Morphological characterization of microporous carbon materials

Abstract

The development of microporous carbon materials has attracted a great deal of attention in recent years. The potential applications of these materials include catalyst support, adsorbents, molecular sieves, porous electrodes and other battery components [1–4]. In addition, these microporous carbon materials have been used as preforms for infiltration with molten silicon or silicon alloys to produce silicon carbide based advanced ceramics for a variety of high temperature applications [5–10]. The melt infiltration kinetics and mechanism of reactions between liquid infiltrants and the porous carbon depend on a number of factors including the preform permeability, which is quite sensitive to small changes in pore structure. The permeability of the porous preform and fluid flow of the infiltrant changes due to volume change accompanied by the silicon-carbon reaction during the infiltration process, with the potential for choking-off and other undesirable results. In order to produce silicon carbide based ceramics with homogeneous second phase distribution and to maintain good control of the melt infiltration process, it is necessary to fabricate microporous carbon preforms with controlled morphology and pore size distributions. The objective of this paper is to study the pore size distribution, pore diameter and other microstructural parameters of three types of microporous carbon materials. In addition, evaluation of the phase separation and the effect of molybdenum on the local graphitization in the microporous carbon materials are presented. A mixture of furfuryl alcohol resin, diethylene and triethylene glycol liquid pore-forming agents, and ptoluene sulfonic acid catalyst were polymerized to form a soild polymer. The solid polymer was heated slowly to 1000 8C to pyrolyse the polymer forming a microporous carbon preform. By varying the ratios of the constituents in the polymer system, a wide range of pore volumes, pore sizes, and morphologies of the microporous carbon materials were obtained [7–10]. Microstructural characterization of the preforms was carried out on a Jeol-JSM 35 scanning electron microscope. The physical density measurements were carried out by measuring the dimensions and weight of the sample. Samples were heated in an oven for 6 –8 h at 160 8C before each measurement. This treatment removes any moisture from the material. The pore size distribution in the preforms was determined by mercury porosimetry using a Micromeritics Auto-Pore 9200 porosimeter. Two porous carbon preforms were selected for study using transmission electron microscopy. Carbon pieces ,3 mm maximum dimension were ground using 600 grit SiC paper and polished with 3 im diamond mylar to produce a flat disc. The carbon was mounted on a copper grid and then ion milled at 5 kV to perforation. Samples were examined at 120 kV using a Philips 400T microscope equipped with a Kevex UTW X-ray spectrometer. Three types of microporous carbon materials were used in this study. For simplicity, they are denoted as A, B, and C in the text and have density of r ˆ 0.71, 0.86, and 0.91 g=cm3, respectively. A scanning electron micrograph of fracture surface of specimen A (Fig. 1a) shows a homogeneous distribution of pores and carbon struts. The average diameter of pores in this material is about ,0.13 im. The microstructure of specimen B (Fig. 1b) has an average pore diameter of ,1.3 im. The fracture surface of specimen C (Fig. 1c) shows an average pore diameter of ,2 im and the carbon struts themselves contain submicron pores. Mercury porosimetry results for specimens A, B, and C are shown in Fig. 2. Specimen A shows no measurable intrusion volume at sizes above 0.19 im. All the pore volume is available between 0.003 and 0.19 im. The physical density of this material is 0.71 g=cm3, while the density obtained by mercury porosimetry is 0.72 g=cm3. Its pore surface area is 40.1 m2=g. In specimen B, the amount of pores with pore diameters higher than 1.75 im is about 3%. The steep rise in the curve indicates a very narrow pore size distribution for these materials. The measured density of specimen B is 0.86 g=cm3 while the density obtained by mercury intrusion is 0.87 g=cm3. The pore surface area of this material is 39.95 m2=g. It is interesting to note that there is no measurable micropore volume in samples A and B. The pore volume deduced from the mercury porosimetry data is almost equal to that obtained by physical density measurements. The density data obtained by mercury porosimetry and physical measurements, respectively are 0.72 g=cm3 and 0.71 g=cm3 for sample A and 0.87 g=cm3 and 0.86 g=cm3 for sample B. Their skeletal densities are also about 1.5 g=cm3. On the other hand, in sample C there is a slight difference in