Abstract
Gas adsorption techniques represent what are perhaps the most common approach to the characterisation of the pore structure of porous materials. Within this general area, the most popular technique by far is to derive surface area and pore size distribution information from nitrogen adsorption isotherms obtained at 77 K. Surface areas are most typically obtained from BET analyses (multipoint or single point), and pore size distributions from a number of different techniques. Mesopore size distributions are usually obtained from techniques based on various adaptations of the Kelvin equation, such as the BJH [1] and DM [2] approaches. A number of different approaches have been proposed for micropore analysis, although it appears that no single technique is applicable to all situations. These range from micropore volume determinations using the so-called t-plot and α-plot methods, involving comparisons of experimental isotherms to standard isotherms, to pore size distribution determinations based on micropore-filling arguments, such as those related to the Dubinin-Radushkevitch equation, and others, like the Horvath-Kawazoe [3] and density functional theory [4] methods. Often, pore size distribution determinations based on micropore filling arguments use CO2 adsorption at 273 K, to circumvent activated diffusion problems which may be encountered by nitrogen at 77 K in the smallest micropores (i.e., the ultramicropores).
The theoretical limitations of these techniques are well documented, and often different techniques result in micropore size distributions which may differ significantly from one another. This state of affairs can be inadequate for a number of potentially important applications of microporous carbons; e.g., for methane storage, for which it has it has been shown theoretically [5] that adsorption is a strong function of pore size.
Small angle scattering (SAS) techniques represent an alternative to gas adsorption methods, with a number of advantages. SAS is sensitive to both closed and open porosity and in many cases offers a more complete picture of porosity. SAS can also be applied to 'wet' samples. The two most common subatomic scattering particles used in SAS are X-rays and neutrons, which can both be produced at suitable wavelengths for small angle scattering analyses. The generally greater accessibility to, and availability of X-ray instruments has made small angle X-ray scattering (SAXS) more common than small angle neutron scattering (SANS).
The physical mechanisms of scattering are completely different for the two types of subatomic scattering particles. For SAXS, interactions with electrons are most important, whereas interactions with nuclei are more important for SANS. There have been no direct comparisons between these two techniques for microporous carbons and it has generally been implicitly assumed that they yield essentially the same information. The purpose of the present communication is to explore this latter assumption via a direct comparison between SAXS and SANS data for a microporous carbon.
SAXS measurements were carried out using a Kratky camera mounted on a fully stabilised Phillips generator, Type PW 1010/1. The generator was operated at 40 kV and 20 mA. A Ni filter was used for the Cuβ radiation. Both entrance and counter slits had slit widths of 250 μm.
Carbon samples were loaded into a 2 mm thick sample holder, 1 cm wide and 1.6 cm in height, covered with kapton film on both sides. The powdered carbon samples were shaken in the holder to ensure homogeneous packing. The measurements were made with a proportional counter linked to an IEEE interface, which controlled the step scanning device to allow exact repetition of the measurements. The scans were made between the 2θ values of 0.1=2θ=3° with steps of 0.05° and counting times of 60 s per point.
SANS was performed at the Intense Pulsed Neutron Source (IPNS) at the Argonne National Laboratory at the small angle neutron diffractometer (SAND). The sample holders were made of Suprasil with a path length of 2 mm. The scattering data were corrected for the scattering from the sample holder and other instrumental backgrounds. Normalisation for the sample thickness and transmission were made and the data were scaled to yield absolute intensities. Detailed descriptions of the SAND instrument and its calibration have been provided by Thiyagarajan et al. [6]. No corrections were made for incoherent scattering because only trace amounts of hydrogen were present. It was found that transmission times of 15 min. and scattering times of 45 min. produced scattering curves at high q with sufficiently low variances to allow meaningful comparisons with corresponding SAXS data.
The sample char was prepared by heating phenolic resin under nitrogen, at a heating rate of 10 Kmin−1, to 900°C. The sample was held at this latter temperature for 1 h and then cooled under nitrogen.
Fig. 1 shows the results of SAXS and SANS intensities from samples of the same phenolic resin char. The SAND instrument at IPNS yields scattering profiles in terms of the absolute differential scattering cross section with units of cm−1 whereas the SAXS instrument yields intensities in arbitrary units. Consequently, the SAXS data has been scaled down to fit onto the same axes. This scaling procedure does not change the shape of the scattering curve. The error bars on the SANS data reflect the variance of the data. The greater scattering intensities of the X-ray data means that the variances are much smaller than the corresponding neutron data.
Original language | English |
---|---|
Pages (from-to) | 1257-1259 |
Number of pages | 2 |
Journal | Carbon |
Volume | 38 |
Issue number | 8 |
DOIs | |
Publication status | Published - 2000 |
Keywords
- porous carbon
- neutron scattering
- X-ray scattering
- electronic properties
- microporosity