Certification of trace metal extractable contents in a sediment reference material (CRM 601) following a three-step sequential extraction procedure

The Scienceof the Total Environment 205(1997) 223-234

Certification of trace metal extractable contents in a sediment reference material (CRM 601) following a three-step sequential extraction procedure Ph. Quevauviller a,*, G. Rauretb, J.-F. tipez-SBnchezb, R. Rubiob, A. Ure’, H. Muntaud aEuropean Commission, Standa&, Measurements and Testing Programme, Rue de la Loi 200, B-1049 Brussels, Belgium bVniversidad de Barcelona, Depatiamento de Quimica Anal&a, Au. Diagonal 647, E-08028 Barcelona, Spain cVniversity of Strathclyde, Department of Pure and Applied Chemistry, Cathedral Street, Glasgow, Gl lm, UK dEuropean Commission, Joint Research Centre, Environment Institute, I-21020 Ispra, Italy

Received 16 May 1997;accepted6 July 1997

Abstract Sequential extraction schemes have been developed in the past 20 years for the determination of binding forms of trace metals in sediment. The lack of uniformity of these schemes, however, did not allow the results so far to be compared worldwide nor the procedures to be validated. Indeed, the results obtained by sequential extraction are operationally defined, i.e. the ‘forms’ of metals are detined by the determination of extractable elements using a given procedure. Therefore the significance of the analytical results is related to the extraction scheme used. Another problem which hampered a good comparability of data was the lack of suitable reference materials which did not enable the quality of measurements to be controlled. Owing to this lack of comparability and quality control, the Community Bureau of Reference (BCR, now Standards, Measurements and Testing Programme) has launched a programme of which one of the aims was to harmonize sequential extraction schemes for the determination of extractable trace metals in sediment. This programme involved the comparison of existing procedures tested in two interlaboratory exercises, and it developed into a certification campaign of extractable trace element contents in a sediment reference material, following a three-step sequential extraction procedure duly tested and adopted by a group of 18 EU laboratories. This paper briefly describes the results of the interlaboratory studies and gives all details on the preparation of the sediment reference material, CRM 601, the homogeneity and stability studies and the analytical work performed for the certification of the extractable contents of some trace elements, following a standardized sequential (three-step) extraction procedure. 0 1997 Elsevier Science B.V. Keywords:

Standardization; Trace metals; Sequential extraction scheme; Sediment; Certified reference material;

BCR

* Corresponding author. 0048-9697/97/%17.00 PII SOO48-9697(97)00205-2

0 1997Elsevier ScienceB.V. All rights reserved.

224

Ph. Quevauviller

et al. / The Science of the Total Environment

1. Introduction The mobility of trace metals, as well as their bioavailability and related eco-toxicity to plants, depends strongly on their specific chemical forms or ways of binding. Consequently, these have to be determined rather than the total element contents in order to asses the toxic effects and study geochemical pathways. However, the determination of specific chemical species or binding forms is difficult and often hardly possible. Therefore determinations of broader forms, e.g. so-called ‘mobile’ or ‘carbonate-bound’ forms, depending on operationally-defined procedures can be a good compromise to give information on environmental contamination risk (e.g. determination of environmentally accessible trace metals upon disposal of sediment on to a soil and evaluation of risk of contamination of ground waters). As a result of this practicality, single and sequential extraction schemes have been designed for the determination of binding forms of trace metals in soil and sediment (Tessier et al., 1979; Salomons and Forstner, 1980) and increasingly used over the last 10 years. This concept has been described recently in detail (Forstner, 1993). The lack of uniformity of these schemes, however, did not allow the results to be compared worldwide or the procedures to be validated which led to critical comments. Indeed, the results obtained, e.g. by sequential extraction are operationally defined, i.e. the ‘forms’ of metals are defined by the determination of extractable elements using a given procedure. Therefore the significance of the anaytical results is related to the extraction scheme Ised. This type of determination has often been ,eferred to as ‘speciation’ although this term trictly speaking cannot be applied to operatioIally defined procedures (the term speciation ather covers the determination of specific forms, .g. oxidation state or organometallic compounds 2uevauviller et al., 1992)). Another problem hich hampered a good comparability of data was te lack of suitable reference materials which did It enable the quality of measurements to be mtrolled. This aspect of quality control related the determination of extractable trace metals LSbeen developed elsewhere (Griepink, 1993).

205 (1997) 223-234

Owing to the likelihood of the many sources of pitfalls which may occur, it was expected that the use of extraction schemes would be progressively abandoned. Consequently, the Community Bureau of Reference (BCR, now Standards, Measurements and Testing Programme) has launched a programme aimed at harmonizing single and sequential extraction schemes for the determination of extractable trace metals in soil and sediment, respectively. This programme started in 1987 with the comparison of existing procedures tested in two interlaboratory exercises (Ure et al., 1992, 1993). The development of this programme was extensively discussed in a workshop organised by the BCR in 1992 (Quevauviller et al., 1993a) which was recognised to be a ‘turning-point’ in the use of these schemes in Europe (Salomons, 19931. This paper presents the results of the certification campaign on extractable trace metals in a sediment candidate reference material, CRM 601.

2. Background

of the project

An initial study of the literature on the ‘speciation’ of metals in soils and sediments by chemical extraction, and a consultation with European experts was carried out by A. Ure on behalf of the BCR (Ure et al., 1992). The outcome of this study was discussed in a meeting of approx. 40 representatives from leading European laboratories in the field of soil and sediment analysis and a programme was adopted involving: (i) the design of single and sequential extraction procedures for the analysis of soil and sediment, respectively; (ii) interlaboratory trials on these extraction schemes; and (iii) the preparation of reference material certified for their extractable trace metals. Therefore a three step extraction procedure was designed based on acetic acid extraction (step l>, hydroxylammonium chloride extraction (step 2) and hydrogen peroxide/ammonium acetate extraction (step 3). This scheme (given in the appendix) was tested in a first interlaboratory trial on Cd, Cr, Cu, Ni, Pb and Zn (Quevauviller et al.,

Ph. Queuauuiiier et al. / The Science ofrhe Total Environment 205 (1997) 223-234

1993b) and the results showed that, while promising, improvements were necessary prior to attempt the certification of a reference material.

Table 1 Results of the first and second interlaboratory 1, 2 and 3) Steps

2.1. Sediment samplesfor the interlaboratory triaks

The sediment sample used in the first interlaboratory trial was collected in Yrseke, The Netherlands, with grab samplers, wet sieved at 2 mm and air dried at room temperature (approx. 20°C). The air-dried material was then ground, sieved at 90 pm and homogenised before bottling. The bulk homogeneity of the sediment was tested by X-ray fluorescence (XRF) by determining major components. The sediment sample used in the second interlaboratory trial was collected in the River Be&s, Spain. The material was sampled with a grab, air-dried, then sieved at 63 pm, homogenised and bottled. Homogeneity and stability studies of extractable trace metals were carried out and the material was found to be homogeneous and stable enough to be used in the intercomparison exercise (Fiedler et al., 1994; Quevauviller et al., 19941. 2.2. Analytical techniquesusedin the intercomparisons

The techniques used to determine metal concentration in extracts were generally FAAS (flame atomic absorption spectrometry) or ETAAS lelectrothermal atomic absorption spectrometry with or without Zeeman background correction). ICPAES (inductively coupled plasma atomic emission spectrometry) and ICP-MS (inductively coupled plasma mass spectrometry) were also used by some laboratories. 2.3. Technical discussionof the results

The results of the two intercomparisons are summarized in Table 1. The effects of shaking type and speed or room temperature on the spread of results were thoroughly discussed. Changes due to temperature effects were not noticeable. However, it was

225

First round-robin Mean

SD.

exercises (steps

Second round-robin CV

Mean

SD.

CV

Cadmium 1 2 3

7.18 3.41 1.03

0.81 0.63 0.20

11.3 18.5 20.1

0.18 0.08 1.02

0.02 0.02 0.17

8.5 16.9 16.6

Chromium 1 2 3

1.36 3.29 76.3

0.20 1.07 10.4

14.7 32.5 13.6

1.28 1.21 866

0.41 0.31 126

32.0 25.6 14.5

Copper 1 2 3

3.69 3.13 63.4

0.76 1.96 13.2

20.6 20.1 20.8

0.23 0.53 90.1

0.08 0.33 7.6

34.8 62.3 8.4

Nickel 1 2 3

9.76 5.79 10.2

4.36 1.54 3.32

44.7 26.6 32.5

13.0 1.80 15.2

2.08 0.23 2.3

16.0 12.7 15.2

Lead 1 2 3

5.06 11.0 6.93

2.50 8.87 4.78

49.4 80.6 69.0

0.30 0.14 47.7

0.17 0.09 7.6

56.7 64.3 15.9

35.1 34.2 9.14

13.4 24.4 10.2

Zinc 1 2 3

262 140 89.7

93.9 79.7 676

16.9 16.9 44

18.0 21.2 6.5

The table lists the mean of the laboratory means along with the standard deviation (SD.) and the coefficient of variation (CV) obtained. The trace metal contents are given in mg kg-l.

found more difficult to standardize the shaking procedures. Six laboratories used end-over-end shaker whereas the other laboratories used a horizontal shaker. No effect of shaker type was suspected in this case; however, effects of shaking speed were suspected as lower results were generally obtained at a speed of less than 40 rpm, whereas higher results corresponded to speeds of up to 150 rpm. The need of maintaining the sediment in suspension during shaking was found to be highly critical. The proposal to use a glass ball during shaking was not accepted as it would lead to possible grinding effects. From these results the University of Barcelona carried out a

226

Ph. Quevauviller

et al. / The Science of the Total Environment

study of the influence of shaking type and speed which demonstrated that the copper extracted in this step using an end-over-end shaker, operated at 30 rpm was 20% higher than when using an horizontal shaker operating a 130 rpm. The participants stressed once more that it is of paramount importance to verify that the sediment is continually in suspension during the extraction. It was agreed to add the following recommendation in the protocol: ‘The sediment should be continually in suspension during the extraction. If this is not verified, the shaking speed should be adapted in order to ensure a continuous suspension of the mixture’. Step 3 was considered to be critical in the presence of high amounts of organic matter (which is the case of the sample used in the second exercise) as the incomplete destruction of organic matter as well the difficulty in oxidizing sulphide, may be the source of a high uncertainty which could explain the spread of results. In some cases, it was difficult to obtain a ‘cake’ after centrifugation and, consequently, fine particles were still present in solution which created problems in ICP analysis (nebulizer clogging). In these circumstances, a filtration step was necessary. Filtration was not recommended to be included in the protocol; however, when ICP is used, the participants strongly recommended this step for ICP users. The protocol was hence modified by adding the following sentence: ‘After centrifugation, a filtration step (0.45 pm> is recommended for ICP determinations’. Later studies performed by the University of Barcelona showed the occurrence of interference due to high contents of Al and the presence of high background signal due to Ca. The effects of interferences can be taken into account at the calibration step by adding both elements to the calibrant. Violent reactions could be observed with hydrogen peroxide with this sediment. The participants recommended special precautions to be taken in the handling of this reagent (i.e. slow addition). Measurements were suspected to be affected by iron interferences and calibration by standard addition was strongly recommended for ETAAS. This recommendation was actually included in

205 (1997) 223-234

the protocol as follows: ‘When ETAAS is the final method of element determination, the method of standard additions is strongly recommended for calibration’. The second interlaboratory exercise on extractable trace metals in sediment showed a consequent improvement in comparison with the results of the first exercise. Furthermore, these collaborative efforts allowed the sequential extraction procedure to be slightly improved by minor amendments (Quevauviller et al., 1994). 3. Preparation

of the candidate reference material

The candidate reference material has been collected in March 1994 from different sampling sites of Lake Maggiore (Italy). Sampling operations were performed using grab collectors. The wet sediment was passed through a Z-mm sieve in order to remove stones and other materials extraneous to sediments. The sieved sediment was placed on stainless-steel free trays and exposed at ambient air temperature for drying, turning the lumps from time to time to accelerate the drying process. The dry sediment with a mean moisture content of 3.5% (calculated by drying a separate portion of sediment at 105°C for 3-4 h) was passed through a tungsten carbide-bladed hammermill and sieved to pass apertures of 90 pm. The < 90pm fraction was collected in a PVC mixing drum (filled with dry argon) and homogenized for 14 days at approx. 48 rpm. Ten subsamples were randomly taken in the drum and analysed by X-ray fluorescence spectrometry in order to assess the bulk homogeneity; the CVs obtained were compared with the CV calculated from 10 replicate measurements performed in one bottle (Table 21. The results obtained indicated that the bulk homogeneity was satisfactory (CV generally lower than 1%) and the bottling operation was therefore carried out. The bottling procedure was performed manually: after an additional period of mixing of 2 days, a first series of 20 bottles was filled and immediately closed with screw caps and plastic inserts. Series of 20 bottles were hence filled, alternating with re-mixing of the powder (2

Ph. Quevauviller et al. / The Science of the Total Environment 205 (1997) 223-234

227

Table 2 Bulk homogeneity testing by XRF

cvs CV drum CV bottle

Si

Al

Ca

K

Mg

Ti

S

P

Pb

Fe

Zn

Cu

Ni

Cr

Mn

0.81 0.33

0.43 0.32

0.67 0.23

0.39 0.14

0.30 0.26

0.49 0.26

1.31 0.51

0.99 0.28

0.65 0.64

0.40 0.13

0.81 0.46

0.71 0.42

1.11 1.26

1.36 0.22

0.61 0.15

CV drum: 10 determinations on samples taken randomly in the drum. CV bottle: 10 determinations on the content of one bottle.

mm). The bottles were stored at ambient temperature. 4. Homogeneity

study

The extractants were prepared as laid out in the appendix. All precautions were taken to avoid contamination during the extraction procedures. The trace element contents (Cd, Cr, Cu, Ni, Pb and Zn) in the extracts were determined by flame atomic absorption spectrometry (FAA% or electrothermal atomic absorption spectrometry with Zeeman background correction (ZETAAS). For the homogeneity study, the six elements were determined in the candidate CRM by analyzing 10 subsamples taken from one bottle (within-bottle homogeneity test) and one subsample in each of 20 different bottles selected during the bottling procedure (between-bottle homogeneity test). The CVs and the total uncertainty Uc, for the extractable trace element contents between (CV,) and within (CV,) bottles, are given in Table 3.

For most of the extractable metal contents, the overlap was within the total uncertainty Ur of the CV (an approximation of the uncertainty U,, of the CV is calculated as follows: U,, = CV/,/2n>. Differences between the within-bottle and between-bottle CVs observed for the step 2 was considered to be rather an analytical artefact than an indication of inhomogeneity which would have been reflected in the spread of results submitted in the certification. The material is then considered to be homogeneous for the stated level of intake (1 g). 5. Stability

study

The stability of the extractable trace element contents was tested to determine the suitability of the sediment as reference material. Sets of bottles were kept at - 20, + 20 and + 40°C during a period of 12 months and the extractable contents of Cd, Cr, Cu, Ni, Pb and Zn were determined (in five replicates) after 1, 3, 6 and 12 months. The detection techniques used were the same as in the homogeneity study.

Table 3 Homogeneity tests Cd

Cr

CU

Ni

Pb

Zn

1st step cvB

cvw 2nd step CVB cvw 3rd step cvB

cvw

1.8 & 0.3 1.5 + 0.3

5.2 i 0.8 5.2 f 1.2

1.1 f 0.2 1.2 + 0.3

6.7 + 1.1 6.1 + 1.4

3.6 i 0.6 3.9 * 0.9

0.6 + 0.1 0.2 * 0.1

3.3 + 0.5 5.2 + 1.2

21.2 5 3.4 15.4 + 3.4

26.1 f 4.1 7.9 + 1.2

12.2 f 1.9 6.1 i 1.4

16.7 f 2.7 7.4 + 1.7

4.2 f 0.7 2.3 + 0.5

6.0 f 0.9 5.3 k 1.2

4.0 + 0.7 4.9 + 1.1

5.1 + 0.8 6.4 + 1.4

7.9 +_ 1.3 7.6 f 1.7

2.7 f 0.4 4.4 * 1.0

4.8 + 0.8 6.1 f 1.4

228

Ph. Quevauviller

et al. / The Science of the Total Environment

Any change in the content of an analyte with time indicates an instability provided that a good long-term analytical reproducibility is obtained. Instability would be detected by comparing the contents of different analytes in samples stored at different temperatures with those stored at a low temperature at the various occasions of analysis. The samples stored at -20°C were used as reference for the samples stored at + 20 and + 40°C respectively. Table 4a-c gives the ratios 03,) of the mean values Czr> of five measurements made at + 20 and + 40°C respectively, and the mean value (x-,,,> from five determinations made at the same occasion of analysis on samples stored at a temperature of - 20°C: R, = j3& 2oec. The uncertainty U, has been obtained from the coefficient of variation (CV) of five measurements obtained at each temperature: UT = CV, + CV! 20.c)1’2. R, In the case of ideal stability, the ratios R, should be 1. In practice, however, there are some random variations due to the error on the measurement. As indicated in the tables, in most of the cases, the value 1 is comprised between R, Ur and R, + U,. The uncertainty on the method can account in most cases of the deviations observed; some risks of instability were, however, suspected at 40°C due to possible changes in the extractability of some elements (e.g. Cu and Pb); these changes induced by the high storage temperature could be related to changes in the status of the organic matter or in the crystallographic compounds of Fe or Mn. Hence, it is recommended to avoid storage at temperature above 20°C. 6. Analytical

methods

Each laboratory that took part in the certification exercise was requested to make a minimum of five independent replicate determinations of each element on at least two different bottles of the two CRMs on different days, following strictly the sequential extraction protocol described in the appendix. In the meeting of the laboratories participating in the certification the sources of error and the measures taken to eliminate them

205 (1997) 223-234

Table 4 Stability tests Element

Time (month)

R, + CJ, (20°C)

R, f u, (4W.I)

1 3 6 12

0.97 1.02 1.00 0.93

f f k f

0.97 1.03 0.98 0.89

Cr

1 3 6 12

1.02 1.41 1.03 1.06

* 0.09 _+ 0.18 + 0.06 f 0.09

1.07*0.11 1.41 * 0.20 1.12 + 0.05 1.01 f 0.12

cu

1 3 6 12

1.06 1.10 1.12 1.20

k * f &

0.01 0.03 0.02 0.09

1.10 1.25 1.39 1.39

* f * +

0.01 0.03 0.03 0.02

Ni

1 3 6 12

1.00 1.01 1.15 0.93

f 0.11 & 0.18 f 0.21 &- 0.11

0.94 0.97 0.90 0.92

+ + * +

0.07 0.15 0.13 0.10

Pb

1 3 6 12

1.03 1.16 0.97 1.01

+ 0.03 + 0.08 * 0.04 _+ 0.03

0.98 1.15 1.24 1.08

f 0.06 f 0.07 -+ 0.04 _+ 0.07

Zn

1 3 6 12

0.97 1.05 1.02 1.01

+ * * *

0.01 0.01 0.01 0.01

0.98 1.06 1.03 0.93

+ 0.01 k 0.01 _+ 0.05 + 0.02

1 3 6 12

1.00 1.00 0.99 0.94

+ f + *

0.02 0.03 0.02 0.03

0.98 0.95 0.97 1.07

f 0.02 * 0.03 _+ 0.02 + 0.03

Cr

1 3 6 12

0.89 0.80 0.92 0.97

+ + + +

0.07 0.12 0.05 0.08

0.88 0.76 0.70 0.92

+_ 0.11 f 0.12 + 0.09 + 0.22

cu

1 3 6 12

0.88 0.92 0.98 1.03

+_ 0.08 + 0.11 + 0.08 * 0.14

0.92 0.87 0.84 0.98

f 0.16 + 0.20 _+ 0.07 + 0.42

Ni

1 3 6 12

0.93 0.98 1.04 0.93

+ k f +

0.09 0.10 0.08 0.15

0.87 0.87 0.73 1.24

+ + f +

Pb

1 3 6 12

1.02 0.74 0.87 0.98

* f + f

0.03 0.09 0.05 0.13

0.97 0.67 0.59 0.86

k 0.06 f 0.09 _+ 0.04 + 0.28

(a> First step Cd

(b) Second Cd

0.01 0.01 0.01 0.03

f + f +

0.02 0.02 0.03 0.01

step

0.04 0.10 0.05 0.14

Ph. Quevauviiler Table

et al. / The Science of the Total Environment

4 (Continued)

Element

Zll

(cl Third Cd

Time (month)

R, + U, (20°C)

R, f U, (40°C)

1 3 6 12

0.98 0.95 0.97 1.00

0.97 0.91 0.79 1.20

+ + + +

0.01 0.02 0.03 0.04

+ * + f

0.03 0.02 0.02 0.05

step 1 3 6 12

0.88 1.06 1.02 1.01

+ + f +

0.07 0.06 0.05 0.07

0.91 1.03 0.92 1.00

f + + f

0.02 0.07 0.04 0.07

Cr

1 3 6 12

0.97 0.95 1.02 1.00

+ & + +

0.04 0.04 0.03 0.04

1.02 0.93 0.96 1.18

+ + + +

0.04 0.05 0.03 0.04

CU

1 3 6 12

1.05 0.93 1.02 0.98

& + f f

0.05 0.04 0.04 0.03

1.02 0.91 0.95 1.03

+ + + +

0.04 0.06 0.03 0.03

Ni

1 3 6 12

1.03 1.04 1.09 0.99

f + + +

0.08 0.07 0.06 0.08

1.15 1.03 1 .oo 0.81

+ * f +

0.08 0.11 0.02 0.06

Pb

1 3 6 12

0.95 1.04 1.01 1.03

f + + *

0.06 0.05 0.04 0.09

1.12 1.02 1.20 0.96

+ * f +

0.07 0.05 0.06 0.07

Zn

1 3 6 12

1.01 1.06 1.05 1.00

+ f + +

0.06 0.08 0.06 0.06

1.09 1.03 0.94 0.75

+ + + +

0.06 0.09 0.03 0.05

were discussed; the laboratories participating in the certification exercise applied their methods correctly, i.e. the determinations were performed only when the method was under control i.e. the standard deviations observed in the laboratory were in accordance with the normal practice of the laboratories.

205 (1997) 223-234

229

detected were rather due to e.g. calibration errors than to the application of the extraction procedure. 8. General statistical

discussion

The sets of results found acceptable on technical and statistical grounds were represented in the form of ‘bar-charts’ of which examples are given in Figs. l-3. In the ‘bar-charts’, the length of a bar corresponds to the 95% confidence interval of the mean of laboratory means. The certified values were calculated as the arithmetic means of laboratory means (taking into account the number of sets accepted for certification after both statistical and technical scrutiny). This value is featured as a vertical dotted line on the bargraphs; its uncertainty is given by the half width of the 95% confidence interval of the mean of laboratory means. The sets of results have been submitted to the following statistical tests: Kolmogorov-Smirnov-Lilliefors tests to assess the conformity of the distributions of individual results and of laboratory means to normal distributions; Nalimov test to detect ‘outlying’ values in the population of individual results and in the population of laboratory means; Bartlett test to assess the overall consistency of the variance values obtained in the participating laboratories; Cochran test to detect ‘outlying’ values in the laboratory variances ($1; one-way analysis of variance W-test) to compare and estimate the between- and the within-laboratory components of the overall variance of all individual results. A summary of the statistical evaluation is given in the certification report (Quevauviller et al., 1997). 9. Certified values

7. Technical

discussion

At the technical meeting, it was recalled that strict observance of the extraction protocol would be a criterion for considering the results for discussion. The participants recommended that a tolerance of k 30% be included in the extraction protocol for the shaker speed. Most of the errors

The certified values (unweighted mean of p accepted sets of results) and their uncertainties (half width of the 95% confidence intervals) are given in Tables 5a-d as mass fractions of the respective extracts obtained at the first, second and third steps (based on dry mass) in mg/kg.

230

Ph. Quevauviller BAR-GRAPHS

FOR LABORATORY 2.8

07

et al. / The Science of the Total Environment

205 (1997) 223-234

MEANS AND 95% CI

3.2

3.6

4.0

4.4

+.....+......+.....+.....+......+.....+......+.....+.....+......+.....+ I FAAS I

10 FAAS

<-------------*-------------,

4.8

<-*->

I I <-*-->

11 FAAS 05 ETAAS

I i------------------*.-----------------,

12 ETAAS

<---*--->

I

13 ETAAS

I

< - - _ _ - - _ * -----__

>

I

02 ICPAES 08

I

<-*->

15 ETAAS

<-----*----->

I

ICPAES

!

09 ICPAES

<-*>

<---*

--->

I I <----*---->

16 ICPMS

I <-------M------>

1MEANS 1

Fig. 1. Cadmium - first step in mg/kg.

10. Availability The certified reference material can be purchased from the Institute for Reference Materials and Measurements (IRMM), Management of Reference Materials Unit, Retiesweg, 2440 Gee1 (Belgium). Each bottle is accompanied by a certificate and a report (Quevauviller et al., 1997) describing the work performed (preparation of the material, homogeneity and stability studies, analytical methods used and individual results). Acknowledgements The CRM was collected and prepared by the

Joint Research Centre, Environment Institute in Ispra (Italy); the homogeneity and stability studies were carried out by the Universidad de Barcelona, Departamento de Quimica Analitica (Spain). The following laboratories participated in the certification campaign: Agriculture and Food Development Authority (Wexford, Ireland); Agricultural Research Centre, Inst. for Crops and Soil (Jokioinen, Finland); Bund. fiir Materialforschung und Priifung (Berlin, Germany); Chalmers University of Technology (Giiteborg, Sweden); Esta@o Agronomica National (Oeiras, Portugal); Estacion Experimental de1 Zaidin, CSIC (Granada, Spain); Institut National d’Agronomie, Lab. de Chimie Analytique (Paris,

Ph. Quevauviller BAR-GRAPHS

FOR

2.4

et al. /The

LABORATORY

MEANS

2.6

Science of the Total Environment AND

95%

2.8

231

205 (195’7) 223-234

CI

3.0

3.2

+ . . . ..+.....f....+.....+.....+......+~....+.....

3.4

3.6

+....+.....+.....+.....+

1 <-----------.-*------------,

07 FAAS

I 10

FAAS

05

ETAAS

12

ETAAS

13

ETAAS

15

ETAAS

<------

.----*--------.--,

< ___________

t ___-_-_----

>

I <.----.-.-*‘--.---->I I *-,- - - - - - - - - - - - - >

<--------------

I < _ _- - __ - * _- __ __ - ,

02 ICPAES

08 ICPAES I

09

ICPAES

16

ICPMS

<--*.->

I

I I+---*----->

I I <---------M-------->

1MEANS 1

Fig. 2. Cadmium - second step in mg/kg.

France); Institut National de Recherche Agronomique (Arras, France); Joint Research Centre, Environment Institute (Ispra, Italy); Laboratoire Central des Ponts et ChaussCes, Division Eau (Bouguenais, France); The Macaulay Land Use Research Institute (Aberdeen, United Kingdom); Universidad de Barcelona, Dept. de Quimica Analitica (Barcelona, Spain); Universidade Nova de Lisboa (Monte da Caparica, Portugal); Universiteit Gent, Lab. of Auaiytical and Agro-Chemistry (Gent, Belgium); University of Strathclyde, Dept. of Pure and Applied Chemistry (Glasgow, United Kingdom); University of Reading, Dept. of Soil Sciences (Reading, United Kingdom).

Appendix 1 Sequential extraction procedure Apparatus All laboratory-ware shall be of borosilicate glass, polyethylene, polypropylene or PTFE, except for the centrifuge tubes, which will be of borosilicate glass of PTFE. Clean vessels in contact with samples or reagents with HNO, 4 mol/l (overnight) and rinse with distilled water. Determine the blank as follows: to one vessel from each batch, taken through the cleaning procedure, add 40 ml of acetic acid (solution A, see below). Analyse this

232

Ph. Queuauuiller

et al. / The Science

of the

blank solution along with the sample solutions from step 1 described below. Use a mechanical shaker, preferably of the horizontal rotary or the end-over-end type, at a speed of 30 rpm and record the speed. Carry out the centrifugation at 1500 x g.

Water: Glass-distilled water is normally suitable; simple de-ionised water may contain organically complexed metals and should not be used. Analyse a sample of distilled water with each batch of step 1 extracts.

FOR

LABORATORY

MEANS

AND

95%

205 (1997) 223-234

Solution A (acetic acid 0.11 mol/l): Add in a fume-cupboard, 25 k 0.2 ml of redistilled glacial acetic acid (or for example Analar of Suprapur grade acetic acid without distillation) to approx. 0.5 1 of distilled water in a l-l polyethylene bottle and make up to 1 1 with distilled water. Make up 250 ml of this solution (acetic acid 0.43 mol/l> with distilled water to 1 1 to obtain an acetic acid solution of 0.11 mol/l. Analyse a sample of each batch of solution A. Solution B (hydroxylamine hydrochloride or hydroxyammonium chloride 0.1 mol/l): Dissolve 6.95 g of hydroxylamine hydrochloride in 900 ml of distilled water. Acidify with HNO, to pH 2 and

Reagents

BAR-GRAPHS

Total Environment

CI

1.0 1.4 1.8 +........+.......+........+........c..~.*...-..~-*.-...~...-....+..-.....~

2.2

I <--------------*-

03

FAAS

07

FAAS

10

FAAS

11

FAAS

05

ETAAS

06

ETAAS

12

ETAAS

13

ETAAS

15

ETAAS

02

ICPAES

09

ICPAES

------------->

I *------>

<---‘-I c-------..-*----------->

I I <*->

<---*-->

/ I

<------ * _- _- - - - >

I I <-----*----->

<----------*---------> 1 <---*-..>

I

I I

(MEANS

1

<------*----->

I I <-------H--------, Fig. 3. Cadmium - third step in mg/kg.

2.6

Ph. Quevauviller

et al. / The Science of the Total Environment

Table 5 Certified contents of extractable contents of Cd, Cr, 01, Ni, Pb and Zn Certified value (w/kd

Uncertainty (mg/kd

p

Table number in Appendix A

(a) First step Cd Cr CU Ni Pb Zn

4.14 0.36 8.32 8.01 2.68 264

0.23 0.04 0.46 0.73 0.35 51

11 12 9 10 11 12

1 2 3 4 5 6

(b) Second step Cd Ni Pb Zn

3.08 6.05 33.1 182

0.17 1.09 10.0 11

10 11 9 12

7 8 9 10

(c) Third step Cd Ni Pb

1.83 8.55 109

0.20 1.04 13

11 9 12

11 12 13

205 (1997)

223-234

233

the contents of the sample bottle, with the F’TFE ball supplied in the bottle, for 3 min. Dry a separate l-g sample of the sediment in a layer of approx. l-mm depth in an oven at 105°C for 2 h and weigh. From this a correction ‘to dry mass’ is obtained and applied to all analytical values reported (quantity per g dry sediment). Perform the extractions by shaking in a mechanical shaker at 20 + 2°C. Measure the temperature of the room at the start and at the end of the extraction procedures. The sediment should be continually in suspension during the extraction. If this is not verified, the shaking speed should be adapted in order to ensure a continuous suspension of the mixture. Perform the sequential extraction procedure according to the steps described below: Step 1

p = number of data sets.

make up to 1 1 with distilled water. Prepare this solution on the same day as the extraction is carried out. Analyse a sample of each batch of solution B. Solution C (hydrogen peroxide solution 300 mg/g, i.e. 8.8 mol/l): Use the H,O, as supplied by the manufacturer, i.e. acid-stabilized to pH 2-3. Analyse a sample of solution C. Solution D (ammonium acetate 1 mol/l): Dissolve 77.08 g of ammonium acetate in 900 ml of distilled water, adjust to pH 2 with HNO, and make up to 1 1 with distilled water. Analyse a sample of each batch of solution D.

Add 40 ml of solution A to 1 g of sediment (as received) in a loo-ml centrifuge tube and extract by shaking for 16 h at ambient temperature (overnight). No delay should occur between the addition of the extractant solution and the beginning of the shaking. Separate the extract from the solid residue by centrifugation and decantation of the supernatant liquid into a high pressure polyethylene container. Stopper the container and analyse the extract immediately or store it at 4°C prior to analysis. Wash the residue by adding 20 ml of distilled water, shaking for 15 min and centrifuging. Decant the supernatant and discard, taking care not to discard any of the solid residue. Break the ‘cake’ obtained upon centrifugation by using a vibrating rod prior to the next step. Step 2

Sequential extraction procedure Determine the extractable contents of the following trace metals, Cd, Cr, Cu, Ni, Pb and Zn using the procedure below. Carry out all extractions on the sediment as received in the glass bottle. Before subsampling the sediment, with a suitable plastic (see apparatus above) spatula, shale

Add 40 ml of solution B to the residue from step 1 in the centrifuge tube and extract by shaking for 16 h at ambient temperature (overnight). No delay should occur between the addition of the extractant solution and the beginning of the shaking. Separate the extract from the solid residue by centrifugation and decantation as in step 1. Retain the extract in a stoppered polyethy-

234

Ph. Quevauuiller

et al. /The

Science

of the

lene tube, as before, for analysis. Wash the residue by adding 20 ml of distilled water, shaking for 15 min, and centrifuging. Decant the supernatant liquid and discard, taking care to avoid discarding any of the solid residue. Retain the residue for step 3. Break the ‘cake’ obtained upon centrifugation with a vibrating rod prior to the next step. Step 3 Add carefully, in small aliquots to avoid losses due to violent reaction, 10 ml of solution C to the residue in the centrifuge tube. Cover the vessel with a watch glass and digest at room temperature for 1 h with occasional manual shaking. Continue the digestion for 1 h at 85°C and reduce the volume to a few ml by further heating of the uncovered vessel in a steam bath or equivalent. Add a further lo-ml aliquot of solution C. Heat the covered vessel again to 85°C and digest for 1 h. Remove the cover and reduce the volume of the liquid to a few ml. Add 50 ml of extracting solution D to the cool moist residue and shake for 16 h at ambient temperature (overnight). No delay should occur between the addition of the extractant solution and the beginning of the shaking. Separate the extract by centrifugation and decant into a high pressure polyethylene tube. Stopper and retain as before for analysis. Important .

The calibrant solutions should be made up with the appropriate extracting solutions. With each batch of extractions a blank sample (i.e. a vessel with no sediment) shall be carried out through the complete procedure. The sediment should be continually in suspension during the extraction. If this is not verified, the shaking speed should be adapted in order to ensure a continuous suspension of the mixture.

.

.

Note: After centrifugation, a filtration step (0.45 is recommended for ICP determinations.

pm)

Total Environment

205 (1997) 223-234

Where ETAAS is the final method of element determination, the method of standard additions is strongly recommended for calibration. References

Fiedler HD, tipez-Sknchez JF, Quevauviller Ph, Ure AM, Muntau H, Rubio R, Rauret G. The study of the stability of extractable trace metal contents in a river sediment using sequential extraction. Analyst 1994;119:1109. Fiirstner U. Metal speciation-General concepts and applications. Int J Environ Anal Chem 1993;51:5. Griepink B. Some considerations with regard to the quality of results of analysis of trace element extractable contents in soils and sediments. Int J Environ Anal Chem 1993;51:123. Quevauviller Ph, Donard OFX, Maier EA, Griepink B. Improvements of speciation analyses in environmental matrices. Mikrochim Acta 1992;109:169. Quevauviller Ph, Rauret G, Griepink B. Conclusions of the workshop: Single and sequential extraction in sediments and soils. Int J Environ Anal Chem 1993;51:231. Quevauviller Ph, Ure A, Muntau H, Griepink B. Improvement of analytical measurements within the BCR-programme: Single and sequential extraction procedures applied to soil and sediment analysis. Int J Environ Anal Chem 1993;51:129. Quevauviller Ph, Rauret G, Muntau H, Rubio R, tipezSgnchez JF, Fiedler H, Griepink B. Evaluation of a sequential extraction procedure for the determination of extractable trace metal contents in sediments. Fresenius J Anal Chem 1994;349:808. Quevauviller Ph, Rauret G, tipez-Stichez JF, Rubio R, Ure A, Muntau H. EUR Report. CEC Brussels, 17554, 1997:53. Salomons W. Adoption of common schemes for single and sequential extractions of trace metal in soils and sediments. Int J Environ Anal Chem 1993;51:3. Salomons W, Fijrstner U. Trace metal analysis on polluted sediments. II. Evaluation of environmental impact. Environ Technol Lett 1980;1:506. Tessier A, Campbell PGC, Bisson M. Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 1979;51:844. Ure A, Quevauviller Ph, Muntau H, Griepink B. EUR report. CEC Brussels, 14 763, 1992:85. Ure A, Quevauviller Ph, Muntau H, Griepink B. Speciation of heavy metals in soils and sediments. An account of the improvement and harmonization of extraction techniques undertaken under the auspices of the BCR of the Commission of the European Communities. Int J Environ Anal Chem 1993;51:135.