1442
Langmuir 1999, 15, 1442-1448
Determination of the Specific Surface Area and the Pore Size of Microporous Carbons from Adsorption Potential Distributions Michal Kruk and Mietek Jaroniec*,† Department of Chemistry, Kent State University, Kent, Ohio 44240
Kishor P. Gadkaree Corning, Inc., Corning, New York 14831 Received June 30, 1998 Porous structures of synthetic active carbons were characterized using nitrogen adsorption at 77 K over a wide range of relative pressures (10-7-0.995), and it was shown that the analysis of adsorption potential distributions (APDs) in many cases allows for estimation of the specific surface area and the average micropore size. The APDs for most of the samples had two peaks, which can be related to the monolayer formation on the micropore surface and to the secondary micropore filling. The minimum between these two peaks was identified as a point of completion of the monolayer, allowing for evaluation of the specific surface area in a good agreement with the calculations made using the advanced DFT Plus software. Assuming the slitlike pore geometry, the obtained specific surface areas were used to calculate average micropore sizes for the synthetic active carbons as well as for commercial high-surface-area porous carbons. The amounts adsorbed per unit surface area (normalized adsorption curves) were calculated for these samples and showed a systematic change with the average micropore size. As the latter increased, the normalized adsorption at low pressures (below ca. 10-3) gradually decreased and for samples with larger micropores it became more and more similar to the normalized adsorption on a macroporous nongraphitized carbon. In addition, a gradual development of the secondary micropore filling was observed. The position of peaks on APDs was related to the average micropore size providing another simple method of micropore size estimation. The observed changes in the low-pressure adsorption behavior correlate well with those predicted by computer simulations and DFT calculations, but the experimentally observed monolayerformation and secondary-micropore-filling transitions were smoothed, which can be attributed to the surface and pore-size heterogeneity.
Introduction Active carbons are commonly used adsorbents for separation of mixtures and purification of liquids and gases.1 The usefulness of active carbons arises from their large surface areas and extensive porous structures of predominantly microporous character, which lead to high adsorption capacities for various substances.1 A significant number of studies have been devoted to application of gas adsorption techniques to evaluate the specific surface area, micropore size, and micropore size distribution for activated carbons and other carbonaceous adsorbents.1-3 However, a reliable determination of these structural parameters on the basis of gas adsorption data is still difficult, especially using a single gas rather than a series of gases of different molecular sizes.2-5 The calculation of the micropore size and/or micropore size distribution from single gas adsorption data is commonly based on the work of Dubinin and co-workers6 or its later modifications.1,2 Dubinin’s theory predicts a one-stage gradual micropore filling.1,2,6 However, recent experimental3,7-9 as well as † E-mail:
[email protected]. Phone: (330) 672 3790. Fax: (330) 672 3816.
(1) Bansal, R. C.; Donnet, J.-B.; Stoeckli, F. Active Carbon; Marcel Dekker: New York, 1988. (2) Jaroniec, M.; Madey, R. Physical Adsorption on Heterogeneous Solids; Elsevier: Amsterdam, 1988. (3) Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area and Porosity; Academic Press: London, 1982. (4) Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem. 1985, 57, 603. (5) Roquerol, J.; Avnir, D.; Fairbridge, C. W.; Everett, D. H.; Hayness, J. H.; Pernicone, N.; Ramsay, J. D. F.; Sing, K. S. W.; Unger, K. K. Pure Appl. Chem. 1994, 66, 1739. (6) Dubinin, M. M. Prog. Surf. Membr. Sci. 1975, 9, 1.
theoretical (density functional theory (DFT) and computer simulation)10-13 studies demonstrated that, in many cases (e.g., nitrogen at 77 K), adsorption in micropores may be a one-stage or two-stage process depending on the micropore size. Namely, micropores which can accommodate not more than two layers of adsorbed molecules exhibit a one-stage micropore filling, whereas adsorption in wider micropores involves two stages: the monolayer formation (ML) and the subsequent secondary micropore filling (SMF) at higher pressures. This two-stage process cannot be adequately described by Dubinin’s theory. More recently, the Horvath-Kawazoe (HK) method to calculate micropore size distributions14 has attracted much attention. The HK calculations are based on the condensation approximation (CA),2 which does not appear to be realistic and leads to a significant distortion of the actual pore size distribution (PSD), which may manifest itself as, for instance, an appearance of artificial peaks.15 New promising approaches for characterization of microporous carbons, for instance the commercially available DFT Plus software,16 are based on advanced computational methods (7) Kaneko, K.; Ishii, C.; Rybolt, T. Stud. Surf. Sci. Catal. 1994, 87, 583. (8) Kenny, M.; Sing, K.; Theocharis, C. In Fundamentals of Adsorption; Suzuki, M., Ed.; Kodansha: Tokyo, 1993; p 323. (9) Li, Z.; Kruk, M.; Jaroniec, M.; Ryu, S.-K. J. Colloid Interface Sci. 1998, 204, 151. (10) Seaton, N. A.; Walton, J. P. R. B.; Quirke, N. Carbon 1989, 27, 853. (11) Lastoskie, C.; Gubbins, K. E.; Quirke, N. Langmuir 1993, 9, 2693. (12) Olivier, J. P. J. Porous Mater. 1995, 2, 217. (13) Setoyama, N.; Suzuki, T.; Kaneko, K. Carbon 1998, 36, 1459. (14) Horvath, G.; Kawazoe, K. J. Chem. Eng. Jpn. 1983, 16, 470. (15) Kruk, M.; Jaroniec, M.; Choma, J. Adsorption 1997, 3, 209.
10.1021/la980789f CCC: $18.00 © 1999 American Chemical Society Published on Web 01/14/1999
Microporous Carbons
of modeling of adsorption in porous media, such as DFT and computer simulations.10,16,17 The latter are employed to provide state-of-the-art adsorption isotherms for micropores of assumed sizes and geometries, which are used in the integral equation for the overall adsorption to describe the local adsorption behavior, allowing one to invert the latter equation with respect to PSD. This concept is very powerful, but the local adsorption isotherms employed so far have been obtained under assumption of surface homogeneity and simple pore geometry, which may lead to considerable inaccuracy for strongly heterogeneous surfaces and/or pores of different geometries.16,18,19 Comparative analysis of adsorption data3,20 is also worth mentioning, since it provides a possibility of a simple qualitative estimation of the micropore size on the basis of the shape of the comparative plot curves.7,8 For wider micropores, the latter curves can also be analyzed to assess the specific surface area, which in turn can be used to estimate the micropore size assuming a certain simple micropore geometry (e.g., slitlike).7,13 In many cases, surface heterogeneity effects can easily be accounted for in the comparative analysis by a proper choice of the reference adsorbent.18,19 The specific surface area for microporous adsorbents is usually estimated using the BET method.2,3 However, adsorption in micropores is appreciably enhanced in comparison to the adsorption on a flat surface and proceeds via the micropore filling mechanism (either one-stage or two-stage) rather than the multilayer adsorption and therefore one cannot expect the BET calculations to be accurate. Indeed, it was recognized that the BET monolayer capacity is often close to the total micropore volume and the derived specific surface area may be unrealistically high.21,22 Recently, the development of the DFT software16 based on the NL DFT local adsorption isotherms provided a new method of surface area evaluation, which is independent from the standard BET method. It is thus clear that new approaches for simple and yet reliable evaluation of specific surface area and micropore size would be highly desirable. Recent experimental and model studies of nitrogen adsorption at 77 K on porous carbons drew our attention to a possibility of extracting structural information from adsorption potential distributions (APDs).9,15,23-27 The APDs can simply be evaluated by numerical differentiation of adsorption isotherms with respect to the adsorption potential. It was reported26,27 that minima on APDs between monolayer formation peaks and multilayer formation regions correspond to the completion of monolayer for carbon blacks. It was also postulated9 that the shape and position of peaks on APDs reflects the pore size of microporous carbons, which was previously demonstrated in a model study.15 The aim of the current work was to develop simple and practical methods to evaluate the specific surface area and the pore (16) Olivier, J. P. Carbon 1998, 36, 1469. (17) Lastoskie, C.; Gubbins, K. E.; Quirke, N. J. Phys. Chem. 1993, 97, 4786. (18) Sayari, A.; Kruk, M.; Jaroniec, M. Catal. Lett. 1997, 49, 147. (19) Kruk, M.; Jaroniec, M.; Choma, J. Carbon 1998, 36, 1447. (20) Jaroniec, M.; Kaneko, K. Langmuir 1997, 13, 6589. (21) Foley, H. C. Microporous Mater. 1995, 4, 407. (22) Lamond, T. G.; Marsh, H. Carbon 1964, 1, 281. (23) Jaroniec, M.; Kruk, M.; Choma, J. In Characterization of Porous Solids IV; McEnaney, B., Mays, T. J., Rouquerol, J., Rodriguez-Reinoso, F., Sing, K. S. W., Unger K. K., Eds.; Royal Society of Chemistry: London, 1997; p 163. (24) Kruk, M.; Jaroniec, M.; Gadkaree, K. P. J. Colloid Interface Sci. 1997, 192, 250. (25) Kruk, M.; Li, Z.; Jaroniec, M.; Betz, W. R. Langmuir 1999, 15, 1435. (26) Choma, J.; Olivier, J.; Jaroniec, M. Biul. WAT 1996, 45, 7. (27) Choma, J.; Jaroniec, M. Pol. J. Chem. 1997, 71, 380.
Langmuir, Vol. 15, No. 4, 1999 1443
size of microporous carbons on the basis of adsorption potential distributions. Experimental Section Materials. The synthetic active carbons were prepared at Corning Research Center (Corning, NY) by controlled carbonization of phenolic precursors on ceramic supports using proprietary processing methods.24 The ACF-20 and M-30 carbons were manufactured by Osaka Gas (Osaka, Japan). Nitrogen adsorption characterization of ACF-20 was reported in detail elsewhere.9 The MAXSORB carbon was manufactured by Kansai Coke and Chemicals (Amagasaki, Japan).28 Measurements. Nitrogen adsorption measurements were carried out at 77 K using an ASAP 2010 volumetric adsorption analyzer from Micromeritics (Norcross, GA). Before the measurements, the carbons were outgassed under vacuum for 2 h at 473 K. Calculations. The BET specific surface areas SBET2,3 were evaluated using adsorption data in the relative pressure range from 0.04 to 0.14. The total pore volumes Vt were estimated from a single point adsorption at ca. 0.995p/p0.3 The micropore volumes Vmi and the external surface areas Sex were calculated using the Rs-plot method3,7,8,20 using data in the Rs range from 1.5 to 2.5. The detailed information about the calculation procedure and the reference adsorption isotherm used can be found elsewhere.24 The specific surface areas SDFT and pore size distributions were obtained using the DFT Plus software (Micromeritics, Norcross, GA).16 The average micropore sizes wDFT (defined as average sizes of pores of the width below 4 nm) were evaluated from the incremental pore size distributions. The adsorption potential distributions (APDs) were calculated by numerical differentiation of the adsorption isotherms: X(A) ) - dv(A)/dA, where v(A) is the adsorbed amount (expressed in cm3 of the liquid adsorbate per gram) as a function of the adsorption potential A. The latter is defined as the change in the Gibbs free energy of adsorption ∆G with the minus sign: A ) -∆G ) RT ln(p0/p), where R is the universal gas constant, T is the absolute temperature, p0 is the saturation vapor pressure, and p is the equilibrium vapor pressure. The conversion from the amount (volume) of the adsorbed gas expressed in cm3 STP/g to the corresponding volume of the liquid adsorbate (in cm3/g) was made by multiplying the former volume by 0.001 546 8. The specific surface area was calculated from the APD data assuming that the cross-sectional area of nitrogen molecule is equal to 0.162 nm2. For the sake of simplicity, it was assumed in the calculations that both the density of nitrogen and the cross-sectional area of its molecules are independent from the micropore size. It needs to be noted that, for the synthetic active carbons under study, the structural parameters, such as surface areas and pore volumes, were calculated for a unit mass of carbon in the samples rather than for a unit mass of the whole samples.
Results and Discussion Nitrogen Adsorption Isotherms and Comparative Plots. Structural parameters for selected synthetic active carbons under study are listed in Table 1. Nitrogen adsorption isotherms for a series of samples with systematically increasing BET specific surface areas and total pore volumes (samples A-E, G, and J) are shown in Figure 1. It can be seen that the shape of adsorption isotherms gradually changes. The adsorption isotherm for the sample A approaches its maximum adsorption capacity at low relative pressures. The other samples exhibit more significant increase in adsorption at higher pressures, which can be seen from gradually developing knees in the relative pressure range up to 0.05-0.3. An increase in the amount adsorbed in this pressure region can be attributed to the secondary micropore filling (SMF), which takes place in micropores wide enough to accommodate more than two layers of adsorbed molecules. The pressure at which the secondary micropore filling takes place was shown to (28) Otowa, T.; Nojima, Y.; Miyazaki, T. Carbon 1997, 35, 1315.
1444 Langmuir, Vol. 15, No. 4, 1999
Kruk et al.
Table 1. Selected Structural Properties of the Synthetic Active Carbons under Studya sample
SBET (m2/g)
Vt (cm3/g)
Vmi (cm3/g)
Sex (m2/g)
SDFT (m2/g)
wDFT (nm)
A B C D E F G H I J
670 850 1040 1150 1490 1450 1840 1570 1200 2320
0.30 0.38 0.46 0.52 0.68 0.65 0.85 0.73 0.62 1.13
0.29 0.36 0.44 0.49 0.64 0.63 0.80 0.68 0.52 1.04