Designation: C1832 − 16

Standard Test Method for

Determination of Uranium Isotopic Composition by the
Modified Total Evaporation (MTE) Method Using a Thermal
Ionization Mass Spectrometer1
This standard is issued under the fixed designation C1832; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.

1. Scope 2. Referenced Documents

1.1 This test method describes the determination of the 2.1 ASTM Standards:2
isotope amount ratios of uranium material as nitrate solutions C753 Specification for Nuclear-Grade, Sinterable Uranium
by the modified total evaporation (MTE) method using a Dioxide Powder
thermal ionization mass spectrometer (TIMS) instrument. C776 Specification for Sintered Uranium Dioxide Pellets
1.2 The analytical performance in the determination of the C787 Specification for Uranium Hexafluoride for Enrich-
235
U/238U major isotope amount ratio by MTE is similar to the ment
(“classical”) total evaporation (TE) method as described in Test C833 Specification for Sintered (Uranium-Plutonium) Diox-
Method C1672. However, in the MTE method, the evaporation ide Pellets
process is interrupted on a regular basis to allow measurements C859 Terminology Relating to Nuclear Materials
and subsequent corrections for background form peak tailing, C967 Specification for Uranium Ore Concentrate
perform internal calibration of a secondary electron multiplier C996 Specification for Uranium Hexafluoride Enriched to
(SEM) detector versus the Faraday cups, peak centering, and Less Than 5 % 235U
ion source refocusing. Performing these calibrations and cor- C1008 Specification for Sintered (Uranium-Plutonium)
rections on a regular basis during the measurement, improves DioxidePellets—Fast Reactor Fuel (Withdrawn 2014)3
precision, and significantly reduces uncertainties for the minor C1068 Guide for Qualification of Measurement Methods by
isotope amount ratios 234U/238U and 236U/238U as compared to a Laboratory Within the Nuclear Industry
the TE method. C1128 Guide for Preparation of Working Reference Materi-
als for Use in Analysis of Nuclear Fuel Cycle Materials
1.3 In principle, the MTE method yields major isotope C1156 Guide for Establishing Calibration for a Measure-
amount ratios without the need for mass fractionation correc- ment Method Used to Analyze Nuclear Fuel Cycle Mate-
tion. However, depending on the measurement conditions, rials
small variations are seen among sample turrets; therefore, a C1347 Practice for Preparation and Dissolution of Uranium
small correction based on measurements of a certified refer- Materials for Analysis
ence material is recommended to improve consistency. The C1411 Practice for The Ion Exchange Separation of Ura-
uncertainty around the mass fractionation correction factor nium and Plutonium Prior to Isotopic Analysis
usually includes unity. C1625 Test Method for Uranium and Plutonium Concentra-
1.4 Units—The values stated in SI units are to be regarded tions and Isotopic Abundances by Thermal Ionization
as the standard. When non-SI units are provided, they are for Mass Spectrometry
information only. C1672 Test Method for Determination of Uranium or Pluto-
1.5 This standard does not purport to address all of the nium Isotopic Composition or Concentration by the Total
safety concerns, if any, associated with its use. It is the Evaporation Method Using a Thermal Ionization Mass
responsibility of the user of this standard to establish appro- Spectrometer
priate safety and health practices and determine the applica- D1193 Specification for Reagent Water
bility of regulatory limitations prior to use.
2
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
1
This test method is under the jurisdiction of ASTM Committee C26 on Nuclear contact ASTM Customer Service at [email protected]. For Annual Book of ASTM
Fuel Cycle and is the direct responsibility of Subcommittee C26.05 on Methods of Standards volume information, refer to the standard’s Document Summary page on
Test. the ASTM website.
3
Current edition approved Jan. 15, 2016. Published January 2016. DOI: 10.1520/ The last approved version of this historical standard is referenced on
C1832-16. www.astm.org.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States

1
 C1832 − 16
E2655 Guide for Reporting Uncertainty of Test Results and 3.3.18 RSD—relative standard deviation—SD (see below)
Use of the Term Measurement Uncertainty in ASTM Test divided by the mean value of the observations in repeated
Methods sampling
E2586 Practice for Calculating and Using Basic Statistics 3.3.19 RSE—relative standard error—SE (see below) di-
vided by the mean value of the observations in repeated
3. Terminology sampling.
3.1 The terminology standard C859 contains terms, 3.3.20 SD—standard deviation—according to Practice
definitions, descriptions of terms, nomenclature, and explana- E2586, 3.1.30: the square root of the sum of the squared
tions of acronyms and symbols specifically associated with deviations of the observed values in the sample divided by the
standards under the jurisdiction of Committee C26 on Nuclear sample size minus one.
Fuel Cycle. 3.3.21 SE—standard error—according to Practice E2586,
3.2 Definitions: 3.1.29: standard deviation of the population of values of a
3.2.1 abundance sensitivity, n—in isotope amount ratio sample statistic (that is, the mean value) in repeated
measurements, the ratio of the measured intensity of an ion measurements, or an estimate of it.
beam at a mass, m, to the measured intensity from the same 3.3.21.1 Discussion—According to Practice E2586, 3.1.30:
isotope measured at one mass unit difference (for example, If the standard error (SE, see above) of a statistic is estimated,
m 6 1). it will itself be a statistic with some variance that depends on
3.2.1.1 Discussion—Abundance sensitivity is a measure of the sample size, that is, the number of observed values in the
the magnitude of the peak tailing correction. For measuring sample (Practice E2586, 3.1.26).
uranium on thermal ionization mass spectrometer (TIMS) and 3.3.21.2 Discussion—According to Guide E2655, 5.8.4.1:
inductively coupled plasma mass spectrometry (ICP-MS) From statistical theory, a 95 % confidence interval for the mean
instruments, the abundance sensitivity is typically calculated as of a normal distribution, given n independent observations x1,
the ratio of the measured signal intensities at masses 237 and x2, ..., xn drawn from the distribution is, x̄ 6 t × SD / √n, where
238 using a suitable uranium sample. x̄ is the sample mean, SD is the standard deviation of the
3.2.2 total evaporation, TE, n—analytical method for deter- observations (see above), and t is the 0.975 percentile of the
mination of isotope amount ratios of uranium or plutonium, as Student’s t distribution with n-1 degrees of freedom. Because
described in Test Method C1672, also called “classical” total Student’s t distribution approaches the Normal as n increases,
evaporation in this test method. the value of t approaches 1.96 as n increases. This is the basis
for using the (coverage) factor 2 for expanded uncertainty. The
3.2.3 turret, n—holder for sample filaments. standard error (SE) of the mean value of a series of n
3.2.3.1 Discussion—Alternate names for turret are carousel, independent repeated measurements can be derived from that
magazine, wheel. by using t = 1, so the standard error (SE) is given by SD ⁄ √n.
3.3 Acronyms: 3.3.22 SEM—secondary electron multiplier.
3.3.1 cpm—counts per minute 3.3.22.1 Discussion—In the scientific literature the acronym
3.3.2 cps—counts per second SEM is also used for Scanning Electron Microscope, but
within this test method SEM represents Secondary Electron
3.3.3 CRM—certified reference material
Multiplier.
3.3.4 dark noise—observed count rate on an ion counting
3.3.23 SGAS—Safeguards Analytical Services Laboratory
detector measured without incident ion beam
of the IAEA
3.3.5 DS—double spike
3.3.24 TIMS—thermal ionization mass spectrometry
3.3.6 DU—depleted uranium
4. Summary of Test Method
3.3.7 EU—European Union
3.3.8 FAR—Faraday Cup 4.1 The modified total evaporation method has been devel-
oped on the basis of the “classical” total evaporation technique
3.3.9 HEU—high enriched uranium (1-4)4, also described in Test Method C1672. By using the total
3.3.10 IAEA—International Atomic Energy Agency evaporation technique, the mass fractionation is minimized by
3.3.11 ICPMS—inductively coupled plasma mass spectrom- evaporating the entire sample amount loaded on the filament,
etry in contrast to the “conventional” technique as described in Test
Method C1625. The MTE method has already been described
3.3.12 IRMM—Institute for Reference Materials and Mea- in detail in Refs (5) and (6). If necessary, uranium is separated
surements from plutonium and other elements (to eliminate isobaric
3.3.13 ITU—Institute for Transuranium Elements interferences) by selective extraction, anion exchange (see
3.3.14 JRC—Joint Research Centre Practice C1411), or extraction chromatography. The purified
uranium fraction as nitrate solution is loaded onto a degassed
3.3.15 LEU—low enriched uranium
3.3.16 MTE—modified total evaporation 4
The boldface numbers in parentheses refer to a list of references at the end of
3.3.17 NBL—New Brunswick Laboratory this standard.

2
 C1832 − 16
filament (made of metals such as rhenium, zone-refined samples, the minor isotopes are only measured using Faraday
rhenium, or tungsten with high evaporation temperature) and cups with amplifier resistors of 1012 Ω yielding a favorable
converted to an oxide by controlled heating of the filament signal-to-noise ratio.
under atmospheric conditions. For the modified total evapora- 4.3 Similar to the TE analysis, the isotope ion beams of the
tion method, the uranium load is in the range of about 2 to 6 µg, major isotopes 235U and 238U are integrated over the course of
which is about one to two orders of magnitude higher than that the analysis, and the summed intensity for 235U is divided by
typically used for the “classical” total evaporation method by the summed intensity for 238U to yield the major isotope
TIMS. The sample filament is mounted on the sample turret amount ratio. The result of the major isotope amount ratio is
using a double-filament configuration. This configuration con- corrected for mass fractionation using the measurement of a
sists of an evaporation filament (Re or W) on which the sample CRM analyzed on the same sample turret.
is loaded, and an ionization filament (Re filament with no 4.4 The minor isotope amount ratios are corrected for mass
sample). For a measurement in the mass spectrometer, the fractionation for each integration step individually. This is
ionization filament is heated to a temperature of about 1800 to accomplished in an internal manner, the magnitude of the mass
1900°C, sufficient to generate 187Re ion beam signals of 150 to fractionation for the minor ratios is calculated from the
400 mV (corresponding to ion currents of 1.5 to 4 × 10-12 A measured mass fractionation of the major ratio. The peak
using an amplifier resistor of 1011 Ω). The intensity of the tailing contributions are determined at two positions, slightly
optimized 187Re signal depends on the Re material (zone- below and above the isotope masses of interest. If applicable,
refined or non-zone-refined), thickness and the measurement the SEM-versus-Faraday calibration is also performed for each
conditions, but it is expected to be the same for all filaments on integration step individually.
the same sample turret. The 187Re ion current is also used for 5. Significance and Use
the initial ion beam focusing. The evaporation filament is
heated next. After ion beam re-focusing and mass re- 5.1 Uranium material is used as a fuel in certain types of
nuclear reactors. To be suitable for use as nuclear fuel, the
adjustment initially using a small sum intensity (sum of 234U,
235 starting material shall meet certain specifications such as those
U, 236U, and 238U) of about 1 to 3 V, the data acquisition
described in Specifications C753, C776, C787, C833, C967,
begins under computer control to yield a predefined target sum
C996, and C1008, or as specified by the purchaser. The isotope
intensity of 20 to 30 V (corresponding to ion currents of 2 to 3 amount ratios of uranium material can be measured by mass
× 10-10 A using an amplifier resistor of 1011 Ω). This target spectrometry following this test method to ensure that they
value depends on the amount of uranium loaded on a filament. meet the specification.
The MTE analysis takes between 2 and 5 h per sample filament
and is about three to ten times longer than typical (“classical”) 5.2 The MTE method can be used for a wide range of
sample sizes even in samples containing as low as 50 µg of
TE analyses in spite of the higher intensities at which the
uranium. If the uranium sample is in the form of uranium
analyses are performed. To ensure that the whole sample is
hexafluoride, it can be converted into a uranium nitrate solution
completely evaporated and analyzed before the ionization
for measurement by the MTE method. The concentration of the
filament breaks as a result of overheating, the MTE analysis loading solution for MTE has to be in the range of 1 to 6 mg/g
routine is programmed to increase the target sum intensity to allow a sample loading of 2 to 6 µg of uranium. A minimum
during the course of the analysis if necessary. The occurrences loading of 3 µg uranium per filament is recommended. This is
of outliers due to technical glitches, for example, as a result of needed to have a suitable ion signal especially for the two
termination before complete sample evaporation or because of minor isotopes (234U and 236U) thus enabling the internal
early sample exhaustion, are minimized by a dynamic target calibration of SEM versus the Faraday cups during the mea-
intensity control feature through manipulation of the target sum surement.
intensity depending on the actual measurement conditions.
5.3 Until now, the instrument capabilities for the MTE
4.2 The sample amount to be loaded for MTE analyses is method have only been implemented on the TRITON™ TIMS
limited to a range of about 2 to 6 µg to achieve ion beam instrument.5 Therefore all recommendations for measurement
signals of about 20 to 30 V for the major isotope – 238U for parameters in this test method are specified for the TRITON™
DU, NU, and LEU samples and 235U for HEU samples – cor- TIMS instrument. The manufacturers of other TIMS instru-
responding to a 234U intensity of 1 to 10 mV. This causes the ments (for example, IsotopX and Nu Instruments) have plans
234
U ion beam to be suitable for an internal cross calibration of to implement the modifications needed in their instruments to
the SEM versus the Faraday cups through the entire measure- use the MTE method.
ment time. This also allows the 236U/238U isotope amount ratio 5.4 The MTE method described here can also be extended
to be measured in a wide dynamic range from 10-2 down to to measurement of elements other than uranium. Note that the
10-10 using a Faraday cup or an SEM in combination with an MTE method has already been implemented for plutonium and
energy filter for improved abundance sensitivity. For certain calcium.
DU samples, the 236U ion beam is used for the cross calibration
5
and the 234U is measured on the SEM (known as “reverse” The TRITON™ (Plus) Multicollector Thermal Ionization Mass Spectrometer is
a trademark product of Thermo Fisher Scientific, http://www.thermoscientific.com/
MTE, see 13.8.8). For all samples with minor ratios 234U/238U content/tfs/en/product/triton-plus-multicollector-thermal-ionization-mass-
and 236U/238U higher than 4 × 10-5, which also includes HEU spectrometer.html#sthash.stbDyJBG.dpuf.

3
 C1832 − 16
6. Interferences 7.1.6 A sample turret with at least ten filament positions to
238
6.1 Isobaric nuclides such as Pu interfere in the uranium allow automatic measurement sequences of at least five repli-
measurements. The removal of interferences is generally ac- cate filament loadings per sample and at least five replicate
complished by chemical separation leading to ionization of filament loadings for a calibration standard (preferentially a
uranium only and improved precision of measured isotope CRM);
amount ratios. 7.1.7 A pumping system that is able to attain a vacuum of
<4.0 × 10-5 Pa (3 × 10-7 torr) in the ion source, the analyzer,
6.2 It has to be ensured that samples are not contaminated and the detector is required. Tailing corrections are dependent
by environmental uranium. The level of effort required to on the vacuum levels inside the mass spectrometer. Analyzer
minimize contamination shall be based upon the sample size pressures below 7.0 × 10-7 Pa (5 × 10-9 torr) are preferred;
and the levels of contamination present in the analytical 7.1.8 A mechanism to scan masses by varying the magnetic
facility. For extremely small samples or extremely low 236U field or the accelerating voltage, or both;
abundances, residual uranium from chemicals used for sample 7.1.9 An optical pyrometer for determining the tempera-
dissolution and sample preparation are possible sources for tures of the filaments;
bias in the isotopic data. 7.1.10 A computer for control of the data acquisition ac-
6.3 Samples shall be chemically purified to assure reliable cording to a predefined sequence.
analyses by TIMS. Impurities, especially alkali elements, 7.2 Special MTE Capabilities—The mass spectrometer soft-
produce unstable ion emission leading to poor precision in the ware must be flexible enough to implement a user-defined
isotope amount ratios. Organic contaminants or oxide layers on filament-heating program for MTE.6
the filaments also adversely influence TIMS analyses. Isobaric
7.3 A separate filament degassing device for cleaning of the
interferences, if not removed, will bias the isotope amount
filaments before sample loading is recommended.
ratios. Contaminants in reagents, lab, ware, or filament material
are also sources for bias in the isotope amount ratios. 7.4 Microsyringe—Syringe to transfer microlitre volumes of
solutions.
6.4 The performance of the instrument can be adversely
affected by changes in the environmental conditions of the 8. Reagents and Materials
laboratory, that is, temperature and humidity. For this reason, 8.1 Purity of Reagents—Reagent-grade chemicals shall be
controlled laboratory environmental conditions should be used in all sample preparations. Unless otherwise indicated, it
maintained (within the manufacturer’s specifications) during is intended that all reagents conform to the specifications of the
instrument operation. Committee on Analytical Reagents of the American Chemical
Society where such specifications are available.7 Other grades
7. Apparatus may be used, provided it is first ascertained that the reagent is
7.1 The suitability of the mass spectrometer for carrying out of sufficiently high purity and low uranium concentration to
measurements by the MTE method shall be evaluated by permit its use without lessening the accuracy of the MTE
means of performance tests. The relevant instrument charac- measurements.
teristics are as follows: 8.2 Ultra-high purity reagents are necessary for processing
7.1.1 A thermal ionization source for using double filament small sample amounts or samples with extremely small isotope
assemblies with rhenium or tungsten filaments, or both; amount ratios. The level of uranium contamination from
7.1.2 A mass analyzer sufficient to resolve adjacent masses chemicals, water, and the sample handling environments shall
in the mass-to-charge range being studied, m/z = 233 to 238 for be determined to ensure that the materials and the environment
U+. Resolution shall be greater than 350 (full width at 1 % of are sufficiently pure for the samples being analyzed.
peak height) and the abundance sensitivity at mass 237 for ions 8.3 Nitric Acid (HNO3, 15.8 M)—Concentrated nitric acid.
of 238U less than 5 × 10-6;
7.1.3 A multiple Faraday collector system to allow simulta- 8.4 Nitric Acid (HNO3, 1 M)—One volume of concentrated
neous detection of isotope beams from m/z = 233 to 238 for U+ nitric acid (HNO3, 15.8 M) brought to 15.8 volumes with
ions; water.
7.1.4 For the Faraday cups used to measure the major ion 8.5 Purity of Water—Unless otherwise indicated, references
beams of 235U and 238U, there shall be current amplifiers to water shall be understood to mean laboratory accepted
equipped with 1011 Ω resistors and, for the minor ion beams of demineralized or deionized water as described by Type I of
234
U and 236U, there shall be current amplifiers equipped with Specification D1193.
at least 1011 Ω, and preferably 1012 Ω (or even 1013 Ω) resistors
to improve the signal-to-noise ratio; 6
The sole source of supply of the apparatus known to the committee at this time
7.1.5 The detection system shall include at least one sec- is Thermo Fisher Scientific Inc., 81 Wyman St., Waltham, MA 02451. If you are
aware of alternative suppliers, please provide this information to ASTM Interna-
ondary electron multiplier (SEM) operated in ion counting tional Headquarters. Your comments will receive careful consideration at a meeting
mode or a Daly detector, or similar detector, with a dark noise of the responsible technical committee,1 which you may attend.
7
<0.2 cps. The SEM shall be equipped with an energy filter to Reagent Chemicals, American Chemical Society Specifications, American
Chemical Society, Washington, DC. For suggestions on the testing of reagents not
improve the abundance sensitivity, which shall be better than
listed by the American Chemical Society, see Analar Standards for Laboratory
2 × 10-8 at mass 237 for ions from 238U+ with the energy filter Chemicals, BDH Ltd., Poole, Dorset, U.K. and the United States Pharmacopeia and
in operation; National Formulary, U.S. Pharmacopeial Convention, Inc. (USPC), Rockville, MD.

4
 C1832 − 16
8.6 Filaments—High-purity filaments, for example, zone degassing procedure when a vacuum pressure of <4.0 × 10-4 Pa
refined Re, are required. The size and configuration of the (3 × 10-6 torr) is reached. Recommended filament currents for
filament is instrument dependent. Tungsten might be used with degassing are in the range of 4.0 to 5.0 A, corresponding to
minor modifications to the heating script. All filaments must be temperatures in the range of 1700 to 2000°C. Perform the
degassed before use. For measurement of uranium present only degassing for a duration of at least 30 min.
at the trace levels in the sample, contamination of the samples 11.2 “Initialization” of the Sample Turret—Adjust, if
by interferences from elemental or molecular species within needed, the position of the sample turret in the ion source and
the m/z = 233 to 238 range need to be determined to ensure that verify proper electrical connections for both the evaporation
biases due to contributions from the filament material will not and ionization filament for each sample loading position. It
bias the analysis results. In case a significant count rate >1 cps shall be ensured that the electrical contact is not interrupted in
is determined at m/z = 236, a correction has to be applied for case the turret is slightly moved for the purpose of ion beam
236
U/238U ratios < 10-6 (6, 7). focusing.
9. Hazards 11.3 Electronics Test—Modern mass spectrometric instru-
ments offer an automated routine for testing the stability and
9.1 TIMS instruments operate at 8 to 10 kV electrical performance of the electronic systems (for example, Faraday
potential. Ensure that the high-voltage is switched off before cup amplifier baselines and gains, high-voltage unit, and
insertion or removal of the sample turret into/from the magnet current supply unit). A report is produced flagging
instrument, and working with the ion source, or accessing other systems or components that are out of specification. Users of
electronic components. the instruments should perform routine checks of the perfor-
9.2 The filaments reach temperatures in excess of 2000°C. mance of the electronic systems and ensure that the perfor-
The filament holders, sample wheel, and ion source parts are mance is within manufacturer’s specifications. The frequency
expected to be hot. Ensure that a sufficient time has lapsed for this test shall be established by the user based on manu-
since the last filament heating before accessing the filaments, facturer recommendations or as specified in the user’s quality
sample turrets, and ion source. assurance program.
9.3 Wear eye protection and suitable gloves when filling 11.4 Amplifier Signal Decay Adjustment—Adjust the signal
cold traps with liquid nitrogen. Protect hands, torso, and feet in decay characteristics of the current amplifiers of the Faraday
the event of splashing or spilling of the liquid nitrogen. cups. This is important for measuring isotope amount ratios
9.4 Handle radioactive materials with appropriate attention with a large dynamic range, especially for the minor isotope
to radiological safety. amount ratios of uranium. Depending on the combination of
the capacitance and resistance of the current amplifier, the
10. Sampling, Test Specimens, and Test Units response time for a sudden change in the ion beam signal to the
Faraday cup can reach up to 5 s. For resistances higher than the
10.1 Isotope Reference Materials—Uranium reference ma- 1011 Ω, longer response times of about 15 s can be expected.
terials used in the analysis should be prepared from certified The MTE method is designed by taking into consideration the
reference materials (CRMs) traceable to SI units. Examples required response times. The amplifier response can be
include the NBL U-Series CRMs (for example, U010 and checked either using a custom-made software module within
U500), IRMM uranium standards IRMM 184-187, IRMM the operating software (when available) or “manually” by
019-029 (to be converted from UF6), IRMM-074, IRMM-075, means of a large ion beam that is abruptly directed into a
and IRMM-3100. See Guide C1128 for additional guidance on Faraday cup to check the signal ingrowth time, or away from
preparation of traceable working reference materials. a Faraday cup (for example, by closing a valve between ion
10.2 Uranium samples to be measured and isotope reference source and analyzer) to check the signal decay time.
materials used for estimation of the mass fractionation correc- 11.5 SEM Plateau Voltage—The SEM plateau voltage, that
tion factor (K factor) or for quality control purposes shall be in is, the dependency of the count rate on the applied high voltage
the same medium and in the same oxidation state. The to the SEM, shall be measured on a regular basis, as recom-
solutions for loading onto the filaments are 1 M nitric acid mended by the manufacturer of the SEM or the operator’s
solutions. The uranium concentrations of these solutions shall quality system. Typical SEM plateau voltages range between
be similar. The loading and drying sequence of the filaments 1700 to 3000 V. The SEM plateau voltage drifts to higher
shall also be similar. values with use and shall be readjusted to restore the original
10.3 In the case of characterization studies of test materials, efficiency. The efficiency should be higher than 90 %.
possible inhomogeneity between test units shall be evaluated 11.6 SEM Dark Noise—SEM detectors are usually operated
statistically and included in the uncertainty calculations of the in ion counting rather than analogue mode to avoid the resistor
isotope amount ratios assigned. noise of the current amplifier. The ion counting mode results in
much smaller uncertainties encountered for measurements of
11. Preparation of Apparatus very low intensity ion beams (10-19 to 10-16 A, equivalent to 0.6
11.1 Filament Degassing—Degass all filaments before us- to 600 cps). However, the dark noise, that is, the observed
ing them for MTE measurements. It is recommended to use a count rate measured without incident ion beam, shall be
separate apparatus (see 7.3) for performing this. Start the measured before an automated MTE measurement sequence to

5
 C1832 − 16
enable corrections to the isotope amount ratios and their performed prior to each automatic MTE sequence. Historical
uncertainties (7). The dark noise shall be below 0.2 cps. gain calibration data can be used to evaluate the stability of the
amplifiers.
11.7 Ion Source and Analyzer Pressure—It is important to
achieve a certain level of vacuum before the isotope amount 12.5 Faraday Cup Effıciency Test—The response of indi-
ratio measurements can begin; see 7.1.7 for the recommended vidual Faraday cups depends on several factors, for example,
pressure. The peak tailing depends strongly on the vacuum extent of usage, manufacturing variability, and can also be
pressure in the detector system since the number of ion affected by an insufficient electron suppression voltage. The
collisions with gas molecules inside the mass spectrometer is a relative response of the Faraday cups, therefore, shall be
direct function of the ambient pressure. Increase in the pressure determined periodically. Usually, the Faraday cups of a multi-
within the ion source caused by the heating of the ionization collector system are only intercalibrated for the current ampli-
and evaporation filaments can be subdued, to a certain extent, fiers connected to them (see 12.4) but not for the differences in
by using a cold trap filled with liquid nitrogen. the efficiencies of the Faraday cups themselves. The efficien-
cies of the Faraday cups are expected to be similar to each
12. Calibration and Standardization other, which means that the relative efficiencies (relative to one
reference cup) are normally close to unity. Note that an
NOTE 1—The measurement method may be qualified following Guide
C1068 and calibrated following Guide C1156. Additional information (electronic) amplifier gain calibration (see 12.4) shall be
regarding mass spectrometer calibrations with regard to the MTE method performed prior to the Faraday cup efficiency test. The Faraday
may be found in Refs (5, 6). cup efficiency test can be performed in several ways, as
described in 12.5.1 – 12.5.4.
12.1 Mass Calibration—The relationship between the
12.5.1 The calibration may be performed by switching a
known atomic masses and the magnetic field necessary to
stable ion beam of 187Re (from a blank filament) between each
direct the isotope beams into the detectors shall be updated on
Faraday cup and a reference Faraday cup. In case a relative
a regular basis. Mass calibration shall be performed at intervals
efficiency between the detectors is significantly different from
specified by the manufacturer or the user’s quality assurance
unity, this result can be used to correct for differences in the
program.
detector response. This procedure can be performed with a
12.2 Peak Centering—The peak centering routine is used as relative uncertainty at the level of <0.1 %.
a fine adjustment to ensure that the ion beam is centered within 12.5.2 A series of peak-jumping measurements between all
the detector. Peak centering usually occurs via fine adjustments Faraday cups and a reference cup to be checked can also be
of the accelerating voltage, and any difference between the performed using a sufficiently large uranium sample and one
value optimized during peak centering from the default accel- large stable ion beam, for example, a 10 to 20 V ion beam of
238
erating voltage requires a readjustment of the mass calibration U from an LEU or natural uranium sample. The drift of the
curve. Peak centering shall be performed for at least three signal intensity shall be corrected for using the operating
uranium masses as part of the mass calibration and also before software. This procedure can be performed with a relative
the start of each MTE measurement sequence. During the MTE uncertainty at the level of <0.01 %.
measurement, peak centering is performed on a regular basis. 12.5.3 A series of comparative neodymium (Nd) isotope
amount ratio measurements can be performed in two different
12.3 Amplifier Baseline Calibration—The baselines of the modes such as the multi-dynamic mode and the static mode
Faraday cup amplifiers, that is, the amplifier responses without with “amplifier rotation” (only for TRITON™ TIMS, also
incoming ion beam to the cup, shall be measured on a regular called “virtual amplifier”: each Faraday cup is connected to
basis and checked for stability. During the MTE measurements, each amplifier for regular time intervals during the measure-
baseline measurements are performed on a regular basis. Note ment). This procedure can be performed with a relative
that the integration time for the baseline measurement has a uncertainty at the level of few ppm (5). It shall be repeated
significant influence on the uncertainty of Faraday cup until all Faraday cups of interest for MTE measurements have
measurements, particularly at lower ion beam intensities. been included.
Therefore, the integration time of the baseline (within a 12.5.4 A series of static measurements can be performed
measurement) shall be comparable to the integration time of using special “multi-isotope” reference materials, such as
the actual ion beam signal integration. The long-term historical IRMM-3100a (233U/235U/236U/238U = 1/1/1/1) (8), IRMM-
baseline data shall be regularly reviewed by the user to assure 072/1, IRMM-074/1, or IRMM-199 (233U/235U/238U = 1/1/1)
that the system performance is within manufacturer specifica- (9), to include all Faraday cups. This procedure can be
tions and quality system requirements. performed with relative uncertainties of about 0.03 %.
12.4 Amplifier Gain Calibration—The stability and re- 12.6 Linearity Test—There are various procedures to check
sponse of individual Faraday detector amplifiers shall be the linearity of an isotope mass spectrometer detection system.
measured and differences between amplifiers corrected for via The procedures described in 12.6.1 and 12.6.2 are mainly
the amplifier gain calibration. Gain calibration is normally applicable for Faraday multi-collector systems (for SEM cali-
performed by sequentially applying a stable calibration current bration and linearity, see 12.8 and 12.9).
to the input of each Faraday cup amplifier and the output is 12.6.1 The linearity of the mass spectrometer is determined
then normalized to a reference value to generate a gain over the working range of the Faraday cups by measuring the
235
calibration factor for each amplifier. A gain calibration shall be U/238U ratios of various reference materials under identical

6
 C1832 − 16
conditions. The mass spectrometer system is linear if the K ≈1 and a 233U/235U ratio of ≈0.01, or the NBL CRM U500,
factor, that is, the ratio of the certified 235U/238U ratio to the characterized by a 238U/235U ratio of ≈1 and a 234U/235U ratio
measured 235U/238U ratio, is independent of the isotopic of ≈0.01 (using new data published in Ref (5)), can be used.
composition of the material. For this procedure, the NBL Peak-jumping measurements are performed for count rates
U-series of reference materials (U005a to U970, 0.5 to 97 % of between 5 × 104 cps and 5 × 105 cps for 235U (similar for 238U).
235
U) is ideal and can be combined with the IRMM-183-187 The 233U/235U ratios (234U/235U in case of NBL CRM U500)
series (0.3 to 5 % of 235U) and the IRMM 019-029 (0.17 to 5 % are internally normalized using the 238U/235U ratios of >1 and
of 235U, to be converted from UF6). This procedure shall be plotted versus the 235U count rate. If the data show a linear
performed sequentially for all Faraday cups of the multi- relationship with no significant intercept, a linear regression is
collector system needed for the MTE analyses. calculated with the intercept being zero. The slope of the
12.6.2 The IRMM-072 and IRMM-074 series of reference regression line can be used to calculate the dead time using:
materials are characterized by 238U/235U ratios of ≈1 and
233
U/235U ratios ranging from ≈1 down to ≈10-6 for the 15 or
10 units, respectively, of the used series. For each unit, the bias
τ 5 $ slope% ⁄ S S D D
1 2
233
U
235
U CERT
(1)

of the measured 238U/235U ratios from the certified ones can be where:
used for internal mass fractionation correction of the measured τ = dead time;
233
U/235U ratios. The comparison of the corrected 233U/235U slope = slope of linear regression calculation
ratios with the certified ones allows the linearity of the based on measured 233U/235U ratios ver-
detection system to be checked over a dynamic range of six sus the 235U count rate; and
orders of magnitude for the ion beam intensity. A detailed (233U/ 235U)CERT = certified isotope amount ratio of used
description of the procedure is given in Ref (9). This procedure CRM.
shall be performed sequentially for all Faraday cups needed for 12.8.1.1 In case the data do not indicate a linear relationship
the MTE analyses. with a non-significant intercept, the SEM detection system is
12.7 Peak Overlap—When a Faraday multi-collector sys- likely to be affected by other sources of nonlinearity, which
tem for the simultaneous detection of several masses is used, it shall be investigated.
needs to be ensured that the peak overlap is acceptable. A mass 12.8.2 For an “external check” of the linearity of the SEM
scan, usually by scanning the magnetic field, shall be per- detection system, a series of reference materials can be
formed by which all ion beams are simultaneously moved measured by MTE, for example, the NBL U-series (5), the
through the respective cups. The measured intensities for all IRMM-183-187 series, the IRMM-019-029 series (converted
detectors shall be plotted versus the mass of a reference from UF6 (7)), or the gravimetrically prepared IRMM-075
detector to make the peak overlap visible. All peaks shall have series (6, 10) with 236U/238U rations of 10-4, 10-5, 10-6, 10-7,
a symmetric shape with a common flat region in the center, 10-8, and 10-9.
with the peak centers reasonably close together, as specified by 12.9 SEM Versus Faraday Cup Inter-Calibration—Since
the manufacturer or the user quality system. After a satisfactory the inter-calibration between a SEM detector with a Faraday
peak overlap is realized (by moving cups relative to one cup depends on the ion source focusing and ion beam shape,
another if necessary), the positions of all detectors shall be this calibration shall be performed in an internal manner during
saved, for example, as a Faraday cup configuration file. The the course of each measurement for MTE measurements. For
positions shall be checked and possibly readjusted, manually or this internal calibration, the ion beam of one of the uranium
using stepping motors, as needed before a new automatic isotopes present in the sample is switched between the SEM
measurement sequence. and one Faraday cup within each integration step during the
12.8 SEM Linearity Check and Correction—The linearity of entire measurement. To allow such an internal calibration,
SEM systems shall be checked carefully on a regular basis there shall be at least one suitable ion beam available for this
according to manufacturer’s specifications or the user’s quality purpose. This ion beam shall be within a certain intensity
assurance program. It is emphasized that there can be more range, between the lower limit of reasonable Faraday cup
than one reason for nonlinearity of a SEM detector. For SEM measurements (typically 1 mV on an amplifier with 1011 Ω,
detectors operated in ion counting mode, the dead time of the which is affected by about the same noise as 0.5 mV on an
pulse-counting electronic system is always one source of amplifier with 1012 Ω) and the upper limit for reasonable SEM
nonlinearity, but this can be easily corrected. As pointed out in measurements (typically 1.0 × 106 cps, equivalent to 16 mV on
Ref (9), there is also the possibility of an intrinsic nonlinearity an amplifier with 1011 Ω). The isotope used for the internal
for SEM detectors, possibly depending on the design of the ion calibration within MTE measurements is for most samples
234
optics, the dynode surfaces, or the electronics, which could U, in few cases 236U instead (“reverse” MTE). As a
cause the linearity investigation and correction to become more consequence, the sample amount to be loaded is limited to a
complicated. But in case the dead time is confirmed to be the certain range as well. For typical TIMS conditions, the sample
only source of nonlinearity of a SEM system, the value can be amount for MTE measurements is about 2 to 6 µg, which
determined in various ways (9) using CRMs, as described in provides ion beam signals of about 20 to 30 V for the sum of
12.8.1 and 12.8.2. the main isotopes 235U and 238U. The choice of the isotope
12.8.1 For this procedure, the IRMM-072/8 or IRMM-074/3 used for the inter-calibration depends on the isotopic compo-
(9) reference materials, characterized by a 238U/235U ratio of sition of the sample (based on process knowledge or sample

7
 C1832 − 16
supplier information, or estimated by a mass scan in the all measured sample ratios. It is important that the reference
m/z = 233 to 238 mass range), in the following way: materials are prepared and measured in the same manner as the
12.9.1 Mainly for DU, NU, and LEU samples with 234U/ samples. For the MTE method, the major ratio 235U/238U is
238
U ratios between 3 × 10-5 and 5 × 10-4 and with 236U/238U used to calculate a correction factor, known as the K factor,
ratios below 3 × 10-5, the 234U ion beam is used for the internal which is then applied also for performing corrections for the
calibration. minor ratios 234U/238U and 236U/238U internally (for details see
12.9.2 Mainly for DU (in principle also in case of NU and Section 14).
LEU) samples with 234U/238U ratios below 3 × 10-5 and with 12.10.2 The mass fractionation correction factor, K, is
236
U/238U ratios higher than 3 × 10-5, the 236U ion beam is used calculated as follows:
for the internal calibration (also called “reverse” MTE). K 5 ~ R c ⁄ R m! (2)
12.9.3 In the rare case that both the 234U/238U and the
236
U/238U ratios are below 3 × 10-5, the larger one of 234U and where:
236
U shall be used. This might only apply to quite highly K = mass fractionation correction factor;
enriched 235U or 238U materials, for example, for spike Rm = average measured isotope amount ratio for the CRM;
materials. In this case, an internal calibration using a suitable and
ion beam from an added 233U spike (11) is another option, Rc = certified isotope amount ratio value for the CRM.
which would require isobaric corrections because of the 12.10.3 To correct individual measured isotope amount
possible 234U and 236U contents of the 233U spike. ratios, calculate the appropriate mass fractionation correction
12.9.4 For all samples, that is, DU, NU, LEU, or HEU, with factor based upon the mass difference between isotopes (=3 for
any values for the 234U/238U and 236U/238U ratios, an MTE the major ratio, the 235U/238U ratio) and multiply the sample
measurement can always be performed using Faraday cups ratio by the applicable mass fractionation correction factor.
only, thus without using the SEM and without the need for a
SEM intercalibration. This shall be carefully considered in 13. Procedure
view of the associated uncertainties originating on one hand 13.1 Sample Preparation:
from the SEM calibration and SEM linearity correction (see 13.1.1 Sample Dissolution—Dissolve an appropriate sample
12.8) and on the other hand from the higher relative amplifier amount to obtain the desired filament loading solution for the
noise in case of low signals on the Faraday cups. In case both mass spectrometric analysis. See Practice C1347 for the
ratios 234U/238U and 236U/238U are higher than 5 × 10-4, the dissolution of uranium.
SEM cannot be used for MTE, only Faraday cups shall be used. 13.1.2 Prepare the sample and any reference material solu-
12.9.5 For an “external” check of the SEM versus Faraday tions as purified nitrates, using identical chemical preparation
intercalibration, a suitable sample with a 236U/238U ratio being and handling steps. For uranium samples hydrolyzed from
in the “overlap” range between SEM and Faraday measure- uranium hexafluoride, it is recommended that the samples are
ments (between 3 × 10-5 and 5 × 10-4) can be measured in a converted to U3O8 before dissolution in nitric acid and analy-
comparative way, first using the SEM and secondly using the sis. The solution concentrations shall be chosen to allow for a
Faraday cup for the 236U ion beam. The results shall agree to convenient filament loading (for example, a 1-mg U/mL
each other within their uncertainties. If this is not the case, the solution yields 1 µg of uranium per µL, see also 13.2).
reason shall be investigated. Using reference materials such as 13.1.3 Sample Purification—Use Practice C1411 or similar
IRMM-187 and NBL CRM 010 with suitable 236U/238U ratios, procedure to separate uranium from plutonium other impuri-
it was shown in Ref (6) that the above described intercalibra- ties.
tion procedure is accurate within uncertainties of about 0.4 %
13.2 Sample Loading and Conditioning—Samples for MTE
(k = 2).
are usually directly loaded on the filament by drop deposition.
12.10 Mass Fractionation Correction: Samples and reference materials shall be prepared for analysis
12.10.1 In theory, the “classical” total evaporation (TE) and, by the same method at similar mass loadings. Drop deposition
therefore, also the MTE methods are expected to yield isotope onto the filament can be accomplished with the use of a
amount ratios that do not need correction for mass fraction- microsyringe fitted with a plastic tip. Change the tip between
ation. In practice, measurable mass fractionation for uranium sample loadings to prevent cross contamination. For filaments
measurements has been observed. To be consistent in the loaded by drop deposition, the solution shall be evaporated by
evaluation of the data and the associated uncertainties, it is passing sufficient electrical current through the filament to
recommended to perform a mass fractionation correction also cause gentle drying without boiling. Samples for MTE are
for measurement sequences in which the ratios seem to be mass usually prepared in a 1 M nitric acid with a uranium concen-
fractionation free, that is, where the K factor (see 12.10.2) is tration between 1 to 6 mg/g, which is equivalent to 1 to 6
equal to unity within its uncertainty. For MTE measurements it µg/µL. Depending on the uranium amount to be loaded, more
is recommended to perform the mass fractionation correction than 1 µL may be needed. Deposit drops very carefully and
by measuring a certified isotope reference material in sequence slowly. It is recommended to keep the filament heated by
with the samples, and calculate a mass fractionation correction passing 0.5 to 0.7 A current and depositing 1-µL-drops at a
factor based upon the deviation of the measured major ratio time. After all drops are loaded, the sample solution on the
from the certified ratio. The mass fractionation correction filament is heated until dryness, for at least one more minute,
factor, adjusted for isotope mass difference, is then applied to and then heated for 10 s at a higher current of 1.5 A for

8
 C1832 − 16
conditioning. Alternatively, a stepped-heating program can be most cases (DU, NU, LEU), this is the 238U beam; for HEU,
used to condition samples, that is, to convert samples to depending on the level of enrichment, it may be the 235U beam.
suitable chemical forms. Avoid quick evaporation of the 13.8.6 Data acquisition is started. The data acquisition is
sample or melting the filament. At different facilities, different performed on a per block basis in which each block consists of
loading and conditioning procedures have been established and a minimum of two and a maximum of five mass cycles and
validated. Each procedure shall be applied in a consistent each mass cycle of (usually) five integration steps. The data
manner for all samples, reference materials, and quality control acquisition is continued until the whole sample is evaporated;
samples. see 13.10.
13.3 Mount all sample (evaporation) filaments and ioniza- 13.8.7 MTE Mass Cycle—The MTE mass cycle consists of
tion filaments on a sample turret and insert the sample turret five steps as shown in Table 1.
into the ion source of the mass spectrometer. 13.8.7.1 The first step is the main integration for all iso-
13.4 Seal the source and start evacuating. topes; the integration time is usually about 16 s. The 236U
13.5 Perform adjustment (also called initialization) of the isotope can be detected by either the center Faraday cup (C) or
sample turret. In case of problems with electrical connections, an SEM detector depending on the 236U/238U ratio in the
the source might have to be vented again to resolve the sample. Usually, for 236U/238U ratios below 3 × 10-5, the 236U
problem. measurement is performed using the SEM. The minor isotope
234
U is always detected using a Faraday cup. In case the
13.6 Evacuate ion source to the manufacturer’s recom- 236
U/238U ratio is higher than 3 × 10-5, the 236U is detected
mended minimum pressure or according to the user’s proce- using the center (C) Faraday cup.
dure.
13.8.7.2 Step 2 of the mass cycle is dedicated to the
13.7 Add liquid nitrogen to the cold trap if desired. intercalibration of the SEM versus the Faraday cup multi-
13.8 Isotope Amount Ratio Measurement—The following collector. This is achieved by measuring the 234U isotope first
steps 13.8.1 – 13.8.3 shall be performed automatically under using Cup L2 in Step 1 and then using the SEM in Step 2. The
computer control depending upon instrument. Very small SEM efficiency is then calculated by dividing the measured
234
samples may require manual control to avoid sample loss. U/235U ratios in Step 2 and Step 1 by each other. In this way,
13.8.1 Perform an amplifier gain calibration for each new the signal drift over time is circumvented, which would not be
automatic sequence. the case if the 234U intensities measured in Step 1 by the
13.8.2 Measure the baseline during the course of each Faraday Cup L2 and Step 2 by the SEM would be divided
measurement of a sample and standard. directly by each other. For the SEM efficiency calculation, the
234
13.8.3 Usually under computer control, the ionization fila- U/235U ratio measured in Step 1 has also to be corrected for
ment is heated up to a temperature of about 1800 to 1950°C. the tailing effect using the background measurements per-
For automatic sequences, usually the magnitude of the 187Re formed in Steps 3 and 4 (see 13.8.7.3). Because the count rate
ion beam is used as the regulated quantity instead of the of 234U is usually >60 000 cps, an integration time of 4 s is
filament temperature by a pyrometer because only the narrow sufficient; an idle time of 6 s is typically used. Even when the
sides of the filaments are oriented towards the pyrometer. Faraday cup and no SEM is used for measuring 236U, this
Typical 187Re ion beams of about 300 to 600 mV (on a 1011 Ω second step is retained within the mass cycle for consistency of
resistor) are used for non-zone refined filaments and lower the timing.
values of 150 to 200 mV are used for zone-refined filaments, 13.8.7.3 In Steps 3 and 4 of the MTE mass cycle, the
depending on the brand. background intensities at the low and high mass side of all
13.8.4 The 187Re ion beam is peak centered and focused. If uranium isotopes of interest are measured. The background for
necessary, the 187Re ion beam size is readjusted towards the 234
U is mainly due to the tailing from 238U and, depending on
target value, which shall be similar (within about 10 %) for the 235U enrichment, also tailing from 235U contributes. As a
each measurement within an automatic sequence, by changing result of the differences in the tailing effect between the low
the filament current. and high mass sides, an interpolation for the 234U/238U ratio
13.8.5 The sample filament is heated to a temperature shall be performed, usually using a linear approximation.
sufficient to yield an ion beam sum intensity from 235U and Depending on the instrument and mass resolution, the peak
238
U of about 1 to 4 V (on 1011 Ω resistor). The beam is areas are located well within 60.35 u of the isotope masses,
focused and peak centered using the largest U ion beam, in which are about 234.05, 235.05, 236.05, and 238.05 u for the

TABLE 1 Mass Cycle for MTE

Idle Time/
Cup L2 L1 SEM (or FAR) H1 H2
Measurement Time
234 235 236 238
Step 1 U U U 237 u U 12 s/16 s
234 235
Step 2 U U 6 s/4 s
234 235 236 238
Step 3 U - 0.35 u U - 0.35 u U - 0.35 u U - 0.35 u 5 s/ 8 s
234 235 236 238
Step 4 U + 0.35 u U + 0.35 u U + 0.35 u U + 0.35 u 5 s/8 s
234 235 236 238
Step 5 U U U 237 u U 3 s/1 s

9
 C1832 − 16
isotopes 234U, 235U, 236U, and 238U, respectively. By measur- 13.8.8 The mass cycle for “reverse” MTE consists of eight
ing the background at mass positions 60.35 u of the isotope steps shown in Table 2.
masses, and thereby closer to the peaks compared to the 13.8.8.1 The first step is used for the data acquisition for all
usually applied “half mass” approach, the interpolation be- isotopes on Faraday cups. In Step 2, 236U is measured on the
comes more accurate. SEM for the intercalibration. In Steps 3 and 4, the background
13.8.7.4 The isotope 236U is a special case. If 236U is intensities for the tailing correction are determined. The
detected using the center (C) Faraday cup, the background correction algorithms are described in 14.5.
measurements performed at 235.7 and 236.4 u shall be 13.8.8.2 Step 5 is used for measuring the 234U in the SEM,
interpolated logarithmically (that is, using the geometric mean) and because of the possibly low abundance of 234U in the
to account for the curvature of the peak tailing from 238U. This sample, Steps 6 and 7 are introduced to measure background
type of logarithmic correction is not possible in the same way intensities also for 234U in this case. Step 8 is needed again
for 234U because the background intensities at 233.7 and 234.4 (similar to Step 5 for the “non-reverse MTE”) to switch the
are quite frequently at the same level as the amplifier baseline; mass back to the main cup configuration.
therefore, individual baseline-corrected intensities and ratios 13.8.8.3 The mass cycles can be extended, for instance if
can be negative (on a statistical basis) and would, therefore, not low levels of 233U have to be determined using the SEM for
allow a logarithmic correction algorithm. The correction algo- specific samples.
rithms are given in 14.4.
13.9 Interblock Actions and Filament Heating for MTE:
13.8.7.5 In case 236U is detected using the SEM, which is
usually combined with an energy filter to improve the abun- 13.9.1 Peak Centering and Baseline—Each measurement
dance sensitivity, the tailing effects for the 236U/238U ratio are block usually consists of a maximum of five mass cycles, with
of the order of 10-9 to 10-10 and can be easily corrected. each cycle having a duration of about 60 to 90 s. Before each
Additionally, sometimes the background at mass 236 u is block a peak centering is performed using the most abundant
dominated by K6+ ions with a mass of about 235.78 u isotope, usually the major isotope 238U (235U for HEU
(39K541K-ions) emitted from the hot ionization filament, and samples). Additionally, an ion source focusing is performed
also by unidentified possibly organic interferences. The K6+ before each block. Typically, every five blocks, the electronic
ions interfere only partially with the 236U isotope at mass baselines of all the Faraday cup amplifiers are remeasured. In
236.05 u and can therefore, not be accurately corrected. Thus, case 1012 Ω resistors are used in some of the amplifiers, the idle
in case of a high background measurement (≥10-9) at mass times for the baseline measurements have to be sufficiently
235.7 u, this value shall not be used for correction. Unknown long, for example, at least 15 s.
organic interferences around mass 236 u shall be investigated 13.9.2 MTE Interblock Heating—The filament heating is
by checking blank filaments or measuring certified isotope controlled by a special program script executed after each
reference materials with 236U/238U ratios in the order of 10-8 to measurement block. This program script first measures the
10-9, for example, the IRMM-075 series (10). In this case, the actual sum intensity (mainly from 235U and 238U) and com-
background at the 236U mass position shall be corrected pares it with the predefined (initial) target sum intensity. This
externally. predefined target sum intensity is typically in the range of 20 to
13.8.7.6 For the measurement of 234U using the SEM 30 V, selected by the operator depending on the sample amount
combined with an energy filter in Step 2, the intensity is usually loaded, the experience with the ion source and the instrument
large enough (>60 000 cps) and the background in the vicinity transmission. Based on the comparison of the actually mea-
of 234.05 u small enough so that additional background sured sum intensity with the target sum intensity, the following
measurements are not needed. filament heating step is calculated. The heating step, that is, the
13.8.7.7 Step 5 is an optional “dummy” step, which might difference between the new evaporation filament current and
be needed to move the mass setting back to the main cup the actual one, is calculated based on the relative difference
configuration (as in Step 1) for checking the ion beam signals between the actually measured and the target sum intensity
at the very end of each block. This is needed to determine the using a quadratic polynomial with empirical parameters:
ion beam intensity before the next heating step (see 13.10). HS 5 P1 3 ~ RD sum intensity! 2 1P2 3 ~ RD sum intensity! 1P3 (3)

TABLE 2 Mass Cycle for “Reverse” MTE

Idle Time/
Cup L2 L1 SEM (or FAR) H1 H2
Measurement Time
234 235 236 238
Step 1 U U U (FAR) 237 u U 12 s/16 s
234 235 236 238
Step 2 U U U (SEM) 237 u U 3 s/8 s
234 235 236 238
Step 3 U - 0.35 u U - 0.35 u U - 0.35 u (FAR) U - 0.35 u 5 s/8 s
234 235 236 238
Step 4 U + 0.35 u U + 0.35 u U + 0.35 u (FAR) U + 0.35 u 5 s/8 s
234 235
Step 5 U (SEM) U 5 s/8 s
234
Step 6 U - 0.35 u (SEM) 5 s/4 s
234
Step 7 U + 0.35 u (SEM) 5 s/4 s
234 235 236 238
Step 8 U U U (FAR) 237 u U 3 s/1 s

10
 C1832 − 16
where: 13.10 Termination of MTE Measurement—The MTE mea-
HS = difference between evaporation filament surement is terminated when the sum intensity is below a
current (in A) after and before the user-defined limit of 1 to 2 V.
heating step (in A); 14. Calculations
RD sum intensity = relative difference between target sum
intensity and measured sum intensity; 14.1 Within this section, the MTE data evaluation and
and calculation of the various correction factors will be discussed.
P1, P2, and P3 = empirical parameters for the quadratic Despite the application of the total evaporation principle, it is
polynomial. recommended for consistency reasons to perform a mass
fractionation correction for MTE measurements. For the major
13.9.2.1 Typical values for P1, P2, and P3 are about 0.35 A, ratio 235U/238U, the mass fractionation correction is performed
0.1 A, and 0 A, respectively, but shall be optimized by the user. externally using a reference material measured on the same
If the measured sum intensity is very close to the target sum sample turret under similar conditions in terms of sample
intensity (for example, within 1 %), the filament current is loading and sample matrix as well as the same MTE interblock
recommended not be changed. heating script. The mass fractionation correction for the minor
13.9.3 Dynamic Target Intensity Concept—This is a pro- ratios 234U/238U and 236U/238U is performed internally based
gramming feature within the MTE heating script that is needed on the result of the corrected 235U/238U ratio using the
to avoid extremely long measurement times. This is achieved exponential fractionation law. The differences between differ-
by automatically readjusting the target sum intensity during the ent fractionation laws, for example, power law, exponential
measurement. If the measurement time becomes too long, for law, linear law, or Rayleigh law are at the level of few parts per
example, 5 h or more, this could cause the ionization filament million for heavy elements like uranium, which is insignificant
to become hotter and thinner over time and possibly break. By for the MTE measurements (6).
means of the “dynamic target intensity concept,” the value for
14.2 The calculations to be performed for MTE measure-
the target sum intensity is itself increased on a block-by-block
ments are rather method specific, so there is currently no
basis as long as the measurement is running below a defined
commercial software available. Up until now, two spreadsheets
filament current threshold (for example, about 1.6 A for Re
exist, one by IRMM (and used/modified also by NBL, ITU,
filaments and 2.2 A for W filaments), which is considered too
and IAEA, regularly updated at IRMM) and another one from
low for a not-too-long measurement time. This increase is
ITU using visual basic for calculations and graphic display.
applied between each two blocks, with the exception of the
Within this test method, only the mathematical equations are
starting period of Block 1 to 5 where the filament current is
presented.
intentionally still low. The stepwise increase of the target sum
intensity is continued until the filament current has exceeded 14.3 Evaluation for Major Ratio 235U ⁄ 238U—The total
the defined threshold. This rule is not only applied to one evaporation principle is associated with a particular way of
defined filament current threshold, it is applied cumulatively calculating isotope amount ratios. For uranium measurements
for multiple filament current thresholds, such as 1.6, 1.7, 1.8, using the TE and MTE methods, the major ratio 235U/238U is
1.9, 2.0, 2.1 and 2.2 A, associated with different programmed calculated as:
steps for increasing the target sum intensity, such as 10 %, 5 %,
4 %, 3 %, 2 %, 1.5 %, and 1 % respectively; for more details, S D
235
U
238
U TE
5
Σ ALL_INTEGRATIONS I ~ 235U !
Σ ALL_INTEGRATIONS I ~ 238U !
see Ref (6).
13.9.4 In addition to the implementation of the dynamic
5
Σ ALLINTEGRATIONS F S ~~ !! D
I
I
U
U
235

238 ·I ~ 238U ! G (4)

target intensity concept, the MTE program script reacts on Σ ALL_INTEGRATIONS I ~ 238U !
unusually low or high ion beam intensities from individual
sample filaments. This is achieved by increasing the target sum where:
intensity by a certain percentage (for example, 10 %) if the I(235U) and I(238U) = signal intensities (in V) for 235U and
238
starting filament current at block one is extremely low (for U measured on Faraday cups, cor-
example, below 0.8 A) for a sample appearing as “too large,” rected for the Faraday cup amplifier
or by decreasing the target intensity if the starting filament gains, baselines, and backgrounds on
current is extremely high (for example, more than 2.0 A) for a the low and high mass sides.
sample appearing as “too small.” 14.3.1 Therefore according to Eq 4, the 235U/238U ratio is
13.9.5 The MTE heating program script shall be adapted for calculated as the sum of all 235U ion beam intensity integra-
different instruments and different types and brands of tions during the measurement divided by the respective sum of
filaments, for example, rhenium versus zone-refined rhenium the 238U ion beam intensity integrations. This is the same as the
versus tungsten. For example, if tungsten is used for the average of all measured I(235U)/I(238U) signal intensity ratios
evaporation filaments, which require much higher filament during the measurement, weighted by the 238U ion signal
currents to achieve the same sample temperatures, the script intensities. This algorithm was already applied for the TE
shall be adapted accordingly. The script parameters shall be method for which the signal integration continues without
optimized to obtain reproducible MTE measurements in terms interruption until the sample is consumed. For the MTE
of the number of blocks and mass cycles, the heating, and the method, the major change is that the measurement is inter-
intensity versus time patterns. rupted on a regular basis to perform additional operations

11
 C1832 − 16
during each mass cycle, such as the SEM versus Faraday close to 4⁄3) with, M(mU) being the molar mass of the isotope
m
calibration and background measurements, which mainly serve U.
to improve the accuracy for the measurement of the minor 14.4.1 If the 236U/238U ratio is measured on the (center,
ratios 234U/238U and 236U/238U. These interruptions amount to “C”) Faraday cup, the correction is calculated according to:
about 20 % of the measurement time (preheating time not
included) and, because of their strict regularity, no significant S D 235
U R236

SS D
effects on the 235U/238U ratios from these interruptions have
been observed.
14.3.2 The mass fractionation correction for the 235U/238U
S D236

238
U
U SA2CO
5
1 S D
238

235

238
U
U
U
SA2CO

SA2ME
2 236

238
U
U SA2ME

ratio is calculated from the average values for all replicate

measurements of the reference material used for mass fraction- 2 ŒSS D 235.7
238
U SA2ME
S D DD
236.4
238
U SA2ME
(8)
ation correction and a given sample according to Eq 5 and Eq
Where SA-CO and so forth are the same as defined for Eq 6.
6. For both, the reference material and each sample, a suitable
The exponent R236 is given by R236
number of replicates, typically at least five, shall be measured
5 @ ln ~ M ~ 236U ! ⁄M ~ 238U !! ⁄ln~ M ~ 235U ! ⁄M ~ 238U !! # >0.66526 (which is
on the same turret under similar conditions.
close to 2⁄3) are used.
14.3.3 The K factor is defined as:
14.4.2 If the 236U/238U ratio is measured using the SEM, the

K5
S D
235

238
U
U ST2CE
(5)
relative detection efficiency, G, of the SEM (that is also called
yield) has to be included. The value, G, is calculated according
S D
235

238
U
U to Eq 9, where FAR stands for Faraday cup:

S D
ST2ME 234
U SEM
14.3.4 The K factor is applied as:

S D
235

238
U
U
SA2CO
5 K· 238
U
U
S D 235

SA2ME
(6)
G5
1 S
235

234

235
U FAR
U FAR
U FAR D
Step2

Step1
2
where: S 234
U FAR
238
U FAR D Step1
SA-CO = sample, corrected ratio;
SA-ME
ST-CE
= sample, measured ratio;
= certified ratio of the certified reference material;
SS 234

238
U FAR
U FAR D Step1
2 SS 233.7FAR
238
U FAR D Step3
1 S 234.4FAR
238
U FAR D DD
Step4
⁄2

(9)
and
ST-ME = measured ratio for reference material. 14.4.3 This formula includes the correction for the tailing
14.4 Evaluation for Minor Ratios U ⁄ U and 234 238 contribution from the 235U and 238U ion beams to the 234U
236
U ⁄ 238U for “Non-Reverse” MTE—The corrections for the beam detected on the Faraday cup. On the SEM, no tailing
minor ratios 234U/238U and 236U/238U are performed internally correction for the calibrating 234U ion beam (>60 000 cps) is
on a per mass cycle basis for each reference material and each necessary if an energy filter with an abundance sensitivity of
sample on the same turret. The 234U/238U ratio is corrected for <5 × 10-8 at mass 237 u is used. The corrected 236U/238U ratio
peak tailing from the 235U and 238U ion beams, which is measured using the SEM is given by:
determined using the background measurements at masses
S D 235
U R236

233.7 and 234.4 u performed in Steps 3 and 4 of the same mass

cycle for “non-reverse MTE.” The mass fractionation correc-
tion of the 234U/238U ratio is then performed internally using
S D 236

238
U
U
SA2CO
5
1
G
S D 1 238

235

238
U
U
U
SA2CO

SA2ME
2
the measured ratio (235U/238U)SA-ME from the same mass cycle
and the corrected ratio (235U/238U)SA-CO determined using Eq SS D SS D
236

238
U
U SA2ME
2
235.7
238
U SA2ME
5 and Eq 6. The correction formula for the 234U/238U ratio is
given by: 1 S D D D236.4
238
U SA2ME
⁄2 2 B (10)

S D 235
U R234

1 2
where:
S D
234
U
5
238
U SA2CO
B = (optional) constant background ratio calculated by di-
238
U SA2CO
S D 235

238
U
U
SA2ME
viding a background count rate intensity at mass 236 u,
for example, predetermined using blank filaments or
SS D234

238SS D
U
U SA2ME
2
233.7
238
U SA2ME
suitable reference materials such as IRMM-075 (10), by
the observed average 238U intensity during the MTE
1 S D DD234.4
238
U SA2ME
⁄2 (7) measurement.
14.4.4 Because of relatively low ion beam intensities during
Where SA-CO and so forth are the same as defined for Eq 6. the start and the end periods of each MTE measurement, the
234
The exponent R234 is given by R234 U/238U and 236U/238U ratios are only taken into consider-
5 @ ln ~ M ~ 234U ! ⁄M ~ 238U !! ⁄ln~ M ~ 235U ! ⁄M ~ 238U !! # >1.33617 (which is ation for the main part of the measurement as defined by the

12
 C1832 − 16
sum intensity being above a threshold of 30 % of the maximum 14.6.2 For the uncertainty of the corrected major ratio
235
sum intensity during the measurement. This threshold shall be U/238U calculated according to Eq 6, the relative uncertain-
applied in a consistent way for reference materials and ties of the measured ratios of the reference material used for
samples. Note that this threshold of 30 % only applies to the mass fractionation correction and the sample as well as the
corrected minor isotope amount ratios and not to the major relative uncertainty in the certified value of the reference
ratio 235U/238U. material are added in quadrature to obtain the relative uncer-
14.4.5 The minor ratios 234U/238U and 236U/238U for the tainty of the corrected ratio of the sample. This correction and
reference material(s) are calculated in the same way using Eq the uncertainty calculation are performed on the averages of all
7-10 as for the samples; SA-ME has to be replaced by ST-ME measurements on the same sample turret. Because of the good
and SA-CO by ST-CO. The ST-CO results obtained for the repeatability of MTE measurements (about 0.01 to 0.02 %
minor isotope amount ratios of reference materials can be RSD, independent of the value of the ratio), this uncertainty is
compared to the certified ratios and, therefore, serve as a often dominated by the uncertainty of the known certified ratio
quality control. of the reference material (typically in the range of 0.015 to
14.5 Evaluation for Minor Ratios 234 U ⁄ 238 U and 0.1 %). Typical values for the expanded uncertainty (coverage
236
U ⁄ 238U for “Reverse” MTE—The corrections for the factor k = 2) are about 0.04 to 0.12 % (see 16.1 for additional
minor ratio, 236U/238U, measured on Faraday cups only are details).
calculated according to Eq 8 as before. 14.6.3 A relative uncertainty of 0.020 % (k = 2) for the
14.5.1 If the 234U/238U ratio is measured using the SEM, the external (rather than internal) type of mass fractionation of the
relative detection efficiency, G, of the SEM (that is also called major ratio, 235U/238U, is added in quadrature to the uncertain-
yield) has to be included. The value, G, is calculated according ties of MTE measurements to take potential unaccounted
to Eq 11 where FAR stands for Faraday cup: fractionation effects into consideration, which might be slightly
different between the reference material and sample(s) (6).
S 236
U SEM
D
1 2
235
U FAR
14.6.3.1 If two samples with the same origin but different
Step2
G5 processing are measured on the same sample turret for the
S 236

235
U FAR
U FAR D Step 1
purpose of directly comparing their (uncorrected) 235U/238U
ratios, for the relative difference of the measured 235U/238U
S 236
U FAR
238
U FAR D Step1
ratios between the two samples also an additional uncertainty
of 0.020 % (k = 2) shall be added in quadrature.
SS 236

238
U FAR
U FAR D Step1
2 SS 235.7FAR
238
U FAR D Step3
1 S 236.4FAR
238
U FAR D DD
Step4
⁄2 14.6.4 For the minor ratios, 234U/238U and 236U/238U, the
corrections are performed on a “per mass cycle” basis accord-
(11)
ing to Eq 7-12. Average values and their uncertainties for each
14.5.2 This formula includes the correction for the tailing sample (or the reference material) on the turret are calculated.
contribution from the 235U and 238U ion beams to the 236U The uncertainty of the average value is also known as the
beam detected on the Faraday cup. On the SEM, no tailing standard error (SE), which is calculated as the “internal” SD
correction for the calibrating 236U ion beam (>60 000 cps) is divided by the square root of the number of mass cycles for an
necessary if an energy filter with high abundance sensitivity is individual measurement. The typical (or average) derived SE
used. The corrected 234U/238U ratio measured using the SEM is for a given sample (or the standard) can be compared to the
given by: “external” SD of the replicate measurements for the same
material on the turret. For a sufficient number (≥5) of replicate
S D 235
U R234

1 2
measurements, the “external” SD can be relied upon, but, in
S D 234
U
5
1 238
U SA2CO
case of very few replicates, the average internal SE shall be
238
U SA2CO
G
S D 235

238
U
U
SA2ME
used instead of the “external” SD as a more reasonable
estimate of the uncertainty. The larger of these two values can
SS D SS D
234
U
238
U SA2ME
2
234.7
238
U SA2ME
be used as a conservative estimate of the uncertainty. Within
the MTE evaluation software, this conservative estimate of the
1 S D DD 234.4
238
U SA2ME
⁄2 (12)
uncertainty value is selected, and then divided by the square
root of the number of replicate MTE measurements for the
sample (or reference material) on the turret to obtain the
14.5.3 There is no background ratio to be taken into account
uncertainty for the average 234U/238U or 236U/238U ratio for the
for typical count rates observed for 234U, in contrast to those
turret sequence. This uncertainty for the average ratio is then
low count rates encountered for 236U, for example, in case of
combined with other uncertainty components arising from, for
natural samples.
example, mass fractionation correction, SEM-Faraday inter-
14.6 Uncertainty Estimation: calibration, SEM linearity correction, dark noise, and back-
14.6.1 The uncertainties for MTE measurements are esti- ground correction, which are described in detail in Ref (6). To
mated according to the principles described in the Guide to the facilitate the uncertainty calculations for some of these
Expression of Uncertainty in Measurements (GUM) (12), see components, constant (relative) uncertainty contributions can
also Refs (13) and (14). For MTE measurements, the uncer- be assigned, for example, 0.4 % for all SEM-related correc-
tainty evaluation steps are explained in the following. tions (6) (k = 2). For minor ratios of 10-4 or larger, the mass

13
 C1832 − 16
fractionation correction is typically the dominant contribution 15.1.1 From New Brunswick Laboratory (NBL, U.S.DOE):
to the uncertainty. For ratios between 3 × 10-5 and 10-4 the original U-series CRMs: U010, U030A, U050, U100,
measured on the Faraday collector, the uncertainty calculated U200, U500, U750, and U900 and, additionally, CRMs U005a,
from the repeatability, which is due to the amplifier noise, is U045, U630, 112A, 113B, 115, 116A, and 145.
dominant. For ratios close to 3 × 10-5 and below measured 15.1.2 From the Institute for Reference Materials and Mea-
using the SEM, the uncertainty is given by a combination of surements (IRMM-JRC-EU): the IRMM-183-187 series,
counting statistics and the constant uncertainty contribution IRMM-072/1, the IRMM-075/1-6 series, the IRMM-074/1-10
stemming from the SEM-related corrections. For the extremely series, IRMM-199, and the IRMM-019-029 series (converted
low ratios, the uncertainty is a combination of the background from UF6).
correction uncertainty and counting statistics. 15.1.3 These materials cover the range of 235U enrichments
14.6.5 The detection limit for the 236U/238U ratio by MTE from 0.2 to 95 %. The reference materials used for testing the
can be expressed by the sum of the absolute deviation of the MTE method were different for each of the instruments
measured from the true (or certified) value plus the uncertainty depending on the operating facility and the projects for which
(coverage factor k = 2) of the measured value. The value the measurements were performed. Therefore, common gen-
depends on the measurement conditions, for example, the eral observations and some examples for precision and bias are
loaded sample amount, the ion beam intensities, and measure- presented within this test method.
ment time. For 5-µg loadings, a detection limit of about 6 × 15.2 Precision for MTE Measurements of the Major Ratio
10-10 was found (6, 15). For 2.5-µg loadings, it is slightly 235
U ⁄ 238U:
higher (that is, inferior), about 3 × 10-9 (6). 15.2.1 For the un-corrected major ratio, 235U/238U, the
14.7 Calculation of Isotope Amount Fractions (“Atom within-turret precision varies within a range of 0.005 and
Percent”), Isotope Mass Fractions (“Weight Percent”), and 0.030 % relative standard deviation (RSD) for replicate (>5)
Molar Mass (“Atomic Weight”): filament loadings of reference materials analyzed (5-7, 14,
14.7.1 Isotope Amount Fraction for Isotope i (i = 234, 235, 16-18). For uranium nitrate solutions prepared directly from
236, 238): uranium metal reference materials such as CRM112A and
n ~ iU ! n ~ iU ! Ri CRM116A, the precision is usually better (RSD < 0.01 %) (16,
A ~ iU ! 5 5 5 (13) 17) compared to solutions prepared from oxides or other
n~U! Σn ~ j U ! ΣR j
j5234, 235, 236, 238 j5234, 235, 236, 238 uranium compounds (RSD only < 0.03 %). Examples for
14.7.2 Isotope Mass Fraction for Isotope i (i = 234, 235, within-turret precisions obtained in MTE measurements are
236, 238): given in A1.1.1.
15.2.2 For the corrected major ratio, 235U/238U, the
m ~ iU ! R iM i
W ~ iU ! 5 5 (14) between-turret precision varies between 0.005 and 0.01 %
m~U! ΣR j M j
j5234, 235, 236, 238 RSD as reported in Refs (5-7, 13, 14, 16-18). An example for
between-turret precisions obtained in MTE measurements is
14.7.3 Molar Mass:
shown in A1.1.2.
ΣR M
j j 15.2.3 Both the within-turret precision and the between-
j5234, 235, 236, 238
M~U! 5 (15) turret precision are found to be independent of the value of the
ΣR j
j5234, 235, 236, 238 major isotope amount ratio (5-7, 13, 14, 16-18).
where: 15.3 Bias for MTE Measurements of the Major Ratio,
235
n(iU) = number of moles for isotope i; U ⁄ 238U:
n(U) = number of moles for all isotopes, usually 234U, 15.3.1 No significant biases (that is, deviations of measured
235
U, 236U, and 238U; and corrected ratios from the known or certified ratios) for the
i
m( U) = mass for isotope i; major ratios, 235U/238U, were observed for CRMs of the NBL
m(U) = mass for all isotopes, usually 234U, 235U, 236U, and U series CRMs (5, 6, 14).
238
U; 15.3.2 For the IRMM reference materials, the IRMM-183-
Ri = isotope amount ratio (i = 234, 235, 236, and 238) 187 series (19), IRMM-072/1, the IRMM-075/1-6 series, the
relative to one (no matter which) of the isotopes; IRMM-074/1-10 series, and IRMM-199, also no significant
and biases were observed (6, 9, 14). As an example discussed in
Mi = molar mass for isotope i. detail in Ref (9) and with data shown in A1.1.3, MTE
measurements of the 235U/238U isotope amount ratios for the
15. Evaluation and Statements for Precision and Bias gravimetrically prepared mixtures IRMM-072, IRMM-199,
15.1 A total of six thermal ionization mass spectrometers at IRMM-074-1, and IRMM-074-Mix1 are presented within this
IRMM, ITU, NBL, and the SGAS (at the IAEA) were used to test method to serve for a bias statement. The sample loading
determine the precision and bias for uranium isotope amount was 5 µg of uranium for all materials. IRMM-072-1 was used
ratio measurements via the MTE method. Only TIMS instru- as a CRM for the K-factor calculation. According to Table
ments of the type TRITON™ were used for the analyses, A1.3, the results for the corrected 235U/238U ratios for IRMM-
because this is currently the only TIMS instrument with MTE 199, IRMM-074-1, and IRMM-074-Mix1 agree well with their
capabilities. The following uranium certified reference materi- respective certified values, showing that the used MTE method
als were used: provides accurate data. As a conclusion from these results, the

14
 C1832 − 16
maximum bias for MTE measurements of the 235U/238U natural uranium sample with proven low 236U (<10-10) relative
isotope amount ratio can be stated to be 0.02 %. abundance. The “bias” for measuring minor ratios using the
15.3.3 A series of comparative measurements using the MTE method obtained at IRMM and IAEA-SGAS using the
MTE method and the “double spike” (DS) method using a IRMM-075 series (and some further materials with minor
233
U/236U double spike was performed (6) to investigate the ratios confirmed by several laboratories, for example, IRMM-
effect of the external mass fractionation correction used for 184 and samples from the REIMEP-18 interlaboratory com-
MTE measurements in comparison to the internal mass frac- parison (21)) is shown in A1.1.6 by tables and figures (similar
tionation correction, which is used for the much more precise to Fig. 4 and 5 in Ref (6)). In Fig. A1.5, on the ordinate the
DS method. The double spike IRMM-3636 is characterized by absolute values of the (absolute) deviation of the measured
a 233U/236U ratio of about one with a relative uncertainty of
from the certified values are added to the (absolute) measure-
0.016 % (coverage factor k = 2) (20). Because of this low
ment uncertainties (with coverage factor k = 2) and plotted
uncertainty and the much better precision of the DS method
against the values of the measured ratios. This parameter on the
(RSD < 0.003 %) compared to MTE (RSD < 0.03 %), the
comparison between the MTE and the DS results is also a ordinate was introduced as the “performance” of MTE mea-
measure of the bias of the MTE method. As shown in Ref (6), surements in Ref (6), and is more meaningful to characterize
the relative difference between MTE and DS results for the the MTE method in terms of bias or accuracy compared to only
major ratio, 235U/238U, is below 0.02 %, thereby supporting the the absolute (or relative) difference between the measured and
bias statement made in 15.3.2. certified values. The advantage is that by adding the uncertain-
ties (see Section 16) to the difference, the full range of possible
15.4 Precision for the Minor Ratios 234U ⁄ 238U and
236 deviations or biases according to the measurement conditions
U ⁄ 238U:
is included. All ratios below about 3 × 10-5 are measured using
15.4.1 In contrast to the major ratios, the precision for the
minor ratios 234U/238U and 236U/238U strongly depends on the the SEM detector; all ratios above this limit were measured
value (in particular on the order of magnitude) of the ratios, as using Faraday cups only.
well as on the loaded sample amount and the detector used and 15.5.2 For comparative TE measurements on a TRITON™
is discussed in detail in several publications (5-7, 14, 16, 17, TIMS and a predecessor MAT261 TIMS, the performances (as
18, 21). There is no difference between the within-turret and defined in 15.5.1) are almost ratio independent at a level of
the between-turret precisions because all minor ratios are about 10-6 (that is, 1 ppm) and 10-5 (that is, 10 ppm),
corrected for mass fractionation internally using the (corrected) respectively, which is more than a factor of 30 to 10 000 higher
major ratios. (that is, inferior) compared to the performance of the MTE
15.4.1.1 An example for precisions of the minor ratios, measurements. The reason for the performance for TE mea-
234
U/238U and 236U/238U, obtained in MTE measurements surements being almost independent of the ratio is given by the
using Faraday cups (equipped with 1012 Ω resistors) is shown fact that peak tailing corrections were not performed for the TE
in A1.1.4. Data are taken from Ref (7). The precision is measurements and no SEM detector with energy filter for
characterized by the (Johnson’s) noise of the amplifier resis- improved abundance sensitivity was used.
tors. Therefore, the absolute precision (SD) of repeated mea-
surements is independent on the ratios. As a consequence, the 16. Measurement Uncertainty
relative precision RSD is decreasing (and thereby improving)
logarithmically with increasing value of the ratios. The slope in 16.1 Major Isotope Amount Ratios 235U ⁄ 238U—The pro-
the double logarithmic plot in Fig. A1.2 is, therefore, close to portions of the uncertainty contributions (see 14.6.2) depend on
minus one. the precisions and number of replicates of the MTE measure-
15.4.1.2 An example for precisions for the minor ratios, ments for sample and reference material, as well as on the
234
U/238U and 236U/238U, obtained in MTE measurements uncertainty of the certified major isotope amount ratio of the
using SEM detectors is shown in A1.1.5. Data are taken from reference material. The uncertainties of the 235U/238U major
Refs (7, 10, and 22). In this case, the precision is characterized isotope amount ratios using MTE were found to be indepen-
by the counting (Poisson) statistics; therefore, the RSD is dent of the ratios and their magnitude range between 0.04 and
decreasing (and thereby improving) with the square root of the 0.12 %.
increasing number of counts and, therefore, also with square
root of the increasing ratios (6, 7). The slope in the double 16.2 Minor Isotope Amount Ratios 234U ⁄ 238U and 236U/
238
logarithmic plot in Fig. A1.3 is, therefore, close to minus one U—Considering the minor isotope amount ratios, the uncer-
half. tainty budgets are quite complex because of the large dynamic
range of the isotope amount ratios and the different types of
15.5 Bias for Minor Ratios 234U ⁄ 238U and 236U ⁄ 238U:
15.5.1 The evaluation of biases for the minor ratios, 234U/ detectors used.
238
U and 236U/238U by MTE is primarily performed using 16.2.1 When 234U is measured on the SEM (“reverse”
CRMs for which also the minor ratios are assigned using a MTE), usually the main source of uncertainty is the contribu-
gravimetrical preparation. One available reference material of tion from the SEM calibration versus the Faraday cups using
this kind is the IRMM-075 series (10) for which the 236U/238U the 236U isotope. The same can be observed when 236U is
ratios are directly traceable to the SI because of the gravimetri- measured on SEM and 234U used for the intercalibration
cal mixing of a highly enriched (>99.96 %) 236U spike with a (“non-reverse” MTE), except for low count rates.

15
 C1832 − 16
16.2.1.1 For the calibrations of the SEM versus the Faraday than for the 236U/238U ratios. To avoid this effect, for example,
cups (see 12.9) and the correction in terms of linearity (see for HEU samples with better measurement precision for the
12.8), a (combined) relative uncertainty of 0.4 % (with cover- minor ratios, the uranium isotope amount ratios are usually
age factor k = 2) can be assigned, as explained, for example, in expressed using 235U in the denominator, which means as
Ref (6). This estimation was confirmed by measuring 236U/ 234
U/235U, 236U/235U, and 238U/235U. In this case, the uncer-
238
U ratios of IRMM-187 and NBL CRM U010 using both the tainty contribution from the mass fractionation correction
SEM and Faraday cups at various facilities and instruments becomes smaller as a result of the smaller mass differences for
(see Ref (6)), see also 12.9.5. the minor ratios 234U/235U, 236U/235U compared to the major
16.2.1.2 With decreasing count rate on the SEM for the ratio 238U/235U.
minor isotopes, the counting statistics and a possible back- 16.2.2.3 For minor isotope amount ratios between these two
ground correction become increasingly dominant in the uncer- ranges, that is, between 10-4 and 3 × 10-4, a varying contribu-
tainty budget. tion of 40 and 60 % to the uncertainty budget comes from the
16.2.2 When the minor isotopes are measured on Faraday precision of the measurements and the mass fractionation
cups, the two main contributions come from the mass fraction- correction.
ation correction and the amplifier noise, which influences the 16.2.3 Expanded uncertainties for the minor isotope amount
precision (see 15.4.1.1). Therefore, the uncertainties depend on ratios 234U/238U and 236U/238U (coverage factor k = 2 corre-
the values of the isotope amount ratios. sponding to a level of confidence of about 95 %) are repre-
16.2.2.1 If the isotope amount ratios are below 10-4, the sented versus the isotope amount ratios in A1.1.6 (Table A1.6
main contribution comes from the amplifier noise and and Table A1.7 and Fig. A1.4).
precision, whereas the mass fractionation correction represents
only a minor contribution for the uncertainty budget. 17. Keywords
16.2.2.2 If the minor isotope amount ratios are above 3 × 17.1 certified reference materials; ion counting; isotope
10-4, the contribution of the mass fractionation correction is the amount-ratios; mass fractionation correction; secondary elec-
main source of uncertainty. Because of the larger mass tron multiplier; thermal ionization mass spectrometry; total
difference, this contribution is larger for the 234U/238U ratios evaporation; uranium

ANNEX

(Mandatory Information)

A1. PRECISION AND BIAS FOR MTE MEASUREMENTS

A1.1 A discussion and statements regarding precision and A1.1.5 Precision (% RSD) for MTE measurements of minor
bias for MTE measurements are given in Section 15. Associ- ratios 234U/238U and 236U/238U using a SEM detector, taken
ated tables and figures are presented in A1.1.1 – A1.1.6. from Refs (7, 10, and 22). See Table A1.5 and Fig. A1.3.
A1.1.1 See Table A1.1. A1.1.6 In Table A1.6 and Table A1.7 and Fig. A1.4 and Fig.
A1.1.2 See Table A1.2. A1.5 are shown the MTE measurement uncertainty (coverage
A1.1.2.1 The data are also shown in Fig. A1.1. factor k = 2) and the measurement “performance,” defined as
the sum of the (absolute) deviation of the measured from the
A1.1.3 MTE measurements of the 235U/238U isotope certified value plus the (absolute) uncertainty (see 15.5.1), for
amount ratio for IRMM-072-1, IRMM-199, IRMM-074-1, and MTE measurements of minor ratios 234U/238U and 236U/238U
IRMM-074-Mix 1 (9). See Table A1.3. using a TRITON™ TIMS. The sample loading is 5 µg, for 236U
A1.1.4 Precision (% RSD) for MTE measurements of minor a SEM detector or a Faraday detector is used, and only a
ratios 234U/238U and 236U/238U on Faraday cups (amplifiers Faraday detector for 234U. Additionally, regression lines for the
with 1012 Ω resistors), taken from Ref (7). See Table A1.4 and performance of TE measurements performed on MAT262 and
Fig. A1.2. TRITON™ TIMS instruments are shown for comparison.

16
 C1832 − 16
235
TABLE A1.1 Within-Turret Precision (% RSD) for MTE Measurements of the Major Ratio U/238U
235 238 235
CRM U/ U U/U, % Average RSD, %
Ref (5) in 2003
U005a 0.005 090 0.5 0.028
U010 0.010 140 1.0 0.028
U030A 0.031 367 3.0 0.017
U050 0.052 784 5.0 0.024
U100 0.113 596 10.2 0.016
U200 0.251 259 20.1 0.040
U500 0.999 698 50.0 0.025
U750 3.166 130 76.0 0.022
U900 10.375 704 91.2 0.019
Ref (7) in 2013/2014
IRMM-019 0.00177 0.2 0.014
IRMM-020 0.00296 0.2 0.013
IRMM-021 0.00405 0.4 0.018
IRMM-022 0.00756 0.7 0.009
IRMM-023 0.03381 3.3 0.012
IRMM-024 0.05354 5.1 0.009
IRMM-025 0.02036 2.0 0.010
IRMM-026 0.02579 2.5 0.018
IRMM-027 0.04117 4.0 0.010
IRMM-028 0.03776 3.6 0.022
IRMM-029 0.04452 4.2 0.018
Ref (18) in 2013
CRM 115 0.002034 0.2 0.019
Ref (16) in 2013
CRM112A (from metal) 0.00754 0.7 <0.010
CRM145 (from metal) 0.00754 0.7 <0.010
CRM030A 0.03167 3.0 <0.020
Ref (17) in 2014
CRM116A (from metal) 19.498803 93.2 <0.010
CRM U900 10.375704 91.2 0.010

TABLE A1.2 Between-Turret Precision

235
Relative Deviation of corrected U/238U ratios from certified values. Additional IRMM-187 samples loaded on each turret 1–6 were used for K-factor determination.
Turret IRMM-187, % IRMM-186, % IRMM-185, % IRMM-184, %
1 0.001 0.022 0.016 -0.020
2 -0.001 0.026 0.022 -0.026
3 0.001 0.032 0.016 -0.019
4 0.000 0.023 0.017 -0.011
5 -0.002 0.020 0.016 -0.023
6 0.001 0.019 0.015 -0.018
Summary:
RSDA 0.0012 0.0048 0.0027 0.0050
A
The between-turret precision (% RSD) varies between 0.0012 and 0.0050 % RSD.

17
 C1832 − 16

FIG. A1.1 Between-Turret Precision

235
TABLE A1.3 Bias for MTE Measurements of the Major Ratio U/238U
235 238 235 238
Corrected U/ U Certified U/ U Rel. Dev. from Certified Value
IRMM-072-1 Used for K-Factor 0.991 03(20) N/A
IRMM-199 1.000 18(36) 1.000 15(20) 0.003(42) %
IRMM-074-Mix1 1.000 09(43) 1.000 26(15) -0.017(44) %
IRMM-074-1 1.000 22(32) 1.000 25(15) -0.003(33) %

234
TABLE A1.4 Precision (% RSD) for MTE Measurements of Minor Ratios U/238U and 236
U/238U on Faraday Cups
CRM Ratio Ratio Value Average % RSD
234
IRMM-021 U/238U 0.000 024 85 0.17
236
IRMM-019 U/238U 0.000 036 52 0.25
234
IRMM-022 U/238U 0.000 053 28 0.12
234
IRMM-025 U/238U 0.000 122 45 0.042
236
IRMM-025 U/238U 0.000 148 39 0.030
234
IRMM-026 U/238U 0.000 149 41 0.041
236
IRMM-026 U/238U 0.000 207 30 0.060
234
IRMM-027 U/238U 0.000 231 59 0.037
236
IRMM-020 U/238U 0.000 286 15 0.033
234
IRMM-024 U/238U 0.000 290 75 0.016
234
IRMM-023 U/238U 0.000 339 50 0.016
236
IRMM-027 U/238U 0.000 387 39 0.020
236
IRMM-024 U/238U 0.000 516 96 0.014
234
IRMM-028 U/238U 0.000 610 41 0.0066
234
IRMM-029 U/238U 0.000 844 4 0.014
236
IRMM-028 U/238U 0.005 194 3 0.0015
236
IRMM-029 U/238U 0.010 556 3 0.0022

18
 C1832 − 16

234
FIG. A1.2 Precision (% RSD) for MTE Measurements of Minor Ratios U/238U and 236
U/238U on Faraday Cups

234
TABLE A1.5 Precision (% RSD) for MTE Measurements for Minor Ratios U/238U and 236
U/238U, Using a SEM Detector for 234
U or 236
U
CRM Ratio Ratio Value Average % RSD
236
IRMM-075-6 U/238U 0.000 000 001 088 9 11.5
236
IRMM-075-5 U/238U 0.000 000 010 651 9 0.91
236
S3-1-2, BC 2692 U/238U 0.000 000 011 9 0.33
236
IRMM-021 U/238U 0.000 000 026 6 0.67
236
S3-1-1, BC 2691 U/238U 0.000 000 105 1 0.27
236
IRMM-023 (also IRMM-075-4) U/238U 0.000 000 115 3 0.49
236
IRMM-022 U/238U 0.000 000 241 5 0.35
236
IRMM-075-3 U/238U 0.000 001 040 93 0.12
234
IRMM-019 U/238U 0.000 006 85 0.25
236
IRMM-075-2 U/238U 0.000 011 416 0.073
234
IRMM-020 U/238U 0.000 011 92 0.037

19
 C1832 − 16

234
FIG. A1.3 Precision (% RSD) for MTE Measurements for Minor Ratios U/238U and 236
U/238U, Using a SEM Detector for 234
U or 236
U

TABLE A1.6 MTE Measurement Uncertainty and Measurement Performance

236
CRM, Sample U/238U Rel. U, k = 2, % “Performance” ABS (Dev.) + U (k = 2)
REIMEP 18-C 0.000 982 015 0.036 6.30 × 10-7
REIMEP 18-B 0.000 388 280 0.034 1.89 × 10-7
IRMM-075-1 0.000 104 433 0.050 1.18 × 10-7
IRMM-187 0.000 071 965 0.074 7.57 × 10-8
IRMM-075-2 0.000 011 416 0.23 3.73 × 10-8
IRMM-075-3 0.000 001 040 93 0.30 3.49 × 10-9
IRMM-184 0.000 000 124 46 0.47 7.43 × 10-10
IRMM-075-5 0.000 000 010 651 9 4.03 4.92 × 10-10

TABLE A1.7 MTE Measurement Uncertainty and Measurement Performance

234
CRM, Sample U/238U Rel. U, k = 2, % “Performance” ABS (Dev.) + U (k = 2)
Close to naturalA 0.00005286 0.12 8.70 × 10-8
NBL 005a 0.00003417 0.16 7.47 × 10-8
REIMEP 18-C 0.00007947 0.092 1.11 × 10-7
REIMEP 18-D 0.00020931 0.071 1.96 × 10-7
REIMEP 18-B 0.00033272 0.067 2.33 × 10-7
IRMM-187 0.00038700 0.056 3.15 × 10-7
A
Average from measurements of IRMM-184, IRMM-075-1, NBL U010, and NBL CRM 129A.

20
 C1832 − 16

FIG. A1.4 MTE Measurement Uncertainty

FIG. A1.5 MTE Measurement Performance

21
 C1832 − 16
REFERENCES

(1) Callis, E. L. and Abernathy, R. M., “High-Precision Isotopic Analyses Mixtures,” EUR 22271 EN, 2006.
of Uranium and Plutonium by Total Sample Volatilization and Signal (12) BIPM, “Joint Committee for Guides in Metrology, Evaluation of
Integration,” Int. J. Mass Spectrom. Ion Proc., Vol 103, 1991, pp. Measurement Data—Guide to the Expression of Uncertainty in
93–105. Measurement,” JCGM, Vol 100, 2008.
(2) Fiedler, R. et al.,“Report on Preliminary Experience with Total (13) Bürger, S., Essex, R. M., Mathew, K. J., Richter, S., and Thomas, R.
Evaporation Measurements in Thermal Ionization Mass B., in International Journal of Mass Spectrometry, Vol 294, 2010,
Spectrometry,” Int. J. Mass Spectrom. Ion Proc., Vol 132, 1994, pp. pp. 65–76.
207–215. (14) Bürger, S., Balsley, S. D., Baumann, S., Berger, J., Boulyga, S. F., et
(3) Fiedler, R., “Total Evaporation Measurements—Experience with al, “Uranium and Plutonium Analysis of Nuclear Material Samples
Multi-Collector Instruments and a Thermal Ionization Quadrupole by Multi-collector Thermal Ionisation Mass Spectrometry: Quality
Mass-Spectrometer,” International Journal of Mass Spectrometry, Vol Control, Measurement Uncertainty, and Metrological Traceability,”
146, 1995, pp. 91–97. International Journal of Mass Spectrometry, Vol 311, 2012, pp.
(4) Mathew, K. J., O’Connor, G., Hasozbek, A., and Kraiem, M., “Total 40–50.
Evaporation Method for Uranium Isotope-amount Ratio (15) Richter, S., Alonso, A., De Bolle, W., Wellum, R., and Taylor, P. D.
Measurements,” Journal of Anal. At. Spectrom., Vol 28, 2013, pp. P., in International Journal of Mass Spectrometry, Vol 193, 1999, pp.
866–876. 9–14.
(5) Richter, S. and Goldberg, S. A., “Improved Techniques for High (16) Mathew, K. J., Mason, P., Voeks, A., and Narayanan, U., “Uranium
Accuracy Isotope Ratio Measurements of Nuclear Materials Using Isotope Abundance Ratios in Natural Uranium Metal Certified
Thermal Ionization Mass Spectrometry,” International Journal of
Reference Material 112-A,” International Journal of Mass
Mass Spectrometry, Vol 229, 2003, pp. 181–197.
Spectrometry, Vol 315, 2012, pp. 8–14.
(6) Richter, S., Kühn, H., Aregbe, Y., Hedberg, M., Horta-Domenech, J.,
(17) Mathew, K. J., Essex, R. M., Hasozbek, A., Orlowicz, G., and
et al, “Improvements in Routine Uranium Isotope Ratio Measure-
Soriano, M., “Uranium Isotope-Amount Ratios in Certified Refer-
ments Using the Modified Total Evaporation Method for Multi-
ence Material 116-A-Uranium(enriched) Metal Assay and Isotopic
Collector Thermal Ionization,” Mass Spectrometry Journal of Anal.
Standard,” International Journal of Mass Spectrometry, Vol 369,
At. Spectrom., Vol 26, 2011, pp. 550–564.
2014, pp. 48–58.
(7) Mialle, S., Richter, S., Hennessy, C., Truyens, J., Jacobsson, U., et al,
“Certification of Uranium Hexafluoride Reference Materials for (18) Mathew, K. J., Singleton, G. L., Essex, R. M., Hasozbek, A.,
Isotopic Composition,” 8th International Conference on Isotopes, Orlowicz, G., et al, “Characterization of Uranium Isotopic Abun-
Chicago, Aug. 24–29, 2014, J Radioanal Nucl Chem, Vol 305, 2015, dances in Depleted Uranium Metal Assay Standard,” J Radioanal
pp. 255–266. Nucl Chem., Vol 296, No. 115, 2013, pp. 435–440.
(8) Richter, S., Alonso-Munoz, A., Aregbe, Y., Eykens, R., Jacobsson, U., (19) Richter, S., Alonso, A., De Bolle, W., Kühn, H., Verbruggen, A., et
et al, “IRMM-3100a: A New Certified Isotopic Reference Material al, “Re-certification of a Series of Uranium Isotope Reference
with Equal Abundances of 233U, 235U, 236U and 238U,” International Materials, IRMM-183, IRMM-184, IRMM-185, IRMM-186, and
Journal of Mass Spectrometry, Vol 299, 2011, pp. 120–124. IRMM-187,” International Journal of Mass Spectrometry, Vol 247,
(9) Richter, S., Alonso, A., Aregbe, Y., Eykens, R., Kehoe, F., et al, “A 2005, pp. 37–39.
New Series of Uranium Isotope Reference Materials for Investigating (20) Richter, S., Alonso, A., Eykens, R., Jacobsson, U., Kühn, H., et al,
the Linearity of Secondary Electron Multipliers in Isotope Mass “The Isotopic Composition of Natural Uranium Samples - Measure-
Spectrometry,” International Journal of Mass Spectrometry, Vol 281, ments using the new n(233U)/n(236U) Double Spike IRMM-3636,”
2009, pp. 115–125. International Journal of Mass Spectrometry, Vol 269, 2008, pp.
(10) Richter, S., Alonso, A., Aregbe, Y., Eykens, R., Jacobsson, U., et al, 145–148.
“Certification of a New Series of Gravimetrically Prepared Synthetic (21) Richter, S., Alonso, A., Truyens, J., Kühn, H., Verbruggen, A., et al,
Reference Materials for 236U/238U Isotope Ratio Measurements,” in “Evaluating the Status of Uranium Isotope Ratio Measurements
Proceedings of the 11th International Conference on Accelerator Using an Inter-Laboratory Comparison Campaign,” International
Mass Spectrometry, Rome, Italy, Sept. 2008 and Nuclear Instruments Journal of Mass Spectrometry, Vol 264, 2007, pp. 184-190.
and Methods in Physics Research B, Vol 268, 2010, pp. 956–959. (22) Richter, S., Alonso, A., Kühn, H., Verbruggen, A., Wellum, R., et al,
(11) Richter, S., Alonso, A., Eykens, R., Kühn, H., Verbruggen, A., et al, “Preparation and Certification of Synthetic Uranium Isotope Mix-
“Isotopic Measurements of Highly Enriched 233U, 235U, 236U and tures with 236U/238U Ratios of 10-6, 10-7, 10-8,” Journal of Analytical
238
U Starting Materials for the Preparation of Synthetic Isotope Atomic Spectrometry, Vol 20, 2005, pp. 1381–1385.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org). Permission rights to photocopy the standard may also be secured from the Copyright Clearance Center, 222
Rosewood Drive, Danvers, MA 01923, Tel: (978) 646-2600; http://www.copyright.com/

22