Dr. Wilmer H. Perera is the Lab Manager at CAMAG Scientific, Inc. in Wilmington, NC and he is dedicated to the development of HPTLC methodologies that can be applied to the dietary supplement, food and cosmetic industries, and more to come. Dr. Michael Lelah and Mallory Goggans are with NutriScience Innovations, a dietary ingredient development and distribution company with headquarters in Milford, CT. Dr. David Bom is a consultant for NutriScience. NutriScience is the developer of ButyraGen™.

ButyraGen™ is a new dietary ingredient, a prebiotic direct butyrate generator [1]. The primary active ingredient, tributyrin (glycerol with three butyrate arms), is hydrolyzed in the body to the short chain fatty acid butyrate (butyric acid). Butyrate is a postbiotic involved in supporting digestive health through reducing gut permeability and also is an important gut signaling molecule for the gut-brain axis and other organ support [2]. Although tributyrin itself is an oil, ButyraGen™ is a spray-dried powder. This makes ButyraGen™ a hybrid – the material is a powder but it contains an oil. Dietary ingredients for use in dietary supplements manufactured under cGMP, require testing for identity, purity, strength and composition [3]. The identity test can also be used as a test for adulteration, which is a general concern for dietary ingredients and supplements. The identity test can help confirm whether an ingredient has been adulterated. HPTLC is widely used in the dietary supplement industry for the identification of botanicals, botanical concentrates and botanical extracts. The purpose of this study was to develop an HPTLC method for the identification of ButyraGen™ using the identification of tributyrin, the main active ingredient in ButyraGen™ (> 50% content) as the primary identification marker. The suitability of the method for this purpose was determined using tributyrin as a standard and also by comparing it against other fatty acids and lipids. Suitability is fit for purpose, which is the appropriate standard for the development of an identification method for a dietary ingredient [4]. Commonly used and inexpensive food fatty acids and oils are compared to determine if the method is sufficiently sensitive and specific to distinguish ButyraGen™ and tributyrin from these materials, which may be considered potential adulterants. Additionally, a negative control consisting of the other components of ButyraGen™ (without tributyrin) was evaluated to determine the effect of these other components in the product.

The use of HPTLC for the identification of oils is far less well known although methods for the determination of fatty oils have been developed [5]. Many manufacturers of dietary ingredients and dietary supplements have HPTLC instrumentation in their analytical labs and conduct identity testing of botanicals on a regular basis. Thus, HPTLC is an idealmethod to identify tributyrin in ButyraGen™ but it can be used for many other applications. The HPTLC PRO System boosts the applicability of the technique since it is a fully automated system where multiple samples can be analyzed in sequence, overcoming the environmental effects produced by the previous open system. HPTLC PRO also adds a more rigorous control of the gas phase and although still under development as an analytical tool, it will become a standard and powerful technique for advanced research and quality control purposes.

4.0 mg of tributyrin and triacetin are dissolved in 1.0 mL ofmethanol. The Universal HPTLC Mix (UHM) solution was prepared as described in literature [6] and used as system suitability test (SST).

ButyraGen™and ButyraGen™placebo (ButyraGen™ without the primary active tributyrin), glycerol monostearate, raw cocoa butter and palm kernel oil were prepared at 10.0 mg/mL methanol. 20.0 μL of medium chain triglycerides and linseed oil were dissolved in 980.0 μL of methanol and toluene, respectively. Samples were sonicated for 10 min at room temperature and centrifuged at 3000 rpm for 5 min as needed. The supernatant was used for further analysis.

HPTLC plates silica gel 60 F₂₅₄ (Merck), 20 x 10 cm were used.

10.0 μL of ButyraGen™and ButyraGen™placebo, glycerol monostearate, raw cocoa butter solutions, 2.0 μL of medium chain triglycerides, palm kernel oil and linseed oil solution while 40.0 μL and 20.0 μL of triacetin and tributyrin solutions, respectively, are applied as bands with the HPTLC PRO Module APPLICATION, 15 tracks, band length 8.0 mm, distance from the left edge 20.0 mm, track distance 11.4 mm, distance from the lower edge 8.0 mm. The first rinsing step (bottle 1 solvent) is done with methanol – acetonitrile – iso-propanol – water – formic acid 250:250:250:250:1 (V/V) and the second rinsing step (bottle 2 solvent) is done with methanol – water 7:3 (V/V).

Chromatography Plates were developed with the three developing solvents in the ADC 2 with activation of the plate at 33% relative humidity for 10 min using a saturated solution of magnesium chloride. LPDS (toluene, ethyl acetate 9:1 (V/V)) is used without saturation, whereas MPDS (cyclopentyl methyl ether, tetrahydrofuran, water, formic acid 40:24:1:1 V/V)) and HPDS (ethanol, dichloromethane, water, formic acid 16:16:4:1 (V/V)) are used with 20 min chamber saturation (with saturation pad). The developing distance for all three methods was 70 mm (from the lower edge). Plates were dried for 5 min.

The plate was immersed into primuline (0.05% in acetone – water, 4:1 (V/V)) using the Chromatogram Immersion Device 3, immersion speed 3 cm/s and immersion time 5 s, dried for 5min with cold air.

Images of the plate are captured with the TLC Visualizer 2 in UV 254 nm prior to derivatization and in UV 366 nm after derivatization.

The HPTLC analysis was qualified by using an UHM as system suitability test [6]. Three main quenching zones were observed in short wavelength UV 254 nm for the SST with R_F 0.11 ± 0.04, 0.21 ± 0.04 and 0.76 ± 0.04 in the figure below. The results are quite straightforward, ButyraGen™ (track 4) is identified by the tributyrin reference standard (track 3) and none of the other fatty acids tested should moved to this position. The other materials tested represent a range of food and other fatty acids which potentially could be used as adulterants to replace tributyrin in ButyraGen™. Tributyrin is a triglyceride with a glycerol backbone and three butyrate side chains. Triacetin is a triglyceride with a glycerol backbone and three acetate side chains. Glycerol monostearate is a long chain monoglyceride commonly used as a food emulsifier. The main constituent in cocoa butter is the triglyceride derived from palmitic, oleic and stearic acid. Cocoa butter also contains other unsaturated and saturated fatty acids. Medium chain triglycerides are triglycerides with two or three medium chain fatty acids. Palm kernel oil is high in saturated fats and lauric acid. Linseed oil (also known as flax seed oil) is high in unsaturated diglycerides and triglycerides, including alpha-linoleic acid.

HPTLC analysis of the UHM (track 1) in UV 254, triacetin and tributyrin (tracks 2 and 3), ButyraGen™ and ButyraGen™ placebo (tracks 4 and 5), glycerol monosterate, raw cocoa butter, medium chain triglycerides, palm kernel oil and linseed oil (tracks 6–10) in UV 366 nm post derivatization with primuline solution.

This method of HPTLC chromatographic separation is very specific for the different types of mono-, di-, and triglycerides indicating very good specificity for tributyrin and ButyraGen™. These results indicate the suitability (fit for purpose) of the method for the identification of tributyrin and ButyraGen™. Certainly, for the wide range of pure and naturally occurring complex fatty acid esters tested here, ButyraGen™ and tributyrin are completely and specifically distinguished. In the event that ButyraGen™ was to be adulterated with any of these products, this identity test method will be able to confirm the presence of such an adulterant. This indicates the method as suitable for confirming the presence of a variety of potential adulterants.

[1] https://nutriscienceusa.com/product/butyragen
[2] Canani R.B. et al. World J Gastroenterol 17(12) (2011) 1519-1528.
[3] FDA, Code of Federal Regulations, 21CFR111.70. https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/ cfcfr/CFRSearch.cfm?fr=111.70&SearchTerm=identity
[4] Wenclawiak B. et al. Quality Assurance in Analytical Chemistry (2010) 215-245.
[5] Identification of fixed oils, HPTLC Association https://www.hptlc-association.org/methods/methods_ for_identification_of_herbals.cfm
[6] Do T.K.T. et al. J Chromatogr A 1638 (2021) 461830.

Contact:

Dr. Wilmer H. Perera, CAMAG Scientific, Inc., 515 Cornelius Harnett Drive Wilmington, NC 28401, USA, wilmer.perera@camag.com

Dr. Michael Lelah, NutriScience Innovations, 130 Old Gate Lane, Unit C, Milford, CT 06460, mlelah@nutriscienceusa.com

Corinna Henninger is a Ph.D. student at the Karlsruhe Institute of Technology, under supervision of Adj. Prof. Katrin Ochsenreither. Her research focuses on the enzyme class of phytases, the analysis of the obtained degradation products, and the design of novel phytases using molecular biology. Her work is conducted at the Offenburg University of Applied Sciences, under co-supervision of Prof. Thomas Eisele, an expert in the field of enzyme production.

Phytases (IUBMB Enzyme Nomenclature: EC: 3.1.3.26) catalyze the stepwise dephosphorylation of phytate (myo-inositol-1,2,3,4,5,6-hexakisphosphate or InsP₆), the natural storage component of phosphate in plants. However, phytate shows poor digestibility in non-ruminant animals such as swine, poultry and fish due to their lack or low activity of InsP₆-hydrolyzing enzymes in the gastrointestinal tract. Therefore, phytases are utilized as a feed additive to release the bound phosphate. The analysis of myo-inositol phosphates (InsP_x) is challenging and time consuming, particularly in terms of separation and detection. However, when dealing with a large number of samples in the screening for phytases during protein engineering, having a fast and robust analysis method is crucial to reliably identify promising novel enzymes or target variants.

Considering high sample throughput and separation of all isomeric pools as well as free phosphate, HPTLC is most suitable as a fast and inexpensive screening method. Furthermore, the utilization of an enzyme assisted post-chromatographic derivatization step makes the method highly specific for InsP_x.

1.0 g/L phosphate (Pi, TraceCERT® for IC) in water is utilized. Inositol phosphates Ins(3)P₁ (sodium salt), Ins(2,4)P₂ (sodium salt), Ins(1,4,5)P₃ (sodium salt), Ins(2,3,5,6)P₄ (sodium salt), Ins(1,3,4,5,6)P₅ (sodium salt) Ins(1,2,3,4,5,6)P₆ (sodium salt) are dissolved in water.

Phytic acid (1.66 g/L in 50 mM NaOAc pH 5.5 and 3.6) are digested enzymatically using 10 U/L phytase activity at 37 °C for 24 h. Samples are taken periodically (after 5, 30, 60, 120, 180, 240, 300 min and 24 h) and stopped by heat.

HPTLC Cellulose F (Merck), 20 x 10 cm and 10 x 10 cm are used.

2.0 μL of sample solutions and 2.0–19.0 μL of standard solutions are applied as bands with the Automatic TLC Sampler (ATS 4), 20 tracks, band length 6.0 mm, distance from the left edge 15 mm, track distance 10 mm, distance from the lower edge 10 mm.

Plates are developed in a twin through chamber after chamber saturation for 30 min with 20 mM NaOAc – 10mM NH₄Cl – 2-propanol – 1,4-dioxane – acetic acid 500:520:200:6 (V/V) up to 75 mm (from the lower edge), followed by drying overnight (minimum 12 h) at 105 °C.

1. Enzymatic digest: Still warm plates are sprayed with 1 mL of enzyme solution (250-fold diluted Quantum® Blueliquid 5G in 50 mM NaOAc pH 4.5) using the Derivatizer (pre-cleaned with water). After spraying, the plate is pre-incubated at ambient temperature for 5 min and then transferred to a TLC Plate Heater at 55 °C for 15 min.
2. Molybdate reagent: Plates are sprayed with 0.5 mL of molybdate reagent (5 mL of a 10 g/L ammonium molybdate heptahydrate aq. solution mixed with 200 μL of concentrated sulfuric acid freshly prepared every day) using the Derivatizer. Subsequently the plates are treated with UV light at 254 nm for 15 min.

Images of the plate are captured with the TLC Visualizer in white light.

Absorbance measurement is performed with a DAD scanner [1] and with the TLC Scanner 4 at 774 nm, with a scanning speed of 5 mm/s, a data resolution of 25 μm/step, slit dimension 5.0 mm x 0.3 mm, spectra recording from 200 to 800 nm.

This HPTLC method is suitable for the separation of InsPx pools as well as P_i. The isomers Ins(3)P₁, Ins(2,4)P₂, Ins(1,4,5)P₃ and free phosphate are baseline separated. Ins(2,3,5,6)P₄ and Ins(1,3, 4,5,6)P₅ may be quantified by the peak splitting method. InsP₆ (track 8) shows two bands in a concentration-dependent manner. Presumably, the part that is present as an undissolved salt remains on the application line, while the free base migrates to an R_F value of 0.06 and thus comigrates with Ins(1,3,4,5,6)P₅ (R_F = 0.07).

Acidic conditions or salt containing samples may affect R_F values, however not the overall separation of inositol phosphates. The quantification of the InsP_x isomers can be performed by external standards and linear regression. For free phosphate, two linear ranges were found between 5–15 ng and 20–150 ng with correlation coefficients of 0.99 ([1] by using the Kubelka-Munk equation). Free phosphate was detected with a LOD and LOQ of 5.7 and 6.9 ng respectively.

The method is utilized to study the InsP_x fingerprint of a phytase to evaluate its ability of phytic acid degradation. Our results show that the HPTLC is suitable for a rapid screening of inositol phosphates with a semi-high sample throughput. Accumulation of isomers can be detected as well as a quantitative phosphate release. The presented method is a useful tool for a fast, visual evaluation of novel phytases.

[1] C. Henninger et al., J Sci Food Agric. (2023), https://doi.org/10.1002/jsf2.109.

Contact: Corinna Henninger, Offenburg University of Applied Sciences, Badstrasse 24, 77652 Offenburg, Germany, corinna.henninger@hs-offenburg.de

Contamination of food with mineral oil products is of significant concern to food safety. Mineral oils can enter food either intentionally or unintentionally as contaminants. Mineral oils, known as MOH (mineral oil hydrocarbons) are intricate mixtures of hydrocarbons obtained from crude oil. They are divided into two fractions: 1. mineral oil aromatic hydrocarbons (MOAH), and 2. paraffin oil, also known as mineral oil saturated hydrocarbons (MOSH). MOSH consist of straight and branched open-chain alkanes (paraffins) and alkylated cycloalkanes (naphthenes). The diverse nature of these compounds presents a substantial challenge for analytical methods. For companies involved in milk processing, particular attention is given to detecting milk batches that might be contaminated with paraffin oil.

The screening method employed for this purpose needs to be as straightforward and rapid as possible. HPTLC, due to its ability to simultaneously separate multiple samples, is a suitable technique to fulfill these requirements.

The procedure described here involves isolating non-polar components from milk through liquidliquid extraction. Following the approach of Wagner and Oellig [1], the MOSH fraction is subsequently separated and identified on the HPTLC plate using primuline as derivatization step. This method can detect 5.0 μg/mL of paraffin oil in milk.

To avoid the possibility of contamination from leaching processes, only glass laboratory equipment is utilized for preparation of standards and samples. During method development, technical grade paraffin oil was diluted with toluene to a concentration of 10.0 mg/mL.

For calibration and repeatability, 2.0 mL of each milk sample containing 3.9% milk fat is pipetted into glass centrifuge tubes using a volumetric glass pipette. These samples are spiked with the 10.0 mg/mL standard solution using a 10.0 μL glass syringe. The sample is acidified with 0.4 mL of formic acid (≥98%). After adding 3.0 mL of tert-butyl methyl ether and vortexing for 30 s, the sample is centrifuged for 5min with 2790 x g. The organic supernatant is transferred into a glass vial and used as test solution.

HPTLC plates silica gel 60 F₂₅₄ (Supelco), 20 x 10 cm are used after pre-washing with cyclohexane up to 50 mm and drying for 30 min at 100 °C.

Prior to sample application, HPTLC plates are impregnated with primuline solution (75 mg/L in methanol) using the Chromatogram Immersion Device 3 (time 20 s, speed 1), and dried using the TLC Plate Heater at 100 °C for 30 min.

6.0 μL of sample and standard solutions are applied as bands with the HPTLC PRO Module APPLICATION, band length 6.0 mm, distance from the left edge 18.0 mm, track distance 8.5 mm, distance from the lower edge 8.0 mm. The first rinsing step (solvent bottle 1) is performed with methanol – acetonitrile – isopropanol – water – formic acid 250:250:250:250:1 (V/V) and the second rinsing step (solvent bottle 2) with methanol – water 7:3 (V/V).

In the HPTLC PRO Module DEVELOPMENT, prior to the development the plates are pre-dried for 30 s, activated at 0–5% relative humidity for 10 minutes using a molecular sieve, and pre-conditioned with cyclohexane at a pump power of 35% for 300 s. No conditioning step is used. Development with cyclohexane to the migration distance of 30 mm from the lower edge of the plate, followed by drying for 5 min.

Images of the plates are captured with the TLC Visualizer 2 in UV 366 nm.

A calibration curve ranging from 5.0 μg/mL to 100.0 μg/mL was created by spiking aliquots of a milk sample with paraffin oil. For this purpose, the 2.0 mL samples were spiked with 1.0 –20.0 μL of the 10.0 mg/mL standard solution. Due to the short development distance of 30 mm, the development time is only 2 min. Considering activation, pre-conditioning, and drying, the complete development cycle is 30 min. This means that if 16 samples are applied to one HPTLC plate, the separation time per sample is only 1.8 min (6.7 min including application). As the HPTLC PRO system autonomously moves the plate from one module to the next, there is no time wasted due to manual transfer between the individual HPTLC steps. In the image of the developed plate in UV 366 nm, the first track is the not spiked milk sample, followed by the reference samples for the matrixmatched calibration and, on the last four tracks, the repeatability samples. The clear separation of the paraffin oil fraction from the other extracted fluorescent components is readily apparent. The lowest spike at 5.0 μg/mL is also clearly visible (second track from the left).

HPTLC chromatograms of matrix-matched calibration standards (from left: track 1–7) and repeatability samples (from left: track 8–11) in UV 366 nm after separation on the primuline impregnated HPTLC plate with a migration distance of 30 mm.

Using visionCATS, peak profiles from the image in UV 366 nm were generated for each individual track. Subsequently, the evaluation was conducted based on peak height.

HPTLC chromatograms of matrix-matched calibration standards (from left: track 1–7) and repeatability samples (from left: track 8–11) in UV 366 nm after separation on the primuline impregnated HPTLC plate with a migration distance of 30 mm.

The calibration curve was generated using a Mime- 2 function. The calibration from 5.0 μg/mL to 70.0 μg/mL includes the results obtained for the four spiked samples (blue cross).

Calibration curve (from 5.0 μg/mL to 70.0 μg/mL) from the peak heights using the Mime-2 function.

For the reproducibility of the extraction, an average value of 12.70 μg/mL, and a standard deviation (STDEV) of 0.51 μg/mL were obtained.

Table 1: Results of the reproducibility and recovery tests

This study demonstrates the rapid and simple, yet sensitive detection of paraffin oil in milk.

[Note]: for a proper quantification, a linear calibration curve would be needed.

[1] M. Wagner and C. Oellig, J. Chromatogr. A ,1588 (2019) 48–57, https://doi.org/10.1016/j.chroma.2018.12.043.

Contact: Dr. Tiên Do, CAMAG, Sonnenmattstrasse 11, 4132 Muttenz, Switzerland, tien.do@camag.com

Mercedes G. López and Luis F. Salomé work on agave fructooligosaccharides (aFOS, fructans) at the Research and Advanced Studies Center (CINVESTAV)-IPN in Mexico. Patricia A. Santiago and Ruth E. Márquez also conduct fructan research at the Interdisciplinary Research Center for the Integral Regional Development (CIIDIR)-IPN in Mexico. The group focuses on fructan chemical characterization and their biological effects as prebiotics on the human health.

Fructans are a polydisperse mixture of fructose polymers, and contain only one or no glucose in their structures. They are commonly found in agaves and possess several industrial applications. These molecules have been mainly used as prebiotics and supplements to produce functional foods. Moreover, they are directly correlated with the yield and quality of the alcoholic drink tequila. The most powerful analytical technique for the characterization of fructans is high performance anion exchange chromatography (HPAEC). However, this technique is time consuming taking up to 80 min for a single sample. In this context, the HPTLC technique allows the parallel analysis of up to 17 samples and there are several options of derivatizing reagents for carbohydrate visualization. The solvent consumption is only 70 mL per sample batch, which is climate friendly.

Thus, this study aimed at exploring the potential of the aFOS fraction as a good descriptor of the fructan differentiation in agave species through age, and at the feasibility of HPTLC as a robust fingerprinting platform through multivariate data analysis (MVDA). In this study, HPAEC was used a standard technique for comparison with HPTLC.

The proposed method is rapid, accurate and precise. It is suited as a high-throughput method with a significant reduction in working time, supplies and solvents. Finally, it produces robust data which can be used for multivariate modelling.

2.0 mg of glucose, fructose, sucrose, 1-kestose, 1-nystose, and 1-F fructofuranosylnystose (DP5) are dissolved in ethanol – water 7:3 (V/V).

Agave fibers (Agave potatorum and Agave angustifolia) are extracted once in aqueous ethanol (80 %) for 1 h at 60 ºC, then reextracted twice with pure water. The extracts are defatted with chloroform and the aqueous phase is reduced and spray-dried. Samples are prepared at 7.0 mg/mL. They are firstly dissolved in 0.3 mL of water and then filled to 1.0 mL with absolute ethanol (room temperature).

HPTLC plates silica gel 60 F₂₅₄ (Merck), 20 x 10 cm are used.

Samples and standard solutions are applied as bands with the Automatic TLC Sampler (ATS 4), 17 tracks, band length 6.0 mm, distance from left edge 20.0 mm, distance from lower edge 10.0 mm and 10.0 mm between bands. 5.0 µL for sample solutions and 1.0 µL for standard solutions are applied.

Plates are developed in the ADC 2 with chamber saturation (with filter paper) 20 min and after activation at 47% relative humidity for 10 min using a saturated solution of potassium thiocyanate. The first development is performed with isopropanol – butanol – water – acetic acid 14:10:4:2 (V/V) to the migration distance of 75 mm (from the lower edge), followed by drying for 5 min. The second development is performed with isopropanol – butanol – water – formic acid – acetic acid 14:10:4:1:1 (V/V) to the migration distance of 85 mm (from the lower edge), followed by drying for 5 min.

The plates are immersed into a solution of diphenylamine-aniline-phosphoric acid (referred to as aniline) using the Chromatogram Immersion Device, immersion speed 2 cm/s and immersion time 2 s, dried for 30 s with cold air and heated at 120 ºC for 3 min using the TLC Plate Heater. The same samples are also derivatized using the same conditions with α-naphthol and orcinol. For these two last reagents, the derivatization temperature was 110 ºC. All derivatization reagents were prepared as previously described [1]. All data is extracted as previously reported using information from the RGB channels and gray scale [2].

Images of the plate are captured with the TLC Visualizer in white light.

As expected, the information produced by HPAEC was able to differentiate agave specimens according to their species and age. Moreover, the data was also good for creating supervised models. The HPAEC models indicated a decrease of simpler sugars such as fructose, glucose and sucrose, while fructans with higher degree of polymerization (DP) are synthetized as the agave age increases. Interestingly, the visual inspection of HPTLC chromatograms, independent of the derivatization reagent, showed the same trend. Representative chromatograms of agavins from A. potatorum derivatized with aniline, α-naphthol and orcinol showed this trend. Also, it was observed that α-naphthol and orcinol produced more intense monochromatic bands, while aniline produced bicolor patterns. That is, blue zones indicate glucose containing aFOS and pink zones fructose containing aFOS. Using the standards’ R_F values, DP-11 was determined as the maximum visible countable DP for aFOS.

(Left) Processed HPTLC chromatograms in white light (according to [1]) of representative Agave potatorum samples and (right) negative-HPTLC chromatograms, after derivatization with A aniline, B α-naphthol and C orcinol. STD, standard mixture; RSE, raftilose; RNE, raftiline. Track number indicates agave age expressed in years. Reproduced from [1]. (https://creativecommons.org/licenses/by/4.0/legalcode).

For MVDA, the intensity values of the peak profiles from images (PPID) were inverted during data extraction [1] and negative-HPTLC chromatograms were processed with an open-source-software in all color channels according to [2]. The data was normalized to the quality control sample track in each corresponding plate. Furthermore, the data was scrutinized by principal component analysis (PCA), orthogonal projection to latent structures discriminant analysis (OPLS-DA) and orthogonal projection to latent structures (OPLS) analysis. Subsequently, data was approached by MVDA and a PCA showed a clear separation dictated by species factors along the PC1 (captures the most variation), while samples were separated by age along the PC2 (the second most variation). Moreover, there was a clear subgrouping of samples from 1-3 years old plants and 4-6 years old plants. Thus, samples were classified as younger than four years (YT4) and older than three years (OT3), and they were then analyzed by OPLS-DA. The model was well validated through permutation test (100 permutations, Q² = 0.82) and in a CV-ANOVA test (p = 2.39 x 10^-8). The S-plot of the analysis indicated that A. potatorum samples possessed a higher content of glucose/fructose (R_F 0.57) and DP5 (R_F 0.25 – 0.26) compared to A. angustifolia, which possessed higher contents of sucrose (R_F 0.47) and high-DP aFOS (R_F 0.04 – 0.13).

To further explore metabolic differentiation correlated to age, the data set was approached by OPLS using agave age as a quantitative “Y” variable. The analysis was well validated (Q² = 0.98, p = 2.93 × 10^-7) and indicated that carbohydrate variation, specially increase of DP-7, DP8 and D-P9 aFOS (R_F 0.14, 0.13, 0.11, 0.16, 0.17, 0.18, and 0.10) was correlated with the increase of age in both agave species determined by the Predictive Variable Importance for the Projection (VIP_pred)-plot. It is worth to mention, that models resulting from HPTLC data provided higher Q² and p-values than those obtained from HPAEC data. For instance, the HPAEC model for YT4/OT3 differentiation produced a Q² = 0.66 and p = 8.13 × 10^-15. Furthermore, the same data set produced a Q² = 0.80 and p = 8.13 × 10^-5 for the OPLS model of carbohydrate and age correlation. Here, we only present the best models for each scrutinized factor. Thus, the best PCA model to separate samples according to species and age is that produced from plates derivatized with aniline and extracted in gray channel. The best OPLS-DA model to classify samples according to species is that produced from plates also derivatized with aniline but extracted in blue channel. The best OPLS model for age correlation is produced from plates derivatized with α-naphthol and extracted in green channel.

MVDA of HPTLC data approached by A, PCA colored according to species; B, PCA colored according to age; C, OPLS-DA for plants younger than 4 years old and older than 3 years old; D, S-plot of OPLS-DA for YT4/OT3 differentiation; E, orthogonal projection to latent structures for carbohydrate/age correlation; F, Predictive Variable Importance for the Projection (VIP_pred)-plot of the age OPLS model. Reproduced from [1]. (https://creativecommons.org/licenses/by/4.0/legalcode).

Thus, we concluded that the aFOS fraction was enough to describe agavins metabolism and that HPTLC data was robust enough to be combined with MVDA, which produced even better supervised models than HPAEC.

[1] L.F. Salomé-Abarca et al. (2023). Curr. Res. Food Sci. 100451, https://doi.org/10.1016/j.crfs.2023.100451.
[2] D. Fichou et al. (2016). Anal. Chem. 12494, https://doi.org/10.1021/acs.analchem.6b04017.

Contact: Dr. Mercedes G. López, Department of Biotechnology and Biochemistry, CINVESTAV-Irapuato, 36824, Guanajuato, Mexico, mercedes.lopez@cinvestav.mx

The research team at the University of Western Australia (UWA), Division of Pharmacy, developed an HPTLC based real-time honey assessment tool for beekeepers and packers to determine a honey’s floral source alongside the collation of key phytochemical parameters and bioactivity data for a wide range of Australian honeys. Currently, the team is using HPTLC as a qualitative and quantitative honey analysis tool. They monitor changes over time, and caused by storage and handling conditions.

Honey is a sweet natural product appreciated for its unique flavor, color and bioactivity. Honey is produced by honeybees from flower nectar. Raw honey contains phenolics, flavonoids, proteins, vitamins and minerals, however, it is rarely sold on the market in the raw form. Before being bottled and packaged, honey undergoes several processing steps, including filtration, radiation and/or heating. Excessive or prolonged heating can have detrimental effects to the honey’s quality. It is known to produce potentially toxic chemicals like the Maillard reaction product 5-hydroxymethyl-furfural (HMF), which is suspected to have carcinogenic effects when ingested in high doses. An HPTLC based method can be used for the fast and cost-effective assessment of the HMF content of honey and thus presents a convenient honey quality control tool.

The applied method is rapid, reliable, and repeatable and, therefore, a convenient analytical tool for routine quality control of honey. The method involves a simple dissolution step followed by a short chromatographic development time (9–10 min) without chamber saturation or derivatization. Up to 10 samples can be analyzed on a single plate with only small sample quantities (approx. 1 g) required.

Aqueous 0.008% (w/v) freshly prepared HMF solution.

An artificial (ART) honey is prepared by dissolving 40.5 g fructose, 33.5 g glucose, 1.5 g sucrose and 7.5 g maltose in 17 mL of deionized water. The ART is individually kept at 40 °C, 60 °C and 80 °C. Sampling is done at time of preparation (t₀), 6 h, 12 h, 24 h, 48 h and then over a period of four months. For HPTLC analysis, the collected honey samples are prepared as 1 g/10 mL aqueous solutions.

HPTLC plates silica gel 60 F₂₅₄ (Merck), 20 x 10 cm are used.

Samples and standard solutions are applied as bands with Linomat 5, 15 tracks, band length 8.0 mm, distance from left edge 20.0 mm, distance from lower edge 8.0 mm.

Plates are developed in the ADC 2 without chamber saturation with ethyl acetate as mobile phase to the migration distance of 50 mm (from the lower edge), followed by drying for 5 min.

Images of the plate are captured with the TLC Visualizer 2 in UV 254 nm.

To find out the absorption maximum for HMF, a spectral scan is performed using the TLC Scanner 4 from 220 nm – 850 nm both on the bands of pure HMF and HMF bands produced in artificial honey treated at elevated temperature. Based on these scans, HMF analysis in honey samples is carried out at 290 nm using the TLC Scanner 4.

The following figure shows the HPTLC fingerprints of pure HMF and HMF produced during storage at elevated temperature. During the analysis, the HMF is completely separated from the honey matrix, and both pure and newly produced HMF appear at R_F 0.76. Positive identity of HMF is indicated by spectra comparison of standard and sample. The absorbance maximum is identified at 290 nm.

HPTLC image of HMF in UV 254 nm (left; track 1: Standard HMF and track 2: HMF produced in honey stored at elevated temperature) and UV-VIS spectra of HMF from 220 – 850 nm (right).

For quantification and to prepare the HMF standard curve, 1.0, 2.0, 3.0, 4.0 and 5.0 μL of the respective standard solution are applied. For the analysis of HMF in the honey samples, 10.0 μL of the respective honey solution is applied at a rate of 30.0 nL/s. After development, Peak Profiles from Images (PPI) obtained in UV 254 nm with the TLC Visualizer 2 are compared with Peak Profiles from Scanning Densitometry (PPSD) subsequently measured at 290 nm with the TLC Scanner 4.

Editor´s note: The response of HMF is higher for scanning densitometry compared to image-based evaluation; the working range can be adjusted to the linear working range by reducing the concentration of the sample and standard solutions.

The level of HMF in artificial honey was within the acceptable limit (80 mg/kg of honey) after 4 months of storage at 40 °C. The HMF limit exceeds the acceptable limit after 48 h for artificial honey stored at 60°C. For the honey stored at 80°C, the limit is exceeded already after 24 h of storage. This experiment shows that honeys need to be stored or temperature treated carefully to limit the formation of HMF.

HMF content in artificial honey stored at different temperatures over time

The HPTLC method for the detection and quantification of HMF in honey is easy to perform and offers a convenient quality control tool for the honeybee industry. It allows monitoring the HMF-related changes to the quality of honey during processing (especially temperature treatment) and storage. The absence of any sample pre-treatment steps and post-chromatographic derivatization, a neat solvent as developing solvent, and no chamber saturation and activation are major advantages. The method may also be used for the detection and quantification of HMF in other botanicals and foods with high sugar content.

Contact: Dr. Cornelia Locher, CRC for Honey Bee Products and Division of Pharmacy, School of Allied Health, University of Western Australia, Crawley, Western Australia, 6009, Australia, connie.locher@uwa.edu.au

Amit Palande, application specialist under the guidance of Akshay Charegaonkar (managing director), Tukaram Thite (senior lab manager) and Dr. Saikat Mallick (lab manager) work at Anchrom Enterprises Pvt Ltd, Mumbai, India. The company specializes in instrumental planar chromatography, and develops new, quantitative, and regulatory compliant analytical methods for pharmaceutical formulations, APIs, herbal products, food products, organic intermediates, dyes etc. Mr. Palande benefits from HPTLC because it is a fast, simple, economical and flexible, “visible chromatography” technique. HPTLC is risk-free and multiple detections can be made without repeating chromatography. This is especially helpful for dihydroartemisinin, because it is not UV absorbing and needs derivatization for densitometric detection.

Dihydroartemisinin (also known as dihydroqinghaosu, artenimol or DHA) is a drug used to treat malaria. It is globally recognized for its efficacy and safety in the clinical treatment of malaria for decades. DHA is the active metabolite of all artemisinin compounds (artemisinin, artesunate, artemether, etc.) and is also available as a drug by itself. It is a semi-synthetic derivative of artemisinin and is widely used as an intermediate in the preparation of other artemisinin derived antimalarial drugs. DHA is often combined with piperaquine phosphate (PPQ). Like any formulation, these tablets need to be tested for shelf life i.e. stability.

Shelf-life studies are performed by accelerated or forced degradation studies as per ICH guidelines Q 1 A (R2).

Chemical structures of dihydroartemisinin and piperaquine phosphate

HPTLC is well suited for stability studies, because it is inexpensive and time-saving. Accelerated degradation studies need a very large number of samples to be analyzed. By HPTLC, 15-20 samples can be quantified simultaneously on one plate in about 40–80 minutes. The solvent consumption is only about 20 mL for those 20 samples and little waste is produced. It was separately established, that piperaquine phosphate did not degrade in the studies. Hence a method was developed that kept piperaquine phosphate at the base of the chromatogram but selectively moved DHA and its two degradation products. DHA and the degradation products were detected by simple derivatization and then quantitatively evaluated.

HPTLC is a cost-effective and time-saving technique for the pharma industry, which deals with a heavy load of samples, also competition to introduce a new formulation is very intense. The presented method is a green method, which only uses 20 mL of solvent for 15–20 samples and produces almost no waste. Since the degradation products are unknown, the common procedure of establishing the calibration function using the diluted main substance (DHA in this case) was applied. Many times, as in this case, the samples have to be overloaded to detect the small quantities of impurities/degradation products.

A stock solution of 1.0 mg/mL of dihydroartemisinin in acetonitrile is prepared and diluted 1:20 (V/V) for analysis.

Ten tablets containing 40.0 mg of dihydroartemisinin and 320.0 mg of piperaquine phosphate are milled. Powdered tablets equivalent to 100.0 mg of DHA are transferred in a 20.0 mL volumetric flask and dissolved in acetonitrile. The supernatant after centrifugation is used for application.

HPTLC plates silica gel 60 F₂₅₄ (Merck), 20 x 10 cm are used.

Samples and standard solutions are applied as bands with the Automatic TLC Sampler (ATS 4), 15 tracks, band length 8.0 mm, distance from left edge 20.0 mm, distance from lower edge 8.0 mm. 10 μL for sample and standard solutions are applied.

Plates are developed in a 20 x 10 cm Twin Trough Chamber with chamber saturation (with filter paper) for 20min, development with cyclohexane – ethyl acetate – glacial acetic acid 10:5:1 (V/V) to the migration distance of 90 mm (from the lower edge), followed by drying for 5 min.

Plates are sprayed with 3.0 mL of anisaldehydesulfuric acid reagent (to 170 mL of cooled methanol, 20 mL of acetic acid and 10 mL of sulfuric acid are added, after cooling to room temperature 1mL of anisaldehyde is added) using the CAMAG Derivatizer. After spraying, the plates are heated at 110 °C for 5 min using the TLC Plate Heater.

Images of the plate are captured with the TLC Visualizer in UV 366 nm and white light.

Absorbance measurement at 540 nm (tungsten lamp) is performed with CAMAG TLC Scanner 4 and visionCATS, slit dimension 6.00 mm x 0.45 mm, scanning speed 20 mm/s, evaluation via peak area.

The determination of the degradation product with the mobile phase cyclohexane – ethyl acetate – glacial acetic acid 10:5:1 (V/V) is verified by the positions of the individual drugs, where PPQ remains at the application position and DHA moves to hRF 40. Any other zones at different hR_F are reported as degradation products. The placebo shows no zones in the chromatogram. Prior to derivatization, no zones are visible. Quantitative evaluation is performed after derivatization by densitometric absorbance measurement at 540 nm. The data is recorded and used to calculate the degradation products. Total impurities in the samples are found to be within the 1.5% limit as per specification, defined in house. A representative densitogram of a stability batch sample and standard is shown. The area of each impurity is calculated as follows and then summarized.

[1] CAMAG Application Note A-86.1: Determination of artemisinin in Artemisia annua leaf by HPTLC

Contact: Akshay Charegaonkar, A 101-104, Shree Aniket Apartment, Navghar Road, Mulund, Mumbai 400081, India, hptlc@anchrom.in, www.anchrom.in

In 2022, CAMAG laboratory introduced the concept of complementary developing solvents (CDS) based on the combination of three automated analyses using three developing solvents (DS): one low polarity (LPDS), one medium polarity (MPDS), and one high polarity (HPDS) solvent [1]. With these three DS, any compound is characterized by three R_F values instead of one. Introduced into a database, a large dataset can be compiled from these values, which then may be subjected to data mining and machine learning.

In their paper [2], CAMAG researchers describe how the application of a CDS protocol to a large number of highly diverse individual compounds was used in combination with machine learning to predict the R_F values of individual substances from molecular properties, and to generate proposals for the identity of a zone. Coupled with machine learning, the CDS concept as a very powerful, general, and medium to high throughput technique for routine analysis and sophisticated research, may become the future of HPTLC. It may help replacing common tasks such as visual evaluation and pattern recognition, as well as subjective pass/fail decisions by automated procedures and numerical values generated by suitable algorithms.

Individual standard solutions were prepared at a concentration of 1.0 mg/mL (adjusted when necessary). Methanol was used as solvent for iridoids, coumarins, pharmaceutical drugs, flavonoids, triterpenes, sesquiterpenes, steroids, phospholipids and cannabinoids, 50% aqueous acetonitrile for carbohydrates, 50% aqueous methanol for amino acids, and toluene for monoterpenes. System Suitability Test (SST): the ready to use solution of Universal HPTLC Mix (UHM) was prepared in house according to [3] and applied on each plate.

HPTLC plates silica gel 60 F₂₅₄ (Supelco/Merck), 20 x 10 cm are used.

2.0 μL of sample solutions are applied as bands with the Automatic TLC Sampler (ATS 4), 15 tracks, band length 8.0 mm, distance from left edge 20.0 mm, distance from lower edge 8.0 mm.

Plates were developed with the three developing solvents in the ADC 2 with activation of the plate at 33% relative humidity for 10 min using a saturated solution of magnesium chloride. LPDS (toluene, ethyl acetate 9:1 (V/V)) is used without saturation, whereas MPDS (cyclopentyl methyl ether, tetrahydrofuran, water, formic acid 40:24:1:1 (V/V)) and HPDS (ethanol, dichloromethane, water, formic acid 16:16:4:1 (V/V)) are used with 20 min chamber saturation (with saturation pad). The developing distance for all three methods was 70 mm (from the lower edge). Plates were dried for 5 min.

Derivatization with anisaldehyde sulfuric acid reagent (10.0 mL of sulfuric acid were carefully added to the ice-cold mixture of 170.0 mL of methanol and 20.0 mL of acetic acid. To this solution, 1.0 mL of anisaldehyde was added) by spraying (Derivatizer, blue nozzle, 3.0 mL, spraying level 3) was followed by 3 min of heating at 100°C. Images of plates derivatized with Fast blue salt B reagent (250.0 mg of Fast blue salt B (o-dianisidine bis(diazotized) zinc double salt) were dissolved in 10.0 mL of water and mixed with 25.0 mL of methanol and 15.0 mL of dichloromethane) were captured within 2 min after spraying (Derivatizer, green nozzle, 3.0 mL, spraying level 5).

For the derivatization with NP reagent (1.0 g of diphenylborinic acid aminoethylester was dissolved in 200 mL of ethyl acetate ) / PEG (10 g of polyethylene glycol 400 (macrogol) were dissolved in 200 mL of dichloromethane), the plates were heated at 100 °C for 3 min, cooled to room temperature, then sprayed with the mixture NP/PEG 1:1 (V/V) (Derivatizer, green nozzle, 3.0 mL, spraying level 3), and dried for 2 min. Derivatization by immersion (Immersion Device, speed 5, time 0) with toluene sulfonic acid reagent (10% of p-toluene sulfonic acid in ethanol) was followed by heating at 150°C for 3 min.

TLC Visualizer in UV 254 nm, UV 366 nm, and white light prior to derivatization, and UV 366 nm, and white light after derivatization (as needed).

For the UHM, TLC Scanner 4 and visionCATS, absorbance measurement at 254 nm, slit dimension 5.00 mm x 0.20 mm, scanning speed 50 mm/s, and in fluorescence mode at 366>/400 nm.

The open-source software KNIME (version 4.6) was used. The “RDKit KNIME Integration” was applied for curation of the databases and conversion of the chemical structures. 178 chemicals of the learning set were then used to benchmark various machine learning models.

R_F values obtained from peak profiles from images (PPI) or scanning densitometry (PPSD), were used with the Random Forest regressor algorithms including 100 trees.

For building a powerful model, four steps were taken:

Overview of the machine learning pipeline and its workflow

The first step was the collection of data. For this, a training set consisting of 178 known individual substances was selected from various chemical classes, covering molecular weights (MW) ranging from 75.1 g/mol to 1131.3 g/mol, and computed octanol/water partition coefficients (SlogP) in the range of -7.53 to 13.98. Using the open source software KNIME and its extensions, molecular descriptors (e.g. MW, SlogP, topological polar surface area (TPSA)…) were computed for each substance. In addition, each substance was chromatographed with the CDS, generating 178 x 3 = 534 R_F values. In the second step, the dataset was cleaned by filtering all descriptors for null variance. The third step included the training of the model. For this, the performance of three regressors was evaluated according to their capacity to predict the R_F within the training set. The Random Forest, trained with 100 trees, yielded the best correlation coefficients R² 0.55, 0.72, and 0.64 for the LPDS, the MPDS, and the HPDS, respectively.

For testing of the model, a test set was created with 20 other substances. The suitability of the selected substances was verified by demonstrating that the chemical space of the test set was within the chemical space of the training set.

Chemical space (2D t-SNE projection) covered by the 178 chemicals belonging to the training set (black dots) and the 20 chemicals belonging to the test set (orange squares)

The model was used to predict the R_F values of the compounds in the test set. Most predicted R_F differ by less than 0.1 units from the measured values. R_F in the MPDS and the HPDS are both predicted within the correct range and with very small errors, leading to R² of 0.87, and 0.71, respectively. The variance for each individual prediction (LPDS, MPDS, and HPDS) remains smaller than 10%, except for a few compounds.

Measured and predicted R_F values of the compounds in the test set

A reverse test was also performed. The query molecule defined by its R_F values in LPDS, MPDS and HPDS was compared to the database for a number of rows (four each) matching the specified similarity. To calculate the similarity, the Euclidean distance was selected and the four nearer neighbors (most similar) were displayed in an additional column.

Use of the database for proposal of potential matches

The examples above illustrate the potential of the CDS and its combination with machine learning. In this study, R_F values can be predicted, emphasizing that this feature is encoded within the chemical structure of the molecules. Moreover, the link between chemical structures and R_F allows to generate a list of four molecules likely to correspond to an unknown zone in complex mixtures. This prediction would be even more useful, if additional data such as mass and UV-VIS spectra were added to the database.

[1] T.K.T. Do, M. Schmid, I. Trettin, M. Hänni, E. Reich, Complementary developing solvents for simpler and more powerful routine analysis by high-performance thin-layer chromatography, JPC – J. Planar Chromatogr. – Mod. TLC. (2022). https://doi.org/10.1007/s00764-022-00185-1.
[2] T.K.T. Do, I. Trettin, M. Hänni, E. Reich, Applying machine learning to the data obtained with the complementary developing solvents protocol, J. Liq. Chromatogr. Relat. Technol. (2023).
[3] T.K.T. Do, M. Schmid, M. Phanse, A. Charegaonkar, H. Sprecher, M. Obkircher, E. Reich, Development of the first universal mixture for use in system suitability tests for High-Performance Thin Layer Chromatography, J. Chromatogr. A. 1638 (2021) 461830. https://doi.org/10.1016/j.chroma.2020.461830.

Contact: Dr. Tiên Do, CAMAG, Sonnenmattstrasse 11, 4132 Muttenz, Switzerland, tien.do@camag.com

Dr. Kashyap Thummar, Assistant Professor at Graduate School of Pharmacy (GSP), Gujarat Technological University (GTU), India, employs chromatographic separation techniques, especially HPTLC, to develop new and improved quantitative analytical methods for determination of drugs, impurities, adulteration and naturally occurring compounds in a variety of sample matrices. He prefers HPTLC because it is flexible, inexpensive, time-saving and does not produce toxic waste.

Phthalates are esters of phthalic acid that are commonly added to plastics to improve their flexibility, transparency, durability, and longevity. These phthalates are easily released into the environment from plastic and can harm all living organisms. Phthalates enter the body by contact to plastics, e.g. through ingestion, inhalation, skin absorption, and intravenous injection.

For the detection of extractable and leachable phthalates in pharmaceutical products, a simple, rapid, precise, and accurate HPTLC method was developed. It simultaneously estimates the presence of four different phthalates in various pharmaceutical products and containers: dimethyl phthalate (DMP), diethyl phthalate (DEP), dibutyl phthalate (DBP) and di(2-ethylhexyl) phthalate (DEHP). The method was successfully applied for determination of extractables and leachables from 12 parenteral products packed in plastic containers.

10.0 mg of each phthalate were individually dissolved in 10.0 mL of methanol and subsequently combined to prepare a working standard solution at a concentration of 0.1 mg/mL of each.

2.0 g of the plastic container (extractable) and 30.0 mL of the product packed in it (leachable) are extracted 3 times with 30.0 mL of n-hexane by 10 min of sonication. The organic layers are collected, evaporate to dryness and reconstituted with 1.0 mL of methanol.

HPTLC plates silica gel 60 F₂₅₄ (Merck), 20 x 10 cm are used.

Samples and standard solutions are applied as bands with the Linomat 5, 20 tracks, band length 5.0 mm, distance from left edge 10.0 mm, distance from lower edge 8.0 mm. 20.0 μL for sample solutions and 1.0–14.0 μL for standard solutions (8 points for calibration) are applied.

Plates are developed in a saturated Twin Trough Chamber (15 min, with filter paper) with n-hexane – ethyl acetate 9:1 (V/V) to a migration distance of 90 mm (from the lower edge), followed by drying with cool air for 5 min.

Images of the plate are captured with the TLC Visualizer in UV 254 nm.

Absorbance measurement at 240 nm is performed with the TLC Scanner 4 and winCATS, slit dimension 4.00 mm x 0.30 mm, scanning speed 20 mm/s.

HPTLC chromatogram in UV 254 nm (left) and densitogram measured at 240 nm (right)

A representative densitogram of samples is shown. During evaluation of the chromatogram, phthalates in simulated and test samples give the same R_F values as the standard and are well separated from matrix components. A simulated sample is a laboratory-made sample that is designed to mimic a real-world product or material.

3D profile of scanned sample and standard tracks (measured at 240 nm)

DMP, DEP, DBP and DEHP are separated at R_F values of 0.23, 0.31, 0.44 and 0.60 respectively. The limit of quantification was in the range of 41.7 to 99.8  ng/band for the four phthalates and linearity was established between 100.0 and 1400.0  ng/band. The presence of individual phthalates was found in 12 pharmaceutical products (all parenteral formulations) in significant amounts. Importantly, of the four phthalates, DEHP was found in all tested samples as an extractable and leachable. The HPTLC method for detection of phthalates is cost-effective as compared to available analytical methods such as HPLC, LCMS/ MS, GCMS/MS etc. as these techniques require more solvent consumption, power consumption, analysis time, and involve complex sample preparation methods. The method can be universally applied to other samples apart from pharmaceutical products like water, food items etc. which are sold in plastic containers.

[1] T. H. Broschard et al. (2016) Regulatory Toxicology and Pharmacology 81, 201–211.
[2] K. Thummar and N. Sheth, Publication date: 2022/6/7, Patent office: IN, Patent number: 398670.
[3] K. Thummar et al. (2020) Analytical Chemistry Letters 10 (1), 93–103.

Contact: Dr. Kashyap Thummar, Assistant Professor, Graduate School of Pharmacy, Gujarat Technological University, Gandhinagar, Gujarat, India, ap_kashya@gtu.edu.in