doi:10.1016/j.chroma.2006.03.112 
Copyright © 2006 Elsevier B.V. All rights reserved.
Effect of temperature on the retention of ionizable compounds in reversed-phase liquid chromatography: Application to method development
Sabine Heinisch
, a, 
, Guillaume Puya, Marie-Pierre Barriouleta and Jean-Louis Roccaa
aLaboratoire des Sciences Analytiques (CNRS UMR 5180), Université Claude Bernard, 43 Bd du 11 Novembre 1918, 69622 Villeurbanne Cedex, France
Received 6 December 2005;
revised 28 March 2006;
accepted 29 March 2006.
Available online 27 April 2006.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
The analysis of pharmaceutical compounds is often a difficult challenge which requires mathematical tools to improve the quality of the separation method. This work is an attempt to rationalize the anomalous variation of the logarithm of the retention factor with temperature in case of ionizable compounds. The effect of temperature on ionizable compounds was studied within a large range of temperature, ranging from 30 to 130 °C. The determination of the so-called chromatographic pKa and the study of its variation with temperature allow to explain why the forms of the van’t Hoff curves are so different depending on the type of solute, the type of buffer and the type of the mobile phase. A retention model along with a computation procedure is proposed to optimize both temperature and mobile phase composition and to provide good and robust conditions as shown by illustrative examples.
Keywords: Liquid chromatography; Method development; High temperature; Ionizable compounds; Retention behaviour
Fig. 1. Effect of the type of solute on the variation of 
with the reciprocal of temperature. Benzoic acid with citrate–acetonitrile 80–20 (v/v) (●); N,N-dimethylaniline with citrate–acetonitrile 80–20 (v/v) (
); codeine with phosphate–acetonitrile 80–20 (v/v) (♦); chloroprocaine with phosphate–acetonitrile 80–20 (v/v) (■).
Fig. 2. Variation of 
vs. 1/T for amitriptriptylline with different buffer–acetonitrile compositions: 70–30 (v/v) (●); 60–40 (v/v) (); 50–50 (v/v) (♦); 40–60 (v/v) (■); 30–70 (v/v) (×). Phosphate buffer (a) and Tris buffer (b).
Fig. 3. Variation of the dissociation rate vs. 1/T for amitriptylline with different buffer–acetonitrile composition: 70–30 (v/v) (●); 60–40 (v/v) (); 50–50 (v/v) (♦); 40–60 (v/v) (■); 30–70 (v/v) (×). Phosphate buffer (a) and Tris buffer (b).
Fig. 4. Variation of the experimental log(k) vs. 1/T for different solutes on a Nucleodur gravity C18 column: diphenhydramine (■); chloroprocaine (♦); quinine (); codeine (●); phenol (×). 
6.2; phosphate 15 mM–acetonitrile 70–30 (v/v).
Fig. 5. Variation of the experimental log(k) vs. 1/T for different solutes on a Hypercarb column: diphenhydramine (■); codeine (●); phenol (×). 
6.2; phosphate 15 mM–acetonitrile 70–30 (v/v).
Fig. 6. Variation of the experimental log(k) vs. 1/T for imipramine on a RP-Xterra C18 column with a buffer–acetonitrile 50–50 (v/v) mobile phase at different 
: 6.0 (♦); 7.0 (■); 8.0 (); 9.1 (●). Phosphate buffer (a) and Tris buffer (b).
Fig. 7. Response surface vs. 1/T and percentage of acetonitrile for the separation of six pharmaceutical compounds (codeine, phenol, quinine, chloroprocaine protriptylline, diphenhydramine) on a RP-Xterra C18 column. Conditions are given in Table 2.
Fig. 8. Simulated (a) and experimental (b) separations for the optimum conditions of Fig. 7. Uracile (1), codeine (2), phenol (3), quinine (4), chloroprocaine (5), protriptylline (6) and diphenhydramine (7).
Fig. 9. Response surface vs. 1/T and percentage of acetonitrile for the separation of six pharmaceutical compounds (N-clozapine, amoxapine, imipramine, benzene, amitriptylline, clozapine) on a RP-Xterra C18 column. Conditions are given in Table 2.
Fig. 10. Comparison of experimental separations of six solutes with two sets of different conditions given by Fig. 9: 30 °C, 30% acetonitrile and 1 mL/min (a) and 76 °C, 45% and 3 mL/min (optimum conditions) (b).
Fig. 11. Experimental separations with conditions corresponding to the four corners of the required robustness window for the optimum separation of Fig. 10: 43% and 78 °C (a); 43% and 74 °C (b); 47% and 78 °C (c); 47% and 74 °C (d). Flow-rate: 3 mL/min.
Table 1.
Predicted vs. experimental separation of basic compounds at different column temperatures (see text for additional explanations)
Table 2.
Experimental conditions for the examples of optimization given in Fig. 7 and Fig. 9
Abstract
Purchase PDF (434 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (11)
Purchase PDF (512 K)
