Environ Monit Assess DOI 10.1007/s10661-013-3354-5
Comparison of USEPA digestion methods to heavy metals in soil samples Ygor Jacques Agra Bezerra da Silva & Clístenes Williams Araújo do Nascimento & Caroline Miranda Biondi
Received: 18 February 2013 / Accepted: 15 July 2013 # Springer Science+Business Media Dordrecht 2013
Abstract The use of appropriate analytical methods is of paramount importance for risk assessment and monitoring of potentially toxic metals in soils. In this sense, the objective of this study was to compare the effectiveness of two sample digestion methods, recommended by the Brazilian legislation for the management of contaminated areas (CONAMA 2009), aiming at the determination of environmentally available metal concentrations (USEPA 3050B, USEPA 3051A), as well as a total digestion method (USEPA 3052). Samples from 10 classes of soils were analyzed for Cu, Zn, Cd, Pb, Ni, and Hg. The results showed that the USEPA method 3051A is more efficient than the USEPA method 3050B in the extraction of levels considered environmentally available of Zn, Cu, Cd, Pb, and Ni. Besides providing a higher recovery of these elements, the method requires shorter digestion time, lower consumption of acids, and reduced risk of contamination. The USEPA method Y. J. A. B. da Silva Departamento de Agronomia, UFRPE, Endereço: Rua Dom Manuel de Medeiros, S/N, Dois Irmãos, Brazil e-mail:
[email protected] C. W. A. do Nascimento (*) : C. M. Biondi UFRPE, Endereço: Rua Dom Manuel de Medeiros, S/N, CEP: 52171-900 Dois Irmãos, Recife, PE, Brazil e-mail:
[email protected] C. W. A. Nascimento e-mail:
[email protected] C. M. Biondi e-mail:
[email protected]
3051A showed greater efficiency in Hg extraction in soils with higher clay content. Therefore, it is suitable for situations where a wide range of soils with different mineralogical characteristics are analyzed or in order to decrease the losses due to volatilization of the element in open systems. Keywords Trace elements . Soil pollution . Chemical extractants . Atomic absorption spectroscopy
Introduction The increasing occurrence of soil contamination by heavy metals has heightened concerns about environment quality, as these elements cause changes in the structure and functioning of ecosystems, as well as posed health risks to humans. In this context, the use of accurate methods for monitoring heavy metals in soils is of great importance for risk assessment (Guven and Görkem 2011). Such studies must carefully consider the attention required in the laboratory analytical determination, which involves digestion methods, choice of adequate instrumentation as well as precision and accuracy in metal determination in the samples. In general, acid digestion procedures are used to convert solid samples into liquid extracts to quantify the total or pseudototal concentrations of metallic elements in soils. This principle consists of releasing the metals present in the solid matrix to the acidic solution during the extraction process, this procedure being necessary for metal determination by conventional methods,
Environ Monit Assess
such as inductively coupled plasma optical emission spectrometry or atomic absorption spectroscopy. Several acid digestion methods for the determination of heavy metals in soils have been described in the literature. They range from milder attacks, such as aqua regia in an open system, to the use of hydrofluoric acid in a closed system, which is considered a total digestion for destroying the silicate matrices (Chen et al. 1998). Due to the large variation in metal content obtained by different methods, the digestion of samples is the principal factor contributing to the uncertainty of analytical results (Kántor 2001; Axelsson and Rodushkin 2001; Belarra et al. 2002; Al-Harahsheh et al. 2009). Currently, this variation in methods hinders the comparison of obtained data; thus, it is essential that the regulatory agencies standardize the method used to determine metal concentrations in soils. In Brazil, the National Environment Council (CONAMA 2009) stipulates that for regulatory purposes, the USEPA methods 3050 and 3051 or their updates must be used in the digestion of soil samples for heavy metal determination. The USEPA methods 3050 and 3050B are considered conventional procedures because they are conducted in an open system, in which the elements in the solid phase are extracted by a heat source in the presence of nitric and hydrochloric acids. This method has the disadvantage of atmospheric contamination risk and loss of more volatile elements (Nieuwenhuize et al. 1991), such as Hg. The USEPA method 3051A is a modification of the method 3051, requiring the addition of hydrochloric acid (HCl) with nitric acid (HNO3), to improve the recovery of Ag, Al, Fe, and Sb (USEPA 1997). Performed in a closed microwave system, this method provides higher temperature and pressure (Berghoff-Tetra 2004), resulting in a faster, safer, and more efficient digestion, in addition to being less susceptible to the loss of volatile elements. However, it should be noted that the methods 3050B and 3051A are not total digestion techniques, because they do not recover 100 % of the element in the soil sample (Sawhney and Stilwell 1994). The USEPA method 3052, also carried out in a microwave, is recommended for a total digestion, by promoting the sample’s total decomposition due to the hydrofluoric acid (HF) presence (USEPA 1995b). Appreciable differences in the recovery of metals are observed between these methods (Scancar et al. 2000; Chen and Ma 2001; Campos et al. 2003; Tighe et al. 2004; Chander et al. 2008), many times with poor
correlations between them, indicating a possible dependence of the metal recovery with the soil mineralogical composition and the nature of the metal. Therefore, it is important to evaluate these digestion methods using soils with different characteristics. In this sense, the objective of this study was to compare the effectiveness of three digestion methods (USEPA 3050B, 3051A, and 3052) for determination of Cu, Zn, Cd, Pb, Ni, and Hg in 10 soil samples with different chemical, physical, and mineralogical characteristics, as well as to provide subsidies to the Brazilian legislation on the subject (CONAMA Resolution no. 420 of 28 Dec. 2009), with respect to analytical methodologies.
Materials and methods Samples were collected from the surface horizons of 10 soil classes in several municipalities in the state of Pernambuco: Mollisols (Nazaré da Mata), Oxisols (Rio Formoso), Ultisols (Camutanga), Ultisols (Aliança), Histosols (Ipojuca), Gleysols (Ipojuca), Oxisols (Caruaru), Ultisols (Garanhuns), Entisols Fluvents (Ibimirim), and Vertisols (Bodocó). The physical and chemical characteristics of these soils are shown in Table 1. The air-dried samples were sifted on a 2-mm mesh nylon sieve. The aliquot was macerated in an agate mortar and sifted with a stainless steel 0.3-mm mesh sieve (ABNT no. 50), in order to avoid contamination. Three different methods of sample digestion were assessed, which are described below. All digestions were performed in duplicate. USEPA 3050B (USEPA 1996) Pulverized soil samples (0.5 g) were transferred to Teflon beakers, where 10 mL of 50 % HNO3 was added. The solutions were heated on a hot plate at 95 °C±5 with a ribbed watch glass, allowing them to evaporate (without boiling) to about 5 mL, for 2 h. Subsequently, 2 mL of ultrapure water and 3 mL of hydrogen peroxide (30 % H2O2) were added to the beakers. The solutions were again heated until the effervescence reduced; aliquots of 1 mL of 30 % H2O2 were added until the effervescence was minimal or the sample’s appearance suffered no further changes. After, the heating procedure was repeated, thus evaporating (without boiling) the solutions to about 5 mL, for 2 h. Finally, 10 mL of concentrated HCl was added to the solutions, followed by hot plate heating (95 °C±5) for 15 min.
Environ Monit Assess Table 1 Physical and chemical characteristics of the soil samples P
Depth
Sanda
Silta
Claya
g kg−1
cm
pH (H2O)
Al3+b
Ca2+ + Mg2+b
K+c
Na+c
cmolc dm−3
(1:2.5)
Pc
Fed
mg dm−3
mg kg−1
Mnd
OCa g kg−1
1
0–30
498
214
288
5.2
0.2
5.3
0.27
0.16
31
16.92
140.02
24.6
2
0–10
498
74
428
5.1
1.6
0.4
0.1
0.06
3
45.87
51.10
22.4
3
0–17
381
338
282
4.8
0.2
2.7
0.58
0.07
3
20.70
186.10
16
4
0–28
792
73
135
5.7
0.2
1.5
0.08
0.05
38
3.64
73.65
7.2
5
0–20
222
373
405
4.2
6.5
10.6
0.14
0.14
6
11.52
60.20
380
6
0–20
61
226
713
5.8
0.2
17
0.17
0.04
85
21.96
242.5
160
7
0–12
520
90
390
4.2
2.2
1.7
0.24
0.07
5
24.32
46.18
40.5
8
0–15
660
60
280
4.7
1.1
0.6
0.09
0.01
3
16.71
30.15
35
9
0–25
491
262
246
7.9
0
18.2
1.26
0.97
102
14.96
283.93
8.5
10
0–25
418
189
393
5.7
0
19.8
0.35
0.02
3
36.48
327.35
11.6
P. Profile, 1 Mollisols, 2 Oxisols, 3 Ultisols, 4 Ultisols, 5 Histosols, 6 Gleysols, 7 Oxisols, 8 Ultisols, 9 Entisols Fluvents, 10 Vertisols a
Embrapa (1997)
b
KCl 1 mol L−1 (De Filippo and Ribeiro 1997)
c
Mehlich-1(De Filippo and Ribeiro 1997)
d
USEPA 3051A (EPA 1998)
USEPA 3051A (USEPA 1998) Pulverized soil samples (0.5 g) were transferred to Teflon tubes, where 9 mL of HNO3 and 3 mL of HCl were added. They were kept in a closed system, a microwave oven (MarsXpress) for 8 min 40 s on the temperature ramp, the necessary time to reach 175 °C; then this temperature was maintained for an additional 4 min 30 s. USEPA 3052 (USEPA 1996) Pulverized soil samples (0.5 g) were placed in Teflon tubes, where 9 mL of Fig. 1 Mean concentration of heavy metals extracted by the USEPA methods 3050B, 3051A, and 3052 in 10 soil samples. Means with the same letter are equally significant by the Tukey test (5 % probability)
HNO3 and 3 mL of concentrated HF, of high analytical purity, were added. After, samples were submitted to microwave irradiation for 5.5 min to reach 180 °C, attaining a maximum pressure of 16 atm, and 4.5 min digestion with constant temperature and pressure. After digestion, all extracts were transferred to 50-mL certified flasks (NBR ISO/IEC), filling with ultrapure water (Millipore Direct-Q System) and filtering in a slow filter paper (Macherey Nagel®). Highpurity acids were used in the analyses (Merck PA).
Environ Monit Assess
Glassware was cleaned and decontaminated in a 5 % nitric acid solution for 24 h and then rinsed with distilled water. Calibration curves for metal determination were prepared from standard 1,000 mg L−1 (Titrisol®, Merck). Sample analyses were performed only when the r2 of the calibration curve was higher than 0.999. After initial calibration, it was checked again after 10 samples were analyzed; in the case of more than 10 % of deviation, the equipment was recalibrated.
Table 2 Hg (in microgram per kilogram), Zn, Ni, Pb, Cu, and Cd (in milligram per kilogram) concentrations in samples of 10 soils digested by the methods 3050B, 3051A, and 3052
Profile
3050B
The concentrations of Zn, Ni, Cu, Pb, and Cd in the extracts of the three methods were determined by an atomic absorption spectrophotometer (PerkinElmer AAnalyst™ 800) using the flame technique. Hg was determined in the same equipment coupled to a hydride generator (FIAS 100/Flow Injection System/PerkinElmer) with an electrodeless discharge lamp. The results were submitted to descriptive statistics (mean and standard deviation), Pearson’s linear correlation between the metals in each method, and analysis of
3051A
3052
Hg
3050B
3051A
3052
Pb
1
34±0.0
36±1.4
40±8.5
11.3±0.6
12.2±1
15.8±0.7
2
135±0.0
135.2±7.4
151.2±3.2
9.4±0.6
26.7±1.7
28.2±1.1
3
19.7±2.5
32.75±1.1
38.2±0.3
43.6±1.3
61.3±3.4
61.5±0.3
4
39±2.8
45.7±3.2
47.2±8.1
6.8±0.3
7.4±0.2
11.9±0.2
5
120.7±0.3
132.5±2.1
139±3.5
26.1±0.01
39.3±0.2
33.5±2.4
6
115.7±2.5
146.7±3.2
143.2±0.3
37.4±0.1
58.1±1
54.6±1.3
7
200.7±1.1
196.5±7.1
237±1.4
15.6±0.8
46.8±0.4
37.5±1.5
8
33.7±1.1
32.2±0.2
41.5±5.7
11.8±1.4
23.9±1.2
26.4±0.7
9
10±0.7
11.7±3.9
10.5±0
12.6±0.4
13.4±0
30.4±0.3
33±0
21.8±0.2
31.3±0.8
28.1±1.2
10
23.5±0.7
30.7±0
Zn
Cu
1
27.4±0.2
45.4±3.2
57.9±5.8
3.7±0.1
4.9±0.6
5.7±0.6
2
10.6±3.3
29±1.4
51.6±1.6
2.0±0.3
6.3±0.3
11.7±0.1
3
54.5±5.3
82.1±9.0
16.2±0.2
25.4±11.1
21.6±0.2
4
15.9±0.9
49±12.4
35.3±6.0
47.2±0.5
7.1±0.3
7.9±0.0
10.8±1.1
5
59.4±4.4
102.7±3.2
28±0.1
34.1±0.5
36.7±0.1
6
59.1±9.0
110.5±16.3
129.7±17.3
23±0
32.4±1.1
35.9±1.9
7
13.3±5.0
30.4±3.2
50.6±1.9
3.6±1.4
5.8±0.9
4.9±0.5
8
13.6±2.7
34.9±3.9
55.7±13.8
4.9±3.3
5.7±0.2
9
31.7±3.6
50.9±2.3
92.7±5.3
12.2±0.0
16.5±0.4
92.5±12.0
90.5±2.1
19.3±0.2
33.0±0.3
36.9±0.2
10
21±1.0
110±21.2
Ni
7.3±0.2 17.15±0
Cd
1
7.8±0.2
4.7±0.1
15.4±0.5
1.5±0.1
2.3±0.1
5.6±0.3
2
7.4±0.2
7.5±0.5
24.5±0.3
1.3±0.1
3.4±0.1
5.2±0.1
3
11.8±0.6
12.8±0.8
28.6±0.7
1.3±0.1
2.8±0.2
5.5±0.0
Results are expressed as mean ± standard deviation. Detection limits (in microgram per kilogram): Hg (0.009), Pb (15), Zn (1.5), Cu (1.5), Ni (6), Cd (0.8)
4
5.0±0.0
5.8±0.4
15.9±0.2
1.1±0.2
1.9±0.1
5.3±0.1
5
17.5±0.1
22.3±1.1
27.6±1.3
2.8±0.5
3.4±0.1
5.1±0.2
6
12.9±0.4
19.2±0.6
30.2±2.5
2.8±0.0
3.2±0.1
5.8±0.4
7
4.9±0.1
7.6±0.2
13.9±0.1
2.4±0.1
3.1±0.0
6.1±0.1
1 Mollisols, 2 Oxisols, 3 Ultisols, 4 Ultisols, 5 Histosols, 6 Gleysols, 7 Oxisols, 8 Ultisols, 9 Entisols Fluvents, 10 Vertisols
8
5.1±0.6
6.1±0.2
16.6±0.7
1.7±0.0
2.62±0.1
6.1±0.0
9
17.4±0.1
15.9±0.6
30.7±1.1
1.8±0.1
2.5±0.0
6.3±0.3
10
20.2±0.7
28.9±0.2
42.3±0.6
1.9±0.0
3.5±0.2
6.2±0.2
Environ Monit Assess
variance using the F-test. The Tukey test at 5 % (P
