India has around 28.52 billion tons of hematite and magnetite ore reserves, over 20% of which is in the form of fines. Also during the mining, crushing and washing of ore, huge quantities of fines are generated, majority of which end up into tailing ponds owing to their high alumina contents (similar to 10-15%). Selective dispersion flocculation is a promising route for beneficiation of such fine ores as opposed to conventional gravity and magnetic separation techniques; however, one requires selective reagents for effective beneficiation. We are engaged in the design and development of selective reagents for beneficiation of iron ores based on a molecular modelling approach. In the present study we have employed density functional theory (DFT) to calculate the interaction energy of hematite (0 0 0 1), goethite (1 0 0), gibbsite (0 0 1) and kaolinite (0 0 1) surfaces with carboxymethyl cellulose (CMC). The magnitudes of the computed interaction energies follow the trend: Hematite > Goethite - Gibbsite - Kaolinite Al-OH terminated > Kaollnite Si-O terminated. Electronic structure analysis revealed that the O-atoms in the hydroxyl and carboxyl groups of CMC form coordinate covalent bonds with surface Fe-atoms of hematite while forming only weak hydrogen bonds with goethite, gibbsite and kaolinite surfaces thus explaining the trends in the interaction energies. These DFT findings were validated by experimental flocculation results which show that a final concentrate of 64.4% Fe, 4.2% Al2O3, and 1.9% Silica with a yield of 56% can be obtained from a feed containing 56.5% Fe, 7.0% Alumina, and 4.9% Silica. Mineral wise estimation of chemical assay shows that CMC can be selective in flocculating hematite mineral particles over goethite, gibbsite and kaolinite. Our work thus highlights the important role of molecular modeling techniques such as DFT in screening and design of selective reagents for beneficiation of alumina-rich iron ore slimes.