TOP 7 ACTIVITIES OF CEN AND ISO STANDARDIZATION OF METHODS FOR ANALYSIS OF ANIMAL FEEDS

TOP 7 ACTIVITIES OF CEN AND ISO STANDARDIZATION OF METHODS FOR ANALYSIS OF ANIMAL FEEDS Preparation, registration & labelling of samples for analysis in laboratories - Presentation of the revised EN ISO 6498: Animal feeding stuffs - Guidelines for Sample Preparation Analysis of selenium, arsenic and mercury Presentation of EN 16159 (Se), EN 16206 (As) and EN 16277 (Hg) Activities of CEN & ISO 1

TOP 7 ACTIVITIES OF CEN AND ISO STANDARDIZATION OF METHODS FOR ANALYSIS OF ANIMAL FEEDS Preparation, registration & labelling of samples for analysis in laboratories - Presentation of the revised EN ISO 6498: Animal feeding stuffs - Guidelines for Sample Preparation Analysis of selenium, arsenic and mercury Presentation of EN 16159 (Se), EN 16206 (As) and EN 16277 (Hg) Activities of CEN & ISO 2

TOP 7 Analysis of: selenium, arsenic and mercury Presentation of European Standards: EN 16159 (Se) EN 16206 (As) EN 16277 (Hg) 3

Selenium, Arsenic, Mercury EN 16159 Animal feeding stuffs Determination of selenium by hydride generation atomic absorption spectrometry (HGAAS) EN 16277 Animal feeding stuffs Determination of mercury by coldvapour atomic absorption spectrometry (CVAAS) EN 16206 Animal feeding stuffs Determination of arsenic by hydride generation atomic absorption spectrometry (HGAAS) after microwave pressure digestion (extraction with 65 % nitric acid and 30 % hydrogen peroxide) 4

Selenium, Arsenic, Mercury Same extraction procedure total contents for Se, Hg and As, but organic As species of marine origin are hard to digest = temperature should be > 280 C like EN 15550 for Lead (Pb) and Cadmium (Cd) by graphite furnace atom absorption spectrometry (GFAAS) and EN 15621 for minerals and trace elements by ICP-AES EN 16279 Fluorine (F) = extraction with 1 M hydrochloric acid (HCl) for 20 minutes by ion sensitive electrode (ISE) to release 1M HCl soluble fluorides or analysis of water-soluble chlorides calculated as salt (NaCl) 5

General Se, As, Hg Reagents (1) The concentration of the trace elements in the reagents and water used shall be low enough not to affect the results of the determination. A blank should be measured simultaneously with the test samples at every day of analysis to control contamination and carry over with arsenic in the reagents and apparatus used. Use water conforming to grade 2 of EN-ISO 3696. NOTE High purity is essential to avoid potential contamination. Therefore only use reagents available with high purity or perform an digestion by a subboiling distillation for nitric acid (4.2). 6

Se, As, Hg Reagents (2) Nitric acid, not less than 65 % (mass fraction), of approximately ρ (HNO3) = 1,4 g/ml Hydrogen peroxide, not less than 30 % (mass fraction), of approximately ρ (H2O2) 1,1 g/ml Hydrochloric acid, > 30 % (mass fraction), of approximately ρ (HCl) about 1,15 g/ml Diluted hydrochloric acid, e.g. about 3 % (mass fraction), used as carrier solution in the flow injection procedure and for dilution of the arsenic stock solution to the 1 mg/l standard solution and furthermore to the calibration solutions. EXAMPLE Dilute approximately 90 ml of hydrochloric acid (4.4) to 1 l with water. 7

Regents for pre-reduction Se, As, Hg Reagents (3) a) Potassium iodide/ascorbic acid solution Dissolve 2,5 g of potassium iodide and 2,5 g of L-ascorbic acid in water and dilute to 100 ml. Prepare a fresh solution at the day of analysis. b) Hydrochloric acid, > 30 % (mass fraction), of approximately ρ (HCl) about 1,15 g/ml For pre-reduction of Selenium ions to Se(IV): Reagent a) For pre-reduction of Arsenic ions to As (III): Reagents a) und b) 8

Reducing agent to convert Se (IV) to selenium hydride (SeH2) As(III) to arsenic hydride (AsH3) Hg to elemental volatile Hg Se, As, Hg Reagents (4) c) Sodium borohydride solution, e.g. c = 2 g/l Dissolve 2 g of sodium hydroxide pellets in water, add 2 g of sodium borohydride and dilute to 1 000 ml with water into 1000 ml flask (5.4). Prepare a fresh solution daily and, when necessary, filter before use. When the analysis procedure is of longer time it is recommended to cool the sodium borohydride solution, i.e. with ice around the flask, during its use in the HGAAS measurement. 9

Se, As, Hg Reagents (5) Alternative reducing agent instead of sodium borohydride for Hg d) Tin(II) chloride solution, c(sncl2 2H2O) = 100 g/l. Dissolve 50 g of tin(ii) chloride, in approximately 100 ml of hydrochloric acid (4.4) in a 500 ml volumetric flask (5.3) and dilute to the mark with water. Prepare a fresh solution daily. 10

Se, As, Hg Preparation of the test solution (1) - NOTE 1 - The following digestion procedure leads in most cases to results for arsenic and for other minerals and trace elements which correspond to the total contents of these elements. For some specific problems, like incomplete mineralization of organic arsenic in marine compounds, check whether modifications of the digestion program or other acid mixtures are necessary. 11

Se, As, Hg Preparation of the test solution (1) Weigh test portions of 0,2 g to 0,5 g in reaction vessels of the microwave pressure digestion system, considering the part of organic matter of the sample and the size of the reaction vessel. 12

Se, As, Hg Preparation of the test solution (2) Add 5 ml nitric acid (HNO 3 65 %) and 2,5 ml hydrogen peroxide (H 2 O 2 30 %) using reaction vessels of 100 ml size 13

Se, As, Hg Preparation of the test solution (3) Digest using an adequate temperature program considering the manufacturers manuals NOTE With a digestion temperature of 200 C a sufficient digestion of arsenic (and other elements) is obtained. In general, it applies that the quality of the digestion will become better with increasing digestion temperature. Digestion temperatures up to 300 C can be necessary in this procedure for a complete mineralization of all organic arsenic species from compounds of marine origin otherwise too low results of total arsenic may occur (see Table A.1 fish feed). 14

Se, As, Hg Preparation of the test solution (4) Cool down the reaction vessels, de-aerate carefully, then open and let pass off brown (nitrose) gases by using an ultrasonic or a water bath. 15

Se, As, Hg Preparation of the test solution (5) Transfer quantitatively the digestion solution to a graduated flask (not necessary when reaction vessels are graduated) and fill up with water. 16

Se, As, Hg Preparation of the test solution (6) Filter or centrifuge the test solution or allow particles to settle to the bottom, if necessary 17

Se, As, Hg Preparation of the test solution (7) Transfer the clear test solution in vials of PTFE, PFA or FEP for measuring selenium by HGAAS (hydride generation atom absorption spectrometry) arsenic by HGAAS (or ICP-MS) mercury by CVAAS (cold vapour AAS) REMARK Direct Mercury Analyzer (DMA) do not need an extraction procedure 18

Se, As, Hg Preparation of the test solution (8) Dilute the stock solution to a standard solution and prepare for selenium at the day of analysis 5 pre-reduced calibration solutions for arsenic at the day of analysis 5 pre-reduced calibration solutions with potassium iodide/ascorbic acid solution and hydrochloric acid (37 %) and wait for 60 min. Dilute with hydrochloric acid (3 %), wait again for 60 min before measuring. for mercury 5 calibration solutions 19

Se, As, Hg Preparation of the test solution (9) For As take an aliquot of the test solution for pre-reduction with potassium iodide/ascorbic acid solution and hydrochloric acid (37 %) and wait for 60 min. Dilute with hydrochloric acid (3 %), wait again for 60 min before measuring. For Se if necessary pre-dilute the test solutions with diluted nitric acid, then pre-reduce the test solutions with hydrochloric acid (30 %) in a water bath for 20 min at 80 C. 20

Se, As, Hg Preparation of the test solution (10) Prepare the reduction and the carrier solution and adjust the AAS system as specified in the operating manual of the manufacturer and optimise the typical settings. 21

Se, As, Hg Preparation of the test solution (11) Se - measure the calibration, test, blank and reference solutions at 196,0 nm by flow-injection HGAAS. As - measure the calibration, test, blank and reference solutions at 193,7 nm by flow-injection HGAAS. Hg - measure the calibration, test, blank and reference solutions at 253,7 nm by flow-injection CVAAS. Calculate the sample results considering the reference and blank samples for control. 22

Contents Selenium Page Foreword... 3 1 Scope... 4 2 Normative references... 4 3 Principle... 4 4 Reagents... 5 5 Apparatus and equipment... 6 6 Procedure... 7 7 Calculation... 10 8 Precision... 10 9 Test report... 11 Annex A (informative) Results of the inter-laboratory test... 13 Annex B (informative) Flowchart... 14 Annex C (informative) Alternative digestion procedure with the same digestion efficiency: Acid digestion with a mixture of 65 % nitric acid and 70 % perchloric acid (7:3 by volume) at atmospheric pressure... 15 Bibliography... 17 23

Selenium 1 Scope This European Standard specifies a method for the determination of selenium in animal feeding stuffs by hydride generation atomic absorption spectrometry (HGAAS) after microwave pressure digestion. The method was successfully tested by an inter-laboratory study of CEN/TC 327/WG 4 in the range of 0,25 mg/kg to 74 mg/kg. The limit of quantification is 0,5 µg/l of the test solution which corresponds to the calibration standard 2. Using a test portion of 0,5 g and a volume of the test solution of 25 ml after pressure digestion the limit of quantification is calculated as 0,125 mg/kg in the feed material. NOTE A lower limit of quantification could be achieved each laboratory has to proof it. 24

Selenium 1 Principle Selenium is determined in the test solution by hydride generation atomic absorption spectrometry (HGAAS) after microwave pressure digestion and a pre-reduction step. The homogenised feeding stuff test sample is digested by nitric acid and hydrogen peroxide under pressure and high temperatures in a microwave-heated pressure digestion system. Selenium ions of the test solution are reduced with hydrochloric acid to selenium (IV) and converted to selenium hydride (SeH 2 ) by sodium borohydride. This selenium hydride is transferred by a gas stream into a heated measurement cell and decomposed. The absorption at the selenium line at 196,0 nm corresponds to the amount of selenium. NOTE Selenium (VI) is not determined by the hydridisation as described here. It is therefore necessary to adjust the digestion conditions and to exercise a pre-reduction step with hydrochloric acid to yield only selenium (IV). Other digestion procedures with the same digestion efficiency or other measurement systems like Fl-HGAAS or HG-ICP-AES are possible (see Annex D). 25

Selenium Table 1 Recommended calibration solutions for the determination of selenium Selenium (Se) Concentration of calibration solution (4.10) after pre-reduction procedure Aliquot of selenium standard solution (4.9) transferred in 100 ml flasks (5.4) (pre-reduction step) Calibration standard 1 0 µg/l 0 µl Calibration standard 2 0,5 µg/l 50 µl Calibration standard 3 2,5 µg/l 250 µl Calibration standard 4 5 µg/l 500 µl Calibration standard 5 10 µg/l 1 000 µl 26

Selenium Apparatus and equipment Microwave-heated pressure digestion apparatus with inert reaction vessels Pipettes, volumetric and/or graduated, 100 µl, 250 µl, 600 µl, 1 000 µl, 1 500 µl, 2 ml, 2,5 ml and 10 ml Flasks, 25 ml, 50 ml, 100 ml, 500 ml and 1 000 ml Flow-injection hydride system, with sample loop, i.e. 500 µl Atomic absorption spectrometer (AAS), with measurement recording system, background correction, heated quartz cell and accessories for the hydride procedure Element-specific lamp for selenium Ultrasonic bath and/or water bath 27

Selenium 6.3 Measurement of the test solution 6.3.1 Pre-dilution of the test solution It is very important that the acid concentration of the (diluted) test solution corresponds to that of the calibration solutions because the signal height for measuring selenium by HGAAS depends on the acid matrix. Therefore when the measured selenium concentration of a sample exceeds the linear range of the calibration function, a dilution of the test solution (6.2) with nitric acid (4.3) instead with water is necessary. 28

Selenium 6.3.2 Pre-reduction of the (pre-diluted) test solution Dilute 10 ml of the test solution (6.2) or the pre-diluted test solution (6.3.1) into a 50 ml flask (5.4), add 5 ml hydrochloric acid (4.5) and heat for 20 min in a water bath at 80 C with opened flask. After cooling down to room temperature the flask is filled up with water. 29

Selenium 6.3.3 Spectrometer settings of the flow-injection hydride generation atomic absorption spectrometer (HGAAS) Table 2 Typical settings of HGAAS for measuring selenium Temperature of the cell 900 C Wave length Slit width Signal processing a) Smoothing Integration time 196,0 nm 2,0 nm Peak height with background correction 0,5 s 15 s a) Nearest to the limit of quantification a signal processing by peak area is recommended. 30

Selenium 6.3.4 HGAAS determination The pre-reduced test solutions (6.3.2), if necessary pre-diluted previously (6.3.1), and the selenium calibration solutions (4.10) are measured directly with an atomic absorption spectrometer with electrically heated quartz cell coupled to a flow-injectionhydride-system. Use of a 500 µl sample loop is recommended. 31

Selenium 6.3.4 HGAAS determination The apparatus should be programmed in such a way that first the sample loop is filled with the pre-reduced test or prereduced calibration solution. Then the test- or calibration solution is transferred to a mixing unit with diluted hydrochloric acid (4.6) and mixed with sodium borohydride solution (4.7). The resulting gas/liquid mixture is separated by an argon-flowed separator. The argon steam sorts out the metal hydrides to the quartz cell for atomisation reaction and measuring the atomic absorption of selenium. 32

Selenium 6.3.4 HGAAS determination Firstly the selenium calibration solutions (4.10) are measured, then the (pre-diluted) test solutions (6.3.12). Check the linear range of the calibration function. If the concentration of the test sample is outside the linear range dilute with nitric acid (4.3) and not with pure water. When carrying out prolonged series of measurements, it is advisable to check the zero and the calibration at intervals. 33

Selenium 6.3.4 HGAAS determination Significant background signal appears in the case of the hydride generation technique, either by matrix effects or by using higher concentrations of nitric acid or hydrochloric acid. Add amido sulphuric acid after the pre-reduction step when disturbances from nitric acid appear. 34

Selenium 6.3.4 HGAAS determination Copper concentrations in the test solution of more than 750 µg/l could lead to a signal depression. Measuring copper and selenium concentrations simultaneously by ICP-AES could be useful: If the selenium concentration of the test solution by ICP- AES is higher to that of HGAAS and if the copper concentration is high a depression effect by copper seems to be possible. Then dilute the test solution (6.3.1) with diluted nitric acid (4.3) or take 1 ml of a 0,5 % solution of 1,10 - phenanthroline into 10 ml of the test solution (= 0,05 % of 1,10 - phenanthroline within the test solution) for complexation of copper to measure selenium by HGAAS without a signal depression (see 6.3.2). 35

Selenium 6.3.4 HGAAS determination For unknown matrix effects use the standard addition procedure. As an analytical control, reference samples having reliable known selenium contents shall be analysed parallel with all the series of samples analysed, the reference samples being subjected to all the steps in the method starting from digestion. Blank solutions prepared by subjecting them to all the steps in the method shall also be determined. 36

Selenium precision data Matrix tested Subsamples fully prepared for direct weighing Cow feed Turkey feed Wheat Mineral pig feed Mineral cow feed Turkey feed Mineral pig feed yes yes yes yes yes no no No. of labs 19 19 19 19 19 18 18 No. of outlier labs 1 4 3 0 4 4 1 No. of non-compliant labs 0 0 0 3 2 0 2 No. of valid labs 18 15 16 16 13 14 15 Mean value, mg/kg 0,305 0,313 0,049 6,87 73,6 0,246 6,76 s r, mg/kg 0,026 0,019 0,018 0,69 2,5 0,016 0,58 r, mg/kg 0,072 0,054 0,050 1,94 6,9 0,045 1,62 RSD(r), % 8,5 6,1 36,3 10,1 3,4 6,5 8,6 S R, mg/kg 0,069 0,048 0,027 1,51 12,3 0,037 1,57 R, mg/kg 0,192 0,134 0,080 4,23 34,5 0,103 4,39 RSD(R), % 22,5 15,3 58,3 22,0 16,7 14,9 23,2 HORRAT(R) 1,2 0,8 2,3 1,8 2,0 0,8 1,9 37

Selenium flowchart Weigh test portions of 0,2 to 0,5 g in reaction vessels of the microwave pressure digestion system, considering the part of organic matter of the sample and the size of the reaction vessel. Add 5 ml nitric acid (65 %) and 2,5 ml hydrogen peroxide (30 %) using reaction vessels of 100 ml size. Digest using an adequate temperature program considering the manufacturers manuals (digestion temperature should achieve 200 C). Cool down the reaction vessels, de-aerate carefully, then open and let pass off brown (nitrose) gases by using an ultrasonic or a water bath. Transfer quantitatively the digestion solution to a graduated flask (not necessary when reaction vessels are graduated) and fill up with water. Filtrate or centrifuge the test solution or let drop particles down, if necessary. Transfer the clear test solution in vials of PTFE, PFA or FEP for measuring selenium by HGAAS. Dilute the selenium stock solution to a standard solution. Prepare at the day of analysis 5 pre-reduced calibration solutions with hydrochloric acid (30 %) in a water bath for 20 min at 80 C. If necessary pre-dilute the test solutions with diluted nitric acid, then pre-reduce the test solutions with hydrochloric acid (30 %) in a water bath for 20 min at 80 C. Prepare the reduction and the carrier solution and adjust the AAS system as specified in the operating manual of the manufacturer and optimise the typical settings. Measure the calibration, test, blank and reference solutions at 196,0 nm by flow-injection HGAAS. Calculate the sample results considering the reference and blank samples for control. 38

Selenium Annex C (informative) Alternative digestion procedure with the same digestion efficiency: Acid digestion with a mixture of 65 % nitric acid and 70 % perchloric acid (7:3 by volume) at atmospheric pressure 39

Selenium Bibliography [15] Validation of an analytical method to determine selenium (Se) in animal feeding stuffs - Final report of the collaborative study - Animal feeding stuffs Determination of selenium by hydride generation atomic absorption spectrometry (HGAAS) after microwave digestion (Digestion with 65 % nitric acid and 30 % hydrogen peroxide), Jürgen Danier (Project leader) c/o Bioanalytic Weihenstephan, Research Centre for Nutrition and Food Science (ZIEL) of TUM, 85350 Freising, Germany: http://www.wzwbioanalytik.de/download_e.php 40

Contents Arsenic Page 1 Scope... 3 2 Normative references... 3 3 Principle... 3 4 Reagents... 4 5 Apparatus and equipment... 6 6 Procedure... 7 7 Calculation... 10 8 Precision... 10 9 Test report... 11 Annex A (informative) Results of the inter-laboratory test... 12 Annex B (informative) Flowchart... 13 Annex C (informative) Alternative digestion procedure with the same digestion efficiency to ensure complete mineralization of all organic and inorganic arsenic species for HGAAS measurement: dry ashing with magnesium oxide and magnesium nitrate as ashing reagents... 15 Bibliography... 16 41

Arsenic 1 Scope This European Standard specifies a method for the determination of total arsenic in animal feeding stuffs by hydride generation atomic absorption spectrometry (HGAAS) after microwave pressure digestion. The limit of quantification is 0,5 µg/l of the test solution. Using a test portion of 0,5 g, a volume of the test solution of 25 ml and an aliquot of 5 ml for pre-reduction the limit of quantification is 0,125 mg/kg in the feed material. NOTE For feed materials containing organic arsenic species from compounds of marine origin (i.e. arsenobetaine and tetramethylarsine oxide) a higher digestion temperature of the microwave system up to 300 C may be necessary in order to enable the hydridisation of these arsenic compounds and in order to determine all different kinds of arsenic species in the corresponding feeding stuffs. Alternatively the digestion procedure of Annex D can be used if the microwave system does not reach higher temperatures up to 300 C to ensure complete mineralization for HGAAS determination. 42

Arsenic 1 Principle Arsenic is determined in the test solution by hydride generation atomic absorption spectrometry (HGAAS) after microwave pressure digestion and a pre-reduction step. The homogenised feeding stuff test sample is digested by nitric acid and hydrogen peroxide under pressure and high temperatures in a microwave-heated pressure digestion system. Arsenic ions of the test solution are reduced with a potassium iodide/ascorbic acid solution and hydrochloric acid to arsenic (III) and converted to arsenic hydride (AsH 3 ) by sodium borohydride. Arsenic hydride is transferred by a gas stream into a heated measurement cell and decomposed. The absorption at the arsenic line at 193,7 nm corresponds to the amount of arsenic. Since arsenic (III) and arsenic (V) show a different sensitivity with the hydride technique, it is necessary to reduce arsenic (V) to arsenic (III) in order to avoid incorrect measurements. Other digestion procedures with the same digestion efficiency are possible in order to mineralize completely all arsenic species like organic arsenic species from compounds of marine origin for HGAAS determination (see Annex D). NOTE 1 When using i.e. perchloric acid as alternative digestiondigestion procedure to ensure complete mineralisation of all organic and inorganic arsenic species for HGAAS determination you must use NaI/L-ascorbic acid because KI results in precipitation of potassium perchlorate. NOTE 2 Alternatively an ICP-MS for measuring can be used where an incomplete mineralization is not of importance. 43

Arsenic 6.3 Measurement of the test solution 6.3.1 Pre-reduction of arsenic (V) to arsenic (III) Dilute 1 ml of the test solution (6.2) into a 25 ml flask (5.4), add 2,5 ml potassium iodide/ascorbic acid solution (4.7) and 2,5 ml hydrochloric acid (4.4) and mix thoroughly. Let the solutions stand at room temperature for 60 min. Finally fill up with hydrochloric acid (4.5) and wait again 60 min at room temperature before the test solutions are measured (Table 2). = Table 1 - Calibration solution concentrations after pre-reduction. 44

Arsenic Table 1 - Calibration solution concentrations after pre-reduction. Arsenic (As) Concentration of calibration solution (4.10) after pipetting 1 ml from the 50 ml flasks (5.4) into 25 ml flasks (5.4) for pre-reduction Aliquot of arsenic standard solution (4.9) transferred in 50 ml flasks (5.4) Calibration standard 1 0 µg/l 0 ml Calibration standard 2 1 µg/l 1,25 ml Calibration standard 3 2 µg/l 2,50 ml Calibration standard 4 6 µg/l 7,50 ml Calibration standard 5 10 µg/l 12,5 ml 45

Arsenic 6.3.1 Pre-reduction of arsenic (V) to arsenic (III) Table 2 Recommended procedure for the pre-reduction of test solutions to measure arsenic by flow-injection HGAAS Flask (volume of test solution after pressure digestion) 25 ml 100 ml Aliquot of test solution (6.2) a 1 ml to 5ml (6.2) Flask for pre-reduction (final volume) Potassium iodide/ascorbic acid solution (4.7) Hydrochloric acid (4.4) Incubation time Temperature 25 ml 2,5 ml 2,5 ml 60 min Room temperature Fill up with: Diluted hydrochloric acid (4.5) Incubation time 60 min a Dilution factor for pre-reduction 1:25 (if aliquot of the test solution = 1 ml) The aliquot of the test solution could be reduced or increased, depending from the concentration of arsenic in the test sample. 46

Arsenic 6.3.2 Settings of the atomic absorption spectrometer (HGAAS - procedure) To devise a test schedule, first adjust the apparatus as specified in the operating manual of the manufacturer, then optimise the settings, paying particular attention to gas flow times and the amounts of sodium borohydride introduced. Typical settings are listed in Table 3. 47

Arsenic 6.3.2 Settings of the atomic absorption spectrometer (HGAAS - procedure) Table 3 Typical settings of HGAAS for measuring arsenic Temperature of the cell 900 C Wave length 193,7 nm Slit width 0,7 nm Signal processing a Peak height with background correction Smoothing 0,5 s Integration time 15 s a Nearest to the limit of quantification a signal processing by peak area is recommended. 48

Arsenic 6.3.3 HGAAS determination The pre-reduced test solutions (6.3.1) and arsenic calibration solutions (4.10) are measured directly with an atomic absorption spectrometer with electrically heated quartz cell coupled to a flowinjection-hydride-system. Use of a 500 µl sample loop is recommended. 49

Arsenic 6.3.3 HGAAS determination The apparatus should be programmed in such a way that first the sample loop is filled with the pre-reduced test or pre-reduced calibration solution. Then the test- or calibration solution is transferred to a mixing unit with diluted hydrochloric acid (4.5) and mixed with sodium borohydride solution (4.6). The resulting gas/liquid mixture is separated by an argon-flowed separator. The argon steam sorts out the metal hydrides to the quartz cell for atomisation reaction and measuring the atomic absorption of arsenic. 50

Arsenic 6.3.3 HGAAS determination Firstly the arsenic calibration solutions (4.10) are measured, then the pre-reduced test solutions (6.3.1). Check the linear range of the calibration function. If the concentration of the test solution is outside the linear range dilute with hydrochloric acid (4.5) and not with pure water. When carrying out prolonged series of measurements, it is advisable to check the zero and the calibration at intervals. 51

Arsenic 6.3.3 HGAAS determination As an analytical control, reference samples having reliable known arsenic contents shall be analysed parallel with all the series of samples analysed, the reference samples being subjected to all the steps in the method starting from digestion. Blank solutions prepared by subjecting them to all the steps in the method shall also be determined. 52

Arsenic precision data Matrix tested Turkey feed Hay Mineral piglet feed Grass silage Bentonite Montmorillonite Fish feed Turkey feed Mineral piglet feed Subsamples fully prepared for direct weighing yes yes yes yes yes yes no no No. of labs 18 18 20 20 20 18 18 20 No. of outlier labs 2 3 5 4 2 0 3 2 No. of non compliant labs 0 0 0 0 0 11 0 0 No. of valid labs 16 15 15 16 18 7 15 18 Mean value, mg/kg 0,100 0,220 2,01 3,58 6,44 3,31 0,101 1,94 s r, mg/kg 0,010 0,019 0,08 0,15 0,28 0,23 0,009 0,07 r, mg/kg 0,027 0,053 0,23 0,43 0,78 0,65 0,025 0,21 RSD(r), % 9,6 8,6 4,1 4,3 4,3 7,0 8,9 3,8 S R, mg/kg 0,021 0,034 0,23 0,29 0,96 0,36 0,017 0,26 R, mg/kg 0,058 0,096 0,64 0,81 2,69 1,00 0,047 0,73 RSD(R), % 20,6 15,7 11,3 8,1 14,9 10,7 16,5 13,4 HORRAT(R) 0,9 0,8 0,8 0,6 1,2 0,8 0,7 0,9 53

Arsenic flowchart Weigh test portions of 0,2 g to 0,5 g in reaction vessels of the microwave pressure digestion system, considering the part of organic matter of the sample and the size of the reaction vessel. Add 5 ml nitric acid (65 %) and 2,5 ml hydrogen peroxide (30 %) using reaction vessels of 100 ml size. Digest using an adequate temperature program considering the manufacturers manuals Note digestion temperature should achieve 200 C - for organic arsenic species of compounds from marine origin up to 300 C. Cool down the reaction vessels, de-aerate carefully, then open and let pass off brown (nitrose) gases by using an ultrasonic or a water bath. Transfer quantitatively the digestion solution to a graduated flask (not necessary when reaction vessels are graduated) and fill up with water. Filtrate or centrifuge the test solution or let drop particles down, if necessary. Transfer the clear test solution in vials of PTFE, PFA or FEP for measuring arsenic by HGAAS (or ICP-MS). Dilute the arsenic stock solution to a standard solution and prepare at the day of analysis 5 pre-reduced calibration solutions with potassium iodide/ascorbic acid solution and hydrochloric acid (37 %) and wait for 60 min. Dilute with hydrochloric acid (3 %), wait again for 60 min before measuring. Take an aliquot of the test solution for pre-reduction with potassium iodide/ascorbic acid solution and hydrochloric acid (37 %) and wait for 60 min. Dilute with hydrochloric acid (3 %), wait again for 60 min before measuring. Prepare the reduction and the carrier solution and adjust the AAS system as specified in the operating manual of the manufacturer and optimise the typical settings. Measure the calibration, test, blank and reference solutions at 193,7 nm by flow-injection HGAAS. Calculate the sample results considering the reference and blank samples for control. 54

Arsenic Annex C (informative) Alternative digestion procedure with the same digestion efficiency to ensure complete mineralization of all organic and inorganic arsenic species for HGAAS measurement: dry ashing with magnesium oxide and magnesium nitrate as ashing reagents 55

Arsenic Bibliography A[12] Validation of an analytical method to determine arsenic (As) in animal feeding stuffs - Final report of the collaborative study - Animal feeding stuffs - Determination of arsenic by hydride generation atomic absorption spectrometry (HGAAS) after microwave digestion (digestion with 65% nitric acid and 30% hydrogen peroxide), Jürgen Danier (project leader) c/o Bioanalytic Weihenstephan, Research Centre for Nutrition and Food Science (ZIEL) of TUM, 85350 Freising, Germany: http://www.wzwbioanalytik.de/download_e.php 56

Mercury Contents Page Foreword... 3 1 Scope... 3 2 Normative references... 3 3 Principle... 3 4 Reagents... 4 5 Apparatus and equipment... 6 6 Procedure... 6 7 Calculation... 8 8 Precision... 9 9 Test report... 10 Annex A (informative) Results of the interlaboratory tests... 11 Annex B (informative) Flowchart Determination of mercury by CVAAS after microwave digestion... 12 Annex C (informative) Alternative digestion procedure with the same extraction efficiency: Acid digestion with a mixture of 65 % nitric acid and 70 % perchloric acid (7:3 v/v) at atmospheric pressure... 13 Annex D (informative) Alternative digestion procedure with the same extraction efficiency: Acid digestion with a mixture of 65 % nitric acid, 37 % hydrochloric acid and 30 % hydrogen peroxide under reflux... 15 Bibliography... 16 57

Mercury 1 Scope This European Standard specifies a method for the determination of mercury in animal feeding stuffs by coldvapour atomic absorption spectrometry (CVAAS) after microwave pressure digestion. The limit of quantification in the test solution should be 0,25 µg/l or lower. Using a test portion of 0,5 g and a volume of the test solution of 25 ml a limit of quantification of 0,0125 mg/kg or lower should be obtained. 58

Mercury 1 Principle Mercury is determined in the test solution by cold-vapour atomic absorption spectrometry (CVAAS) after microwave pressure digestion. The homogenised feeding stuff test sample is digested with nitric acid and hydrogen peroxide under pressure and high temperatures in a microwave-heated pressure digestion system. The test solution is transferred to the reaction vessel of the mercury analysis unit, and the mercury is reduced with sodium borohydride or tin(ii) chloride to elemental volatile mercury and flushed into the cell of the AAS instrument using a carrier gas stream. As an option with an additional amalgamation step sensitivity could be increased and matrix effects could be decreased. The absorption at 253,7 nm (mercury line) is used as a measure of the mercury concentration in the cell. Other digestion procedures with the same extraction efficiency (see Annex D and Annex E) or other measurement systems like Fl-CVAAS (flow injection cold-vapour atomic absorption spectroscopy) or CV-ICP- AES (cold vapour atomic emission spectroscopy) are possible. 59

Mercury Table 1 Recommended calibration solutions (4.9) for the determination of mercury. Mercury (Hg) Concentration of calibration solution (4.9) µg/l Aliquots of standard solution (4.8), transferred in 50 ml flasks Calibration standard 1 0 0 Calibration standard 2 0,5 25 Calibration standard 3 1 50 Calibration standard 4 5 250 Calibration standard 5 10 500 µl 60

Mercury Apparatus and equipment Microwave-heated pressure digestion apparatus with inert reaction vessels Pipettes, Flasks, 25 ml, 50 ml, 100 ml, 500 ml, 1 000 ml. Flow injection cold-vapour system with sample loop, i.e. 500 µl. Atomic absorption spectrometer (AAS), with a heated quartz cell and optionally with an amalgamation system. Element-specific lamp for mercury (NOTE An electrodeless discharge lamp would provide a higher sensitivity compared to a hollow-cathode lamp.) Ultrasonic bath or water bath. Analytical balance, accurate to 0,1 mg. 61

Mercury 6.2 Measurement of the test solution 6.2.1 Spectrometer settings of the cold-vapour atomic absorption spectrometer (CVAAS) To devise a test schedule, first adjust the apparatus as specified in the operating manual of the manufacturer, then optimise the settings, paying particular attention to gas flow times and the amounts of tin(ii) chloride or sodium borohydride introduced. Typical settings are listed in Table 2. 62

Mercury 6.2 Measurement of the test solution 6.2.1 Spectrometer settings of the cold-vapour atomic absorption spectrometer (CVAAS) Table 2 Typical settings of CVAAS for measuring mercury Temperature of the cell 100 C Wave length Slit width Signal processing a) Smoothing Integration time 253,7 nm 0,7 nm Signal height 0,5 s 15 s a) Nearest to the limit of quantification a signal processing by signal area is recommended. 63

Mercury 6.2.2 CVAAS determination The test and calibration solutions are measured directly with an atomic absorption spectrometer with electrically heated quartz cell coupled to a flow injection cold-vapour system. Use of a 500 µl sample loop is recommended. 64

Mercury 6.2.2 CVAAS determination The apparatus should be programmed in such a way that first the sample loop is filled with the test or calibration solution. Then the test or calibration solution is transferred to a mixing unit with diluted hydrochloric acid (4.5) and mixed with sodium borohydride solution (4.6.3) or with tin(ii) chloride solution (4.6.2). The resulting gas/liquid mixture is separated by an by an argon-flow separator. The argon steam sorts out the elemental volatile mercury to the quartz cell for measuring the atomic absorption of mercury. Optionally an amalgamation system could be used to increase sensitivity and to decrease matrix effects. 65

Mercury 6.2.2 CVAAS determination Firstly the calibration solutions are measured, then the test solutions. Check the linear range of the calibration function. If the concentration of the test sample is outside the linear range dilute it with diluted hydrochloric acid (4.5) and not with pure water. When carrying out prolonged series of measurements, it is advisable to check the zero and the calibration at intervals. Although a correction is seldom necessary in the case of the coldvapour technique, whether the background correction is necessary or not shall be checked for every type of sample. 66

Mercury 6.2.2 CVAAS determination As an analytical control, reference samples having reliable known mercury contents shall be analysed parallel with all the series of samples analysed, the reference samples being subjected to all the steps in the method starting from digestion. Blank solutions prepared by subjecting them to all the steps in the method shall also be determined. NOTE Other elements in high concentration levels like selenium, iodide and copper extracted from the test sample matrix (like e.g. from premixtures) can interfere and lead to lower measured mercury values of the test solution. In cases with complex or unknown matrices the standard addition method can be used. More detailed information about the tolerable concentration levels of interfering elements in the test solution (mostly in the range of in maximum 0,5 mg/l to 1 mg/l) is given in table 1 of EN 1483:2007 at point 4.2. 67

Matrix tested Subsamples fully prepared for direct weighing Mineral piglet feed Mercury precision data Fish meal Fish feed Rabbit feed Dicalcium phosphate Brewers grains Fish feed Mineral piglet feed yes yes yes yes yes yes no no No. of labs 20 20 20 20 20 20 20 20 No. of outlier labs 3 2 2 3 3 3 2 4 No. of labs without outliers Mean value, mg/kg Repeatability standard deviation s r, mg/kg Repeatability limit r, mg/kg Coefficient of variation of Repeatability, CV r, % Reproducibility standard deviation s R, mg/kg Reproducibility limit R, mg/kg Coefficient of variation of Reproducibility, CV R, % HORRATvalue (R) *) Thompson 17 18 18 17 17 17 18 16 0,005 0,087 0,174 0,754 2,003 0,464 0,092 0,006 0,001 0,003 0,005 0,034 0,085 0,024 0,004 0,001 0,003 0,008 0,015 0,095 0,237 0,066 0,010 0,002 17,6 3,4 3,1 4,5 4,2 5,1 3,8 12,3 0,002 0,013 0,021 0,085 0,146 0,055 0,011 0,002 0,005 0,035 0,060 0,237 0,410 0,154 0,032 0,007 31,6 14,3 12,3 11,2 7,3 11,9 12,5 37,8 1,4 * 0,7 * 0,6 Workshop ANKARA 0,7 TOP 0,5 7 0,7 0,6 * 1,7 * 68 (2000)

Mercury flow chart Weigh test portions of 0,2 g to 0,5 g in reaction vessels of the microwave pressure digestion system, considering the part of organic matter of the sample and the size of the reaction vessel. Add 5 ml nitric acid (65 %) and 2,5 ml hydrogen peroxide (30 %) using reaction vessels of 100 ml size. Extract using an adequate temperature program considering the manufacturers manuals (digestion temperature should achieve 200 C). Cool down the reaction vessels, release pressure carefully, then open and let pass off brown (nitrose) gases; use of an ultrasonic or a water bath is recommended. Transfer quantitatively the extraction solution to a graduated flask (not necessary when reaction vessels are graduated) and fill up with water. Filter or centrifuge the test solution or allow particles to settle to the bottom, if necessary. Transfer the clear test solution to vials of PP, PFA or FEP for measuring mercury by CVAAS. Dilute the mercury stock solution to a standard solution and prepare 5 calibration solutions. Prepare the reduction and the carrier solution and adjust the AAS system as specified in the operating manual of the manufacturer and optimise the typical settings. Measure the calibration, test, blank and reference solutions at 253,7 nm by flow injection CVAAS. Calculate the sample results considering the reference and blank samples for control. 69

Mercury Annex C (informative) Alternative digestion procedure with the same extraction efficiency: Acid digestion with a mixture of 65 % nitric acid and 70 % perchloric acid (7:3 v/v) at atmospheric pressure Annex D (informative) Alternative digestion procedure with the same extraction efficiency: Acid digestion with a mixture of 65 % nitric acid, 37 % hydrochloric acid and 30 % hydrogen peroxide under reflux 70

Mercury Bibliography [19] Validation of an analytical method to determine mercury (Hg) in animal feeding stuffs - Final report of the collaborative study - Animal feeding stuffs Determination of mercury by coldvapour atomic absorption spectrometry (CVAAS) after microwave digestion (extraction with 65 % nitric acid and 30 % hydrogen peroxide), Jürgen Danier (Project leader) c/o Bioanalytik Weihenstephan, Research Centre for Nutrition and Food Science (ZIEL) of TUM, 85350 Freising, Germany: http://www.wzwbioanalytik.de/download_e.php 71

EN 16278 Inorganic As Scope This document describes a procedure for the determination of inorganic arsenic in animal feeding stuffs of marine origin by solid phase extraction (SPE) and hydride generation atomic absorption spectrometry (HG-AAS). The method has been successfully tested in a collaborative trial with a working range from 0,19 mg/kg to 2,7 mg/kg (HORRATvalues < 2). The LOQ of the method is usually approximately 0.1 mg/kg or lower. 72

EN 16278 Inorganic As Principle Inorganic arsenic consists of arsenite, As(III) and arsenate, As(V). This standard describes a method for the determination of inorganic arsenic (=sum of As(III) and As(V)). 73

EN 16278 Inorganic As Principle A representative test portion of the sample is treated with a diluted hydrochloric acid and hydrogen peroxide solution using microwave assisted heating. Hereby inorganic arsenic species are extracted and As(III) is oxidised to As(V). The inorganic arsenic is selectively separated from other arsenic compounds using solid phase extraction (SPE) and the concentration of inorganic arsenic is determined by HG-AAS in the SPE eluate. The gaseous hydride is transferred into a heated measuring cell (cuvette) by means of a carrier gas stream and decomposed. 74

EN 16278 Inorganic As Principle Since arsenic (III) and arsenic (V) show a different sensitivity with the hydride technique, it is necessary to reduce arsenic (V) to arsenic (III) in order to avoid incorrect measurements. 75

TOP 7 Analysis of: selenium, arsenic and mercury Presentation of European Standards: EN 16159 (Se), EN 16206 (As), EN 16277 (Hg) Any Questions? 76