Environ. Sci. Technol. 2004, 38, 276-284
Enhancing Activated Carbon Adsorption of 2-Methylisoborneol: Methane and Steam Treatments K I R K O . N O W A C K , * ,† FRED S. CANNON, AND DAVID W. MAZYCK‡ Department of Civil and Environmental Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802
This research investigated methods for tailoring a commercial, lignite-based granular activated carbon (GAC) to enhance its adsorption of 2-methylisoborneol (MIB) from natural water. Tailoring efforts focused on heat treatments in gas environments comprising steam and/or methane, since these gases can alter GAC pore structure and surface chemistry. Heat treatments that combined methane and steam enhanced MIB adsorption considerably, causing a 4-fold improvement (over untreated GAC) in fixedbed adsorption performance relative to initial MIB breakthrough. These favorable effects, observed in rapid small-scale column tests, occurred following simultaneous and separate (sequential) applications of methane and steam. Moderately low temperature steam treatments also improved MIB uptake in fixed-bed adsorption tests but to a lesser extent (approximately 1.5-fold). In contrast, methane treatments alone, at various temperatures, led to significant carbon deposition within the GAC pore structure. As a result, total pore volume was reduced and MIB adsorption performance declined.
Introduction Taste-and-odor control has long been an important aspect of drinking water treatment. Although many naturally occurring tastes and odors pose no discernible health threats, their presence can greatly undermine public confidence in a water supply. Two common sources of offensive flavors and smells in natural surface waters are 2-methylisoborneol (MIB) and geosmin. These musty-smelling organic compounds are microbial byproducts that are easily detected by human olfactory senses. In fact, many people will notice an earthy smell and/or flavor if MIB concentrations exceed 7-15 ng/L (parts per trillion) or if geosmin concentrations exceed 4 ng/L (1-3). Thus, effective taste-and-odor control necessitates a technology that is capable of lowering MIB and geosmin concentrations to below the detection limit of modern chromatographic equipment. Several studies have shown that conventional treatment processes such as aeration, rapid sand filtration, and chemical oxidation cannot accomplish this (4-6). In some cases, ozone and advanced oxidants (ozone + hydrogen peroxide) can effectively elimi* Corresponding author phone: (201)797-7400; fax: (201)797-4558; e-mail:
[email protected]. † Present address: Malcom Pirnie, Inc., 17-17 Route 208 North, Fair Lawn, NJ 07410. ‡ Present address: Department of Environmental Engineering Sciences, The University of Florida, Gainesville, FL 32611. 276
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 1, 2004
nate MIB and geosmin, although the usefulness of these treatments depends largely on the background water quality (7-11). Notably, activated carbon has repeatedly proven to be a highly effective means for controlling MIB- and geosminrelated tastes and odors in a variety of natural waters (4, 12-16). Activated carbon is a versatile material that is widely used for removing organic compounds, including MIB and geosmin, from water. Its porous structure is ideal for adsorbing small, nonpolar chemicals. However, activated carbon has a finite capacity for adsorption, and this strongly impacts its application. In systems where powdered activated carbons (PACs) are used, PAC dosages are a function of adsorption capacity. In fixed bed adsorbers that utilize granular activated carbons (GACs), adsorption capacity dictates the frequency at which spent GAC must be replaced or reactivated. PAC usage and GAC reactivation/replacement represent major costs for water utilities. Hence, it would be of great benefit to the drinking water industry to enhance the MIB and geosmin adsorption capacity of commercial activated carbons. For the research herein, the authors employed 14C-MIB (see Materials and Methods) as a surrogate for MIB and geosmin since (a) these two compounds have a similar structure, (b) MIB has usually (but not always) been more difficult to adsorb than geosmin, so if an activated carbon exhibits superior MIB adsorption, it probably would also adsorb geosmin effectively, and (c) 14C-MIB can be monitored with greater ease and accuracy than nonradiolabeled MIB. Interestingly, although the drinking water industry is heavily dependent on activated carbon for taste-and-odor control, there have been few efforts to maximize activated carbon performance for this application. The authors herein have sought to develop techniques for improving the MIB adsorption capacity of a commercial lignite-based activated carbon. In work discussed elsewhere (17, 18), researchers at Penn State discovered that high-temperature treatments in pure hydrogen greatly enhanced MIB uptake. However, the safety issues associated with storing and transferring pure hydrogen would probably discourage an activated carbon manufacturer from using it. Thus, the authors sought to identify alternatives to hydrogen gas treatment that achieved the same favorable effects. It was discovered herein that a combination of methane and steam treatment improved 14CMIB uptake to a much greater extent than had been observed following hydrogen treatment. The literature has shown that at high temperatures, methane and steam will react to form hydrogen if a catalyst is present. However, the favorable effects of methane and steam treatment occurred even when these gases were applied sequentially so as to preclude hydrogen formation. These favorable effects, then, resulted from separate interactions of methane and steam with the activated carbon sample. A review of literature pertaining to catalysis and gas-phase carbon reactions gives some indication of how methane and steam interact with each other and with activated carbon surfaces at high temperatures and how these might lead to the favorable effects observed herein. Chemical vapor deposition of hydrocarbon gases/vapors has often been used to modify the pore structure of activated carbon materials (19-26). For instance, Kawabuchi et al. (23-25) have loaded benzene onto various commercial porous carbons (via chemical vapor deposition) to create molecular sieving materials capable of separating CH4 and CO2. They discovered that when benzene was applied onto activated carbon fibers, the CO2 selectivity of these fibers increased significantly while their CO2 adsorption capacity 10.1021/es026397j CCC: $27.50
2004 American Chemical Society Published on Web 11/21/2003
was relatively unaffected. In constrast, when benzene was applied onto a chemically activated carbon made from petroleum coke, or onto a thermally activated carbon made from coconut shells, CO2 selectivity improved slightly, but the CO2 adsorption capacity of these materials was almost completely eliminated. In addition, Kawabuchi et al. (25) reported that temperature strongly affected the nature of benzene deposition onto the micropore walls of activated carbon fibers. They discovered that an activated carbon fiber subjected to benzene deposition at 900 °C (900-ACF) lost its capacity for CO2 adsorption. When this same fiber was subjected to benzene deposition at 725 °C (725-ACF), its selectivity for CO2 increased, and its capacity for CO2 uptake was not seriously reduced. Following a microscopic analysis of the crosssectional surfaces of these treated fibers, Kawabuchi et al. (25) determined that benzene had deposited evenly throughout the pore structure of 725-ACF, whereas it mainly coated the outer surfaces of 900-ACF. This anisotropic distribution of carbon within 900-ACF was attributed to the thermal decomposition of benzene in the gas phase. At 900 °C, benzene molecules can combine to form large hydrocarbons that cannot fit into micropores. Instead they deposit on external surfaces and within large pores, thereby effectively blocking entrances to smaller pores. A similar phenomenon has been observed for methane deposition onto activated carbons, as discussed below. Under the right conditions, steam and methane can react to form hydrogen gas according to the following reaction, termed “steam methane reforming” (27-30).
H2O + CH4 / 3H2 + CO This reaction is highly endothermic and has only been found to occur in the presence of certain metal catalysts (27-30). Steam methane reforming has long been used to produce hydrogen on an industrial scale, and steam methane reforming reactors usually contain nickel- or platinum-based catalysts heated to 450-950 °C (27-30). It was initially surmised that steam methane reforming might be one way to generate hydrogen while heat-treating carbon to enhance its MIB removal performance. A mixture of methane and steam could be applied to the reaction chamber, and subsequent steam methane reforming reactions, catalyzed by the activated carbon present, would produce hydrogen. However, as described below, activated carbon is not an ideal material upon which to initiate steam methane reforming reactions. Commercial activated carbons do not contain high levels of noble metals (31) and thus are not naturally catalytic for this particular reaction. Furthermore, there are several mechanisms whereby steam methane reforming catalysts can be “deactivated,” namely sintering, sulfur poisoning, and carbon deposition, and these are likely to occur. In any case, the authors soon learned that the MIB adsorption characteristics of an activated carbon could be improved by the sequential application of methane followed by steam, and the two gases need not be used concurrently, as discussed below. At the elevated temperatures required for steam methane reforming, metal (catalyst) particles on carbon surfaces can sinter, thereby reducing total catalyst surface area. Since catalyst activity depends on surface area, sintering could severely inhibit the rate of steam methane reforming (27, 30). A second source of inhibitory effects on the catalysis of steam methane reforming reactions is sulfur poisoning. Sulfur can bond to the surface of catalytic metals (often irreversibly), thereby occupying sites that would otherwise be available for methane and steam transformations (27, 28, 32). Since H2S and COS are commonly released during coal gasification (32), there is a great potential for sulfur to poison catalytic
sites within activated carbon. Cannon et al. (33) similarly observed that sulfur poisoned the iron-catalyzed gasification of GACs during thermal reactivation. Finally, carbon deposition onto catalysts presents a problem similar to sulfur poisoning. As methane interacts with the catalyst surface, a fraction of the resultant free carbon atoms combine to form filamentous, graphitic carbon (27-30). The tendency for carbon deposition increases with increasing carbon coverage, so the process is autocatalytic. Carbon deposits coat the catalyst surface and prevent further steam methane reforming. Because carbon-rich surfaces tend to promote carbon deposits, the authors herein surmise that activated carbon pore walls could not act as sites for steam methane reforming reactions. Indeed, it is apparent that steam methane reforming reactions are not likely to occur when methane and steam are combined in the presence of activated carbon. Instead, methane and steam are likely to participate in separate reactions. Linares-Solano and Walker (34) carefully examined the interactions of methane and steam with various carbons surfaces. They observed that for an activated Saran char (Saran is a copolymer of polyvinylidene chloride and poly(vinyl alcohol)), the rate of mass change in an H2O/N2 (wet N2) environment at 900 °C was -45%/hr (i.e., the char lost mass). In contrast, when a mixture of H2O/CH4/N2 (wet 80% N2-20% CH4) was employed, the mass change was initially +180%/hr before tapering off (i.e., the char gained mass). Results for an activated lignite char were substantially different. The rate of mass change in these same environments at 900 °C was -350%/hr and -190%/hr, respectively (i.e., the chars lost mass in both cases). Linares-Solano and Walker (34) proposed that these results reflect the fact that carbons have different activities for methane cracking and their interactions with steam. These results also support the idea that steam and methane react separately, where methane undergoes deposition reactions while steam gasifies the existing carbon surface and/or the methane deposits. If a significant fraction of the steam and methane present in the above-mentioned experiments underwent steam methane reforming reactions to form hydrogen, one would not expect to see the extreme mass changes that were actually observed since hydrogasification exhibits much slower kinetics and only targets highly (re)active carbon atoms (35, 36). LinaresSolano and Walker (34) also pointed out that the rapid tapering off of the deposition rate observed for the activated Saran char indicated that methane deposits had rendered the micropores of that sample inaccessible. They explained that if the pore volume had been completely filled with deposited carbon, the sample would theoretically gain 88% mass, whereas the actual mass gain was 5.5%. Similar pore blockage effects resulting from methane deposition were similarly observed by Kamishita et al. (37). As previously mentioned, Linares-Solano and Walker (34) reported steam gasification rates of -45%/hr and -350%/hr for two activated chars heated to 900 °C. It was noted that the higher rate corresponded to the activated lignite char, which contained a much higher level of metallic impurities than did the activated Saran char (which essentially had none). In comparison, McKee and Chatterji (38) observed that the gasification rate for pure graphite in steam at 1000 °C was less than -1.5%/hr. This indicates that steam gasification reactions are more likely to target highly disorganized regions within a carbon sample as opposed to organized, “graphite-like” regions (if both are present). It is well established that steam gasification or “burnoff” can lead to enhanced porosity (surface area and pore volume) in carbon materials (39-47). Importantly, steam gasification that is catalyzed by calcium and other inorganics can cause a loss of micropores (48-50), and the authors herein regarded this as an unfavorable effect. VOL. 38, NO. 1, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
277
Steam can also act as a source of hydrogen via the watergas shift reaction. Water molecules can combine with carbon monoxide to form hydrogen and carbon dioxide, via the reaction
H2O + CO / H2 + CO2 As previously mentioned, hydrogen gas treatments can greatly improve the MIB adsorption capacity of an activated carbon, and so the water-gas shift reaction is a particularly important consideration for the work herein. Notably, the equilibrium constant for this reaction is 0.7 at 1000 °C, 1.0 at 800 °C, and 16.3 at 375 °C (51), indicating that the thermodynamic driving force for hydrogen formation via the water-gas shift reaction increases with decreasing temperature.
Objectives The primary objectives of this work were to (a) determine the extent to which heat treatments can improve the MIB removal performance of a lignite-based activated carbon when it operates in batch and fixed-bed adsorption modes and (b) characterize the physical/chemical changes brought about by these treatments that enhanced performance. The work discussed in this paper specifically focuses on developing a heat treatment protocol that employs methane and/or steam to enhance MIB uptake. As described elsewhere (17, 18), it was observed that hydrogen gas treatment doubled the MIB adsorption performance of an activated carbon operating in fixed-bed adsorption mode, and the main goal herein was to replicate or even exceed these results without using pure hydrogen. Additional discussion of results pertaining to objective (b) can be found in the dissertation that accompanies this research (17) and will appear in forthcoming papers (52, 53).
Materials and Methods The majority of materials and methods employed herein are more fully described in the above-mentioned dissertation (17). 14C-Labeled MIB. MIB adsorption studies were conducted using 14C-labeled MIB, purchased from American Radiolabeled Chemicals, Inc. (St. Louis, MO). This material was artificially synthesized such that it contained 14C and exhibited a specific activity of 55 mCi/mM. Consequently, the 14C-MIB concentration of a water sample is directly proportional to its radioactivity. Radioactivity was measured by means of a scintillation counter (Wallac 1217 Rackbeta - Wallac Inc., Gaithersburg, MD), and the detection limit of this instrument was approximately 5 disintegrations per minute (dpm), or about 3 ng/L 14C-MIB. Minicolumn Studies. Pirbazari et al. (3) determined that the constant diffusivity equations developed by Crittenden and co-workers (54-56) are appropriate for designing smallscale columns intended to simulate full-scale (fixed-bed) MIB adsorption performance. Hence, the columns employed herein were constructed according to the constant diffusivity equations. These columns were designed based on the fullscale parameters of the filter-bed contactors located at the Norristown water purification facility of the PennsylvaniaAmerican Water Company (Hershey, PA). As addressed elsewhere (17, 18), replicate minicolumn runs exhibited excellent reproducibility, to the extent that the authors felt comfortable performing single runs for many of the experiments (including some of those presented herein). Activated Carbons. As described above, this work focuses primarily on the use of heat treatments in methane and/or steam environments to improve the MIB adsorption performance of a commercial lignite-based activated carbon [Norit Americas Inc. (Atlanta, GA)] (designated HD4000 in 278
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 1, 2004
the text and figures). This material had previously been washed (by the manufacturer) in hydrochloric acid and thus contained low levels of inorganic ash. Importantly, the calcium content of the test carbon was about 0.12%, and this was low enough to preclude significant calcium catalysis of gasification reactions involving steam and the carbon surface (48, 49). It should be mentioned that all carbon samples utilized herein were ground and sieved to a 170 × 200 mesh size (geometric mean particle size ) 0.082 mm) prior to heat treatment. As discussed in the accompanying dissertation (17) and in a forthcoming paper (18), this was an appropriate grain size for minicolumn testing. Heat Treatments. Heat treatments were performed in a Cahn thermogravimetric analyzer (TGA - Cahn Instruments, Inc., Cerritos, CA). Treatment temperatures ranged from 375 °C to 1000 °C, and gas applications typically lasted 5 to 60 min. In general, the methane and/or steam application rates ranged from 1.0 to 14.0 g of gas per gram of carbon per hour. A list of all the experimental carbons tested herein, and the conditions under which they were prepared, is given in Table 1. It is important to note that steam is not easily introduced into a TGA. For the instrument utilized in this work, only a portion of the reaction chamber (a quartz glass tube) was heated, and treatment gases entered the vertical chamber (flowing upward) well below this region. The authors discovered that steam generated outside the TGA would condense before it reached the heated zone. In response, a novel approach to steam application was employed, wherein a stainless steel tube was installed inside the reaction chamber such that treatment gases flowed through this tube and entered the bulk atmosphere of the chamber within the heated zone. When a steam environment was required, distilled water was pumped through this tube, and as it entered the heated zone, the water converted to steam. As discussed below, initial trials of steam generation with this system indicated that the steam generation rate was not always constant when both water and methane were applied simultaneously. Natural Water Samples. As described elsewhere (17, 18), adsorption experiments were conducted using clarified water from the Norristown water purification plant. During the course of this research, samples were collected on two occasions and designated Water 1 and Water 2. The tests discussed in this paper utilized Water 2 exclusively, and it will be identified herein simply as “Norristown water.” Norristown water exhibited a pH of 7.8 throughout the entire study. Interestingly, although the water samples were stored in the dark at 4 °C, the background total organic carbon (TOC) concentration dropped from 3.7 (initially) to 2.6 mg/L over a period of 18 months. It was determined that the apparent loss of TOC over time did not significantly alter the MIB adsorption performance of the activated carbons tested herein. As shown in Figure 1, the MIB breakthrough performance of HD4000 in fresh Norristown water (TOC ) 3.7 mg/L) was the same as its performance in Norristown water that had been stored for 18 months (TOC ) 2.6 mg/L). Pore Size Analysis. Pore volume and pore size distribution for the experimental carbons were determined using a Micromeritics ASAP 2000 or 2010 Pore Analyzer, and these analyses were performed according to the protocol of Moore et al. (57). The Micromeritics instruments generate argon adsorption isotherms that are converted to porosity data by means of the Density Functional Theory (58, 59). Slurry pH. The slurry pH of a carbon sample was determined using a method based on the procedure of Mene´ndez et al. (35). Powdered carbon was combined with distilled-deionized water in a glass scintillation vial at a ratio of 10% (by weight) carbon to water. The slurry was purged
TABLE 1. Experimental Carbons - Treatment Conditions carbon
treatment conditions
HD4000 N-850 H-1025 M-1000 (x% gain) M-1000 (70:3,10% gain)
untreated (commercial lignite-based activated carbon) nitrogen at 850 °C for 1 hour hydrogen at 1025 °C for 1 hour methane at 1000 °C until x% mass gain (i.e., 8, 29, 33-36, or 108% mass gain) 70:3 (by volume) nitrogen-methane mixture at 1000 °C until 10% mass gain
M-725 (4% gain) S-375/N-850 MS-600
methane at 725 °C until 4% mass gain steam at 375 °C for 1 h followed by temperature ramping to 850 °C in nitrogen methane and steam (mixed) at 600 °C for 60 min
MS-850
methane and steam (mixed) at 850 °C for 20 min
MS-1000a, b
methane and steam (mixed) at 1000 °C for 25 min
M/S-1000a, f
methane at 1000 °C until 10% mass gain followed by steam at 1000 °C until 23% mass loss (relative to pyrolyzed mass), cool in steam + nitrogen methane at 1000 °C until 10% mass gain followed by steam at 1000 °C until 23% mass loss (relative to pyrolyzed mass), cool in nitrogen only
M/S-1000b-e
FIGURE 1. MIB breakthrough profiles for carbons treated in methane + steam at 1000 °C [MS-1000a,b; M/S-1000a]. The profiles for virgin [HD4000] carbon are included for comparison and to show that MIB breakthrough performance was unaffected by changes in the background TOC concentration of Norristown water. Influent MIB concentration ) 135 ng/L. with nitrogen for several minutes prior to sealing the vial, and the mixture was then agitated on a rotating tumbler. After 24 h, the slurry pH was measured and recorded as the “slurry pH” of the carbon sample. Mene´ndez et al. (35) recommended that the work of Noh and Schwarz (60) justifies using slurry pH as an accurate indicator of the pH of zero surface charge (pHpzc), although others have not adopted this notion.
Results and Discussion As a culmination of the research activity discussed herein, the authors were able to tailor an activated carbon such that it processed about 4 times more water than did commercial HD4000 prior to initial 14C-MIB breakthrough (Figure 1). This was accomplished by heating HD4000 to 1000 °C and either (a) exposing it to a mixture of methane and steam [MS-1000a] or (b) first exposing it to methane and then to steam [M/S1000a]. It would appear that this modified activated carbon could be of considerable benefit to the drinking water industry. The following narrative describes the steps leading to this favorable outcome. These included the following: (a) thermal treatment in hydrogen, as discussed elsewhere (17, 18), (b)
gas flow rates (g gas/g GAC‚h) N/A N2 ) 23.9 H2 ) 1.7 CH4 ) 8.1-9.4 N2 ) 14.0 CH4 ) 0.3 CH4 ) 9.7 H2O ) 10.0 CH4 ) 1.0 H2O ) 8.2 CH4 ) 7.7 H2O ) 9.2 CH4 ) 6.8 H2O ) 8.2 CH4 ) 6.1-7.5 H2O ) 7.3-9.0 CH4 ) 3.0-7.1 H2O ) 3.6-8.6
thermal treatment in pure methane (>99.99 vol %), (c) lowtemperature steam treatments followed by a ramped temperature phase in nitrogen, (d) thermal treatment in mixtures of pure methane and steam or involving separate phases of methane and steam exposure, and (e) high-temperature steam treatments, as discussed in the dissertation that accompanies this work and in a forthcoming paper (17, 52). Methane Treatment. As discussed elsewhere (17, 18), Mene´ndez and co-workers (35) indicated that the formation of CH2 radicals during high temperature hydrogen treatments leads to the inactivation (“healing”) of reactive sites on the carbon surface. These inactivated sites do not subsequently react to form oxygenated functional groups when the GAC is later stored in air or used in water treatment. Since methane can dissociate to form CH2 radicals (27-30), the authors herein hypothesized that high-temperature methane treatments could eliminate reactive sites within the experimental carbon and thus improve its MIB adsorption performance. To test this hypothesis, several samples of HD4000 were treated in a flow of pure methane at 1000 °C. Treatment times varied, as it was found that upon switching to a methane environment (from nitrogen) in the furnace, the carbon samples steadily gained mass until the treatment ended (Figure 2). It should be noted that the mass gain shown in Figure 2 is reported as a percentage of the pyrolyzed mass of the sample. In every gas heat treatment described herein, carbon samples were heated to the desired treatment temperature in a flow of pure nitrogen. For the nomenclature herein, “pyrolyzed mass” refers to the sample mass immediately following this ramped temperature phase and just prior to methane and/or steam treatment. Referring to Figure 2, the mass gain illustrated there was due to chemical vapor deposition of methane within the pore structure of the sample carbon. Chemical vapor deposition is commonly used for pore size control in porous carbonaceous materials, as discussed above, and the effects of chemical vapor deposition on MIB uptake were not known when these samples were prepared. Figure 3 shows MIB breakthrough profiles for several samples that were treated in pure methane at 1000 °C. The samples are identified according to their mass gain (above pyrolyzed mass) during treatment. It was discovered that the breakthrough performance for these carbons was inversely related to mass gain. In other words, the time to MIB breakthrough decreased as mass gain increased. Based on pore size distribution analyses, it appeared that pore blockage/pore filling was largely responsible for the observed loss in adsorption capacity. As shown in Figure 4, a sample that VOL. 38, NO. 1, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
279
FIGURE 2. Mass and temperature profiles during a representative trial of methane treatment (mass is shown as a percentage of the “pyrolyzed mass”, where “pyrolyzed mass” refers to the mass immediately following the temperature ramping phase and prior to methane treatment).
FIGURE 3. MIB breakthrough profiles for carbons treated in (a) pure methane at 1000 °C (percentages correspond to the mass gained during methane treatment), (b) pure methane at 725 °C, and (c) a 70:3 (by volume) nitrogen-methane mixture at 1000 °C. The profile for virgin carbon [HD4000] is also included. Influent MIB concentration ) 135 ng/L. gained 8% mass [M-1000 (8% gain)] during methane treatment at 1000 °C contained significantly less pore volume than HD4000, and a sample that gained 108% mass [M-1000 (108% gain)] did not contain any measurable pores. This loss of pore volume was also manifested in the mass profile shown in Figure 2. The declining rate of mass gain during methane treatment indicates that the internal surface area of the carbon sample gradually became inaccessible. Assuming that the rate of chemical vapor deposition per unit surface area remained constant, a declining rate could only occur if the surface area available for chemical vapor deposition was reduced. As discussed below, there was sufficient evidence to suggest that this apparent loss of internal surface area (pore volume) was due to pore blockage, rather than pore filling effects. Comparing the pore size distributions of M-1000 (8% gain) and HD4000, it first appeared that methane treatment had uniformly affected all the pores within M-1000 (8% gain). Notably, when the pore volume distribution data for HD4000 was multiplied by 0.701 (the ratio of pore volume contained in
