FINAL TECHNICAL REPORT

September 1, 1995, through August 31, 1996

Project Title: NOVEL CARBONS FROM ILLINOIS COAL FOR NATURAL GAS STORAGE

ICCI Project Number: 95-1/4.2A-4

Principal Investigator: Massoud Rostam-Abadi, ISGS

Other Investigators: Jian Sun and Mark Rood, University of Illinois, Anthony A. Lizzio, ISGS

Project Manager: Daniel Banerjee, ICCI

ABSTRACT

The goal of this project was to produce low-cost adsorbent carbons from Illinois coal and evaluate the suitability of these materials for natural gas storage.

Granular (-20+100 mesh) and pellet (0.25"x0.125-0.25") activated carbons were produced by physical activation with steam and CO₂ and chemical activation with KOH, H₃PO₄, and ZnCl₂ from an Illinois bituminous coal (IBC-106). The products were characterized for their BET surface area, micropore volume, bulk density, and gravimetric/volumetric methane adsorption/storage capacities. Vm/Vs (volume of stored methane at STP per volume of storage container ) of some of the granular carbons produced by steam activation were about 70 cm³/cm³ which is comparable to that of BPL, a commercial activated carbon. Vm/Vs was found to be related to the pore surface area and micropore volume on the unit volume basis (per unit volume). Activated carbons produced by carbon dioxide activation had lower nitrogen surface area and micropore volume than those of the steam-activated carbons. Chemical activation was not an effective method for producing activated carbon from IBC-106 for natural gas storage. Densification to increase the carbon's bulk density was also evaluated. Activated carbon prepared by steam activation of pellets made from oxidized IBC-106 had a bulk density 50% higher than that of granular steam activated carbon and Vm/Vs methane storage capacity of 83 cm³/cm^3.

The production cost of the carbon pellets were estimated to be about $336/ton. These low-cost, highly microporous carbons could find potential application in adsorbed heat pumps, gas purification and separation, and energy storage systems. A market study is in progress to identify organizations, both in the U.S. and abroad, to test and evaluate the various uses and applications of the products.

Background

An adsorbed natural gas (ANG) storage system is being considered as an on-board storage technology for natural gas vehicles. An ANG storage vessel is less expensive, lighter, occupies less space, costs less to refuel, and has much greater storage density at low pressure (< 35 atm) than that of compressed natural gas storage. Carbon-based adsorbents have been found to have the most favorable gas storage density compared to other adsorbents, e.g., zeolites. A key concern in commercial development of an ANG storage technology is the availability of low cost carbon adsorbents (< $2/lb) and natural gas storage density of > 150 Vm/Vs (volume of stored methane at STP per volume of storage container) compared to current 100-120 Vm/Vs.

Goals and Objectives

The overall goal of this project was to develop a technology for producing low-cost adsorbent carbons from Illinois coal and evaluate the potential application of these materials for use in low pressure ANG vehicles.

Results and Discussion

Carbon Production - The adsorbent carbons were produced using IBC-106 coal from the Illinois Basin Sample Program. The parent coal was ground and sieved from -8 mesh to 20+100 and -100 mesh. The Free Swelling Index of IBC-106 coal is 4.5.

Activated carbons were prepared by physical (from the -20+100 mesh coal) and chemical activation (from the -100 mesh coal) methods in a bench-scale tubular reactor. In the physical activation method, a three-step process was applied: coal oxidation in air at 225C, devolatilization of oxidized coal in nitrogen at 400-500C, and char activation in 50% steam, or 50% CO₂, balance nitrogen at 800-850C.

Chemical activation was performed with KOH, ZnCl₂, and H₃PO₄. In the KOH activation, coal was mixed with granular KOH (coal/KOH ratio 1:1) and ground into a gel-like solid. The mixture was activated at 800 C in nitrogen for 1.25 hours.

In the chemical activation with phosphoric acid and zinc chloride, coal was pre-oxidized in air at 225C for 2 hours, then the oxidized coal was devolatilized in nitrogen at 400C for 1 hour. One gram of the resultant char was mixed with 0.1 gram of chemical and the mixture was dried at room temperature. The impregnated char was steam activated at 825C.

Bulk density, pore volume, and surface area of activated carbons - The surface area and micropore volume of the steam activated carbon products increased with carbon burn-off. The maximum values obsorved for the surface area and micropore volume were 1056 m²/g and 0.41 cm ³/g, respectively. The bulk density of the steam activated carbon decreased with the extent of reaction (total weight loss during Preoxidation, devolatilization, and activation steps) as a result of porosity development. The surface area and micropore volume when calculated on the unit volume basis (surface area x bulk density) initially increased and reached a maximum at about 70% weight loss (490 m ²/cm ³ and 0.17 cm ³/cm ³, respectively), then dropped off due to the decrease in the carbon bulk density. Carbon micropore volume showed a similar trend. The surface area and micropore volume of the activated carbon made by carbon dioxide activation were 200 m²/g and 20%, respectively, lower than those of a steam activated carbon.

The BET surface area and micropore volume of the KOH activated carbon were 1478 m²/g and 0.62 cm³/g, respectively. These values are significantly higher than those of the steam activated carbons. However, the bulk density of the sample was 0.27 compared to 0.44 g/cm³ for that of the steam activated carbon. The activated carbons produced by phosphoric acid and zinc chloride activation methods had lower nitrogen surface area (500 to 950 m²/g) and micropore volume (0.22 to 0.38 cm³/g) than those of the steam-activated carbons.

Table 1 summarizes the physical properties of Illinois coal-derived granular activated carbon produced by various physical and chemical activations. A commercial activated carbon, BPL manufactured by Calgon Carbon Corp., was used as a reference adsorbent. BPL is a well characterized and widely used commercial product.

Table 1. Physical properties of select Illinois coal-derived carbons.

A: Granular carbon made from IBC-106 (-20+100 mesh)

B: made with KOH chemical activation

C: made with H₃PO₄ chemical activation

D: made with ZnCl₂ chemical activation

BPL: (Commercial GAC)

Methane storage capacity - Methane storage capacity was evaluated both experimentally and using a correlation based on the micropore volume and bulk density of the activated carbon. For the steam activated carbon, the Vm/Vs leveled off at about 65% coal weight loss, corresponding to surface area and micropore volume of 900 m²/g and 0.32 cm³/cm³, respectively. The highest Vm/Vs observed was 76 cm³/cm³ (see table 1). Vm/Vs was linearly proportional to the surface area and micropore volume on the unit volume basis (surface area x bulk density; micropore volume x bulk density).

KOH activated carbon had the largest surface area and micropore volume among the activated carbon sample made in this study. However, its Vm/Vs was lower than that of steam activated carbon. This is attributed to the lower bulk density of KOH carbon compared to that of the steam activated carbon (see table 1). KOH activation of Illinois coal could be an effective method for producing carbon adsorbent for gas storage provided that the carbon's bulk density can be increased. Vm/Vs values for ZnCl₂ and H₃PO₄ activated carbon were 61 and 67 cm³/cm³, respectively. The Vm/Vs methane adsorption capacities of carbons from Illinois coal are comparable to that of BPL.

Carbon Densification studies - Because a high methane storage capacity (Vm/Vs>150) is essential for ANG powered vehicles, the bulk density of the adsorbent becomes important. This is due to the space limitations in ANG tank. Pelletization is a commonly used technique to increase the bulk density of activated carbon. Pelletization greatly reduces the void spaces between granular carbon particles and could also destroy large pores (macropores) in the material through compression.

Carbon pellets were made from Calgon BPL, an Illinois coal-derived steam-activated carbon, and an Illinois coal-derived KOH-activated carbon. Densification increased the bulk density of the activated carbons over 100%. However, micropore volumes were reduced nearly 100%, resulting in significant losses in methane storage capacity. The elimination of micropore volume is most likely due to the large amount of binder (up to 50 wt%) used to make the pellets. It is believed that the binder can block access to the micropores by covering or filling the pores. Efforts to make pellets with < 40% binder were unsuccessful.

Carbon pellets were also made from IBC-102 and 106 coal and oxidized IBC-106 coal using a suitable binder. Coal pellets were pyrolyzed in nitrogen at 400C and steam-activated at 825 and 850C. Table 2 summarizes the physical properties of granular and pellet activated carbons. The carbon product had a substantial high bulk (random packing) density of 0.58 g/cm³ for IBC-102 derived carbon and 0.65 g/cm³ for IBC-106 derived carbon, compared with 0.44 g/cm³ for that of the granular products. The difference between the pellet bulk densities is due to their different geometry. Vm/Vs methane storage capacity of carbon pellets from IBC-102 was around 60 cm³/cm³, which is 15% lower than that of the granular carbon. The lower Vm/Vs is due to about 40% lower micropore volume of the pellet compared with that of the granular activated carbon. IBC-102 coal pellets were not oxidized prior carbon production. Although no particle agglomerating was observed during activation, it is possible that the inherent caking property of IBC-102 led to the deterioration of microporosity.

BET surface area and microporosity of the carbon pellets prepared from oxidized IBC-106 coal are comparable to those of granular carbons. Preoxidation of the coal pellets reduced the caking and preserved the original microporosity in these products. As a result, the Vm/Vs of the pellets was about 10% higher than that of the granular carbon. Pre-oxidation

of the coal pellets prior to activation is recommended to obtain a carbon product with microporosity comparable to a granular carbon.

Table 2. Comparison of physical properties of carbon pellets and granular carbons.

G1 and G2: Granular activated carbon made from IBC-106 (-20+100 mesh);

P1, P2 and P3: Pellet carbon made from IBC-102 pellet;

P4: Pellet carbon made from IBC-106 pellet;

P5: Pellet carbon made from oxidized IBC-106 coal (1.5h).

Cost study - Based on a previous ISGS economic study, the production cost for an 80,000 ton/year plant to produce granular steam activated carbon from an Illinois coal is about $326 per ton. This economic study was performed based on data obtained from a pilot-scale carbon production test at Svelada Industries, Oak Creek, WI, in 1994. Assuming an additional $10/ton for making coal pellets, the production cost of the activated carbon would be $336/ton, or 16.6 cents per pound. This cost is substantially less than the target cost of <$2/lb for a suitable adsorbent for natural gas storage. The methane storage capacity of the activated carbons produced in this study are about 50% of the target storage capacity. However, in certain applications where space limitation is not a critical issue, the low-cost, highly microporous Illinois coal- derived activated carbon would be preferable to the expensive high capacity carbon products.

Marketing study - Arrangement have been made to ship a selected number of activated carbon samples prepared from IBC-106 coal to Professor David Quinn of Royal Military College of Canada for further extensive application evaluations. A market study is in progress to assess the application of the Illinois coal-derived activated carbon for ANG system, adsorbed heat pumps, hydrogen storage, and other applications where a low-cost, highly microporous activated carbon is required. A number of companies and organizations involved in these R&D activities, both in the U.S. and abroad, have been contacted. The goal of the market study is to identify an industrial participant to test and evaluate various use and application of activated carbons developed in this research program.