FINAL TECHNICAL REPORT

September 1, 1997, through February 28, 1999

Project Title: CARBON-BASED SORBENTS FOR CLEANUP OF COAL-DERIVED GASES IN IGCC SYSTEMS

ICCI Project Number: 97-1/5.2A-5

Principal Investigator:  Anthony A. Lizzio, ISGS

Other Investigators: Mark A. Kelly, ISGS

Project Manager:  Ronald H. Carty, ICCI

ABSTRACT

Low cost processes for removal of hydrogen sulfide (H₂S) from hot coal gas are needed to accelerate the successful demonstration and commercialization of advanced coal gasification systems worldwide.  A review of the literature has shown that every sorbent examined to date for hot gas cleanup has had at least one major deficiency that prevents its widespread use.  Because these sorbents are only beginning to be used commercially, there is still ample time for the development of improved sorbents and systems for hot gas cleanup.

The goal of this study was to determine the commercial potential of Illinois coal-based sorbents in hot gas cleanup processes.  Carbon-based sorbents were produced from Illinois coal that adsorbed as much as 20% sulfur at 550^oC.  Fixed-bed adsorption tests performed on activated chars indicated that zinc-loaded chars prepared from IBC-102 coal has H₂S adsorption capacities comparable to those of some commercial metal-based sorbents.  Insight was also gained into the mechanism of H₂S removal by activated char.  The presence of CO₂ in the gas stream enhanced removal, while H₂ and CO inhibited H₂S removal by activated char.  Also, over the temperature rance of 400-600^oC, there was little or no change in the H₂S adsorption capacity of activated char unlike some metal-based sorbents which tend to show a substantial decrease in H₂S removal performance at 400^oC.

Char optimization studies were also performed.  Activated chars were prepared by zinc chloride (ZnCl₂) activation in a one-step process that increased surface area as well as added zinc to the char.  Carbon-based sorbents having surface areas up to 1300 m²/g and up to 10% Zn were produced.  Our best ZnCl₂ activated char, although it adsorbed less H₂S than zinc titanate, adsorbed significantly more H₂S under simulated hot coal gas conditions than any other commercial activated carbon tested.  Several methods to regenerate sulfided chars were examined, including heat treatment in H₂, steam, N₂, or 1-21% O₂, and the use of aqueous solutions of H₂O₂, HNO₃, and KOH.  Regeneration with H₂ at 800^oC or 1-3% O₂ at 300-500^oC worked best.  The carbon-sulfur bond formed during sulfidation is comparable in strength to that of organic sulfur in coal, which is very stable unless a high temperature treatment in H₂ is applied.  This limits the usefulness of carbon sorbents in hot gas cleanup processes.  Nonetheless, activated char could serve well in removing residual H₂S and possibly other air toxics such as mercury, selenium and arsenic from hot coal gas.  Conceivably, the next generation hot cleanup process for IGCC processes could include activated char working in tandem with zinc-based sorbents to remove H₂S as well as air toxics from hot coal gas.

Pages 5 to 27contain proprietary information.

EXECUTIVE SUMMARY

Integrated Gasification Combines Cycle (IGCC) processes are emerging as the most promising technology to convert high sulfur coal into electricity.  In these processes, over 99% of the sulfur and particulates need to be removed from the coal gas before it enters the turbine.  To achieve maximum operating efficiency, hydrogen sulfide should be removed from the fuel gas while hot.  Although H₂S can be removed quite effectively by cooling the hot gases to room temperature, removal of H₂S at 500-800^oC could lead to a 3% increase in overall process efficiency.  Since Illinois coal is a proven feedstock in IGCC processes, e.g., the Destec process, more efficient IGCC processes should ultimately expand markets for high sulfur Illinois coal.

Numerous metals and mixed metal compounds have been studied as possible desulfurization sorbents.  Current leading sorbents include zinc titanate and Z-Sorb (a proprietary zinc-oxide-based sorbent).  However, not only are these sorbents expensive (up to $7-15/lb), they are also prone to chemical and/or physical degradation during cycling.  Zinc titanate suffers from spalling due to formation of sulfide and sulfate which have 2 to 3 times higher molar volume than oxide.  The formation of sulfate also prevents further adsorption of H₂S on the stable zinc sulfate and thus lowers H₂S adsorption capacity of the sorbent.  Z-Sorb degrades in the presence of steam present in coal gas.  Both sorbents sinter during regeneration and typically their reactivity can drop by up to 50 percent in just 50 cycles.

One may conclude by reviewing the literature on hot gas cleanup sorbents that while research on sorbent materials has been extensive, continuing efforts are important because these sorbents are not yet used commercially in coal gasification processes, and there is still time for the development of improved sorbents.  Another reason to encourage research in this area is that not all gasification systems are alike, and because of these differences in operating conditions and requirements, it is likely that more than one type of sorbent will be needed to satisfy the market.  Every sorbent examined to date has had at least one major deficiency that prevents its widespread use.  The fact that there is still ample opportunity to develop new types of sorbents for hot gas cleanup provides incentive for research on new types of materials.

One material that should be examined in more detail is carbon.  At first glance, carbon may appear unfit for coal gas environments, however, it is well known that carbon will not react (gasify) to any appreciable extent with the gases found in a reducing atmosphere (CO, H₂O, H₂) at temperatures less than 800^oC.  A well documented advantage of carbon sorbents, e.g., carbon molecular sieves (CMS), over other materials used in gas separation, e.g., zeolites, is its exceptional resistance to chemical attack in harsh process environments.  Carbon molecular sieves in the form of pellets can withstand attrition for long periods of time, and often least for more than ten years in air separation plants.  In a pressure swing adsorption process, the CMS is exposed to severe changes in pressure as well as temperature, and with typical cycle times of two minutes or less, thousands of adsorption/desorption cycles must be performed before the sorbent is replaced.  Another key advantage of carbon over metal-based sorbents is that carbon, itself, adsorbs H₂S, meaning that carbon could be used as an active support for metals such as zinc and copper, which also adsorb H₂S.  Most metal-based sorbents have an inert support matrix, sometimes constituting up to 60% of the mass of the sorbent.  Metal-based sorbents, such as zinc titanate and Phillips Z-Sorb, have an inert support matrix that does not actively participate in the H₂S adsorption process.

The goal of this research was to prepare a regenerable carbon-based sorbent from Illinois coal suitable for use in hot gas cleanup, i.e., removal of H₂S from hot coal gas to concentrations less than 20 ppm at temperatures between 400 and 600^oC.  The project consisted of six tasks.  In Task 1, activated chars and metal-impregnated chars were produced from both size-graded and pelletized Illinois coal.  The pore structure and surface chemistry of the carbon sorbents were tailored for H₂S removal.  Various metals such as zinc and copper that are known to be good H₂S adsorbers were added to activated char by impregnation to incipient wetness or ion exchange.  In Task 2, the physical/chemical properties of the carbon sorbents were determined to gain further insight into their H₂S removal capabilities.  Surface areas, pore size distribution, oxygen and sulfur contents, bulk density and attrition resistance were determined.  In Task 3, the H₂S removal capabilities of activated and metal promoted chars were determined using a fixed-bed reactor coupled to a quadrupole mass spectrometer.  A simulated coal gas mixture containing 0.5% H₂S and representative concentrations of CO₂, N₂, H₂O, CO and H₂ was used.  Breakthrough curves were obtained at a space velocity of 2000 h^-1.  In Task 4, sulfided chars were regenerated by a variety of methods including reaction with air and hydrogen, the goal was to fully restore the H₂S adsorption capacity of activated char while minimizing carbon consumption through gasification or attrition.  If suitable H₂S removal performance in the second and third adsorption cycles was observed, fluidized tests involving higher flow rates and smaller particle sizes would be performed with that carbon.  In Task 5, the H₂S removal capabilities of optimized chars were compared to those of other sorbents being considered for hot gas cleanup.  If a suitable char was identified, the technical and economic feasibility of using it on a commercial scale would be evaluated.  In Task 6, monthly, mid-year and final technical and management reports were prepared and submitted to the ICCI.

Results from the first year project (1996-1997) showed that carbon-based sorbents prepared from Illinois coal adsorbed up to 20% sulfur from a gas stream containing 0.5% H₂S in CO₂ and N₂ at 550^oC.  Fixed bed breakthrough curves were obtained for several chars using a simulated coal gas stream containing 0.5% H₂S at a space velocity of 2000 hr^-1.  Breakthrough times (200 ppm H₂S) ranged from 2 to 6 h, depending on the char used.  These breakthrough times were comparable to those of some metal-based sorbents being considered for hot gas cleanup.  Results of these H₂S adsorption tests indicated that carbon had a potential as a hot gas cleanup sorbent.  However, further improvements in both sulfidation capacity and regenerability were still needed.

The effect of reaction temperature on H₂S removal by activated char was also studied.  Over a temperature range of 400-600^oC there was little or no change in the H₂S adsorption capacity of one activated char.  The ability of this activated char to perform just as well at lower temperatures, i.e., 400^oC, indicated another possible advantage over metal-based sorbents such as zinc titanate, which show a substantial decrease in performance at lower temperatures.  Further insight into the mechanism of H₂S removal by activated char was also gained by determining the effects of CO₂, H₂, CO and H₂O in the simulated coal gas on H₂S removal performance.  It was found that CO₂ enhances, while H₂ and CO inhibit H₂S removal by activated char.  A mechanism for H₂S removal by activated char was proposed that includes the contributions of carbon active sites (C), carbon-oxygen (C-O) complexes and metal (M) atoms such as zinc or copper to overall H2S adsorption capacity.

C + H₂S ------> C-S + H₂

C-O + H₂S -----> C-S + H₂O

C-M + H₂S -----> C-M-S + H₂

Continuing efforts to optimize the H₂S adsorption capacity of activated char led to activated chars having surface areas and zinc contents greater than those made previously in this study.  The zinc retained int he sample was controlled by the extent of water washing and by the temperature at which the coal/ZnCl₂ mixture was activated.  At higher temperatures (>600^oC), the zinc becomes more volatile and less is retained in the char.  Activated IBC-102 chars having up to 10 mole % Zn were produced.  A steam activation treatment of ZnCl₂ activated IBC-102 char increased surface area from 500 to 1300 m²/g.  A large batch of Zn exchanged carbon was also prepared from a nitric acid treated commercial carbon.  These carbons were tested in a 1/2 in. ID quartz tube reactor operated in a fixed-bed mode under the following conditions:  5 g in. char, 538^oC, 1 atm, space velocity of 2000 h^-1, 0.562% H₂S., 11.7% H₂, 12.5% CO, as.5% CO₂, 3.7% H₂O, balance N₂.  Our best IBC-102 char adsorbed one fourth the amount of H₂S as zinc titanate and one half of that H2S capacity was restored by regeneration with H₂ at 800^oC, whereas zinc titanate could be fully regenerated in air at 700^oC.  A Zn exchanged Calgon F400 carbon (derived from steam activated bituminous coal) had an H2S adsorption capacity in the first adsorption cycle only 25% less than that of zinc titanate, but this same carbon retained less than 5% of its original H₂S capacity in subsequent cycles after regeneration in H₂ at 800^oC.  A regeneration method involving the use of a 1-3% O₂, 50% CO₂, balance N₂ applied at 427^oC was used to restore more than half of the original H₂S adsorption capacity of the Zn exchanged Calgon F400 carbon, a significant improvement over H₂ regeneration.  The CO₂ was added to the regeneration gas to inhibit carbon loss due to combustion.

It remains to be determined how the pore structure and surface chemistry of the char can be further adjusted to optimize H₂S adsorption capacity and regenerability.  The mechanism of H₂S removal by carbon also needs further clarification.  We have proposed an overall mechanism based on experimental results suggesting that carbon active sites, chemisorbed oxygen and Zn metal all play a role in H₂S removal by activated char.  Scale up work could involve production of pound quantities of ZnCl₂ activated char in a continuous rotary tube kiln.  Before scale up we would want to be able to further develop a suitable char regeneration process since all indications are that high temperature H₂ treatment is not a practical means to regenerate activated char.  Regeneration of sulfided char with low partial pressure of oxygen showed more promise, but further work is needed to inhibit carbon combustion by oxygen.  Another unique aspect of this work is the concept of producing a dual purpose sorbent optimized to remove H₂S from hot coal gas and then mercury from coal combustion flue gas.  Results from the literature suggest that sulfur-laden chars produced under hot gas cleanup conditions may be ideal for mercury capture.  Such a dual purpose sorbent might be a more economical alternative to regeneration.

The remainder of this report contains proprietary information and is not available for distribution except to the sponsor(s) of this project.