1100

'Measured by the N2 BET method.

"Calgon Carbon Corp., P.O. Box 717, Pittsburgh, PA 15230. . "Barnebey & Sutcliffe Corp., P.O. Box 2526, Columbus, OH 43216. dCarbon Activated, 1662 W. 139th Street, Gardena, CA 90249. 'Westvaco Corp., Carbon Department, Coventon, VA 24428. fRohn and Haas Co., 5000 Richmond Street, Philadelphia, PA 19137.

The adsorption capacity of an adsórbate on activated carbon is related to the adsorbate's concentration in solution. Equilibrium adsorption isotherm tests were conducted to determine the relationship between the concentration of sulfur in methanol and the activated carbon dosage. These tests used the coal-based Calgon Filtrasorb 400 AC and three coconut-based carbons: Calgon Type PC and B&S Types SE and MI. As expected, the sulfur concentration in the AC-treated methanol decreased with increasing dosage. At 1-g carbon per 100-mL methanol, AC removed approximately 45% of the benzothiophenes and approximately 97% of the dibenzothiophenes. This value increased to 97-99% for benzothiophenes and essentially 100% for dibenzothiophenes at a dosage of 4-g AC/100-mL methanol.

The data obtained from the equilibrium adsorption isotherm tests were interpreted by using the empirical Freundlich equation (1):

where X = amount of sulfur impurity adsorbed, g m = mass of activated carbon, g

C = sulfur concentration in methanol after equilibrium, mg/L

All four adsorption isotherms (Fig. 1) in the plots of log (X/m) versus log C show two linear regions. Adsorption characteristics of a solution containing two solutes measured by a gross concentration parameter are reflected in an isotherm plot consisting of two straight lines (2). The adsorption isotherms in Fig. 1 suggest that the benzothiophene group in methanol has different adsorptivity on activated carbon from the dibenzothiphene group, and the two types of sulfur species compete for adsorption sites in the pore structure of the activated carbon.

Figure 1 also shows that the coal-based Filtrasorb 400 AC has higher sulfur adsorption capacities than the coconut-based ACs for methanol solutions that have sulfur concentrations greater than about 4.0 mg/L (5 ppmW). However, the reverse is true for sulfur concentrations less than 4.0 mg/L. Among the. three coconut-based ACs tested, B&S Type MI has the highest sulfur adsorption capacity.

A preliminary assessment of the dynamic performance of these ACs was obtained in a bench-scale, dynamic batch test. In this test, Calgon Filtrasorb 400 and B&S Type ML activated carbons were packed in a Pyrex glass column of I.77-cm ID x 91.44-cm length (0.5-in. ID x 36-in. length). The carbon has a particle size of 1.64 x 0.84 mm (-12 +20 U.S. Standard Testing Sieve No.) and was tested with diesel-fuel-contaminated methanol (containing 30 ppmW sulfur) at a flow rate of 5 mL/min. The empty bed contact time was 24 min. A 5-mL effluent sample was collected at the end of the column for every time interval tested. The test with Filtrasorb 400 lasted 16 hours, and the test with Type MI lasted 30 hours.

The sulfur breakthrough curves for both tests are shown in Fig. 2. Within experimental and analytical errors, both ACs behave similarly in terms of their breakthrough characteristics. During the first 3.0 h of continuous flow testing, they removed 100% of the 30 ppmW sulfur in the methanol feed. The breakthrough curve levels off at an effluent sulfur concentration of 24 ppmW, which is 80% of the sulfur concentration in the methanol feed. Apparently, the sulfur compounds in the feed had not totally broken through the column after the 30-h test (i.e., the column has not reached dynamic equilibrium with the feed). A GC/MS analysis of the effluent sample collected at 29 h showed approximately 89% of total benzothiophenes and approximately 29% of total dibenzothiophenes in the methanol feed.

The breakthrough data shown in Fig. 2 were mathematically analyzed by the LUB/equilibrium section concept (3). On the basis of this concept, the dynamic adsorbate loading (or adsorption capacity) of the activated carbons was estimated to be 0.79 mg S/g AC for the 80% adsorption capacity level. By extrapolation, the dynamic adsorption capacity of these two ACs would be 1.0 mg S/g AC if these activated carbons were tested until the complete breakthrough of sulfur from the column. Figure 1 indicates that the equilibrium isotherm adsorption capacity for Calgon Filtrasorb 400 and B&S Type MI activated carbons is 3.5 and 2.8 mg S/g AC, respectively. Therefore, the dynamic adsorption capacity of 1.0 mg S/g AC for these two activated carbons is about one-third of their equilibrium isotherm adsorption capacities.

>— CALGON FJLTBASOBB 400 CALGON TYPE PCB "ö—BiS TYPE SE -•CK-BiS TYPE MI

0 0

Post a comment