Selecting Equivalent Columns

Routine RP LC assay procedures are often carried out over periods of months or years, as well as in different laboratories and different parts of the world. During the application of such a procedure over time, many columns may be required. For various reasons, it may prove difficult or impossible to obtain a replacement column with sufficiently similar selectivity from the original source. In such cases, it is necessary to locate an equivalent replacement column from a different source - or a column of different part number from the same source. During the 

The need to find a column that is of similar selectivity as an original column is therefore fairly common. 

A column-comparison fimction based on values of H, S, etc. for columns 1 and 2 has been derived [12]  

Hj and H2 refer to values of H for columns 1 and 2, respectively (and similarly for values of S and S2, etc.). can be regarded as the distance between two columns whose values of H, S, etc. are plotted in five-dimensional space, with weighting factors (12.5,100, etc.) added for a sample of average composition. It was found [12] that if 3 for two columns 1 and 2, variations in a should be 3%, so that the two columns are likely to provide equivalent selectivity and separation for different samples and conditions. 

An example of the application of Eq. (3) is shown in Fig. 2, for the separation of a mixture of neutral, basic, and acidic compoimds on four different columns. Values of Fj from Eq. (3) are shown for the three columns of Fig. 3b-d, each of which is compared with the Discovery Cg column of Fig. 3a. The values of F for the ACE Cg (b) and Precision Cg (c) columns are relatively small (Fg 4), and separation on these columns is therefore expected to be (and is) quite similar to that on the Discovery Cg column. For the Inertsil Cg column, F = 38, indicating that this column has a selectivity that is very different from that of the Discovery Cg column, which indeed it does note, for example, the co-elution of bands nos. 1/2 and 8/9 in Fig. 3d. 

---

Figure 3.25. RPLC column selectivity chart. X-axis is an indication of the hybrophobicity and Y-axis is an indication of the silanophilic activity of the bonded phase. This comparative chart is useful for selecting equivalent or dissimilar columns from various vendors. Diagram courtesy of Waters Corporation.

Most of the stationary phase types in Table 2.8 were derived from popular packed column stationary phases and subsequently modified to allow for immobilization and improved thermal stability. Their selectivity equivalence can be ascertained by comparison of system constants for stationary phases of nominally similar chemical composition in Table 2.6 and 2.8 [125]. There are small differences in selectivity for... 

The sample of Fig. 2 was chosen from the test solutes used to measure values ofH, S, etc. according to Eq. (3). A better test of Eq. (3) and values ofH, S, etc. for measuring column selectivity has recently been reported [18]. Twelve different routine RP LC separations were selected from three different pharmaceutical laboratories, whereupon (successful) attempts were made to select equivalent replacement columns for each separation on the basis of Eq. (3). One of these separations is illustrated in Fig. 3, where the original separation on an ACE Cg column (a) is compared with that on three other columns (b-d) with 1.3 < F < 248. The separation with the Discovery Cg column (b) is seen to be reasonably similar to that in (a), as expected from its value of Fg = 1.3. However, separation on the... 

Solid-phase extraction (SPE) cartridge, LC Si 2-g, 12-mL (Supelco), or equivalent ABC Laboratories Model SP 1000 gel permeation chromatograph system equipped with a 2.5 x 32.0 cm glass column of Bio-Beads S-X3 Select 200-400 mesh (ca 50 g, Bio-Rad Laboratories) preconditioned with ethyl acetate-cyclohexane (1 1, v/v), or equivalent... 

To evaluate compound solubility, a /.iPLC system equipped with a cartridge containing 24 parallel columns (80 x 0.5 mm (inner diameter equivalent)) was employed. Sets of calibration standards were prepared for 24 compounds at different concentrations (in a 50 50 CH3CN H20 solvent). A maximum standard concentration of 500 jt/M was selected to maintain the amount of DMSO co-solvent in all samples and standards below 5% v/v to minimize possible solubility enhancements due to the presence of DMSO when working with stock solutions provided at 10 mM in DMSO. Standards were added to the appropriate wells of a 384-well plate. The plate was covered with a heat seal foil and transferred to the /./PI.C system for analysis. Figure 6.26 depicts the process for preparation of standards 95 /./I. of a buffer of desired pH were added to the appropriate wells. An additional 5, uL of each compound at a concentration of lOmM (in DMSO) was added to the corresponding wells. The plate was shaken for 90 min and centrifuged at 4000 rpm for 3 min. 

If two rows are equivalent under criterion 1, select the row that has a nonzero element in the column for which the total (vertically) of nonzero elements is the greatest. 

As with other types of solutes, chromatographers have attempted to improve the speed and efficiency of analysis of bases by fhe use of smaller particle (e.g., sub-2 p,m) or monolithic columns. Small particle columns have not yet been fully evaluated for the analysis of bases, to determine whether they give equivalent selectivity, and reduction in plate height commensurate with the reduction in particle size, as has been demonstrated for neutral compounds. Commercial silica monolith columns give reasonable performance for the analysis of bases at low pH, but show evidence of... 

---