AmericanChemicalSociety.com

Dibenzothiophenes (DBTs) are sulfur-containing aromatic compounds found in diesel and other fuel stocks. The traditional hydrodesulfurization (HDS) treatment process does not remove these compounds efficiently. Therefore, HDS is not suitable for producing fuels that meet the more stringent standards for Ultra-Low Sulfur Diesel. Oxidative desulfurization (ODS) is currently receiving significant attention as an alternative to HDS. ODS is a two-step process where chemical oxidation of the sulfur compounds is followed by extraction of the oxidized products from the fuel. The reaction of the DBTs with oxidants typically requires the use of a catalyst. The goal of this project is a detailed understanding of the metalloporphyrin catalyzed oxidation of DBTs.

Work on this project can be divided into several main areas.

1. Identification of suitable catalytic conditions

2. Determination of the scope of the reaction

3. Model fuel studies

Metalloporphyrins are known to catalyze reactions of organic substrates by a variety of oxidants. For this work, dibenzothiophene was the first substrate studied. Hydrogen peroxide was chosen as the desired oxidant for our initial studies because of the non-hazardous nature of its reaction by-product, water. Iron porphyrins were surveyed as potential catalysts since most biological porphyrins (hemes) contain iron. The iron derivatives of several porphyrins were employed as catalysts and iron tetrakis-(pentafluorophenyl)porphyrin chloride [Fe(PFPP)Cl] was identified as a promising catalyst. Slow addition of the oxidant was required to prevent excessive catalyst degradation during the reactions. Reproducible results were observed for reactions containing 1% or more catalyst. The sulfoxide and sulfone derivatives of DBT were identified as the products of the reaction by GC-MS. The percent conversions were determined by HPLC analysis of the reaction mixtures.

Previous metalloporphyrin studies from the literature had shown that solvent composition often plays a significant role in reactions catalyzed by Fe(PFPP)Cl. Therefore, reactions were performed in the presence and absence of alcoholic cosolvents. A mixture of dichloromethane and acetonitrile was suitable for dissolving the catalyst and DBT. Reasonable (>75%) DBT conversions were observed regardless of the additive. Nevertheless, relative to a CH₂Cl₂/CH₃CN mixture, additional water was observed to decrease catalytic activity and slightly higher conversions were observed in the presence of alcohols. In additional experiments, the reaction was monitored by UV-visible spectroscopy and DBT was consumed approximately 3-times faster in the presence of methanol compared to butanol. Based on these results, standard reaction conditions were identified. The reaction conditions for subsequent experiments were performed in a CH₂Cl₂/CH₃CN/methanol mixture with 50 mM substrate, 5 equivalents of hydrogen peroxide and 1% catalyst.

The influence of substrate structure on the reactivity was investigated. Reactions of benzothiophene, 4-methyldibenzothiophene and 4,6-dimethyldibenzothiophene were performed. All three substrates were oxidized under the standard reaction conditions. Conversions of >99%, 98%, 97% and 46% were observed for BT, DBT, 4-MDBT and 4,6-DMDBT, respectively.  For all substrates, both the sulfoxide and sulfone products were observed by GC-MS, and minimal conversion was observed in the absence of the catalyst.

The results discussed so far demonstrate that Fe(PFPP)Cl catalyzes for the oxidation of DBTs.  The next step of this project investigated the use of Fe(PFPP)Cl as the catalyst under conditions that are more relevant to the processing of real fuels. In this portion of the study, octane replaced CH₂Cl₂/CH₃CN as the main solvent. For reactions where the DBT concentration was 50 mM, high conversions were still observed using only a minor excess of oxidant. Concentrations around 200 ppm for DBT, BT, 4-MDBT and 4,6-DMDBT are more typical for real fuels. Therefore, model fuel studies will involve reactions containing all four substrates at a concentration of 200 ppm. Preliminary reactions with the model fuel have shown that larger excesses of oxidant are required to achieve good conversions in reasonable time periods. Experiments aimed at optimizing conversion of substrate in the hydrocarbon solvent (model fuel conditions) are currently underway.  Submission of a manuscript detailing this work is planned upon completion of the model fuel experiments.

Funding of this project was extremely important for the PIs overall research program. Three University of Wisconsin-Eau Claire (UWEC) students without previous research experience began work on this project in September, 2008. All three students gained experience performing catalytic reactions, and the analysis of the reaction mixtures by GC-MS and HPLC. Two of these three students are strongly considering graduate school in chemistry. The third student is considering a career in the area of environmental science, or pursuing a graduate degree in the field of environmental health. 

This project provides a suitable research experience for non-UWEC students with additional funding provided by the PIs NSF-REU grant that started in May 2009. A fourth student from Itasca Community College was supported by a combination of PRF (supplies) and REU (stipend and housing) for 10 weeks during summer 2009. She is continuing her education at the University of North Dakota. It is anticipated that new UWEC students hired in Fall 2010 as well as the PI's REU students in 2010 and 2011 will also contribute to this project. Therefore, PRF funding from this project is integral to the PIs research program and will provide support for UWEC and non-UWEC students (in combination with NSF funds) through the summer of 2011.