ISO/TR 18336:2016

TECHNICAL ISO/TR
REPORT 18336
First edition
2016-02-15
Guidelines for good XRF laboratory
practice for the iron ore industry
Lignes directrices de bonnes pratiques de laboratoire de
spectrométrie de fluorescence de rayons X pour l’industrie du
minerais de fer
Reference number
ISO/TR 18336:2016(E)
©
ISO 2016

---------------------- Page: 1 ----------------------
ISO/TR 18336:2016(E)

COPYRIGHT PROTECTED DOCUMENT
© ISO 2016, Published in Switzerland
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized otherwise in any form
or by any means, electronic or mechanical, including photocopying, or posting on the internet or an intranet, without prior
written permission. Permission can be requested from either ISO at the address below or ISO’s member body in the country of
the requester.
ISO copyright office
Ch. de Blandonnet 8 • CP 401
CH-1214 Vernier, Geneva, Switzerland
Tel. +41 22 749 01 11
Fax +41 22 749 09 47
copyright@iso.org
www.iso.org
ii © ISO 2016 – All rights reserved

---------------------- Page: 2 ----------------------
ISO/TR 18336:2016(E)

Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Reagents . 1
3 Apparatus . 2
4 Fused glass beads . 5
4.1 General . 5
4.2 Storage . 5
4.3 Disc making precision . 5
4.4 Bead distortion (curvature and flatness) . 6
5 Quality control . 6
5.1 Selection of QC samples and frequency of preparation . 6
5.2 Analysis of QC and analytical samples . 7
5.3 Control charts . 7
5.4 Participation in proficiency test programs . 8
Annex A (informative) Results for flux loss on ignition testing . 9
Annex B (informative) Procedure to check disc making precision .10
Annex C (informative) Method to determine relationship between height and concentration.11
Annex D (informative) Production of a height adjustable cup .12
Annex E (informative) Bead measurement apparatus .13
Annex F (informative) Flow sheet for fused bead quality .14
Annex G (informative) Microsoft Excel program for calculating disc precision .15
Annex H (informative) Data input screen for calculating disc precision .19
Annex I (informative) Loss of accuracy with no loss of precision .20
Annex J (informative) Loss of accuracy with loss of precision .22
Annex K (informative) Loss of accuracy and precision — Loss of bead making precision .23
Annex L (informative) Results for spectrometer precision test .25
Annex M (informative) Drift correction .26
Bibliography .27
© ISO 2016 – All rights reserved iii

---------------------- Page: 3 ----------------------
ISO/TR 18336:2016(E)

Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out
through ISO technical committees. Each member body interested in a subject for which a technical
committee has been established has the right to be represented on that committee. International
organizations, governmental and non-governmental, in liaison with ISO, also take part in the work.
ISO collaborates closely with the International Electrotechnical Commission (IEC) on all matters of
electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are
described in the ISO/IEC Directives, Part 1. In particular the different approval criteria needed for the
different types of ISO documents should be noted. This document was drafted in accordance with the
editorial rules of the ISO/IEC Directives, Part 2 (see www.iso.org/directives).
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. ISO shall not be held responsible for identifying any or all such patent rights. Details of
any patent rights identified during the development of the document will be in the Introduction and/or
on the ISO list of patent declarations received (see www.iso.org/patents).
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation on the meaning of ISO specific terms and expressions related to conformity
assessment, as well as information about ISO’s adherence to the WTO principles in the Technical
Barriers to Trade (TBT), see the following URL: Foreword — Supplementary information.
The committee responsible for this document is ISO/TC 102, Iron ore and direct reduced iron,
Subcommittee SC 2, Chemical analysis.
iv © ISO 2016 – All rights reserved

---------------------- Page: 4 ----------------------
ISO/TR 18336:2016(E)

Introduction
This Technical Report is intended for use in conjunction with other International Standards for the
chemical analysis of iron ores. Although it was written for a high through-put iron ore laboratory,
the procedures described can be modified to suit other industry or laboratory requirements. Some
laboratories may find the recommended frequency of testing recommended by this Technical Report
to be excessive for their situation or the precision required by them. In this case, the operator may use
their informed discretion to adapt the recommendations of the guidelines to their situation.
© ISO 2016 – All rights reserved v

---------------------- Page: 5 ----------------------
TECHNICAL REPORT ISO/TR 18336:2016(E)
Guidelines for good XRF laboratory practice for the iron
ore industry
1 Scope
This Technical Report specifies recommended quality control procedures for XRF laboratories
operating within the iron ore industry.
2 Reagents
All reagents (including fusion fluxes and calibration reagents) should be purchased from reputable
suppliers and should meet the minimum requirements for purity as listed in ISO 9516-1. All reagents
should have a batch number and, where available, a certificate of analysis. Details of purchased reagents
(supplier, amount purchased, quality, and batch number) should be recorded. These records should
include what the reagents are used for. For batches of flux, the records should indicate which samples
were analysed with a particular batch.
2.1 Fusion flux
As the levels of contamination may vary from batch to batch of flux, the purity of fusion fluxes should
be checked prior to use. This can be achieved by fusing duplicate beads of pure silica and iron with
the new flux, and analysing these along with beads prepared using a previously tested (certified) flux.
Background concentrations should not exceed 10 ppm to 20 ppm (as compared to a certified batch of
flux) for each of the following oxides Mn O , SnO , V O , Cr O , Co O , NiO, CuO, ZnO, As O , PbO, BaO,
3 4 2 2 5 2 3 3 4 2 3
Na O and Cl and the sum of the positive differences should not exceed 40 ppm to 50 ppm.
2
The concentrations of the oxides should not exceed 0,01 % for each of the following oxides Fe O , SiO ,
2 3 2
CaO, Al O , TiO , MgO, K O and P O , (the absolute sum of the differences should not exceed 0,02 %),
2 3 2 2 2 5
while backgrounds should not differ by more than 0,01 %. Sulfur (reported as SO ) can frequently vary
3
by 0,05 %. Where flux does not conform to specifications, a second duplicate set of beads (made with old
and new flux) should be prepared by a different operator on the same day, or by the same operator on a
different day. If the material fails to meet the minimum specifications, the supplier of non-conforming
flux should be contacted and a replacement batch obtained and tested.
Where non-significant deviations are observed for major and trace elements between flux batches,
these beads can be used to update calibration intercepts. In all cases, records of calibration prior and
after amendment should be kept.
Prior to calibration amendment, the concentration levels of all previously analysed blank beads (prior
to calibration amendment) should be plotted, and trends noted. If consecutive sets of duplicate beads
show consistent positive concentration increases, previous beads should be refused and re-run, and the
trends confirmed or negated.
Where laboratories elect to use additive fluxes (oxidizing, release agents or internal standard
compounds), the homogeneity of the flux should be tested, assessed and compared against the quoted
quality or against a flux batch that is known to be homogenous. Testing methods include direct
measurement of added analytes, or indirect measurement of a quality parameter (ignition loss). An
example of flux testing results can be found in Annex A.
As calibrations would have been amended, trends will be seen as negative values progressing towards
a more positive result. If past expected trends cannot be replicated, the XRF instrument (calibration,
monitor) should be inspected. If previously seen trends are repeated, flux suppliers should be contacted,
and the problem discussed.
© ISO 2016 – All rights reserved 1

---------------------- Page: 6 ----------------------
ISO/TR 18336:2016(E)

2.2 Calibration reagents
Reagents used for XRF recalibration should be checked in a similar manner to that used to check flux
(by preparing both the old reagent and the new reagent in the same type of flux). Here the level of
contaminants should not exceed those reported on the reagent supplier’s certificate of analysis. Please
note that it is common for reagent suppliers to report reagent purity based on an “as difference” basis.
Consequently, a 99,999 % reagent may have only been analysed for a single contaminant whose content
is less than 1 ppm. However, this reagent may contain other contaminants which have not been analysed
whose concentration may be significant.
Alternatively, if no in-house high purity reagent is available, small quantities of analysed reagents may
be obtained from reputable laboratories. In addition, where reagents are suspected of contamination,
they can be externally analysed.
3 Apparatus
All equipment used to prepare and measure fused glass beads should be checked on a regular basis in
accordance with the schedule set out in Table 1. The frequencies defined in Table 1 are those required by
a high through-put iron ore laboratory. As this Technical Report is a guideline rather than a prescriptive
standard, laboratories with lesser demands can modify the figures accordingly.
3.1 Bead preparation equipment
All equipment used for the production of fused glass beads (such as balances, fusion ware, fusion
furnaces and sample drying equipment) should be installed, maintained and operated in accordance to
the manufacturer’s recommendations. The external surfaces of all fusion furnaces should be inspected
at the commencement of each shift for excessive dust and glass fragments or glass spills. If any spillage
is detected it should be cleaned up before further use. On a weekly basis, or if spillage has occurred,
fusion furnaces should be allowed to cool and the interior should be inspected for faults (broken or
loose furnace linings that may contaminate samples) and cleaned (and repaired if necessary). Fusion
furnace temperatures should be checked weekly for accuracy and uniformity of temperature using a
calibrated reference thermocouple. As a general rule, an independent performance audit of bead making
equipment (including any environmental factors) should be performed yearly.
Table 1 — Summary of frequency of tests and procedures
Frequency Test
a
Each shift Run Monitors . Monitor data should be checked for drift (3.2.4). If problems/issues
are suspected within the instrument, then increased monitor samples should be run
throughout the day.
At the commencement of every shift, all laboratory personnel involved in the produc-
tion of XRF fusion beads should prepare at least one quality control sample (reference
or certified reference material) in duplicate (5.1).
At the commencement of every shift, all laboratory personnel involved in the produc-
tion of XRF fusion beads should prepare at least one quality control sample (reference
or certified reference material) in duplicate (5.1). These specimens should be meas-
ured and the results evaluated before any unknown specimens are validated.
Monitor chilled water flow rate and temperature (3.2.2).
Monitor compressed air pressure (3.2.2).
Monitor detector gas flow rate and pressure (3.2.2).
Monitor instrument error logs (3.2.2).
Inspect fusion furnaces and clean if necessary (3.1).
a
Monitor is used exclusively for drift correction. Where the power settings have not changed more than 1 kW from the
nominal operating power and the spectrometer has been shown to have achieved stability over a desired time frame (3.2).
Monitor updates can be performed for the corresponding time frame.
2 © ISO 2016 – All rights reserved

---------------------- Page: 7 ----------------------
ISO/TR 18336:2016(E)

Table 1 (continued)
Frequency Test
Weekly Check resolution and pulse shift for flow detectors (3.2.2).
Perform disc making precision tests for new operators. Once competent, perform
checks monthly (4.3).
Clean interior and exterior of fusion furnaces (3.1).
Check furnace temperature and uniformity of temperature (3.1).
Fortnightly Back up spectrometer data files, calibrations and configurations (3.2.3).
Measure flatness of moulds and casting dishes (4.4).
Monthly Perform disc making precision tests for competent operators (4.3) or automated
equipment.
Clean monitors (3.2.4).
Three-monthly Prepare synthetic calibration standards (SynCals) in quadruplicate and compare
results to calibration data (5.3).
Half-yearly and prior to Perform precision tests in ISO/TR 18231.
calibration or after
Perform long-term stability tests in ISO/TR 18231.
major repairs
Perform detector linearity tests in ISO/TR 18231.
Yearly Check all bead preparation equipment (balances, fusion ware, fusion furnaces and
environmental conditions) (3.1).
a
Monitor is used exclusively for drift correction. Where the power settings have not changed more than 1 kW from the
nominal operating power and the spectrometer has been shown to have achieved stability over a desired time frame (3.2).
Monitor updates can be performed for the corresponding time frame.
3.2 XRF Spectrometer
3.2.1 General
All XRF instruments should be tested to ensure conformance to ISO/TR 18231 for instrument precision
and detector linearity. Precision testing should be performed twice yearly, or whenever major repairs
to an XRF (replacement of detector, X-ray tube, generator or measurement electronics) or to the chilled
water circuit has been performed.
For all instrument tests, a relative precision of better than 0,03 % coefficient of variation is required
for the analysis of Fe in iron ores, and so spectrometer precision and linearity tests should be carried
7
out using accumulated count rates of 2 × 10 counts for each measurement. Where separate counting
channels are used for the different elements (simultaneous instruments), or where the detector is
changed, the dead time of each channel and each detector should be determined independently, using
an appropriate wavelength for the detector/crystal combination.
Long term (24 h) XRF stability tests should be performed using various kV and mA settings, detector,
collimator, and crystal combinations. All tests should be conducted biannually at a significance level of
7
0,03 % (2 × 10 counts), and immediately prior to performing an XRF calibration. Instruments should
exhibit short-term stability in the order of 0,03 % over a 6 h to 12 h period. If an XRF fails to meet
desired precisions, instrument manufacturers should be consulted and the cause of poor instrument
stability rectified.
Additional spectrometer tests can be found in Annexes I, J and L.
3.2.2 Each shift/weekly spectrometer checks
As the operation and performance of an XRF spectrometer is highly dependent on the quality of external
services such as chilled water (flow rate and temperature), compressed air (pressure) and detector
gas (flow rate and pressure), the status of these supplies should be monitored on a daily basis and the
© ISO 2016 – All rights reserved 3

---------------------- Page: 8 ----------------------
ISO/TR 18336:2016(E)

results recorded in a check sheet. In addition, error logs should be checked daily and corrective action
taken when faults are reported. As spectrometers possess internal services (such as chilled water flow,
temperature and conductivity, internal spectrometer temperature and instrument vacuum levels), the
status of these should be monitored. Note that as all modern XRF software possesses spectrometer
status screens, these should be configured to be open in a minimised window so that the status of
spectrometer services can be monitored.
As the performance and condition of spectrometer flow detectors are highly dependent on the quality
of detector gas, the resolution and pulse shift should be checked on a weekly basis. Checks should be
performed with regard to the manufacturer’s recommended procedure. Minor changes in pulse shift
(≥2, ≤5) should be corrected (note as spectrometers typically have different detector voltage settings
for each crystal, all should be checked when tests and amendments are performed). Significant changes
in detector resolution or pulse shift position should be investigated as these may be caused by window
failure or gas contamination (leaks).
3.2.3 Hardware and software backup
A backup of spectrometer data files, calibrations and configurations should be performed every two
weeks. This backup should be stored on a remote computer or shared data storage area. XRF software
can be stored in a secure area within the laboratory.
It is also highly recommended that XRF software and spectrometer data (calibration files and
configuration settings) be loaded on several computers (laptops preferred) which possess the necessary
hardware (serial ports) required to connect and communicate with the spectrometer. These computers
will serve as a hardware and software backup.
3.2.4 Instrument monitors
To compensate for drifts in X-ray tube output intensity, all X-ray measurements should be made
relative to a monitor disc. Although different monitor discs could be used for each component, it is
most convenient to use a single disc, containing all components to be measured. The requirement of
the monitor disc is that it be stable over a long period. Also, the monitor should contain sufficient of
each element to ensure that the intensity of each analytical line is much higher than the intensity of
the background and can be measured with the required precision in a reasonable time. Suitable stable
monitor discs made for the analysis of iron ore are available commercially.
The essential property of the monitor is its long-term stability. It should be flat so that it can be placed
reproducibly in the sample holder of the XRF, and the analytical surface should be polished for easy
cleaning. Monitors should be cleaned in accordance with manufacturer’s specifications or every two
months with ethanol or acetone. In-house prepared borate monitors are not recommended as they are
unstable and deteriorate over time.
The accumulated counts for the various elements should be such that they are higher than those from
iron ores, so that precision is not limited by the monitor. For the major elements, including Fe, the count
rate may be less than observed from iron ores, but it should be high enough that, in a short counting
time, the counting error is sufficiently small. Monitors should be run every 4 h to 6 h during, and prior
to the commencement of routine analyses.
Monitor count times can be determined at the time of calibration using the formula below.
The monitor count time TM (in seconds) for each element is given by Formula (1):
RS ×T
TM =×2 (1)
RM
where
RS is the intensity, c/s, from the calibration standard SynCal, measured for 10 s
(see ISO 9516-1);
4 © ISO 2016 – All rights reserved

---------------------- Page: 9 ----------------------
ISO/TR 18336:2016(E)

For Fe, the intensity should be measured using 100 % Fe O .
2 3
T is the required counting time (in seconds), for each element as derived in ISO 9516-1:2003, 7.2.6;
RM is the intensity, c/s, from the monitor for each analyte, measured for 10 s.
Whenever XRF spectrometer power settings are changed more than 1 kW, instrument tube instability
may occur. Depending on the type of spectrometer and X-ray tube design, this instability may persist
for 30 min to 60 min. Consequently, prior to measurement, X-ray equipment should be powered up for a
suitable time and allowed to stabilise.
Where spectrometers are maintained at a standby setting that is less than 0,5 kW of the nominal
operating power, monitors need not be run prior to the commencement of analysis. Monitor acceptance
control charts should be plotted on Shewhart charts using the limits specified in the control chart
section of this Technical Report.
3.2.5 Calibration validation
Once a calibration has been performed in accordance to ISO 9516-1, the method should be validated
(trueness testing and conformance to method precision) using appropriate reference standards. The
certified reference and other reference materials used to validate a calibration should be chosen using
ISO 16042 for concentration and matrix elements.
In addition to validating a calibration using reference materials, silicon dioxide and iron oxide, pure
and binary calibration standards and synthetic calibration standards (SynCals) should be analysed as
unknowns to validate the calibration. Note, as the calibration and calculation algorithms are different,
a slight bias between the regressed and measured chemical concentrations may occur. This bias should
be not more than 1,5 times the counting statistical error (measurement precision) for 10 % of the
analysed samples.
4 Fused glass beads
4.1 General
All samples, regardless if they are calibration, quality control or for routine analysis should be prepared
with equal care and consideration. Consequently, all samples should be prepared using the procedures
listed in ISO 9516-1:2003, 7.1.2 to 7.1.7 using competent operators. To ensure that the quality of analyses
are not operator compromised or biased, all bead preparations should be operator traceable.
4.2 Storage
Beads should not be stored in desiccators containing any drying agents (such as silica gel) or in drying
ovens as erroneous dust particles will contaminate the analytical surface. It is recommended that beads
be clearly labelled and placed in clean re-sealable PVC bags prior to analysis and for long term storage.
Where beads are to be analysed within 30 min of preparation, cooled beads should be transferred
directly from the casting dish/mouldable into a clean spectrometer sample tray.
4.3 Disc making precision
All personnel/equipment involved in the production of XRF samples should be tested for disc making
precision. Initially, operators should be taught by an operator who has proven experience in the
preparation of fused beads. Disc making should be checked weekly, until such a time as beads can be
made consistently to conform to required laboratory precisions (normally, three consecutive runs).
Thereafter, disc making precision should be checked monthly. A general procedure for checking disc
making can be found in Annex B, with a calculation program for measurement precision in Annex G.
Where automated bead making facilities are used (weigh stations and fusion lines), all preparation
lines should be tested.
© ISO 2016 – All rights reserved 5

---------------------- Page: 10 ----------------------
ISO/TR 18336:2016(E)

As a general guide, reading errors should not exceed statistical counting errors, while disc making
errors should not exceed reading errors by more than 1,5 % of the theoretical counting error. If
reading errors (measurement precision) exceed those expected for counting times, this is indicative of
instrument precision errors (non-stability of detectors, tube, collimators, loading errors). Alternatively,
this could indicate insufficient instrument warm up time (stabilisation time). If consecutive runs
(performed at longer instrument warm up periods) fail to meet desired instrument precisions,
instrument manufacturers should be consulted and the cause of poor instrument stability rectified.
If disc making errors (precision) are in excess of read errors by 1,5 %, beads should be broken and
refused/re-poured using exactly the same fusion ware (moulds and mouldables) used to prepare the
beads. If individual re analysed bead error trends correlate previously analysed beads, this is indicative
of an inherent problem in the composition of the bead. This could be due to incomplete drying of flux
or oxidant or incomplete ignition of sample. Alternatively, operators may have balance problems, poor
weighing technique, bead curvature problems or quantitative loss of material during transfer operations.
If disc making is found to improve following refusion and reanalysis (XRF reading errors are constant)
this is indicative that the originally cast beads were not homogeneous or were poorly prepared beads
(cracked, undissolved material crystallisation). If this occurs, the operation of the fusion equipment
should be checked to ensure that the temperature within the furnace, the agitation and the bead cooling
operation are all uniform. Where a significant loss in the ability to cast beads is encountered, inspections
and tests of weighing and casting equipment should be performed (see Annex K). If these are found to
be in order, the fusion parameters (agitation time, speed etc.) should be adjusted and the bead making
precision rechecked using freshly weighed samples. Note that when manually operated fusion equipment
is used, differences in the pouring time and cooling rates have been found to affect precision.
...