Determination of the molar extinction coefficient for

Analytical Biochemistry xxx (2011) xxx–xxx

Contents lists available at ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

Determination of the molar extinction coefficient for the ferric reducing/antioxidant power assay William A. Hayes a, Daniel S. Mills a,⇑, Rachel F. Neville a, Jenna Kiddie b, Lisa M. Collins c a

Department of Biological Sciences, University of Lincoln, Riseholme Park, Lincoln LN2 2LG, UK Department of Veterinary Clinical Sciences, Royal Veterinary College, Hatfield, Hertfordshire AL9 7TA, UK c School of Biological Sciences, Queen’s University Belfast, Belfast BT9 7BL, Northern Ireland, UK b

a r t i c l e

i n f o

Article history: Received 28 March 2011 Received in revised form 17 May 2011 Accepted 18 May 2011 Available online xxxx Keywords: Antioxidant Calibration FRAP

a b s t r a c t The FRAP reagent contains 2,4,6-tris(2-pyridyl)-s-triazine, which forms a blue–violet complex ion in the presence of ferrous ions. Although the FRAP (ferric reducing/antioxidant power) assay is popular and has been in use for many years, the correct molar extinction coefficient of this complex ion under FRAP assay conditions has never been published, casting doubt on the validity of previous calibrations. A previously reported value of 19,800 is an underestimate. We determined that the molar extinction coefficient was 21,140. The value of the molar extinction coefficient was also shown to depend on the type of assay and was found to be 22,230 under iron assay conditions, in good agreement with published data. Redox titration indicated that the ferrous sulfate heptahydrate calibrator recommended by Benzie and Strain, the FRAP assay inventors, is prone to efflorescence and, therefore, is unreliable. Ferrous ammonium sulfate hexahydrate in dilute sulfuric acid was a more stable alternative. Few authors publish their calibration data, and this makes comparative analyses impossible. A critical examination of the limited number of examples of calibration data in the published literature reveals only that Benzie and Strain obtained a satisfactory calibration using their method. Ó 2011 Elsevier Inc. All rights reserved.

The purpose of this investigation was to determine the correct molar extinction coefficient of the complex ion formed between ferrous ions and 2,4,6-tris(2-pyridyl)-s-triazine in the FRAP (ferric reducing/antioxidant power)1 assay of Benzie and Strain [1,2]. This value is currently unknown. This molar extinction coefficient will enable users of the FRAP assay to check whether their calibration has been carried out correctly. The FRAP reagent was originally devised by Benzie and Strain to measure the ‘‘antioxidant power’’ of blood plasma by its ability to reduce added ferric ion to ferrous ion [1]. In this early work, Benzie and Strain used the acronym FRAP to indicate the ferric reducing ability of plasma [1]. Benzie and Strain later changed the meaning of the acronym to ferric reducing/antioxidant power after using the FRAP reagent to measure the total antioxidant capacity of other biological fluids [2]. There are three versions of the FRAP assay presented by Benzie and Strain [1,2]. Version 1 is an automated assay carried out at 37 °C where the FRAP reagent has water added as a diluent [1,2]. Version 2 is a manual assay carried out 37 °C with no added water as diluent [2]. Version 3 is a slower manual assay carried out at room temperature with no added water as diluent [2]. ⇑ Corresponding author. Fax: +44 1522 895437. E-mail address: [email protected] (D.S. Mills). Abbreviations used: FRAP, ferric reducing/antioxidant power; TPTZ, 2,4,6-tris(2pyridyl)-s-triazine; UV–Vis, ultraviolet–visible. 1

Version 1 requires an automated Cobas Fara centrifugal analyzer, an instrument not commonly found in analytical laboratories. The majority of workers have used version 2 of the FRAP assay or a variant with added water. In the current work, a manual version of the FRAP assay (version 4) was used to nearly match the conditions used in the automated FRAP assay so as to permit comparison with published absorbance data [1,2] for the automated assay. Benzie and Strain calibrated the FRAP reagent with ferrous sulfate heptahydrate solutions [1,2]. Although calibration absorbances are presented for the automated version of the FRAP assay [1,2], no calibration absorbances are given for the two manual versions of the assay. The absorbances from a Cobas Fara centrifugal analyzer are different from those that would be obtained using a manual method with a conventional spectrophotometer. This is a problem because users of manual assays do not know what absorbances to expect. Obtaining a molar extinction coefficient for the automated FRAP assay is possible, but one must look beyond the reports by Benzie and Strain for the dimensions of the Cobas Fara cuvette. In addition, ferrous sulfate heptahydrate crystals are unstable and tend to form the tetrahydrate slowly during storage. This problem is examined in this work. Some workers using the manual version of the FRAP assay have assumed incorrectly that Beer’s law does not apply during calibration by presenting linear equations with intercepts for the relation

0003-2697/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2011.05.031

Please cite this article in press as: W.A. Hayes et al., Determination of the molar extinction coefficient for the ferric reducing/antioxidant power assay, Anal. Biochem. (2011), doi:10.1016/j.ab.2011.05.031

2

Molar extinction coefficient for FRAP assay / W.A. Hayes et al. / Anal. Biochem. xxx (2011) xxx–xxx

between ferrous sulfate heptahydrate concentration before it is added to the cuvette and absorbance [3–5]. These calibrations disagree with each other. In general, FRAP assay calibrations are underreported in the literature, placing most of the published FRAP values for biological fluids beyond critical appraisal. There is also a need for a single manual protocol with an established molar extinction coefficient to harmonize future work. Materials and methods Chemicals Concentrated sulfuric acid and ammonium acetate were purchased from Aldrich. Diphenylamine sulfonic acid, barium salt, was obtained from BDH Chemicals (Poole, UK). Ferrous sulfate heptahydrate, concentrated (85%) orthophosphoric acid, concentrated hydrochloric acid SG 1.18, glacial acetic acid, hydroxylammonium chloride, and sodium acetate trihydrate were obtained from Fisher Scientific (Loughborough, UK). TPTZ (2,4,6-tris(2-pyridyl)-s-triazine) was purchased from Fluka. Potassium dichromate was obtained from Griffin & George (Loughborough, UK). 1,10Phenanthroline monohydrate, ferrous ammonium sulfate hexahydrate, and ferric chloride hexahydrate were obtained from Sigma–Aldrich (Dorset, UK). Where possible, all chemicals were of analytical reagent grade or better. Instruments A WPA Biowave II Ultraviolet–visible (UV–Vis) spectrophotometer (Biochrom, Cambridge, UK) was used for absorbance readings. The pH measurements were taken with an HI9321 Microprocessor pH meter (Hanna Instruments, Leighton Buzzard, UK) that was calibrated at pHs 4.0 and 7.0 with buffer solutions supplied by Scientific Laboratory Supplies (Nottingham, UK). Reaction temperatures of 37 °C were obtained using a stirred temperature-controlled water bath (Griffin & George, London, UK). Samples were mixed using a VM20 Vortex Mixer (Chiltern Scientific Instrumentation, Wendover, UK).

tion of ferrous ion with 0.001 N potassium dichromate was carried out using the method of Sarver and Kolthoff [8] with 0.005 N (3.17 g/L) diphenylamine sulfonic acid, barium salt, as the redox indicator. The indicator correction was 0.25 ml of 0.001 N potassium dichromate [8]. Potassium dichromate (5 g) was dried to constant weight at 130 °C. Potassium dichromate (0.001 N) was prepared by dissolving 294.2 mg in 15 MX water and diluting to 1 L. Then 30 ml of stock 200 mg/L ferrous ion solution was pipetted into a 100-ml conical flask. The following were added: 2.941 ml of 85% orthophosphoric acid, 1.2 ml of concentrated sulfuric acid, 2.059 ml of 15 MX distilled water, and 25 ll of redox indicator. The clear solution was titrated to a light purple endpoint. After subtracting the indicator correction, the expected titer was 17.90 ml.

Checking accuracy of redox titration with 1,10-phenanthroline As a quality check of the redox titrations, a stock solution of ferrous ammonium sulfate, after correcting the ferrous ion concentration by titration, was further assayed with 1,10-phenanthroline. The ferrous ion procedure of the American Public Health Association was followed [9]. Here 0.1% (w/v) 1,10-phenanthroline solution was prepared by dissolving 100 mg of 1,10-phenanthroline monohydrate in 15 MX water, to which two drops of concentrated HCl had been added, and diluting to 100 ml. Ammonium acetate buffer solution was prepared by dissolving 50 g of ammonium acetate in 30 ml of 15 MX distilled water, and adding 140 ml of glacial acetic acid. Stock 200 mg/L ferrous ion solution prepared from ferrous ammonium sulfate was pipetted (0–1257 ll) into a series of 50-ml volumetric flasks so that the final ferrous ion concentration would range from 0 to 90 lM in 15-lM steps. Then 10 ml of 1,10phenanthroline solution and 5 ml of ammonium acetate buffer solution were added and mixed. The flasks were diluted to 50 ml with 15 MX water. The absorbances were measured at 510 nm at room temperature within 5–10 min. The molar extinction coefficient of the complex ion formed from ferrous ions and 1,10phenanthroline was compared with the literature value of 11,100 [10].

Correction of absorbances for stray light The spectrophotometer used in this work had been checked for equal to or less than 0.5% stray light with 10 g/L NaI at 220 nm and 50 g/L NaNO2 at 340 nm [6]. To obtain accurate molar extinction coefficients, all calibration absorbances measured in this work were corrected for stray light using the stray light equation of Fleming [7] assuming the transmittance of the sample to stray light to be unity. Preparation of stock ferrous ion solutions (200 mg/L Fe2+) Two types of stock ferrous ion solution were used in this work, each containing a nominal 200-mg/L ferrous ion. For stock ferrous sulfate solution, 0.9956 g of ferrous sulfate heptahydrate (MW = 278.02) was dissolved in 800 ml of 15 MX distilled water and diluted to 1 L. For stock ferrous ammonium sulfate solution, 20 ml of concentrated sulfuric acid was slowly added to 50 ml of 15 MX distilled water and allowed to cool. Then 1.404 g of ferrous ammonium sulfate hexahydrate (MW = 392.14) was added to dissolve. The solution was diluted to 1 L. Redox titration of stock ferrous ion solutions Redox titrations were used to determine the precise concentrations of the prepared stock ferrous ion solutions. The redox titra-

Molar extinction coefficient of Fe(TPTZ)22+ complex ion in total iron assay method of Collins and coworkers The method of Collins and coworkers [10] was used without attempting to purify the reagents of extraneous iron. Here 10% hydroxylammonium chloride was prepared by dissolving 2.5 g of hydroxylammonium chloride in 22.5 ml of 15 MX distilled water. Sodium acetate–acetic acid buffer (2 M) was prepared by dissolving 27.216 g of sodium acetate trihydrate and 11.5 ml of glacial acetic acid in 15 MX distilled water and diluting to 100 ml. TPTZ (0.001 M) was prepared by dissolving 31.2 mg of 2,4,6-tris(2-pyridyl)-s-triazine in 100 ll of concentrated HCl and diluting to 100 ml. In the original method, iron standards were prepared from electrolytic iron dissolved in hydrochloric acid. Stock 200 mg/L ferrous ion solution prepared from ferrous ammonium sulfate was pipetted (0–838 ll) into a series of 50-ml volumetric flasks so that the final ferrous ion concentration would range from 0 to 60 lM in 10-lM steps with an additional standard at 2 lM. The following were then added in order: 10 ml of TPTZ, 2 ml of hydroxylammonium chloride (to reduce any ferric ion), and 10 ml of buffer. The flasks were diluted to 50 ml with 15 MX water. The absorbances were measured at 593 nm at room temperature. The molar extinction coefficient of the Fe(TPTZ)22+ complex ion was calculated from the slope of the calibration curve.



3

Molar extinction coefficient of Fe(TPTZ)22+ complex ion in the FRAP assay of Benzie and Strain

Molar extinction coefficient of complex ion Fe(TPTZ)22+ in assay for total iron

A modified version of the manual FRAP assay of Benzie and Strain [2] was used. The FRAP reagent was prepared by mixing 20 ml of 300 mmol/L acetate buffer (pH 3.6), 2 ml of 10 mmol/L TPTZ in 40 mmol/L HCl, 2 ml of 20 mmol/L FeCl36H2O, and 2.4 ml of 15 MX distilled water as diluent. Then 10 standard ferrous ion concentrations in the range of 200–2000 lM in 200-lM steps were prepared by dilution of stock 200 mg/L ferrous ion solution with 15 MX distilled water. A 30-ll ferrous ion standard was pipetted into a 1.5-ml-capacity Eppendorf tube. The reaction was started by adding 1000 ll of FRAP reagent and vortexing for 10 s. The Eppendorf tube was transferred to a 37 °C water bath. The absorbance reading was taken at 593 nm at 4 min in a 1.4-mlcapacity polystyrene cuvette of 1 cm path length. Three blank reaction mixtures were prepared using 30 ll of 15 MX distilled water in 1000 ll of FRAP reagent. The mean blank absorbance was subtracted from the absorbance of each ferrous ion standard to obtain the Fe(TPTZ)22+ absorbance.

The complex ion formed between ferrous ion and two molecules of TPTZ in the total iron assay of Collins and coworkers [10] is the same as that formed in the FRAP assay (i.e., Fe(TPTZ)22+). Although free ferrous ion and unprotonated TPTZ are in equilibrium with the complex ion [14], the formation constant of Fe(TPTZ)22+ is so high that virtually all of the ferrous ion is chelated [14] in the pH range of 3.4–5.8 [10,15]. Although Beer’s law is followed in this pH range [10], sources disagree on the molar extinction coefficient of the complex ion, with reported values ranging from 21,600 to 22,600. Buchanan and coworkers found a value of 22,300 at 594 nm [14]. In kinetic studies, Pagenkopf and Margerum [16] used a lower value of 21,600 at 593 nm, but the origin of this low value is unspecified. Fraser and coworkers [12] gave a range of 22,400 ± 200 at 593 nm. Collins and coworkers are the only authors to report their raw absorbance data [10], which indicate a value of 22,200. It was found in the current work that the molar extinction coefficient in the assay for total iron was 22,230 ± 90 (n = 3), in close agreement with the data of Collins and coworkers [10].

Results and discussion Efflorescence of ferrous sulfate heptahydrate It was noted that calibrations of the FRAP reagent with ferrous sulfate heptahydrate were giving progressively higher absorbances, with a 7.5% rise over 2 months. The ferrous sulfate heptahydrate crystals in this experiment were visually compared with a new batch and found to contain paler blue crystals. Redox titration of a stock solution of the new batch indicated the ferrous ion content to be 2% higher than expected. In both instances, the crystals had undergone efflorescence (i.e., loss of water of crystallization). Mitchell [11] found that ferrous sulfate heptahydrate (FeSO47H2O, MW = 278.02) dehydrates to ferrous sulfate tetrahydrate (FeSO44H2O, MW = 223.97) when the relative humidity is less than 65% or on heating at 40 °C. Mitchell also found that the heptahydrate from various commercial sources usually contained traces of tetrahydrate. This tendency to efflorescence makes the ferrous sulfate heptahydrate calibrator recommended by Benzie and Strain [1] unsuitable as a calibrator for the FRAP assay. To eliminate problems due to efflorescence, ferrous ammonium sulfate hexahydrate in dilute sulfuric acid replaced ferrous sulfate heptahydrate in the preparation of calibration solutions. In past investigations of the ferrous ion–TPTZ reaction, workers have typically used ferrous ammonium sulfate solution acidified with sulfuric acid [12] or electrolytic iron dissolved in hydrochloric acid [10]. In solution, ferrous ions form hydrolyzed ferrous species that are far more prone to spontaneous oxidation than the parent ferrous ion [13]. The benefit of acidification is that at pH values less than approximately 4.0, these species are present at low concentration and the solution oxidizes far more slowly [13].

Checking accuracy of redox titration with 1,10-phenanthroline To check the accuracy of the redox titration in obtaining the correct ferrous ion concentration, the colorimetric reaction between stock ferrous ammonium sulfate and 1,10-phenanthroline was used to obtain an experimental molar extinction coefficient. The good agreement between the experimental molar extinction coefficient of 11,120 (n = 1) and the literature value of 11,100 [10] indicated that the redox titration was accurate.

Molar extinction coefficient of complex ion Fe(TPTZ)22+ in FRAP assay In this work the molar extinction coefficient of the complex ion found in the FRAP assay was found to be 21,140 ± 90 (n = 3). This value is approximately 5% lower than the value of 22,230 ± 90 found with the total iron assay. This difference may be due to differences in temperature and composition between the two reagents but was not explored further. Stratil and coworkers reported a molar extinction coefficient at 593 nm of 19,800 [17] for a calibration made up to 1 mM ferrous sulfate. Benzie and Strain presented absorbances graphically for the reaction of FRAP reagent with ferrous sulfate using a Cobas Fara centrifugal analyzer [2]. Although a molar extinction coefficient was not given, it can be calculated for this instrument using a modified form of Beer’s law:

e¼

A CSA ; 0:001x v

where e is the molar extinction coefficient of the complex ion, A is absorbance, CSA is the cross-sectional area of the cuvette (cm2), x is the ferrous sulfate concentration before mixing with FRAP reagent (mM), and v is sample volume (ml). Interpreting a digital picture of the graph presented by Benzie and Strain [2] by counting pixels obtained a spot value of 0.850 for absorbance A at a ferrous sulfate concentration x of 1 mM. The same article [2] gave 0.01 ml as the value for v. A value of 0.25 cm2 for the cross-sectional area of the Cobas Fara cuvette was not given by Benzie and Strain but was obtained from McMaster and coworkers [18]. From the equation, the molar extinction coefficient e at 593 nm and 37 °C is estimated to be 21,250. Benzie and Strain’s molar extinction coefficient is in reasonable agreement with the value found in the current work. The lower value found by Stratil and coworkers may be due to a spectrophotometer stray light problem causing a negative deviation from Beer’s law with increasing concentration. Pulido and coworkers [3], Goraca and coworkers [4], and Rivero-Perez and coworkers [5] have supplied linear equations relating change in absorbance to ferrous sulfate concentration (before mixing with FRAP reagent). To account for differences in manual assay methodology, these equations are expressed in the following common format:

epred ¼ econst þ ðk=cÞ; where epred is the predicted molar extinction coefficient at concentration c, econst is the molar extinction coefficient constant term, c is


4


Table 1 Predicted molar extinction coefficients for reaction between ferrous sulfate and FRAP reagent at midrange ferrous ion concentration in reaction volume of 30 lM. Data source

Pulido et al. [3]

RiveroPerez et al. [5]

Goraca et al. [4]

econst

22,100 0.02051 2.941– 58.820 595 21,416

21,716 +0.00893 6–48

16,188 0.02725 2.5–60.0

593 22,013

593 15,280

K Range of ferrous ion concentrations in reaction volume (lM) Wavelength (nm) Predicted molar extinction coefficient epred at ferrous ion concentration of 30 lM in reaction volume

and report the molar extinction coefficients for their calibrations for quality control purposes. Table 2 shows that there are compositional differences in the FRAP reagent components in the manual [2] and automated [1,2] versions of the FRAP assay of Benzie and Strain due to the addition of water as a diluent in the automated assay. In addition, Benzie and Strain also suggested the manual assay may be performed at the same temperature as the automated assay (37 °C) or at room temperature [2]. Differences in reagent composition and reaction temperature may influence the molar extinction coefficient. Therefore, it is recommended that future workers using a manual assay adopt the manual procedure presented in the current work that has been designed to nearly match the automated assay. Acknowledgment

Table 2 Composition of reaction mixtures in manual and automated FRAP assays. Type of FRAP assay

Assay volumes Volume of sample (ll) Volume of water as diluent (ll) Volume of reagent (ll) Total assay volume (ll)

Manual assay: Benzie and Strain [2]

Automated assay using Cobas Fara centrifugal analyzer: Benzie and Strain [1,2]

Manual assay: this work

100

10

30

0

30

0

3000

300

1000a

3100

340

1030

Ingredient concentrations in final assay volumes 806.5 735.3 Reagent TPTZ concentration (lM) Reagent ferric 1613.0 1470.6 chloride concentration (lM) 241.9 220.6 Reagent pH 3.6 acetate buffer concentration (mM) 30:1 30:1 Ratio of neat reagent to sample

735.5

1471.0

220.7

30.3:1

a Water is incorporated directly into the reagent as a diluent as described in Materials and methods.

the concentration of ferrous ion in the reaction volume (M), and k is a constant. Table 1 lists the econst and k terms in the three derived equations and the ferrous ion concentrations in the reaction volume over which the original equations were obtained. Table 1 also includes the predicted molar extinction coefficient at a midrange ferrous ion concentration in the reaction volume of 30 lM. At a midrange ferrous ion concentration, two of the predicted molar extinction coefficients are a little higher than expected, but all three groups failed to appreciate that the calibration should follow Beer’s law. The calibration of Goraca and coworkers stands out from the other two because absorbances are much lower than they should be [4]. The FRAP values based on this calibration, therefore, should be treated with caution. Unfortunately, the majority of authors do not disclose FRAP assay calibration data, and so the reliability of their FRAP values remains uncertain. It is recommended that future authors calculate

This work was funded by the charity Dog’s Trust, United Kingdom, as part of the project ‘‘Assessment of Quality of Life in Kennelled Dogs.’’ References [1] I.F.F. Benzie, J.J. Strain, The ferric reducing ability of plasma (FRAP) as a measure of ‘‘antioxidant power’’: the FRAP assay, Anal. Biochem. 239 (1996) 70–76. [2] I.F.F. Benzie, J.J. Strain, Ferric reducing/antioxidant power assay: direct measure of total antioxidant activity of biological fluids and modified version for simultaneous measurement of total antioxidant power and ascorbic acid concentration, in: L. Packer (Ed.), Methods in Enzymology, vol. 299: Oxidants and Antioxidants, Academic Press, Orlando, FL, 1999, pp. 15–27. [3] R. Pulido, L. Bravo, F. Saura-Calixto, Antioxidant activity of dietary polyphenols as determined by a modified ferric reducing/antioxidant power assay, J. Agric. Food Chem. 48 (2000) 3396–3402. [4] A. Goraca, E. Ciejka, A. Piechota, Effects of extremely low frequency magnetic field on the parameters of oxidative stress in heart, J. Physiol. Pharmacol. 61 (2010) 333–338. [5] M.D. Rivero-Perez, P. Muniz, M.L. Gonzalez-Sanjose, Antioxidant profile of red wines evaluated by total antioxidant capacity, scavenger activity, and biomarkers of oxidative stress methodologies, J. Agric. Food Chem. 55 (2007) 5476–5483. [6] Biochrom, WPA Biowave II and Biowave II+ User Manual, version 2.3, Biochrom, Cambridge, UK, 2006, p. 66. [7] P. Fleming, Stray light, in: C. Burgess, T. Frost (Eds.), Standards and Best Practice in Absorption Spectrometry, Blackwell Science, Oxford, UK, 1999, pp. 81–98. [8] L.A. Sarver, I.M. Kolthoff, Indicator corrections for diphenylamine, diphenylbenzidine, and diphenylamine sulfonic acid, J. Am. Chem. Soc. 53 (1931) 2906–2909. [9] American Public Health Association, Standard Methods for the Examination of Water and Wastewater, 21st ed., APHA, Washington, DC, 2005, 3500-Fe B, Phenanthroline method, p. 3-77. [10] P.F. Collins, H. Diehl, G.F. Smith, 2,4,6-Tripyridyl-s-triazine as a reagent for iron, Anal. Chem. 31 (1959) 1862–1867. [11] A.G. Mitchell, The preparation and characterization of ferrous sulphate hydrates, J. Pharm. Pharmacol. 36 (1984) 506–510. [12] F.H. Fraser, P. Epstein, D.J. Macero, The formation constant of Bis[2,4,6-tri(2pyridyl)-1,3,5-triazine]iron(II): evidence for formation of a protonated Bis complex, Inorg. Chem. 11 (1972) 2031–2034. [13] Y. Gendel, O. Lahav, Accurate determination of Fe(II) concentrations in the presence of a very high soluble Fe(III) background, Appl. Geochem. 23 (2008) 2123–2129. [14] E.B. Buchanan Jr., D. Crichton, J.R. Bacon, Study of the factors which affect the reaction of the ferrous ion with tripyridyltriazine, Talanta 13 (1966) 903–909. [15] P.F. Collins, H. Diehl, Determination of iron in wine using 2,4,6-tripyridyl-striazine, Anal. Chim. Acta 22 (1960) 125–127. [16] G.K. Pagenkopf, D.W. Margerum, Formation and dissociation kinetics of Bis(2,4,6-tripyridyl-s-triazine)iron(II), Inorg. Chem. 7 (1968) 2514–2519. [17] P. Stratil, B. Klejdus, V. Kuban, Determination of total content of phenolic compounds and their antioxidant activity in vegetables: evaluation of spectrophotometric methods, J. Agric. Food Chem. 54 (2006) 607–616. [18] D. McMaster, N. Bell, P. Anderson, A.H.G. Love, Automated measurement of two indicators of human selenium status, and applicability to population studies, Clin. Chem. 36 (1990) 211–216.
