58th Annual Report on Research 2013 Under Sponsorship of the ACS Petroleum Research Fund

The following projects were completed during the final year of the grant period. All of the photodecomposition experiments described were done with visible and near-UV light, generally by using a 345 nm cutoff filter, in order to simulate solar irradiation.

A.    Photocatalysis of Chloroform Decomposition

1.      RuCl₆^2-

The hexachlororuthenate(IV) ion catalyzes the photodecomposition of chloroform in a manner very similar to that of the hexachlororhodate(III) ion, though not as efficiently, producing approximately 50 equivalents of COCl₂ (later converted to CO₂) after a half an hour of irradiation with a 100-W Hg lamp.

2.      Dowex 2-X8 (Clˉ form) as a heterogeneous photocatalyst

This was a result of another “failed” experiment, in which the photocatalytic activity of a metal complex supported on an anion exchange resin turned out to be 100% attributable to the resin and not to the metal complex. Only the chloride form of the resin was photocatalytically active. The mechanism is considerably different than that shown above in Eqs (1) through (5), and has not been completely worked out. It appears, though, that CCl₄ is formed in a reaction that reduces dioxygen to peroxide.

Here Eq. (10), in which the resin cationic sites are shown as R⁺,  probably consists of two or more steps. Hydroxide radicals, generated from peroxide, then react with chloroform to produce CCl₃ radicals, which generate phosgene as in Eqs. (3) through (5), with Eq. (2) a chain process. This was the first heterogeneous catalyst we discovered capable of effecting the photodecomposition of chloroform with near-UV light.

3.      CuCl₄^2- heterogenized on Dowex 2-X8

The tetrachlorocuprate(II) ion functions well as a heterogeneous catalyst for chloroform decomposition (though not for dichloromethane decomposition) when supported on a polystyrene anion exchange resin. Interestingly, it functions even better than an equivalent amount of CuCl₄^2- in solution. The reaction does not occur as we had hypothesized, through the photodissociation of chlorine atoms as in Eq. (1). Nevertheless chlorine atoms are generated, from the photodissociation of CCl₄, which is the product formed from chloroform initiated by excited state CuCl₄^2-. Chloroform then decomposes by the chain process described above in Eqs. (2) through (5).

B.     Photocatalysis of Dichloromethane Decomposition

4.      FeCl₄ˉ on Dowex 2-X8

The Dowex anion exchange resin in the chloride form catalyzes chloroform photodecomposition, as discussed above, but does not work with dichloromethane. However, the tetrachloroferrate(III) ion heterogenized on this resin is photocatalytically active. While the decomposition process is initiated as we had predicted, by photodissociation of a chlorine atom, the overall process is more complicated than we had predicted based on Eqs (1) through (5), and seems to involve the reduction of O₂ to a peroxide intermediate.

5.      FeCl₃ on silica gel

We hypothesized that iron(III) chloride, which in dichloromethane is the dimer Fe₂Cl₆, would be photocatalytically active in a manner similar to heterogenized FeCl₄ˉ, that is, through photodissociation of a chlorine atom. In laboratory experiments this proved to be true and the two catalysts had similar efficiencies. When both were tested in sunlight, however, the FeCl₃/silica catalyst was at least ten times more effective. This was ascribed to a better overlap of the FeCl₃ absorption spectrum with sunlight than with the laboratory illumination.

C.    Methods Development

6.      Determining the concentration of HCl and other strong acids in chloroform

In order to measure the amount of hydrogen chloride produced during photodecomposition reactions, we developed a spectrophotometric technique making use of the large equilibrium constant for the protonation of tetraphenylporphyrin,

K_eq for this reaction is approximately 10⁹ and although this seems to imply a complete reaction, that is not quite true for micromolar concentrations of HCl. We developed an experimental method to correct for this.

7.      Variation of photodecomposition rates with the mass of catalyst used

In our own experiments and those of others using heterogeneous photocatalysts we noticed two distinct behaviors when the amount of catalyst was increased. For some catalysts the decomposition rate increased towards an asymptotic limit, but for others the decomposition rate increased to a maximum and then decreased. We found an explanation for this in an equation expressing the fraction of light absorbed by the catalyst as a function of the amount of catalyst in suspension. The critical factor is the absorptivity of the catalyst. When much more light is reflected than absorbed, more catalyst particles eventually increase the rate of back-reflection faster than the rate of absorption and the net absorption decreases. When, however, the absorptivity is high, back-reflection is negligible (or smaller than absorption) no matter how many particles are present; however, when the absorbance is high nearly all incoming light is absorbed and doubling the number of particles (for example) can have very little effect, thus the approach to an asymptotic limit.

D.    Dialkyl Carbonates

8.      Green synthesis of dialkyl carbonates using sunlight and no reagent phosgene

Dialkyl carbonates are commonly manufactured through the alcoholysis of phosgene. We developed a method to accomplish this by generating phosgene photochemically in situ, obviating the requirement to buy and store an extremely toxic chemical. Conditions are such that sunlight can be used for the light source and the catalyst, an Amberlite anion exchange resin, is extremely cheap. Diethyl carbonate is of particular interest due to proposals to use it as an oxygenating gasoline additive in place of MTBE. Although this particular project developed as a spin-off from work of ours on photocatalysis of chlorinated hydrocarbons by anion exchange resins, and not directly from the project proposed for this PRF grant, in fact it fits the criteria for PRF support quite well, illustrating how basic research can point towards practical results in unforeseen directions.