Mobile phase layer thickness effects

Evidence concerning the identity of the mobile species can be obtained from observation [406,411—413] of the dispositions of product phases and phase boundaries relative to inert and immobile markers implanted at the plane of original contact between reactant surfaces. Movement of the markers themselves is known as the Kirkendall effect [414], Carter [415] has used pores in the material as markers. Product layer thickness has alternatively been determined by the decrease in intensity of the X-ray fluorescence from a suitable element which occurs in the underlying reactant but not in the intervening product layers [416]. 

The particle size and size distribution of adsorbents for preparative purposes are higher and wider, respectively, compared to analytical ones. In addition, the adsorbent layer is much thicker and effectively overloaded with the compoimds. These items make resolution difficult, which must even be better than for quantitative separations as discussed in Section 5.1. These facts necessitate an excellent and superior strategy to hud the best separation, i.e., the mobile phase with the best selectivity (see Chapter 4). It was also shown that plates with a thickness gradient, called Uniplate-T taper plate [5], could improve resolution in the lower-Mp range. 

Single linear developments are mostly employed in the vertical mode. The apph-cabihty of the horizontal mode is discussed in Chapter 6. For circular and anticircular developments, the movement of the mobile phase is two-dimensional however, from the standpoint of sample separation it is a one-dimensional technique. Circular developments result in higher hRp values compared to linear ones imder the same conditions, and compoimds are better resolved in the lower-AR range. The same effect is noticed on plates with a layer thickness gradient (see Section 5.2.1). On the other hand, using antieircular developments, compounds are bettCT resolved in the upper-M range. 

Both the absorption and the resonant tunneling experiments find quantization effects for layer thicknesses of 50 A or less. It is, however, not immediately obvious why the quantum states should be observed even in these thin layers. The discussion of the transport in Chapter 7 concludes that the inelastic mean free path length is about 10-15 A at the mobility edge. The rapid loss of phase coherence of the wavefunction should prevent the observation of quantum states even in a 50 A well, but there are some factors that may explain the observations. The mean free path increases at energies above the... 

One zone is normally kieselguhr, 3 cm long and 150 pm thick, which has comparatively poor ad-sorptive properties. Thus, any size of spot placed on tiiis layer and run in the mobile phase will become a sharp band before it gets to the analytical silica gel layer. Anotiier form of plate for special applications is one with a pre-concentration zone of octadecyl-silica and an analytical layer of silica. These plates simplify sample application and improve sensitivity, but are very expensive compared with conventional plates. Approximately the same effect can be obtained using conventional plates and running them first in methanol for 0.5 cm. This converts all the spots to thin bands which can then be run in the solvent of choice. 

The concentration of the electrolyte used in the mobile phase affects the value of the zeta potential and hence the flow. Knox et al. (13] investigated the effect of NaNOv concentration and pH on the zeta potential using ODS particles and found that it decreased at lower pHs and higher electrolyte concentration. From these results it would appear that a 0.001-0.01 M concentration range is the most acceptable for CEC. Very dilute solutions would give better zeta potentials, but would increase the thickness of the double layer and limit the particle size to a minimum of 1-2 pm 14(. Wan 14 extended Rice and Whitehead s theoretical model (15] of EOF in an open tube to predict the double-layer overlap effects in packed columns. The results published agreed with Knox s earlier work, the main conclusion being that electrolyte concentration has a major influence on EOF for low values of particle diameter and inter-particle porosity. 

As noted earlier in this chapter, the apparent Km values of immobilized enzymes vary with the thickness of the diffusion layer surrounding the particles. In packed-bed enzyme reactors, the thickness of this layer varies with the mobile phase flow rate. Faster flow rates produce smaller diffusion layers and therefore K m values that more closely approximate the true Km of the enzyme. This effect has also been observed with the ficin-CM-cellulose reactor, and plots of K m against flow rate Q obtained at different mobile phase flow rates are shown in Figure 4.14. 

Each particle in a bed of porous particles is surroimded by a laminar sublayer (Figure 5.4), through which mass transfer takes place only by molecular diffusion. On one side, this layer is exposed to the flowing mobile phase and is entirely accessible. On the other side, it wraps the particle wall and is accessible from the particle inside only at the pore openings. The thickness of this layer, hence the mass transfer coefficient, is determined by hydrodynamic conditions and depends on the flow velocity. The mass transfer rates can be correlated in terms of the effective mass transfer coefficient, fcy, defined according to a linear driving force equation ... 

Microporous particles (3-10 (im) give columns that are as much as 20 times as efficient as porous layer-bead or pellicular (40 pm) packings. Whilst modem LC is based almost exclusively on microporous packing materials it is informative to relate the advances in particle design with the attempts to eliminate the deleterious effects on column performance since the latter as expressed by H is related to experimental variables, such as, the particle size (dp), the nominal stationary phase thickness (ds) and the mobile phase velocity (u). 

In contrast to RP chromatography, water is the stronger solvent in HILIC mobile phases, where it is usually contained in concentrations of 1-30%. Hence it forms an adsorbed layer on the surface of a polar adsorbent, which is thick enough to induce liquid-liquid partition between the bulk mobile phase and the adsorbed aqueous liquid layer. Consequently, the retention in HILIC is often due to the combination of adsorption and liquid-liquid partition mechanisms with possible additional effects of ion exchange (especially on bonded amino, zwitterionic, or weak ion exchange colunms). The transition between the adsorption and partition retention mechanisms is probably continuous as the water concentration in the mobile phase gradually increases. 

The resulting plates are scanned (off-line) to find the limit at which resolution becomes unsatisfactory. From these experiments the maximum amount of sample for the on-line preparative separation can be predicted, taking the particle size and the volume of the stationary phase into account. In analytical U-RPC and M-RPC the separation distance is 8 cm and the average particle size 11 pm in preparative U-RPC and M-RPC the separation distance is increased by 25% but the particle size is approximately 30% larger. These adverse effects practically cancel each other, so only the layer thickness has to be considered in the scale-up procedure. In our experience, therefore, a factor of 20 is generally appropriate (73). The flow rate of the mobile phase has to be adapted to preparative separation, so that the migration of the a front is as fast as in the analytical separation. 

Stationary phase S = silica gel G impregnated with 3% Primene JM-T So = silica gel impregnated with 5% Primene JM-T Sj = silica gel impregnated with 0.15% Primene JM-T. Mobile phase M, = 0.05 M citric acid M2 = 0.1 M citric acid, M, = 1.0 M citric acid. Conditions Ascending technique, layer thickness 0.5 mm, loading volume 5 /ul of 2.5% metal ion solution. Remarks Investigation of effects of eluent and impregnant concentrations on the Rf values of metal ions. An inverse correlation between Ry-values and the %E (E = extraction) of the metal ions is observed. S1-M2 is identified as the best system. 

---