J Chromatogr A 1653, 462442 (2021). Samples were peptides obtained through tryptic hydrolysis of the 5 most abundant milk proteins: α-lactalbumin (α-LA), β-lactoglobulin (β-LG), α-, β- and κ-casein (CA). As standards, synthetic whey and pea (Pisum sativum, Fabaceae) peptides (selected based on the in silico tryptic digest of α-LA, β-LG, legumin A, and vicilin with one or zero miscleavages) were only used in the last assay for prediction of the RF values of peptides with known amino-acid (AA) sequences. Two-dimensional HPTLC on silica gel (pre-washed with methanol and activated 10 min at 100°), first with basic mobile phase sec-butanol – pyridine – ammonia – water 39:34:10:26, and (after 12h drying) in the orthogonal direction with acidic mobile phase sec-butanol – pyridine – acetic acid – water 11:8:2:5. Derivatization for peptides and proteins by immersion into fluorescamine (0.05 % in acetone); visualization under UV 254 nm and 365 nm. Computer-assisted determination of the x- and y-coordinates of the derivatized zones. Repeatability (n=8) of the 2D-HPTLC was statistically tested with the Kolmogorov-Smirnov test for normal distribution and with Dixon’s Q test for outliers. Relative standard deviation (RSD) for the RF values was 12.9 % for the first dimension (y-coordinates) and 16.5 % for the second dimension (x-coordinates). According to their higher intensity and sharpness, 15 – 20 detected zones from each protein hydrolyzate were selected, manually scraped from the derivatized layer, dissolved in formic acid solution (0.1 % in acetonitrile – water 3:2), mixed with an equal volume of matrix (dihydroxybenzoic acid 2 % in acetonitrile – water 3:7), crystallized on air on a ground steel target, before being desorbed by the laser beam of the MALDI-TOF-MS/MS (matrix-assisted laser desorption/ionization time-of-flight tandem mass spectrometry). Direct hyphenation of HPTLC to MS was not performed, to avoid zone diffusion during plate coating with the matrix and to circumvent the stronger binding of polar peptides on the layer.  The MS spectra were acquired in positive reflector mode in m/z range 340 – 4000 (10 – 2500 for fragments), using an external peptide as calibration standard. Identification of 51 from the 85 selected peptides according to AA sequences was performed, using software programs allowing m/z calculation of protein fragments and estimation of cleavage sites. Correlation of the retention behaviour of the peptides with their properties (molecular weight MW, isoelectric point IEP, charges, polarity) was tested with Student’s two-sided t-test after calculation of Pearson’s correlation coefficients. The correlation was significant with IEP, percentages of anionic AA and of non-polar AA; but not with the following properties: MW, percentages of cationic AA and of uncharged polar AA. Finally, based on the correlation results, regression formulas were found to calculate the x- and y-coordinates of any known peptide from the percentage of non-polar AA (or vice-versa). The prediction power of these formulas was verified by repeating the complete 2D-HPTLC-MS experiment with the standard peptides of whey and of peas, and measuring the absolute and relative deviations between the actual x- and y-coordinates and the predicted values. The absolute deviations were higher in the lower RF zones. The average, relative RF value deviations (range 22.1 – 25.7 %) were not different between whey and pea peptides.

J. Liq. Chromatogr. Relat. Technol. 44, 820-828 (2021). HPTLC of Gentiana extracts from in vitro cultures on silica gel with acetate - methanol - water 4:1:1 in sandwich mode. Detection under UV light at 254 nm and fluorescence at 312 nm (emission above 370 nm). Principal component analysis (PCA), hierarchical cluster analysis and the novel proposal—nonnegative PCA was performed to identify common and distinct features.

J Chromatogr Sci, 60 (4), 372-386 (2022). Investigation of the dual retention mechanism in TLC taking place on three stationary phases of different polarity (C-18, silica gel and DIOL)  using binary mobile phases composed of acetonitrile as the main component and water, or methanol as a modifier. The test analytes were 12 compounds of pharmaceutical importance and considerably different chemical structure, i.e. the imidazoline and serotonin receptor ligands, and their related compounds. Determination of retention of each analyte in each investigated chromatographic system in a wide enough range of the mobile phase composition, with volume fraction of the mobile phase modifier ranging from 0.10 to 0.90. Calculation of the exact turning point values as a proof of occurrence of the reversed-phase hydrophilic interaction chromatography (HILIC/RP) retention mechanism based on the multimodal retention model. Analysis of the dual retention mode with the use of the volume fraction of the mobile phase modifier, the total polarity and the total solubility models, allowing the dual (HILIC/RP) retention mechanism for the DIOL, C-18 and silica gel stationary phase to be confirmed. The observed retention mechanism was more complicated than the dual HILIC/RP one in the case of the DIOL stationary phase and acetonitrile/methanol mobile phase.

J. Chromatogr. Sci. 52 (9), 1095-1103 (2014). Investigation of the retention behavior for a series of amine neurotransmitters, their precursors, metabolites and structurally related drugs on RP-18, RP-8, RP-2, cyano and diol phases with mixtures of phosphate buffer and methanol. According to the computed lipophilicity of the neutral form of the investigated compounds the most lipophilic compound is dobutamine (log PN = 3.78), while the least lipophilic is norepinephrine (log PN = -0.14). Dobutamine was the most lipophilic compound on RP-18 (RM0 = 1.58) and CN phase (RM = 1.21). Norepinephrine on RP-8 phase was the least lipophilic one (RM0 = -0.70). A similar chromatographic behavior could be assumed for RP-18, RP-8 and cyano phases. The mean of the RM values and PC1/RM (scores of the first principal component) experimental lipophilicity indices showed a high correlation with the computed ones.

Chromatographia 27, 297-300, (1989). Study of the effects of acidity of the media on the adsorption of phenol and cresols. Investigation with TLC on silica, silica containing 3% of active carbon and silica containing 6% of active carbon with various solvent systems, using Rf values as a measure of the adsorption content of the solvent mixture on the adsoption of phenol and cresols on the active carbon layer.

J. Planar Chromatogr. 3, 417-421 (1990). Retention versus eluent composition relationships of 36 drugs and poisons have been determined using silica gel and 20 binary eluents - mixtures of four weakly polar diluents (heptane, chloroform, dichloromethane, and trichloroethylene) and four polar modifiers (MEK, ethyl acetate, ether, diisopropyl ether) or mixtures of the four polar modifiers with methanol. In accordance with the Snyder-Soczewinski model of adsorption, linear relationships of the type RM = RM - m log cmod were obtained in almost all cases.

I. Basic aspects and relationship between slope and intercept of TLC equations. J. Chromatogr. 662, 341-361 (1994). Review with 50 references on the determination of lipophilicity by RP. TLC equation based on the linear relationship between RM values and organic solvent concentrations in the mobile phase. Study of the correlation between experimental and extrapolated RM values, and effect of the nature of the organic solvent on the value of lipophilicity. Discussion of the structural meaning of chromatographic congenerics of the compounds based on analysis of more than 700 TLC equations.

J. Planar Chromatogr. 9, 418- 424 (1996). Use of a simple equation to describe solute retention in different types of liquid chromatography with binary solvent systems. The displacement mechanism of solute distribution between mobile and stationary phases has been assumed and the effect of stationary phase composition on solute retention has been ignored. Two types of thin-layer chamber and column liquid chromatography technique were used in the experiments.