Real Structure of Column Surfaces

the adsorption enthalpies evaluated by the procedures which have been in use to date do roughly correlate with the vaporization or sublimation parameters of bulk compounds. 

The mere observation of the above correlations has been widely used to judge the parameters of macro volatility of the (oxy)halides of TAEs. Not enough attention has been paid to the fact that the evaluated — Aa(js// values fall between the corresponding Avap H and Asub H. It is not at all obvious that it must be so be if we imagine a molecule interacting with bare flat surface of fused silica and compare the situation with that of a molecule in the condensed phase, where it is tightly surrounded by a number of its own replicas see Fig. 5.4. 

These are the most important findings. The linearity of the correlations is not of major concern — it becomes questionable when taking the narrower region of the A ads values of truly molecular liquids. 

This uneasy situation, with such important quantities, calls for a search of the physicochemical rationale of the regularities. The poor understanding of the absolute values of adsorption enthalpy gives little ground to any far-going conclusions from the experimental data. It concerns, for example, the manifestation of relativistic effects in chemical properties of the new elements, which is believed to be evidenced if a TAE compound is more volatile (less adsorbable) then that of a lighter homolog. 

The clue to the similarity of micro and bulk volatility parameters seems to be in the true structure of the column surface. There is overwhelming evidence that real surfaces of solids are, almost in all respects, heterogeneous rather than homogeneous. The starting point is the common roughness, which increases the real surface area, compared with the smooth case. Most in-depth studies evidence a broad 

---

Other Experimental Methods. It is probably suitable to discuss here column porous structure. Porous space of a conventional packed column consists of the interparticle volume (Vip—space around particles of packing) and pore volume (Vp— space inside porous particles). The sum of those two constitutes the column void volume. The void volume marker ( unretained ) should be able to evenly distribute itself in these volumes while moving through the column. Only in this case the statistical center mass of its peak will represent the true volume of the Uquid phase in the column. In other words, its chromatographic behavior should be similar to that of the eluent molecules in a monocomponent eluent. If a chosen void volume marker compound has some preferential interaction with the stationary phase compared to that of the eluent molecules, it will show positive retention and could not be used as void marker. If on the other hand it has weaker interaction, it will be excluded from the adsorbent surface and will elute faster than the real void time, meaning that it also could not be used. For any analytical applications (when no thermodynamic dependences are not extracted from experimental data), 10% or 15% error in the determination of the void volume are acceptable. It is generally recommended to avoid elution of the component of interest with a retention factor lower than 1.5. Accurate methods for the determination of the column void volume are discussed in Chapter 2. 

---