Reduced sulfur compounds such as methanethiol (MSH), dimethylsulfide (DMS) and dimethydisulfide (DMDS) are nauseous by-products produced by a great number of industrial processes. Oxidation of these reduced sulfur compounds in polluted atmospheres and hence the decrease of their harmful and malodorous effects is thus a matter of concern in numerous industrial and water treatment plants. Photocatalytic treatment of gaseous flow polluted by these sulfur compounds has been actively investigated for the last few years.The first part of the paper is devoted to a literature review on the different TiO2-based photocatalytic processes designed for the oxidation of these gaseous compounds. The comparison of their efficiency is done according to the process parameters: batch or flow reactors, photocatalytic materials, residence time, gas flow, pollutant nature and concentration, relative humidity, . . . . Special attention is paid to the poisoning of the photocatalytic material and to its possible recycling.In the second part of the paper, alternative materials based on aromatic photosensitizers (9,10-dicyanoanthracene, 9,10-anthraquinone) deposited or grafted on silica matrices are then presented and their efficiency compared to more conventional TiO2-based materials. It is demonstrated that the oxidation products are totally different from those obtained with TiO2. With the photosensitizing materials, singlet oxygen addition is shown to be the major pathway, leading to sulfoxide and sulfone starting from DMS and to methyl methanethiosulfonate starting from DMDS. With TiO2-based materials, in the absence of water and hence of hydroxyl radicals, products arising of C-S and S-S bond cleavage are mainly obtained: disulfide from DMS and CH3SSSCH3 together with CH3SCH2SSCH3 from DMDS. These latter products may be accounted for by electron transfer from sulfide or disulfide to photogenerated holes, leading to radical mechanisms. Mineralization to CO2 and H2O is also shown to occur with DMS, but is not favoured under these conditions, due to the absence of water and of a too fast gas hourly space velocity (GHSV).The advantages-drawbacks of the two kinds of materials are presented. (c) 2007 Elsevier B.V. All rights reserved.

The synthesis and complete characterization of layered double hydroxides (LDHs) where 4-benzoyl benzoate is either intercalated or adsorbed has been achieved and their photo-sensitizing efficiency for the oxidation of di-n-butylsulfide has been compared. XRD data are noticeably different for the two materials and the presence of an intercalated new phase is obvious for the first material. Elemental analysis and thermogravimetric analysis (TGA) results indicate that intercalated LDH contains much more benzoyl benzoate than adsorbed LDH, while TGA further suggests stronger interactions between the host and the organic anion in the former case. The Fourier transform/infra red (FT/IR) spectrum of the intercalated sample shows no modification of the carbonyl vibration of the benzoyl benzoate moiety upon intercalation within LDH, whereas the diffuse reflectance UV (DRUV) spectrum is strongly modified relative to that of the adsorbed sample.Both intercalated and adsorbed LDHs proved to be efficient and selective sensitizers for the photo-oxidation of di-n-butylsulfide in oxygenated acetonitrile solution. The intercalated photo-sensitizer was efficiently recycled for three successive runs. According to X-ray photoelectron spectroscopy (XPS) analysis of the material after irradiation, the major drawback of these supported sensitizers is the partial replacement of the benzoate anion within the LDH by sulfate and sulfonate anions arising from sulfide oxidation. (c) 2005 Elsevier Inc. All rights reserved.

Two different photosensitizers, 9, 10-dicyanoanthracene (DCA) and benzophenone (BzO) or a silica bound derivative (BzO-Si) have been compared for the photooxidation of di- n -butyl sulfide and di- n -butyl disulfide. With either photosensitizer, sulfide photooxidation in acetonitrile leads very efficiently to sulfoxide, with sulfone and disulfides as by-products. Although an electron transfer mechanism has previously been established starting with DCA, our results are indicative of two competitive mechanisms using BzO as the photosensitizer, instead of singlet oxygen addition and electron transfer. The more sluggish photooxidation of disulfides leads to a complex mixture of products, among which n -butyl butanethiosulfonate and strong acids (alkylsulfonic and sulfuric) are the major ones. The relative ratio thiosulfonate: acids depends, among other factors, on the medium polarity with acid formation favored starting with BzO or BzO-Si in a methanol-water mixture. An electron transfer mechanism only can account for the observed products. Superoxide anion, the formation of which is much easier starting from BzO than from DCA, is suggested to play a crucial role in this oxidative radical pathway. Starting from disulfides, grafted benzophenone is more efficient for acid formation than its soluble counterpart. As this photosensitizer can easily be recycled, an easy and smooth way to acid formation is thus available, provided that the reaction solvent is properly chosen.

Different spectroscopic methods are compared to quantitatively determine organic compounds adsorbed or grafted on silica. The studied example is a derivative 1 of a well known photosensitizer, benzophenone. Transmission FTIR, diffuse reflectance FTIR (DRIFT) and UV (DRUV) spectroscopy are used to measure the concentration of adsorbed 1 on silica. The diffuse reflectance spectra are treated according to the Kubelka–Munk theory. It is shown that for such a compound absorbing in the UV range, DRUV spectra actually display a linear correlation between the remission function F(R) and the concentration of adsorbed 1. As this method neither implies any sample preparation nor any problems arising from the supporting silica, it proves to be fast, efficient and with a low detection threshold. For the IR spectra, it is necessary to numerically substrate the spectrum of silica which absorbs the light in this range. Provided that some care is paid to this numerical treatment which takes the sample heterogeneity into account, transmission and DRIFT spectra also display a linear correlation between the absorbance or the F(R) function and the concentration of adsorbed 1. However, sample preparation is much easier for DRIFT spectra as no pellets are made. Moreover, the DRIFT spectra of ground samples appear easier to process than the transmission spectra because of a better baseline and resolution. Although less resolved under 1500 cm−1, the DRIFT spectra of non-ground samples can also be processed and give satisfactorily and rapid results without any possible perturbation of the structural integrity of the sample.