Characterization of charcoals for helium cryopumping in fusion devices B. W. Sedgley and A. G. Tobin Gmmmun Corporation, Berhpage. Neuv York 11764 T. H. Batzer and W. R. Call University of California, Lau~rettce Livermore firionul Luboruto~y, Livermore, Ca'alijbrniu 94550 (Received 25 September 1986; accepted 3 November 1986) The capability of charcoal as a sorbent for helium at cryogenic temperatures depends upon charcoal characteristics that are not well understood. Previous work by the authors has indicated that the charcoals' pumping capability for helium depends as much on their source as on their particle size distributions. To develop a correlation between the physical characteristics of charcoal and helium pumping performance, different charcoals based on wood, coal, coconut, and a petroleum by-product were obtained from commercial sources. They were bonded to an aluminum substrate, and cooled to liquid-helium temperatures in a vacuum chamber. The helium pumping speed at constant throughput versus quantity of helium absorbed was measured for each charcoal grade. Porosimetry measurements on each charcoal grade using nitrogen as the sorbent gas were made that included total surface area, adsorption and desorption isotherms, and pore area and pore voiume distributions. Significant differences in helium pumping performance and in pore size distribution were observed. Comparisons are made between helium pumping performance and charcoal characteristics and a possible correlation is identified. I. INTRODUCTION A study'.2 performed since I983 to develop cryopumps for fusion application has shown significant improvements in helium vacuum pumping speed using cryogenically cooled, activated charcoals compared to other methods. The results showed in a limited test series that performance of cocorlut charcoal is superior to that of coal charcoal; that a coconut (named PCB from Calgon, Pittsburgh, PA) cbarcoal of ap- proximately 1.7-0.6 mm ( 12 X 30 U.S. standard mesh) size, which was in the midrange of all sizes tested, produced the highest performance; and that performance was strongly in- fluenced by the bonding methods and materials used to at- tach the charcoal to the cryogenically cooled substrate. A charcoalhraze bond and a charcoal/inorganic (copper) ce- ment bond yieided helium pumping performance signifi- cantly greater than that of a conventional charcoal/epoxy bond combination. It was appropriate therefore, based on &he success of this program, to identify those charcoal characteristics which would account for its potential for cryopumping helium. The value of charcoal as a sorbent comes from its extensive internal surface in the form of a network of fine intercon- nected pores. These surfaces typically measure hundreds of square meters per gram. The test program showed, however, that the charcoal with the largest total overall pore area per gram of charcoal was not that which yielded the highest helium pumping speed or adsorption capacity for helium. In this study, six charcoals derived from coconut, coal, and other sources were evaluated, and the following infos- mation was obtained for each material: helium-pumping speed and capacity, helium and nitrogen isotherms (quanti- ty sorbed versus equilibrium pressure at constant sorbent temperature), cumulative pore area as a function of pore size, and cumulative pore volume as a function of pore size. !I. SURVEY OF SELECTED CHARCOALS Gas adsorption analyses were performed on the charcoals selected for the study to determine surface area, pore volume and pore size. and isotherms for nitrogen. Nitrogen is used as a standard gas in the analyzer used in the program (Digisorb 2600 by Micromeritics Instrument Corporation) and was used for convenience. The algorithm used in the Digisorb 2600 for determining pore volume as a function of pore size is based on an adaptation by Micromeritics of the BJH meth- od developed by Barrett, Soyner, and Hale~da.~ The rela- tionship for determining pore surface area is the BET equa- tion after Rrunauer, Emmett, and Telle~.~ Pore size and pore surface area are both related to relative pressure, which is the ratio of the pressure (P) on the char- coal for a particular volume of gas absorbed to the ultimate saturation pressure (Po) of the absorbate on the sample. The relationship for determining pore size versus volume and area is not valid for relative pressures below 0.05, i.e., for pore sizes below approximately 10-A diameter; therefore, data presented in terms of pore size are limited in this study to values greater than 14 A. This limit, however, does not aEect the validity of relative pressure data below 0.05. Table I shows the charcoal grades selected for the survey and includes total surface area (N, BET) which is a com- mon!~ cited specification for sorbents. A11 charcoal samples were obtained as irregular particles, with the exception of the grade 965 which was a pelletized granule. This petroleum by-product from acid coke sludge is no longer available. Par- ticle sizes were selected to match as closely as possible the 12X 30 U.S. standard mesh size (particles from 1.74.6 mm) of the PCB coconut charcoal which was used earlier in helium pumping tests. ' Adsorption and desorption data for nitrogen on liquid- nitrogen-cooled charcoal were obtained. Figure 1 shows a 2572 d. Vac. Sci. Technol. A 5 (41, JuISAug 1987 0934-2101/87/042552-05501.00 @ 1987 American Vacuum Society 25722576 Sedgley etal.: Characterization of charcoals for helium cryopurnping in fusion devices 2576 SAMPLE: BPL HELIUM ACCUM.: 1.64 TORR * LITERICM' MINUTES AFTER STQPPING FLOW FIG. 10. Pressure in appendage chamber after cessation of helium flaw. PCB demonstrated the highest capacity (greater than 3 Torr l/cm2) for helium in the vacuum pumping tests com- pared to 0.6-0.7 Torr B/cm"or WV-B and GAC 102 GA, and 1.3-1.7 Torr I/cm2 for the remaining three chsrcoals. The data shown in Fig. 4 is the basis for the pore area and pore volume versus pore size profiles shown in Figs. 2 and 3. Although the relationships below 14-lf pore size are not de- terminable, the data at low relative pressures give an indica- tion of the distributions of small pores in the charcoal sam- ples. For example, the relative pressure ratio at 100 cm3/g nitrogen adsorbed at STP (constant cumulative volume) is 0.0017% for PCB; 0.00384.0048% for BPL, grade 945, and GAC 1240; and 0.0064-0.0078% for