ANALYTICAL SCIENCES JANUARY 2018, VOL. 34 103

Introduction

The prevalence and incidence of kidney stone disease are increasing across the world1 because of increases in extreme heat and water shortages due to climate change, which is increasing our risk for dehydration-associated kidney diseases.2 Kidney stone disease affects up to 20% of the general population worldwide.3 Some studies report that up to 12% of men and 6% of women will develop a renal stone at some point in their life in the United States.4 In Thailand, kidney stone disease is a longstanding medical illness and remains a common public health problem. A  report from The Ministry of Health in Thailand found that the incidence of kidney stones in the urinary tract of patients increased from 99.25 per 100000 people in 2007 to 122.46 in 2010. Furthermore, the highest prevalence was found in the Northeast provinces, where the prevalence was up to 16.9% for the age range of 40 – 50 years, the incidence was more common in males than females by up to 2 times, and the rates of recurrence within 2 years after the surgery were 39%.5,6 Kidney stone disease usually recurs, and it poses difficulty in management and burdensome medical costs.7 Recurrence rates as high as 50% within 10 years have been documented.8 The major constituent of a kidney stone is calcium oxalate, with more than two-thirds of other renal stone constituents (i.e., uric acid and calcium phosphate). Normal levels of oxalate for a healthy human are 10 – 30 mg/24 h

(100 – 300 μM) in urine9 and 0.8 – 2.50 μM in plasma.10 A  urinary oxalate concentration higher than 300 μM can indicate the occurrence of initial kidney stones. Kidney stones can cause renal function deterioration and kidney failure in end-stage renal disease. The occurrence of kidney stones has many causes, such as reduced fluid intake, reduced urine output, high dietary consumption and an abnormal metabolic system. Therefore, an accurate, efficient, convenient, fast and affordable method to measure oxalate is important to help patients and increase general public awareness of the risk of kidney stone disease. Kidney stones can also be prevented before they become severe. Many methods have been reported for measuring oxalate in food, urine or plasma, such as spectrophotometry,11 fluorometry,12,13 capillary electrophoresis,14 ion chromatography,15 high-performance liquid chromatography16 and gas chromatography.17 Although these methods can detect oxalate with high sensitivity and accuracy, they often require large sample volumes, special equipment, and/or an expert technician. Hence, these methods cannot be applied for monitoring urinary oxalate at the point of care. The colorimetric method for oxalate determination is an alternative strategy for point-of-care monitoring due to its conveniences of visual observation and simple operations. Colorimetric methods allow direct analysis by the naked eye without costly instruments, compared with the other methods. In 2013, Ruan and coworkers reported the determination of the total oxalate contents of a great variety of foods commonly available in southern China using an oxalate oxidase method (colorimetric detection).18 Then, Tang and Liu developed a new chemosensing ensemble for the colorimetric detection of oxalate in water at pH 7.4.

2018 © The Japan Society for Analytical Chemistry

† To whom correspondence should be addressed.E-mail: wijitar.dun@kmutt.ac.th

A Simple Paper-based Colorimetric Device for Rapid and Sensitive Urinary Oxalate Determinations

Piyakorn WORRAMONGKONA,* Kanyarat SEEDA,* Phatharatanathorn PHANSOMBOON,* Nalin RATNARATHORN,* Orawon CHAILAPAKUL,** and Wijitar DUNGCHAI*†

* Department of Chemistry, Faculty of Science, King Mongkut’s University of Technology Thonburi, Prachautid Road, Thungkru, Bangkok, 10140, Thailand

** Center of Excellence for Petroleum, Petrochemicals, and Advanced Materials, Chulalongkorn University, Patumwan, Bangkok, 10330, Thailand

In this work, a simple and inexpensive paper-based colorimetric device (cPAD) was developed for oxalate measurements. The colorimetric assay is based on the formation of formazan via the reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) by oxalate decarboxylase and formate dehydrogenase on the paper device. After the sample was spotted on the device, MTT changed color from yellow to purple, and the purple color intensity correlated with the oxalate concentration. The quantitative detection of oxalate ranged from 10 – 1000 μM, with a linear equation, y = 0.0086x + 34.978, and a correlation coefficient (R2) = 0.9994. The detection limit was 10 μM by the naked eye. The recoveries of oxalate spiked in artificial urine samples evaluated by our cPADs were in the range of 81 – 92%. This simple cPAD for rapid and sensitive oxalate determination should be useful for diagnostic urinary oxalate measurements for point-of-care monitoring.

Keywords Oxalate, paper-based colorimetric device, colorimetric assay

(Received August 30, 2017; Accepted September 22, 2017; Published January 10, 2018)

104 ANALYTICAL SCIENCES JANUARY 2018, VOL. 34

The ensemble was constructed by a copper complex, CuL, and Chrome Azurol S.19 However, these devices for colorimetric detection of oxalate still require a UV-vis spectrometer.

Paper-based colorimetric devices (cPADs) have the potential to be better alternatives for point-of-care testing over the traditional colorimetric method because they are easy to use, inexpensive, require small volumes of reagents and samples, provide rapid analysis, are disposable, can be made from renewable substrate materials, and are portable. Therefore, cPADs have been used for monitoring samples in many fields, such as diagnosis (i.e., glucose, protein, pH, and G6PD), food safety (i.e., copper and nitrite) and environmental analysis (i.e., cadmium and lead). In this work, we present a simple and rapid cPAD for the determination of urinary oxalate using the formation of formazan via the reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) by oxalate decarboxylase and formate dehydrogenase (as shown in Fig. 1). After the sample was spotted on the device, MTT changed color from yellow to purple, and the purple color intensity correlated with the oxalate concentration. The optimal conditions including reagent concentration and analysis time were investigated and are shown in the following sections.

Experimental

Reagents and chemicalsAll chemicals used in the experiment were of analytical

reagent (AR) grade. Creatinine, formate dehydrogenase, 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT), nicotinamide adenine dinucleotide (NAD+), oxalate decarboxylase, 1-methoxy-5-methylphenazinium methyl sulfate (PMS), and urea [Co(NH2)2] were obtained from Sigma-Aldrich (St. Louis, MO). Monobasic sodium phosphate (NaH2PO4·2H2O), dibasic sodium phosphate (Na2HPO4), and magnesium sulfate (MgSO4·7H2O) were bought from Panreac (Barcelona, Spain). Ammonium chloride (NH4Cl), calcium chloride (CaCl2·2H2O), citric acid (C6H8O7·H2O), potassium chloride (KCl), sodium chloride (NaCl), sodium oxalate (Na2C2O4), tri-sodium citrate (Na3C6H5·2H2O) and uric acid were bought from Fisher Scientific (Pittsburgh, PA). Sodium hydrogen carbonate (NaHCO3) was purchased from Merck (Darmstadt, Germany). Whatman No. 1 filter paper was bought from Cole-Parmer

(Vernon Hills, IL). All glassware was thoroughly cleaned with freshly prepared 3:1 HCl/HNO3 and rinsed with Milli-Q water prior to use. NAD+ was dissolved in phosphate buffer, pH 7.4, whereas sodium oxalate (Na2C2O4) was prepared using a citrate-phosphate buffer, pH 5.0. Milli-Q water from a Millipore system (R ≥ 18.2 MΩ cm–1) was used throughout the experiments.

ApparatusThe results of the colorimetric detection were recorded with a

scanner (Scx-3405, Samsung) using 600-dpi resolution, and the color of the whole spot was analyzed by ImageJ (RGB color mode).20,21

Preparation of paper-based colorimetric device (cPAD)The pattern of the cPAD was created in Corel Draw designs

(Fig. 1), and these patterns were transferred to a wax printer (ColorQube, Xerox, Thailand) using Whatman No. 1 paper. Then, they were baked at 110°C in an oven for 5 min. The backside of the cPADs was attached with adhesive tape.

Optimization of the measurement conditionsPrior to the quantification of artificial urinary oxalate samples,

the optimal conditions, including 1-methoxy-5-methyl-phenazinium methyl sulfate (PMS) at 1, 3 and 5 mM, 3-(4,5- dime thylthiazol-4-yl)-2,5-diphenyl tetrazolium bromide (MTT) at 0.1, 0.3, 0.4 and 0.5 M, and nicotinamide adenine dinucleotide (NAD+) at 5, 10, 25, 40 and 57 mM, were investigated at various time points, 5, 15, 30 and 45 min. The reactants were dropped on the center of the cPAD: 1 μL of each reagent was sequentially dropped, followed by 3 units/mL oxalate decarboxylase, 40 mM of NAD+, 50 units/mL formate dehydrogenase, and MTT, and the droplets were allowed to dry prior to the addition of the next reagent. During the optimization, 5 μL of 200 μM oxalate were mixed with PMS (ratio 4:1) and spotted on the loading zone. The sample or standard solutions were incubated for 30 min at room temperature. After incubation, the cPADs were recorded with the scanner (Scx-3405, Samsung) using 600-dpi resolution, and then the intensity of the colors were analyzed with ImageJ (RGB color mode) using the whole spot (3 mm). The intensity difference was calculated by the purple color intensity generated from the sample or standard oxalate solution after subtraction of the background color intensity.

Fig. 1　Pattern of cPAD for determination of oxalate.

ANALYTICAL SCIENCES JANUARY 2018, VOL. 34 105

Analytical performanceThe linear relationship between the oxalate concentration and

color intensity was investigated using the optimized conditions and an oxalate standard concentration range of 0 – 1000 μM. Since our assay is related to the formate pathway (Fig. 1), formate concentrations in the range of 0 – 2000 μM were also studied to quantify formate interference. The precision of our cPAD was then investigated using replicated detection of 200 and 400 μM oxalate (n = 10) solutions under the optimized conditions.

ApplicationsTo evaluate the accuracy and utility of our proposed cPAD, the

oxalate concentrations in artificial urine samples were quantified by our cPADs. The artificial urine sample without oxalate was prepared by the method of Chutipongtanate and Thongboonkerd22 and contained 0.44 g of CaCl2·2H2O, 0.45 g of creatinine, 2.25 g of KCl, 0.50 g of MgSO4·7H2O, 1.48 g of Na3C6H5·2H2O, 3.17 g of NaCl, 0.17 g of NaHCO3, 0.05 g of Na2HPO4, 0.05 g of NaH2PO4·2H2O, 1.29 g of Na2SO4, 0.80 g of NH4Cl, 12.13 g of urea and 0.17 g of uric acid in 1000.00 mL of deionized water (Solution A). Artificial urine samples of healthy people (200 μM oxalate, S1 – S2) were prepared by weighing 0.026 g of Na2C2O4, and artificial urine samples of unhealthy people (400 μM oxalate, S3 – S4) were prepared by weighing 0.053 g of Na2C2O4. All samples were adjusted with Solution A  to a final volume of 1000.00 mL. Prior to analysis, the spiked samples were diluted 5 times with citrate-phosphate buffer, pH 5.0, to adjust the pH of the sample. The analysis of each sample was replicated 3 times (n = 3).

Results and Discussion

Determination of the oxalate concentrationPaper-based colorimetric devices (cPADs) have the potential

to be better alternatives for use as point-of-care devices or health care devices over the traditional colorimetric method. They are gaining a great of deal attention due to their applications in many interdisciplinary areas. The advantages include their hydrophilic nature, rapid fabrication, easy of transferring, ease of handling, the ultra-low cost of the materials, high wicking properties, high porosity, disposability, non-toxicity, biodegradability, biocompatibility, ease of use, requirement for small volumes of reagents and samples, rapid analysis time and portability. The white color of the filter paper and its hydrophilic nature facilitates its usage in rapid colorimetric assays without any requirement of external forces to drive the analyte into the detection zone.

Two previous enzymatic systems were reported for colorimetric oxalate detection: oxalate oxidase and oxalate decarboxylase. In the presence of oxalate, oxalate oxidase produces hydrogen peroxide, whereas oxalate decarboxylase produces formate and carbon dioxide. Oxalate decarboxylase degraded oxalic acid faster than oxalate oxidase.23 Furthermore, oxalate decarboxylase was not as sensitive to chlorate and chlorite as was oxalate oxidase. Up to 4 mM chlorate ions, as reported in previous studies as the highest concentration tested, had no inhibitory effects on oxalate decarboxylase, whereas this concentration had an effect on oxalate oxidase activity. Thus, we chose the oxalate decarboxylase enzymatic system for the cPAD, as shown in Fig. 1. The concentration of enzymes was 3 units/mL oxalate decarboxylase and 50 units/mL formate dehydrogenase, which are higher concentrations than those used in solution tests. Although we tried to reduce the concentration of enzymes to

minimize analysis cost, the color change could not be noticed by the naked eye between the absence and presence of 10 mM oxalate ion at less than 3 units/mL oxalate decarboxylase and 50 units/mL formate dehydrogenase. The activity of the immobilized enzyme was much lower compared to free enzyme in solutions, due to damage during immobilization and mass transfer limitations inside the solid support system. However, the cost per test is still lower for the cPAD than for a solution test, according to the microliter volume of enzyme used in our cPAD.

The MTT assay is normally applied for a colorimetric assay for assessing cell metabolic activity. NAD(P)H-dependent cellular oxidoreductase enzymes may, under defined conditions, reflect the number of viable cells present. These enzymes are capable of reducing the tetrazolium dye MTT 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide to insoluble formazan, which has a purple color. The insoluble purple formazan was selected for our cPAD because this compound reduces the distribution of non-equal flow along the edges of the loading zones. Moreover, the yellow-to-purple color change in the presence of oxalate is easy to discern by the naked eye. In  the next sections, the optimization of the PMS, MTT and NAD+ concentrations were studied.

Effects of the PMS, MTT and NAD+ concentrationsThe concentrations of PMS, MTT and NAD+ have a direct

effect on the purple color intensity, and thus we optimized the PMS concentration by comparing 1, 3, and 5 mM. The PMS solution was saturated at concentrations above 5 mM. An increase in the intensity difference was observed as the PMS concentration increased from 1 to 3 mM, as shown in Fig. 2. The intensity difference was not significantly different between 3 and 5 mM PMS, so we selected 3 mM as the optimized PMS concentration.

The optimal MTT concentration was also investigated from 0.1 to 0.5 M. When the MTT concentration increased from 0.1 to 0.3 M, the intensity difference dramatically increased, but the intensity difference remained constant at concentrations above 0.3 M MTT, as shown in Fig. 3. Thus, we chose an MTT concentration of 0.3 M for further experiments.

The NAD+ concentrations were varied from 5 to 57 mM. The results showed that the intensity difference increased as the NAD+ concentration increased. The optimized NAD+ concentration was selected at 40 mM because the intensity difference was not significantly different between 40 and 57 mM, as shown in Fig. 4.

Fig. 2　Optimization of PMS concentration at 0.3 M MTT, 57 mM NAD+, and 200 μM oxalate. The color intensity was recorded at 30 min analysis time.

106 ANALYTICAL SCIENCES JANUARY 2018, VOL. 34

Effect of analysis timeThe chemical and enzymatic reactions require time to reach

their equilibria; therefore, the analysis time after sample loading requires optimization. Images of the resulting colors were recorded at 5 min and then recorded every 15 min until 45 min of the reaction time had elapsed. It was observed that longer reaction times gave higher intensity differences (Fig. 5). We selected 30 min of analysis time as a compromise between a rapid and sensitive detection of oxalate.

Analytical performanceAfter determining the optimal conditions, the linearity between

the intensity difference and the oxalate concentration (0 – 1000 μM) was demonstrated. Quantitative detection of oxalate was found for concentrations from 10 to 1000 μM, with the linear equation y = 0.0086x + 34.978 and correlation coefficient (R2) = 0.9994 (Fig. 6). The threshold levels of oxalate in human urine are 100 – 300 μmol/L and over 300 μM, respectively, for healthy and unhealthy individuals. Our device should therefore be adequate for the determination of oxalate concentration in human urine samples. The detection limit was found to be 10 μM by the naked eye. Our linear range was wider, and our LOD was 5 times lower than those of a paper-based device with an oxalate peroxidase coating (100 – 2500 μM, LOD = 50 μM) as shown in Table 1.24 Moreover, our cPAD required lower sample and reagent volumes than the device

described in a previous report.24

Next, we studied formate interference from the related formate pathway (Fig. 1) using real urine samples containing formate (0 – 400 μM in healthy humans).25 Furthermore, formate is derived from the metabolism of formaldehyde, and several other industrial compounds and some drugs may elevate the formate concentration in urine above the normally expected values. The formate concentrations in the urine samples of humans who had been exposed to chemicals were reported as 100 – 2000 μM.25 Hence, formate concentrations in the range of 0 – 2000 μM were also studied to quantify formate interference. We found that the intensity difference was not significantly different among all formate concentrations (average intensity difference of all formate conc. = 25.249 ± 5.273, n = 12) due to the rapid evaporation of formic acid (Fig. 7). Although a high concentration of formate was dropped on the cPAD, the enzymatic rate of formate dehydrogenase is slower than the evaporation rate of formic acid. Hence, our cPAD can be used for the determination of the urinary oxalate concentration without formate interference under the optimized conditions. The precision is expressed as the relative standard deviation (RSD%), obtained from a series of 12 standard samples

Fig. 3　Optimization of MTT concentration at 3 mM PMS, 57 mM NAD+, and 200 μM oxalate. The color intensity was recorded at 30 min analysis time.

Fig. 5　Analysis time study at 0.3 M MTT, 3 mM PMS, 200 μM oxalate, and 57 mM NAD+.

Fig. 4　Optimization of NAD+ concentration at 0.3 M MTT, 3 mM PMS, and 200 μM oxalate. The color intensity was recorded at 30 min analysis time.

Fig. 6　The linearity response under the optimized conditions.

ANALYTICAL SCIENCES JANUARY 2018, VOL. 34 107

containing 200 and 400 μM of oxalate, and was less than 7%. According to the AOAC Guidelines for Standard Method Performance Requirements, the %RSD should be lower than 8%.26

ApplicationsTo evaluate the paper-based colorimetric device with real

samples, three determinations of the oxalate concentration in artificial urine samples spiked with a known oxalate standard solution were performed using the optimized conditions. Artificial urine samples were spiked with known oxalate concentrations at 200 and 400 μM to represent healthy and unhealthy people, respectively. The recoveries of spiked oxalate in the artificial urine samples evaluated by our cPADs were 81 – 92% (Table 2). Our recoveries were lower than 100% in all artificial urine samples because some salts have an effect on oxalate decarboxylase activity.23 However, the AOAC recommends % recoveries in the range of 80 – 110%, so our device is acceptable and can be applied for oxalate determination in artificial urine samples.

Conclusions

In this study, we developed a paper-based colorimetric device that is easy to use for the determination of urinary oxalate with highly sensitivity (LOD = 10 μM), acceptable repeatability

(<7% RSD) and rapidity (analysis time = 30 min). Moreover, the analysis was inexpensive and did not require complex equipment or extensive training of users. Our cPAD was successfully applied to determine the levels of oxalate in artificial urine samples (81 – 92% recoveries). Our device, therefore, represents a path for the development of a portable diagnostic assay that may be useful in remote settings and important for the diagnosis of kidney stone disease in point-of-care monitoring.

Acknowledgements

We gratefully acknowledge the financial support from Thailand Research Fund through MRG6080171, and Post-doctoral Fellowship, Research, Innovation and Partnerships Office, King Mongkut’s University of Technology Thonburi.

References

1. V. Romero, H. M. D. Akpinar, G. M. D. Dean, and M. D. Assimos, Rev. Urol., 2010, 12, e86.

2. R. J. Johnson, P. Stenvinkel, T. Jensen, M. A. Lanaspa, C. Roncal, Z. Song, L. Bankir, and L. G. Sánchez-Lozada, J. Am. Soc. Nephrol., 2016, 8, 2247.

3. A. Ramello, C. Vitale, and M. Marangella, J. Nephrol., 2000, 13, S45.

4. K. K. Stamatelou, M. E. Francis, C. A. Jones, L. M. Nyberg, and G. C. Curhan, Kidney Int., 2003, 63, 1817.

5. M. Yanagawa, J. Kawamura, T. Onishi, N. Soga, K. Kameda, and P. Sriboonlue, Int. J. Urol., 1997, 4, 537.

6. P. Chanapa, Songkla Med. J., 2011, 29, 299.

Fig. 7　Study of formate interference.

Table 2　Determination of urinary oxalate using our cPAD under the optimized conditions

Artificial urine sample

Spiked oxalate/μM

Detected oxalate/μM ± SD (n = 3)

Recovery, %

S1 200 161 ± 11.1 81S2 200 165 ± 9.82 83S3 400 369 ± 8.54 92S4 400 354 ± 9.70 89

Table 1　Comparison of oxalate determination with different material/method

Material/method LOD/μM Linear range/μM Real sample Instrument Portability Ref.

OXD/CO2-sensing electrode 40 200 – 10000 Urine pH meter Yes 27OXO/HRP TiO2/TB-CPE 90 100 – 2000 Spinach Potentiostat No 28OXO/PB/self-doped polyaniline film 80 80 – 450 NA Potentiostat No 29Pd nanoparticles modified CPE 20 20 – 10000 NA Potentiostat No 30An electrode modified with Fe(III)-tris-(2-thiopyridone) borate

1000 3000 – 50000 NA FIA-potentiostat No 31

TiO2 nanoparticles/multiwalled carbon nanotubes modified electrode

33 100 – 10001 × 103 – 1 × 105

Spinach Potentiostat No 32

A carbon ionic liquid electrode (CILE) modified with TiO2-Fe nanoparticles

23 50 – 3000 Urine Potentiostat No 33

Oxalate peroxidase coated on paper-based device

50 100 – 2500 NA Instrument free Yes 24

Paper-based colorimetric device 10 10 – 1000 Urine Instrument free Yes Our method

OXO, Oxalate oxidase; OXD, oxalate decarboxylase; PB, Prussian Blue; HRP, horseradish peroxidase; TB, Toluidine Blue; CPE, carbon paste electrode. NA = no analysis.

108 ANALYTICAL SCIENCES JANUARY 2018, VOL. 34

7. W. G. Robertson, Urol. Res., 2006, 4, 157. 8. A. Trinchieri, F. Ostini, R. Nespoli, F. Rovera, E. Montanari,

and G. Zanetti, J. Urol., 1999, 162, 27. 9. C. S. Pundir, M. Thakur, and Satypal, Clin. Chem., 1998,

44, 1364. 10. C. S. Pundir, N. K. Kuchhal, M. Thakur, and Satypal, Ind.

J. Biochem. Biophys., 1998, 35, 120. 11. P. M. Ladwing, R. R. Liedtke, T. S. Larson, and J. C.

Lieske, Clin. Chem., 2005, 51, 2377. 12. Z. Shengrui, W. Qin, T. Guanghui, and G. Hongguang,

Mater. Lett., 2014, 115, 233. 13. Q. Y. Chen, D. H. Li, Q. Z. Zhu, H. Zheng, and J. G. Xu,

Anal. Chim. Acta, 1999, 381, 175. 14. J. A. Muñoz and M. Lopez-Mesas, Talanta, 2010, 81, 392. 15. J. K. Kristen, Anal. Chim. Acta, 1990, 237, 299. 16. J. Perello, P. Sanchis, and F. Grases, J. Chromatogr. B,

2005, 824, 175. 17. H. Li, X. S. Chai, N. DeMartini, H. Zhan, and S. Fu, J.

Chromatogr. A, 2008, 1192, 208. 18. Q. Y. Ruan, X. Q. Zheng, B. L. Chen, Y. Xiao, X. X. Peng,

D. W. M. Leung, and E. E. Liu, J. Food Compos. Anal., 2013, 32, 6.

19. L. J. Tang and M. H. Liu, Bull. Korean Chem. Soc., 2010, 31, 3159.

20. T. Ferreira and W. Rasband, Fiji 1.46, 2012, 1. 21. T. M. G. Cardoso, P. T. Garcia, and W. K. T. Coltro, Anal.

Methods, 2015, 7, 7311.

22. S. Chutipongtanate and V. Thongboonkerd, Anal. Biochem., 2010, 402, 110.

23. P. Cassland, A. Sjöde, S. Winestr, L. J. Jönsson, and N. O. Nilvebrant, Appl. Biochem. Biotechnol., 2010, 161, 255.

24. J. Manassawee, R. Poomrat, C. Orawon, and S. A. Monpichar, in Proceeding of Pure and Applied Chemistry International Conference 2017 (PACCON2017), 2017, Bangkok, Thailand, 124.

25. F. B. Mark, Am. Ind. Hyg. Assoc. J., 1987, 48, 900. 26. Office Methods of Analysis, “Appendix F: Guidelines for

Standard Method Performance Requirements”, 2016, AOAC International, Rockville, MD, USA.

27. R. K. Kobos and T. A. Ramsey, Anal. Chim. Acta, 1980, 121, 111.

28. E. F. Âtima Perez, G. de Oliveira Neto, and L. T. Kubota, Sens. Actuators, B, 2001, 72, 80.

29. P. A. Fiorito and S. I. Córdoba de Torresi, Talanta, 2004, 62, 649.

30. L. G. Shaidarova, I. A. Chelnokova, A. V. Gedmina, G. K. Budnikov, S. A. Ziganshina, A. A. Mozhanova, and A. A. Bukharaev, J. Anal. Chem., 2006, 61, 375.

31. J. A. Rodriguez, P. Hernandez, V. Salazar, Y. Castrillejo, and E. Barrado, Molecules, 2012, 17, 8859.

32. A. R. Fakhari, B. Rafiee, H. Ahamar, and A. Bagheri, Anal. Methods, 2012, 4, 3314.

33. M. Akhond, G. Absalan, A. Tafakori, and H. Ershadifar, Anal. Bioanal. Chem. Res., 2016, 3, 73.