Arsenic quantification techniques and ISO/IEC 17025 accreditation in Brazil


 The importance of arsenic (As) quantification in environmental compartments is due to its risks to ecosystems and public health. There are reports of high concentrations of this metalloid in Brazil and technological differences between states are observed. The objective of this work was to present and discuss current scenarios of accreditation and compare the limit of quantification (LOQ) of As by analytical technique in Brazil. Data from accredited laboratories were collected on Inmetro website and in state metrological networks and then grouped and analyzed by state, matrix and analytical technique. There are large discrepancies between the number of laboratories per state and a good correlation with gross domestic product (GDP). Almost all laboratories have a LOQ less than the environmental limits. The observed list of techniques sorted from lowest to highest LOQ values is: for liquid samples ICP MS (inductively coupled plasma mass spectrometry), ET AAS (electrothermal atomic absorption spectrometry), HG AAS (hydride generation combined with atomic absorption spectrometry) or HG ICP OES (hydride generation combined with inductively coupled plasma optical emission spectrometry) and UV VIS (visible ultraviolet spectroscopy); for solids samples HG ICP OES, ICP MS, HG AAS, ET AAS and FAAS (flame atomic absorption spectrometry); and for bioindicators ICP MS, HG ICP OES. Analysis of As species is accredited in only one laboratory, but does not include all species.



Introduction
The environmental risk of arsenic (As) to ecosystems and public health is a fact in a large number of countries around the world . The metalloid can be responsible for various dermatological, cardiovascular, pulmonary, reproductive, neurological and tumorigenesis complications in many parts of the body (Abdul et al., 2015). Analytical techniques for quantification of As on environmental samples typically include atomic absorption, atomic emission and mass spectrometry and, X-ray fluorescence and visible-ultraviolet spectroscopy techniques (Ma et al., 2014;Sankararamakrishnan and Mishra, 2018). These techniques may include flame atomizers, hydride generators and electrothermal devices, such as the graphite furnace (Sankararamakrishnan and Mishra, 2018). Analytical techniques s approved by USEPA include inductively coupled plasma optical emission spectrometry (ICP OES), inductively coupled plasma mass spectrometry (ICP MS), hydride generation combined with atomic absorption spectrometry (HG AAS) and electrothermal atomic absorption spectrometry (ET AAS), with quantification limits ranging from 0.5 to 50 μg L − ¹ (Ma et al. 2014).
In addition to total As quantification, speciation techniques are important to detect and quantify many of As species, which may occur in organic or inorganic forms and in different valence states, with implications on their bioavailability and toxicity (Abdul et al., 2015;Moe et al., 2016;Liu et al., 2018). Some hyphenated techniques consider chromatography and capillary electrophoresis (Khan et al., 2015;Ma et al., 2016;Cheng et al., 2018;García-Rico et al., 2019); ion chromatography (Zhu et al., 2017), ultraviolet spectroscopy and capillary electrophoresis (CE UV) (Lee et al., 2018) for the separation of species.
The maximum stablished limit for drinking water, groundwater and class 1 freshwater is 10 µg L -¹ (Brasil, 2005(Brasil, , 2009(Brasil, , 2017. For effluents, the limit is 500 µg L -¹ (Brasil, 2011). In soils, the value of prevention (VP) is 15 mg kg -¹, the value of agricultural investigation is 35 mg kg -¹, residential investigation is 55 mg kg -¹ and the industrial value is 150 mg kg -¹ (Brasil, 2009). The resolutions of National Environment Council (CONAMA), which enforce the accepted limits, also require that the analyzes for characterization and monitoring be performed by laboratories accredited by the National Institute of Metrology, Standardization and Industrial Quality (INMETRO) (Brasil, 2009(Brasil, , 2011. However, there are no standards for bioindicators, air quality and As species. Brazil has historical regional differences that are reflected in the access to technologies (Santos et al., 2014). According to Grochau et al. (2017) there is a strong correlation between the gross domestic product (GDP) of counties in the Americas and the number of ISO 17025 accredited laboratories.
In Brazil, the main high As concentrations occur in Minas Gerais state , especially correlated to gold deposits in the Iron Quadrangle (Borba et al., 2003;Varejão et al., 2011;Costa et al., 2015;Silva et al., 2018) and in the surroundings the of city of Paracatu (Ono et al., 2012;Rezende et al., 2015;Bidone et al., 2018). In these regions the presence of As is associated with natural deposits of gold, where it is predominantly contained in sulfide minerals such as arsenopyrite and pyrite (Deschamps et al., 2002).
Keeping this in view, the objective of this work was to present and discuss current scenarios of accreditation and compare the limit of quantification (LOQ) of As by analytical techniques in Brazil.

Results and discussion
In Brazil, there are 1081 accredited laboratories, of which 385 offer environmental tests and only 102 analyze As. The highest number of environmental laboratories is concentrated in the Southeast and South regions, which accounts for 89% of the total. For instance, São Paulo (SP) state respond for almost 53%. The number is even higher when considering As analysis, with 96% of the laboratories located in the South and Southeast regions, 67% in SP. Northeast region has only one laboratory in Bahia (BA) while the North has no. Figure 1 shows the distribution of laboratories and GDP by state.
A strong positive correlation (Pearson correlation) was found between the number of accredited laboratories and GDP, as also observed by (Grochau et al., 2017). The highest correlation occurred with the testing laboratories (r = 0.97), followed by testing environmental laboratories (r = 0.94) and finally those performing As analyses (r = 0.92). HG ICP OES or HG AAS are most common techniques used for quantify As in small content (Table 1), corresponding to 80.3% of the accreditations. The ones Rev. Bras. Gest. Amb. Sustent., 2019, vol. 6, n. 14, p. 803-817.
with larger investments and with higher analytical sensitivity, such as ICP MS and ET AAS, accounted for 17.2% and those with lower sensitivity, such as flame atomic absorption spectrometry (FAAS) and UV VIS, only 2.5%. More sensitive techniques are present in only 5 states (SP, MG, RJ, RS and PR). Table 1. Analytical techniques for As quantification of As used in accredited laboratories in different states from Brazil.

Technique
States Considering only laboratories accredited for As analysis, 89% have accreditation in liquid matrices, 42% in solids, 10% in gases and 2% in bioindicators. That distribution probably stems from a greater concern with water potability. Fig. 2 shows the LOQ (Limit of quantification) for As per analytical technique, in the three different matrices: liquid, solid and bioindicator samples.
Almost all laboratories presented a LOQ in accordance with environmental threshold for water, effluent and soil. Many laboratories (31) has a value equal to the potability standard (Brasil, 2017) ( Figure 2a), potentially indicating that it is used as a target. Only one laboratory presented a LOQ greater than the potability standard and is therefore not applicable for this purpose (Figure 2a). This laboratory utilizes the HG AAS, but the occurrence does not prove to be a limitation since the other laboratories and other studies (Borba et al., 2004;Akter et al., 2005;Waterlot and Douay, 2015) presented considerably lower values. Three laboratories presented LOQ values greater than the VP (15 mg kg -1 ) (Brasil, 2009) (Figure 2b), of which two used HG ICP OES and one FAAS. As with other laboratories, which obtained LOQ values below the VP using HG ICP OES, the occurrence is not a limitation of the technique. Only one laboratory uses FAAS and presents LOQ values above the VP. Other authors also obtained LOQ values near or even above the VP, such as 12.8 mg kg -1 with ICP OES (Paye et al., 2010) and 33.3 mg kg -1 with PXRF (García-Rico et al., 2019).
Techniques sorted from lowest to highest LOQ, considering the lowest LOQ, was similar among the matrices. In the case of liquid matrices, it was: ICP MS (40 ηg L -1 ), ET AAS (50 ηg L -1 ), HG AAS or HG ICP OES (100 ηg L -1 ) and UV VIS (4 µg L -1 ) (Fig. 2a). In solid matrices: HG ICP OES (1 µg kg -1 ), ICP MS (1.5 µg kg -1 ), HG AAS (10 µg kg -1 ), ET AAS (50 µg kg -1 ) and FAAS (100 mg kg -1 ) (Figure 2b). In bioindicators: ICP MS (10 µg kg -1 ) and HG ICP OES (50 µg kg -1 ) (Figure 2c). In a study comparing practically the same analytical techniques, Hung et al. (2004)    showed in studies involving As quantification in the three matrices considered in the present data survey.   (Oliveira et al. 2016) In our data survey, it was possible to observe that the ICP MS presented a LOQ ranging from 0.04 to 4.00 µg L -1 in liquids, 1.50 to 1,000 µg kg -1 in solids and 10.00 μg kg -1 in bioindicators (Table 2) The ET AAS showed a LOQ between 0.05 and 10.00 μg L -1 in liquid samples and between 50.0 and 1,000 μg kg -1 in solids (Table 2) The two techniques that use HG presented similar LOQ values. The HG AAS had a LOQ ranging from 0.1 to 11 μg L -1 in liquid samples and from 10 to 6,000 μg kg -1 in solid samples; while the HG ICP OES ranged from 0.1 to 10 μg L -1 in liquid samples, from 1.0 to 20,000 μg kg -1 in solid samples and from 50 to 5,000 μg kg -1 in bioindicators (Table 2). Other studies using HG presented values within the same ranges found. When using HG AAS to analyze water, Akter et al. (2005)  100,000 μg kg -1 in solid samples (Table 2). These techniques are also less used in scientific works, which in addition to the techniques previously discussed, use techniques that involve fluorescence, microwave induced plasma, electron capture detector and X-ray in diverse matrices. In soil samples, Pierangeli et al. (2015) Table 3 present the LOQ for As species obtained using different analytical techniques and compare it with that showed in studies involving As quantification in the three matrices considered in the present data survey. Only one laboratory is accredited for As speciation. However, it does not cover all species analyzed in scientific studies, quantifying only the species As III, As V, Monomethyl Arsenic (MMA), Dimethyl Arsenic (DMA) and Arsenic Betaine (AsB). Species such as trimethylarsine oxide (TMAO) found in plants (Bergqvist and Greger, 2012) or Arsenocoline (AsC) found in marine animals (Raber et al. 2012), are not accredited by any Brazilian laboratory.
The LOQ for As species in liquid samples was 0.25 μg L -1 (Table 3) for As III, As V, MMA, DMA and AsB. These values are similar to other reports which used similar techniques to determine As species. One study presented a lower LOQ when using solid phase extraction (SPE) coupled to ET AAS, obtaining 0.006 μg L -1 in the determination of As(III) and As(V), but it should be noted that other species were not determined (Hassanpoor et al. 2015). Using high performance liquid chromatography (HPLC) combined with ICP MS, Liu e Cai (2013) obtained values between 0.13 and 1.66 μg L -1 , Komorowicz and Barałkiewicz (2016) between 0.22 and 0.39 µg L -1 and Cheng et al. (2018) between 3.00 and 5.90 μg L -1 for the determination of As III, As V, MMA, and DMA. Where in addition to the previous species, Komorowicz and Barałkiewicz (2016) also determined AsB and Cai (2013) quantified ROX andAsA. Akter et al. (2005) used HG AAS and obtained between 0.33 and 0.63 μg L -1 for determination of As III and DMA. Waterlot and Douay (2015) found 0.41 μg L -1 when determining As III and V by the same technique. Using LC ICP MS, Akter et al. (2005) obtained between 0.33 and 0.67 μg L -1 for the determination of As III, As V, DMA and MMA. Some studies have higher LOQ values because they use less sensitive techniques, such as Gürkan et al. (2015) that obtained 3.00 μg L -1 for As V when using UV VIS. With CE UV Lee et al. (2018) obtained values between 7.90 and 20.0 μg L -1 for As III, As V, MMA and DMA and Akter et al. (2005) reported between 300 and 1600 μg L -1 for As III, As V and DMA.
For As species in solid samples and bioindicators, the LOQ was 12.5 μg kg -1 ( Table 3) for As III, As V, MMA, DMA and AsB. The LOQ of the laboratories was considerably lower than those achieved by Wang et al. (2018) which obtained values between 400 and 1000 μg kg -1 for the determination of As III, V, DMA and MMA in Eisenia fetida using HG AFS. Other studies obtained smaller LOQ values in several matrices and analytical techniques, mostly using HPLC HG ICP OES, as in Khan et al. (2015) with LOQ between 0.06 and 0.20 μg kg -1 for MMA, AsB, As V, DMA, AsC and As III in algae. For quantification of As(III), As(V), MMA and DMA using same technique, García-Rico et al. (2019) found values ranging from 0.40 up to 0.70 μg L -1 in urine. Ma et al. (2016) found values between 0.70 and 3.00 μg kg -1 in rice and Al-Assaf et al. (2009) reported values between 1.30 and 3.30 μg L -1 in soils. Using HPLC ICP MS, Wolf et al. (2011) obtained 0.05 μg L -1 for As III and V in soils. Working with rice samples,  showed As values ranging from 0.70 and 3.00 μg kg -1 with HG AFS. Zhu et al. (2017) found lower values between 0.10 and 0.20 μg L -1 with ion Rev. Bras. Gest. Amb. Sustent., 2019, vol. 6, n. 14, p. 803-817. chromatography (IC) combined with ICP MS. Using HG AAS, Santos et al. (2018) obtained values between 0.07 and 0.10 μg L -1 for As III and V in fish.