid decreased in proportion to the increase in OH- radical scavenger concentration. ESR spectroscopy has confirmed an abundance of OH-radicals produced upon the photoexcitation of Ti02 [15,16], As discussed earlier, Cunningham and Srijaranai [19] noted a photocatalytic kinetic isotope effect using DzO instead of H20, but not when an isotopic exchange was made on the organic substrate.

The OH- radicals that are produced at the catalyst surface may attack organic substrate that is also adsorbed on the catalyst surface or they may desorb and then react with the organic in the bulk solution [7,43], Recently Lawless et al. [90] conducted studies to identify the role and importance of organic degradation by free versus Ti02 surface-bound OH- radicals.

Externally produced hydroxyl radicals were quickly adsorbed by Ti02 particles, producing a Ti02 surface that had properties similar to those of the surface under PCO conditions. This suggests that surface-bound OH- radicals are formed during PCO. Their results also imply that the organic oxidation occurs via these adsorbed OH- species.

Minero et al. [91] noted that decafluorobiphenyl (DFBP) that is strongly adsorbed onto A1203 is not photocatalytically oxidized when mixed with illuminated Ti02. However, degradation of adsorbed DFBP does occur when hydroxyl radicals are supplied by H202 in the presence of UV light. This evidence also suggests that the organic substrate must be in close contact with the Ti02 surface for degradation to occur and that the OH- radicals that are formed do not travel far into the bulk solution.

Minero et al. [36] compared the diffusion rate of hydroxyl radicals in aqueous solution with experimentally determined organic oxidation rates. Based on the assumption that these rates are equal, it was shown that the OH- radicals could not diffuse far from the photocatalyst into the bulk solution before reacting, even at very low concentrations of the organic substrate. Thus photocatalytic degradation processes occur either on or very near (within a few monolayers) the particle surface.

Others have proposed that OH- radicals formed by oxidation of surface water or hydroxo groups eventually enter the solution to oxidize organics [16,92]. The formation of aqueous H202 [through reactions (7b) and (7d)] and hydroxylated organic intermediates during photo-catalysis provided the basis for their conclusion.

Peterson et al. [67] conducted a series of experiments using Ti02 immobilized on a conducting carbon paste to study some of the reactions suggested to occur in a photoelectrochem-ical slurry cell. If hydroxyl radicals were formed only at the Ti02 surface, they would act as recombination centers and form H202, producing an anodic photoresponse in their system. However, if the OH- radicals were escaping into the bulk solution, a cathodic photoresponse would be expected. A cathodic response was confirmed in the cell, leading to the conclusion that OH- radicals do escape into the bulk solution.

Turchi and Ollis [7] have presented possible reaction mechanisms based on four types of interactions between photocatalytically formed OH- and the organic reactants:

1. The OH- radical species and the organic substrate are both adsorbed on the catalyst surface.

2. The OH- radical migrates to the bulk solution (free OH- species) and reacts with the organic substrate in solution.

3. The adsorbed OH- radical reacts with the bulk solution organic substrate.

4. The free OH- radical reacts with the surface-adsorbed organic substrate.

Considering rates of recombination between holes and electrons, illumination intensity, and additional information on catalyst physical properties, a mechanistic rate expression was derived for the PCO process. The resulting expression was in the form of Equation (11). The rate parameter k was found to be constant for all four examined mechanisms and predicted to be a function of the catalyst properties and reaction conditions; it was found to be essentially independent of the organic substrate involved in the photocatalytic oxidation process. Experimental results have confirmed this prediction of k being reactant-independent [7].

Turchi and Ollis proposed that, for the mechanisms in which the organic substrate is present in the bulk solution, the parameter K is a function of the second-order reaction rate constant for reaction of the OH- radical and the dissolved organic substrate. Conversely, if the organic substrate is adsorbed on the catalyst surface, K is also proportional to the equilibrium adsorption constant of the organic substrate.

Mechanisms of degradation by photocatalysis have been presented for several organic substrates. Compounds such as hydroquinone, pyrocatechol, 1,2,4-benzenetrioI, pyragallol, and

2-hydroxy-1,4-benzoquinone have been isolated as intermediates for the photocatalytic oxidation of phenol using Ti02 [24], Such hydroxylated compounds are suggestive of hydroxyl radical attack. Catechol and quinone were also detected in the PCO of phenol by Augugliaro et al. [33]; hydroquinone and a phenol dimer were observed by Tseng and Huang [34], A mechanism involving hydroxyl radicals from illuminated TiOz for 2,4,5-trichlorophenoxyacetic acid and 2,4,5-trichlorophenol was postulated by Barbeni et al. [41]. Chlorinated quinones, multihydroxylated chlorobenzenes, and other aldehyde and ketone chloroaromatics were isolated. Final products were C02 and HC1. Methyl catechol, methyl resorcinol, and methyl hydroquinone, all hydroxylated cresols, were isolated during the PCO of o-, m- and p-cresols [43], A detailed examination of the PCO of fluorinated phenols was presented by Minero et al. [36]. Stoichiometric production of C02 and F~ was found; fluoride appearance was much more rapid than that of C02 and nearly coincides with the disappearance of the parent compounds, suggesting that defluorination occurs early in the degradation process. Intermediate products that were identified include mono- and dihydroxylated derivatives of the parent compounds.

A plausible mechanism for phenol oxidation to catechol is that of reaction (21) [24], Similar mechanisms can be invoked for the formation of hydroquinone from phenol and for reactions of chlorinated and fluorinated phenols in which substituted catechol and hydroquinone compounds have been detected. Continued hydroxyl radical attack should follow mechanisms similar to those presented where OH- radicals are formed from sources other than photocatal-ysis, i.e., high-pH ozone decomposition or radiolysis.


A detailed mechanism for the degradation of triazine herbicides has been published [45]. The primary degradation pathway is through alkyl chain oxidation. The final product for the herbicide PCO is cyanuric acid, which is stable and not subject to further degradation [44]. At this point, cyanuric acid is one of only a few organic compounds that are photocatalytically unreactive.

Several publications have investigated the PCO of other nitrogen-containing heterocyclic organic compounds. Low et al. [58] monitored the concentrations of ammonium, nitrate, and carbon dioxide from several of these compounds. The appearance of ammonia was very rapid, indicating quick cleavage of the ring to release the nitrogen. Rates of N03~ production suggested that its formation was entirely due to the oxidation of previously synthesized ammonium. Ammonium was also detected in the PCO of nitrophenols [57], indicating a reduction of the substrate nitro group.

As with all organic oxidation processes, there is concern over the formation of intermediate compounds that may be more toxic than the parent material. Unknown intermediate compounds from the partial PCO of pentachlorophenol and 2,4-dichlorophenol demonstrated an increase in toxicity to an activated sludge system over that of the parent compound [93].


Nevertheless, long-term PCO of these compounds, as well as PCO of methyl vinyl ketone, decreased toxicity to activated sludge.

In summary, the mechanistic work with PCO has demonstrated the formations of hydrox-ylated intermediates, indicative of hydroxyl radical oxidative pathways. Less understood are the formation and interaction of hydroxyl radicals and organic substrate, either at the photo-catalyst surface or at some small distance into the solution. More studies are needed for complete comprehension of these mechanisms.

0 0

Post a comment