Uncertainty in Chemistry Vol. 3 Examples

Nineta Hrastelj
Ricardo Bettencourt da Silva Editors
Traceability,
Validation and
Measurement
Uncertainty in
Chemistry: Vol. 3
Practical Examples
Traceability, Validation and Measurement
Uncertainty in Chemistry: Vol. 3
Nineta Hrastelj Ricardo Bettencourt da Silva
•
Editors
Traceability, Validation
and Measurement
Uncertainty in Chemistry:
Vol. 3
Practical Examples
123
Editors
Nineta Hrastelj Ricardo Bettencourt da Silva
European Association for Chemical Faculdade de Ciências
and Molecular Sciences (EuCheMS) Centro de Química Estrutural
Brussels, Belgium Universidade de Lisboa
Lisbon, Portugal
ISBN 978-3-030-20346-7 ISBN 978-3-030-20347-4 (eBook)

https://doi.org/10.1007/978-3-030-20347-4
© Springer Nature Switzerland AG 2019
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Foreword
Discendo docebis, docendo disces

By learning you will teach; by teaching you will learn
Latin Proverb
Chemical and bio-analytical measurements are omnipresent and often very impor-
tant in our society. Just think of the quality of the food we eat, the air we breathe, the
role of these measurements in health care, in trade and in research. In all these cases,
people strive to get reliable data. There is an international standard for assuring the
quality of measurement data, namely EN ISO/IEC-17025. It contains particular
management as well as technical requirements. These technical requirements are
linked to the science behind the measurements, meaning that metrological issues
such as traceability, uncertainty and validation are at the heart of this.
So as to provide Life-Long Learning in this area, the TrainMiC® programme
[www.trainmic.org] was conceived in 2001 by the Institute for Reference Materials
and Measurements of the European Commission Joint Research Centre. First, it
addressed the need arising with those countries wanting to enter the EU at that
moment. Rather than approaching this training in an anecdotal way and organizing
ad hoc events, a programme was set up—called TrainMiC®—to create harmonised
training material as well as to disseminate knowledge in the various countries via a
network of authorised TrainMiC® trainers. Afterwards, the TrainMiC® programme
spread to the rest of the EU and Europe’s largest Life-Long Learning programme in
this area was created. Up to now, 20 national TrainMiC® teams have been set up
and more than 9000 people have been trained all across Europe at the end of 2014.
Trainers quickly realised the importance of having suitable examples for their
training events. They realised soon that creating examples adapted to the various
audiences is quite a labour-intensive activity. For this reason, sharing such exam-
ples proved to be an attractive proposition. Today, examples are reviewed and then
published mainly in an e-collaboration environment which is only available for the
authorised trainers. It was decided to publish part of the examples also in the format
of a series of books, with a first volume that appeared in 2010, a second one in 2011
and this now being the third volume.
v
vi Foreword
In the meantime, there is a lot of experience in using these examples for training
purposes. First and foremost these examples should be seen as a vehicle for
problem-based learning. Their intention is not to be a definitive guidance document,
explaining a unique way to approach a given problem. Typically, they are used in a
classroom context, where small groups of trainees will try to provide answers to the
questions raised. It is recommended that they have group discussions, because
discussion creates learning. Later on, the different groups present the outcome in a
session in plenum. Also from this, there is learning. The answers provided in the
book, are most likely one of the possibilities. As we all know, a lot depends on
interpretation.
Brussels, Belgium Philip Taylor
Philip Taylor works for the European Commission Joint

Research Centre. As an analytical chemist, he has been involved
in many different research areas, techniques and type of activities,
ranging from atomic to mass spectrometry, from fundamental
constant work to applied industrial analysis, from isotopes to
nuclear safeguards, from capacity building to training and
teaching.
His current research interests lie in measurement and testing
linked to aviation security and detection of CBRNE security
threats (Chemical, Biological, Radioactive, Nuclear and
Explosives).
There has however been one recurring theme which links up
all these areas: how to ensure quality of measurement data. This is
very important in the context of ISO/IEC-17025 laboratory
accreditation and European Regulations.
For this reason, in 2001 he created a European Training
platform for metrology in chemistry (TrainMiC®) involving
experts from all across the EU. This was later followed by a
teaching initiative, the Euromaster Measurement Science in
Chemistry international master programme.
Preface
The publication of the ISO Guide to the Expression of Uncertainty in Measurement

in 1993, known as the GUM, has set the ground for the development of metrology
in physics, chemistry and biology, in particular, for the generalisation of the
evaluation of the uncertainty of routine measurements. This guide, subject to minor
revisions in 1995 and 2008, presented the tools and defined the conventions for
using the measurement uncertainty concept in the resolution of daily societal
problems and demands.
Although this document established a solid base for the development of
metrology, the science of measurement, it was only after the publication of the
ISO/IEC 17025 Standard in 2000, which imposed the evaluation of the uncertainty
of accredited measurements, that GUM started to be used in routine laboratories.
Fortunately for chemists, also in 2000, Eurachem published the second edition of a
guide for the quantification of measurement uncertainty. The new revision of the
guide presented alternatives to the detailed assessment of measurements: the
so-called “top-down” approaches. These approaches allow laboratories to apply the
GUM principles to routine measurements in chemistry in a cost-effective way.
Nowadays, many other guidance documents have been published to help analysts in
the evaluation of measurement uncertainty based on Eurachem’s pragmatic inter-
pretation of the implementation of GUM principles in chemistry.
The latest revision of the ISO/IEC 17025 standard attracts even more attention to
the measurement uncertainty concept by asking laboratories to use it in conformity
assessment and in quantifying the risk of wrong conformity decisions.
Nineteen years after important milestones in metrology in chemistry were attained,
there is a consensus that accredited laboratories are experienced in “top-down”
evaluations of measurement uncertainty. However, most of them are not familiar with
the way to define and assess the traceability of measurements in chemistry, in par-
ticular, when the specific analytical field has no adequate and stable references for
measurements and, therefore, the analysts should be more careful when planning to
compare results from different laboratories.
vii
viii Preface
Laboratories not involved in highly regulated analytical fields, such as many of

those dedicated to fundamental and applied research, are frequently not aware of the
ways to both select references for their measurements and to take information about
the measurements’ quality in the interpretation of results. The inadequate selection
of references for measurements and the underestimation of relevant effects on
measurement quality can have severe impacts on the outcomes of the studies
undertaken.
Since the state of the art seems to be able to solve most challenges faced by
measurements in chemistry, the goal now should be to ensure that this knowledge
reaches those who need it.
TrainMiC® Life-Long learning programme has been teaching Metrology in
Chemistry successfully since 2001. In 2007, a Portuguese team of 7 authorised
trainers was set up and, since then, it has organised 13 courses that trained 211
Portuguese and Brazilian trainees. The courses delivered by the Portuguese team
have been extremely appreciated due to the quality of the contents as well as for the
demonstrated dedication and competence.
In 2011, during the celebration of the tenth anniversary of the TrainMiC®
programme, the Portuguese team was awarded the best Mini-Case elaborated to
increase the interactivity between trainers and trainees. This award led to the
invitation of a member of the Portuguese team to be the Guest Editor of the third
volume of TrainMiC® examples.
The authors of this book of examples wish this document will be useful to many
analysts to improve their measurements and keep them alert to the latest devel-
opments in this field.
Lisbon, Portugal Ricardo Bettencourt da Silva

Contents
Measurement of Total Reactive Phosphorus in Natural Water

by Molecular Spectrophotometry (SMEWW 4500-P D) . . . . . . . . . . . . . 1
Ricardo Bettencourt da Silva,
Maria da Ascensão Rebelo da Silva Trancoso,
Paula Alexandra Lourenço Teixeira,
Cristina Maria Roque Ramiro de Oliveira, Alice Isabel de Jesus Mosca,
Florbela Aura do Sacramento Dias
and Maria Filomena Gomes Ferreira Crujo Camões
Measurement of Concentration of Nitrate in Drinking Water . . . . . . . . 21
Andreja Drolc
Measurement of Moisture Content (Water Content) in Edible Oil
Using the Volumetric Karl Fischer Method According to
ISO 8534:1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Ivo Leito and Lauri Jalukse
Measurement of Polyphosphates in Meat Products
by Spectrophotometric Method According Laboratory-Developed
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Tidža Muhić-Šarac
Task for a Laboratory Measurement of Mass Fraction of TiO2
and Fe-tot in Ilmenite and Titanium Slag . . . . . . . . . . . . . . . . . . . . . . . . 99
Jurij Pustinek, Karmen Rajer Kanduč and Nineta Hrastelj
ix
Editors and Contributors
About the Editors
Nineta Hrastelj works as the General Secretary of the

European Association for Chemical and Molecular
Sciences (EuCheMS). She studied at the University of
Ljubljana, Slovenia. Her research topics were horizon-
tal across analytical chemistry, chemometrics and
metrology. She is author, co-author or editor of about
250 contributions in research, teaching and related
topics of general concern. In the recent years, her expert
work is mostly about quality of analytical measure-
ments, where, amongst others, she has been charing the
TrainMiC Editorial Board and is a founding member of
both, TrainMiC and Measurement Science in
Chemistry (previously AcadeMiC) programme of the
European Commission JRC.
Ricardo Bettencourt da Silva completed his B.Sc. in

chemistry at the Faculty of Sciences of the University
of Lisbon (FCUL), his M.Sc. in bromatology at the
Faculty of Pharmacy of the University of Lisbon and
his Ph.D. in analytical chemistry—metrology in chem-
istry at FCUL. The last two academic degrees were
completed in parallel with his full-time professional
experience as analyst, in official, public and private
laboratories, of the different inorganic and organic
analytes in various types of matrices using classical and
instrumental methods of analysis.
This analytical experience was focused on the
detailed validation of the measurement procedure, test
xi
xii Editors and Contributors
quality control and evaluation of measurement uncer-

tainty. Since 2002, Ricardo has worked regularly as an
assessor of the Portuguese Accreditation Body (IPAC)
and as a trainer and consultant for the accreditation of
chemical laboratories. In 2009, Ricardo was contracted
as a researcher by the Centre for Molecular Sciences
and Materials of the Faculty of Sciences of the
University of Lisbon where he has been continuing
his research work on metrology in chemistry while
collaborating in teaching at national and foreign
universities. Ricardo’s research includes the develop-
ment of approaches for the detailed evaluation of the
uncertainty associated with complex measurements and
the assessment of the sources of lack of comparability
of measurements in some analytical fields. Ricardo has
been a member of the IPAC Accreditation of Chemical
Laboratories Working Group since 2006, the
Eurachem/CITAC Measurement Uncertainty and
Traceability Working Group since 2010, the
Portuguese TrainMiC® team since 2008 and the
TrainMiC®. Editorial Board since 2010.
Contributors
Maria Filomena Gomes Ferreira Crujo Camões Centro de Química Estrutural,

Departamento de Química e Bioquímica, Faculdade de Ciências da Universidade de
Lisboa, Lisbon, Portugal
Maria da Ascensão Rebelo da Silva Trancoso Laboratório Nacional de Energia
e Geologia, Amadora, Portugal
Ricardo Bettencourt da Silva Centro de Química Estrutural, Departamento de
Química e Bioquímica, Faculdade de Ciências da Universidade de Lisboa, Lisbon,
Portugal
Alice Isabel de Jesus Mosca AIM - Consultoria, Formação, Auditoria Unipessoal
Lda, Lisbon, Portugal
Cristina Maria Roque Ramiro de Oliveira Centro de Química Estrutural,
Departamento de Química e Bioquímica, Faculdade de Ciências da Universidade de
Florbela Aura do Sacramento Dias Instituto Português da Qualidade, Costa da
Caparica, Portugal
Andreja Drolc National Institute of Chemistry, Ljubljana, Slovenia
Editors and Contributors xiii
Nineta Hrastelj EuCheMS, Brussels, Belgium

Lauri Jalukse Institute of Chemistry, University of Tartu, Tartu, Estonia
Karmen Rajer Kanduč Cinkarna Celje, Celje, Slovenia
Ivo Leito Institute of Chemistry, University of Tartu, Tartu, Estonia
Tidža Muhić-Šarac Faculty of Science, Department of Chemistry, University of
Sarajevo, Sarajevo, Bosnia and Herzegovina
Jurij Pustinek Cinkarna Celje, Celje, Slovenia
Paula Alexandra Lourenço Teixeira Instituto Superior Técnico, Universidade de
Measurement of Total Reactive
Phosphorus in Natural Water
by Molecular Spectrophotometry
(SMEWW 4500-P D)
Ricardo Bettencourt da Silva, Maria da Ascensão Rebelo da Silva

Trancoso, Paula Alexandra Lourenço Teixeira,
Cristina Maria Roque Ramiro de Oliveira, Alice Isabel de Jesus
Mosca, Florbela Aura do Sacramento Dias
and Maria Filomena Gomes Ferreira Crujo Camões
1 Brief Description of the Analytical Procedure
1.1 Introduction
Phosphorus occurs in natural waters almost solely as phosphates. These are clas-
sified as free phosphate (orthophosphate), condensed phosphates (pyro-, meta-, and
other polyphosphates) and organically bound phosphates.
Various classes of phosphorus compounds can be quantified by molecular
spectrophotometry after different sample pre-treatment (Fig. 1).
The free phosphate and some condensed phosphates are quantified by molecular
spectrophotometry performed on an untreated aliquot (direct spectrophotometry).
This phosphorus compounds fraction is known as “reactive phosphorus” in specific
experimental conditions. Most of the condensed phosphates and some organically
bound phosphate are also analysed by spectrophotometry after the hydrolysis of the
Electronic supplementary material The online version of this chapter (https://doi.org/10.1007/

978-3-030-20347-4_1) contains supplementary material, which is available to authorized users.
R. B. da Silva (&) C. M. R. R. de Oliveira M. F. G. F. C. Camões

Centro de Química Estrutural, Departamento de Química e Bioquímica, Faculdade de
Ciências da Universidade de Lisboa, Lisbon, Portugal
e-mail: rjsilva@fc.ul.pt
C. M. R. R. de Oliveira
e-mail: cmoliveira@fc.ul.pt
M. F. G. F. C. Camões
e-mail: mfcamoes@fc.ul.pt
© Springer Nature Switzerland AG 2019 1

N. Hrastelj and R. Bettencourt da Silva (eds.), Traceability, Validation
and Measurement Uncertainty in Chemistry: Vol. 3,
https://doi.org/10.1007/978-3-030-20347-4_1
2 R. Bettencourt. da Silva et al.
sample aliquot with sulphuric acid at boiling water temperature. This phosphorus
compounds fraction excluding reactive phosphorus is known as acid-hydrolyzable
phosphorus in conditions prescribed by the measurement procedure. Finally, the
total phosphorus can be quantified in the samples after more drastic digestion
conditions (Fig. 1).
Therefore, the presented analytical procedure does not allow the discrimination
of the phosphorus compounds as their chemical classes.
Water sample
Direct 1. H2SO4 hydrolysis; 2. Spectrophotometry 1. Digestion

spectrophotometry 2. Spectrophotometry
Reactive Total
(A+B)*
phosphorus (A) phosphorus (C)
Acid- Organic
hydrolyzable phosphorus (D)
phosphorus (B)
[D=C-(A+B)]
[B=(A+B)-A]
* Reactive and acid hydrolyzable phosphorus
Fig. 1 Steps for the analysis of phosphorus fractions
M. da Ascensão Rebelo da Silva Trancoso

Laboratório Nacional de Energia e Geologia, Amadora, Portugal
e-mail: maria.trancoso@lneg.pt
P. A. L. Teixeira
Centro de Química Estrutural, Instituto Superior Tecnico da Universidade de Lisboa, Lisbon,
Portugal
e-mail: paula.lourenco.teixeira@gmail.com
A. I. de Jesus Mosca
AIM - Consultoria, Formação, Auditoria Unipessoal Lda, Lisbon, Portugal
e-mail: alice.mosca@aim-consultoria.com
F. A. do Sacramento Dias
Instituto Português da Qualidade, Costa da Caparica, Portugal
e-mail: florbelaD@ipq.pt
Measurement of Total Reactive Phosphorus in Natural Water … 3
1.2 Determination of Reactive Phosphorus in Natural Water

by Molecular Spectrophotometry
Principle: Molybdophosphoric acid is formed and reduced by stannous chloride to

intensely coloured molybdenum blue.
Interferences:
Negative interferences are caused by arsenate, fluoride, thorium, bismuth, sulphide,
thiosulphate, thiocyanate, or excess of molybdate. These species are unlikely to be
present in most natural waters.
Procedure:
(i) Preliminary sample treatment: To 100 mL of sample, add one drop of phe-
nolphthalein indicator. Discharge pink colour with a strong acid. If more than
0.25 mL (5 drops) is required, take a smaller sample aliquot, discharge the
pink colour with acid and dilute to 100 mL with distilled water.
(ii) Colour development: Add, with thorough mixing after each addition, 4.0 mL
molybdate reagent (0.02 mol L−1 ammonium molybdate; 5.2 mol L−1 sul-
furic acid) and 0.5 mL stannous chloride reagent (0.11 mol L−1 stannous
chloride in glycerol). Hold samples, calibrators, and reagents within 2 °C of
one another and in the temperature range between 20 ºC and 30 °C.
After 10 min, but before 12 min, using the same specific interval for all
determinations, measure colour photometrically at 690 nm and compare with
a calibration curve, using a purified water blank.
(iii) Preparation of calibrators:
Calibrators were prepared from KH2PO4 salt using procedure described in
Fig. 2.
Fig. 2 Preparation of the calibrators

(iv) Measurement model (to calculate the measured quantity value):
mðlg P) MðP2 O5 Þ
cðmg L1 of P2 O5 Þ ¼
VðmL) 2MðPÞ
c(mg L−1) Reactive phosphorus (RP) mass concentration in the sample expressed as
P2O5;
m(µg P) RP mass interpolated in the calibration curve expressed as P;
V(mL) Sample volumetric aliquot;
M(P2O5) Molar mass of phosphorus pentoxide;
M(P) Molar mass of phosphorous.
2 The Customer’s Requirements Concerning Quality

of the Measurement Result
The measurement aims at assessing the compliance of surface water, intended for
the abstraction of drinking water, with limits settled by Directive 75/440/EEC. This
monitoring must be performed using a measurement procedure meeting require-
ments of Directive 79/869/EEC.
Directive 75/440/EEC excerpt:

Directive 79/869/EEC excerpt:
Directive 79/869/EEC defines “limit of detection” as “the minimum value of the

parameter examined which it is possible to detect”, “precision” as “the range within
which 95% of the results of measurements made on a single sample, using the same
method, are located” and “accuracy” as “the difference between the true value of the
parameter examined and the average experimental value obtained”.
3 Validation of the Measurement Procedure – Relevant

Equations and Measurement Data
The validation of the measurement procedure involved the following studies and
the evaluation of the linearity of the spectrophotometer response (not included in
this document): (Tables 1, 2, 3, 4 and 5).
Table 1 Blank signals obtained in repeatability conditions (absorbance units, a.u.)

0.002 0.001 0.000 −0.001 0.002
0.002 0.003 0.000 0.000 0.001
Mean value: 0.001 a.u.
Standard deviation: 0.001247 a.u.
Table 2 Results from the analysis of a 2 µg P calibrator (µg P units) obtained under within-lab
reproducibility conditions (i.e. intermediate precision conditions)
1.8 2.1 1.9 2.3 1.9
1.9 2.0 2.1 2.1 2.0
1.7 2.2 2.2 – –
Mean value: 2.015 µg P
Standard deviation: 0.1725 µg P
Variance: 0.02974 (µg2 P)
Table 3 Impact of the temperature on the absorbance of 10 µg P calibrator (results obtained

under repeatability conditions)
Temperature (°C) 20 22 24 26 28 30 32
Absorbance (a.u.) 0.105 0.107 0.109 0.112 0.113 0.116 0.119
Mean absorbance increment (°C-1): 0.0117
Table 4 Analysis of a certified reference material

Observed analyte recovery (%)
106.7 110.0 103.4 108.4 100.3
108.7 100.7 101.2 95.6 102.1
100.1 91.7 109.1 97.9 97.1
110.3 97.0 103.8 98.3 –
Result from the analysis of a certified reference material (CRM)
CRM producer: LGC standards
Reference: RTC(QCI-042-1)
Matrix: Water (spiked water)
Certified phosphate value: (0.7410 ± 0.0076) mg L−1 of P (gravimetric value) (k = 2)
Mean value: 102.23%
Standard deviation: 5.43%
Table 5 Dispersion of duplicate results of unknown samples obtained under repeatability

conditions
Result (µg P)
Sample Code Replicate 1 Replicate 2 Mean Difference
434/15 37.3 37.1 37.20 0.2
435/15 23.0 22.7 22.85 0.3
436/15 37 37 37 0
501/15 27.2 27.1 27.15 0.1
516/15 23.5 23.4 23.45 0.1
735/15 37.3 37.1 37.20 0.2
799/15 40.9 41.0 40.95 −0.1
Mean 32.3 32.2 32.2 0.114
Standard Deviation 7.5 7.6 7.5 0.134
4 Measurement Uncertainty of the Result: Relevant

4.1 Equations
1. Interpolation standard uncertainty, uInt, estimated by the linear least squares

model (applicable to homoscedastic data; simplified model)
rffiffiffiffiffiffiffiffiffiffiffi
sy=x 1
uInt ¼ þ1
b n
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uP
u ðyi ^yi Þ2
t i
sy=x ¼
n2
sy/x Residual standard deviation of the calibration curve;

b Slope of the calibration curve;
n Number of calibrators readings;
yi Signal of calibrator reading i with RP mass xi;
^yi Signal of calibrator with RP mass xi estimated by the regression line.
2. Calibrators value relative uncertainty
uðfStd Þ uSt2 0:0332lg P

¼ ¼ ¼ 0:0166
fStd 2 2lg P
3. Sample volume uncertainty
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s ffi
tol 2 V DT aH2 O 2
uð V Þ ¼ pffiffiffi þ ðsr Þ þ 2
pffiffiffi
3 3
tol Tolerance associated with pipette nominal volume (0.08 mL);

sr Repeatability of pipette manipulation (0.014 mL);
aH20 Volume expansion coefficient for water (2.1 10−4 °C−1).
DT Temperature range of the solution centred in the reference temperature of the
volumetric material calibration (4 °C).
4. Recovery uncertainty
2 ffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
Þ
uð R sobs uðcCRM Þ 2 sR 2 uðcCRM Þ 2
pffiffiffiffi þ pffiffiffiffi þ
¼
R cobs N cCRM
¼
R N cCRM
Mean recovery observed in the analysis of the CRM;

R
Standard uncertainty associated with R;
u(R)
sobs Standard deviation of the estimated mass concentrations of the CRM;
cobs Mean of the estimated mass concentrations of the CRM;
N Number of replicated analysis of the CRM;
u(cCRM)/cCRM Relative standard uncertainty associated with the certified mass
concentration;
sR Standard deviation of the observed recoveries.
4.2 Experimental Data of Sample Analysis
Spectrophotometric signals (absorbance units, a.u.)
Calibrators (µg P in
Signal (a.u.) Repression parameters
100 mL volume)
0 0.002 Intercept (a.u.) 0.00206
2 0.023 Slope (a.u. µg-1) 0.0102
3 0.032 sy/x (a.u.) 0.00110
5 0.052
7.5 0.080
10 0.103 Linear correlation parameter
R(Pearson) 0.99966
Analysed Sample signal (a.u.)

sample 0.067
Exercise 1: Establishing Traceability in Analytical

Chemistry
1. Specifying the analyte and measurand
Analyte Reactive phosphorus
Measurand Mass concentration of reactive phosphorus, estimated by measurement
procedure SMEWW 4500-P D [1], in Tagus river GPS coordinates N39º 4′
0.26, W8º 45′ 44.44 at 0.2 m depth and on 12 August 2018 at 8h20
Units mg L−1 of P2O5
2. How would you demonstrate traceability of your result?

1 By analysing an adequate Certified Reference Material (CRM) and correcting
measurement result of unknown samples for observed analyte recovery. In this case,
measurement result would be traceable to the value embodied in the CRM
2 Through the accurate application of the measurement procedure including the use of
calibrated equipment and chemical references traceable to adequate references. In this
case, measurement result would be traceable to the value defined by the operationally
defined measurement procedure SMEWW 4500-P D
3
3. Any other comments, questions …

.
.
.
.
.
Exercise 2: The Customer’s Requirements Concerning

Quality of the Measurement Result
1. Specify the scope
Matrix Natural fresh waters
Measuring range 0.02 mg L−1 to 1 mg L−1 of P2O5
2. Does the analytical procedure fulfil the requirement(s) for the intended
use?
Intended use of the To check compliance of water with limits set by Directive 75/440/EEC
results for water intended for the abstraction of drinking water. The
measurement performance requirements are laid down in Directive 79/
869/EEC
Parameters to be Value requested by the Value obtained during validation
evaluate customer
LOD LOD 0.02 mg L−1 LOD = (3sy/x/b)/V = 0.0032 mg L−1 of P;
of P2O5 0.0032 2.29 = 0.0073 mg L−1 of P2O5
(where sy/x, b and V are the residual standard
deviation and the slope of a least squares
calibration, and sample aliquot volume
respectively; 2.29 = M(P2O5)/(2 M(P))
LOQ Not specified(a)
(continued)
(continued)
Repeatability 2 standard Instrumental signal repeatability:
deviation 0.04 mg 2 (sy/x/y) (m/V) 2.29 =
L−1 of P2O(b)
5 ¼ 2 (0.00110/0.103) (10/100) 2.29 =
¼ 0.0049 mg L−1 of P2O5; (where y is the
signal estimated for the largest mass of the
calibration interval, m, 10 µg of P which
corresponds to 0.23 mg L−1 of P2O5)
or
Measurement repeatability estimated from the
standard deviation of the difference, sd, of
duplicate sample results:
2 (0.00134 mg L−1 of P/sqrt(2)) 2.29 =
¼ 0.0043 mg L−1 of P2O5
(Table 5 of yellow pages)
Within-lab 2 standard 2 0.001725 2.29 = 0.0079 mg L−1 of
reproducibility deviation 0.04 mg P2O5
L−1 of P2O(b)
5 (estimated for a mass concentration of
0.046 mg L−1 of P2O(c) 5 ; Yellow pages:
Table 2)
(a)
Since LOQ = LOD/0.3, acceptable LOD would result in an acceptable LOQ
(b)
According to Directive 79/869/EEC “the range within which 95% of the results of measurements
made on a single sample, using the same method, are located”, defined as precision, shouldn’t be
larger than 10% of the limit of P2O5 in the water set in Directive 75/440/EEC: 0.4 mg L−1
(10% 0.4 mg L−1 = 0.04 mg L−1 of P2O5). Assuming the normal distribution of results, the half
range of results specified in the Directive is estimated by two times the standard deviation of
measurements precision. The Directive does not specify precision conditions. Therefore, this limit
is used for estimated repeatability and within-lab reproducibility (i.e. intermediate precision)
standard deviations
(c)
Ideally this information should be estimated for 0.4 mg L−1 of P2O5
(…)
Parameters to be Value requested by the Value obtained during validation

evaluate customer
Trueness Absolute mean The mean analyte recovery is 102.23%
error 0.08 mg L−1 (Table 4 of Yellow pages) which corresponds
of P2O(d)
5 to a mean error of 0.0089 mg L−1 at
0.4 mg L−1 of P2O5 (|102.23–100|/
100 0.4 = 0.0089 mg L−1 of P2O5)
Measurement Expanded uncertainty, U = 0.025 mg L−1 for 0.40 mg L−1 of P2O5
uncertainty U 0.10 mg L−1(e)
Other-state
The analytical procedure is fit for the intended use:
Yes No
According to Directive 79/869/EEC “the difference between the true value of the parameter
(d)
examined and the average experimental value obtained”, defined as accuracy, shouldn’t be larger
than 20% of the limit of P2O5 in the water set in Directive 75/440/EEC: 0.4 mg L−1
(20% 0.4 mg L−1 = 0.08 mg L−1 of P2O5). In this Directive accuracy term is used for the
absolute mean error
(e)
Assuming random and systematic effects affecting measurement results are adequately quantified
by the intermediate precision and mean analyte recovery respectively, the target (maximum
admissible) uncertainty is 0.10 mg L−1 of P2O5. The following equations are used to combine the
target values for precision and trueness to estimate the target expanded uncertainty, Utg [2]
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
0:042 0:082
U ¼ ku ¼ 2u ¼ 2
tg tg tg
2
p ffiffi
þ 3 ¼ 0:10 mg L1 of P2 O5
where utg and k are the target standard uncertainty and coverage factor respectively. The first term
within the square root is the square of the intermediate precision standard deviation and the second
term is the square of the bias standard uncertainty assuming mean error has a rectangular
distribution. This equation assumes that the reference value of recovery tests is affected by a
negligible uncertainty
Exercise 3: Enlarge the Analytical Method Scope

1. Specify the additional validation work needed to enlarge the scope of
analysis to the analysis of wastewaters
Intended use of the results: to estimate eutrophication risk of a river by wastewater discharge.
The reactive phosphorus provides an estimation of the most readily bioavailable phosphorus
Parameters to be evaluted Value requested by the customer
(relative to the request presented on Exercise 2) (“=” –
Same; “<” – Smaller value; “>” – Larger value)
LOD Maximum LOD “>” (0.02 mg L−1 of P2O5)
LOQ See LOD
Selectivity Check if known interferences can be present in analysed
wastewater
Linearity “=” (the same for wastewaters analysed in the same
calibration range)
Homogeneity of variances of “=” (the same for wastewaters analysed in the same
instrumental response calibration range)
Repeatability Maximum repeatability standard deviation “>” 0.02 mg L−1
of P2O5 (0.04/2)
Within-lab reproducibility Maximum within-lab reproducibility standard deviation
“>” 0.02 mg L−1 of P2O5 (0.04/2)
Reproducibility (between Lab) Not to be included in in-house validation study
Trueness Maximum absolute mean error “>” 0.08 mg L−1 of P2O5
Robustness Not particularly relevant if within-lab reproducibility
(i.e. intermediate precision) is studied in significantly
different experimental and operational conditions and
measurement procedure is not to be transferred to
another location
(continued)
(continued)
Participation in PT schemes The z-score should be estimated with a reference
standard deviation (“>”) 0.05 mg L−1 of P2O5 (half of
the target expanded uncertainty). Satisfactory z-score are
within the interval [−2, 2]
Measurement uncertainty Target expanded uncertainty “>” 0.10 mg L−1 of P2O5
Other-state
2. State how should be validated the selectivity
Exercise 4: Building an Uncertainty Budget

1. Specify the measurand and units
Measurand Mass concentration of reactive phosphorus, estimated by measurement
procedure SMEWW 4500-P D, in Tagus river GPS coordinates N39º4′0.26,
W8º45′44.44 at 0.2 m depth and on 12 August 2018 at 8h20.
Unit mg L−1 of P2O5
2. Provide the model equation used to evaluate the measurement uncertainty

Measurement model:
mðlg P) MðP2 O5 Þ
cðmg L1 Þ ¼
VðmL) 2MðPÞ
c(mg L−1) Reactive phosphorus (RP) mass concentration in the sample expressed as
P2O5;
m(µg P) RP mass interpolated in the calibration curve expressed as P;
V(mL) Sample volumetric aliquot;
M(P2O5) Molar mass of phosphorus pentoxide;
M(P) Molar mass of phosphorus.
3. Identify (all possible) sources of uncertainty
RP mass interpolated in the calibration curve, m(µg P) (statistical interpolation and
calibrator mass uncertainty components)
sample volumetric aliquot, V(mL)
and CRM certified value uncertainty components)
Recovery (mean analyte recovery, R,
Dilution of sample volumetric aliquot by acid addition, fdil (negligible component for most
samples)
Molar mass of phosphorus pentoxide (g mol−1) (negligible component)
Molar mass of phosphorus (g mol−1) (negligible component)
3.1. Build cause/effect diagram
Updated measurement model:
mðlg P) fstd MðP2 O5 Þ 1

cðmg L1 Þ ¼
VðmL) 2MðPÞ R
mean analyte recovery;

R
fstd unitary factor for accounting for calibrator mass uncertainty.
4. Evaluate values of each input quantity
Input Value Unit Remark
quantity
m 6.39 µg (0.067 a.u. − 0.00206 a.u.)/(0.0102 a.u. lg−1)
V 100 mL –

R 1 – is set equal to 1 since analyte recovery is metrologically
R
equivalent to 100%(a)
fstd 1 – –
M(P2O5) 141.94 g mol−1 negligible uncertainty component
M(P) 30.97 g mol−1 negligible uncertainty component
(a) Þ t(N1; 0:05Þ ¼ t(18; 0:05Þ , 1:63 2:10 (see following
Significance test: ðjR 1jÞ=uðR
section)
5. Evaluate the standard uncertainty of each input quantity

Input Standard Unit Remark
quantity uncertainty
m 0.117 µg This uncertainty component represents the statistical
interpolation of sample signal in the calibration curve(a)
V 0.0684 mL See following calculations

R 0.0135 – See following calculations
(continued)
(continued)
fstd 0.0166 – The standard uncertainty associated with fstd, u(fstd), is
estimated from the RP mass relative standard uncertainty
of the calibrator with lowest quantity except the blank,
uStd2/2 (i.e. u(fstd) = fstd uStd2/2) [3]
M(P2O5) negligible g mol−1
M(P) negligible g mol−1
(a)
The estimated mass of RP in the diluted sample is affected by both the statistical interpolation of
sample signal in the calibration curve and by calibrator mass uncertainty. The ratio of the masses
of RP of any pair of calibrators should have a negligible uncertainty given the instrumental signal
precision to allow the reliable estimation of the interpolation uncertainty from least squares
regression model [2]
Calculations:
fstd:
uðfStd Þ uStd2 0:0332lg P

¼ ¼ ¼ 0:0166 and, since fstd = 1, u(fstd) = 0.0166.
fStd 2 2lg P

R:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2ffi
Þ ¼ R
s R u ð c Þ
uð R pffiffiffiffi þ CRM
N
R cCRM
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 0:0076 2
5:43
¼ 102:2 pffiffiffiffiffi þ 2
¼ 1:35%½0:0135
102:2 19 0:7410
V:
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s ffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
tol 2 V 4 aH2 O 2 0:08 2 100 4 0:00021 2
uðV Þ ¼ pffiffiffi þ ðsr Þ þ 2
pffiffiffi ¼ pffiffiffi þ ð0:014Þ þ 2
pffiffiffi
3 3 3 3
¼ 0:0684
tol Tolerance associated with the pipette nominal volume (0.08 mL);
sr repeatability of pipette manipulation (0.014 mL);
aH20 volume expansion coefficient for water (2.1 10−4 °C−1).
m:
rffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffi
sy=x 1 0:00110 1
uInt ¼ þ1 ¼ þ 1 ¼ 0:117lg P
b n 0:0102 6
6. Calculate the value of the measurand, using the model equation
mðlg P) fstd MðP2 O5 Þ 1

cðmg L1 Þ ¼
VðmL) 2MðPÞ R
0:0670:00206
1 141:94 1 6:39
¼ 0:0102
¼ 2:291 ¼ 0:14590 mg L1 of P2 O5
100 2 30:97 1 100
7. Calculate the combined standard uncertainty (uc) of the result and specify
units
Using: Mathematical solution; Spreadsheet Approach; Commercial
Software
Input quantity Value Standard uncertainty Unit Remark

m 6.39 0.117 µg
V 100 0.0684 mL

R 1 0.0135 –
fstd 1 0.0166 –
M(P2O5) 141.94 – g mol−1
M(P) 30.97 – g mol−1
Calculations:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 2 uðV Þ2 uðR Þ 2

uðfStd Þ 2

Int
uðcÞ ¼ c þ þ þ
m V R fStd
s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2 ffi
0:117 0:0684 0:0135 2 0:0166 2
uðcÞ ¼ 0:14590 þ þ þ
6:39 100 1 1
uðcÞ ¼ 0:0041 mg L1 of P2 O5
8. Calculate the expanded uncertainty (Uc) & specify the coverage factor k
and the units
Uc ðcÞ ¼ 0:0082 mg L1 of P2 O5
For a confidence level of approximately 95% considering a coverage factor, k, of 2.

Result to be reported: (0.1459 ± 0.0082) mg L−1 of P2O5 (for k = 2 and 95%)
9. Analyse the uncertainty contribution & specify the main input quantity
contributing the most to Uc
1 Statistical interpolation component: 42.2%
2 Calibrator mass component: 34.7%
3 Analyte recovery component: 23.0%
4 Sample aliquot: 0.06%
5
6
Calculations:
Example: Percentage contribution, PInt, of the statistical interpolation uncertainty to
the combined standard uncertainty:
u 2
Int
¼ m
PInt
uInt 2 uðfStd Þ 2 uðV Þ 2 uR 2
þ þ þ
m fStd V R
ð0:0183Þ2
PInt ¼
ð0:0183Þ þ ð0:0166Þ2 þ ð0:000684Þ2 þ ð0:0135Þ2
2
PInt ¼ 0:422½42:2%
Graphic representation of the percentage contribution of the uncertainty components:

10. Prepare your Uncertainty Budget Report

(…)
Conclusion: Measurement is fit for its intended use since the reported expanded
uncertainty (0.0082 mg L−1 of P2O5) is smaller than the target expanded uncer-
tainty (0.1 mg L−1 of P2O5).
References
1. APHA, AWWA, WEF, Standard Methods for the Examination of Water and Wastewater, 22th.
Ed., USA, 2012.
2. R. J. N. B. Silva, Water 5 (2013) 1279–1302.
3. R. J. N. B. Silva, M. F. G. F. C. Camões, Anal. Lett. 43 (2010) 1257–1266.
Ricardo Bettencourt da Silva completed his B.Sc in chemistry

at the Faculty of Sciences of the University of Lisbon (FCUL),
his M.Sc in bromatology at the Faculty of Pharmacy of the
University of Lisbon, and his Ph.D. in analytical chemistry—
metrology in chemistry at FCUL. The last two academic degrees
were completed in parallel with his full-time professional expe-
rience as analyst, in official, public and private laboratories, for
the analysis of different inorganic and organic analytes in various
types of matrices using classical and instrumental methods of
analysis.
This analytical experience was focused on the detailed vali-
dation of the measurement procedure, test quality control and
evaluation of measurement uncertainty. Since 2002, Ricardo has
worked regularly as an assessor of the Portuguese Accreditation
Body (IPAC) and as a trainer and consultant for the accreditation
of chemical laboratories. In 2009, Ricardo was contracted as a researcher by the Centre for
Molecular Sciences and Materials of the Faculty of Sciences of the University of Lisbon (recently
‘Centro de Química Estrutural’) where he has been continuing his research work on metrology in
chemistry while collaborating in teaching at national and foreign universities. Ricardo’s research
includes the development of approaches for the detailed evaluation of the uncertainty associated
with complex measurements and the assessment of the sources of lack of comparability of mea-
surements in some analytical fields. Ricardo has been a member of the IPAC Accreditation of
Chemical Laboratories Working Group since 2006, the Eurachem/CITAC Measurement
Uncertainty and Traceability Working Group since 2010, the Portuguese TrainMiC® team since
2008 and the TrainMiC®.Editorial Board since 2010.
Maria A. Trancoso started her career in 1980 working in ana-

lytical chemistry applied to environmental geochemistry. She
gained her Ph.D. in Chemistry at the Technical University of
Lisbon.
Maria Trancoso has been the head of accredited laboratories
since 2001 and quality manager from 1990. She has a large
experience in laboratory management and the training of staff
members concerning quality assurance and quality control in
analytical chemistry.
Maria Trancoso is currently working as a senior research
scientist at Bioenergy Unit of the National Laboratory for Energy
and Geology and is head of the Biofuel and Environmental
laboratory. Her main interests are related to the development and
validation of analytical procedures including traceability and
measurement uncertainty evaluation in the fied of environmental
science and materials for energy. She is involved in the development and organization of external
proficiency testing in wastewater and sludge contracted with RELACRE.
Maria A. Trancoso is author of several scientific papers and is a trainer on issues related to
laboratory accreditation and metrology in chemistry. She is actively collaborates with the
Portuguese Accreditation Institute (IPAC) as technical assessor.
In 2007 joined to TrainMic® Programme as a member of the Portuguese team of authorised
TrainMIC® trainers.
Paula Teixeira completed her Master Degree in Applied

Analytical Chemistry in 2004 at Faculty of Sciences, University
of Lisbon and her Ph.D. in Energy and Bioenergy at Faculty of
Science and Technology, New University of Lisbon in 2012.
Since 2016, develop research activities at ‘Instituto Superior
Técnico’ of ‘Universidade de Lisboa’ (‘Centro de Química
Estrutural’) and between 2001 and 2015 developed research
activities at ‘Laboratório Nacional de Energia e Geologia’. She
is author or co-author of scientific papers, reports and conference
proceedings. The main research areas include: (1) Ca-looping,
(2) CO2 capture and storage, (3) sorbents characterization: tex-
tural, morphological and mineralogical properties,
(4) physic-chemical characterization of fuels and ashes, (5) study
of slagging, fouling and ash agglomeration problems during the
thermochemical conversion of several solid fuels, (6) metrology
in chemistry, (7) environmental issues associated with the fuels utilization and thermochemical
conversion.
She held several training courses in validation of analytical methods, statistics for laboratories,
measurement uncertainty in physical chemical analysis and sampling uncertainty.
Since 2007 Paula Teixeira has been an authorized TrainMiC® trainer and member of the
Portuguese TrainMiC® team.
Cristina Oliveira received her master degree and her Ph.D.

degree at the Faculty of Sciences of the University of Lisbon
where she works as Researcher since 1989. Her Ph.D. was in the
field of Ion Selective Electrodes (ISE), particularly potassium ion
selective electrodes. Since then her scientific research work has
been carried out in Analytical Chemistry, at the Centre for
Molecular Sciences and Materials of the Faculty of Sciences of
the University of Lisbon (recently ‘Centro de Química
Estrutural’). Her research is mainly in the fields of analysis of
atmospheric aerosols by Ion Chromatography as well as the
determination of metals in soils, leaves and vegetables by Atomic
Absorption Spectrometry with Graphite Furnace. For the last few
years her research work also involves the metrology in chemistry.
Additionally Cristina has been collaborating in teaching at the
University and acting as a mentor for undergraduate and post-
graduate students. Integrating the Portuguese TrainMiC® team in 2007 she became the National
TrainMiC® team leader for Portugal in 2011.
Alice Mosca completed her B.Sc in Agricultural Engineering at

the University of Evora (UE) and her M.Sc in Chemistry applied
to the Cultural Heritage at the Faculty of Science of the
University of Lisbon (FCUL).
She worked as Analytical Manager in a private accredited
laboratory for ten years (1994–2014), and was responsible for
different instrumental methods of analysis (GC, HPLC, FTIR,
among others). Her responsibilities included the implementation,
validation and definition of the quality control schemes of the
methods. She was also co-responsible for the implementation and
maintenance of both management and technical requirements of
the EN ISO/IEC 17025.
She started working as an independent consultant for the
accreditation of laboratories, as a trainer in different technical and
management areas, as an internal auditor (for the management
and technical requirements of the EN ISO/IEC 17025 standard) and as an assessor of the
Portuguese Accreditation Body (IPAC) in 2004. She founded her own company (AIM –
Consultancy, training and audits Lda) in 2010 were she still works as General Manager.
She has been collaborating in teaching at national and foreign universities for the last 20 years
and became a part of the Portuguese TrainMiC® team since 2006.
Florbela Aura do Sacramento Dias completed her graduation

in Chemical Engineering at the Faculty of Sciences and
Technology of the New University of Lisbon (FCT-UNL), her
Ph.D. in Chemistry—Chemical-Physics at the same University,
and her Post-Doc in the field of Chemistry—Thermodynamics at
Faculty of Sciences of the University of Lisbon (FC-UL). The
last two years of Ph.D. were completed in parallel with her
partial-time professional experience as professor of chemistry in
the FCT-UNL.
Since 2002, Florbela Dias has worked in full-time in
Portuguese Institute for Quality as Head of Amount of Substance
and Electrochemistry Area. Since 2002 she has been Contact
person at EURAMET/METCHEM Gas Working Group and at
the Consultative Committee for Amount of Substance
(CCQM-BIPM). Since 2005, Florbela Dias has worked regularly
as an assessor of the Portuguese Accreditation Body (IPAC), of the Israel Accreditation Body
(ISRAC) and of the Turkish Accreditation Body (TURKAK). Since 2007 Florbela Dias has been
in Portuguese TrainMiC team and since 2011 has been process manager and GLP inspector in the
scope of GLP principles of OECD.
Maria Filomena Gomes Ferreira Crujo Camões graduated in

Physics and Chemistry from the Faculty of Sciences of the
University of Lisbon (FCUL) in 1966, completed her Ph.D. in
Physical Chemistry at the University of Newcastle-upon-Tyne—
UK, in 1973, while on leave from FCUL where she was a
Chemistry Teaching Assistant. From 1974 she was promoted to
Assistant Professor, Associate Professor and Professora
Associada com Agregação, after her Habiltation in Analytical
Chemistry, in 1993. In this capacity she is also coordinator of the
Research Group on Analytical and Environmental Chemistry—
CQE@FCUL. Examples of recent Projects have been: PAHLis—
PTDC/AMB/65699/2006, GEOTRACES- National Contact;
IAPWS-Task Group on “pH of Seawater” of the Subcommittee
on Seawater (Giles Marion-chair, WG127); COST Action
ES0801-The ocean chemistry of bioactive trace elements and
paleoclimate proxies; Env05-Metrology Acidity and salinity—EMRP- REG2, which gave rise to
11 Ph.D. Thesis, 20 M.Sc. Thesis, more than 50 Graduation Projects, 200 scientific publications in
national and international (SCI) journals and more than 400 lectures and presentations at scientific
meetings.
She was nominated for various national and international functions, namely (a) President of the
IUPAC Division Analytical Chemistry (2012–2013) while Chair of the Subcommittee on pH
(2010–….) and Member of the WG on “pH Standards and Recommendations” (1998–2002);
(b) EURACHEM President (2002–2004), Delegate to EA/European Accreditation (2004–2006)
and National Delegate (1991–2002); (c) President of the Division of Analytical Chemistry of the
Portuguese Chemical Society (1999–2001); VP (1997–1999); PP (2001–2003); (d) Coordinator of
the EuroMaster® on Measurement Science in Chemistry; e) Ambassadeur and Member of
TrainMiC® Advisory Board, (f) Advisor to the Minister of Education on National Examinations
(Chemistry and Environmental Sciences), 1996–2010.
She was in charge of the organisation of the 37th IUPAC General Assembly (1993) and chaired
the organization of EUROANALYSIS XI (2000).
Measurement of Concentration
of Nitrate in Drinking Water
Andreja Drolc
1 Description of the Analytical Procedure
For nitrate we have here below text extracts of an ISO method and also the
requirements in the directive for nitrate.
International Standard EN ISO 10304-1 Water quality – Determination of dis-
solved anions by liquid chromatography of ions – Determination of bromide,
chloride fluoride, nitrate, nitrite, phosphate and sulphate.
1.1 Scope
1.1.1 General
The method is specified for the determination of nitrate in water (drinking water,
ground water, surface water, waste water, leachates and marine water) in the fol-
lowing range: <0.1 mg L−1 nitrate. The lower limit of application depends on the
matrix and the interferences encountered.
In certain cases, the range of application may be changed by variations in the
working conditions (e.g. sample volume, dilution, separating columns, pre-
concentration techniques, sensitivity ranges of detectors, etc.).

A. Drolc (&)
National Institute of Chemistry, Hajdrihova 19, 1000 Ljubljana, Slovenia
e-mail: andreja.drolc@ki.si

https://doi.org/10.1007/978-3-030-20347-4_2
22 A. Drolc
1.1.2 Interferences
Some organic acids (malonic acid, maleic acid, malic acid) may interfere in the
determination of inorganic anions if they are present in high concentration. Sulfite
may cause positive bias for sulphate due to autooxidation. In this case the sample
may be adjusted to pH 10 and formaldehyde solution is added in order to stabilize
sulfite, if necessary.
In the presence of metals possibly interfering with the anions of interest, should
be checked and eliminated with the aid of special cation exchangers.
Cross sensitivity (lack of resolution) may occur in the case of large differences in
concentration between the anions determined
Ratio of the mass concentrations Maximum tolerable absolute

solute/interfering ion concentration of interfering ions, mg L−1
NO3/Cl− 1:500 Cl− 500
NO3/Br− 1:100 Br− 100
NO3/SO42− 1:500 SO42− 500
NO3/SO32− 1:50 SO32− Can always interfere
1.2 Principle
Liquid chromatographical separation of ions by means of a separating column. Use

of a low capacity anion exchanger as the stationary phase, and usually aqueous
solutions of salts of weak monobasic and dibasic acids as mobile phases (eluent).
Detection by a conductivity detector, sometimes employed in combination with a
suppressor device (e.g. cation exchanger) which decreases the conductivity of the
eluent and converts the separated anions into their corresponding acids.
1.3 Reagents
1.3.1 Eluents
Different eluents are used, their choice depending on the type of separating column
and detectors (follow column manufacturer’s instructions).
1.3.2 Water
Deionised water with electrical conductivity of <0.1 lS/cm and shall be free of
particulate matter with diameter >0.45 lm.
Measurement of Concentration of Nitrate in Drinking Water 23
1.3.3 CRM – Nitrate in Water
NIST Standard Reference Material 3185.
1.4 Apparatus
Usual laboratory equipment (graduated flasks, pipettes, filtering apparatus, weigh-

ing equipment) and ion chromatographic system, consisting of the following
components:
– eluent reservoir,
– pump having a very low pulsation effect,
– sample injection system,
– pre-column,
– separating column with the required separating performance,
– conductivity detector with or without suppressor device,
– recording device.
1.5 Quality Requirements of the Separating Column
The peak resolution shall not fall below R = 1.3
R ¼ 2ðtR2 tR1 Þ=ðW1 þ W2 Þ
where
tR1 is the retention time, in seconds, of the 1st peak
tR2 is the retention time, in seconds, of the 2nd peak
W1 is the peak width, in seconds, of the 1st peak
W2 is the peak width, in seconds, of the 2nd peak.
1.6 Sampling and Samples
Use vessels made of PTFE or polyethylene for sampling. After arrival of the sample
in the laboratory, filter it through a membrane filter (of pore size 0.45 lm). Stabilize
the sample by cooling it to between 4 and 6 °C or by deep freezing until the
analysis is performed.
24 A. Drolc
1.7 Procedure
Set up the ion chromatograph according to manufacturer’s instructions.

Identify the anions by comparing the retention times with those of the standard
solutions. Retention times may be dependent on concentration and matrix. Different
working ranges can be applied, e.c. 0.1–1 mg L−1 NO3, 1–10 mg L−1 NO3.
The area (or height) of the peak (signal) is proportional to the concentration of
the anion.
1.8 Calculation
When the analytical system is first evaluated, and at intervals afterwards, establish a
calibration function for the measurement according to ISO 8466-1 (Water Quality –
Calibration and evaluation of analytical methods and estimation of performance
characteristics).
Estimate the mass concentration, in milligrams per litre, of the nitrate in the
solution using the peak areas or peak heights as follows:
A B0
C¼
B1
where A, B1 and B0 are measured area of the sample chromatographic peak, slope of
the linear least square calibration curve and calculated blank.
Slope of the linear least square calibration curve B1 and calculated blank B0 are
calculated from equations:
P
n

ðAi AÞ
ðCi CÞ
i¼1
B1 ¼ P
n
2
ðCi CÞ
i¼1
B1 C
B0 ¼ A
X
n
¼1
A Ai
n i¼1
X
n
¼1
C Ci
n i¼1
where Ci and Ai are concentration of reference solution on ith level (C1, …, Ci, …, Cn)
and areas of chromatographic peaks of ith reference solution (A1, …, Ai, …, An),
respectively.
Take into account all the dilution steps.
Report the results to a minimum of three significant figures.

Extract from the EU Water Directive1

The parametric value (max limit) for nitrate in drinking water is 50 mg L−1.
The requirements for the analyses are the following:
Parameter Trueness % of Precision % of Limit of detection % of

parametric value parametric value parametric value (Note 3)
(Note 1) (Note 2)
NO3 10 10 10
Note 1 (*): Trueness is the systematic error and is the difference between the mean value of the
large number of repeated measurements and the true value
Note 2 (*): Precision is the random error and is usually expressed as the standard deviation (within
and between batch) of the spread of results about the mean. Acceptable precision is twice the
relative standard deviation. (*) These terms are further defined in ISO 5725
Note 3: Limit of detection is either:—three times the relative within batch standard deviation of a
natural sample containing a low concentration of the parameter, or—five times the relative within
batch standard deviation of a blank sample

In the present case study, methodology for validation of measurement procedure

EN ISO 10304-1 (nitrate in water by liquid chromatograhy) is presented. For the
calculation part, the emphasis is on the parameters that are required by the cus-
tomer. In this particular case, these parameters are:
– LOD
– within-laboratory reproducibilty
– trueness.
1
98/83/EC, of 3 November 1998, on the quality of water intended for human consumption.
26 A. Drolc
Measurements and intermediate calculations
5.41
5.31
5.37
5.41
B0 14,967
B1 112,837
Abl 5104
5216
5072
Cobserved 4.56 mg L−1
Cmatrix 2.32 mg L−1
Cspiked 2.22 mg L−1
Equations
P
n
xi
i¼1
x ¼
n
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
ðxi xÞ2
s¼
n1
ALOD ¼ Abl þ 3sbl
ALOD b0
LOD ¼
b1
cobserved cmatrix
R¼
cspiked
4 Measurement Uncertainty of the Result – Relevant

Equations and Measurement Data2
In the present case study, methodology for evaluation of measurement uncertainty

of result of nitrate ion determination in drinking water is presented. Nitrate ion was
determined using ion chromatography (EN ISO 10304-1, 1996). The necessary
relevant information was obtained from the method validation data, the quality
2
Here you may also want to include the relevant certificates.
control data and equipment calibration certificates. The method of measurement is

described together with the measurement equation, selected traceable reference
standards and the associated measurement uncertainty. The major sources of
uncertainty of the result of measurement were identified and the combined uncer-
tainty was calculated. Identification of the main uncertainty sources represent basis
for target operation for reducing the measurement uncertainty of this determination.
Experimental protocol
pre-analytical phase
Collection of
water sample
from a tap
Preparation and
storage of sample
analytical phase
Separation:
liquid
chromatography
Preparation of
calibration
Conductometric Conductometric
detection detection
Result
28 A. Drolc
Measurement data and intermediate results
Quantity Unit Ref. Value std-unc RSu

Standards C1 mg L−1 0.488 0.002 0.39%
C2 mg L−1 0.976 0.004 0.36%
C3 mg L−1 1.952 0.007 0.36%
C4 mg L−1 2.928 0.011 0.38%
C5 mg L−1 3.904 0.014 0.36%
C6 mg L−1 4.880 0.019 0.39%
C7 mg L−1 5.856 0.021 0.36%
C8 mg L−1 6.832 0.025 0.37%
Areas A1 – 51,437 76 0.15%
A2 – 98,583 204 0.21%
A3 – 200,037 151 0.08%
A4 – 305,822 312 0.10%
A5 – 418,793 315 0.08%
A6 – 531,318 712 0.13%
A7 – 645,226 716 0.11%
A8 – 767,727 295 0.04%
A – 594,350 4220 0.71%
Recovery R – 1 0.01 1%
Peak area, average Aaver – 377,367 146
Concentration, average Caver mg L−1 3.477 0.016
B1 – 112,837 831
Equations
A B0
C¼
B1
P
n

ðAi AÞ
ðCi CÞ
i¼1
B1 ¼ P
n
2
ðCi CÞ
i¼1
X
n
¼1
A B1 C
Ai B0 ¼ A
n i¼1
X
n
¼1
C Ci
n i¼1
A B0 cobserved cmatrix
CNO3 ¼ R¼
B1 R cspiked
Explanation of all the input quantities mentioned in model equation

C concentration of nitrate (mg L−1)
A measured area of the chormatographic peak for sample
B0 calculated blank
B1 slope of the linear least square calibration curve
Ci concentration of calibration solutions (mg L−1)
Ai areas of chromatographic peaks for the calibration solutions
CNO3 final result for concentration of nitrate (mg L−1)
R recovery
Cobserved concentration of the analyte in the spiked sample
Cspiked concentration of the spike added to the sample
Cmatrix concentration of the analyte in unspiked sample.
Equations for measurement uncertainty calculation

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðB0 Þ ¼ 2 þ ðuðB1 CÞÞ
ðuðAÞÞ 2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Þ
uðB1 C uðB1 Þ 2 uðCÞ 2
¼
B1 C B1
þ
C
s ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

uð c Þ uðB1 Þ 2 uðRÞ 2 uðA B0 Þ 2
¼ þ þ
c B1 R A B0
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðA B0 Þ ¼ ðuðAÞÞ2 þ ðuðB0 ÞÞ2
30 A. Drolc

Chemistry

Analyte Nitrate
Measurand Nitrate in drinking water, rain water, ground water and surface water
Units mg L−1
2. Choosing a suitable measurement procedure with associated model

equation
Measurement procedure EN ISO 10304-1 Water quality – Water quality – Determination
of dissolved anions by liquid chromatography of ions – Determination of
bromide, chloride fluoride, nitrate, nitrite, phosphate
and sulphate.
Type of calibration Standard curve Standard addition Internal standard
Model equation
The concentration of nitrate, C, is calculated by:
A B0
C¼
B1
where A, B1 and B0 are measured area of the sample chromatographic peak, slope of
the linear least square calibration curve and calculated blank. Slope of the linear
least square calibration curve B1 and calculated blank B0 are calculated from
equations:
P
n

ðAi AÞ
ðCi CÞ
i¼1
B1 ¼
P
n
2
ðCi CÞ
i¼1
B1 C
B0 ¼ A
X
n
¼1
A Ai
n i¼1
X
n
¼1
C Ci
n i¼1
respectively.
3. List the input quantities according to their influence on the uncertainty of

the result of the measurement (first the most important ones). At this point,
your judgement should be based on your previous experience only
1 Uncertainty of concentration of reference solutions
2 Uncertainty of volumes
3 Uncertainty of weighing
4 –
5 –
4. List the reference standards needed and state the information regarding
traceability of the reference value
For the analyte
1 Name/ChemicalFormula/producer: NIST Standard Reference Material 3185

‘Nitrate anion solution’
2 Name/ChemicalFormula/producer: –
For the other input quantities
1 Quantity/Equipment/calibration: Volume/calibrated glassware and

e.g. mass/balance/calibrated by NMI, U = xx pipettes/with established traceability
(k = 2), see also data yellow sheet
2 Quantity/Equipment/calibration: Mass/calibrated balance/with
established traceability
3 Quantity/Equipment/calibration:
5. Estimating uncertainty associated with the measurement

Are all important parameters included in the model equation? Yes No
Other important parameters are: Matrix effects
6. How would you prove traceability of your result?

1 Analysis of matrix CRM
2 Participation in a proficiency testing scheme (Aquacheck,…)
3 –
7. Any other comments, questions …

32 A. Drolc
Exercise 2: Single Laboratory Validation of Measurement

Procedures
Part I: General Issues

1. Specify the measurement procedure, analyte, measurand and units
The measurement Measurement procedure is based on EN ISO 10304-1 Water quality –
procedure Determination of dissolved fluoride, chloride, nitrite, orthophosphate,
bromide, nitrate and sulphate ions, using liquid chromatography of
ions – Part I: Method for water with low contamination
Analyte nitrate
The measurand Nitrate in drinking water
Unit mg L−1
2. Specify the scope

Matrix Drinking water
Measuring range >0.1 mg L−1
3. Requirement on the measurement procedure

Intended use of the results
Mark the customer’s requirements and Parameters to be Value requested by
give their values validated the customer
LOD 5 mg L−1
LOQ
Repeatability
Within-lab 5 mg L−1
reproducibility
Trueness
Measurement 5 mg L−1
uncertainty
Other-state
4. Origin of the Measurement Procedure

Validation
New In-House Method Full
Modified Validated Method Partial
Official Standard Method Confirmation/Verification
Part II: Parameters to be Validated

5. Selectivity/Interference/Recovery
Where yes, please give further information e.g. which CRM, reference method
CRM/RM: analysis of available CRM or RM

Further information: Promochem QCP 051 – Anions
Spike of pure substance
spiking of samples with pure substances and calculation of recovery
Compare with a reference method
Selectivity, interferences
Test with different matrices
Other – please specify
6. Measuring range
Linearity
Upper limit
LOD
LOQ
7. Spread – Precision
Repeatability
Reproducibility (within Lab)
Reproducibility (between Lab)
8. Robustness
Variation of parameters
9. Quality Control
Control charts
Participation in PT schemes
34 A. Drolc
10. Other parameters to be tested

Working range and testing of homogeneity of variances
R square
Residual standard deviation
Standard deviation of the analytical procedure
Coefficient of variation of the analytical procedure
Measurement uncertainty
Part III: Some Calculations and Conclusions

11. Calculation of parameters requested by the customer
Parameters requested to be Calculations
validated
LOD LOD = 0.5 mg L−1
Calculations
Abl (aver) 33,869
Sbl 915.4294
ALOD = Abl + 3Sbl 36615.29
LOD 0.5 mg L−1
LOQ
Repeatability
Within-lab reproducibilty s = 0.06 mg L−1
Calculations
c (aver) 5.40
s 0.06 mg L−1
Trueness R = 1.005
Significant test was used to show that R is not
significantly different from 1
Other-please state
use?
Parameter Value requested Value obtained The requirement
by the customer during validation is fulfilled
(the same as stated Yes/No
in question 3)
LOD 5 mg L−1 0.5 mg L−1
LOQ – –
Repeatability – –
−1
Within-lab 5 mg L 0.06 mg L−1
reproducibility
Trueness 5 mg L−1 0 (R = 1)
Measurement – –
uncertainty
Other – –

Yes No
For Measurement Uncertainty and Traceability refer to the corresponding sheets.

36 A. Drolc

Measurand Nitrate in drinking water
Unit mg L−1
2. Describe the measurement procedure and provide the associated model

equation
Measurement procedure:
Concentration of nitrate in water is measured by liquid chromatography (calibration
curve).
For the final result, recovery is taken into account.
Model equation:
A B0
CNO3 ¼
B1 R
where A, B1, B0 and R are measured area of the sample chromatographic peak,
slope of the linear least square calibration curve, calculated blank and method
recovery, respectively. Slope of the linear least square calibration curve B1 and
calculated blank B0 are calculated from eqns:
P
n

ðAi AÞ
ðCi CÞ
B1 ¼ i¼1 B1 C
B0 ¼ A
P
n
2
ðCi CÞ
i¼1
X
n X
n
¼1
A Ai ¼1
C Ci
n i¼1
n i¼1
cobserved cmatrix
R¼
cspiked
respectively.
Uncertainty of concentration of reference solutions
Uncertainty of measurements of peak area
(continued)
(continued)
Method bias
Matrix effect
Other:
Other:

Input quantity Value Unit Remark
A 594,350 –
B0 −14,967 –
B1 112,837 –
R 1 –

C 3.4770 mg L−1

A 377,367 –

Input quantity Standard uncertainty Unit Remark
A 4220 –
B0 3410 –
B1 831 –
R 0.01 –

C 0.016 mg L−1

A 146 –
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Þ
uðB1 C uðB1 Þ 2 uðCÞ 2 831 2 0:016 2
¼ þ ¼ þ
B1 C B1 C 112837 3:4770
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 5:4237 105 þ 2:1175 105 ¼ 7:5412 105
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 112837 3:4770
uðB1 CÞ 7:5412 105 ¼ 3407
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðB0 Þ ¼ uðAÞ 2 þ uðB1 CÞ 2 ¼ 1462 þ 34072 ¼ 3410
A B0
CNO3 ¼ CNO3 ¼ 5:4 mg L1
B1 R
38 A. Drolc
7. Calculate the combined standard uncertainty (uc) of the result & specify
units
Software
Input Value Standard Unit Remark

A 594,350 4220 – Included in the MU data sheet
B0 −14967 −130 – To be calculated by the participants (for
details see xls green sheet)
B1 112,837 831 – Included in the MU data sheet
R 1 0.01 – Included in the MU data sheet
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

uðcÞ uðB1 Þ 2 uðRÞ 2 uðA B0 Þ 2
¼ þ þ
c B1 R A B0
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2 2
831 0:01 5425:54
¼ þ þ
112837 1 594350 þ 14967
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 23:35235 105
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðA B0 Þ ¼ uðAÞ2 þ uðB0 Þ2 ¼ 42202 þ 34102 ¼ 5425:54
uðcÞ ¼ 0:015281475 5:4 ¼ 0:082 mg L1
u(c) = 0.082 mg L−1
8. Calculate expanded uncertainty (Uc) & specify the coverage factor k and
the units
Uc ¼ 0:2 mg L1 ðk ¼ 2Þ
9. Analyse the uncertainty contribution & specify the main three input
quantities contributing the most to Uc
1 Method recovery
2 Calibration curve
3 Measurement of the peak area of the sample

Addendum I: Measurement Uncertainty Calculation: Spreadsheet Approach (Excel)
STEP 1: Calculation of A(average)

A1= A2= A3= A4= A5= A6= A7= A8=
Values 51437 98583 200037.0 305822.0 418793 531318 645226 767727
Uncertainties 76 204 151 312 315 712 716 295
A1= 51437.0 51513 51437 51437 51437 51437 51437 51437 51437
A2= 98583.0 98583 98787 98583 98583 98583 98583 98583 98583
A3= 200037.0 200037 200037 200188 200037 200037 200037 200037 200037
A4= 305822.0 305822 305822 305822 306134 305822 305822 305822 305822
A5= 418793.0 418793 418793 418793 418793 419108 418793 418793 418793
A6= 531318.0 531318 531318 531318 531318 531318 532030 531318 531318
A7= 645226.0 645226 645226 645226 645226 645226 645226 645942 645226
A8= 767727.0 767727 767727 767727 767727 767727 767727 767727 768022
A-average= 377367.875 377377.375 377393.375 377386.75 377406.875 377407.25 377456.875 377457.375 377404.75
9.5 25.5 18.875 39 39.375 89 89.5 36.875
u(A-average)= 146.4894941 90.25 650.25 356.265625 1521 1550.390625 7921 8010.25 1359.765625
Measurement of Concentration of Nitrate in Drinking Water
39
40
STEP 2: Calculation of C(average)

C1= C2= C3= C4= C5= C6= C7= C8=
Vrednosti 0.488 0.976 1.95 2.93 3.904 4.88 5.856 6.832
Negotovosti 0.0019 0.0035 0.007 0.011 0.014 0.019 0.021 0.025
C1= 0.488 0.4899 0.488 0.488 0.488 0.488 0.488 0.488 0.488
C2= 0.976 0.976 0.9795 0.976 0.976 0.976 0.976 0.976 0.976
C3= 1.95 1.952 1.952 1.959 1.952 1.952 1.952 1.952 1.952
C4= 2.93 2.928 2.928 2.928 2.939 2.928 2.928 2.928 2.928
C5= 3.904 3.904 3.904 3.904 3.904 3.918 3.904 3.904 3.904
C6= 4.88 4.88 4.88 4.88 4.88 4.88 4.899 4.88 4.88
C7= 5.856 5.856 5.856 5.856 5.856 5.856 5.856 5.877 5.856
C8= 6.832 6.832 6.832 6.832 6.832 6.832 6.832 6.832 6.857
C-average= 3.4770 3.4772375 3.4774375 3.477875 3.478375 3.47875 3.479375 3.479625 3.480125
0.0002375 0.0004375 0.000875 0.001375 0.00175 0.002375 0.002625 0.003125
u(C-average)= 0.005316337 5.64062E-08 1.91406E-07 7.65625E-07 1.89062E-06 3.0625E-06 5.64062E-06 6.89062E-06 9.76562E-06
A. Drolc
STEP 3: Calculation of B1
C1= A1= C2= A2= C3= A3= C4= A4= C5=
Values 0.488 51437 0.976 98583 1.952 200037 2.928 305822 3.904
Uncertainties 0.044 76 0.043 204 0.044 151 0.044 312 0.044
C1= 0.488 0.532 0.488 0.488 0.488 0.488 0.488 0.488 0.488 0.488
A1= 51437 51437 51513 51437 51437 51437 51437 51437 51437 51437
C2= 0.976 0.976 0.976 1.019 0.976 0.976 0.976 0.976 0.976 0.976
A2= 98583 98583 98583 98583 98787 98583 98583 98583 98583 98583
C3= 1.952 1.952 1.952 1.952 1.952 1.996 1.952 1.952 1.952 1.952
A3= 200037 200037 200037 200037 200037 200037 200188 200037 200037 200037
C4= 2.928 2.928 2.928 2.928 2.928 2.928 2.928 2.972 2.928 2.928
A4= 305822 305822 305822 305822 305822 305822 305822 305822 306134 305822
C5= 3.904 3.904 3.904 3.904 3.904 3.904 3.904 3.904 3.904 3.948
A5= 418793 418793 418793 418793 418793 418793 418793 418793 418793 418793
C6= 4.88 4.88 4.88 4.88 4.88 4.88 4.88 4.88 4.88 4.88
A6= 531318 531318 531318 531318 531318 531318 531318 531318 531318 531318
C7= 5.856 5.856 5.856 5.856 5.856 5.856 5.856 5.856 5.856 5.856
Measurement of Concentration of Nitrate in Drinking Water
A7= 645226 645226 645226 645226 645226 645226 645226 645226 645226 645226
C8= 6.832 6.832 6.832 6.832 6.832 6.832 6.832 6.832 6.832 6.832
A8= 767727 767727 767727 767727 767727 767727 767727 767727 767727 767727
C-average= 3.477 3.477 3.477 3.477 3.477 3.477 3.477 3.477 3.477 3.477
A-average= 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88
B1= 112837.2713 113250.1549 112831.1121 1.132E+05 1.128E+05 1.130E+05 1.128E+05 1.129E+05 1.128E+05 1.128E+05
412.8836783 -6.159118274 329.2488931 -13.83320764 193.7876004 -6.243467101 56.59986455 -4.644147184 -71.38606829
u(B1)= 831.0006894 170472.9318 37.93473791 108404.8336 191.3576335 37553.63408 38.98088144 3203.544667 21.56810306 5095.970746
% u(B1)= 0.736459399
41
42
STEP 3: Calculation of B1 (cont.)

A5= C6= A6= C7= A7= C8= A8= C-average= A-average=
Values 418793 4.88 531318 5.856 645226 6.832 767727 3.477 377367.88
Uncertainties 315 0.044 712 0.045 716 0.046 295 0.0156 146
C1= 0.488 0.488 0.488 0.488 0.488 0.488 0.488 0.488 0.488 0.488
A1= 51437 51437 51437 51437 51437 51437 51437 51437 51437 51437
C2= 0.976 0.976 0.976 0.976 0.976 0.976 0.976 0.976 0.976 0.976
A2= 98583 98583 98583 98583 98583 98583 98583 98583 98583 98583
C3= 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952 1.952
A3= 200037 200037 200037 200037 200037 200037 200037 200037 200037 200037
C4= 2.928 2.928 2.928 2.928 2.928 2.928 2.928 2.928 2.928 2.928
A4= 305822 305822 305822 305822 305822 305822 305822 305822 305822 305822
C5= 3.904 3.904 3.904 3.904 3.904 3.904 3.904 3.904 3.904 3.904
A5= 418793 419108 418793 418793 418793 418793 418793 418793 418793 418793
C6= 4.88 4.88 4.924 4.88 4.88 4.88 4.88 4.88 4.88 4.88
A6= 531318 531318 531318 532030 531318 531318 531318 531318 531318 531318
C7= 5.856 5.856 5.856 5.856 5.901 5.856 5.856 5.856 5.856 5.856
A7= 645226 645226 645226 645226 645226 645942 645226 645226 645226 645226
C8= 6.832 6.832 6.832 6.832 6.832 6.832 6.878 6.832 6.832 6.832
A8= 767727 767727 767727 767727 767727 767727 767727 768022 767727 767727
C-average= 3.477 3.477 3.477 3.477 3.477 3.477 3.477 3.477 3.4926 3.477
A-average= 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377367.88 377513.88
B1= 112837.2713 1.128E+05 1.126E+05 1.129E+05 1.125E+05 1.129E+05 1.124E+05 1.129E+05 1.128E+05 1.128E+05
3.646846346 -199.3082064 27.08424298 -332.4772175 46.18346366 -460.0447772 26.83450429 -5.955889584 0
u(B1)= 831.0006894 13.29948827 39723.76115 733.5562177 110541.1001 2132.912316 211641.1971 720.0906207 35.47262073 0
% u(B1)= 0.736459399
A. Drolc
STEP 4: Calculation of B0
A-average Caverage B1
Values 377367.8 3.477 112837.270
Uncertainties 146.48 0.005316337 831
A-average 377367.8 377514.28 377367.8 377367.8

Caverage 3.477 3.477 3.482316337 3.477
B1 112837.27 112837.27 112837.27 113668.27
B0= -14967.4 -14820.91 -15567.2687 -17856.7748

146.48 -599.8809535 -2889.387
u(B0)= 2954.635474 21456.3904 359857.1583 8348557.236
STEP 5: Calculation of result C

A B0 B1 Fdil R
Values 594350.0 -14967 112837 1 1
Uncertainties 4220 3389 831 0 0.01
A 594350.00 598570 594350 594350 594350 594350

B0 -14967 -14967 -11578 -14967 -14967 -14967
B1 112837 112837 112837 113668 112837 112837
Fdil 1 1 1 1 1 1
R 1 1 1 1 1 1.01
C= 5.4 5.437374 5.369940711 5.360497238 5.39997519 5.34651008

0.03739908 -0.030034475 -0.039477948 0 -0.053465101
u(C)= 0.1 0.00139869 0.00090207 0.001558508 0 0.002858517
% u(C)= 1.5
ri. % = 20.82 13.43 23.20 0.00 42.55
U(C). (k=2)= 0.2

% u(C)(k=2)= 3.0
Nitrate determination - sources of measurement

uncertainty
B0
B1
Fdil
R
44 A. Drolc
Andreja Drolc works as senior researcher at the National

Institute of Chemistry in Ljubljana, Slovenia. Her research
activities are related mostly to environmental area: application of
classical and instrumental analysis for water samples, metrology
in chemistry (method validation, measurement uncertainty eval-
uation, interlaboratory comparisons) and quality systems
according to international standards (ISO/IEC 17025). She has
been involved in MiC related educational activities both at
national and international level.
Measurement of Moisture Content
(Water Content) in Edible Oil Using
the Volumetric Karl Fischer Method
According to ISO 8534:1996
Ivo Leito and Lauri Jalukse
1.1 Principle
Karl Fischer (KF) titration is based on the following reaction:
_
I2 + N ⋅SO2 + 2 N + H2O → 2 N H+ I + N ⋅SO3
ð1Þ
_
N ⋅SO3 + ROH → N H + ROSO3 ð2Þ
The reaction is very fast and with strict stoichiometry. Solution of iodine, SO2
and pyridine dissolved the alcohol ROH is the titrant solution. In the classical KF
titrant the alcohol ROH is methanol. In modern commercial titrants ROH is often
methoxyethanol and pyridine is often replaced by imidazole (both because of
potential toxicity). Titration is carried out in ROH or in a mixture of ROH and some
other solvent (if samples are not soluble in ROH).
The end point of the titration is indicated by a small amount of unreacted iodine
in solution. End-point is usually determined voltammetrically: alternating current of
constant strength is applied to a double Pt electrode. Potential difference between

I. Leito (&) L. Jalukse

Institute of Chemistry, University of Tartu, Ravila 14a, 50411 Tartu, Estonia
e-mail: ivo.leito@ut.ee

https://doi.org/10.1007/978-3-030-20347-4_3
46 I. Leito and L. Jalukse
the Pt wires is monitored. Even small quantities of iodine lead to a dramatic drop of
the potential difference.
1.2 Scope
The procedure is suitable for moisture content (water content) determination in

edible oils in the range of 50–2000 mg/kg.
1.3 Procedure
KF titration is carried out in a tightly closed (to minimize sample contamination by

atmospheric moisture) magnetically stirred titration cell in a solvent (mixture of
methoxyethanol and chloroform). The cell is not emptied after each sample.
Samples are titrated one after another in the same cell until the cell is full.
1. Before the first titration the solvent is titrated to the end-point (to remove traces
of water that have diffused into the cell).
2. 5–20 g of oil is weighed in a syringe to nearest 0.01 g.
3. The sample is injected into the titration cell.
4. The empty syringe is weighed again to determine the mass of the oil sample.
5. When the sample has dissolved then the sample is titrated to the end-point.
6. Concentration of water in the sample is calculated from the titration data.
The concentration of KF titrant solution is usually expressed as titer with respect
to water. The titer is determined by titrating either a standard solution of water or a
solution of some salt containing an exact amount of hydrate water in ROH. The titer
of freshly prepared KF titrant is around 5 mg of water per 1 ml of titrant. Every care
is taken to protect the titrant solution from atmospheric moisture. Nevertheless, the
titer of the titrant solution decreases with time and it is necessary redetermine it
daily or every second day.
1.4 Interferences
Many compounds (strong oxidizing and reducing agents, aldehydes, etc.) can
interfere with KF reaction. However, no such compounds are present in edible oils.
Thus within the scope of application there are no interferences.
Measurement of Moisture Content (Water Content) in Edible Oil … 47
1.5 Equipment and Reagents

Titrator
Volumetric KF-titrator capable of air-tight storage of the titrant and dispensing
titrant in 0.002 ml steps, with voltammetric end point detection.
Solvent
Mixture of methoxyethanol and chloroform.
Titrant
Commercial KF titrant with proprietary composition.
1.6 Sampling and Sample Pre-treatment
The oil sample is shaken to ensure its homogeneity and a suitable amount is aspired
into an air-tight plastic syringe.
1.7 Calculation
Calculation of water content in sample is carried out according to the following

equation (mathematical model):
Vt Tt
cwater ¼ 1000 ð3Þ
ms
cwater is concentration of water in oil sample [mg/kg], Vt is volume of titrant used

for titration [ml], Tt is the titer of the titrant (mass of water per unit volume of
titrant) [mg/ml], ms is the sample mass [g], 1000 [g/kg] is the unit conversion factor.
1.8 Results
The result is presented with number of decimal digits corresponding to the obtained
uncertainty.
2 The Customer’s Requirement Concerning Quality

In this case the customer is a producer of refined rapeseed oil. The measurement is
needed to test if the moisture content in the oil under question is below or above the
maximum permissible limit, which is 500 mg/kg. This permissible limit has been
set by the quality standard of the producer.
3 Validation Data of the Measurement Procedure –

Relevant Equations and Measurement Data
3.1 Equations
Pooled standard deviation:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðn1 1Þs21 þ ðn2 1Þs22 þ þ ðnk 1Þs2k
spooled ¼ ð4Þ
n1 þ n2 þ þ nk k
k number of data groups

s 1, s 2, … within group standard deviations
n 1, n 2, … numbers of measurements in groups
3.2 Validation Data
The data are obtained on different days with different samples. From these data the
pooled repeatability standard deviation sr can be calculated. It can be seen that there
is no correlation between the water content and standard deviation. Thus sr can be
assumed to be reasonably constant over the whole concentration range (Table 1).
Table 1 Repeatability data

Sample 1 Sample 2 Sample 3 Sample 4 Sample 5
−1
Values (mg kg ) 77.6 115.6 181.6 249.3 392.3
76.6 125.3 183.6 255.4 389.1
87.1 128.7 187.4 253.1 378.1
Average 80.4 123.2 184.2 252.6 386.5
Std Dev 5.8 6.8 2.9 3.1 7.4
4 Measurement Uncertainty of the Result – Relevant

4.1 Equations for Uncertainty Calculation by the ISO GUM

Modelling (Bottom up) Approach
The equation to use for the ISO GUM modelling approach is similar to the equation
used for calculating the result. There is only one difference: the repeatability
components of Vt and ms are not included in the uncertainties of these quantities but
are taken into account separately by the repeatability factor fr. The reason is the
following: the amount of injected sample is measured gravimetrically and each time
the sample amount is slightly different.
The equation on which the uncertainty calculation is based is the following:
Vt Tt
cwater ¼ fr 1000 ð5Þ
ms
The value of fr is 1 (without units). Its standard uncertainty u(fr) is found as

follows:
sr
uðfr Þ ¼ ð6Þ
cwater
sr can be found from the repeatability data (Table 1).

The combined standard uncertainty of cwater is found as follows:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

uðVt Þ 2 u ð Tt Þ 2 uðms Þ 2 uðfr Þ 2
uc ðcwater Þ ¼ cwater þ þ þ ð7Þ
Vt Tt ms fr
4.2 Measurement Data for Uncertainty Calculation Using

the ISO GUM Modelling Approach
Table 2 presents the data relevant for the particular measurement for which the
uncertainty is calculated.
Table 2 Measurement data

Input quantity Unit Value Uncertainty Type of Comments
or component uncertainty
Vt ml 0.720
u(Vt, cal) ml 0.004 Rectangular
u(Vt, temp) ml 0.002 Standard
u(Vt, rep) ml – Taken into account by fr
Tt mg/ml 4.9987 0.05 Standard
ms g 9.7734
u(ms, drift) g 0.0002 Rectangular
u(ms, nonlin) g 0.0002 Rectangular
u(ms, rep) g – Taken into account by fr
fr Unitless 1 Uncertainty to be found
from repeatability data
4.3 Equations for Uncertainty Calculation Using In-house

Validation Data
The notations are the same is used in the Nordtest handbook for uncertainty
evaluation.1
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uc ¼ uðRw Þ2 þ uðbiasÞ2 ð8Þ
u(Rw) is the uncertainty component that takes into account precision, within lab
reproducibility.
u(bias) is the uncertainty component that takes into account the trueness, i.e. the
(possible) lab bias. The trueness estimate in turn is composed of two components:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðbiasÞ ¼ RMS2bias þ uðCref Þ2 ð9Þ
RMSbias is the root mean of bias.

u(Cref) is the uncertainty of the reference value against which the bias is found.
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
ðbiasi Þ2
RMSbias ¼ ð10Þ
n
1
Handbook for calculation of measurement uncertainty in environmental laboratories (2017)
Nordtest TR 537, 4th ed. Espoo, Finland (available from the Internet: http://www.nordtest.info/).
Table 3 Participation in interlaboratory comparisons

No Ref value (mg kg−1) s (mg kg−1) No of labs Lab (mg kg−1)
1 363 33 7 381
2 88 19 5 96
3 82 22 7 85
4 374 21 9 365
biasi is the bias found on i-th bias determination (difference between lab result and
reference value‚ Table 3).
If interlaboratory comparisons are used then u(Cref) is found as follows:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 2
uðCref Þi
uðCref Þ ¼ ð11Þ
n
n is the number of interlaboratory comparisons where the laboratory has

participated.
U(Cref)i is the uncertainty of the reference value (or consensus value) of the i-th
interlaboratory comparison.
si
uðCref Þi ¼ pffiffiffiffi ð12Þ
ni
si is the standard deviation of participant results of the i-th intercomparison.

ni is the number of participant results from which the si value was obtained.
4.4 Data for Uncertainty Calculation Using In-house

Validation Data
In this case the pooled standard deviation sr (data from Table 1) can be used as an
estimate of u(Rw).
Interlaboratory comparison data (Table 2) are used for finding RMSbias and u
(Cref).

Chemistry
Analyte Water
Measurand Content of water in the sample
Units mg kg−1

equation
Measurement procedure Volumetric Karl Fischer titration
Type of calibrationa Standard curve Standard addition Internal standard
a
The titer of the titrant (mass of water per unit volume of titrant) is determined. This could be considered
as two-point standard curve
Model equation
Calculation of water content in sample is carried out according to the following
equation (mathematical model):
Vt Tt
cwater ¼ fr 1000 ð13Þ
ms
cwater is concentration of water in oil sample [mg/kg], Vt is volume of titrant used

for titration [ml], Tt is the titer of the titrant (mass of water per unit volume of
titrant) [mg/ml], ms is the sample mass [g], fr is the repeatability factor, 1000 [g/kg]
is the unit conversion factor. The reason for inclusion of fr is the following: the
amount of injected sample is measured gravimetrically and each time the sample
amount is slightly different. Thus, the repeatability components of Vt and ms are not
included in the uncertainties of these quantities but are taken into account separately
by the repeatability factor fr.
1 fr
2 Tt
3 Vt
4 ms
5
4. List the reference standards needed and give also the information
regarding traceability of the reference value
For the analyte
1 Name/ChemicalFormula/ Water reference standard solution, e.g. any of the Hydranal

producer: standards available from Sigma-Aldrich
2 Name/ChemicalFormula/
producer:
1 Quantity/Equipment/calibration: Balance calibrated by NMI

e.g. mass/balance/calibrated by NMI, U = xx (k = 2), see
also data yellow sheet

Are all important parameters included in the measurement equation? Yes No
Other important parameters are:

1 By measuring independently prepared reference standards (e.g. other manufacturer) with
reference values traceable to the same point of origin
2
3
7. Any other comments, questions…


Procedures

The measurement procedure Volumetric Karl Fischer titration
Analyte Water
The measurand Content of water in the sample
Unit mg kg−1
2. Specify the Scope

Matrix Edible oils
Measuring range 100–500 mg kg−1

Intended use of the results: Characterisation of edible oil quality
Mark the customer’s LOD No requirement, as this is not trace
requirements and give analysis
their values LOQ 100 mg kg−1
Repeatability 8 mg kg−1 expressed as standard
deviation
Within-lab 16 mg kg−1 expressed as standard
reproducibility deviation
Measurement 20 mg kg−1 expressed as standard
uncertainty uncertainty
Trueness It is important, but is actually taken
into account by measurement
uncertainty
Other-state

Validation
New in-house method Full
Modified validated method Partial
Official standard method Confirmation/Verification
Part II: Parameters to Be Validated

5. Selectivity/Interference/Recovery (where yes, please give further informa-
tion e.g. which CRM, reference method)
Further information:

KF titration is in fact the reference method for moisture determination in edible oil.
However, interlaboratory comparison (even if the other participants have likewise KF
methods) is certainly useful
Selectivity is provided by (1) the chemistry of the KF method and (2) the simplicity and
small variability of the edible oil matrix
6. Measuring range
Linearity
Upper limit
LOD
LOQ
Repeatability
8. Robustness
The following parameters should be varied:
- Stirring speed;
- Titrant flow rate (i.e. rate of piston movement in the burette);
- Age of titrant.
9. Quality Control
Control charts

R square
Standard deviation of the method
Coefficient of variation of the method

These parameters are: LOD LOQ Repeatability
Within-lab reproducibility Uncertainty Trueness
Parameter 1: Repeatability sr is found as pooled standard deviation from the
Repeatability data of Table 1:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
ðn1 1Þs21 þ ðn2 1Þs22 þ þ ðnk 1Þs2k
Spooled ¼ n1 þ n2 þ þ nk k ð14Þ
k number of data groups

s1, s2, … within group standard deviations
n1, n2, … numbers of measurements in groups
sr = 5.5 g kg−1
Parameter 2: There are no direct data available on within-lab reproducibility.
Within-lab However, it can tentatively be assumed that sRW = 2 ∙ sr
reproducibility sRW = 11 g kg−1
Parameter 3: Dedicated LOQ determination was not carried out. However,
LOQ from the available data (Table 1) it is known that the procedure
operates without problems at 80 mg kg−1. So, this value is
tentatively used as an estimate of LOQ
Parameter 4: See separate section on measurement uncertainty
Measurement
uncertainty
12. The method fulfils the requirement for the intended use:
Parameter Value requested by the customer Value obtained during validation
LOD
LOQ 100 mg kg−1 80 mg kg−1
−1
Repeatability 8 mg kg 5.5 mg kg−1
Within-lab reproducibility 16 mg kg−1 11 mg kg−1
Measurement uncertainty uc 20 mg kg−1 ISO GUM Modeling: 7 mg kg−1
Nordtest: 15 mg kg−1
Trueness
Other
Yes No
13. Calculation of other parameters
Here Only the ISO GUM Modelling Approach Is Addressed

Measurand Water content of edible oils
Unit mg kg−1

equation
Karl Fischer (KF) titration is based on the following reaction:
_
I2 + N ⋅SO2 + 2 N + H2O → 2 N H+ I + N ⋅SO3
ð15Þ
_
N ⋅SO3 + ROH → N H + ROSO3 ð16Þ
The reaction is very fast and with strict stoichiometry. Solution of iodine, SO2
and pyridine dissolved the alcohol ROH is the titrant solution. In the classical KF
titrant the alcohol ROH is methanol. In modern commercial titrants ROH is often
methoxyethanol and pyridine is often replaced by imidazole (both because of
potential toxicity). Titration is carried out in ROH or in a mixture of ROH and some
other solvent (if samples are not soluble in ROH).
The end point of the titration is indicated by a small amount of unreacted iodine
in solution. End-point is usually determined voltammetrically: alternating current of
constant strength is applied to a double Pt electrode. Potential difference between
the Pt wires is monitored. Even small quantities of iodine lead to a dramatic drop of
the potential difference.
Model equation:
See Eq. 13.
Uncertainty of concentration of the titrant
Uncertainty of measurements of the titrant volume
Method bias (taken into account within the other input quantities)
Matrix effect
Other: Sample mass
Other: Overall repeatability
Other:

Sample mass, ms 9.7734 g
Titrant volume used 0.720 ml
for titration Vt
Titer of the titrant Tt 4.9987 mg ml−1
Repeatability fr 1 – Repeatability is taken into account using a
multiplicative factor as the overall repeatability of
the procedure. This means that the repeatability
contributions of the other input quantities are not
taken into account with those quantities

Input quantity Standard Unit Remark
uncertainty
Sample mass, ms 0.00016 g
Titrant volume used 0.0031 ml
for titration Vt
Titer of the titrant Tt 0.05 mg ml−1
Repeatability fr 0.015 – Evaluated as pooled repeatability
standard deviation (sr) divided by the
actual water content (cwater) in the sample

cwater = 369 mg kg−1
units
Software
Input quantity Value Standard uncertainty Unit Remark

Sample mass, ms 9.77340 0.00016 g
Titrant volume used for titration Vt 0.720 0.0031 ml
Titer of the titrant Tt 4.9987 0.05 mg ml−1
Repeatability fr 1 0.015 –
uc = 6.8 mg kg−1
the units
U = 2 ∙ uc = 14 mg kg−1
1 Repeatability fr
2 Titer of the titrant Tt
3 Titrant volume used for titration Vt
Uncertainty estimation: Water content in refined rapeseed oil

by volumetric Karl Fischer titration (ISO 8534:1996)
Value unit Uncertainty components u
Sample mass drift(±) repeatability non-linearity(±)

ms 9.7734 g 0.0002 0 0.0002 0.000163
Titrant volume used for titration cal(±) repeatability temp(su)

Vt 0.720 ml 0.004 0 0.00199532 0.0031
Titer of the KF titrant

Tt 4.9987 mg(H2O)/ml 0.05000
Repeatability
fr 1.000 0.015
V⋅T
cwater = t t ⋅ f r ⋅ 1000
ms
Finding combined uncertainty:
2 2 2 2
⎛ u(V ) ⎞ ⎛ u(T ) ⎞ ⎛ u(ms ) ⎞ ⎛ u( fr ) ⎞
uc (cw ater) = c w ater ⎜⎜ t ⎟⎟ + ⎜⎜ t ⎟⎟ + ⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟
Result:
⎝ Vt ⎠ ⎝ Tt ⎠ ⎝ ms ⎠ ⎝ f r ⎠
u' (m s ) u' (V t ) u' (T t ) u' (f r)
c water 368.2510 % 0.00002 0.0042 0.0100 0.0150 6.8332
Index: 0.0% 5.2% 29.1% 65.7%
Result: Water content in oil: 368 mg/kg

Expanded uncertainty (k = 2): 14 mg/kg
Ivo Leito works as professor of analytical chemistry at

University of Tartu. His research directions are on the borderline
of analytical chemistry with other disciplines: chemistry of
superacids and superbases; metrology in chemistry (MiC); liquid
chromatography and mass spectrometry; sensors and their
metrological characterization; applications of instrumental
methods in analysis of historical objects. He teaches analytical
chemistry and its metrological aspects at UT and has been
involved in setting up several international MiC-related educa-
tional activities.
Lauri Jalukse worked as analytical chemistry research fellow at

University of Tartu and works now at the Estonian Veterinary
and food board. He taught analytical chemistry and metrology in
chemistry at all study levels. He was continuously introducing
innovative and active learning approaches into teaching. His
research work was focused on metrological studies of electro-
chemical and optical sensors, measurements of dissolved oxygen
concentration and moisture content, as well as organization of
interlaboratory comparisons.
Measurement of Polyphosphates in Meat
Products by Spectrophotometric Method
According Laboratory-Developed
Method
Tidža Muhić-Šarac
1.1 Task Description
1.1.1 Polyphosphates as Additive to Meat Products
According to the EU directive 95/2/EC and regulations from the Statute on Quality
of Meat Products (B&H), polyphosphates can be used in production of all kinds of
sausages (except those with long shelf life) and in production of canned meat and
other products which are thermally processed. The main condition for their usage is
that the quantity in the final product does not exceed w = 0.5% or w = 0.3%
(m m−1) in relation to the quantity of the product, expressed as P2O5 content.
Polyphosphates (condensed phosphates) can be obtained by heating of primary or
secondary orthophosphates, which dehydrate and get condensed at the same time.
Only cyclical polyphosphates are used in meat processing and most frequently in the
form of: disodium phosphate (Na2HPO4; E339), mono-sodium phosphate (NaH2PO4;
E339), sodium-meta-phosphates (Na3P3O9; E452), tetrasodium-pyrophosphate
(Na4P2O7; E450) and acid disodium pyrophosphate (Na2H2P2O7; E450).
They function in the direction of protein hydration and puffing up of muscle
tissue, and better homogeneity, to achieve a firm consistence and stability of the
meat products. They also help in improvement of taste, and mistakes in production
are reduced to a minimum. From the point of health the possibility of harmful

T. Muhić-Šarac (&)
Faculty of Science, Department of Chemistry, University of Sarajevo, Zmaja od Bosne 33-35,
71000 Sarajevo, Bosnia and Herzegovina
e-mail: tidzamuhic@gmail.com

https://doi.org/10.1007/978-3-030-20347-4_4
64 T. Muhić-Šarac
effects of polyphosphates to human health is generally excluded if utilized in

quantity proscribed by the Statute for Quality of Meat Products. However numerous
experiments on animals as well as experience of people show that increased con-
sumption of polyphosphates can harm human health.
Since polyphosphates often fall into a group of additives (food emulsifier) in
meat products, their content need to be listed by producers and its presence has to
be inspected by a verified laboratory.
1.2 Principle of the Measurement Method
The spectrophotometric determination of 12-molybdophosphoric acid (12-MPA) is

most frequently used for the determination of phosphorus in the range of 10−3–
10−4%, according following reaction:
3 Na þ þ 12 MoO2
4 þ H3 PO4 þ 21 H
þ
¼ ðNaÞ3 PO4 ðMoO3 Þ12 þ 12 H2 O
Determination of phosphates in meat products is performed by molecular absorp-

tion spectrophotometric method. Phosphate forms a blue-colored complex with
sodium molybdate (Na2MoO4 2H2O) in an acid medium. Additional reagents as
ascorbic acid (C6H8O6) provide the proper reduction medium and potassium
pyrosulphate (K2S2O7) provide the proper transform into orthophosphate.
1.3 Analytical Procedure
1.3.1 Translation Samples in Solution
1. Weigh approximately 5.0 g ± 0.1 mg (m1) of the sample (meat product) and
transfer it on ceramic hot plate.
2. The digestion of organic issue (mineralization of samples) was done by calci-
nations with long lasting (3 h) and gradual heating in the furnace, starting from
room to t = 530 °C temperature.
3. The dissolution of ash, which consists of oxides P2O5, K2O, CaO, MgO, Fe2O3,
and SiO2, chlorides and sulfates was done with 6 mol L−1 HCl and quantita-
tively taken over into a volumetric flask 100 mL (V1). This is stock solution of
sample (meat product).
1.3.2 Intermediate Checks of Spectrophotometer-in One Point
According to the requirements of ISO/IEC 17025:2005, (5.5.10) stated: »When

intermediate checks are needed to maintain confidence in the calibration status of
the equipment, these checks shall be carried out according to a defined procedure«.
Measurement of Polyphosphates in Meat Products … 65
Spectrophotometer at which the measurement is made of phosphate content was

calibrated by the manufacturer, HACH Company. Data from the calibration are
stored in the software program 3025th.
To maintain confidence in the status of the spectrophotometer calibration must
be carried out necessary checks according to a defined procedure. Intermediate
checking the spectrophotometer is performed using a standard solution of potas-
sium dihydrogen phosphate (KH2PO4). The concentration of phosphate is in the
area of interest.
Preparation of Standard Solutions

1. Stock potassium hydrogen phosphate standard solution
Potassium dihydrogen phosphate (KH2PO4) is dried at 105 °C for 2 h. Allow to

cool in a desiccator. Dissolve 426.6 mg of dried potassium dihydrogen phosphate
in 100 mL of distilled water taken in a 1000 mL volumetric flask. Mix to dissolve
and make up to 1000 mL with distilled water. Solution kept in dark bottle at room
temperature. Is stable for three months.
The phosphates (PO−3 −1
4 ) concentration in this solution is 0.3 mg mL .
2. Working potassium hydrogen phosphate standard solution
The calibration potassium dihydrogen phosphate (KH2PO4) standard solution are

prepared by dilution of stock potassium dihydrogen phosphate standard solution.
Into 200 mL volumetric flask transfer 1.0 mL of stock potassium dihydrogen
phosphate standard solution and dilute each to 200 mL with distilled water. The
working standard solution contain 0.0015 mg mL−1 phosphates (PO−3 4 ). Solution
kept in dark bottle at room temperature. Is stable for one weak.
Procedure for Intermediate Checks of Spectrophotometer
1. Transfer 10 mL of standard work solution (concentration 0.0015 mg mL−1

phosphates) in the spectrophotometric cuvette.
2. Add the “Reagent pillows” for developing colour. Reagent pillows was pulver
mixture of sodium molybdate (Na2MoO4), ascorbic acid (C6H8O6) and potas-
sium pyrosulphate (K2S2O7).
3. Prepare the blank probe in the way that you add the “Reagent pillows” to 10 mL
distilled water.
4. After two minutes of reacting time measured the concentration of phosphate in
the standard work solution at 890 nm.
5. If you are a good confirmation of the status of equipment (spectrophotometer)
approach to measuring the content of phosphate from the solution of the meat
product (stock solution of sample).
66 T. Muhić-Šarac
Model Equation
Intermediate checks of spectrophotometer-in one point

Concentration of PO−3
4 in working standard solution:
mðKH2 PO4 Þ P 10 1000 V2

cPO3 ðmg L1 Þ ¼
4 V4 V3 V1 F
where:
C Concentration of PO−3 4 in working standard solution mg L
−1
m1 Mass KH2PO4 (±0.1 mg)

P Purity of KH2PO4 (99.95% m m−1)
V1 Volumetric flask 1000 mL (±0.15 mL)
V2 Pipette 1 mL (±0.01 mL)
F Concentration factor M(KH2PO4)/M(PO−3 4 ) = 1.43
10 Factor of calculation
1000 Factor for conversion (L ! mL)
1.3.3 Measurement
1. Transfer 1.0 mL (V2) from the stock solution samples (meat products) to a
200 mL (V3) volumetric flask, and add distilled water till the mark.
2. Transfer 10.0 mL (V4) from the above solution into the spectrophotometric
cuvette. Add the “Reagent pillows” for developing color. Allow this solution to
stay at least two minutes to color developing.
3. Parallel prepare the blank probe in the way that you add the reagent pillow to
10 mL blank probe, without meat product.
4. Measuring of the optical density of the obtained blue-colored complex which is
in a linear dependence on the concentration of the phosphate was done by a
UV-VIS spectrophotometer product of the HACH, USA. type DR/4000U
according to program: 3025, for concentration area within 0–2.5 mg L−1 PO−3 4
in a 2 cm cuvette and with k = 890 nm wavelength.
5. Read the content of phosphate in mg L−1.
6. Calculate the content of phosphorous-pentoxide given by the formula.
Calculation of the Result
Mass fraction wP2 O5 in %
c1PO3 ðmg L1 Þ 0:75 2

WP2 O5 ð%Þ ¼ 4
msample ðgÞ R
where:
0.75 Concentration factor M(P2O5)/2M(PO−3
4 )
Comprehensive Equation for Mass Fraction wP2O5 in %

c1 V4 V3 V1 100 F Rrep: Icheck
WP2 O5 ð%Þ ¼
10001 10 V2 m1 10002 R
where:
wP2O5 Mass fraction wP2O5 in %
C1 Concentration of PO−3 −1
4 (mg L ) in meat product
m1 Mass of sample (g) (±0.1 mg)
F Concentration factor M(P2O5)/2M(PO−34 ) = 0.75
100 Factor for conversion (mg ! %)
10001 Factor for conversion (mg L−1 ! mg mL−1)
10002 Factor for conversion (g ! mg)
R Recovery
Rrep Repeatability
Icheck Intermediate checks of spectrophotometer-in one point

Is defined as the maximum concentration of eligibility:

w(P2O5) = 0.30% (m m−1) – Statute on Quality of Meat Products (B&H) and
w(P2O5) = 0.50% (m m−1) – EU Directive 95/2/EC
68 T. Muhić-Šarac

3.1 Limit of Detection (LOD) and Limit of Quantification

(LOQ)
Results for the measurements of eight (separately prepared) blank samples
No. of blank probe Concentration of PO−3 −1

4 (mg L ) in blank probe
1 0.008
2 0.007
3 0.008
4 0.007
5 0.008
6 0.007
7 0.008
8 0.007
Mean
Sb
Signal = b0 + b1 c
Ybl = Signal of the ‘blank’;
Sbl = stdev of the ‘blank’ in signal domain
YLOD ¼ Ybl þ 3 Sbl ! LOD ¼ ðYLOD b0 Þ=b1

YLOQ ¼ Ybl þ 10 Sbl ! LOQ ¼ ðYLOQ b0 Þ=b1
Eqution of calibration curve is:
Y ¼ 0:5727 X b0 ¼ . . .. . .. . .. . .. . .. . .. . . and b1 ¼ . . .. . .. . .. . .. . .. . .::

Yb1 ¼ . . .. . .. . .. . . mg L1 and Sbl ¼ . . .. . .. . .. . .. . .:: mg L1
YLOD ¼ . . .. . .. . .. . .. . .. . .. . .. . .. . . ! LOD ¼ . . .. . .. . .. . .. . .. . .. . .. . .. . . mg L1 PO3

4
YLOQ ¼ . . .. . .. . .. . .. . .. . .. . .. . .. . . ! LOQ ¼ . . .. . .. . .. . .. . .. . .. . .. . .. . . mg L1 PO3
4
3.2 Repeatability of Measurements on Spectrophotometer
Stability of measurement on spectrophotometer – repeatability of concentration

PO−3
4 (mg L )
−1
No. of probe Concentration of PO−3 −1

4 (mg L ) (C1)
1 1.220
2 1.218
3 1.222
4 1.223
5 1.218
6 1.221
7 1.216
8 1.219
Mean of C1 (PO−3 −1
4 (mg L ))
Standard deviation (S)
Variance (S2)
Coefficient of variation (C.V. (%))
3.3 Repeatability to Analytical Procedure, Including Sample

Treatment
The standard uncertainty of the measurement of phosphate content in the meat

products is expressed as standard deviation after eight consecutive analyses, and
according to the method that was described. Repeatability of concentration P2O5
in % (m m−1)
No. of probe Measured of PO−3 −1

4 (mg L ) (C1) Amount of P2O5 in % (m m−1)
1 1.20 0.360
2 1.29 0.387
3 1.19 0.357
4 1.29 0.387
5 1.26 0.378
6 1.29 0.387
7 1.22 0.366
8 1.17 0.351
Mean of PO−3 −1
4 (mg L ) and P2O5 (%)
Variance (S2)
Coefficient of variation (C.V. (%))
3.4 Within-Lab Reproducibilty
Intermediate precision was calculated in conditions of intra-laboratory repro-

ducibility. Intermediate precision condition of measurement: same measurement
70 T. Muhić-Šarac
procedure, same location, and replicate measurements on the same objects over a
short period of time, involving changes operators (Analyst)
No. of probe Analyst I Analyst II Analyst III

% (m m−1) % (m m−1) % (m m−1)
P2O5 P2O5 P2O5
1 0.360 0.388 0.388
2 0.387 0.350 0.353
3 0.357 0.382 0.368
4 0.387 0.356 0.383
5 0.378 0.352 0.386
6 0.387 0.386 0.375
7 0.366 0.383 0.392
8 0.351 0.350 0.352
Mean of P2O5 (%)
Variance (S2)
Coefficient of variation (C.V.
(%))
3.5 Linearity of Method
The linearity of method was evaluate for concentration rang of phosphates from
0.30 to 2.50 mg L−1 PO−34 . The using standard working solution is KH2PO4,
concentration 0.0015 mg mL−1 PO−34 .
Linearity of method
1,6
1,4 y = 0,5727x
R2 = 0,999
1,2
Absorbance
0,8
0,6
0,4
0,2
0
0 0,5 1 1,5 2 2,5 3
3-
Concentration of PO4 (mg/L)
3.6 Recovery
The Recovery was evaluate for seven concentration level with additional method.
Standard working solution is added in meat samples (matrix).
cobserved cmatrix
R¼
cspike
No. of Concentration Added conc. Total amount Recovery

probe PO−3
4 (mg L )
−1
PO−3 −1
4 (mg L ) (stand. Observed
(matrix-meat) solution) PO−3 −1
4 (mg L )
1 1.22 0.00 1.25 –

2 1.22 0.30 1.52 1.0000
3 1.22 0.60 1.81 0.9833
4 1.22 0.90 2.16 1.0444
5 1.22 1.20 2.46 1.0333
6 1.22 1.50 2.75 1.0200
7 1.22 1.80 3.05 1.0166
Mean of R
Standard deviation of R, (S)
Variance of R, (S2)
Coefficient of variation of R, (C.V. (%))

Equations and Measurement Data1
See Scheme 1.
1
Here you may also want to include the relevant certificates.
72 T. Muhić-Šarac
DRY KH2PO4 A -TYPE
VEIGH KH2PO4
(m1)
B -TYPE
SOLVE SALT A -TYPE
DILUTE (V1)
(mass/1000mL) A/B -TYPE
TAKE ALIQUOT (V2)

(1mL) B/A/B -TYPE
DILUTE (V3)
(1mL/200mL) B/A/B -TYPE
TAKE ALIQUOT (V4)

(10mL/measur. cuvette) B/A/B -TYPE
INTERMEDIATE A -TYPE
CHECKS
Scheme 1 Identification of uncertainty sources and type of standard uncertainty (INTERMEDIATE

CHECKS OF SPECTROPHOTOMETER)
4.1 Combined Standard Uncertainty Calculation

for Intermediate Checks of Spectrophotometer
mðKH2 PO4 Þ P 10 1000 V2
cPO3 ðmg L1 Þ ¼
4 V4 V3 V1 F
where:
−1

See Table 1 and Scheme 2.
Table 1 Uncertainty elements for intermediate checks of spectrophotometer – in one point

P 0.9995 0.0003 – Purity of KH2PO4
V1 1000.0 0.507 mL Volume of stock standard solution
KH2PO4
V2 1.0 0.004 mL Volume of aliquot standard solution
KH2PO4
V3 200.0 0.117 mL Volume of working standard solution
KH2PO4
V4 10.0 0.043 mL Volume of aliquot working standard
solution KH2PO4
m1 426.6 0.085 mg Mass of KH2PO4
F 1.43 0.000013 – Concentration factor M(KH2PO4)/M
(PO−3
4 )
74 T. Muhić-Šarac
A -TYPE
SAMPLE
A -TYPE
HOMOGENISE
B - TYPE
WEIGH (m1)
TERMIC
DESTRUCTION A - TYPE
SOLVE OF ASH A - TYPE
DILUTE (V1)
B/A/B-TYPE
(ash/100 mL)
TAKE ALIQUOT (V2)

(1mL) B/A/B -TYPE
DILUTE (V3)
(1mL/200mL) B/A/B -TYPE
A/B - TYPE TTYPE
INTERMED. TAKE ALIQUOT (V4)

CHECKS (10mL/meas. cuvette) B/A/B -TYPE
MEASUREMENT
A - TYPE
Scheme 2 Identification of uncertainty sources and type of standard uncertainty (ANALYTICAL

PROCEDURE)
4.2 Combined Standard Uncertainty Calculation

for Analytical Procedure
(a) Relevant equation to calculate combined uncertainty
c1PO3 ðmg L1 Þ 0:75 2

WP2 O5 ð%Þ ¼ 4
msample ðgÞ R
where:
4 )
(b) Comprehensive equation (containing all sources of uncertainty)

WP2 O5 ð%Þ ¼
10001 10 V2 m1 10002 R
where:
wP2O5 Mass fraction wP2O5 in %
R Recovery
Rrep Repeatability
See Table 2.
76 T. Muhić-Šarac
Table 2 Uncertainty elements for analytical procedure

Input Value Unit Standard Remark
m1 5000.0 mg 0.085 Mass of sample
(sample)
C1 1.22 mg L−1 0.002 Concentration of PO−3 −1
4 (mg L ) in
meat product
V1 100.0 mL 0.068 Volume of volumetric flask
V2 1.0 mL 0.004 Volume of sample solution aliquot
V3 200.0 mL 0.117 Volume of volumetric flask
V4 10.0 mL 0.043 Volume of pipette
F 0.75 – 0.00001 Concentration factor M(P2O5)/2M
(PO−3
4 )
R 1.016 – 0.0117 Recovery
Rrep 1 % 0.015 Repeatability
Icheck 1 mg L−1 0.009 Intermediate checks of
spectrophotometer

Chemistry
Analyte Phosphate expressed as phosphorus penthoxide (P2O5)
Measurand Phosphate content expressed as total phosphorus penthoxide in meat product
Units % (m m−1)

equation
Measurement Spectrophotometric method
procedure
Type of standard curve standard addition internal standard
calibration
Model equation
Phosphate concentration in working standard solution:
1
mðKH2 PO4 Þ P 10 1000 V2
cPO3 mg L ¼
4 V4 V3 V1 F
where:
−1

Analytical procedure
Mass fraction wP2 O5 in % (m m−1) in meat products:

WP2 O5 ð%Þ ¼
10001 10 V2 m1 10002 R
where
wP2 O5 Mass fraction wP2O5 in % (m m−1)
78 T. Muhić-Šarac

R Recovery
Rrep Repeatability

1 Repeatability
2 Recovery
3 Volume of the analysed solution (pipettes and volumetric flasks) (mL)
4 Concentration of standard solution-purity of KH2PO4 (99.95%)
5 Mass of homogenized sample (g)
6
For the analyte
1 Name/ChemicalFormula/producer: Purity of KH2PO4, Merck, min. 99.95%

2 Name/ChemicalFormula/producer:
1 Quantity/Equipment/calibration: e.g. Balance-calibrated by CRM etalones

mass/balance/calibrated by NMI. U = xx (certified mass pieces) in laboratory
(k = 2).
2 Quantity/Equipment/calibration: Volumetric flask-class A or B quality
(certificate) and calibration with water in
laboratory

Other important parameters are:

1 By the use of purity reagents (p.a. quality)
2 By the use of calibrated balance
3 By the use of volumetric flasks-class A or B quality calibrated in laboratory
7. Any other comments. questions…

80 T. Muhić-Šarac

Procedures

1. Specify the measurement procedure. analyte. measurand and units
The measurement Spectrophotometric method
procedure
Analyte Phosphate expressed as phosphorus penthoxide (P2O5)
The measurand Phosphate content expressed as total phosphorus penthoxide in
meat product
Unit % (m m−1)
in %

Matrix Meat (proteins, fats)
Measuring range 0.02–2.5 mg L−1 PO−3 −1
4 or 0.006–0.75% (m m ) P2O5

Intended use of the results
LOD Not defined
LOQ Not defined
Repeatability Not defined
Within-lab Not defined
reproducibility
Trueness Not defined
Measurement Not defined
uncertainty
Recovery Not defined

VALIDATION
Part II: Parameters to be Validated

Where yes. please give further information e.g. which CRM. reference method

Further information:
Purity of KH2PO4 (value of certificate 99.95% m m−1)
.
Selectivity. interferences
.
.
Recovery
Test for recovery with added standard solution KH2PO4 in matrix
6. Measuring range
Linearity
Upper limit
LOD
LOQ
Repeatability
8. Robustness
9. Quality Control
Control charts
82 T. Muhić-Šarac

Recovery

Parameters requested to be Calculations
validated
LOD YLOD ¼ Ybl þ 3 Sbl ! LOD ¼ ðYLOD b0 Þ=b1
YLOD ¼ 0:0075 þ 3 0:001 ¼ 0:0105
! LOD ¼ ð0:0105 0:0Þ=0:5727 ¼ 0.0183 mg L1 PO3
4
! LOD ¼ 0.0055% ðm m1 Þ P2 O5
LOQ YLOQ ¼ Ybl þ 10 Sbl ! LOQ ¼ ðYLOQ b0 Þ=b1
YLOQ ¼ 0:0075 þ 10 0:001 ¼ 0:0175
! LOQ ¼ ð0:0175 0:0Þ=0:5727 ¼ 0.0305 mg L1 PO3
4
! LOQ ¼ 0.0091% ðm m1 Þ P2 O5
rPffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Repeatability
Sr ¼ 0.0497 mg L1 PO3
n
ðx xÞ2
S¼ n¼1 i 4
n1 Sr ¼ 0.015% P2 O5
Within-lab SWR ¼ 0.0161% P2 O5
reproducibility sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðn1 1Þ s21 þ ðn2 1Þ s22 þ ðn3 1Þ s23
sWR ¼
n1 þ n2 þ n3 3
Trueness
Measurement U ¼ k uðxÞ; k ¼ 2 95% a level of confidence
uncertainty
U ¼ 2 0:0078 ¼ 0.016% ðm m1 Þ
Linearity Equation; Y = 0.5727X
Linearity rang; C = 0.02–2.5 mg L−1 PO−3 4
wðP2 O5 Þ ¼ 0.006 0.75% ðm m1 Þ
Coefficient of correlation; R = 0.999
cmatrix
Recovery R ¼ cobserved
cspike R ¼ 1.016 0.0117
use?
Parameter Value requested by Value obtained during The
the customer validation requirement is
(the same as stated fulfilled
in question 3) Yes/No
LOD Not defined 0.0183 mg L−1 PO−3
4 Yes
0.0055% (m m−1) P2O5
LOQ Not defined 0.0305 mg L−1 PO−3
4 Yes
0.0091% (m m−1) P2O5
Repeatability Not defined ±0.015% (m m−1) P2O5 Yes
Within-lab Not defined 0.0161% (m m−1) P2O5 Yes
reproducibility
Trueness Not defined
Measurement Not defined ±0.016% (m m−1) P2O5 Yes
uncertainty
Recovery Not defined 1.016 ± 0.0117 Yes
Yes No
Note:
This new analytical method is validated in one laboratory.
Requested by the customer is not defined, except the maximum limit in the content
of phosphate (% m m−1 P2O5) in meat products.
Evaluation of the obtained values for the elements of validation may be given to the
statement: “The analytical procedure is fit for the intended use.”
Low LOD, LOQ. Satisfactory: linearity of measurement in the measuring range of
interest, repeatability and within-lab. reproducibility, and the value of Recovery
test.
For Measurement Uncertainty and Traceability refer to the corresponding sheets
Building an Uncertainty Budget

Measurand Phosphate content expressed as total phosphorus penthoxide in meat product
Unit % (m m−1)
84 T. Muhić-Šarac

equation
2.1 Procedure of intermediate checks of spectrophotometer-in one point:
1. Transfer 10 mL of standard work solution (concentration 0.0015 mg mL−1

phosphates) in the spectrophotometric cuvette.
2. Add the “Reagent pillows” for developing colour. Reagent pillows was pulver
mixture of sodium molybdate (Na2MoO4), ascorbic acid (C6H8O6) and potas-
sium pyrosulphate (K2S2O7).
3. Prepare the blank probe in the way that you add the “Reagent pillows” to 10 mL
distilled water.
4. After two minutes of reacting time measured the concentration of phosphate in
the standard work solution at 890 nm.
5. If you are a good confirmation of the status of equipment (spectrophotometer)
approach to measuring the content of phosphate from the solution of the meat
product (stock solution of sample).
Model equation:
mðKH2 PO4 Þ P 10 1000 V2

cPO3 ðmg L1 Þ ¼
4 V4 V3 V1 F
where:
−1

2.2 Analytical procedure
1. Weigh approximately 5.0 g ± 0.1 mg (m1) of the sample and transfer it on

ceramic hot plate.
2. Thermally decomposed test sample of meat at a temperature of 530 °C.
3. Dissolve the resulting ash with 6 mol L−1 HCl and quantitatively transfer it to
the volumetric flask of 100 mL (V1). This is stock solution of sample (meat
product).
4. Transfer 1.0 mL (V2) from the stock solution sample (meat products) to a
200 mL (V3) volumetric flask, and add distilled water till the mark.
5. Transfer 10.0 mL (V4) from the above solution into the spectrophotometric
cuvette. Add the “Reagent pillows” for developing color. Allow this solution to
stay at least two minutes to color developing.
6. Parallel prepare the blank probe in the way that you add the reagent pillow to
10 mL blank probe, without meat product.
7. Measure the concentration of phosphate (C1 in mg L−1) in the UV-VIS spec-
trophotometer by 3025 program and wavelength of k = 890 nm.
8. Calculate the content of phosphorous-pentoxide given by the formula.
Model equation:
(a) Mass fraction wP2 O5 in % (m m−1)
c1PO3 ðmg L1 Þ 0:75 2
WP2 O5 ð%Þ ¼ 4
msample ðgÞ R
where:
4 )
(b) Comprehensive equation for mass fraction wP2 O5 in % (m m−1)

WP2 O5 ð%Þ ¼
10001 10 V2 m1 10002 R
where:
wP2O5 Mass fraction wP2O5 in % (m m−1)
F Concentration factor M(P2O5)/2M(PO−34 )= 0.75
86 T. Muhić-Šarac

R Recovery
Rrep Repeatability

Uncertainty of concentration of reference solutions
Repeatability
Uncertainty of recovery
Intermediate checks of spectrophotometer
Other: Volume of the analysed solutions
Other: Measurements of sample mass
Other:


P 0.9995 – Purity of KH2PO4
V1 1000.0 mL Volume of stock standard solution KH2PO4
V2 1.0 mL Volume of aliquot standard solution KH2PO4
V3 200.0 mL Volume of working standard solution KH2PO4
V4 10.0 mL Volume of aliquot working standard solution KH2PO4
m1 426.6 mg Mass of KH2PO4
F 1.43 – Concentration factor M(KH2PO4)/M(PO−3 4 )

m1(sample) 5000.0 mg Mass of sample meat product
C1 1.22 mg L−1 Concentration of PO−3 −1
V1 100.0 mL Volume of volumetric flask
V2 1.0 mL Volume of sample solution aliquot
V4 10.0 mL Volume of aliquot
F 0.75 – Concentration factor M(P2O5)/2M(PO−34 )
R 1.016 – Recovery
Rrep 1 % Repeatability
Icheck 1 mg L−1 Intermediate checks of spectrophotometer


P 0.0003 – Purity of KH2PO4
V1 0.507 mL Volume of stock standard solution KH2PO4
V2 0.004 mL Volume of aliquot standard solution KH2PO4
V3 0.117 mL Volume of working standard solution KH2PO4
V4 0.043 mL Volume of aliquot working standard solution
KH2PO4
m1 0.085 mg Mass of KH2PO4
F 0.000013 – Concentration factor M(KH2PO4)/M(PO−3 4 )

m1(sample) 0.085 mg Mass of sample meat product
C1 0.002 mg L−1 Concentration of PO−3 −1
4 (mg L ) in meat
product
V2 0.004 mL Volume of sample solution aliquot
V4 0.043 mL Volume of aliquot
F 0.00001 – Concentration factor M(P2O5)/2M(PO−34 )
R 0.0117 – Recovery
Rrep 0.015 % Repeatability
Icheck 0.009 mg L−1 Intermediate checks of spectrophotometer
88 T. Muhić-Šarac
6. Calculate the value of the measurand using the model equation

Intermediate checks of spectrophotometer – in one point
Concentration of PO−3
4 in working standard solution:
mðKH2 PO4 Þ P 10 1000 V2

cPO3 ðmg L1 Þ ¼
4 V4 V3 V1 F
Comprehensive equation for mass fraction wP2 O5 in %:

WP2 O5 ð%Þ ¼
10001 10 V2 m1 10002 R
units
Software

P 0.9995 0.0003 – Purity of KH2PO4
V1 1000.0 0.507 mL Volume of stock standard solution
KH2PO4
V2 1.0 0.004 mL Volume of aliquot standard solution
KH2PO4
V3 200.0 0.117 mL Volume of working standard solution
KH2PO4
V4 10.0 0.043 mL Volume of aliquot working standard
solution KH2PO4
m1 426.6 0.085 mg Mass of KH2PO4
F 1.43 0.000013 – Concentration factor M(KH2PO4)/M
(PO−3
4 )
u(Int.checks) = 0.009 mg L−1 PO−3
4

m1 5000.0 0.085 mg Mass of sample meat product
(sample)
C1 1.22 0.002 mg L−1 Concentration of PO−3 −1
4 (mg L ) in
meat product
V1 100.0 0.068 mL Volume of volumetric flask
V2 1.0 0.004 mL Volume of sample solution aliquot
V3 200.0 0.117 mL Volume of volumetric flask
V4 10.0 0.043 mL Volume of aliquot
F 0.75 0.00001 – Concentration factor M(P2O5)/2M
(PO−3
4 )
R 1.016 0.0117 – Recovery
Rrep 1 0.015 % Repeatability
Icheck 1 0.009 mg L−1 Intermediate checks of
spectrophotometer
u(w(P2O5) = 0.0078% (m m−1)
the units
UðwðP2 O5 ÞÞ ¼ k uðwðP2 O5 ÞÞ ¼ 2 0.0078 ¼ 0.016% (m m1 )
1 Repeatability – contributing 47.52% to the expanded uncertainty
2 Recovery – contributing 27.37% to the expanded uncertainty
3 Uncertainty of intermediate checks of spectrophotometer – contributing 17.11% to the
expanded uncertainty

90
Addendum I: Measurement Uncertainty Calculation: Spreadsheet Approach (Excel)

Unit Parameter P(KH2PO4) m1 V1 V2 V3 V4 F

value 0,9995 426,6 1000 1 200 10 1,43
uncertainty 0,0003 0,085 0,507 0,004 0,117 0,043 0,000013
P(KH2PO4) 0,9995 0,9998 0,9995 0,9995 0,9995 0,9995 0,9995 0,9995

mg m1 426,6 426,6 426,685 426,6 426,6 426,6 426,6 426,6
mL V1 1000 1000 1000 1000,507 1000 1000 1000 1000
mL V2 1 1 1 1 1,004 1 1 1
mL V3 200 200 200 200 200 200,117 200 200
mL V4 10 10 10 10 10 10 10,043 10
F 1,43 1,43 1,43 1,43 1,43 1,43 1,43 1,430013
1,490862587 1,490863 1,490863 1,490863 1,490863 1,490863 1,490863
mg/L C (phosph.) 1,490863 1,49131007 1,49116 1,490107 1,496826 1,489991 1,484479 1,490849
u(y,xi) 0,000447483 0,000297 -0,00076 0,005963 -0,00087 -0,00638 -1,4E-05
u(y)2 7,79E-05 2,00241E-07 8,82E-08 5,71E-07 3,56E-05 7,6E-07 4,07E-05 1,84E-10
%u(xi)2 0,256956081 0,113234 0,732416 45,63541 0,974958 52,28679 0,000236
u(C(phos.)) 0,008828 0,000447483 0,000297 0,000755 0,005963 0,000872 0,006383 1,36E-05
%u(xi)e2
V4
P
F
m( KH2 PO4 ) P 10 1000 V2
m1
cPO 3 (mg / L)
4
V4 V3 V1 F
V1
V3
T. Muhić-Šarac
V2
Unit Parameter m1 c1 V1 V2 V3 V4 F R Rrep. Icheck

value 5000 1,22 100 1 200 10 0,75 1,016 1 1
uncertainty 0,085 0,002 0,068 0,004 0,117 0,043 0,00001 0,0117 0,015 0,009
mg m1 5000 5000,085 5000 5000 5000 5000 5000 5000 5000 5000 5000
mg/L c1 1,22 1,22 1,222 1,22 1,22 1,22 1,22 1,22 1,22 1,22 1,22
mL V1 100 100 100 100,068 100 100 100 100 100 100 100
mL V2 1 1 1 1 1,004 1 1 1 1 1 1
mL V3 200 200 200 200 200 200,117 200 200 200 200 200
mL V4 10 10 10 10 10 10 10,043 10 10 10 10
F 0,75 0,75 0,75 0,75 0,75 0,75 0,75 0,75001 0,75 0,75 0,75
R 1,016 1,016 1,016 1,016 1,016 1,016 1,016 1,016 1,0277 1,016 1,016
% Rrep. 1 1 1 1 1 1 1 1 1 1,015 1
Icheck 1 1 1 1 1 1 1 1 1 1 1,009
0,360236 0,360236 0,360236 0,360236 0,360236 0,360236 0,360236 0,360236 0,360236 0,360236
%(m/m) P 2O 5 0,360236 0,360230 0,360827 0,360481 0,358801 0,360447 0,361785 0,360241 0,356135 0,36564 0,363478
Measurement of Polyphosphates in Meat Products …
u(y,xi) -6,1E-06 0,000591 0,000245 -0,00144 0,000211 0,001549 4,8E-06 -0,0041 0,005404 0,003242
u(y)2 6,14E-05 3,75E-11 3,49E-07 6E-08 2,06E-06 4,44E-08 2,4E-06 2,31E-11 1,68E-05 2,92E-05 1,05E-05
%u(xi)2 6,1E-05 0,567613 0,097663 3,352465 0,072281 3,905248 3,75E-05 27,37478 47,52195 17,1079
u(%P2O5) 0,007838 6,12E-06 5,91E-04 2,45E-04 1,44E-03 2,11E-04 1,55E-03 4,80E-06 4,10E-03 5,40E-03 3,24E-03
%u(xi) 2
m1 V1 V2 V4
Icheck c1 V3 F c1 V4 V3 V1 100 F Rrep. I check
R
WP2O5 (%)
10001 10 V2 m1 10002 R
Rrep.
91
92 T. Muhić-Šarac
Addendum II: Measurement Uncertainty Calculation –

GumWorkbench
Evaluation of Standard Uncertainty for Separate Parameters

of the Intermediate Checks of Spectrophotometer
Preparation of the standard solution

V1 – 1000 mL volumetric flask
The volume has two major influences; calibration and temperature effects.
ffiffi ¼ 0:16 mL
(i) Calibration u ¼ 0:4pmL
6
(ii) Temperature ð1000 4 2:1 104 Þ ¼ 0:84 mL
0:84 mL
u¼ pffiffiffi ¼ 0:5 mL
3
The two contributions are combined to give the standard uncertainty u(V) of
the V.
Finally, total standard uncertainty of the 1000 mL volumetric flask is:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðV1 Þtotal ¼ 0:162 þ 0:52 ¼ 0:507 mL
V2 – 1 mL pipette
The volume has three major influences; calibration, repeatability and temperature
effects.
ffiffi ¼ 0:004 mL
6
(ii) Repeatability u ¼ S ¼ 0.0013 mL
(iii) Temperature ð1 4 2:1 104 Þ ¼ 0:0008 mL
0:0008 mL
u¼ pffiffiffi ¼ 0:0005 mL
3
The three contributions are combined to give the standard uncertainty u(V) of
the V.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðV2 Þtotal ¼ 0:0042 þ 0:00132 þ 0:00052 ¼ 0.004 mL

effects.
ffiffi ¼ 0:06 mL
6
0:168 mL
u¼ pffiffiffi ¼ 0:097 mL
3
the V.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðV3 Þtotal ¼ 0:062 þ 0:0252 þ 0:0972 ¼ 0.117 mL
V4 – 10 mL pipette
effects.
ffiffi ¼ 0:04 mL
6
0:0084 mL
u¼ pffiffiffi ¼ 0:005 mL
3
the V.
uðV4 Þtotal ¼ 0:042 þ 0:0152 þ 0:0052 ¼ 0.043 mL
m1 – mass of the sample

For the precision the scale the producer lists the value of ±0.1 mg. The rectangle
division of the dispersed values is being evaluated. and calculation of the standard
uncertainty is carried out in the following way:
0:1 mg
u ¼ pffiffiffi ¼ 0:06 mg
3
Finally the combined uncertainty acquires the following value:

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðm1 Þ ¼ 2 ð0:06Þ2 ¼ 0.085 mg
(0.06 multiplied with 2 because we have m(tare) and m(weight))

94 T. Muhić-Šarac
P – purity of KH2PO4
The purity of the KH2PO4 is given on the certifikate as 99.95 ± 0.05% m m−1.
Purity is therefore 0.9995 ± 0.0005. Because there is no additional information
about the uncertainty value, a rectangular distribution is assumed:
0:0005
u(P) ¼ pffiffiffi ¼ 0:0003
3
F – concentration factor M(KH2PO4)/M(PO−3 4 )

From the latest IUPAC table, the atomic weights and listed uncertainties for the
constituent K, H, P and O of KH2PO4 and PO−34 are:
Element Atomic weight Quoted uncertainty Standard uncertainty

K 39.0983 ±0.0001 0.000058
H 1.00794 ±0.00007 0.000040
P 30.9738 ±0.0005 0.00029
O 15.9994 ±0.0003 0.00017
For each element, the standard uncertainty is found by the IUPAC quoted
uncertainty as forming the bounds of a rectangular distribution.
The separate element contributions to the molar mass, together with the uncer-
tainty contribution for each, are:
Element Calculation Result Standard uncertainty

K 1 39.0983 39.0983 0.000058
H2 2 1.00794 2.01588 0.000080
P 1 30.9738 30.9738 0.00029
O4 4 15.9994 63.9976 0.00068
This gives a molar mass for KH2PO4 of
MðKH2 PO4 Þ ¼ 39:0983 þ 2:01588 þ 30:9738 þ 63:9976 ¼ 136:0856 g mol1 ;

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðMðKH2 PO4 ÞÞ ¼ 0:0000582 þ 0:000082 þ 0:000292 þ 0:000682
¼ 0:0007 g mol1

P 1 30.9738 30.9738 0.00029
O4 4 15.9994 63.9976 0.00068
This gives a molar mass for PO−3

4 of

M PO3
4 ¼ 30:9738 þ 63:9976 ¼ 94:9714 g mol1 ;
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u M PO3
4 ¼ 0:000292 þ 0:000682 ¼ 0:0007 g mol1
Combined uncertainty of concentration factor is:

M KH2 PO4
F¼ ¼ 1:43
M PO34
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u0 12
u u M 3 2
uð F Þ u
A þ uðMKH2 PO4 Þ
PO
¼ t@
4
F MPO3 4
MKH2 PO4
u(F)
¼ 0:000009
1:43
u(F) ¼ 1:43 0:000009 ¼ 0.000013
Evaluation of Standard Uncertainty for Separate Parameters

of the Analysis

effects.
ffiffi ¼ 0:04 mL
6
0:084 mL
u¼ pffiffiffi ¼ 0:05 mL
3
the V.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðV1 Þtotal ¼ 0:042 þ 0:022 þ 0:052 ¼ 0.068 mL
V2 – 1 mL pipette
effects
96 T. Muhić-Šarac
ffiffi ¼ 0:004 mL
6
0:0008 mL
u¼ pffiffiffi ¼ 0:0005 mL
3
the V.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðV2 Þtotal ¼ 0:0042 þ 0:00132 þ 0:00052 ¼ 0.004 mL

effects
ffiffi ¼ 0:06 mL
6
0:168 mL
u¼ pffiffiffi ¼ 0:097 mL
3
the V.
uðV3 Þtotal ¼ 0:062 þ 0:0252 þ 0:0972 ¼ 0.117 mL
V4 – 10 mL pipette
effects.
ffiffi ¼ 0:04 mL
6
0:0084 mL
u¼ pffiffiffi ¼ 0:005 mL
3
the V.

uðV4 Þtotal ¼ 0:042 þ 0:0152 þ 0:0052 ¼ 0.043 mL
m1 – mass of the sample

For the precision the scale the producer lists the value of ±0.1 mg. The rectangle
division of the dispersed values is being evaluated. and calculation of the standard
uncertainty is carried out in the following way:
0:1 mg
u ¼ pffiffiffi ¼ 0:06 mg
3
Finally, the combined uncertainty acquires the following value:

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
uðm1 Þ ¼ 2 ð0:06Þ2 ¼ 0.085 mg
(0.06 multiplied with 2 because we have m(tare) and m(weight))

F – concentration factor (M(P2O5)/2M(PO−3 4 ))
From the latest IUPAC table, the atomic weights and listed uncertainties for the
constituent P and O of PO−3
4 and P2O5 are:
Element Atomic weight Quoted uncertainty Standard uncertainty

P 30.9738 ±0.0005 0.00029
O 15.9994 ±0.0003 0.00017
For each element, the standard uncertainty is found by the IUPAC quoted
uncertainty as forming the bounds of a rectangular distribution.

P 1 30.9738 30.9738 0.00029
O4 4 15.9994 63.9976 0.00068
This gives a molar mass for PO−3

4 of
MðPO3 1
4 Þ ¼ 30:9738 þ 63:9976 ¼ 94:9714 g mol ;
u M PO3
4 ¼ 0:000292 þ 0:000682 ¼ 0.0007 g mol1
98 T. Muhić-Šarac

P2 2 30.9738 61.9476 0.00058
O5 5 15.9994 79.9970 0.00085
This gives a molar mass for P2O5 of
M ðP2 O5 Þ ¼ 61:9476 þ 79:9970 ¼ 141:9446 g mol1

uðMP2O5 Þ ¼ 0:000582 þ 0:000852 ¼ 0:001 g mol1
Combined uncertainty of concentration factor is:
MðP2 O5 Þ
F¼ ¼ 0:75
2 M PO3 4
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u 0 12
u
uð F Þ u u M PO4 3
uð M P 2 O 5 Þ 2
t
¼ 2x @ A þ
F MPO3 4
MP2 O5
u(F)
¼ 0:000012
0:75
u(F) ¼ 0:75 0:000012 ¼ 0:000009 g mol1 ¼ 0.00001 g mol1
Tidža Muhić-Šarac started work as a researcher at the motor

factory Famos, Sarajevo (Bosnia and Herzegovina). She finished
her Ph.D. study in 1999 at the Faculty of Science, Department of
Chemistry, University of Sarajevo (Bosnia and Herzegovina)
with her thesis ‘Extractability of amounts of metals (Fe, Mn, Cu
and Zn) from soils in Bosnia and Herzegovina’. She has con-
tinued her analytical work in the quality control of metals, alloys,
water, soils and food. Currently, Tidža Muhić-Šarac is employed
at the Faculty of Science, Department of Chemistry, University
of Sarajevo (Bosnia and Herzegovina), as an full professor.
A major part of her work is teaching analytical chemistry,
environmental chemistry, quality in analytical chemistry and
metrology in chemistry. Her research interests include experi-
mental and theoretical studies of species metals from water and
soil. Tidža Muhić-Šarac is the TrainMiC® team leader for Bosnia
and Herzegovina and since 2000 she has also been associated to BATA, the Institute of
Accreditation BiH, as technical assessor and technical expert for measurement uncertainty issues.
Task for a Laboratory Measurement
of Mass Fraction of TiO2 and Fe-tot
in Ilmenite and Titanium Slag
Jurij Pustinek, Karmen Rajer Kanduč and Nineta Hrastelj
1.1 Purpose
This procedure describes titrimetric determination of mass fraction of titanium

dioxide (TiO2) and total iron content (Fe-tot) in Ilmenite and Titanium Slag.
Measuring range:
Ilmenite Titanium Slag

TiO2 from 30 to 60% from 60 to 90%
Fe-tot from 20 to 40% from 5 to 20%

J. Pustinek (&) K. R. Kanduč

Cinkarna Celje, Kidričeva 26, 3000 Celje, Slovenia
e-mail: jurij.pustinek@cinkarna.si
K. R. Kanduč
e-mail: karmen.kanduc@cinkarna.si
N. Hrastelj
EuCheMS, Rue du Trone 62, 1050 Brussels, Belgium
e-mail: nineta@euchems.eu

https://doi.org/10.1007/978-3-030-20347-4_5
100 J. Pustinek et al.
1.2 Principle
The method is comprised of two parts. In the first the solution of potassium
dichromate is standardized by using the certified reference material (CRM Ilmenite
or CRM Titanium Slag), and in the second part the mass fraction of TiO2 and Fe-tot
is determined in a sample of Ilmenite or Ti-Slag by titration. Determination of the
equivalence points is potentiometric using a gold indicator electrode and an Ag/
AgCl reference electrode.
The sample of Ilmenite or Ti-Slag is dried at 110 °C to constant mass. Dried
sample is fused with melted potassium pyrosulfate and the melt is dissolved in
hydrochloric acid. By addition of the excess of the solution of chromium(II)
chloride Ti4+ is reduced to Ti3+ and Fe3+ is reduced to Fe2+ in the inert atmosphere
of carbon dioxide.
With potassium dichromate is titrated the firstly exceeded chromium(II) chloride
(1st equivalence point), then Ti3+ (2nd equivalence point) and then Fe2+ (3rd
equivalence point).
1.3 Chemicals
P.a. chemicals and deionised water are used.

1.3.1. Hydrochloric acid, 37% (q 1.19 g mL−1)
1.3.2. Sulphuric(VI) acid, 96% (q 1.84 g mL−1)
1.3.3. Sulphuric(VI) acid, diluted (10%)
• Approximately 500 mL deionised water is added in 1000 mL flask.

• Slowly and carefully 58.0 mL of concentrated sulphuric(VI) acid is added
(1.3.2.).
• The solution is cooled, flask is filled to the mark with deionised water and all is
well mixed.
1.3.4. Solution of chromium(III) chloride, c(CrCl3 6 H2O) = 0.3 mol L−1
• 79.93 g of chromium(III) chloride hexahydrate is added to 1000 mL flask.

• 300 mL of deionised water and 100 mL of hydrochloric acid (1.3.1.) are added,
to dissolve all chromium(III) chloride.
• The solution is cooled, flask is filled to the mark with deionised water and all is
well mixed.
Task for a Laboratory Measurement of Mass Fraction … 101
1.3.5. Zinc amalgam

WARNING! Mercury is poisoned. You should be careful not to get it in
touch with skin and you should avoid breathing its vapour. Zinc
amalgam is therefore prepared in fumehood.
• 50 mL of Hg is put with a measuring cylinder in porcelain vessel.

• Surface of Hg is then covered with diluted sulphuric(VI) acid (1.3.3.); the
porcelain vessel is put on waterbath in fumehood.
• 20–30 g of pure zinc in small granules is added.
• We stir on occasions. When needed, diluted acid is replaced by deionised water.
• When all solid zinc is used, amalgam is removed from the waterbath and we left
it in the fumehood during the night.
• Amalgam is filtrated through the filter paper white band, in which we made a
small hole.
• Amalgam is stored in 250 mL laboratory flask, under a layer of diluted sulphuric
(VI) acid (1.3.3.).
• Amount of zinc amalgam that we get from 50 mL Hg is sufficient for a
reduction of approximately 800 mL of chromium(III) chloride (1.3.4.). Used
zinc amalgam can be cleaned and re-used. Instead of 50 mL of Hg we add in a
porcelain vessel 50 mL of used zinc amalgam and continue the procedure as
described above (1.3.5.).
1.3.6. Paraffin oil
1.3.7. Solution of chromium(II) chloride, c(CrCl2) = 0.3 mol L−1
• Zinc amalgam (1.3.5.) is poured in Nakazon’s reductor.

• Amalgam is separated from sulphuric(VI) acid. Amalgam, which is on the
bottom of the reductor, is poured in measuring cylinder, and then sulphuric(VI)
acid is poured in the laboratory bottle.
• Amalgam is poured in Nakazon’s reductor.
• 200 mL of chromium(III) chloride (1.3.4.) is put in a beaker and solution is
heated on an electrical heater to 60 °C.
• The hot solution is transferred into reductor and blowed with CO2 for 3 min.
• We close the valve, which connects reductor with air. Then we close the valve,
which brings CO2 into reductor. This is how an inert atmosphere is maintained.
• Solution in a reductor is shaken for about 5 min. When a reduction is finished,
the colour of a solution is changed from green to intensive blue.
• CO2 is transferred into reductor.
• Amalgam is poured in a laboratory bottle which contained a diluted sulphuric
(VI) acid.
Last drops of amalgam are collected in a special vessel for waste.
• Chromium(II) chloride is poured under a layer of paraffin oil (3.6.) in a brown
titration flask.
• If needed, additional amounts of chromium(III) chloride can be reduced and

aliquots collected in a titration bottle.
Solution of chromium(II) chloride should be prepared one day before the
titration (measurement).
Solution of chromium(II) chloride under paraffin oil is usable for one month.
1.3.8. Solution of potassium dichromate, c(K2Cr2O7) = 0.01 mol L−1
• Solid potassium dichromate is dried at 110 ± 2 °C to the constant mass (min

2 h).
• Dried potassium dichromate is cooled in the exicator.
• 5.8836 g dried potassium dichromate, weighted to 0.1 mg is quantitatively
transferred to 2000 mL flask.
• Potassium dichromate is dissolved in deionised water; the flask is filled with
deionised water up to the mark and well mixed.
1.3.9. Solution of potassium dichromate, c(K2Cr2O7) = 0.02 mol L−1
• Solid potassium dichromate is dried at 110 ± 2 °C to the constant mass (min

2 h).
• Dried potassium dichromate is cooled in the exicator.
• 11.7672 g dried potassium dichromate weighted to 0.1 mg is quantitatively
transferred to 2000 mL flask.
• Potassium dichromate is dissolved in deionised water; the flask is filled up to the
mark with deionised water and well mixed.
1.3.10. Potassium pyrosulfate (K2S2O7), fine powder

1.3.11. Carbon dioxide (CO2), gas of purity 4.8 (99.998%)
1.3.12. CRM Ilmenite SARM 59, Mintek
1.3.13. CRM Titanium Slag SARM 58, Mintek
1.4 Equipment
• Analytical balance, accuracy: ±0.0003 g (calibrated by NMI)

• Precise balance, accuracy: ±0.03 g (calibrated by NMI)
• Drying oven, 110 ± 2 °C
• Exicator
• Erlenmeyer flasks, 300 mL
• Bunsen’s burner
• Electrical heater
• PTFE watch glass
• Metal nets with mesh of a diameter of 7.6 cm
• Nakazon’s reductor
• Automatic titrator Mettler DL70ES connected to PC and printer (calibrated by
producer)
• Gold (Au) indicator electrode
• Reference Ag/AgCl electrode
• Glass titration flasks, 250 mL
• Burettes Mettler (calibrated by producer)
• Gas flow meter, 400 ± 20 mL/min
• Volumetric flask − class A quality
• Volumetric pipette − class A quality
• Fumehood
1.5 Sample and CRM Preparation
• Sample and certified reference material are dried at 110 ± 2 °C to the constant
mass (min 2 h).
• It is cooled in the exicator to the room temperature.
1.6 Procedure
1.6.1. Fusion – Dissolution
• 30 ± 1 g of potassium pyrosulfate (1.3.10.) is added in 300 mL Erlenmeyer

flask.
• Potassium pyrosulfate is melted above the fire and cooled.
• 0.500–0.525 g of sample is weighted.
• Sample is transferred on the layer of melted potassium pyrosulfate and its mass
is noted down to 0.1 mg.
• Sample is equally distributed over the whole area of melted potassium pyro-
sulfate by gentle shaking.
• Erlenmeyer flask is covered with PTFE watch glass, put on a net and gently heat
above the fire.
• When potassium pyrosulfate melts, the heating is slowly increased until the
maximum is reached.
• Erlenmeyer flask is occasionally gently shaken.
• Heating is on-going as long as sample is not fused (dissolved).
• At the edge of Erlenmeyer flask, the melt starts to get solid fast, and hence we
continue heating directly above the flame with two burners.
• We let it cool down.
• 70 mL of 37% hydrochloric acid (1.3.1.) and 140 mL of deionised water are

added.
• It is heated on the electrical heater until it boils and we let it boil as long as the
melt is not dissolved.
• We let the solution to cool a bit.
• Still warm solution is quantitatively transferred into 500 mL flask and let it cool
to room temperature.
• Flask is filled with deionized water and well mixed.
1.6.2. Standardization of Potassium Dichromate

Solution of potassium dichromate (1.3.8., 1.3.9.) is daily standardised with a cer-
tified reference material in four parallels.
• Following the above mentioned procedure 5. (CRM preparation) and 1.6.1.

(Fusion – Dissolution) 4 dissolutions of certified reference materials are made
(1.3.12.) – for Ilmenite and (1.3.13.) – for Titanium Slag.
• From 500 mL flasks 100 mL of solution is pipetted in titration flasks.
• 50 mL 10% H2SO4 (1.3.3.) and 5 mL 37% HCl (1.3.1.) is added in each flask.
• A burette is mounted on a titrator; burette is filled with 0.02 mol L−1 K2Cr2O7
(1.3.9.) – for Ilmenite, and with 0.01 mol L−1 K2Cr2O7 (1.3.8.) – for
Titanium Slag. Another burette with CrCl2 (1.3.7.) is added and well rinsed.
• Gold indicator electrode, a reference electrode and a mixer are connected to the
titrator.
• CRM solution is heated on the electrical heater to (50–70) °C.
• Titration flask is fixed on the titrator and wrapped with Al-folio.
• Gas flow (1.3.11.) is set to 400 mL/min.
• Standardization is done via PC with 4 dissolutions of a certified reference
material.
Titrator automatically calculates average value of four dissolutions expressed as:
g TiO2 mL−1 K2Cr2O7 (value H8) and g Fe mL−1 K2Cr2O7 (value H9) – for
Ilmenite and
g TiO2 mL−1 K2Cr2O7 (value H11) and g Fe mL−1 K2Cr2O7 (value H12) – for
Titanium Slag.
1.6.3. Determination
• 100 mL of sample solution is transferred from 500 mL flask to titration flask.

• 50 mL 10% H2SO4 (1.3.3.) and 5 mL 37% HCl (1.3.1.) are added.
• Sample solution is heated on an electrical heater to (50–70) °C.
• Titration flask is fixed on titrator and wrapped with Al-folio.
• Gas flow (1.3.11.) is set to 400 mL/min.
• Determination of mass fraction of TiO2 and Fe-tot in sample is done via PC
according to the methods IL01 – for Ilmenite and ZL01 – for Titanium Slag.
1.6.4. Control of the Procedure

Control is done by a certified reference material in each series of measurements in
two parallel determinations. Result of the average value is inserted in a control
chart.
1.7 Reporting of Results
Samples are analysed in two parallels.

Average value of % TiO2 and % Fe-tot is given as a result to one decimal
place.
1.8 Measurement Uncertainty
Measurement uncertainty is calculated and given in the valid corresponding report.
1.9 Training and Quality Control
Training and checking competence of the personnel is done according to the cor-
responding internal procedure.
Quality is assured by:
• participation in a proficiency testing scheme

• quality control charts
• determinations in parallel
• checking competence of the personnel.
1.10 Safety
Safety instructions are used when working in the laboratory.


Relative expanded uncertainty (k = 2)

Ilmenite (%) Titanium slag (%)
TiO2 <1 <1
Fe-tot <5 <5
The value of the relative expanded uncertainty is defined as the level of uncertainty
that is broadly accepted as appropriate by the end-users of the data for the sector of
application of the results. A suggested definition of this »fitness-for-purpose« is
that it should comprise the uncertainty that minimized the combined costs of
analysis and the financial penalties associated with incorrect decisions multiplied by
their probabilities of occurrence.

The accuracy of test procedure was checked and evaluated by measuring the
appropriate CRM:
• CRM SARM 59 – Ilmenite
• CRM SARM 57 – Titanium Slag (Table 1).
Table 1 Reproducibility of TiO2 and Fe-tot measurements on CRM (from control charts)
CRM SARM 59 Ilmenite CRM SARM 57 Titanium
Slag
TiO2 [%] Fe-tot [%] TiO2 [%] Fe-tot [%]
n* 30 30 30 30
Average 48.81 35.23 85.34 8.24
SD 0.069 0.060 0.068 0.053
RSD [%] 0.14 0.17 0.08 0.64
CV** 48.8 35.2 85.4 8.25
95% CL*** 48.6–49.1 34.9–35.6 85.1–85.7 8.0–8.4
Recovery [%] 100.0 100.1 99.9 99.9
*Number of measurements
**Certified Value
***95% Confidence Limits
The laboratory repeatability of TiO2 and Fe-tot measurement was evaluated by

measuring different test samples of Ilmenite and Titanium Slag under repeatability
conditions.
The laboratory reproducibility of TiO2 and Fe-tot measurement was evaluated by
measuring different test samples of Ilmenite and Titanium Slag under repro-
ducibility conditions (Tables 2, 3, 4 and 5).
Table 2 Repeatability (reproducibility) of TiO2 measurements on Ilmenite test samples

TiO2 [%]
Sample 1 2 3 4 5 6 7**
n* 8 6 6 6 6 6 8
Average 54.40 52.60 52.11 52.19 54.95 52.28 53.75
SD 0.103 0.062 0.075 0.068 0.028 0.168 0.098
RSD [%] 0.19 0.12 0.14 0.13 0.05 0.32 0.18
**Reproducibility conditions
Table 3 Repeatability (reproducibility) of Fe-tot measurements on Ilmenite test samples

Fe-tot [%]
Sample 1 2 3 4 5 6 7**
n* 8 6 6 6 6 6 8
Average 30.10 32.01 32.17 32.13 27.38 32.98 30.61
SD 0.084 0.081 0.121 0.082 0.061 0.115 0.085
RSD [%] 0.28 0.25 0.38 0.25 0.22 0.35 0.28
Table 4 Repeatability (reproducibility) of TiO2 measurements on Ti-Slag test samples

TiO2 [%]
Sample 1 2 3 4 5 6**
*
n 8 6 6 6 6 8
Average 80.23 79.19 79.29 78.92 78.87 76.66
SD 0.098 0.076 0.147 0.069 0.095 0.066
RSD [%] 0.12 0.10 0.19 0.09 0.12 0.09
Table 5 Repeatability (reproducibility) of Fe-tot measurements on Ti-Slag test samples

Fe-tot [%]
Sample 1 2 3 4 5 6**
*
n 8 6 6 6 6 8
Average 7.44 8.07 8.07 6.03 8.02 7.49
SD 0.035 0.036 0.048 0.050 0.053 0.032
RSD [%] 0.48 0.45 0.60 0.84 0.66 0.42
Measurements were tested with the Grubbs test and were found to have no
outliers.
Laboratory reproducibility of TiO2 and Fe-tot measurement does not differ
substantially from laboratory repeatability.
Standard deviations of measurement for different samples do not differ sub-
stantially, that is why we can calculate spooled:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ðn1 1Þ s21 þ ðn2 1Þ s22 þ
spooled ¼
ðn1 1Þ þ ðn2 1Þ þ
The comparability of the TiO2 content results in Ilmenite and Ti-Slag was
checked with the inter laboratory comparison, organized by the Quality Control
department of Cinkarna Celje. 5 laboratories from Europe participated in this
inter-laboratory comparison (“Measurements of various parameters in some TiO2
based materials, March 2011”).
With the inter laboratory comparison we have confirmed the comparability of the
TiO2 content results in Ilmenite and Ti-Slag (z = −0.32 and z = 0.06) (Tables 6
and 7).
Table 6 Results of measurements of TiO2 content in Titanium Slag – inter laboratory comparison
Assigned value Xa 76.61%
Uncertainty on the assigned value u(Xa) 0.068%
Standard deviation of the proficiency assessment rpa 0.77%
Participant’s Result Standard Number of Method Reported z− f−
Code (Sample – X [%] deviation measurements standard score score
ID) [%] uncertainty [%]
CC-02/10 76.51 0.160 9 EN ISO 591-1 0.123 −0.13 −0.69
CC-02/11 76.20 0.046 3 UV-VIS 0.09 −0.53 −3.61
CC-02/12 76.70 0.086 3 Titration 0.07 0.11 0.88
CC-02/13 76.71 0.155 3 XRF 0.2 0.13 0.49
CC-02/14* 76.66 0.066 8 Potentiometric 0.1 0.06 0.38
titration
*Our result
Table 7 Results of measurements of TiO2 content in Ilmenite – inter laboratory comparison

Assigned value Xa 53.92%
Uncertainty on the assigned value u(Xa) 0.060%
Standard deviation of the proficiency assessment rpa 0.54%
Participant’s Result Standard Number of Method Reported z− f−
Code (Sample – [%] deviation measurements standard score score
ID) [%] uncertainty [%]
CC-03/10 53.90 0.151 8 EN ISO 591-1 0.126 −0.03 −0.12
CC-03/11 53.87 0.050 3 UV-VIS 0.09 −0.12 −0.61
CC-03/12 54.00 0.052 3 Titration 0.07 0.14 0.85
CC-03/13 54.04 0.081 3 XRF 0.2 0.21 0.55
CC-03/14* 53.75 0.098 8 Potentiometric 0.1 −0.32 −1.50
titration
*Our result

Mass fraction of TiO2 and Fe-tot is calculated by:
Ilmenite
ðVEX1 þ VEQ2 ÞH8 V500 mCRM W CRM V100

WTiO2 ¼ ms V100 100 ½% H8 ¼ ðVEX1 þ VEQ2TiOÞV
2
500 100
½g mL1 g TiO2 mL1 K2 Cr2 O7
WFe-tot ¼
ðVEX2 þ VEQ3 ÞH9 V500
100 ½% H9 ¼
mCRM WFe
CRM
-tot V100 ½g mL1 g Fe mL1 K2 Cr2 O7
ms V100 ðVEX2 þ VEQ3 ÞV500 100
Ti-Slag

2
500 100
½g mL1 ðg TiO2 mL1 K2 Cr2 O7 Þ
ðVEX2 þ VEQ3 ÞH12 V500 m WFe-tot V100
CRM
1
1

WFe-tot ¼ ms V100 100 ½% H12 ¼ ðVEX2CRM
þ VEQ3 ÞV500 100 ½gmL g Fe mL K2 Cr2 O7
VEX1 excess volume of K2Cr2O7 1st titration [mL]

VEX2 excess volume of K2Cr2O7 2nd titration [mL]
VEQ2 volume of K2Cr2O7 used for Ti3+ titration (2nd titration) [mL]
VEQ3 volume of K2Cr2O7 used for Fe2+ titration (3rd titration) [mL]
V500 volume of dissolved sample or CRM (500 mL flask) [mL]
V100 volume of sample aliquot or CRM aliquot (100 mL pipette) [mL]
m mass of sample or CRM [g]
W mass fraction [%]
H standardization parameter [g mL−1].
The repeatability of the entire process combines all the repeatability of indi-
vidual sources of uncertainty. Because we conduct all the analysis in two parallel
determinations, we have divided the standard deviations of measurements spooled
pffiffiffi
with 2 to obtain the standard uncertainty of the double determination.
Determination of mass fraction of TiO2 in Ilmenite [%]
Input quantity Value Unit Standard

uncertainty
V100 Volume of sample aliquot 100 mL 0.0489
V500 Volume of dissolved sample 500 mL 0.2086
ms Mass of sample 0.5101 g 0.00015
H8 Standardization of K2Cr2O7 with 0.009585 g mL−1 0.0000298
CRM Ilmenite
VEX1 þ VEQ2 Volume of K2Cr2O7 used for 5.7155 mL 0.00836
titration
Determination of mass fraction of Fe-tot in Ilmenite [%]

uncertainty
CRM Ilmenite
titration
Determination of mass fraction of TiO2 in Titanium Slag [%]

uncertainty
CRM Titanium Slag
titration
Determination of mass fraction of Fe-tot in Titanium Slag [%]

uncertainty
CRM Titanium Slag
titration
4.1 Equations
Ilmenite
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s
uðWTiO2 Þ uðms Þ 2 uðH8 Þ 2 uðV100 Þ 2 uðVEX1 þ VEQ2 Þ 2 uðV500 Þ 2 uðPTiO2 Þ 2
¼ þ þ þ þ þ
WTiO2 ms H8 V100 ðVEX1 þ VEQ2 Þ V500 WTiO2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

uðWFe-tot Þ uðms Þ 2 uðH9 Þ 2 uðV100 Þ 2 uðVEX2 þ VEQ3 Þ 2 uðV500 Þ 2 uðPFe-tot Þ 2
¼ þ þ þ þ þ
WFe-tot ms H9 V100 ðVEX2 þ VEQ3 Þ V500 WFe-tot
Ti-Slag
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
uðWTiO2 Þ uðms Þ 2 uðH11 Þ 2 uðV100 Þ 2 uðVEX1 þ VEQ2 Þ 2 uðV500 Þ 2 uðPTiO2 Þ 2
¼ þ þ þ þ þ
WTiO2 ms H11 V100 ðVEX1 þ VEQ2 Þ V500 WTiO2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
uðWFe-tot Þ uðms Þ 2 uðH12 Þ 2 uðV100 Þ 2 uðVEX2 þ VEQ3 Þ 2 uðV500 Þ 2 uðPFe-tot Þ 2
¼ þ þ þ þ þ
WFe-tot ms H12 V100 ðVEX2 þ VEQ3 Þ V500 WFe-tot
Excercise 1: Establishing Traceability in Analytical

Chemistry
Analyte TiO2
Fe-tot
Measurand Mass fraction of TiO2 in Ilmenite
Mass fraction of TiO2 in Titanium Slag
Mass fraction of Fe-tot in Ilmenite
Mass fraction of Fe-tot in Titanium Slag
Units %

equation
Measurement The method is comprised of two parts. In the first the solution of potassium
procedure dichromate is standardized by using the certified reference material (CRM
Ilmenite or CRM Titanium Slag), and in the second part the mass fraction of
TiO2 and Fe-tot is determined in a sample of Ilmenite or Ti-Slag by titration.
Determination of the equivalence points is potentiometric using a gold indicator
electrode and an Ag/AgCl reference electrode
chloride Ti4+ is reduced to Ti3+ and Fe3+ is reduced to Fe2+ in the inert
atmosphere of carbon dioxide
equivalence point)
Type of calibration*
standard curve standard addition internal standard
*None of the above
Model equation
Ilmenite

2
500 100
½g mL1 ðg TiO2 mL K2 Cr2 O7 Þ
ðVEX2 þ VEQ3 ÞH9 V500 m WFe-tot V100
CRM
1
1

WFe-tot ¼ ms V100 100 ½% H9 ¼ ðVEX2CRM
þ VEQ3 ÞV500 100 ½g mL g Fe mL K2 Cr2 O7
Ti-Slag

2
500 100
ðV þ V ÞH12 V500 mCRM WFe
CRM
V
WFe-tot ¼ EX2 msEQ3 V100 100 ½% H12 ¼ ðVEX2 þ VEQ3-ÞVtot 100
500 100
½g mL1 ðg Fe mL K2 Cr2 O7 Þ

W mass fraction [%]
H standardization parameter [g mL−1].

1 Standardization of K2Cr2O7 with CRM Ilmenite or CRM Titanium Slag – H
2 Volume of K2Cr2O7 used for titration – VEQ
3 Repeatability – P
4 Volume of dissolved sample – V500
5 Volume of sample aliquot – V100
For the analyte
1 Name/Chemical Formula/Producer: Ilmenite SARM 59/Mintek

2 Name/Chemical Formula/Producer: Titanium Slag SARM 58/Mintek
1 Quantity/Equipment/Calibration: Analytical and precise balance – calibrated

e.g. mass/balance/calibrated by NMI, by NMI
U = xx (k = 2), see also data yellow
sheet
2 Quantity/Equipment/calibration: Automatic titrator Mettler DL70ES with
gold indicator electrode and Ag/AgCl
reference electrode – calibrated by producer
(continued)
(continued)
3 Quantity/Equipment/calibration: Burettes Mettler – calibrated by producer
4 Quantity/Equipment/calibration: Volumetric flask – class A quality
5 Quantity/Equipment/calibration: Volumetric pipette – class A quality

Other important parameters are: Within-lab repeatability or
reproducibility

1 Analysis of matrix CRM
2 Participation in a proficiency testing scheme
7. Any other comments, questions…


Procedures

The measurement Titrimetric determination of mass fraction of titanium dioxide (TiO2)
procedure and total iron content (Fe-tot) in Ilmenite and Titanium Slag
Analyte TiO2
Fe-tot
The measurand Mass fraction of TiO2 in Ilmenite
Unit %

Matrix Ilmenite and Titanium Slag
Measuring range Analyte Ilmenite Titanium Slag
TiO2 from 30 to 60% from 60 to 90%
Fe-tot from 20 to 40% from 5 to 20%

Intended use of the results Quality control of Ilmenite and Titanium Slag (raw
materials)
LOD
LOQ
Repeatability
Within-lab
reproducibility
Trueness
Measurement TiO2: Urel < 1%
uncertainty Fe-tot: Urel < 5%
Other-state

VALIDATION
Part II: Parameters to Be Validated

Where yes, please give further information e.g. which CRM, reference method

Further information: CRM SARM 59 – Ilmenite, CRM SARM 57 – Titanium Slag
Participation in the inter laboratory comparison organized by the Quality Control
Department of Cinkarna Celje, “Measurements of various parameters in some TiO2
based materials, March 2011”
6. Measuring range
Linearity
Upper limit
LOD
LOQ
Repeatability
8. Robustness
9. Quality Control
Control charts

R square

Parameters requested to be validated Calculations
LOD
LOQ
Repeatability
Within-lab reproducibility
Trueness
Other-please state
use?
Parameter Value requested by the customer (the Value obtained The
same as stated in question 3) during validation requirement is
fulfilled
Yes/No
LOD
LOQ
Repeatability
Within-lab
reproducibility
Trueness
Measurement TiO2: Urel < 1% Ilmenite: Yes

uncertainty Fe-tot: Urel < 5% TiO2: Urel = 0.74%
Fe-tot: Urel = 1.2%
Ti-Slag:
TiO2: Urel = 0.54%
Fe-tot: Urel = 2.9%
Other

Yes No
For Measurement Uncertainty and Traceability refer to the corresponding sheets
Building An Uncertainty Budget

Measurand Mass fraction of TiO2 in Ilmenite
Unit %

equation
The method is comprised of two parts. In the first the solution of potassium
dichromate is standardized by using the certified reference material (CRM Ilmenite
or CRM Titanium Slag), and in the second part the mass fraction of TiO2 and Fe-tot
is determined in a sample of Ilmenite or Ti-Slag by titration. Determination of the
equivalence points is potentiometric using a gold indicator electrode and an Ag/
AgCl reference electrode.

chloride Ti4+ is reduced to Ti3+ and Fe3+ is reduced to Fe2+ in the inert atmosphere
of carbon dioxide.
equivalence point).
Model equation:
Ilmenite

2
500 100
½g mL1 g TiO2 mL1 K2 Cr2 O7
ðV þ V ÞH9 V500 m WFe
CRM
-tot V100 1
1

WFe-tot ¼ EX2 msEQ3 V100 100 ½% H9 ¼ ðVEX2CRM
þ VEQ3 ÞV500 100 ½g mL g Fe mL K2 Cr2 O7
Ti-Slag

2
500 100
mCRM WFe
ðVEX2 þ VEQ3 ÞH12 V500 tot V100
CRM
WFe-tot ¼ ms V100 100 ½% H12 ¼ -
ðVEX2 þ VEQ3 ÞV500 100 ½g mL1 g Fe mL1 K2 Cr2 O7

W mass fraction [%]
H standardization parameter [g/mL]

Repeatability
Volume of sample aliquot
Volume of dissolved sample
Mass of sample
Standardization of K2Cr2O7 with CRM
Volume of K2Cr2O7 used for titration


V100 100 mL Volume of sample aliquot
V500 500 mL Volume of dissolved sample
ms 0.5101 g Mass of sample
H8 0.009585 g mL−1 Standardization of K2Cr2O7 with CRM Ilmenite
VEX1 þ VEQ2 5.7155 mL Volume of K2Cr2O7 used for titration

H9 0.006835 g mL−1 Standardization of K2Cr2O7 with CRM Ilmenite

quantity
H11 0.004862 g Standardization of K2Cr2O7 with CRM Titanium
mL−1 Slag

quantity
H12 0.003480 g Standardization of K2Cr2O7 with CRM Titanium
mL−1 Slag
5. Evaluate the standard uncertainty of each contribution


P 0.068 % Repeatability
V100 0.0489 mL Volume of sample aliquot
V500 0.2086 mL Volume of dissolved sample
H8 0.0000298 g Standardization of K2Cr2O7 with CRM
mL−1 Ilmenite

mL−1 Ilmenite

mL−1 Titanium Slag

(continued)
(continued)
mL−1 Titanium Slag

Ilmenite – TiO2
ðVEX1 þ VEQ2 Þ H8 V500

WTiO2 ¼ 100
ms V100
ð0:1483 mL þ 5:5672 mLÞ 0:009585 g mL1 500 mL 100
¼ ¼ 53:70%
0:5101 g 100 mL
Ilmenite – Fe-tot

WFe-tot ¼ 100
ms V100
ð0:1008 mL þ 4:4724 mLÞ 0:006835 g mL1 500 mL 100
¼ ¼ 30:64%
0:5101 g 100 mL
Titanium Slag – TiO2

WTiO2 ¼ 100
ms V100
ð0:1407 mL þ 16:1476 mLÞ 0:004862 g mL1 500 mL 100
¼ ¼ 78:41%
0:5050 g 100 mL
Titanium Slag – Fe-tot

WFe-tot ¼ 100
ms V100
ð0:1184 mL þ 1:7092 mLÞ 0:003480 g mL1 500 mL 100
¼ ¼ 6:30%
0:5050 g 100 mL
units
Software
Ilmenite
TiO2 Fe-tot
Unit Value u urel Value u urel
P % – 0.068 0.00127 – 0.064 0.00209
V100 mL 100 0.0489 0.00049 100 0.0489 0.00049
V500 mL 500 0.2086 0.00042 500 0.2086 0.00042
ms g 0.5101 0.00015 0.00029 0.5101 0.00015 0.00029
H g/mL 0.009585 0.0000298 0.00311 0.006835 0.0000360 0.00527
VEX þ VEQ mL 5.7155 0.00836 0.00146 4.5732 0.00836 0.00183
W % 53.70 0.200 0.00373 30.64 0.184 0.00600
Titanium slag
TiO2 Fe-tot
Unit Value u urel Value u urel
P % – 0.066 0.00084 – 0.030 0.00476
V100 mL 100 0.0489 0.00049 100 0.0489 0.00049
V500 mL 500 0.2086 0.00042 500 0.2086 0.00042
ms g 0.5050 0.00015 0.00030 0.5050 0.00015 0.00030
H g/mL 0.004862 0.0000105 0.00216 0.003480 0.0000344 0.00989
VEX þ VEQ mL 16.2883 0.0174 0.00107 1.8276 0.0163 0.00892
W % 78.41 0.208 0.00265 6.30 0.089 0.0142
the units
k=2 ILMENITE TITANIUM SLAG
TiO2 Fe-tot TiO2 Fe-tot
U* % 0.40 0.37 0.42 0.18
*U is expanded uncertainty calculated using a coverage factor of 2, which gives a level of
confidence of approximately 95%
1 Standardization of K2Cr2O7 with CRM Ilmenite or CRM Titanium Slag – H
2 Volume of K2Cr2O7 used for titration – VEQ
3 Repeatability – P

Ilmenite:
• (53.7 ± 0.40) % TiO2 (k = 2)*

• (30.6 ± 0.37) % Fe-tot (k = 2)*
Titanium Slag:
• (78.4 ± 0.42) % TiO2 (k = 2)*

• (6.3 ± 0.18) % Fe-tot (k = 2)*
*The reported uncertainty is expanded uncertainty calculated using a coverage

factor of k = 2, which gives a level of confidence of approximately 95%.
Addendum I. Measurement Uncertainty Calculation
ILMENITE TITANIUM SLAG

TiO2 Fe-tot TiO2 Fe-tot
Value u urel Value u urel Value u urel Value u urel
P [%] – 0.068 0.00127 – 0.064 0.00209 – 0.066 0.00084 – 0.030 0.00476
V100 [mL] 100 0.0489 0.00049 100 0.0489 0.00049 100 0.0489 0.00049 100 0.0489 0.00049
V500 [mL] 500 0.2086 0.00042 500 0.2086 0.00042 500 0.2086 0.00042 500 0.2086 0.00042
ms [g] 0.5101 0.00015 0.00029 0.5101 0.00015 0.00029 0.5050 0.00015 0.00030 0.5050 0.00015 0.00030
H [g/mL] 0.009585 0.0000298 0.00311 0.006835 0.0000360 0.00527 0.004862 0.0000105 0.00216 0.003480 0.0000344 0.00989
VEX + VEQ [mL] 5.7155 0.00836 0.00146 4.5732 0.00836 0.00183 16.2883 0.0174 0.00107 1.8276 0.0163 0.00892
W [%] 53.70 0.20 0.00373 30.64 0.18 0.00600 78.41 0.21 0.00265 6.30 0.089 0.0142
U 0.40 0.37 0.42 0.18
ri [%] ri [%] ri [%] ri [%]
Task for a Laboratory Measurement of Mass Fraction …
P 11.53 12.14 10.10 11.32

V (100) 1.72 0.67 3.41 0.12
V (500) 1.25 0.48 2.48 0.09
m (s) 0.62 0.24 1.26 0.04
H 69.50 77.18 66.48 48.75
V (EX) + V (EQ) 15.38 9.30 16.27 39.68
W 100.00 100.00 100.00 100.00
125
H8 H9 H11 H12
Value u urel Value u urel Value u urel Value u urel
V100 [mL] 100 0.0489 0.00049 100 0.0489 0.00049 100 0.0489 0.00049 100 0.0489 0.00049
V500 [mL] 500 0.2086 0.00042 500 0.2086 0.00042 500 0.2086 0.00042 500 0.2086 0.00042
mCRM [g] 0.5088 0.00015 0.00029 0.5088 0.00015 0.00029 0.5054 0.00015 0.00030 0.5054 0.00015 0.00030
VEX + VEQ [mL] 5.1810 0.00836 0.00161 5.2405 0.00836 0.00160 17.4827 0.0174 0.00100 1.9490 0.0163 0.00836
WCRM [%] 48.8 0.125 0.00256 35.2 0.175 0.00497 84.1 0.150 0.00178 6.71 0.035 0.00522
H [g/mL] 0.009585 0.0000298 0.00311 0.006835 0.0000360 0.00527 0.004862 0.0000105 0.00216 0.003480 0.0000344 0.00988
ri [%] ri [%] ri [%] ri [%]
V (100) 2.47 0.86 5.12 0.24
Task for a Laboratory Measurement of Mass Fraction …
V (500) 1.80 0.63 3.72 0.18

m (CRM) 0.90 0.31 1.89 0.09
V (EX) + V (EQ) 26.94 9.17 21.20 71.63
W (CRM) 67.89 89.03 68.08 27.86
H 100.00 100.00 100.00 100.00
127
Further Readings
1. ISO 591-1: 2000 Titanium dioxide pigments for paints – Part 1: Specifications and methods
of test
2. J. D. Norris; Analyst, November 1984, Vol. 109: Determination of titanium in titanium
dioxide pigments, paints and other materials by chromium(II) chloride reduction and
automatic potentiometric titration
3. ISO 21748: 2010 Guidance for the use of repeatability, reproducibility and trueness estimates
in measurement uncertainty estimation
4. ISO TR 13587: 2012 Three statistical approaches for the assessment and interpretation of
measurement uncertainty
5. ISO 7870-2: 2013 Control charts – Part 2: Shewhart control charts
6. ISO 2854: 1976 Statistical interpretation of data – Techniques of estimation and tests relating
to means and variances
7. EURACHEM/CITAC Guide CG4: Quantifying Uncertainty in Analytical Measurement,
Third Edition, 2012
8. Eurolab Technical Report No. 1/2002: Measurement Uncertainty in Testing
9. NORDTEST Report TR 537: Handbook for Calculation of Measurement Uncertainty in
Environmental Laboratories, Edition 3.1, 2012
10. R. Bettencourt da Silva, E. Bulska, B. Godlewska-Żyłkiewicz, M. Hedrich, N. Majcen, B.
Magnusson, S. Marinčić, I. Papadakis, M. Patriarca, E. Vassileva, P. Taylor; Analytical
measurement: measurement uncertainty and statistics, European Commission, Joint Research
Centre, Institute for Reference Materials and Measurements, © European Union, 2012
Jurij Pustinek works in Cinkarna Celje, one of the largest

chemical-processing companies in Slovenia. He is holding the
position of Head of the Analytical laboratory of the Quality
Control Department, where the main activities are quality
control of raw materials, intermediate products, various prod-
ucts of Cinkarna and also environmental samples. He is
responsible for the development and validation of analytical
methods and estimation of uncertainty of measurements. He is
also involved in implementation and maintenance of the
technical requirements of EN ISO/IEC 17025 in Analytical
laboratory.
Karmen Rajer Kanduč works in Cinkarna Celje, one of the

largest chemical-processing companies in Slovenia. She is
holding the position of Quality Manager of the Quality Control
Department, where the main activities are sampling, quality
control of raw materials, intermediate products and various
end-product of Cinkarna. As the Quality Control Manager she
is also responsible for the quality control of measurements in
the field of testing of samples from environmental. Early work
in Cinkarna included quality control of titanium dioxide white
pigment, the development and validation of analytical methods
and estimation of uncertainty of measurements. She was
involved in implementation of the technical requirements of EN
ISO/IEC 17025 in testing laboratories in Quality Control
Department. Currently she is focusing on implementation and
maintenance of ISO standards in Cinkarna Celje.
Nineta Hrastelj works as the General Secretary of the

European Association for Chemical and Molecular Sciences
(EuCheMS). She studied at the University of Ljubljana,
Slovenia. Her research topics were horizontal across analytical
chemistry, chemometrics and metrology. She is author,
co-author or editor of about 250 contributions in research,
teaching and related topics of general concern. In the recent
years, her expert work is mostly about quality of analytical
measurements, where, amongst other, she has been charing the
TrainMiC Editorial Board and is a founding member of both,
TrainMiC and Measurement Science in Chemistry (previously
AcadeMiC) programme of the European Commission JRC.

Uncertainty in Chemistry Vol. 3 Examples

Uploaded by

Copyright:

Available Formats

You might also like

Uncertainty in Chemistry Vol. 3 Examples

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Uncertainty in Chemistry Vol. 3 Examples

Uploaded by

Copyright:

Available Formats

Nineta Hrastelj

Ricardo Bettencourt da Silva Editors

ISBN 978-3-030-20346-7 ISBN 978-3-030-20347-4 (eBook)

Discendo docebis, docendo disces

Brussels, Belgium Philip Taylor

Philip Taylor works for the European Commission Joint

The publication of the ISO Guide to the Expression of Uncertainty in Measurement

Laboratories not involved in highly regulated analytical ﬁelds, such as many of

Lisbon, Portugal Ricardo Bettencourt da Silva

Measurement of Total Reactive Phosphorus in Natural Water

About the Editors

Nineta Hrastelj works as the General Secretary of the

Ricardo Bettencourt da Silva completed his B.Sc. in

quality control and evaluation of measurement uncer-

Maria Filomena Gomes Ferreira Crujo Camões Centro de Química Estrutural,

Nineta Hrastelj EuCheMS, Brussels, Belgium

Ricardo Bettencourt da Silva, Maria da Ascensão Rebelo da Silva

1 Brief Description of the Analytical Procedure

Electronic supplementary material The online version of this chapter (https://doi.org/10.1007/

R. B. da Silva (&) C. M. R. R. de Oliveira M. F. G. F. C. Camões

© Springer Nature Switzerland AG 2019 1

Direct 1. H2SO4 hydrolysis; 2. Spectrophotometry 1. Digestion

* Reactive and acid hydrolyzable phosphorus

Fig. 1 Steps for the analysis of phosphorus fractions

M. da Ascensão Rebelo da Silva Trancoso

1.2 Determination of Reactive Phosphorus in Natural Water

Principle: Molybdophosphoric acid is formed and reduced by stannous chloride to

Fig. 2 Preparation of the calibrators

(iv) Measurement model (to calculate the measured quantity value):

2 The Customer’s Requirements Concerning Quality

Directive 75/440/EEC excerpt:

Directive 79/869/EEC excerpt:

Directive 79/869/EEC deﬁnes “limit of detection” as “the minimum value of the

3 Validation of the Measurement Procedure – Relevant

Table 1 Blank signals obtained in repeatability conditions (absorbance units, a.u.)

Table 3 Impact of the temperature on the absorbance of 10 µg P calibrator (results obtained

Table 4 Analysis of a certiﬁed reference material

Table 5 Dispersion of duplicate results of unknown samples obtained under repeatability

4 Measurement Uncertainty of the Result: Relevant

1. Interpolation standard uncertainty, uInt, estimated by the linear least squares

sy/x Residual standard deviation of the calibration curve;

uðfStd Þ uSt2 0:0332lg P

3. Sample volume uncertainty

tol Tolerance associated with pipette nominal volume (0.08 mL);

Mean recovery observed in the analysis of the CRM;

4.2 Experimental Data of Sample Analysis

Spectrophotometric signals (absorbance units, a.u.)

Analysed Sample signal (a.u.)

Exercise 1: Establishing Traceability in Analytical

2. How would you demonstrate traceability of your result?

3. Any other comments, questions …

Exercise 2: The Customer’s Requirements Concerning

Parameters to be Value requested by the Value obtained during validation

Exercise 3: Enlarge the Analytical Method Scope

2. State how should be validated the selectivity

Exercise 4: Building an Uncertainty Budget

2. Provide the model equation used to evaluate the measurement uncertainty

3.1. Build cause/effect diagram

Updated measurement model: