The First

International Proficiency Testing Conference

Sinaia, România
11th − 13th October, 2007

PROFICIENCY TESTING: EXPERIENCES FROM CROATIA ON THE

ISSUE OF HEAVY METALS DETERMINATION IN MARINE
SEDIMENTS

Marijana Matek Saric, Judita Grzunov

PUBLIC HEALTH INSTITUTE ZADAR, Kolovare 2, Zadar, Croatia
marijana.matek-saric@zd.htnet.hr; jgrzunov@zjz-zadar.hr

Abstract

Laboratories may check the accuracy of heavy metal measurements in marine

sediments by taking part in an inter-laboratory comparison. Measurement methods
for all metals must be validated and laboratories must implement good laboratory
practices to avoid the risk of contamination and/or loss of analytes during the
pretreatment processes. The risk of contamination should be avoided by using
reagents that are supra-pure or that are of analytical grade as well as by using Milli-Q
deionized water (<1µSi) for preparing all lab ware and samples. Sediment samples
are dry samples, therefore samples should be homogenized (shaking 2 min) before
weighing. It is important to wait a few minutes and accurately weigh the samples in a
labeled Teflon vessels before opening the bottles. For moisture content determination
at least two samples should be prepared (in separated cups, not in digestion
vessels). At least one certified reference material should be used and prepared in
duplicate for each digestion batch (12 places). The reference digestions are to be
prepared in a similar manner as the actual samples. It is best to use a reference
material that is of similar composition and concentration range as the actual samples.
At least two blanks should be prepared for each batch of analytes. Sediment samples
must be treated in closed Teflon vessels with hydrofluoric acid (HF) in combination
with nitric acid (HNO3). The concentration of the standard solution must be calculated
so that they bracket the concentrations of the samples and the reference materials. In
order to check the accuracy of the prepared curve an independent standard is
prepared.

Key words

Proficiency testing, inter-laboratory comparison, marine sediment, heavy metals

177
 Marijana Matek Sarić, Judita Grzunov: Proficiency testing: Experiences from Croatia on the
issue of heavy metals determination in marine sediments

1 INTRODUCTION

Proficiency testing (PT) is a kind of inter-laboratory comparison used for

determination of the competence of individual laboratories for specific tests and to
assist in resolving of inter-laboratory differences. Inter-laboratory comparison is the
organization, performance and evaluation of tests of the same items or identical
portions of an effectively homogeneous material in two or more different laboratories
in accordance with predetermined conditions [1, 2, 3]. Laboratories are encouraged
to participate in PT to control the accuracy of their results. The procedure for PT is as
follows: each participant in a PT receives a sample prepared by the organizer and
has to determine a list of analytes; the results are sent back to the organizer and the
collected data are analyzed [4].
A test material distributed by a PT provider must be similar to the material that is
routinely analyzed in testing laboratories with respect to the composition of the
matrix, the concentration range, and/or the quantity of the analyte. As a general
principle, procedures used by testing laboratories participating in the comparison
should simulate those used in their routine analytical work [1, 3].
When the number of testing laboratories participating in PT is a large enough (more
than 20-30), the assigned value of the test material can be calculated from analytical
results of the laboratories as a consensus value (as the mean of the median of the
results) while the performance of the participating laboratories is assessed based on
the differences between their results and the assigned value. A performance score is
calculated for each laboratory using statistical schemes detailed in [1, 2, 3]. The
competence of PT schemes providers is assessed according to the guidelines [5].
PT has dramatically pervaded throughout the world of chemical measurement over
the last few decades. A lot of nationally and internationally recognized PT providers
are active in different analytical fields.
This paper describes the inter-comparison exercise in our laboratory (Health Ecology
Service of Public Health Institute Zadar) for the determination of trace elements (Cu,
Cr, Fe, Mn, Ni and Zn) in a marine sediment sample (IAEA-158). The sample was
prepared at the International Atomic Agency, Marine Environmental Studies
Laboratory, Monaco (IAE-MEL). All participants were requested to determine as
many elements as they normally analyze from the following suite: As, Ag, Cd, Co, Cr,
Cu, Fe, Hg (total and methyl Hg), Mn, Pb, Se, Sn, Sr, Zn.

2 MATERIAL AND METHODS

Measurement methods for all metals must be validated and the laboratories must
implement good laboratory practices to avoid the risk of contamination and/or the loss
of analytes during pretreatment process.

Validation of methods for AAS determination of metals must include validation criteria
described in Table 1 [6].

178
 Marijana Matek Sarić, Judita Grzunov: Proficiency testing: Experiences from Croatia on the
issue of heavy metals determination in marine sediments

Table 1-List of validation criteria

Validation parameters Criteria of acceptance
Linearity k≥0,99
Precision
-reputability of measurement RSD≤ 1 %
-repeatability of sample measurement RSD≤ 2 %
intermediary precision RSD≤ 5 %
Accuracy 100 ± 10 %
stability of measurement standard information
Detection limit information
Quantification limit information

The risk of contamination should be avoided by using reagents that are supra-pure or
of analytical grade as well as by using Milli-Q deionized water (<1µSi) for preparing
all lab ware and samples. Water conductivity (Milli-Q deionized and Milli-R) must be
controlled daily.
Cleaning procedure-The best way to clean Teflon vessels and their caps (used in
digestion) is to soak them overnight in a detergent solution (micro solution 2% in tap
water) followed by rinsing them thoroughly with tap water, and then with Milli-R water
(< 60 µSi). Afterward the vessels should be filled with 5 mL of HNO3 (conc.), closed
and placed in a microwave-oven. The oven program for cleaning may be the same as
the digestion program (Table 2). Allow vessels to cool to room temperature, release
pressure carefully by opening the valve, and only then open the vessels. Empty the
vessels (note that the acid can be recycled for more cycles of cleaning) and rinse
them carefully with Milli-Q water. Leave them to dry under a laminar flow hood. Once
Teflon vessels are dry, they should be closed and packaged in polyethylene bags to
prevent the risk of contamination prior to use. This procedure should in fact be used
for cleaning all plastic lab ware in general, but instead of microwave-oven cleaning
digestion procedure plastic lab ware in general should be left to stand in 20% (v/v)
HNO3 conc. solution at least 6 h. If affordable, it is always better to use unused/new
lab ware. Our laboratory follows the cleaning procedure for cleaning lab ware as
described in SOP Cleaning procedure, RU 5.5/1-3 [9, 10].
Weighing-Sediment samples are dry samples, therefore they should be homogenized
(shaking 2 min) before weighing. Before opening the bottle, wait a few minutes and
accurately weigh (4 digit analytical balance: linearity ± 0,2 mg; repeatability SD ±0,03
mg) the sample in labeled teflon vessel.
sture content-The material can easily pick up moisture during storage. Therefore it is
necessary to determine the water content of the material at the time of analysis. To
do so a separate sub-sample (not the same sample as used for analysis) is dried to a
constant weight at 105 °C (usually at least 24 hours).
Analytical quality control-At least one certified reference material should be used and
prepared in duplicate for each digestion batch (12 places). The digestion of the
reference material is prepared in a similar manner as the samples. The reference
material used should be similar in composition and concentration range to the
samples used. For the presented study we used IAEA-433 as a quality control.
Digestion-At least two blanks should be prepared for each batch of analysis. They
should be prepared in a similar manner as a sample, except that no sample is added
to the digestion vessels.

179
 Marijana Matek Sarić, Judita Grzunov: Proficiency testing: Experiences from Croatia on the
issue of heavy metals determination in marine sediments

Sample digestion is a pretreatment step that may provide contamination risk and/or
loss of analytes. Therefore sediment samples must be treated in closed Teflon
vessels with hydrofluoric acid (HF) in combination with HNO3. The use of HF is
essential because it is the only acid that completely dissolves silicate lattices and
therefore releases all material. This digestion procedure is suitable for later
determination of metals like Fe, Cu, Cr, Zn, Ni and Mn.
All vessels must be closed tightly and the appropriate digestion program must be
used (Table 2). Polypropylene graduated tubes with known quantities of boric acid
(0,8 g) should be prepared and weight in order to later neutralize HF and HNO3 used
for the digestion of samples.

Table 2 - Digestion program for determination of metals in marine sediment [7].

Step Time Power Press Temp 1 Temp 2

1 00:05:00 250 0 0 00
2 00:05:00 400 0 0 0
3 00:05:00 650 0 0 0
4 00:05:00 250 0 0 0

When the digestion is completed, samples should be allowed to cool to room

temperature, followed by the release of pressure by carefully opening the valve. Only
then vessels can be opened. Samples are to be transferred into the above mentioned
polypropylene graduated tubes (already containing dissolved boric acid). The
emptied Teflon vessels previously containing the samples should immediately be
rinsed with Milli-Q water into graduated tubes (at least 3 times). Those graduated
tubes must be diluted with Milli-Q water to the mark.
Preparation of metal standard solutions for the calibration curve-The calibration curve
must be generated from at least 3 points (standard solutions of different
concentration) plus a zero calibration. The concentration of the standard solutions
must be calculated so that they bracket the concentrations of the samples as well as
the concentration of the reference material. If the concentration of a sample exceeds
the limit of the calibration curve, the sample must be either diluted to a concentration
within the calibration curve, or the calibration curve must be extended by using an
additional and higher concentration standard. If, on the contrary, the concentration of
the samples is lower than the lowest calibration curve's point, a new calibration curve
must be prepared.
In order to prepare a calibration curve, add approximately 10 mL of Milli-Q water into
a clean polypropylene tube (50 mL) and add all reagents needed to obtain a similar
matrix as the matrix of the samples (e.g. Boric acid, HNO3, HF, etc). Add the
appropriate quantity of standard solution using a micropipette, dilute to the mark with
Milli-Q water and shake well. The thereby produced solution can be kept for a few
days if stored in a refrigerator (+4°C). Depending on the element, it may be
necessary to prepare an intermediate stock standard solution.
In order to check the accuracy of the prepared calibration curve an independent
standard is prepared. The concentration of the external calibration verification (ECV)
should be within the calibration curve. This solution is prepared as described above
but a different stock standard solution (1000 µg/mL) is used [9].
Analytical conditions may change for each element, so it is necessary to first carefully
read the analytical methods book of the AAS before starting an analysis [8].

180
 Marijana Matek Sarić, Judita Grzunov: Proficiency testing: Experiences from Croatia on the
issue of heavy metals determination in marine sediments

3 RESULTS

Reporting of results-Make at least three, but preferable six independent

determinations for each element and report the results as the arithmetic mean and
the standard deviation in mg/kg on a dry weight basis for both, sample (Table 3) and
RM (Table 4).
Table 3-Table of analyses. Moisture content is 1,47%. Inter-comparison Sample
Number 158, number of bottle analyzed No 19
Ele Unit Average Mean Stan- Detecti Quanti Lineari Repeatabili
me sample Value dard on fication ty ty of
nts weight devia- limit* limit* sample
(mg) tion* measure-
ment
RSD<2%
Cr mg/kg 266,758 62,55 0,989 0,0003 0,0010 0,9997 1,58
Cu mg/kg 266,758 49,93 0,952 0,0003 0,0010 0,9996 1,90
Fe mg/kg 266,758 28,15 0,391 0,0015 0,005 1,0000 1,39
Mn mg/kg 266,758 376,26 2,487 0,0021 0,007 0,9998 0,66
Ni mg/kg 266,758 39,37 0,441 0,0018 0,006 0,9994 1,12
Zn mg/kg 266,758 136,55 2,470 0,0003 0,0010 0,9999 1,81

Table 4-Table of analyses of RM. Moisture content is 1,07%. IAEA 433

Ele Unit Average Mean Stan- Lineari Repea 95% Accuracy
me sample Value dard ty tability Confide 100±10 %
nts weight deviati of nce
(mg) on* sample Interval
measu (mg/kg)
rement
RSD<2
%
Cr mg/kg 273,294 135,76 0,850 0,9997 0,626 134-138 99,82
(136)
Cu mg/kg 273,294 30,582 0,272 0,9996 0,888 30,2- 99,29
31,4
(30,8)
Fe mg/kg 273,294 40,864 0,568 1,0000 1,389 40,3- 100,15
41,3
(40,8)
Mn mg/kg 273,294 315,26 4,067 0,9998 1,290 312-320 99,77
(316)
Ni mg/kg 273,294 38,53 0,206 0,9994 0,534 38,7- 97,79
40,1
(40,8)
Zn mg/kg 273,294 101,72 1,414 0,9999 1,390 99-103 100,71
(101)
*Standard deviation and detection limit should be reported in the same unit as results

181
 Marijana Matek Sarić, Judita Grzunov: Proficiency testing: Experiences from Croatia on the
issue of heavy metals determination in marine sediments

4 CONCLUSIONS

The accuracy of the results of heavy metals analysis is influenced by several

procedures, including: lab ware cleaning, quality of used water, sampling, sample
preparation, sample processing, and sample analysis. Measurement methods for all
metals must be validated and laboratories must implement good laboratory practices
to avoid the risk of contamination and/or loss of analytes during pretreatment
process. Simultaneous and paired analysis of RM and test portion is necessary for
PT purpose in order to check results.

REFERENCES

[1] Thompson M, Wood R. International harmonized protocol for proficiency testing

of (chemical) analytical laboratories. J AOAC, 76:926-940, (1993)
[2] ISO Proficiency testing by interlaboratory comparison. Part 1: Development
and operation of proficiency testing schemes. ISO/IEC Guide 43-1, ISO,
Geneva, (1997)
[3] Kuselman I, Pavlichenko M. Designs of experiment for proficiency testing with
a limited number of participants. Accredit Qual Assur, 9:387-390,(2004).
[4] Lafargue E, Biogeaud S, Rutledge DN, Feinberg MH. Proficiency testing
schemes: solutions from homogenity control. Accredit Qual Assur, 9:387-
390,(2004)
[5] ILAC. Guidelines for the requirements for the competence of providers of
proficiency testing schemes. ILAC G13:2000, International Laboratory
Accreditation Co-operation (ILAC), (2000)
[6] Lazarić K. Validacija analitičkih metoda. In: Validacija metoda, seminar i
raspravljaonica; ed. Gašljević V, Hrvatsko mjeriteljsko društvo, str.1-8.(2001).
[7] Milestone acid digestion cookbook for Microwave laboratory system MLS 1200
MEGA, January (1996)
[8] Atomic Absorption Spectrophotometry cook book, Basic Conditions for Analysis
of Principle of Atomic Absorption Spectrometry, Shimadzu (1995)
[9] Laboratory Procedure Book, Training workshop on the analysis of trace metals
in biological and sediment sample. International atomic Energy Agency-Marine
Environmental Laboratory (2003)
[10] SOP, RU 5.5/1-3 Pranje posuđa za kemijske analize, Zavod za javno zdravstvo
Zadar, Služba za zdravstvenu ekologiju (2007)

182