Displacer-based desorption

Displacer based desorption is practiced more often in liquid-phase processes, particularly in ion-exchange resin based processes. Conventional adsorption based liquid-phase processes are not particularly suitable for displacer based processes. Further, displacer based processes are invariably used in multicomponent separations (Rhee and Amundson, 1982) and are properly considered under chromatographic processes (see Section 7.1.5) (Frenz and Horvath, 1985). 

The analyte must be efficiently recovered. The usual mechanism for solvent desorption is selective displacement of the analyte. Selective displacement occurs as a more polar solvent displaces a less polar one on charcoal, just as a more active ion displaces a less active one on ion exchange resins. CS2 is frequently used to recover substances from charcoal, but simple alcohols cannot be displaced from charcoal by CS2, and it is necessary to add l%-5% of another alcohol to the CS2 to facilitate desorption. Frequently, low recoveries can be increased by increasing the quantity of solvent, if analytical sensitivity permits. Prospective solvents may be chosen based on polarity or solubility of the analyte. 

Some preliminary runs were made on the Al O support in the absence of Pt. Heating this alumina without CO pulsing showed a small peak appearing near 100°C. This was identified as N adsorbed as an impurity from the carrier gas. CO completely displaces the on pulsing at room temperature and, as previously reported, probably results from the presence of Lewis acid-base sites gn the alumina (8). The amount of this peak was approximately 10 moles per g of Al O which represents less than 5% of the CO desorbing from a well dispersed Pt catalyst of 0.4 wt% loading. Runs without the in-situ H O and 0 trap had shown that C0 desorbed with the CO. However, with the In-situ trap minimal CO was observed. This was confirmed by a pair of runs shown in Figure 1 with and without Al O (CO2 trap) placed at the exit of the desorption tube. 

The latest developments in the issue are indicating that the view based on the H adsorption model is subject of some revision. In References 23 and 24 the voltammetric contribution of some specifically adsorbed anions (acetate, oxalate, chloride and bromide) was studied in the case of Pt(lll) electrodes by means of experiments involving the displacement of the adsorbed species by CO in acidic medium. The conclusion of this study was that the usual states correspond to the reversible adsorption/desorption of hydrogen, whereas the so-called unusual states would correspond to the adsorption/ desorption of anions. 

Since after adsorption of MM the displacement of the preadsorbed water was observed as change in the intensity of DTG desorption peaks for water about 353 K and DMDS 473 K (Fig. 31), and following the assumption that either H2O or DMDS are adsorbed only in pores smaller than 5 nm, the data was normalized based on the volume of these pores. Fig. 32 shows the relationship between the normalized amount of DMDS and water. The correlation coefficient and slope are equal to 0.89 and 0.99, respectively. The slope represents the density of DMDS (1.06 g cm. The small discrepancy is likely related to the fact that not all pores are filled by oxidation products owing to the existence of some physical hindrances (blocked pore entrances). Tte thin line represents theoretical limit of adsorption assuming real densily of DMDS and H2O. The fact that almost all points are located below this line validates hypothesis about the active" pore volume [7--10]. All points used for this correlation represent equilibrium data If equilibrium conditions, ftir instance for adsorption of water, are not fulfilled tte amount of DMDS is usually small and the point moves ikim the established dependence line. 

Fainerman and Miller [35] found that displacement of an initially adsorbed surfactant by a second, more surface-active species allowed measurement of the desorption rate of the former. For example, competitive adsorption of sodium decyl sulfate and the nonionic Triton X-165 gave a desorption rate constant for the former of 40 s". Mul-queen and coworkers [36] recently developed a diffusion-based model to describe the kinetics of surface adsorption in multicomponent systems, based upon the Ward-Tor-dai equation. Experimental work with a binary mixture of two nonionic alkyl ethoxy-late surfectants [37] showed good agreement with the model, demonstrating a similar temporal adsorption profile to that found by Diamant and Andehnan [34],... 

Both these effects displace the isotherms of adsorption and desorption towards smaller relative pressure as compared with the system of unrelated pores. Hence it follows l) the pore size distribution, calculated in the frameworks of unrelated pores model, gives the decreased values of pore radii 2) the distribution obtained on the bases of adsorption isotherm, differs from the distribution obtained on the bases of desorption isotherm. Cooperative effects can be taken into account by means of network models, reflecting the special features of the pore structure more fully, than a system of unrelated pores. 

---