Capacity factor, optimization

Figure 9-17 plots flood capacity versus flow parameter. The FP values of 0.4-0.7 are estimated by Kister, el al. [136] in absence of data. The plots show that for low and moderate pressures the flood capacity factor versus FP correlates the effects of liquid rate and pressure on the optimized tray capacity [136]. At higher pressures an additional effect of pressure on capacity shows a decline of optimized tray capacity. 

For optimization of chromatographic separations the ratio of the time spent by the solute in the stationary phase to the time it spends in the mobile phase is more fundiwentally i tortant. This ratio is called the solute capacity factor and is given by equation (1.8)... 

Any optimization strategy that considered only efficiency is inadequate to describe accurately resolution, which is a strong function of the capacity factor at low capacity factor values. The... 

Solvent optimization in reversed-phase liquid chromatography is commenced by selecting a binary mobile phase of the correct solvent strength to elute the seuaple with an acceptable range of capacity. factor values (1 < k <10 in general or 1 < k < 20 when a larger separation capacity is required). Transfer rules (section 4.6.1) are then used to calculate the composition of other isoeluotropic binary solvents with complementary selectivity. In practice, methanol, acetonitrile and tetrahydrofuran are chosen as the selectivity adjusting solvents blended in different... 

Figure 4.37 Preparative-scale separation of bilirubin laomarm by high pressure liquid chronatography. The analytical separation was optimized to maximize the separation factor at the smallest practical value for the capacity factor and then the saiq>le size scaled-up to that allowed by the larger amount of packing in the preparative column. (Reproduced with permission from Perkin-Elmer Corporation).

The isocratic reversed phase solvent system consists of water (polarity, p = 10.2), the most polar solvent in RPLC, as a primary solvent to which water-miscible organic solvents such as methanol (p = 5.1), acetonitrile (p = 5.8), or tetrahydrofuran (p = 4.0) are added. In order to optimize the speed of separation for an analyte pair, the proportions of water to nonpolar solvent are chosen such that the capacity factor of the last-eluting analyte of interest has a value of about 2.13... 

Other than selecting the column and mobile phase for the correct mode of separation, optimizing different HPLC parameters (injection volume, run time, wavelength, and detector) is equally important for achieving acceptable capacity factor (k ), resolution ( R), and tailing factor (T). 

Other results obtained from the ruggedness test are the definition of optimized method conditions for the factors and of system suitability criteria for a number of responses. System suitability parameters [6,17] are defined as an interval in which a response can vary for a rugged method. The system suitability criteria are the range of values between which a response (e.g. retention time, capacity factor, number of theoretical plates, resolution) can vary without affecting the quantitative results of the analysis. For instance, a design is performed and the retention time of the main substance varies between 200 s and 320 s without affecting the quantitative determination of the substances. The system suitability criteria for the retention time is then defined as the interval 200 s - 320 s. 

However, for very highly retained solutes direct measurement of the capacity factor ks is not possible, and this parameter must be predicted on the basis of retention data determined with a stronger mobile phase. The determination of VB is an essential step in the optimization of trace enrichment and clean-up procedures. 

Obviously, there are many ways to influence the capacity factors. However, the effects described above are predictable (see section 4.2.3) and in a sense trivial. It is worth noticing at this point that certain parameters do not at all affect the capacity factor and therefore do not at all affect chromatographic selectivity. These parameters include column length, flow rate and the diameter of packed columns. This renders these parameters irrelevant to the selectivity optimization process. In some cases they may be considered as parameters... 

The optimal working range in terms of capacity factors will not only be determined by the considerations of analysis time given in the previous section, but also by the number of peaks present in the chromatogram. The theoretical peak capacity (np of a chromatogram can be found from [104]... 

Once we have realized optimum capacity factors and optimized the selectivity, we can use the resolution equation to calculate the number of plates that is required to achieve baseline resolution (11,= 1.5). The required number of plates will to a large extent determine the kind of column and instrumentation needed to perform the separation. This will be briefly discussed in chapter 7. 

Where possible, we will derive simple relationships between retention and the relevant parameters. For reasons of clarity, we will express all equations in terms of the capacity factor (k). Obviously, the simplest possible equations will be most useful for optimization purposes. Ideally, we will be looking for linear relationships, since straight lines allow straightforward interpolation. 

Clearly, for the purpose of selectivity optimization, the capacity factor (fc) is greatly to be preferred to the retention index (I). 

To find a compromise for a mobile phase with neither too large a chain length (because of slow equilibration) nor too high a modifier content (because of the suppression of ionization), but yet optimum capacity factors and stable operating conditions is an optimization problem on its own. 

The capacity parameters allow a variation of the capacity factor (and hence the resolution) independent of the selectivity. However, all these parameters are difficult to vary, since they almost always require new columns to be used. Moreover, the range of variation offered by these parameters is too limited for them to be generally useful in optimization schemes (see also section 4.2.3). 

---