J Chromatogr A, 1602, 458–466 (2019). HPTLC of caffeine, physostigmine (alkaloids) and hydroethanolic extract of Peganum harmala seeds (Nitrariaceae, Zygophyllaceae) on silica gel prewashed twice with methanol – water 3:1, followed by 1 h drying at 120 °C. Separation, after 5 min chamber saturation, with ethyl acetate – methanol – ammonia (25%) 85:11:4 (basic mobile phase) or ethyl acetate – toluene – formic acid – water 16:4:3:2 (acidic mobile phase, requiring neutralization with phosphate-citrate buffer). Derivatization with Dragendorff’s reagent and with anisaldehyde sulfuric acid. Effect-directed analysis by spraying A) with Gram-negative bioluminescent Aliivibrio fischeri suspension for antibacterial activity (caffeine was used as standard); B) with acetyl- and butyryl-cholinesterase (AChE / BChE) solutions for enzymatic inhibition. For AChE and BChE asssays, classical immersion into the enzyme solutions was also used for comparison, and inhibition densitometry for active analytes was performed by inverse scan measurement (fluorescence without optical filter) at 546 nm using a mercury lamp; activity was expressed as physostigmine equivalents. Active bands were eluted (only after basic MP) with methanol through the oval elution head of a TLC-MS interface pump, into a quadrupole-Orbitrap mass spectrometer. Full scan mass spectra (m/z 50−750) in positive ionization mode were recorded using heated electrospray ionization (HESI, spray voltage 3.5kV, capillary temperature 270°C). By comparison to literature, AChE inhibitors (also active against A. fischeri) were assigned to be harmine, harmaline and ruine (β-carboline alkaloids), and BChE inhibitors were harmol (same class) and vasicine and deoxyvasicine (quinazoline alkaloids, also called peganine and deoxypeganine). Piezoelectric spraying had the following advantages over automated immersion: (1) it covered the whole plate surface; (2) required much lower volumes of solutions; (3) applied always fresh enzyme or reagent solutions, thus avoiding gradual inactivation; (4) avoided zone distortions, shifts or tailings occurring during immersion or withdrawal of the plate, or due to the hydrophilicity of compounds.

J Chromatogr A, 1629, 461511 (2020). HPTLC of methanolic extracts of female inflorescences from ten hemp varieties (Cannabis sativa, Cannabaceae) on silica gel with toluene – ethyl acetate 1:1 or (for yeast assays) on RP-18W with toluene – ethyl acetate 7:3. When intended for MS experiments, layers were previously washed twice with methanol – formic acid 10:1, once with acetonitrile – methanol 2:1 and air-dried. Chromatograms were documented under white light, UV 254 nm and for fluorescence detection (FLD) at 366 nm. Afterwards, 6 derivatization assays were performed with the following reagents, either without heating: primuline; or requiring heating 5 min at 120 °C: p-aminobenzoic acid; anisaldehyde sulfuric acid; diphenylamine aniline phosphoric acid; ninhydrin; vanillin sulfuric acid. Besides, 8 effect-directed assays (EDA) were performed for free radical (DPPH•) scavengers, for antimicrobial compounds (Gram-positive Bacillus subtilis assay, Gram-negative Aliivibrio fischeri bioluminescence assay), for phytoestrogens (planar yeast estrogen assay), for inhibitors of the following enzymes: acetyl-cholinesterase (AChE), α- and β-glucosidase, tyrosinase. AChE assay was performed by immersion (speed 3.5 cm/s, time 5 s) into AChE solution (666 units in TRIS buffer 0.05 M, with bovine serum albumin 0.1 %, pH 7.8), incubation 25 min at 37 °C, spraying with substrate solution, and heating 2 min at 50 °C. Two AChE substrate solutions were used: A) α-naphthyl acetate 0.1 % and chromogenic reagent Fast Blue salt B 0.18 % in ethanol – water 1:2, giving white inhibition bands visible on purple background under white light; B) with 3-indoxyl-3-acetate, giving black inhibition bands on blue background under UV 254 nm, which was useful to prevent false negatives when Fast Blue Salt B formed colored bands with analytes. Two bands of multipotent compounds were eluted from normal-phase layer with methanol through the oval elution head of a TLC-MS interface pump, into a quadrupole-Orbitrap mass spectrometer. Full scan mass spectra (m/z 50−750) in the positive and negative ionization modes were recorded using heated electrospray ionization (HESI, spray voltage 3.5 kV, capillary temperature 270 °C). By comparison to literature and standards, they were identified as cannabidivarinic acid (hRF 55) and cannabidiolic acid (hRF 60-70).

J Chromatogr A, 1616, 461434 (2020). Samples were acetonic extracts of Malus domestica fruit peels (Rosaceae) and of Salvia officinalis, Thymus vulgaris and Origanum vulgare spice powders (Lamiaceae), as well as standards of maleic acid (dicarboxylic acid), carvacrol, thymol (phenolic monoterpenes), rosmanol (phenolic diterpene), betulinic acid, corosolic acid (CA), maslinic acid (MA), oleanolic acid (OA) and its isomer ursolic acid (UA) (triterpenes). HPTLC on silica gel, when intended for MS and NMR experiments, layers were prewashed twice with methanol – water 3:1, followed by 30 min drying at 120 °C. When intended for quantitative densitometry, start zones were submitted to prechromatographic derivatization with iodine solution (10 g/L in chloroform) allowed to migrate up to 12 mm, incubated 10 min at 27 °C and dried under cold air stream; this allowed separation of isomeric triterpenes. Separation with toluene – methanol – ethyl acetate 17:2:1 after 5 min chamber saturation at 50 % relative humidity. CA coeluted with MA, and OA with UA. Four hyphenations: A) Quantitative HPTLC densitometry for active analytes was performed by measuring absorption at 665 nm with a tungsten lamp after immersion of the chromatograms in anisaldehyde sulfuric acid reagent and heating 5 min at 110 °C. Linear range was obtained at 25 - 200 ng/band for OA and 100 - 400 ng/band for UA. B) Effect-directed analysis by immersing the chromatograms into Gram-positive Bacillus subtilis suspension for antibacterial activity and into acetyl-cholinesterase and tyrosinase solutions for enzymatic inhibition. C) Active bands were eluted with methanol through the oval elution head and in-line filter frit of a TLC-MS interface pump, into a quadrupole-Orbitrap mass spectrometer. Full scan mass spectra (m/z 100−1000) in the positive and negative ionization modes were recorded using heated electrospray ionization (HESI, spray voltage 3.5 kV, capillary temperature 270 °C, probe heater temperature 200 °C). D) With higher amounts applied, preparative HPTLC, by scraping the multipotent band corresponding to OA and UA, and dissolving these analytes in methanol, for NMR analyses (1H raw or deconvoluted, and 2D 1H–13C Heteronuclear Single Quantum Coherence). Both isomers were distinguished by their allylic H-18 protons and separately quantified by applying PULCON method (PUlse Length-based CONcentration). LOQ was 267 μM for OA and 173 μM for UA; optimal range was 300 – 4600 mM, corresponding to 126 - 2090 μg of triterpenes.

J Chromatogr A, 1640, 461929 (2021). Study of the impact of different strains, culture media and parameters (temperature, time, rotational speed, and glucide and amino-acid supply) on the metabolite profile of bacteria. Samples were cultivation broths of Bacillus subtilis, B. licheniformis, B. pumilus and B. amyloliquefaciens, as well as their respective supernatant liquid-liquid extracts (apolar solvents only or QuEChERS method with acetonitrile and MgSO4 – NaCl mixture 4:1). HPTLC on silica gel (normal phase and RP-18), either as bands (for small volumes of extracts) or as areas for supernatants and bigger volumes of extracts. Extract areas were focused with a three-step procedure (up to 20mm with acetone, and twice with methanol); unextracted supernatants were focused twice with methanol and once with tetrahydrofuran, but the application zone of the plate had to be cut before development, due to the high matrix load. Development with ethyl acetate – methanol – water at different ratios after activation of the plate surface with magnesium chloride (33% relative humidity), evaluation in white light and UV. Detection of antibacterial compounds with Aliivibrio fischeri bioassay. Derivatization with primuline (for lipophilic substances) and diphenylamine aniline sulfuric acid reagent (for saccharides). This method allowed a fast comparison: A) of the patterns of the different strains (presence /absence and intensity of detected or antibacterial bands); B) of cultivation parameters: the number of metabolites increased with time, rotational speed (oxygen level), and at 37°C (vs. 30°C), whereas a minimal medium allowed the detection of more metabolites, due to the lower matrix load; C) of the impact of the extraction parameters: choice of the solvents (QuEChERS method had no advantage here), solvent – supernatant ratio (1:3 showed richer patterns than 1:1); D) of the HPTLC parameters used (better separation and resolution with normal phase vs. RP18 layers).

J Am Soc Mass Spectrom 31(9), 1981-1993 (2020). Low-temperature plasma-mass spectrometry was studied for comparison between direct desorption (DD) and diode laser assisted desorption (LD) in terms of quantitative and qualitative analysis of compounds from cellulose vs. silica gel TLC layers. Compounds (the 20 common amino acids, propofol, nicotine, cotinine, salicylamide, acetylsalicylic acid, paracetamol, caffeine, valprolactone and its isomer 4-ene-valproic acid) were applied on the TLC plates (without development) at different concentrations; a commercial mixture of acetylsalicylic acid, paracetamol and caffeine was also applied on TLC plates, developed with dichloromethane – ethyl acetate 1:50, detection at UV 254 nm and quantitative MS. In general, DD provided good results on cellulose, where LODs where between 0.01 and 2.55 ng/mm2, whereas several compounds remained undetected on silica gel. LD however provided LODs on silica gel from 0.3 to 84 pg/mm2. Tandem MS with collision-induced dissociation was implemented to improve signals, LODs and to characterize the other analytical figures-of-merits (including detection of the main fragment ions, determination of optimal laser beam width and irradiance depending on the compounds). For the two metabolites of valproic acid, the ions and fragments had identical values; therefore, a mix of the two isomers had to be applied and separated with dichloromethane – methanol 50:1 before MS; one half of the plate was visualized for control by dipping into potassium permanganate reagent (7.5g KMnO4, 50g K2CO3, 0.75g NaOH in 1L water).

J. Planar Chromatogr. 33, 51-57 (2020). HPTLCS (high-performance thin-layer chromatography scanning) of ethyl p-methoxycinnamate in the rhizome of Kaempferia galanga on silica gel with hexane - ethyl acetate 18:1. Quantitative determination by absorbance measurement at 309 nm. The hR_F value for ethyl p-methoxycinnamate was 29. Linearity was between 2 and 6 ng/zone. Intermediate precisions were below 2.4 % (n=3). Mean recovery rate was 100.1 %. Results were compared with a validated HPLC method and the HPTLCS method was more environmetnally friendly..

J. Planar Chromatogr. 33, 59-70 (2020). HPTLC of metformin (1), saxagliptin (2) and dapagliflozin (3) on silica gel with chloroform - methanol - water - acetic acid 740:260:50:1. Quantitative determination by absorbance measurement at 224 nm. The hR_F values for (1) to (3) were 14, 50 and 66, respectively. Linearity was between 30 and 350 µg/mL for (1), 140 and 1500 µg/mL for (2) and 50 and 600 µg/mL for (3).  Intermediate precisions were below 2 % (n=6). LOD and LOQ were 7 and 23 µg/mL for (1), 39 and 130 µg/mL for (2) and 14 and 47 µg/mL for (3), respectively. Recovery rate was between 98.9 and 100.5 % for (1), 99.2 and 100.5 % for (2) and 99.2 and 100.7 % for (3). Comparable results were obtained when compared with a HPLC method.

Food Chem. 288, 1-7 (2019). HPTLC of sesamin (1) and sesamolin (2) in sesame seed extracts on silica gel with n-pentane - diethyl ether 3:2. Quantitative determination by absorbance measurement at 290 nm. Linearity was between 25 and 150 µg/mL for (1) and 12.5 and 75 µg/mL for (2). Intermediate precision was below 3 % (n=6). The LOD and LOQ were 1.3 and 4 µg/mL for (1) and 0.4 and 1 µg/mL for (2), respectively. Recovery rate was  between 95 and 105 % for (1) and (2). Correlation of the HPTLC and a HPLC-PDA method was notably high.