PQRI Approach for Selecting Columns of Equivalent Selectivity

When replacing a used column for a particular reversed-phase chromatographic (RPC) separation, two problems may arise. First, that particular column may not be readily available (for any reason), or second, a new column of the same designation (e.g., Waters Symmetry C18) may not result in the same (or acceptable) separation. In either case, the chromatographer needs to be able to select a different column (e.g., Phenomenex Luna C18) that will be equivalent to the original column.

The PQRI procedure for choosing an equivalent column that can be substituted for the original column was developed over a 10-year period from 1998 to the present time. 1 2 On the basis of almost 3000 retention-time measurements for 150 different test-compounds and 100 different alkylsilica columns, it was possible to show that five column properties completely account for column selectivity:

Column hydrophobicity Measured by the Column Parameter H

Column steric interaction

" "


Column hydrogen-bond acidity

" "


Column hydrogen-bond basicity

" "


Column ion-exchange capacity

" "


Values of these parameters do not vary much with other column conditions, except that C depends on mobile phase pH. Therefore, it is necessary to specify whether a high-pH or low-pH mobile phase is used for the separation. The USP program allows a choice of pH = 2.8 (low) or 7.0 (high), whichever value is closest to the actual mobile phase pH. Column selectivity for carboxylic acids is measured by the parameter B, while selectivity for basic compounds is measured by C. If it is known that the sample to be separated contains acids, it is necessary to check the "acids present" box in the program. Similarly, if the sample contains bases, the "Bases present" box should be checked. If it is not known whether acids or bases are present in the sample, both boxes should be checked.

When two columns are compared by the USP-PQRI procedure, their similarity is indicated by a quantity F (shown in the second column of values). After the original column is entered into the "Select column for comparison" box, similar columns will be selected by the program from a database of 368 different columns, and displayed in order of decreasing similarity. Columns which have values of F ≤ 3 are very likely to give an equivalent and acceptable separation for any sample. When the original separation is relatively "easy," as indicated by widely separated peaks (resolutions Rs » 2), acceptable separation on the replacement column may result for values of F > 3. In any case, the column with the smallest value of F is most likely to provide a similar and adequate separation of the sample.

Previous collaborative studies have shown that our column parameter values (H, S*, etc.) are repeatable when measured by different laboratories,3 and values of F appear effective for choosing equivalent columns.4 Further verification of the latter is desirable, so users of the present USP program are urged to share their experience with this program.


  • L. R. Snyder, J. W. Dolan and P. W. Carr, J. Chromatogr. A, 1060 (2004) 77-116. ("The "Hydrophobic-subtraction" Model of Reversed-phase Column Selectivity")
  • L. R. Snyder, J. W. Dolan and P. W. Car, Anal. Chem., 79 (2007) 3255-3261. ("A New Look at the Selectivity of Reversed-phase HPLC Columns")
  • L. R. Snyder , A. Maule, A. Heebsch, R. Cuellar, S. Paulson, J. Carrano, L. Wrisley,, C. C. Chan, N. Pearson, J. W. Dolan and J. Gilroy, J. Chromatogr. A, 1057 (2004) 49-57.
  • J. W. Dolan, A. Maule, L. Wrisley,, C. C. Chan, M. Angod, C. Lunte, R. Krisko, J. Winston, B. Homeierand, D. M. McCalley and L. R. Snyder, J. Chromatogr. A, 1057 (2004) 59-74.