J. Chromatogr. A 1577, 101-108 (2018). In order to overcome the difficulty of the interpretation of the MS signals present at a low intensity for unknown degradation products or impurities, a new strategy and open-source software called eicCluster was developed. It offered unsupervised machine learning algorithms and powerful interactive visualization tools that made data processing fast and intuitive. The low-intensity HPTLC-HRMS signals were highlighted in a stressed formulation by using eicCluster. Thus, even compound ions present at low intensities were separated in subclusters from background signals (in silico highlighting). The respective preprocessing led to intensity-agnostic signals and the t-SNE algorithm clustered mass signals based on their similarity. The resulting 2D maps allowed a new view on the data set to such low-intensity target molecules in complex mixtures. Moreover, the targeted on-surface synthesis of degradation products (in situ highlighting) was shown to support a fast structure elucidation, when standards are not commercially available. It allowed a better understanding of the proposed degradation reactions in the formulation. Comparison with the results of stressed samples as well as the proposed degradation products of on-surface synthesis proved that in silico and in situ signal highlighting substantially eased structure elucidation and data processing.

Phytochem. Anal. 29, 452-462 (2018). HPTLC of apigenin(1), luteolin (2), kaempferol (3), quercetin (4), kaempferol‐3‐O‐glucoside (5), quercetin‐3‐O‐glucoside (6), isorhamnetin‐3‐O‐neohesperidoside (7) and rutin (8) in the fruits and aerial parts of Foeniculum vulgare (fennel), Pimpinella anisum (anise), Carum carvi (caraway), Cuminum cyminum (cumin), Coriandrum sativum (coriander), Apium graveolens (celery), Petroselinum crispum (parsley), and Anethum graveolens (dill) on silica gel with ethyl acetate – methanol – water – acetic acid 3:1:1:1. Quantitative determination by absorbance measurement at 254 nm. Heat maps and hierarchical clustering were performed for image processing. The hRF values for (1) to (8) were 97, 90, 83, 73, 30, 17, 7 and 2, respectively. LOD and LOQ were 54 and 166 ng/zone for (1), 52 and 157 ng/zone for (2), 58 and 177 ng/zone for (3), 37 and 112 ng/zone for (4), 57 and 172 ng/zone for (5), 54 and 164 ng/zone for (6), 55 and 169 ng/zone for (7) and 58 and 178 ng/zone for (8), respectively. The intermediate precision was <2 % (n=6). Average recovery was 98.4 % for (1), 92.3 % for (2), 95.3 % for (3), 96.8 % for (4), 94.5 % for (5), 95.6 % for (6), 95.1 % for (7) and 97.3 % for (8).

Phytochem. Anal. 30, 121-131 (2019). HPTLC bioautography of essential oils of oregano, rosewood and cumin on silica gel with ethyl acetate – petroleum ether 1:19 for cumin, ethyl acetate – petroleum ether 1:9 for oregano and ethyl acetate – petroleum ether 1:3 for rosewood. 2D resolution by eluting for a second time along the orthogonal axis. TLC-direct bioautography by spraying with a bacterial suspension of methicillin‐sensitive S. aureus NCTC 1298 (MSSA), methicillin‐resistant S. aureus NCTC 12497 (MRSA), E. coli NCTC 8003, ciprofloxacin‐resistant E. coli (clinical isolate), P. aeruginosa NCTC 6749, P. aeruginosa (clinical isolate), vancomycin‐sensitive Enterococcus faecium and vancomycin‐resistant E. faecium NCTC 12202, followed by incubation at 37 ºC for 4 h. Antimicrobial fractions were further characterized by NMR spectroscopy, GC-MS and LC-MS.

J. Planar Chromatogr. 31, 397-403 (2018). HPTLC of clozapine (1) and aripiprazole (2) on silica gel with toluene ‒ methanol ‒ ethyl acetate ‒ ammonia 65:25:10:1. Quantitative determination by absorbance measurement at 218 nm. The hRF values for (1) and (2) were 43 and 60, respectively. Linearity was between 200 and 1600 ng/zone for (1) and 100 and 800 ng/zone for (2). LOD and LOQ were 34 and 103 ng/zone for (1) and 20 and 60 ng/zone for (2), respectively. The intermediate precision was <2 % (n=3). Recovery ranged between 98.2-99.9 % for (1) and 98.7-99.5 % for (2).

(Decontamination of insecticides in natural drugs with supercritical carbon dioxide.) Planta Med. 50, 171-173 (1984). TLC of insecticides, i.a. DDT and HCH on silica with hexane. Detection by UV 254 and 366 nm. New "fluid extraction TLC" technique.

1. Separation on silica by thin-layer and high-performance liquid chromatography. J. Chromatogr. 328, 279-287 (1985). HPTLC of bis-quaternary aminosteroids (pipecuronium bromide and related steroids) on silica with methanol - acetonitrile - conc. NH3 514:386:100 containing 0.005 mole/l ammonium chloride and 0.08 mole/l ammonium carbonate. Detection by immersion of the plate in Dragendorff reagent. Densitometry by absorbance at 525 nm.

(Detection limits of some analgesics in blood and urine with TLC analysis using preadsorption columns Y29.) Pharmazie 40, 358-239 (1985). Preadsorption using Y29 resin followed by TLC on silica with chloroform - acetone 9:1; staining with potassium ferricyanide and with Dragendorff reagent; detection limits of the procedure for acetylsalicylic acid, phenacetine and similar drugs 0.4 to 18 mg/ml.

Nachweis von Verfäschungen durch DC und GC. Strukturrevision der Hauptkomponenten des aetherischen Öls. (Radix pimpinellae and its adulterations. Identification by TLC and GC. Revision of the structure formula of the main constituent in the essential oil.) Dtsch. Apoth. Ztg. 125, 399-402 (1985). TLC of essential oils on silica with hexane - ether 8:2. Detection by spraying with anisaldehyde reagent and heating for 5 min. at 120 °C, and TLC of coumarins on silica with toluene - ether 1:1. Detection with 10 % acetic acid and by UV 366 nm.