Professional Documents
Culture Documents
Participation of The Laboratory of Neutron Activation Analysis
Participation of The Laboratory of Neutron Activation Analysis
Abstract
The Nuclear Technology Development Centre/Brazilian Commission for Nuclear Energy, CDTN/CNEN,
is making an effort to meet the ISO/IEC 17025 requirements aiming at laboratory accreditation. It means to
determine, recognize and promote technical competence. The invitation to participate in the Proficiency Tests
ISE/IPE 2013-1, through WEPAL (Wageningen Evaluating Programs for Analytical Laboratories), in the
framework of the IAEA TC Project RLA0037, was a very good opportunity to evaluate the performance of the
Laboratory for Neutron Activation Analysis, applying the k0-method. The results of reference certified materials
GBW 0805, Tea, and BCR-320R, Channel Sediment, showed that related to some results of z-score, GBW 0805,
the performance were questionable, however, concerning En-number, all results were acceptable. Related to
BCR-320R, both scores pointed out the good performance of analyse. The summary of the statistical evaluation
of the performance of the IPE (plants samples) and ISE (geological samples) analyses, z and Zeta-scores, show
that for ISE, geological samples, the performance was better than IPE, plant samples, although the results for
plants were also good.
1. INTRODUCTION
The technical competence of a laboratory to achieve accurate and reliable results depends on
many factors such as qualification, training and experience of its staff; adequate equipment,
properly calibrated and maintained; suitable quality assurance procedures; proper sampling
practices; appropriate testing procedures; valid test methods; traceability of measurements to
national standards; accurate recording and reporting procedures; suitable testing facilities.
The ISO/IEC 17025, “General requirements for the competence of testing and calibration
laboratories” [1], guides the laboratories to implement a quality system aimed at improving
their ability to consistently produce valid results. It is also the basis for accreditation, a formal
recognition of a demonstration of competence. A prerequisite for the accreditation of a
laboratory is to have a documented quality management system. The usual contents of the
quality manual follow the outline of the ISO/IEC 17025 standard [2].
The ISO/IEC 17025 determines internal and external Quality Control (QC) and Quality
Assurance (QA) clauses. To meet its quality requirements, however, the precise way about
proving consistency and reliability of analysis is not defined. On the other hand, accreditation
bodies have built some prescriptive clauses to facilitate meeting the requirements of
ISO/IEC17025 in an effective manner. An externally provided proficiency testing program is
a useful tool; however, participating in this program will not necessarily mean that all quality
assurance aspects have been met.
The laboratory should participate in proficiency testing programs for all types of analyses
undertaken in that laboratory, whenever applicable. The laboratory staff will profit from
participating in such programs, gaining confidence in their abilities, and awareness of their
capabilities. The laboratory staff will gradually improve its knowledge and maintain its level
of competence. The difficulties related to analyses can be detected and training needs
196
identified. The laboratory staff will know that it is playing a valuable role to ensure product
quality and safety, which reinforces teamwork [3].
CDTN Laboratory for Neutron Activation Analysis (LAN) is part of the Unit of Nuclear
Chemistry and Radiochemistry, Division of Reactor and Analytical Techniques,
CDTN/CNEN. The activities of the Laboratory comprise the delayed fission neutron
activation analysis, instrumental (relative and k0-methods) and radiochemical / chemical
methods. These methods are responsible for relevant percentage of CDTN’s analysis demand,
meeting the clients' analytical needs and researches developed by the Laboratory, by CDTN
and by other institutions. Over the years its work has been linked to the goals of the country
and its institutions. Neutron activation analysis is responsible for a high percentage of the
analytical demand. The k0-Instrumental method [5] is used in the majority of the analyses
performed to meet clients’ request and researches in order to determine many elements in
several matrixes and a large range of concentrations. In Brazil, CDTN is the only Institute that
fully masters the instrumental neutron activation analysis k0-method using its own nuclear
reactor. In this context, it is of utmost importance to release analytical results with quality.
At the end of 2012, the LAN was invited to participate in a new round of Proficiency Tests
ISE/IPE 2013-1, through WEPAL (Wageningen Evaluating Programs for Analytical
Laboratories), in the framework of the IAEA TC project RLA0037. Several neutron activation
analysis (NAA) laboratories participated in two consecutive proficiency testing rounds
(ISE/IPE 2011-4/2012-1) by interlaboratory comparison organized by WEPAL in conjunction
with the TU Delft (Expert Peter Bode).
In order to meet the guidelines of the institute and the requirements of analytical demand, the
laboratory is being prepared to get future accreditation. Due to this, LAN is concerned about
QC/QA and has procedures routinely applied. The procedures comprise the use of several
reference certified materials (CRM), replicate of samples, determination of background and
blank, checking the HPGe detector (energy calibration, FWHM) and others. Therefore, to
participate in the WEPAL Proficiency Test is a way to comply with the ISO/IEC 17025.
2. EXPERIMENTAL PART
The samples - four geological samples and four plant samples were provided in each round for
analysis sent by WEPAL, named ISE 2013/1 and IPE 2013/1, respectively. They arrived at
197
CDTN/CNEN on December 7th 2012 and were prepared on December 10th to be analyzed in
January, 2013.
The method for dry mass determination was to weigh 1g of each sample (triplicate), to keep
in an oven at 105ºC until constant mass (6 days), weigh again and calculate the dry mass.
The k0-standadized method is routinely applied at LAN. To analyse the samples, the neutron
flux monitor used was Al-(0.1%)Au foil, cut into discs (about 7 mg, 6 mm in diameter and 0.1
mm thick). The alloy Al-(0.1%)Au is a certified reference material IRMM-530R, from
Institute for Reference Materials and Measurements, Belgium.
The sample geometry was a vial 4.3 mm inner height and 9.6 mm inner diameter for
geological samples and another vial 7 mm inner height and 9.6 mm inner diameter for
biological samples.
As internal standard, certified reference materials were analysed together: GBW0805, Tea [6],
and BCR-320R, Channel Sediment [7].
The samples, RCM and monitors were inserted in another polyethylene vial, stacked in
“sandwich” form.
2.1.1. Irradiation
The samples and the RCM were irradiated for 8 hours in the TRIGA MARK I IPR-R1 reactor
at 100 kW in the carousel, on January 7th 2013, in the IC-7 and IC-8.
The thermal neutron fluence rate in the irradiation channels was 6.35 × 1011 cm-2·s-1, and the
parameters f, 22.32 and α, −0.0022 in the IC-7. In the IC-10, the thermal neutron flux rate was
5.99 × 1011 cm-2· s-1, f was 20.65 and α − 0.0022 [8].
Related to method for flux gradient correction, in axial direction, the correction was based on
Au monitors measurements. The monitors were irradiated together with the samples in
sandwich form; in radial direction corrections were not necessary, once the monitor and the
sample have about the same diameter.
The gamma measurements started on January 9th and were carried out in an HPGe detector,
vertical dipstick, 50% relative efficiency. The full energy photopeak efficiency calibration
approach was done using point sources (198Au, 241Am, 109Cd, 203Hg, 137Cs, 65Zn). The dead
time correction was made by KayWin software, V.2.42 [9] – during elemental concentrations
- based on the percentage of dead time. The software used for gamma spectra acquisition was
Genie 2K, CANBERRA.
198
The sample-detector end-cap distances during the gamma spectrometry were:
- measurement 1: 2 -3 days after irradiation, at 10 or 20 cm distance sample-detector
- measurement 2: 4 -7 days after irradiation, at 2 or 5 cm distance sample-detector
- measurement 3: 15 -35 days after irradiation, at 2 cm distance sample-detector
The spectra acquired were analysed by HyperLab software [10] producing net peak area
(*.PTF) and spectrum (*.SPE) files. These files were used in the software package Kayzero
for Windows, v. 2.42 [9], for elemental concentrations calculation. The calculations were
based on the same gamma lines of the radionuclides in all samples.
The uncertainty calculation, Unc, was according to the Eq. 1 for n = 2, and according to Eq. 2,
for n = 3. It means, if a radionuclide was detected in the triplicate aliquots, n = 3, in 2 of
triplicate, n = 2 and in only one aliquot, the uncertainty will be the value calculated by the
software. The equations are:
1
UncAvg = 2
Unc21 +Unc22 (1)
To evaluate the performance of the laboratory in the proficiency testing, there are a number of
different scoring systems available. Performance criteria are associated to each scoring system
which allows participants to assess whether their performance is acceptable. There are many
score systems but the z-score is the most frequently used in schemes covering chemical
analysis.
A z-score is calculated by comparing the difference between a participant’s result and the
assigned value for the PT sample with a standard deviation for proficiency assessment. Zeta-
score (ζ-score) takes into account both the uncertainty in the assigned value and the
uncertainty in participants’ results [11].
The analytical performance of k0-INAA at LAN-CDTN was evaluated expressing the results
through z-score, Zeta-score and En-number. For the samples in the WEPAL IPE/ISE
proficiency tests, the performance was calculated considering the assigned values for
reference materials through z and Zeta-scores.
199
z-score:
ValueExperimental − ValueAssigned
z=
St.Dev.Assigned (3)
To compare the results – experimental and assigned values - the criterion was applied,
meaning that the evaluation of the performance of the method was satisfactory if |z| ≤ 2,
questionable, if 2 < |z| <3 and unsatisfactory, if |z| ≥ 3. The interpretation is based on the
properties of the normal distribution of data. In a normal distribution, approximately 95% of
values are expected to lie within ±2 standard deviations of the mean value.
Zeta-score:
where
2 2
U Experimental = u AREA + umethod (5)
In Equation 5, uAREA is the uncertainty of the net peak area in the gamma spectrum and umethod
is the overall uncertainty of k0-NAA established at the CDTN/CNEN as 3.5% with a coverage
factor k = 1.
En-number:
ValueExperimental − ValueAssigned
En =
2
U LabExpand 2
+ U Assigned (6)
where ULabExpand and UAssigned are the expanded uncertainties (k = 2) of the experimental result
and the assigned result, respectively, and
U LabExpand = 2 ⋅U Experimental
(7)
U Assigned = 2 ⋅ St.Dev.Assigned
(8)
En-number was also calculated to measure the agreement between the experimental result and
the assigned value, taking into account the expanded uncertainty (k = 2) of the both values. To
compare the results the criterion |En| ≤ 1 was applied, meaning that the evaluation of the
200
performance of the method was satisfactory and if |En| > 1, the performance was
unsatisfactory.
3. RESULTS
The internal quality control materials used were GBW0805, Tea, and BCR-320R, Channel
Sediment, analysed in duplicate.
Tables 1 and 2 show the experimental results and certified and non-certified values for GBW
0805 and for BCR-320R. For both reference materials, Z-score and the En-number are also
displayed.
201
Observing z-score column, several results point out that the analysis had satisfactory
performance, but others, in bold, are between 2 and 3; it means that the performance is
questionable. Observing En-number, all results are acceptable. It means that the experimental
results reference materials are inside the expanded uncertainties.
Table 3 shows the results for ISE-2013-1 PT and Table 4, the results of IPE-2013-1 PT. In
both tables the values of WEPAL are also displayed.
202
TABLE 4. RESULTS OF IPE-2013-1 PT
203
TABLE 5. RESULTS OF ISE-2013-1 PT
204
Tables 6 and 7 shows the results of the statistical evaluation of the results obtained by LAN-
CDTN.
Note: #, less than 8 values, no mean and Sd calculated; *, straggler; **, outlier; -, not calculated
205
TABLE 7. Z- AND ZETA-SCORES FOR ISE-2013-1 ROUND
870 890 919 961
(Clay from river basin) (Sandy soil) (Sandy soil) (Clay)
z-score ζ-score z-score ζ-score z-score ζ-score z-score ζ-score
El (WEPAL) (CDTN) (WEPAL) (CDTN) (WEPAL) (CDTN) (WEPAL) (CDTN)
Ag <# - <# - <# - <# -
As +0.84 +0.78 +0.93 +0.88 +1.16 +1.09 +0.96 +0.75
Ba −1.80 −1.25 +0.23 +0.22 −0.31 −0.28 −0.40 −0.30
Br +0.66 +0.60 +0.69 +0.66 +0.88 +0.82 +0.70 +0.49
Ca < - < - < - < -
Cd < - < - < - < -
Ce +1.04 +0.91 +0.81 +0.66 +0.71 +0.57 +1.60 +0.99
Co +1.52 +1.27 +0.51 +0.49 +0.42 +0.40 +1.53 +0.88
Cr +1.74 +1.53 +1.52 +1.38 +1.99 +1.39 *+2.27 +1.40
Cs +1.18 +1.07 +1.19 +1.11 +1.31 +1.20 +1.21 +0.95
Cu **−4.19 - **−4.53 - **−3.88 - < -
Fe +1.63 +1.27 +1.77 +1.58 +1.69 +1.53 *+2.35 +1.06
Ga +1.20 +1.06 −0.26 −0.25 +0.38 +0.37 +0.67 +0.62
Hg < - < - < - < -
K +0.82 +0.65 *+2.75 +2.12 *+2.44 +1.81 *+2.27 +1.02
La +0.76 +0.61 +1.30 +0.96 +0.61 +0.47 *+2.24 +1.04
Mo < - +1.80 +0.85 < - *+2.36 +1.61
Na +1.76 +1.42 +1.71 +1.57 *+2.46 +2.10 **+3.35 +1.41
Nd +0.73 +0.68 −0.05 −0.04 +1.47 +0.92 *+2.60 +1.45
Rb +1.56 +1.21 +1.55 +1.31 +1.59 +1.38 +1.98 +0.91
Sb +0.83 +0.73 +0.61 +0.55 +0.24 +0.24 +0.69 +0.58
Sc +1.51 +1.25 +0.97 +0.88 *+2.22 +1.66 +1.77 +1.09
Se <# - <# - <# - <# -
Sn < - <# - < - < -
Sr +1.47 +1.04 **+3.03 +1.90 *−2.06 −1.94 **+4.62 +2.34
Th +0.84 +0.77 +0.81 +0.75 −0.59 −0.25 +1.19 +0.94
U −0.06 −0.05 +0.30 +0.29 +0.22 +0.16 +0.35 +0.31
§ §
W +0.14 +0.08 +1.10 +1.05 **+183 **+177 −0.06 −0.05
Zn *+2.86 +0.99 - - - - +0.92 +0.54
Note: #, less than 8 values, no mean and Sd calculated; *, straggler; **, outlier; -, not calculated, § , typing error when filling
in the form.
206
In Table 8, a summary of all results obtained from statistic evaluation is displayed. At first
glance, it is possible to observe that more elements were determined in geological matrixes,
which was expected as they are more complex matrixes in terms of activatable elements.
Related to outlier percentage values, z- and zeta-scores, ISE samples presented better results,
pointing out better performance.
4. CONCLUSIONS
The Laboratory for Neutron Activation Analysis, CDTN/CNEN, accepted the invitation to
participate in the WEPAL PT, ISE and IPE-2013-1 round, in spite of the changes in to routine
infrastructure conditions due to lab renovation and reactor process licensing. The gamma
spectrometry system was moved to another laboratory off the controlled area. A special
license from radioprotection staff was needed in order to participate in this proficiency
testing.
The samples – four plants and four geological matrixes – were analysed by k0-standardized
method according to usual procedure applied at the laboratory.
The results of reference certified materials GBW 0805, Tea Leaves, and BCR-320R, Channel
Sediment, showed that related to some results of z-score, GBW 0805, the performance were
questionable, however, concerning En-number, all results were acceptable. Related to BCR-
320R, both score pointed out the good performance of analyse.
The summary of the statistical evaluation of the performance of the IPE and ISE analysis
shows that for ISE, geological samples, the performance was better than IPE, plant samples,
although the results for plants were also good. It can be explained by the following: all the
samples were irradiated in just one irradiation; it is usually done in two irradiations, just to
allow more time to carry out the gamma measurements. Time was also short to perform all
steps before the deadline to send the results to WEPAL. Another period of 8 hours irradiation
due to restricted use of the reactor could not be performed. The alternative was to analyse all
samples together. Other suggestion is possible non-homogeneity of plants samples. The non-
homogeneity of the samples can be explained why the results were not as good as for the
geological samples.
The actions for the next participation of PT, will be: to perform 2 irradiations period, one for
geological material and another for plants matrixes; to carry out gamma spectrometry for
longer time in order to minimize uncertainties in the results, to perform 5 replicates, to
decrease possible interferences related to inhomogeneity. The future action plan toward
sustainability of quality is to keep on participating in proficiency testing rounds.
To participate in this PT was a very good opportunity to evaluate the performance of the
Laboratory of Neutron Activation, CDTN/CNEN, not only related to PT itself, but, mainly, in
this specific round, because there were only participants were applying the same technique,
neutron activation analysis, and, among them, several applying the same k0-method.
ACKNOWLEDGEMENT
To Dr. Peter Bode, TU Delft, the Netherlands, and IAEA (TC Project RLA0037) for the
invitation, to Direction of CDTN/CNEN and Radioprotection staff for understanding the
importance of this participation and allowing the analyses. The staff of the reactor for the
special irradiation of the samples (supervisors Fausto Maretti Júnior, Luiz Claudio A. Souza
and operators Paulo Fernando Oliveira e Luiz Otávio I. Sette Camara). Also to Dr. Radojko
207
Jaćimović, Institute Jožef Stefan, Slovenia, for his valuable suggestions mainly related to
statistical evaluations.
REFERENCES
208