Fast and accurate screening of heavy metal contaminants in soils

Introduction Loose powder soil samples are notoriously hard to analyze accurately because of the sample representation and the mineralogy of these sample types. But we’ve got an instrument that can – the Epsilon 1, a benchtop energy dispersive X-ray fluorescence (ED XRF) spectrometer. Our Epsilon 1 sees and analyzes trace elements in soils and sediments that are prepared as loose samples – and does it with close to 100% accuracy. In this application note, we’ll show you just how the Epsilon 1 can quickly and precisely identify trace elements in your loose powder samples.

Instrumentation For the measurements a Malvern Panalytical single position Epsilon 1 EDXRF spectrometer was used, equipped with a 50 kV silver (Ag) anode X-ray tube, 6 software-selectable filters and a high-resolution SDD10 silicon drift detector.

Sample preparation Thirty-three soil and sediment certified reference standards were used to create the application. This set included the GSS- and GSD- series of geochemical reference materials originating from the Institute of Geophysical and Geochemical Prospecting (PRC). The standards and samples were oven dried and then pulverized in a planetary ball mill. All samples were analyzed in the form of loose powders. 4 grams of powder was put in a P1 sample cup with a 3.6 um Mylar Foil.

Measurement procedure Two different measurement conditions were used, both optimized to excite a particular group of elements (Table 1). The total measurement time was 200 seconds per sample. Silicon oxide was used as a balance compound. A typical spectrum of a soil sample is shown in Figure 1.

Elements & Compounds kV uA Filter Meas. time (s) K 2 O, CaO, TiO 2 , V, Cr, Mn, Fe 2 O 3 , Ni, Co 20 500 Al-thick 100 Cu, Zn, As, Se, Br, Rb, Sr, Zr, Y, Nb, Mo, U, Th, Pb 50 200 Ag 100 Table 1. Measurement conditions

Figure 1: XRF spectrum obtained with the first condition.

Calibration results Figures 2 and 3 show the resulting calibration graphs for As and Pb in the soil and sediment standards, respectively. The graphs show very good correlation between the certified concentrations and the measured intensities. Detailed calibration results for all analyzed compounds and trace elements are listed in Table 2. The RMS (Root Mean Square) value is equivalent to 1 sigma standard deviation.

To determine the lower limits of detection (LLD) the standard MAC000 from Malvern Panalytical’s Pro-trace solution was used. The composition of this sample contains 75% SiO 2 , 20% Al 2 O 3 and 5% Fe 2 O 3 , and is seen therefore as a real representative blank for the soil application. The sample is measured 10 times consecutively using the application conditions in Table 1. The LLDs shown in Table 2 are calculated by taking 3 times the standard deviation. For most trace elements the LLDs are around 1 ppm; only for V a slightly higher detection limit was observed, caused by the line overlap with TiO 2 .

Compound

Concentration range

RMS*

Correlation Coefficient

LLD

As (ppm)

0.7 - 412

2.3

0.9996

0.9

Br (ppm)

0.4 – 26

0.5

0.9956

0.2

CaO (wt%)

0.10 – 13.12

0.15

0.9989

NA

Co (ppm)

2.7 – 97

3.8

0.9751

0.3

Cr (ppm)

3.7 – 1090

4.1

0.9961

1.8

Cu (ppm)

2.8 – 1230

3.3

0.9999

1.2

Fe 2 O 3 (wt%)

1.46 – 18.76

0.22

0.9982

NA

K 2 O (wt%)

0.20 – 4.16

0.11

0.9911

NA

Mn (ppm)

173 – 2490

44.5

0.9956

1.2

Mo (ppm)

0.15 – 10

0.2

0.9959

0.7

Nb (ppm)

2.7 – 72

0.9

0.9985

0.3

Ni (ppm)

2.3 – 349

2.9

0.9990

0.6

Pb (ppm)

7 - 2690

1.5

0.9999

0.9

Rb (ppm)

16 – 408

3.4

0.9988

0.9

Se (ppm)

0.1 – 8.75

0.3

0.9844

0.2

Sr (ppm)

17.9 – 3430

7.2

0.9999

0.3

Th (ppm)

0.4 – 27.5

1.5

0.9721

0.3

TiO 2 (wt%)

0.13 – 2.02

0.03

0.9980

NA

U (ppm)

0.1 – 9.1

0.9

0.9110

0.9

V (ppm)

16.8 - 332

37.0

0.8654

14.1

Y (ppm)

8.1 – 52.8

1.2

0.9906

0.2

Zn (ppm)

18 - 2600

5.0

0.9996

0.6

Zr (ppm)

57 - 524

8.4

0.9963

1.2

Table 2: Calibration details (* RMS: The more accurate calibrations have the smaller RMS values).

NA: Not applicable. LLD values for major and minor compounds are irrelevant

Figure 2: Calibration graph for As in soils and sediments

Figure 3: Calibration graph for Pb in soils and sediments

Accuracy and precision The accuracy and instrument precision were tested by measuring one soil standard, CRM GSS-02, as an unknown, 10 times consecutively. The certified and average measured concentrations of 18 compounds and elements in the sample, including RMS and relative RMS values are presented in Table 3. The data demonstrates excellent accuracy and precision for the set of measured elements. The V result is affected by of the line overlap with the minor TiO 2 .

Compound

Certified conc.

Average conc.

RMS

Rel. RMS (%)

As (ppm) 13.7 13.9

0.9

6.5

CaO (wt%)

2.36

2.39

0.0

0.2

Co (ppm)

8.7

9.3

0.0

0.3

Cr (ppm)

47

49.3

0.8

1.6

Cu (ppm)

16.3

16.8

0.4

2.4

Fe 2 O 3 (wt%)

3.52

3.51

0.0

0.1

Mn (ppm)

510.0

513

2.3

0.4

Mo (ppm)

1.0

0.9

0.2

21.3

Ni (ppm)

19.4

20.2

0.6

3.0

Pb (ppm)

20.0

19.9

0.4

2.0

Rb (ppm)

88.0

89.1

0.9

1.0

Se (ppm)

0.16

0.22

0.1

27.3

Sr (ppm)

187.0

188.6

0.7

0.4

Th (ppm)

16.6

`13.4

0.2

1.5

TiO 2 (wt%)

0.271

0.264

0.001

0.3

V (ppm)

62.0

100.1

8.0

Zn (ppm)

42.0

41.8

0.5

1.2

Zr (ppm)

219

227.7

1.0

0.4

Table 3. Repeatability results

Conclusion The results demonstrate the capability of Epsilon 1 for quick and accurate screening of trace elements in soil and sediment samples. The high-resolution and outstanding sensitivity of the SDD10 silicon drift detector, combined with powerful software deconvolution algorithms, make it possible to quantify compounds and trace elements in an accurate and precise way.

Malvern Panalytical: making the impossible possible and the invisible visible.

Fast and accurate screening of heavy metal contaminants in soils

Login

Popular links

Support and services

Company profile

Legal information