"030
N. 1L LAINX,l?.J.
\-ASTOLA, AND
P. L. WALXER,JR.
\-o1. 67
THE INYOKTANCE OF ACTIVE SURFACE AREA I N TH33 CAKBOK-OXYGES REACTIOW2 BY N.R. LAINE,F. J. VASTOLA, AND P. L. WALKER,JR. Department of Fuel Il'echnology, The Pennsylvania State University, Universitg Park, Pennsylvania Received March 4, 1963 Graphon, a highly graphitixed vttrhon black, was oxidized to seven levels of burn-off between zero and 35?& The avtivc surfaw area of these oxidized samples was determined by measuring the coverage of the surfare a i t h oxygen complex upon exposure of the samples to 02a t 300" for 24 hr. After removal of the complex, the Graphon saniplbs were sobsequbntly treated with additions1 02. The amount of oxygen complex formed on the surfarc during reaction could be folloRed by a material balance and at the end of a ruh by outgassing. From these data, unoccupied active surface areas were calculated. Rate constants for 0 2 depletion and product formation, calculated on the basis of unoccupied active area, were essentially constant over most of the burn-off range studied. On the other hand, rate constants, calculated on the basis of B.E.T. surface area, showed a considerable increase with burn-off.
Introdliction In the vast literature on the carbon-oxygen reaction,3 there are literally as many different reactivities reported as there are carbon materials used. Further, the reactivity of a particular carbon material can vary with carbon b ~ r n - o f f . These ~ variations in reactivity have been ascribed to the fact that different carbons have, initially, different amounts of active surface areas (ASA) and that their ASA cliange to different extents with burn-off. The extent of AS-4 is thought to be a function of such properties of the carbon as cryst.allite size, crystallite orientation, vacancy concentration in the basal plane, and impurity concentration, type, and location. Until recently, tlierc was little success achieved a t measuring ASA of carbon. Graharr~,~ however, showed that the ASA of graphitized carbon blacks could be estimated from the lorn-coverage end of iV2 isotherms obtained a t 78'K. The graphitized carbon blacks which he studied were found to have relatively small fractions of their total surface area (TSA) as ASA. This is attributed to the fact that the surface of grstphitiaed carbon blacks is composed almost entirely of the basal planes of carbon crystallites.6,6 These basal pIanes are not thought to be the source of strong sites, other than where they contain vacancies. Graham concluded that the strong site area in graphitized carboh blacks is primarily accounted for by the intersection of the basal plane sbrfaces coinposihg the polyhedral-shapcd particles. Recently, the authors' have shown that the ASA of a graphitized carbon black can also be measured by exposure of the black to O2 a t 300' for a prolonged period. At this low temperature, the 0 2 interacts with the carbon surface, forming carbon-oxygen complex on the strong sites and a negligible amount of gaseous CO and COS. Upon heating the carbon to 950°, (1) Bnscd u n a 1'11.1). thesis subniitted by N. K. Laine to tlie Graduate School of T h e l'ennsylvania State University, June, 1962. (2) This work was supgorted by the National Science Foundation Grant (X3023 and tlia A t o m i c Energy Commission on Contract No. AT(30-1)1710. (3) P. L. Walker, Jr., I?. Rusinko, dr., and I,. G. Austin. Aduan. Calalysis, 11, 133 (1959). ( 4 ) D. Graham, J . Ph2/a. Chem., 61, 1310 (1957). ( 5 ) 11. Akamatri and 11. Krlroda, "Procerdings of the Fourth Carbon Conference," I'crguinon Press, New Tork, N. Y.,1961, pp. 3R5-369. ( G ) 12. A . ICmrtko, "Proceedinss of tho First nnd Second Carbon Conferences," U. of I3uNalo. 1956, pp. 21-30. (7) N. 11. Laine, 1:. .J. Vnstoln, and 1'. I,. Walker, .Jr., "t'roccedizigs of tlii: I'ifth Carbon Conf?rrnrr.," Vol. 11, ~'i*rb?:ifllonI'rwn, h'rw Ywk, x. I,., 1963,i ) p . 21 1--21i.
the cohplex is recovered, with its volhme related to tlic ASA of the carbon. I n this paper, the ambitious undertaking of correlating the structural properties of different carboiis t o their ASA will not be considered. Rather, the chaiigt: in ASA with burn-off for one carbon will be measured, and these ASA will be used to calculate meaningful rate constants for the carbon-oxygen reaction (for this particular carbon). Experimental The carbon used in this investigation ivus (;raplion,' rvhicli 1%2s produced by heat treatment of the channel black, Spheron-ti, In the absence of oxidizing gases to 2800". The original Graphon was oxidized to seven levels of burn-off between zero and 35% a t temperature of 625' and an initial 0 2 pressure of 500 p , in order to obtain samples with different ASA. The sample bed depths used in these initial treatments were sufficiently sliallow- to assiir(' :t uniform rate of oxidation through the bed.3 These samples were first heated a t 050", in vacuo, for 3 hr. to remove essentially all of the surface complex which was present as a result of their prior oxidation. The TSA of these samples were measured by the B.E.T. method using S2 adsorption a t 78°K. The total ASA of these samples were determined from the amount of surface oxygen complex formed in 24 hr. a t 300" using an initial 0 2 pressure of 500 p. The complex formed was recovered at 950" and converted to equivalent oxygen concentration. The assumption was made that the complex consists of one oxygen atom prr edge carbon atom. Further, i t was assumcd that the edge carbon a t o m lie in th: (100) plane; that is, each carbon atom occupies an area of 8.3 A2 Reaction rate runs were then made a t temperatures of 575, 628, and 675" and an initial 0 2 pressure of 39+ 0.5 p , employing the reaction rate apparatus previously described.* At the bed depths used, the rates were not controlled by the rate of oxygen difueion through the bed. A CI> kh, upon some activation of the 8.5 ... 4.9 ... Graphon to produce a significant amount of ASA, the 14.4 4.0 4.8 6.7 over-all rate can be approximated closely by the first 20.8 ... 5.1 ... term in the above expression. Such appears to be the 25.8 4.2 5.0 6.5 case for the Graphon samples having burn-offs above 34.9 ... 5.1 ... 3.3%, as is seen in Table 111. An estimate of kb can then be made from the rate data for the original Discussion Graphon sample by assuming that ke has the same It appears from adsorption studies with Nz a t 78OK. value as that for the larger burn-offs. This calculathat the original sample of Graphon used in this study tion, for the three reaction temperatures, yields a had a somewhat more homogeneous surface thah that value for k e / k b of ea. 2000. used by Graham in his early studies.* First, from low The answer to the question of why edge carbon atoms coverage adsorption studies, using the intercept method are more reactive to oxygen than are basal plane cardeveloped by Graham, our Graphon sample had the bon atoms would appear to lie with two possible factors smaller ASA per unit of TSA.g Second, the B.E.T. area -a geometric factor and/or an impurity factor. Reof our Graphon was 76 m.2/g., compared to 82 m.z/g. garding the geometric factor, the important point to for Graham’s sample. These results are consistent with be considered is the relative likelihood of edge us. the fact that our Gra,phon was heat treated to a higher basal plane carbon atoms being able to form bonds with temperature (2800’) than was Graham’s sample (2700’). chemisorbed oxygen. The formation of such bonds is Polley, et ccZ.,l8 have shown that the surface area of thought to be a required step in the conversion of oxygen Spheron 6 carbon black decreases with increasing heat to gaseous carbon oxides. From recent studies on the treatment temperature. Beebe and YounglQ have change in thermoelectric power of graphite upon the shown, from heats of adsorption studies, that the surchemisorption of oxygeqZ6it appears that a conducface homogeneity of carbon black increases with intion electron (r-electron) participates with an in-plane creasing heat treatment temperature. a-electron of the carbon to form a carbon-oxygen Apparently, Smith and PolleyZ0were the first to double bond. Edge carbon atoms have unpaired urecognize the potential of using carbon blacks heated to electron^,^^^^* which are available to form bonds with various temperatures to study reactivity and activachemisorbed oxygen; basal plane carbon atoms pretion processes in the carbon-oxygen system. They sumably have their a-electrons tied up in chemical found that the reactivity of a thermal black to air bonds with adjacent carbon atoms. decreased sharply with increasing heat treatment temConsidering the impurity factor, it is known that perature. They attributed this t o the increase in certain impurities in small amounts can strongly crystallite size and the reorientation of crystallites catalyze the reaction of carbon with oxidizing g a s e ~ . ~ , ~ ~ upon heat treatment to yield a surface composed inFor example, the addition of 100 p.p.m. of iron to creasingly of the basal plane of the crystallites. Upon spectroscopic graphite can increase its rate of gasificaoxidation, they suggested that attack primarily occurs tion by COZ ea. 150-f0ld.~~It is estimated that the a t the intersection of the basal plane surfaces (the location of the strong site area according to Graham4 total ash content of the Graphon used in this study is
