Molecular Modelling Based Design of Selective Depressants for Beneficiation of Dolomitic Phosphate Ores

Processing of dolomitic phosphate ores remains one of the most challenging problems in phosphate ore beneficiation. The beneficiation plant at Jhamarkotra in Rajasthan, India for example is currently using a conventional double stage flotation process which involves first a fatty acids flotation to produce a combined dolomite-apatite concentrate (bulk concentrate) from associated silicate gangue. Dolomite is then separated from this bulk concentrate by flotation with fatty acids collectors and depressing apatite with phosphoric acid and sulphuric acid. This process requires addition of large dosages of reagents (H3PO4 and H2SO4) to depress apatite and to maintain highly acidic pH during flotation. Furthermore several cleaning and scavenging stages are needed to obtain the product of desired quality. We have developed and employed a molecular modelling based methodology of designing selective reagents for apatite-dolomite separation. We have been able to take into ac-count the differences in the distribution of calcium cations on these two surfaces in our design and hence develop novel phosphonic acid based depressants for apatite-dolomite separation with con-ventional fatty acids as flotation collectors. We have demonstrated the utility and power of molecular modelling based reagents design approach through this work on beneficiation of Jhamarkotra dolomitic phosphate ore using our novel phosphonic acids based depressants at pH?s close to natural pH. It was possible to reduce the MgO content of the ore from as high as 13.3 % to 1.5% in a single stage flotation. Our flotation separation results on Jhamarkotra ore with these new phosphonic acids based tailor-made depressants for apatite-dolomite flotation separation are discussed in this paper along with results of corresponding molecular modelling computations on the same system.