IMPORTANCE OF THE ACCURACY OF EXPERIMENTAL-DATA IN THE NONLINEAR CHROMATOGRAPHIC DETERMINATION OF ADSORPTION ENERGY-DISTRIBUTIONS

Adsorption energy distributions (AEDs) are calculated from the classic al, fundamental integral equation of adsorption using adsorption isoth erms and the expectation-maximization method of parameter estimation. The adsorption isotherms are calculated from nonlinear elution profile s obtained from gas chromatographic data using the characteristic poin ts method of finite concentration chromatography. Porous layer open tu bular capillary columns are used to support the adsorbent. The perform ance of these columns is compared to that of packed columns in terms o f their ability to supply accurate isotherm data and AEDs, The effect of the finite column efficiency and the limited loading factor on the accuracy of the estimated energy distributions is presented. This accu racy decreases with decreasing efficiency, and approximately 5000 theo retical plates are needed when the loading factor, L(f), equals 0.56 f or sampling of a unimodal Gaussian distribution. Increasing L(f) furth er increases the contribution of finite efficiency to the AED and caus es a divergence at the low-energy endpoint if too high. This occurs as the retention time approaches the holdup time. Data are presented for diethyl ether adsorption on porous silica and its C-18-bonded derivat ive. Both the frontal-analysis-by-characteristic-points (FACP) and the elution-by-characteristic-points (ECP) methods of gas chromatography are presented. The FACP experiment yielded divergent results at high c olumn loadings for this system. The results indicate that the accuracy of estimated AEDs with respect to the true, underlying distribution o f the surface is poor for most studies of this nature.