Klára Valkó

Retention in reversed-phase liquid chromatography as a function of mobile-phase composition

The dependence of solute retention (k′) on mobile phase composition (%B for binary-solvent mixtures A–B) is reviewed and compared with various empirical and theoretical equations that have been proposed for this relationship. Because the functional dependence of k′ on organic modifier composition varies from one system to another, it is not possible from these data to draw any overall conclusions as to the nature of the retention process in reversed-phase chromatography. Likewise, there is probably no one best equation for extrapolating all retention data to 0 %B for purposes of predicting log Po/w values from chromatographic data. The relative change in k′ with change in %B can be described in terms of the parameter S = −d(log k′)/dϕ (ϕ = 0.01 %B). Values of S as a function of solute structure, mobile-phase composition, column type and experimental conditions are of interest for several reasons: insight into the retention process or “mechanism”, mobile phase optimization, etc. Previous work relating to this question is reviewed here and some conclusions are presented.

New chromatographic hydrophobicity index (ϕ0) based on the slope and the intercept of the log k′ versus organic phase concentration plot

Abstract A new chromatographic hydrophobicity index (ϕ 0 ) is suggested as a measure of the lipophilic character of compounds in reversed-phase high-performance liquid chromatography (RP-HPLC). The parameter ϕ 0 is defined as the organic phase concentration (methanol or acetonitrile) in the mobile phase which is required for log k ′ = 0 (retention time is double the dead time), that is, the molar fraction of the compound is identical in the mobile and the stationary phases. The ϕ 0 values therefore range from 0 to 100%, and the higher the value the more hydrophobic is the compound. It is shown that the value of ϕ 0 is characteristic for a compound and depends only on the type of organic modifier, pH and temperature. It is independent of the RP column type and length, flow-rate and the mobile phase compositions where the actual retention measurements are carried out. The other advantages of ϕ 0 are that it can be precisely measured, as it has a concrete physical meaning, namely the organic phase concentration of the mobile phase at which the retention time is exactly double the dead time (not like log k ′ values extrapolated to water as mobile phase), and it is independent of the linear or quadratic function of the log k ′ versus ϕ 0 relationships. The ϕ 0 values not only reflect the hydrophobic character of compounds but also provide a valuable means for method development in RP-HPLC as they reveal a mobile phase composition with known retention time values. The ϕ 0 values for over 500 compounds were calculated and are presented on the basis of their published retention data. The ϕ 0 values obtained with methanol and acetonitrile showed an excellent correlation with each other. Significant correlations were found between the ϕ 0 values and the logarithm of 1-octanol—water partition coefficients (log P ).

Trace analysis of impurities in 3′-azido-3′-deoxythymidine by reversed-phase high-performance liquid chromatography and thermospray mass spectrometry

Abstract An analytical method has been developed for the detection of trace amounts of impurities in 3′-azido-3′-deoxythymidine referred to herein as AZT (Zidovudine). A sample extract was preconcentrated by normal-phase high-performance liquid chromatography (HPLC) with subsequent on-line reversed-phase HPLC-thermospray mass spectrometry (TSP-MS). During the sample extraction and concentration step, carried out by semipreparative normal-phase chromatography, the preliminary separation of the impurities from the AZT takes place. The organic solvent (dichloroethane-acetonitrile, 40:60) is evaporated from the collected fractions and the compounds are redissolved in a smaller volume of the reversed-phase mobile phases for a further degree of concentration. The collected fractions are then subjected to reversed-phase HPLC-TSP-MS. The influence of acetonitrile concentration and pH on the reversed-phase separation together with the sensitivity of the TSP-MS detection have been examined to maximise detection levels. The 3′-azido-3′-deoxy-5′-O-tritylthymidine, triphenyl methanol and 3′-chloro-3′-deoxythymidine, which are route-indicative impurities formed during the synthesis can be detected in the 50–100 ppb (w/w) range.

Two-dimensional high-performance liquid chromatographic method for assaying S-adenosyl-l-methionine and its related metabolites in tissues

S-Adenosyl-L-methionine (SAM) is a methyl-donor compound which is actively involved in a variety of biochemical reactions. An assay has been developed permitting the quantitative measurement of SAM and its related metabolites (S-adenosylhomocysteine, decarboxylated SAM, methylthioadenosine, adenosine and adenine) in liver and cell cultures. As gradient reversed-phase chromatographic or cation-exchange chromatographic methods often resulted in overlapping peaks, a two-dimensional high-performance liquid chromatographic (HPLC) procedure was developed involving gradient reversed-phase chromatographic separation followed by ion-exchange chromatography. After precipitating large molecules in the sample by perchloric acid, gel permeation was carried out on a Sephadex G 25 column to separate small water-soluble metabolites from proteins and membrane fragments. The freeze-dried sample was injected onto an ODS column and a 0-10% acetonitrile gradient in 10 mM ammonium formate buffer (pH 2.9) (20 min, linear) was applied. The relevant fractions were collected and injected onto a cation-exchange column (Partisil SCX, 10 microns, 250 mm x 4.6 mm I.D.). Elution and quantification were carried out using ammonium formate buffers of various concentration (15-400 mM), pH 2.9. The detector response (254 nm) as a function of concentration was linear over the concentration range 30-500 pmol. The detection limits of the compounds after the two-dimensional chromatographic procedure ranged from 10 to 60 pmol and the recovery was higher than 70%. The reproducibility of the results obtained from given samples was within 9-22% for rat liver and 6-24% for mast cells.

Lipidic peptides. XI: Quantitative structure-activity relationships of a series of lipidic amino acid conjugates of β-lactam antibiotics

Abstract A series of lipidic amino acid conjugates of β-lactam antibiotics were synthesized and the in vitro and in vivo activity was determined against a variety of Gram-positive and -negative bacteria. The chemical structures of the compounds were characterized by indicator variables showing the presence or absence of a substituent at a certain position, and by their lipophilicity as the calculated logarithmic value of octanol/water partition coefficient (clog P ). Stepwise linear regression analysis was applied for revealing quantitative structure-activity relationships (QSAR). The in vitro activities of the compounds did not correlate with in vivo activities indicating the influence of absorption processes. The in vivo oral and subcutaneous activities were influenced by the lipophilicity of the compounds. The optimum clog P values for high in vivo oral and subcutaneous activity were around 6 and 8, respectively. The presence of the long lipidic side chain at position 1 decreased the activity against E. coli and Pseudomonas aeruginosa . Loss of activity against the non-β-lactamase-producing S. aureus was observed for compounds with a lipidic side chain at position 2.

Effect of the eluent pH on the thermospray molecular ion intensity of nucleosides

Abstract The protonated molecular ion intensities of 15 nucleosides obtained by thermospray ionisation have been measured using 0.1 M ammonium acetate mobile phase at neutral and acidic pH. To explain the dependence of the molecular ion intensity on the mobile phase pH, the hydrophobicity, the pK a values and reversed-phase high-performance liquid chromatographic retention data (log k ′) at neutral and acidic pH values were studied. Significant correlations (above 95% probability level) were found between the change in the protonated molecular ion intensity and the hydrophobicity as well as the p K a values of the compounds. The reversed-phase chromatographic retention parameter (log k ′) obtained at pH 3.5, showed significant correlation together with the p K a values to the molecular ion intensity change caused by decreasing the mobile phase pH. None of the investigated nucleosides showed an increased molecular io intensity change at low pH when more than 5% methanol was present in the mobile phase.

Chapter 2 Retention prediction of pharmaceutical compounds

Publisher Summary This chapter discusses the retention prediction of pharmaceutical compounds. It focuses on those pharmaceutical compounds that are structurally unrelated. The retention prediction is most widely applied is reversed-phase high-performance liquid chromatography (HPLC). Several approaches for retention predictions are known. The most widespread measure of the degree of the retention is the capacity ratio or retention factor. The retention is expressed relative to the unretained component in the system. Therefore, the exact dead time determination is essential for the accurate determination of the retention factor. The retention time determination is usually carried out by an integrator or other data handling computer program. In HPLC, the retention of a compound is determined by the strength of the solute–stationary phase and solute–mobile phase interaction. The close relationship between the retention factor and the equilibrium constant allows extracting thermodynamic information from the chromatogram that can be used for understanding and predicting the retention.

Relationships between the chromatographic retention data and the effects of nucleoside derivatives in highly metastatic 3LL cells

The effect of 21 nucleoside derivatives on the [3H]-thymidine cellular uptake and on the incorporation into DNA of highly metastatic 3LL (Lewis lung carcinoma) cells has been measured. Hydrophobic and hydrophilic molecular parameters (the adsorption capacity, specific adsorption surface, lipophilicity and specific hydrophobic surface area) have been determined by using thin-layer chromatography. Stepwise linear regression analysis and principal component analysis have been applied in order to reveal the relationships between the molecular parameters and the effect of the nucleoside derivatives on highly metastatic 3LL cells. The first principal component obtained from the measured activity data could be attributed to the change of [3H]-thymidine cellular uptake caused by the nucleoside, while the second principal component could be regarded as the measure of the effect on the DNA incorporation of [3H]-thymidine. The effect of the nucleosides on the [3H]-thymidine uptake could be explained by the specific hydrophobic and adsorption surface area of the nucleoside, on the other hand the effect on the DNA incorporation could be described by the adsorption characteristics (specific hydrophilic surface area and adsorption capacity) of the derivatives.

Structure-retention relationships of diastereomeric mixtures of lipidic amino acid conjugates on reversed-phase stationary phases

Abstract The retention parameters of diastereomeric mixtures of 19 lipidic amino acid conjugates have been determined on Spherisorb ODS and Supelcosil LC-ABZ stationary phases by changing the acetonitrile concentration in acetonitrile-0.1% trifluoroacetic acid-water mobile phases. In general, better resolution of the diastereomeric mixtures was obtained on the Supelcosil LC-ABZ stationary phase but in some cases the Spherisorb ODS column showed better separation. The presence of silanol groups improved the separation of the diastereomers in those cases. The log k ′ values were plotted against the acetonitrile concentrations. In the case of basic derivatives parabolic relationships could be observed on the Spherisorb ODS stationary phase which contains free silanol groups, while always straight lines were obtained on the Supelcosil LC-ABZ column. The slope and the intercept values were calculated for the straight lines obtained. The octanol-water partition coefficients (log P ) and the dissociation constants (p K a ) of the compounds were also calculated and were related to the chromatographic parameters. It was found that the chromatographic parameters obtained on the Spherisorb ODS stationary phase showed slightly better correlations with the calculated hydrophobicity and p K a parameters than those obtained on the Supelcosil LC-ABZ column, which shows that different retention mechanism can be expected on the two stationary phases.