The role of mesopores in MTBE removal with granular activated carbon

This activated carbon research appraised how pore size and empty-bed contact time influenced the removal of methyl tert-butyl ether (MTBE) at part-per-billion (ppb) concentrations when MTBE was the sole organic impurity. The study compared six granular activated carbons (GACs) from three parent sources; these GACs contained a range of pore volume distributions and had uniform slurry pHs of 9.7-10.4 (i.e. the carbons' bulk surface chemistries were basic). Several of these activated carbons had been specifically tailored for enhanced sorption of trace organic compounds. In these tests, MTBE was spiked into deionized-distilled water (similar to pH 7); MTBE loading was measured by isotherms and by rapid small-scale column tests (RSSCTs) that simulated full-scale empty-bed contact times of 7, 14, and 28 min. The results showed that both ultra-fine micropores and small-diameter mesopores were important for MTBE adsorption. Specifically, full MTBE loading during RSSCTs bore a strong correlation (R-2 = 0.94) to the product (mL/g x mL/g) of pore volume <= 4.06 angstrom wide and pore volume between similar to 22 angstrom and similar to 59 angstrom wide. This correlation was greater than for the product of any other pore volume combinations. Also, this product exhibited a stronger correlation than for just one or the other of these two pore ranges. This multiplicative relationship implied that both of these pore sizes were important for the optimum GAC performance of these six carbons (i.e. favorable mass transfer coupled with favorable sorption). The authors also compared MTBE mass loading during RSSCTs (mu g MTBE/g GAC) to isotherm capacity (mu g MTBE/g GAC). This RSSCT loading "efficiency" ranged from 28% to 96% for the six GACs; this efficiency correlated most strongly to pores that were 14-200 angstrom wide (R-2 = 0.94). This correlation indicated that only those carbons with a sufficient volume of 14-200 angstrom pores could adsorb MTBE to the extent that would be predicted from isotherm data. (C) 2014 Elsevier Ltd. All rights reserved.