On The Accuracy And The Repeatability Of Results Of Benzene Vapour Adsorption Measurements
ON THE ACCURACY AND THE REPEATABILITY OF RESULTS OF BENZENE VAPOUR ADSORPTION MEASUREMENTS
B.J. Trznadel
Military Institute of Chemistry and Radiometry, 00-910 Warsaw, Poland
Activated carbons are structurally and energetically heterogeneous adsorbents because they possess a complex porous structure (Jankowska et al., 1991). The porous structure is usually described by the characteristic surface area, total and micropore volume and pore size distribution or energy distribution (Bansal et al., 1988). Additionally surface roughness is analysed as a structural parameter (Pfeifer and Avnir, 1983). Every porous structure characteristic is commonly determined on the basis of the analysis of adsorption-desorption isotherms. Nitrogen, argon or benzene vapour are mainly used as an adsorbate in such measurements (Dubinin, 1975). The UPAC (Sing et al., 1985) recommended the use of nitrogen as a standard adsorptive but in Sing’s opinion (1989), the main disadvantage of nitrogen is untypical size and shape of nitrogen molecule. Thus, nitrogen due to its micropore filling behaviour has been found gives consistently higher values of effective micropore volume than those obtained by the use of other adsorbates. Moreover, it was shown (Trznadel et al., 1999) that the analysis of benzene vapour isotherms give higher reliability of the various structural parameters determined for the porous structure of the activated carbons than nitrogen adsorption measurements. Analysis of adsorption data leads to similar trends in changes of structural parameters (Trznadel and Cwitkowski, 1999) and conclusions are independent of used adsorbate.
The question arising now concerns the accuracy and repeatability of the results obtained on the basis of benzene vapour adsorption measurements. This problem has a great importance, because the proper evaluation of structural parameters of activated carbons affects possibilities of their applications. On the other hand the European standard EN-45001 orders the validation of each testing procedure used in any accreditation laboratory.
The aim of this work is the determination of the accuracy and repeatability of the results of the analysis of benzene vapour adsorption-desorption isotherms on activated carbon.
The activated carbon of AG-5 type produced by HPSDD (Hajnówka, Poland) was used in experiments. Granules of carbon were partially ground to unify shapes and sizes and next the fraction of diameter 0.75-1.0mm was separated by sieving. This procedure was used according to conclusions of the work of Buczek et al. (2000). It was shown (ibid.) that external parts of the granules of activated carbons exhibit less differences of structural parameters for samples of different level of burn-off then internal cores. This elimination of outer parts of granules leads to greater differentiation of the values of structural parameters.
Two methods were used for separation carbon samples: the static and the dynamic. In a case of the static method the samples (designated S1-S15) were taken from central, North, East, South and West points of upper, middle and bottom layers of the storage tank. In a case of the dynamic methods separation for fraction (designated D1-D11) was performed by the elutriation in an air steam of constant flow-rate using apparatus proposed by Diduszko et al. (2000). The benzene vapour adsorption-desorption isotherms at 293K were careful aquisited for all samples by gravimetric method using a McBain-Bakr vacuum balance.
Three groups of the adsorption models were used in the analysis of isotherms. The first one based on the theory of volume filling of micropores consists Dubinin-Raduszkevich (DR), the Izotova-Dubinin (ID) and Dubinin-Stoeckli (DS) equation. The second one is based on the statistical thermodynamics of sorbed gas molecules on surfaces and the equation of Horvath-Kawazoe (HK) (1983) adapted to the benzene vapour/activated carbon system (Cwitkowski et al., 1995) was used in analysis. The third one, strictly empirical, is based on the Froundlich’s equation (FR) (Boededecker, 1985, Froundlich, 1926).
Additionally the fractal dimension (a quantitative measure of a surface roughness) was calculated for all samples by the method proposed by Ehrburger-Dolle (1994).
The linear, surface and volume structural parameters were evaluated on the basis of the mentioned adsorption models. The obtained results indicate that one can observe real dispersion of structural characteristics in production batch of examined activated carbon, but larger range of changes is observed in a case of dynamic method of sampling.
The statistical analysis of the values of structural parameters indicates that accuracy of measurements is better in a case of the simple than in a case of the complex adsorption models (especially DS equation). The best accuracy and repeatability were deduced for the fractal dimension and the average micropore half-width connected with the power law. This fact is a piece of evidence that fractality really prevails in a case of activated carbons.
Related articles::