913Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... DOI: 10.17344/acsi.2021.6891 Scientific paper Determination of Some Elements in Legumes Using ICP-MS and EDXRF Methodology Applications Lovro Sinkovič,1,* Marijan Nečemer,2 Barbara Pipan1 and Vladimir Meglič1 1 Crop Science Department, Agricultural Institute of Slovenia, Hacquetova ulica 17, SI-1000 Ljubljana, Slovenia 2 Department of Low and Medium Energy Physics, “Jožef Stefan” Institute, Jamova 39, SI-1000 Ljubljana, Slovenia * Corresponding author: E-mail: lovro.sinkovic@kis.si phone: +386 (0)1280 52 78 Received: 08-16-2021 Abstract The current study involves two analytical research techniques, inductively coupled plasma-mass spectrometry (ICP-MS) and energy dispersive X-ray fluorescence (EDXRF) spectroscopy, used to determine the elemental composition of differ- ent legumes usually produced and consumed in Slovenia. Results indicate that data obtained using these methods are in agreement with certified reference materials. In total, nineteen elements were determined from twenty legume samples. An intercomparison between four macro- (P, S, K, Ca) and three microelements (Fe, Zn, Mo) measured using ICP-MS and EDXRF methods showed a strong correlation. The EDXRF was found to be a cheaper, simpler and more environ- mentally friendly method for determination of elements P, S, Cl, K, Ca, Fe, Zn, Mo, Sr, Rb, Ti and Br in legumes, while for the identification and determination of Na, Mg, V, Cr, Mn, Co and Cu content ICP‐MS was the method of choice due to its excellent sensitivity and accuracy. Using principal component analysis (PCA), the samples of the studied legumes were classified into four groups according to their elemental composition. Keywords: EDXRF; elemental composition; ICP-MS; legumes. 1. Introduction Legumes are of prime importance in human or ani- mal nutrition with a great variety of plants.1 These include crops grown for grains (e.g. common bean, runner bean, lupins, lentil, chickpea), fresh vegetables (e.g. snap bean, green pea) and livestock forage (e.g. soybean, field pea). Legumes play a distinct role in agricultural ecosystems with their ability to fix nitrogen symbiotically.2 Grains of food legumes are an important source of elements such as P, Ca, K, N, Fe, Mg and Zn, essential for a human well-be- ing. Legume grains not only play a vital role in many tra- ditional diets worldwide but are valuable for the food and animal feed industries.3 The common bean is the most important grain legume for direct human consumption which provides 10 – 20% of the adult requirements for nu- trients, namely Fe, P, Mg, Mn, and to a lesser degree for Zn, Cu and Ca.4 Chickpea applies the third most impor- tant grain legume for human consumption after beans and peas.5 According to the FAO, in 2019 the total world pro- duction was the highest for soybeans, followed by beans (Phaseolus spp.), peas, chickpeas, lentils, faba beans and lupins.6 The main objective in multi‐elemental analysis of foods is to ensure food quality and safety. Therefore, with the increased crop production to meet the growing de- mands, product quality becomes an important issue.7 El- emental fingerprinting has been proven effective way for quality and authentication of foods.8 Besides, it serves as an important tool for plant breeding programmes and nu- tritional biofortification purposes.9,10,11 Around twenty of the known elements are defined as essential since they act as important phytochemicals and have a significant role in the maintenance of human health. Essential elements are classified into macroelements or major minerals, namely Na, K, Mg, Ca, Cl, P and S, and microelements or trace minerals such as Fe, Cu, Mn, Zn, Co and Mo.12 The levels of these elements are measured to provide valuable nu- tritional information about foods. Over the past decade, with the development of advanced analytical techniques, the element composition of various food samples can be 914 Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... successfully measured using inductively coupled plas- ma-mass spectrometry (ICP-MS), energy dispersive X-ray fluorescence (EDXRF) spectroscopy, among others.13 Al- though ICP-MS is frequently the most accurate technique for elemental analysis due to the simplicity, easily inter- preted spectra and the exceptionally low limits of detec- tion, on the other hand, requires expensive reagents, gases and laborious sample preparation.8 EDXRF spectroscopy is a good alternative to ICP-based methods, in which sam- ple digestions are needed since hardly any sample treat- ment is required to carry out multi-elemental analysis in solid samples. Despite huge progress made in EDXRF instruments few works were published in the field of ele- mental characterization of organic matrices such as plant foods using this technique. Nevertheless, EDXRF is char- acterized by detections limits at the low mg/kg level, which are about three orders of magnitude higher than those ob- tained by the ICP-MS method.14 Within the framework of the current studies at the Agricultural Institute of Slovenia, the capabilities of two analytical techniques, ICP-MS and EDXRF spectroscopy, were compared through the determination of elements in several legume samples. The present research aims to provide data on elemental composition of legumes usually produced and consumed in Slovenia which can be used in the ongoing plant breeding programs (e.g. common bean). The purpose of the study was: (i) to determine the multi-elemental composition of different legumes; (ii) to assess the appropriateness of two analytical methods ICP- MS and EDXRF for element determination of different legumes, thus (iii) evaluating the correlation between used analytical methods. 2. Experimental 2. 1. Materials A set of twenty homogenised plant samples consists of several legume species as presented in Table 1. Analysed samples were as follows: common bean grains (6 samples), common bean pods (3  samples), lupin (3  samples), faba bean (2 samples), lentil (2 samples), chickpea (1 sample), soybean (1 sample), field pea (1 sample) and runner bean (1 sample). Most of the analysed legumes were produced at the experimental fields of Infrastructure Centre Jablje at the Agricultural Institute of Slovenia (304 m a.s.l.; 46.151°N 14.562°E). Chickpea, brown and red lentil samples were purchased from the Slovenian food retail market. The leg- ume grain samples were air-dried after harvest to reduce the moisture content, to levels below 11%. The common bean pods were immediately after harvesting frozen us- ing liquid nitrogen and lyophilized. Before determination of elements, all legume samples were homogenised and powdered using a laboratory ball mill (Retsch MM 400, GmbH) at a high frequency of 30 Hz for 2–5 min. 2. 2. Determination of Elements Using ICP-MS Digestion of plant samples (decomposition of organic matter). The powdered legume samples were digested us- ing a high-pressure microwave oven (Milestone ETHOS 1600). Separate samples were weighed (250 mg) into PTFE vessels and 6 mL of 65% nitric acid (HNO3, SUPRAPUR, Merck) and 2  mL 30% hydrogen peroxide (H2O2, SU- PRAPUR, Merck) was added. The digestion was conduct- Table 1. List of studied legume samples and their origin Sample name Legume species Latin name Variety/ Seed provider Sample accession type KIS_GL1 common bean Phaseolus vulgaris L. Ribnčan Semenarna Ljubljana grains KIS_GL2 common bean Phaseolus vulgaris L. SRGB204 Slovenian Plant Gene Bank grains KIS_GL3 common bean Phaseolus vulgaris L. Zorin Semenarna Ljubljana grains KIS_GL4 common bean Phaseolus vulgaris L. Ribnčan Semenarna Ljubljana pods KIS_GL5 common bean Phaseolus vulgaris L. SRGB204 Slovenian Plant Gene Bank pods KIS_GL6 common bean Phaseolus vulgaris L. Zorin Semenarna Ljubljana pods KIS_GL7 common bean Phaseolus vulgaris L. Etna Semenarna Ljubljana grains KIS_GL8 common bean Phaseolus vulgaris L. Golden gate Semenarna Ljubljana grains KIS_GL9 common bean Phaseolus vulgaris L. SRGB196 Slovenian Plant Gene Bank grains KIS_GL10 white lupin Lupinus albus L. Energy Feldsaaten Freudenberger grains KIS_GL11 faba bean Vicia faba L. var. minor Zoran Agricultural Institute of Slovenia grains KIS_GL12 faba bean Vicia faba L. var. minor Merkur Semenarna Ljubljana grains KIS_GL13 blue lupin Lupinus angustifolius L. Sonet Feldsaaten Freudenberger grains KIS_GL14 yellow lupin Lupinus luteus L. Mister Feldsaaten Freudenberger grains KIS_GL15 red lentil Lens culinaris Medik. rdeča leča food retail market grains KIS_GL16 brown lentil Lens culinaris Medik. rjava leča food retail market grains KIS_GL17 chickpea Cirer arietinum L. čičerika food retail market grains KIS_GL18 soybean Glycine max L. Merr. ES Mentor Saatbau Slovenia grains KIS_GL19 field pea Pissum sativum L. Eso Semenarna Ljubljana grains KIS_GL20 runner bean Phaseolus coccineus L. SRGB222 Slovenian Plant Gene Bank grains 915Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... ed according to the following programme: step 1, 300 W, 5 min at 100 °C; step 2, 400 W, 5 min at 150 °C; step 3, 500 W, 5 min at 180 °C; step 4, 600 W, 7 min at 210 °C; step 5, 550 W, 15 min at 210 °C; step 6, 0 W, 20 min cool- ing. Preparation of test solution. Digested samples were cooled to room temperature and the solution was quan- titatively transferred into 50  mL plastic tubes (Sarstedt, USA) and filled up to full volume with Milli-Q water. Be- fore analysis, the digested samples were diluted by a factor of 20 and consisted of 1% HNO3 (v/v). Determination of elements by ICP-MS. For the deter- mination of elements in samples, the Agilent 7900 ICP-MS was used. This instrument includes a 4th generation col- lision/reaction cell, the Octopole reaction System (ORS4) which provides optimized operational conditions for the removal of polyatomic interferences using helium (He) collision mode. In this manner, smaller and faster analyte ions are separated from larger, slower interference ions us- ing kinetic energy discrimination. The following isotopes were monitored: 23Na, 24Mg, 31P, 34S, 39K, 43Ca, 51V, 52Cr, 55Mn, 56Fe, 59Co, 63Cu, 66Zn and 95Mo. Due to the high sen- sitivity of the 7900 ICP-MS most elements of interest could be measured in He mode, only phosphorus (P) and sul- phur (S) were measured in the high helium (HEHe) mode. The ICP-MS operating conditions were optimized using autotuning functions within the ICP-MS MassHunter software. Other instruments operating conditions were as follows: general-purpose plasma mode; peri-pump sample introduction; micro-mist nebuliser; nickel cones interface; He gas flow was 5 mL/min in He mode and 10 mL/min in HEHe mode; spectrum acquisition mode; one point peak pattern; three replicates and 100 sweeps per replicate. Calibration of the instrument. Quantitative analy- sis was performed with the external calibration method. Calibration standards for most of the elements were pre- pared using IV-STOCK-50 standard solution containing: 1000 mg/L of Na, Mg, K, Ca and Fe and 10 mg/L of V, Cr, Mn, Co, Cu, Zn and Mo (matrix 5% v/v HNO3, Inorganic Ventures). Phosphorus (1000 mg/L P, matrix H2O, CGP1, Inorganic Ventures) and sulphur (1000  mg/L S, matrix H2O, CGS1, Inorganic Ventures) single standard solutions were added separately to the mixture. A five-point cali- bration from 0.05 µg/L to 50 µg/L was carried out for the elements V, Cr, Mn, Co, Cu, Zn and Mo, and from 5 µg/L to 5000 µg/L for the elements Na, Mg, S, P, K and Fe. Only for Ca a six-point calibration was used between 5 µg/L and 10000 µg/L due to the higher concentrations expected in plant samples. Final multi-element calibration solutions were prepared daily and contained 1% HNO3 (v/v). The internal standards used to compensate for sensitivity drift and matrix effects during the analytical run were Sc, Rh, In and Lu (Agilent PN 5188-6525). They were added online in a concentration of 200 µg/L (in 1% v/v HNO3). Quality control. For quality control analytical blanks, independent QC standards and standard reference mate- rial (SRM) were used. In each test series, a blank sample containing only acids was included. The QC standards were prepared in the concentrations of 2 µg/L, 20 µg/L and 2000 µg/L over the analytical range from ICP-MS mul- ti-standard solution VIII (MERCK, Certipur, PN 1.09492) and ICP-MS multi-standard solution XVI (MERCK, Cer- tipur, PN 1.09487) and were analysed in the beginning, in the middle and at the end of each analysis run. Finally, a National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) Tomato Leaves (1573a) was used to check the accuracy of the analytical procedures and recovery. Analytical data was processed using the ICP-MS MassHunter Workstation Software (Rev. C.01.02, G7201C, Agilent technologies, 2015). The software calculates the correlation coefficient of the cali- bration curve (R), the limit of detection (LOD) and the background equivalent concentration (BEC) for each el- ement. Data along with the accuracy (as % recovery) is shown in Table 2 and accuracy data using NIST SRM To- mato Leaves (1573a) in Table 3. The data are expressed as macro- (g/kg) or microelements (mg/kg). Table  2. ICP-MS calibration coefficients, method detection limits and background equivalent concentration data Element R2 LOD BEC Recovery (mg/kg) (mg/kg) (%) 23Na 1.0000 1.07 6.63 82.5 24Mg 1.0000 0.53 0.74 84.9 31P 1.0000 10.8 8.2 92.7 34S 1.0000 223 3360 95.7 39K 1.0000 6.8 318.0 91.4 43Ca 1.0000 48 125 89.9 51V 1.0000 0.0001 0.0041 71.8 52Cr 1.0000 0.085 0.342 86.6 55Mn 1.0000 0.038 0.093 90.9 56Fe 1.0000 0.021 3.522 86.3 59Co 1.0000 0.003 0.035 85.8 63Cu 1.0000 0.041 0.160 98.5 66Zn 1.0000 0.254 0.759 89.7 95Mo 1.0000 0.010 0.217 73.3 R, calibration coefficient; LOD, method detection limit; BEC, back- ground equivalent concentration. 2. 3. Determination of Elements Using EDXRF Identification of twelve elements (P, S, Cl, K, Ca, Ti, Fe, Zn, Br, Rb, Sr, Mo) in a single measurement was car- ried out using non-destructive EDXRF spectrometry. Pel- lets prepared from 0.5 g to 1.0 g of powdered legume sam- ples were set using a pellet die and a hydraulic press. The disc radioisotope excitation sources Fe-55 (25  mCi) and Cd-109 (20 mCi) from Eckert & Ziegler (Germany) was used for fluorescence excitation. The emitted fluorescence 916 Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... radiation was measured using the EDXRF spectrometer with an XR-100 SDD detector (Amptek), a PX5 digital pulse processor (Amptek), and a PC-based, multichannel analyser software package DPPMCA. In Fe-55 mode, the spectrometer was equipped with a vacuum chamber to measure the three light elements (P, S, Cl), and in Cd-109 mode, the multi-element measurement was performed in the air for the nine elements K, Ca, Ti, Fe, Zn, Br, Rb, Sr and Mo. The energy resolution setting of the spectrom- eter was 125eV at 5.9keV. The analysis of complex X-ray spectra was performed using the AXIL spectral analysis program.15 The evaluated uncertainty of this procedure included the statistical uncertainty of measured intensities and the uncertainty of the mathematical fitting procedure. The overall uncertainty of spectral measurement and anal- ysis was in most cases better than 1%. Quantification was performed using the QAES (Quantitative Analysis of Environmental Samples) soft- ware.16 The estimated uncertainty of the analysis was around 5% to 10%, LOD for Zn was from 5 μg/g to 10 μg/g. A high total estimated uncertainty is mainly due to con- tributions of matrix correction and geometry calibration procedures, which include errors of tabulated fundamen- tal parameters, and also contributions of spectrum acqui- sition and analysis. The uncertainty due to the inhomoge- neity of the sample was not included. Accuracy of the data using the 1573a (Tomato Leaves) is presented in Table 3. The quantified twelve elements were expressed as macro- (g/kg) or microelements (mg/kg). 2. 4. Statistical Evaluation Statistical calculations and multivariate analysis were carried out using the XLSTAT software package (Addin- soft, New York, USA). The multivariate analysis involved principal component analysis (PCA). 3. Results and Discussion A common characteristic of both analytical tech- niques applied ICP-MS and EDXRF is their multi-element capability. Preparation of legume samples was simple and non-destructive in the case of EDXRF, while, ICP-MS re- quired skilled personnel and decomposition of samples. ICP-MS was a more sensitive method in this study with LODs in the range of ng/g. Results of the LOD, BEC and accuracy of multi-elemental determination performed by ICP-MS with certified reference material (NIST SRM Toma- to Leaves 1573a) are presented in Tables 2 & 3. The sensitiv- ity of EDXRF according to estimated uncertainty was from 5% to 10% and LODs for the analysed elements in the range from a hundred to a few μg/g. This means that LODs of ICP-MS were approximately two orders of magnitude lower compared to EDXRF. The determination of element Cl by ICP-MS was impossible since it forms negative ions, while EDXRF enables its determination as an essential element (dietary mineral) being one of the main electrolytes in the body. On the other hand, ICP-MS enables the determination of elements Na, Mg, Mn and Cu which are according to the European Food Safety Authority (EFSA) essential required substances as nutrients necessary by the body to perform a variety of functions. Considering the cost per sample, EDX- RF was cheaper, simpler and more environmentally friendly when compared to ICP-MS and much more suitable for de- termination of P, S, Cl, K, Ca, Fe, Zn, Mo, Sr, Rb, Ti and Br in legume samples. However, for determination of Na, Mg, V, Cr, Mn, Co and Cu content ICP‐MS was a method of choice due to its excellent sensitivity and accuracy. The macroelement composition of twenty legume samples determined by ICP-MS and/or EDXRF is present- ed in Table 4 and the microelement composition in Table 5. A total of nineteen elements were determined and divided into six macro- (Mg, P, S, K, Cl, Ca) and thirteen microe- lements (Mn, Fe, Cu, Na, Cr, Co, Zn, V, Rb, Ti, Br, Sr, Mo). The results of macroelements are expressed as g/kg (Table 4) and those of microelements as mg/kg (Table 5). Based on the average values the order is K > P > S > Ca > Mg > Cl of for the macro- and Fe > Mn > Zn > Na > Cu > Rb > Br > Ti > Sr > Mo > Cr > Co > V for the microelements in analysed legume samples. The ranges of individual macroelements in the analysed legumes were as follows: K (7.5–22.6 g/kg), Table 3. Accuracy of the data using the National Institute of Stand- ards and Technology (NIST) Standard Reference Material (SRM) Tomato Leaves (1573a) Element NIST SRM Tomato Leaves (1573a) EDXRF ICP-MS Certified results results values g/kg K 27.80 ± 0.28 24.47 ± 0.04 26.76 ± 0.48 P 1.930 ± 0.023 2.004 ± 0.001 2.161 ± 0.028 S 8.82 ± 0.90 9.19 ± 0.18 9.60* Ca 50.10 ± 0.49 45.37 ± 0.07 50.45 ± 0.55 Mg / 10.18 ± 0.01 12.00* Cl 6.50 ± 0.65 / 6.60* mg/kg Fe 353.0 ± 3.6 317.3 ± 1.5 367.5 ± 4.3 Mn 246.1 ± 2.8 223.8 ± 0.6 246.3 ± 7.1 Zn 29.10 ± 3.01 27.75 ± 0.27 30.94 ± 0.55 Na / 112.3 ± 0.2 136.1 ± 3.7 Cu / 4.63 ± 0.02 4.70 ± 0.14 Rb 15.90 ± 0.21 / 14.83 ± 0.31 Br 12.70 ± 1.30 / 13.00* Sr 83 ± 9 / 85* Mo / 0.34 ± 0.1 0.46* Cr / 1.721 ± 0.011 1.988 ± 0.034 Co / 0.4952 ± 0.0001 0.5773 ± 0.0071 V / 0.599 ± 0.005 0.835 ± 0.034 *noncertified values 917Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... P (2.7–7.5  g/kg), S (0.9–4.7  g/kg), Ca (0.2–5.0  g/kg), Mg (0.8–2.9 g/kg), and Cl (0.1–51.0 g/kg). The highest coeffi- cient of variation was calculated for the Ca (66.49%), fol- lowed by Cl (58.30%) and S (38.40%). Among different leg- ume species, the highest concentrations of K was found in common bean pods (snap beans) and soybean grains and the lowest for lentil. The P concentration was the highest for faba bean and the lowest for chickpea. The lupins and soybean contained the most S, and common bean pods the most Ca. When compared with other legumes, lentil had the lowest concentration of Ca and Mg. The ranges of es- sential microelements in the analysed legumes were the fol- lowing: Fe (37.2–126.0 mg/kg), Mn (8.3–487.7 mg/kg), Zn (17.5–64.6 mg/kg), Na (1.7–91.8 mg/kg), Cu (4.9–16.3 mg/ kg), and Mo (0.2–8.5 mg/kg). Among determined microe- lements the highest coefficients of variation were calculated for the Mn, Cr and Na (> 102.12%) while the lowest for Cu, Zn and Fe (< 38.72%). The highest concentration of Fe was found for soybean and the lowest for field peas. All three lupin samples (white, blue and yellow) were the richest source of microelement Mn compared to other legumes. The Na concentration was the highest for common bean pods and lupins. These elemental compositions are consist- ent with literature data for common bean grains17,18 and pods,19,20 lupin,21,22 faba bean,22,23 lentil,3,24,25 chickpea,3,5 soybean,21 field pea25 and runner bean.26,27 The EFSA set dietary reference values for the fol- lowing fourteen essential elements Ca, Cl, Cu, F, I, Fe, Mg, Mn, Mo, P, K, Se, Na and Zn. Using laboratorial developed ICP-MS multi-element method for plant samples ten (Ca, Cu, Fe, Mg, Mn, Mo, P, K, Na, Zn) of these fourteen ele- ments were determined. The three elements, namely Cl, F and I, belong to halogen elements and for these due to the presence of interferences formed in the Ar plasma oper- ating under conventional conditions or due to the matrix effects, their determination by ICP-MS is still considered a challenging task. Despite quadrupole ICP-MS is by far the most used instrumentation for multi-element determi- nation, there are some drawbacks particularly associated with its application for halogen determination.28 Further- more, the Se is another element which determination by ICP-MS remains particularly difficult. For Se analyses, conventional quadrupole ICP-MS operation suffers from inadequate sensitivity due to the high ionization potential of Se in the plasma as well as isobaric and polyatomic in- terferences. Selenium has six stable isotopes (74Se 0.87%, 76Se 9.02%, 77Se 0.58%, 78Se 23.52%, 80Se 49.82%, 82Se 9.19%), which adds complexity to the analysis.29 When using the EDXRF multi-element method seven (Ca, Cl, Fe, Mo, P, K, Zn) of fourteen above mentioned elements was determined. However, EDXRF is a non-destructive and simultaneous method with simple sample preparation steps.30 After the sample preparation procedure, the EDX- RF analysis presented high analytical frequency and most equipment counts on an automatic sample holder.31 Out of a total of nineteen elements determined us- ing ICP-MS and EDXRF, the following seven P, S, K, Ca, Fe, Zn and Mo were determined by both methods. The Table 4. Macroelement composition of 20 legume samples determined by ICP-MS and EDXRF Sample Legume Macroelements (g/kg) name species K P S Ca Mg Cl ICP-MS ED XRF ICP-MS ED XRF ICP-MS ED XRF ICP-MS ED XRF ICP-MS ED XRF KIS_GL1 CB (grains) 13.9 12.8 4.8 5.0 2.1 1.9 1.0 0.8 1.4 0.2 KIS_GL2 CB (grains) 19.1 17.3 5.9 5.9 2.3 1.9 1.2 1.0 1.6 0.3 KIS_GL3 CB (grains) 14.4 13.1 5.0 4.6 2.2 1.4 0.9 0.9 1.5 0.1 KIS_GL4 CB (pods) 17.1 18.3 3.1 3.5 1.4 1.2 2.5 2.2 1.7 0.7 KIS_GL5 CB (pods) 18.5 18.7 3.1 3.9 1.7 1.5 4.1 3.3 2.0 0.6 KIS_GL6 CB (pods) 20.9 22.6 3.7 4.7 1.5 1.8 5.0 3.7 2.1 1.0 KIS_GL7 CB (grains) 13.5 13.4 4.2 4.5 2.0 1.5 1.3 1.1 1.5 0.1 KIS_GL8 CB (grains) 16.3 15.4 5.9 5.8 2.4 2.1 1.2 1.1 1.7 0.3 KIS_GL9 CB (grains) 17.8 15.9 5.9 5.6 2.2 1.7 0.8 0.9 1.7 0.1 KIS_GL10 white lupin 15.3 11.9 5.3 6.2 2.4 2.5 3.0 1.8 1.7 0.1 KIS_GL11 faba bean 12.5 11.4 6.5 6.5 1.5 0.9 1.1 0.9 1.4 0.6 KIS_GL12 faba bean 12.7 12.9 6.2 6.5 1.6 1.1 1.0 1.0 1.4 0.5 KIS_GL13 blue lupin 11.1 10.4 4.9 4.9 2.2 2.0 2.2 2.1 1.8 0.2 KIS_GL14 yellow lupin 12.1 11.6 7.5 7.0 4.7 4.6 1.9 1.6 2.9 0.6 KIS_GL15 red lentil 9.2 9.1 3.7 3.5 1.8 1.5 0.2 0.2 0.8 0.5 KIS_GL16 brown lentil 7.8 7.5 3.1 3.1 1.6 1.2 0.8 0.8 0.8 0.7 KIS_GL17 chickpea 10.1 9.4 2.9 2.7 2.1 1.4 1.3 1.1 1.3 0.6 KIS_GL18 soybean 18.2 17.1 5.5 4.9 3.0 2.5 1.8 1.6 2.4 0.1 KIS_GL19 field pea 9.1 8.8 3.7 3.6 1.6 1.3 0.9 0.8 1.3 0.5 KIS_GL20 runner bean 17.8 16.3 4.6 5.1 2.0 2.3 1.1 1.2 1.8 0.5 Correlation 0.96 0.94 0.94 0.98 n.a. n.a. Coefficient of variation (%) 26.85 25.54 38.40 66.49 29.27 58.30 CB, common bean; n.a., not applicable. 918 Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... correlations between these two methods for the macroe- lements P, S, K and Ca are presented in Table 4 and for the microelements Fe, Zn and Mo in Table 5. A very strong correlations were detected for the elements Ca (0.98), Zn (0.97), K (0.96), Mo (0.95), P and S (0.94), while strong correlation was observed for Fe (0.72). For the elements K, S, Ca and Zn brought to evidence a tendency that EDXRF values were lower, when compared with ICP-MS values. Phosphorus values were comparable between methods while Fe EDXRF values were higher compared to ICP-MS values. These discrepancies can be due to the inhomoge- neity of the samples. Table 5. Microelement composition of 20 legume samples determined by ICP-MS and EDXRF Sample Legume Microelements (mg/kg) name species Fe Mn Zn Na Cu Rb Br Ti Sr Mo Cr Co V ICP- ED ICP- ICP- ED ICP- ICP- ED ED ED ED ICP- ED ICP- ICP- ICP- MS XRF MS MS XRF MS MS XRF XRF XRF XRF MS XRF MS MS MS KIS_GL1 CB (grains) 64.6 69.2 12.0 28.1 26.6 6.1 8.7 5.9 1.3 1.6 0.9 0.6 0.9 0.2 0.03 0.003 KIS_GL2 CB (grains) 69.9 79.0 12.4 27.9 26.7 3.3 9.5 3.7 1.2 2.2 0.9 3.2 1.9 0.1 0.04 0.000 KIS_GL3 CB (grains) 57.7 91.6 11.0 28.9 22.3 2.4 5.9 9.0 1.7 3.5 1.1 0.9 0.9 0.3 0.03 0.004 KIS_GL4 CB (pods) 54.0 65.8 13.6 21.5 17.5 67.2 7.1 6.2 6.9 3.5 3.0 0.4 1.1 0.1 0.03 0.026 KIS_GL5 CB (pods) 44.2 76.6 14.7 22.0 23.9 53.5 5.8 3.9 2.9 4.4 4.1 0.2 1.3 0.1 0.02 0.002 KIS_GL6 CB (pods) 50.6 68.0 16.7 24.2 25.4 91.8 4.9 8.2 13.9 3.3 4.8 0.4 1.8 0.2 0.02 0.014 KIS_GL7 CB (grains) 77.3 84.0 13.6 26.8 25.1 3.0 8.3 2.4 1.5 3.0 4.6 5.8 4.5 0.2 0.04 0.045 KIS_GL8 CB (grains) 58.7 76.0 11.9 27.6 27.1 38.3 5.8 7.8 1.8 2.4 3.1 8.5 6.6 0.2 0.15 0.019 KIS_GL9 CB (grains) 56.9 84.7 9.4 20.3 19.4 1.7 5.2 6.7 1.6 3.8 0.8 4.7 3.5 0.2 0.05 0.008 KIS_GL10 white lupin 37.2 57.5 487.7 42.4 48.6 58.8 8.0 23.3 2.0 2.0 2.0 2.2 1.6 0.1 0.09 0.008 KIS_GL11 faba bean 43.3 74.7 16.4 49.1 43.1 9.0 16.3 5.6 1.3 2.0 0.9 0.9 1.1 0.2 0.20 0.005 KIS_GL12 faba bean 46.9 84.9 16.8 44.6 43.0 10.0 16.3 3.4 1.5 2.1 1.8 1.0 0.9 0.2 0.16 0.007 KIS_GL13 blue lupin 43.9 58.3 79.5 34.9 33.1 58.6 6.1 8.9 3.5 2.4 5.7 1.7 1.4 1.7 0.04 0.031 KIS_GL14 yellow lupin 64.9 95.6 82.9 63.3 64.6 78.3 10.7 3.3 2.1 4.1 2.2 1.4 1.1 2.9 0.07 0.022 KIS_GL15 red lentil 65.9 87.6 13.6 33.0 33.0 10.7 9.1 4.1 3.8 2.7 1.8 5.3 4.4 0.1 0.03 0.008 KIS_GL16 brown lentil 47.3 70.8 9.6 19.6 22.1 6.9 8.1 3.1 1.3 2.8 1.8 0.5 0.9 0.2 0.04 0.006 KIS_GL17 chickpea 55.5 71.1 25.8 28.6 26.0 33.8 8.0 6.7 10.8 0.6 5.5 2.7 1.7 1.0 0.12 0.027 KIS_GL18 soybean 86.0 126.0 22.7 36.8 35.9 10.3 13.4 7.6 6.6 3.7 2.0 0.7 2.0 2.7 0.13 0.034 KIS_GL19 field pea 46.8 79.3 8.3 26.3 24.3 16.2 7.3 4.9 1.6 1.1 1.8 0.6 1.0 0.2 0.03 0.027 KIS_GL20 runner bean 52.7 77.6 12.6 23.0 20.8 1.8 5.4 3.4 2.6 1.0 2.6 2.4 1.9 0.3 0.04 0.013 Correlation 0.72 n.a. 0.97 n.a. n.a. n.a. n.a. n.a. n.a. 0.95 n.a. n.a. n.a. Coefficient of variation (%) 25.96 232.72 35.83 102.12 38.72 68.38 97.13 39.79 59.72 89.18 146.10 77.30 79.13 CB, common bean; n.a., not applicable. Fig. 1. PCA score plot (left) and discriminant loadings plot (right) 919Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... Statistical evaluation of results was performed on the dataset using PCA to identify the elements responsible for differentiating legume samples. The discrimination score plot and loadings across the original dataset are shown in Fig. 1. When the PCA was applied to the data (20 samples, 19 variables), discriminant functions were obtained where Function  1 explained 22.31% and Function  2 21.90% of the total variance. According to the multi-elemental com- position, the analysed legumes can be divided into four groups as seen in Fig. 1. The first group included samples of common bean pods and chickpea with the most influ- ential elements Ca, Na, Sr and Cl. Second group involved soybean and lupin samples, where the most influential el- ements were Na, Mg, S, Zn and P. The third group consist- ed of runner bean, field pea and lentil samples with most influential elements Mg, S, Zn and P. Finally, the faba bean and common bean grain samples formed the fourth group with most influential elements Mo, Cu and Co. 4. Conclusion There is a need for reliable legumes element con- centration data to provide information about nutritional uptake, especially for essential macro- and microelements. Besides, the multi-elemental composition, it provides im- portant information for the plant breeding programmes and can potentially serve as a base for nutrient bioforti- fication purposes (e.g. Fe, Zn). This paper provides some interesting comparisons between two different techniques (ICP-MS, EDXRF) in the determination of the multi-el- emental composition of different legumes. An intercom- parison of seven elements (P, S, K, Ca, Fe, Zn and Mo) showed satisfactory agreement between both methods. The simple, fast and cheaper EDXRF method when com- bined with ICP-MS, was found the most appropriate tech- nique for determination of Na, Mg, V, Cr, Mn, Co and Cu, and used to provide the first evaluation of multi-elemental composition to differentiate between several legume sam- ples usually produced and consumed in Slovenia. Despite these encouraging data, there remain some limitations and further research base on a larger dataset of legume species will be important in establishing more reliable databases in elemental composition. Acknowledgements This research was a part of the research programme Agrobiodiversity (P4-0072) and the research project L4- 7520 financially supported by the Slovenian research agen- cy (ARRS), Ljubljana, Slovenia, and the Public Service in vegetables and herbs financed by the Ministry of Agricul- ture, Forestry and Food, Republic of Slovenia. The authors would like to thank Peter Kump, PhD, for kindly providing the QAES (Quantitative Analysis of Environmental Sam- ples) software (XRF), and to the national curator of SPGB (Slovene Plant Gene Bank) Jelka Šuštar Vozlič, PhD, for providing the legumes genetic resources. 5. Reference 1. E. Röös, G. Carlsson, F. Ferawati, M. Hefni, A. Stephan, P. Tidåker, C. Witthöft, Renew. Agr. Food Syst. 2020, 35(2), 192– 205. DOI:10.1017/S1742170518000443 2. D. Rubiales, A. Mikić, Crit. Rev. Plant Sci., 2015, 34, 2–3. DOI:10.1080/07352689.2014.897896 3. M. Farooq, M. Hussain, M. Usman, S. Farooq, S. S. Algham- di, K. H. Siddique, J. Agric. Food Chem. 2018, 66(34), 8887– 8897. DOI:10.1021/acs.jafc.8b02924 4. C. Pinheiro, J. P. Baeta, A. M. Pereira, H. Domingues, C. P. Ricardo, J. Food Comp. Anal. 2010, 23(4), 319–325. DOI:10.1016/j.jfca.2010.01.005 5. K. I. Ereifej, G. N. Al-Karaki, M. K., Hammouri, Int. J. Food Prop. 2001, 4(2), 239–246. DOI:10.1081/JFP-100105190 6. FAOSTAT, 2020. Food and Agriculture Organisation (FAO) of the United Nations. Crops. Avaialble form: http://www.fao. org/faostat/en/#data/QC (Accessed on 8/4/2021). 7. G. Alandia, C. Pulvento, M. H. Sellami, N. Hoidal, T. Anem- one, E. Nigussie, J. J. Agüero, A. Lavini, S. E. Jacobsen, in: A. Hirich, R. Choukr-Allah, R. Ragab (Eds.): Emerging Re- search in Alternative Crops, Environment & Policy, vol 58. Springer, Cham, 2020, 25–53. DOI:10.1007/978-3-319-90472-6_2 8. M. Markiewicz-Keszycka, R. Cama-Moncunill, M. P. Casa- do-Gavalda, C. Sullivan, P. J. Cullen, Curr. Opin. Food Sci. 2019, 28, 96–103. DOI:10.1016/j.cofs.2019.10.002 9. S. Kumar, G. Pandey, Heliyon 2020, 6(3), p.e03682. DOI:10.1016/j.heliyon.2020.e03682 10. B. M. Freire, R. M. Pereira, C. N. Lange, B. L. Batista, in: Sus- tainable Solutions for Elemental Deficiency and Excess in Crop Plants. Springer, Singapore. 2020, pp. 135–182. DOI:10.1007/978-981-15-8636-1_7 11. N. Petry, E. Boy, J. P. Wirth, R. F. Hurrell, Nutrients, 2015, 7(2), 1144–1173. DOI:10.3390/nu7021144 12. M. A. Zoroddu, J. Aaseth, G. Crisponi, S. Medici, M. Peana, V. M. Nurchi, J. Inorg. Biochem. 2019, 195, 120–129. DOI:10.1016/j.jinorgbio.2019.03.013 13. H. Zhao, J. Tang, Q. Yang, J. Food Comp. Anal. 2021, p.103900. DOI:10.1016/j.jfca.2021.103900 14. M. Ghidotti, S. Papoci, C. Dumitrascu, T. Zdiniakova, Y. Fiamegos, M. B. de la Calle Gutiñas, Talanta Open, 2021, p.100040. DOI:10.1016/j.talo.2021.100040 15. M. Nečemer, P. Kump, J. Ščančar, R. Jaćimović, J. Simčič, P. Pelicon, M. Budnar, Z. Jeran, P. Pongrac, Spectrochim. Acta B 2008, 63, 1240–1247. DOI:10.1016/j.sab.2008.07.006 16. M. Nečemer, P. Kump, K. Vogel-Mikuš, in: Use of X-ray flu- orescence-based analytical techniques in phytoremediation, Nova Science Publishers, Inc., New York, USA, 2011, 331–358 17. A. P. de Oliveira, B. D. S. O. Mateó, A. M. Fioroto, P. V. de Oliveira, J. Naozuka, J. Food Comp. Anal. 2018, 67, 135–140. DOI:10.1016/j.jfca.2018.01.012 920 Acta Chim. Slov. 2021, 68, 913–920 Sinkovič et al.: Determination of Some Elements in Legumes ... 18. G. E. Guild, N. G. Paltridge, M. S. Andersson, J.C. Stangoulis, Plant Soil 2017, 419(1), 457–466. DOI:10.1007/s11104-017-3352-4 19. G. R. Gomes, G. H. Freiria, L. Vinicius Constantino, S. M. Alves, D. M. Zeffa, L. S. A. Takahashi, Semina: Ciênc. Agrár. 2020, 41(5), 1469–1482. DOI:10.5433/1679-0359.2020v41n5p1469 20. T. Celmeli, H. Sari, H. Canci, D. Sari, A. Adak, T. Eker, C. Toker, Agronomy 2018, 8(9), p.166. DOI:10.3390/agronomy8090166 21. S. Lukin, Vestnik Russian Agric. Sci. 2018, 6, 76–79. DOI:10.30850/vrsn/2018/6/76-79 22. C. I. Lizarazo, M. Yli-Halla, F. L. Stoddard, Nutr. Cycl. Agroe- cosystems 2015, 103(3), 311–327. DOI:10.1007/s10705-015-9743-0 23. H. Khazaei, A. Vandenberg, Agronomy 2020, 10(4), p.511. DOI:10.3390/agronomy10040511 24. Z. Türk, J. Plant Nutr. 2020, 43(4), 563–577. DOI:10.1080/01904167.2019.1685099 25. G. Ciurescu, I. Toncea, M. Ropotă, M. Hăbeanu, Rom. Agric. Res. 2018, 35, 101–108. https://www.incda-fundulea.ro/rar/ nr35/rar35.14.pdf 26. L. X. Lopez-Martinez, in: H. N. Murthy, K. Y. Paek (Eds.) Bioactive Compounds in Underutilized Vegetables and Leg- umes. Reference Series in Phytochemistry. Springer, Cham, 2020, 1–17. DOI:10.1007/978-3-030-44578-2_31-1 27. M. T. Mosisa, Afr. J. Food Sci. 2017, 11(3), 74–81. DOI:10.5897/AJFS2016.1475 28. E. M. Flores, P. A. Mello, S. R. Krzyzaniak, V. H. Cauduro, R. S. Picoloto, Rapid Commun. Mass Spectrom. 2020, 34, p.e8727. DOI:10.1002/rcm.8727 29. M. W. Donner, T. Siddique, Can. J. Chem. 2018, 96(8), 795– 802. DOI:10.1139/cjc-2017-0637 30. G. G. A. de Carvalho, M. B. B. Guerra, A. Adame, C. S. No- mura, P. V. Oliveira, H. W. P. de Carvalho, D. Santos, L. C. Nunes, F. J. Krug, J. Anal. Atom. Spectrom. 2018, 33(6), 919– 944. DOI:10.1039/C7JA00293A 31. E. de Almeida, N. M. Duran, M. H. Gomes, S. M. Savassa, T. N. da Cruz, R. A. Migliavacca, H. W. Pereira de Carvalho, X‐ Ray Spectrometry 2019, 48(2), 151–161. DOI:10.1002/xrs.3001 Except when otherwise noted, articles in this journal are published under the terms and conditions of the  Creative Commons Attribution 4.0 International License Povzetek V raziskavi smo z dvema analiznima tehnikama, induktivno sklopljeno plazemsko-masno spektrometrijo (ICP-MS) in energijsko disperzijsko rentgensko fluorescenčno spektroskopijo (EDXRF), določili elementarno sestavo različnih stročnic, ki se pogosto pridelujejo in uživajo v Sloveniji. Rezultati kažejo, da se podatki, pridobljeni z uporabo teh metod, dobro ujemajo s certificiranimi referenčnimi materiali. Skupno smo v dvajsetih vzorcih stročnic določili devetnajst ele- mentov. Medsebojna primerjava štirih makro- (P, S, K, Ca) in treh mikroelementov (Fe, Zn, Mo) izmerjenih z ICP-MS in EDXRF, je pokazala visoko korelacijo med uporabljenima metodama. EDXRF se je izkazala za cenejšo, enostavnejšo in okolju prijaznejšo metodo za določanje elementov P, S, Cl, K, Ca, Fe, Zn, Mo, Sr, Rb, Ti in Br v stročnicah, medtem ko je za določanje vsebnosti Na, Mg, V, Cr, Mn, Co in Cu ustreznejša ICP-MS metoda, predvsem zaradi visoke občutljivosti in natančnosti. Z analizo glavnih komponent (PCA) smo uspeli razvrstiti vzorce preučevanih stročnic glede na elemen- tarno sestavo v štiri skupine.