Screening for Ricinoleic Acid as a Chemical Marker for Secale

Oct 4, 2016 - Screening for Ricinoleic Acid as a Chemical Marker for Secale cornutum in Rye by High-Performance Thin-Layer Chromatography with Fluores...
0 downloads 10 Views 4MB Size
Article pubs.acs.org/JAFC

Screening for Ricinoleic Acid as a Chemical Marker for Secale cornutum in Rye by High-Performance Thin-Layer Chromatography with Fluorescence Detection Claudia Oellig* Institute of Food Chemistry, University of Hohenheim, Garbenstrasse 28, 70599 Stuttgart, Germany ABSTRACT: Ricinoleic acid as the characteristic fatty acid of Secale cornutum oil is a good marker for Secale cornutum impurities in cereal. The presented screening for ricinoleic acid in rye by high-performance thin-layer chromatography with fluorescence detection (HPTLC−FLD) offers a selective and sensitive method for the determination of Secale cornutum and is very different from existing gas chromatographic analyses. Lipid extraction was followed by transesterification and solid-phase extraction cleanup; thereafter, extracts were selectively derivatized with 2-naphthoyl chloride and analyzed by HPTLC−FLD with silica gel plates and cyclohexane/diisopropyl ether/formic acid (86:14:1, v/v/v) as mobile phase. For quantitation, the enhanced fluorescence was scanned at 280/>340 nm. Limits of detection and quantitation of 0.1 and 0.4 mg ricinoleic acid/kg of rye were obtained, which enables the determination of Secale cornutum far below the maximum admitted level. With near-100% recoveries and low standard deviations at relevant spiking levels, reliable results were guaranteed. KEYWORDS: ricinoleic acid, Secale cornutum, rye, screening, high-performance thin-layer chromatography−fluorescence detection, HPTLC−FLD



INTRODUCTION Ergot (Secale cornutum) is the overwintering form of the parasitic fungus Claviceps purpurea1 and is known to be responsible for different toxicological effects in mammals. Consumption of infested grain can lead to severe intoxications, which are caused by several toxic alkaloids, known as ergot alkaloids.2−6 Thus, the infestation of food and feed grain with Secale cornutum, particularly growing on rye, merits very serious consideration. At present, there is no maximum limit for the ergot alkaloid content in grain and grain-based food in the European Union (EU). There are, however, activities ongoing, and it may be expected that such limits will be defined in the near future.7 In contrast, maximum levels for Secale cornutum in food and feed materials already exist. For feeding stuff containing unground cereals, the maximum level is set to 0.1% Secale cornutum in directive 2002/32/EC8 on undesirable substances in animal feed. For food, a maximum limit of 0.05% Secale cornutum for certain unprocessed cereals, with the exception of corn and rice, is laid down in regulation 2015/1940/EU, 7 amending regulation 1881/2006/EC.9 The physical determination is known to be incorrect because of the variability of size and weight of Secale cornutum and is furthermore not applicable to processed food.10 Therefore, chemical analysis methods are more suitable and urgently needed to monitor the Secale cornutum contamination of cereal. Apart from the toxic ergot alkaloids, ricinoleic acid ((R)-12hydroxy-(Z)-9-octadecenoic acid) is a key component of Secale cornutum oil.10 According to the literature,11−13 ergot contains about 30% of fat wherein ricinoleic acid is dominant with an amount of ∼30%. In contrast to the significant variability of the ergot alkaloid amount in Secale cornutum,10,14−16 with an average of approximately 0.08% in Central Europe (according to a statement of the European Food Safety Authority17), the © 2016 American Chemical Society

ricinoleic acid content is reported to be quite stable with an average amount of 10%.10 Regarding the chemistry, it has to be noted that in Secale cornutum oil the ricinoleic acid is esterified with glycerol, and the characteristic hydroxy group is not freely accessible because it is acylated with the usual long-chain fatty acids to tetra-acid, penta-acid, and hexa-acid triglycerides.13 Besides the high amount in Secale cornutum, ricinoleic acid is also a major component of castor oil.10,18 The contamination of rye with castor oil, however, is not to be expected. The fatty acid composition of rye and other cereal oils, on the other hand, differs considerably from that of Secale cornutum, mainly due to the absence of ricinoleic acid. With this prerequisite, the presence of ricinoleic acid is a good indicator for Secale cornutum impurities in rye, and the analysis of ricinoleic acid as a marker for Secale cornutum contaminations represents a meaningful analytical approach. The only method made available in the literature for ricinoleic acid analysis in rye was described by Franzmann et al.,10 who used gas chromatography coupled to flame ionization detection (GC−FID). The authors ascertained that the ricinoleic acid amount in rye correlates very well with the amount of Secale cornutum impurity. Consequently, the fatty acid was used as a marker substance and was analyzed for the determination of Secale cornutum contaminations in rye. After the hydrolysis of the starch with α-amylase and after saponification with sodium hydroxide in the presence of the internal standard 15-hydroxypentadecanoic acid in n-butanol, the organic phase of the acidified extract was dried, reconstituted in acidified methanol, and finally subjected to Received: Revised: Accepted: Published: 8246

August 27, October 2, October 4, October 4,

2016 2016 2016 2016 DOI: 10.1021/acs.jafc.6b03841 J. Agric. Food Chem. 2016, 64, 8246−8253

Article

Journal of Agricultural and Food Chemistry

(Darmstadt, Germany) and were used without prewashing. Rye samples (German type 997 rye flour, German type 1150 rye flour, whole rye flour, and whole rye) were purchased from local supermarkets. Standard and Derivatization Reagent Solutions. For the standard stock solutions, 4.5 mg of ricinoleic acid and ricinoleic acid methyl ester were individually dissolved in 10 mL of acetonitrile (450 mg/L). For method development, stock solutions were individually diluted 1:10 with acetonitrile, resulting in a concentration of 45 ng/μL. All standard solutions were stored at room temperature in a desiccator until use. For limit of detection and quantitation (LOD/LOQ) studies and for recovery experiments, aliquots of 4 and 25 μL of the diluted ricinoleic acid methyl ester solution were evaporated under a gentle stream of nitrogen at ambient temperature and derivatized as described in the Derivatization section, resulting in standard solutions for HPTLC of 0.18 and 1.125 ng/μL, respectively. Derivatizing reagent solution was prepared in anhydrous methylene chloride, containing DMAP and 2-NCl at concentrations of 40 and 7.5 mg/mL, respectively. The solution was freshly prepared for each series of analyses. Sample Preparation and Extraction. Rye flour, whole rye flour, and whole rye were finely milled in a Tube Mill control for 30 s at 12 000 min−1 (IKA, Staufen, Germany) before analysis. Secale cornutum was also finely milled and sieved (98%, analytical reagent grade) were purchased from Fisher Scientific (Schwerte, Germany). Sodium chloride (≥99%, Ph. Eur), acetone (Rotisolv pestilyse), and paraffin oil (low viscosity) (Ph. Eur.) were obtained from Carl Roth (Karlsruhe, Germany). Ultrapure water (>18 MΩ cm) was supplied by a Synergy System (Millipore, Schwalbach, Germany). Discovery Ag-Ion SPE tubes (polymerically bonded benzenesulfonic acid, coated with silver ions, 6 mL, 750 mg, 50 μm) were purchased from Sigma-Aldrich. Bondesil−PSA (primary secondary amine, 40 μm) was obtained from Varian (Palo Alto, U.S.A.), and SiliaBond carboxylic Acid (Si-CAA) (weak cationic exchanger (WCX), 40−63 μm) was purchased from SiliCycle Inc. (Québec, Canada). HPTLC silica gel 60 plates were from Merck 8247

DOI: 10.1021/acs.jafc.6b03841 J. Agric. Food Chem. 2016, 64, 8246−8253

Article

Journal of Agricultural and Food Chemistry Derivatization. Two hundred μL of derivatizing reagent solution was added to the sample or standard residue, the tube was tightly sealed, and the sample was mixed by vortexing for 10 s. To cover the entire glass inner surface with reagent solution, the tube was horizontally rotated by hand. The sample was stored at room temperature for 30 min. During the reaction time, the tube was vortexed twice for 10 s. After the addition of 800 μL of methylene chloride, the tube was briefly vortexed, 120 mg of PSA and 90 mg of Si-CAA were added, and the mixture was shaken for 1 min at 2200 rpm on a small-size shaking device. The tube was kept in the vertical position for 5 min, and a 200 μL aliquot of the clear methylene chloride solution was transferred into an HPTLC vial with a 400 μL insert. The final extract for HPTLC had a sample concentration of 25 mg/mL. The whole analytical procedure for the determination of ricinoleic acid in rye, including sample extraction, transesterification, cleanup, and derivatization, is outlined in Figure 1.

(CAMAG)), with an immersion speed of 2 and an immersion time of 3, and dried in a stream of cold air for 2 min. Plate images were captured with the TLC Visualizer Documentation System (CAMAG) under UV illumination at 254 nm and UV 366 nm. The plate was scanned by the TLC Scanner 4 (CAMAG) in the fluorescence mode at UV 280/>340 nm using the mercury lamp, with a scanning speed of 20 mm/s, a data resolution of 100 μm/step, and a slit dimension of 4 mm × 0.45 mm. The manual detector mode was selected, applying a quick scan range of 25−35 mm on the track of the highest standard of ricinoleic acid methyl ester. All HPTLC instruments were controlled by the winCATS software 1.4.6.2002 (CAMAG). Statistical Analysis. Statistical analysis was done with Microsoft Excel, version Office Professional Plus 2010. Calculation of 10-point calibration curves was performed by linear regression. Respective means, standard deviations, and relative standard deviations were calculated during validation.



RESULTS AND DISCUSSION A suitable derivatization for ricinoleic acid was evaluated first, before the parameters for HPTLC−FLD were developed. Thereafter, the sample workup was selected and optimized, regarding efficient lipid extraction, (trans)esterification, and cleanup for ricinoleic acid from the rye matrix. After the entire HPTLC−FLD screening was developed, the performance characteristics were assessed by LOD, LOQ, and the precision of the method. In a last step, several rye flours from the German market were analyzed to determine the present contamination with Secale cornutum. Method Development. Derivatization. With the intention to obtain the highest selectivity and sensitivity for ricinoleic acid determination, the characteristic hydroxyl group was used for fluorescence labeling. On the basis of procedures described in the literature,24−35 four commercially available fluorescent labeling reagents5-(dimethylamino)naphthalene-1-sulfonyl chloride (DANSCl), 9-fluorenylmethyl chloroformate (Fmoc), 2-naphthoyl chloride (2-NCl), and anthracene-9carbonyl chloride (ACCl)were tested for the derivatization of the standard with variations regarding the reaction conditions (solvent volume, catalyst, time, temperature, and shaking). For evaluation, the reaction yields were quantitated by HPTLC−FLD, taking into account the optimal detection settings for the respective derivatization product. Very soon, the free acid proved unsuitable due to the formation of several reaction products. Contrarily, a single product was obtained after derivatization of the corresponding methyl ester. Therefore, the methyl ester of ricinoleic acid was selected in this study as the most suitable analyte for derivatization. ACCl as reagent, reported to offer high sensitivity,32−34 proved unsuitable due to unacceptable low reaction yields, even after 24 h of reaction time. Likewise, the use of Fmoc and DANSCl led to unsatisfactory results; reaction yields were unacceptably low, independent of reaction conditions, or sensitivity was low due to unfavorable excitation/emission wavelengths. Contrarily, 2-NCl turned out to be most suitable for fluorescent labeling; the highest reaction yield and sensitivity of the corresponding naphthoate were obtained. Time dependency, influence of reaction temperature, and diverse base catalysts were evaluated for 2-NCl, and the stoichiometric ratio of ricinoleic acid methyl ester/2-NCl/DMAP was adjusted, because this influent parameter was frequently mentioned in the literature for the acylation of various compounds.24,25,32 The final retained procedure for derivatization consists of the addition of 200 μL of derivatization reagent (containing 7.5 mg of 2-NCl and 40 mg of DMAP/mL of methylene chloride) to the completely

Figure 1. Procedure for the determination of ricinoleic acid in rye. HPTLC−FLD. An Automatic TLC Sampler 4 (ATS 4, CAMAG, Muttenz, Switzerland) was used to apply samples and standard as 6 mm bands with the following settings leading to 22 tracks on a 20 cm × 10 cm plate: 8.0 mm distance from the lower edge, 8.0 mm distance from the left edge, and 8.7 mm track distance. Application parameters were set to 12 μL/s filling speed, 200 nL predosage volume, 200 nL retraction volume, 250 nL/s dosage speed, 8 s rinsing vacuum time, 0 s filling vacuum time, 3 rinsing cycles, and 1 filling cycle. Acetone was used as the rinsing solvent. Application volumes were 20 μL for sample extracts, 0.5−10 μL for ricinoleic acid methyl ester standard solution for LOD/LOQ determinations, and 0.5−30 μL for the standard for recovery experiments, respectively. For LOD/LOQ, application volumes resulted in 0.1−1.8 ng/zone, corresponding to 0.2−3.6 mg/ kg of rye, and for recovery experiments application volumes resulted in 0.6−33.8 ng/zone, corresponding to 1.1−67.5 mg/kg of rye, respectively, calculated for 20 μL sample application and expressed as ricinoleic acid. After the application, the start zones were dried at room temperature inside a fume hood for 10 min. Chromatography was performed in the Automatic Developing Chamber (ADC2, CAMAG) with a 20 cm × 10 cm twin-trough chamber (CAMAG). Before development, the plate activity was controlled for 5 min to 33% relative humidity with saturated MgCl2 solution. As the mobile phase, cyclohexane/diisopropyl ether/formic acid (86:14:1, v/v/v) was used up to a migration distance of 70 mm. Drying followed in a stream of cold air for 5 min. The developing time was 30 min. For fluorescence enhancement the plate was dipped into a solution of n-hexane/paraffin oil (2:1, v/v) (TLC Chromatogram Immersion Device III 8248

DOI: 10.1021/acs.jafc.6b03841 J. Agric. Food Chem. 2016, 64, 8246−8253

Article

Journal of Agricultural and Food Chemistry

Figure 2. Separation on silica gel plates with cyclohexane/diisopropyl ether/formic acid (86:14:1, v/v/v) as mobile phase up to a migration distance of 70 mm under UV 254 nm illumination after dipping in n-hexane/paraffin: (A) from left to right, ricinoleic acid (marked by arrow) with 10.4 and 25.9 ng/zone (corresponding to 0.02 and 0.05% Secale cornutum in rye) and Secale cornutum extracts with 10.4 and 25.9 ng of ricinoleic acid/zone (corresponding to 0.02 and 0.05% Secale cornutum in rye); (B) rye sample extracts (German type 1150 rye flour, finely milled whole rye flour, and finely milled whole rye, from left to right), prepared according to the method of Schulte36 and Christie,23 25 mg of rye/mL, 20 μL application volume, showing strong matrix interferences; (C) rye sample extracts prepared by the developed sample workup with the same sample amount and application volume as in (B); (D) Secale cornutum extracts prepared according to the developed sample preparation with the same amount and application volume as in (A); (E) reagent blank for comparison, 20 μL application volume.

dried sample, 30 min of reaction time at room temperature, and finally the addition of 800 μL of methylene chloride. However, for application volumes above 20 μL (for HPTLC−FLD), the zones of application were quite overloaded and poor chromatographic separation was obtained due to the high excess of fluorescent labeling reagent, directly visible under UV illumination at 254 and 366 nm. To ensure an effective reduction of the derivatization reagent in the final extract, a simple dispersive SPE (dSPE) with PSA/Si-CAA (120/90 mg/ mL) was used. After this procedure a regular and interferencefree chromatography was guaranteed. HPTLC−FLD. A ricinoleic acid methyl ester standard and a Secale cornutum extract (prepared according to the method of Schulte36 and Christie23 for lipid extraction and transesterification) were used for method development. Derivatization was performed according to the procedure described above. Silica gel and amino-modified HPTLC material were checked first for the separation of ricinoleic acid methyl ester and Secale cornutum matrix and additionally derivatization reagent, when silica gel offered the best separation. Different solvents and solvent mixtures were applied as mobile phase, taking into account the selectivity of the groups referred to Snyder.38 Separation was directly visible by the fluorescence under UV 254 nm and UV 366 nm illumination after dipping in n-hexane/paraffin for fluorescence enhancement. As a result of the screening study, HPTLC silica gel material delivered the best results with cyclohexane and diisopropyl ether as the mobile phase for selective separation of ricinoleic acid methyl ester from Secale cornutum matrix and the fluorescent labeling reagent. The addition of formic acid delivered a sharper standard zone, and a mixture of cyclohexane/diisopropyl ether/ formic acid (86:14:1, v/v/v) was finally selected for best separation. Plate activity was set to 33% relative humidity because separation and sharpness of the standard zone were influenced by the water content of the layer. These conditions finally resulted in a sharp zone of ricinoleic acid methyl ester at hRF of 33 without interferences for a migration distance of 70 mm shown by ricinoleic acid methyl ester and a Secale cornutum extract (Figure 2 A). Visual detection was feasible down to 5 ng/zone, expressed as ricinoleic acid, corresponding to ∼0.01% Secale cornutum, calculated on 20 μL of sample volume, which is well below the maximum limit for Secale cornutum in rye.7

Thereby, an easy initial estimation is offered whether the permitted level is exceeded. Quantitation was performed after fluorescence detection. The fluorescence spectra of the acylated ricinoleic acid methyl ester in remission mode showed maxima at 242 and 284 nm. Consequently, the intensive mercury lamp with the line of 280 nm was selected for the fluorescence measurements and yielded the highest signal intensity in combination with the 340 nm edge filter. Extraction and Cleanup. Applying HPTLC−FLD to several rye extracts, prepared according to the method of Schulte36 and Christie,23 it became obvious that the chosen workup was not suitable for a successful HPTLC−FLD screening of ricinoleic acid in rye matrix. Some matrix components were separated from ricinoleic acid, but the entire matrix load was too high, and a specific matrix component directly superimposed the analyte zone (Figure 2 B). Therefore, diverse sample workup strategies were evaluated. Different rye flour types were applied to cover the variability of the rye matrix, and ricinoleic acid and Secale cornutum were used as standards for comparison. For lipid extraction, the direct extraction with t-butyl methyl ether in the ultrasonic bath was tested. Several modifications concerning extraction time and conditions (temperature and sonication) were evaluated. Additionally, the method according to Franzmann et al.10 for the ricinoleic acid determination in rye was verified (without the SPE cleanup). Aliquots of the respective lipid or fatty acid extracts (corresponding to 100 mg of rye) were evaporated, and based on the method of Christie,23 transesterification was performed for 3 h with 1 mL of 1% sulfuric acid in methanol (v/v) at 80 °C on a thermal mixer at 800 rpm. To check the differences in matrix loads and recoveries, the extracts were derivatized and analyzed by HPTLC−FLD. High amounts of coextracted matrix compounds were present, independent of the extraction procedure, and were directly visible under UV illumination at 254 and 366 nm after dipping the plate into n-hexane/paraffin oil solution. A successful implementation of the HPTLC−FLD screening was not possible without applying an effective cleanup. Consequently, a suitable cleanup strategy was needed to remove the coextracted matrix. The method of Schulte36 offers a reliable lipid extraction and was therefore selected for evaluation of different purification strategies, while the success was again monitored by HPTLC−FLD after transesterification and 8249

DOI: 10.1021/acs.jafc.6b03841 J. Agric. Food Chem. 2016, 64, 8246−8253

Article

Journal of Agricultural and Food Chemistry

Figure 3. (A) HPTLC chromatogram on silica gel plates under UV 254 nm illumination of ricinoleic acid (marked by arrow) and extracts of Secale cornutum spiked rye samples (German type 1150), spiking level 0.02 and 0.05% Secale cornutum (n = 5), and a blank rye sample extract; (B) corresponding 3D densitogram of the fluorescence scan at UV 280/>340 nm, and (C) resulting calibration graph of ricinoleic acid displayed as 0.6− 33.8 ng/zone.

acylation. For separation of coextracted compounds, the removal of the unsaponified matter was often suggested in the literature.39,40 Methanolic potassium hydroxide was used for saponification, and different solvents and solvent mixtures were tested for the separation of the unsaponified matter by liquid−liquid partition. After acidification, the saponified substance was transferred into t-butyl methyl ether, and the result of the cleanup was evaluated. t-Butyl methyl ether effectively removed the matrix but also ricinoleic acid. In contrast, n-hexane and petroleum ether yielded high recoveries but the cleanup was not sufficient. Solvent mixtures either removed both the matrix and ricinoleic acid or recovery was sufficient but matrix separation was not achieved. With the aim of cleanup procedure being as simple as possible, dSPE was checked next. Graphitized carbon black (GCB) is known to remove planar molecules, especially chlorophyll, carotinoids, and sterols, whereas polyvinylpolypyrrolidone (PVPP) effec-

tively removes phenolic components. Therefore, a combination of both should help to minimize coextracted matrix. Dispersive SPE in the crude lipid extract, and also after removal of the unsaponified matter with petroleum ether, with several amounts of GCB and PVPP, both individually and in combination, did not offer any noticeable cleanup; matrix loads were identical to those of the respective lipid extracts without cleanup. In the literature, 10,19−22 aminopropyl cartridges were often reported to be useful for the separation of lipid classes. SPE according to Franzmann et al.10 and Kaluzny et al.19 was employed for lipid extracts after removal of the unsaponified matter. Additionally, relative composition and volume of rinsing and elution solvents were varied to optimize the cleanup. Unfortunately, in all tested versions aminopropyl material was not able to retain efficiently the high amount of matrix substances, resulting in large matrix loads remaining in the eluate, nearly equal to those of the lipid extraction 8250

DOI: 10.1021/acs.jafc.6b03841 J. Agric. Food Chem. 2016, 64, 8246−8253

Article

Journal of Agricultural and Food Chemistry according to Schulte36 without cleanup. SPE on Ag-ion cartridges is the method of choice for the separation of saturated and unsaturated fatty acids and also of cis/trans fatty acids.37,41−43 According to Shimelis et al.,37 Ag-ion SPE was tested for the transesterified lipid extracts. The separation of some matrix compounds was obtained, but the cleanup was not sufficient. Therefore, the relative composition and volume of the solvent mixtures for rinsing and elution (n-hexane/acetone) were optimized for the best separation of the used fatty acid methyl esters and further rye matrix components, keeping in mind the characteristic hydroxy function of the target analyte. To determine the selectivity, 10 fractions of 6 mL of eluate with an increasing content of acetone (0−40%) were collected individually. Finally, the volumes of the optimal rinsing and elution mixtures (n-hexane/acetone, 70:30, v/v, and 65:35, v/ v) were adjusted. The best results for both cleanup and recovery were obtained by rinsing the cartridge with 18 mL of n-hexane/acetone (70:30, v/v) followed by the elution with 7.5 mL of n-hexane/acetone (65:35, v/v). Matrix components were effectively removed, and ricinoleic acid methyl ester was entirely recovered (Figure 2 C and D). With the aim to determine interferences by contamination of the reagents, a reagent blank was additionally analyzed. Interferences by the SPE material were not detected, and excess of derivatizing reagent was only visible in hRF ranges not relevant for the determination of the ricinoleic acid methyl ester (Figure 2 E). As the result of the cleanup screening study, simple and rapid dSPE proved to be unsuitable for purification. Likewise, SPE on aminopropyl material showed too much remaining matrix, while Ag-ion SPE turned out to be most suitable for the separation of matrix components and the best recovery of ricinoleic acid. In the end, the sample preparation consists of a lipid extraction according to the method of Schulte,36 followed by a transesterification with methanol/sulfuric acid and a simple SPE on Ag-ion material. For sensitive and selective fluorescence detection of ricinoleic acid, an easy and rapid prechromatographic derivatization with 2-NCl was performed. The parameters for HPTLC−FLD were successfully developed for the determination of ricinoleic acid in rye, enabling a fast and easy screening for Secale cornutum contaminations. On a single plate, 22 samples (including calibration standards) were analyzed simultaneously, resulting in a run time of ∼5 min per sample and a solvent consumption of only 0.5 mL per sample. Method Validation. Sensitivity. Calibrations were performed in the range 0.1−1.8 ng of ricinoleic acid/zone, calculated to 0.2−3.6 mg of ricinoleic acid/kg of rye (n = 5) with regard to the sample preparation and the sample application volume of 20 μL. Resulting calibration graphs showed good linearity with high coefficients of correlation (R2 > 0.9990). According to the DIN 32645 calibration method,44 LOD and LOQ were calculated to be 0.06 and 0.2 ng of ricinoleic acid/zone, respectively, corresponding to 0.1 and 0.4 mg of ricinoleic acid/kg of rye. With %RSDs of 3.0% and 2.9%, respectively, the determination was well repeatable. According to the results reported in the literature,10 the low variation of the ricinoleic acid content in Secale cornutum enables a reliable determination of Secale cornutum impurities by analyzing the quantity of ricinoleic acid. Because of this firm relationship, LOD and LOQ were calculated to be 0.0001 and 0.0004% for Secale cornutum in rye, taking into account the sample workup and a ricinoleic acid amount in Secale cornutum of 10.4%.

Consequently, the developed HPTLC−FLD screening enables the quantitation of Secale cornutum impurities in rye far below the currently set maximum limit of 0.05% for certain unprocessed cereals.7 Recoveries. Recovery experiments for ricinoleic acid in rye flour were carried out at levels of 0.02 and 0.05% Secale cornutum in rye (n = 5), calculated and expressed as ricinoleic acid. The ricinoleic acid quantity in the spiked sample was easily calculated by the previously determined ricinoleic acid content in the Secale cornutum that was used for spiking. Recovery experiments were performed three times on different days with different rye flour types (flour of the German type 1150, finely milled whole rye flour, and finely milled whole rye) to cover the variability of the rye matrix. Quantitation by HPTLC−FLD using a ricinoleic acid methyl ester solvent standard is exemplarily shown in Figure 3 for rye flour of the German type 1150 at levels of 0.02 and 0.05% Secale cornutum. For all rye types, the recovery rates were near to 100%, but in most cases exceeded 100% (Table 1). The highest recoveries Table 1. Recoveries of Ricinoleic Acid from Rye Spiked with Fine-Milled and Sieved (