Improvement of precision spectrophotometric method with internal reference and its application to analysis of plutonium solutions

Article abstract of DOI:10.1134/S1066362206020196

A spectrophotometric precision method with internal reference [1, 2] was applied to analysis of straight Pu solutions for certifying Pu reference materials and for studying the PuO₂ solubility in the framework of developing methods for accounting and control of nuclear materials. In the context of the activities mainly concerned with certification of reference materials, a two-channel spectrophotomer and the corresponding technique were further improved in order to decrease the error of the method. This allowed the random component of the relative error of the method to be decreased from 0.1 to 0.04% at the confidence level p = 0.95 and the number of degrees of freedom f = 25. The fixed component of the error of the method was studied in relation to impurities of U, Np and corrosion products of structural materials. Also, the extent of Pu oxidation during sample preparation was studied as influenced by the fluoride ion. The revealed lack of such influence within the limits of the study indicates that the method is suitable for analysis of Pu in mixed solutions. Pleiades Publishing, Inc., 2006.

Full text of DOI:10.1134/S1066362206020196

ISSN 1066-3622. Radiochemistry, 2006, Vol. 48, No. 2, pp. 198 203. Pleiades Publishing, Inc., 2006.
Original Russian Text
A.V. Stepanov, D.A. Stepanov, S.A. Nikitina, T.D. Gogoleva, M.G. Grigor’eva, L.S. Bulyanitsa, Yu.A. Panteleev, M.V. Pen’kin,
E.V. Pevtsova, V.D. Domkin, 2006, published in Radiokhimiya, 2006, Vol. 48, No. 2, pp. 179 184.
Improvement of Precision Spectrophotometric Method
with Internal Reference and Its Application to Analysis
of Plutonium Solutions
A. V. Stepanov* , D. A. Stepanov*, S. A. Nikitina*, T. D. Gogoleva*, M. G. Grigor’eva*,
L. S. Bulyanitsa* , Yu. A. Panteleev*, M. V. Pen’kin**,
E. V. Pevtsova*, and V. D. Domkin*
* Khlopin Radium Institute, Research and Production Association, St. Petersburg, Russia
** International Atomic Energy Agency, Vienna, Austria
Received February 1, 2005
Abstract A spectrophotometric precision method with internal reference [1, 2] was applied to analysis of
straight Pu solutions for certifying Pu reference materials and for studying the PuO2 solubility in the frame-
work of developing methods for accounting and control of nuclear materials. In the context of the activities
mainly concerned with certification of reference materials, a two-channel spectrophotomer and the corre-
sponding technique were further improved in order to decrease the error of the method. This allowed the ran-
dom component of the relative error of the method to be decreased from 0.1 to 0.04% at the confidence level
p = 0.95 and the number of degrees of freedom f = 25. The fixed component of the error of the method was
studied in relation to impurities of U, Np and corrosion products of structural materials. Also, the extent of Pu
oxidation during sample preparation was studied as influenced by the fluoride ion. The revealed lack of such
influence within the limits of the study indicates that the method is suitable for analysis of Pu in mixed
solutions.
PACS numbers: 82.80.Dx
DOI: 10.1134/S1066362206020196
Enhancement of the precision of methods for deter-
mination of U and Pu in various nuclear materials re-
mains urgent in view of the need in their accounting
and control. This implies improving precision meth-
ods for determination of U and Pu in certification of
the corresponding reference materials (RMs).
filtered (white) light simultaneously in two sections of
the spectrum. One of these sections incorporates the
analytical absorption band of the element to be deter-
mined, and the other (reference band) is a transparent
spectral window acting as internal reference. The ana-
lytical signal in this case is the ratio of the intensities
R in the transparent window and in the absorption
band, rather than the absolute value of the intensity of
the light that passed through the solution at the ab-
sorption band.
Development of RMs of Pu certified to both the
isotope composition and mass fraction of Pu in the
material is under way now in Russia, and Radium In-
stitute is actively engaged in this work. These materi-
als will be distributed among Pu-handling enterprises
in Russia to ensure consistent measurements of the
isotope composition and total mass of Pu. Clearly,
such RMs should be certified to the highest possible
precision. In this connection, Radium Institute under-
takes improvement of the existing methods and in-
struments, in particular, the spectrophotometric proce-
dure with internal reference (IRSP) [1, 2].
The internal referencing principle was implemented
with a two-channel spectrophotometric analyzer de-
signed at the Radium Institute [1]. In this instrument,
the light sent by an incandescent lamp passes through
a cell containing the test solution, whereupon it is
split into two monochromatic beams: The absorption
band is cut with an MDR-23 monochromator, and
the reference band, with a narrow-band interference
filter having a transmission band width of 5 nm.
This method consists in measuring the light absorp-
tion of the test solution when illuminated with non-
Two light guides send the light to a measurement
channel switching unit. This is a rotating disk with
two slitlike slots (one slot for each measurement
Deceased.
198

IMPROVEMENT OF PRECISION SPECTROPHOTOMETRIC METHOD
199
channel), sharing the case with the measurement sys-
tem consisting of a photomultiplier, one-electron
pulse amplifier, and signal formation and synchroniza-
tion units. The electric background of the instrument
is measured at the disk position when both light
guides are shielded. The measurements are carried out
in a cyclic (each cycle lasts 2.2 s) mode under preset
control. The measurement in one cycle yields the ratio
of the light intensities R, corrected for the dead time
of the instrument.
where R , R , and R are the light intensity ratios at
the absorption and reference bands for the test and
x
1
2
two reference (calibration) solutions; C , Pu concen-
x
tration in the test solution; and C and C , Pu concen-
1
2
trations in the two reference (calibration) solutions,
chosen so that C < C < C .
1
x
2
The three (two calibration and one test) solutions
are subjected to photometeric measurements in the
same 1-cm cell (after washing) within the total time of
20 s. During one working day, 2 unknown solutions
can be analyzed with 5 6 observations for each of
them (the instrument achieves the working regime
1.5 h after turning on).
The variability of the intensities in both spectral
sections can be due to instability of the light source,
drift of the amplification coefficient of the amplifica-
tion tract, poor reproducibility of the geometric posi-
tion of the cell, variability of the temperature of the
solution, etc. To achieve as rigorous correlation bet-
ween the signals as possible, the measurements are
carried out virtually simultaneously, using the same
amplification tract, and the measurement cycles are
many times replicated. To ensure identical conditions
of the photomultiplier operation when recording the
light intensities at the analytical and reference bands,
the ratio R should be close to unity; this is achieved
by setting neutral light filters in the reference channel.
We utilized the internal referencing principle in our
previous studies on precision determination of Pu, U,
Rh, and Nd [1, 2] based on their intrinsic absorption,
as well as for development of a method of precision
determination of Th and Pu in Arsenazo III solutions
[4]. The RRMSD of the results of determining the
concentrations of these elements with these techniques
lies within 0.1 0.3%.
Determination of Pu(VI) in a nitric acid solution
[1] utilizes a very intense band at 830.2 nm (herein-
after, 830 nm for brevity) with the molar absorption
In view of the fact that all the instrumental errors
are mutually correlated, the relative random error of
measuring the ratio of the light intensities R at close
intensities can be determined by the expression [3]
1
1
coefficient 500 l mol cm , and a transparent seg-
ment at 700 800 nm (in this study, 720 nm) served
as the reference band.
Along with the random error of measuring the light
intensities, the method is characterized by the fixed
error. It has the following sources: certification of the
reference material; weighing in the sample prepara-
tion; nonlinear response of the recording system to
the change in the analytical parameter, Pu concentra-
tion (this nonlinearity is mainly due to inaccurate
determination of the coefficient taking into account
the dead time of the recording system and the intensity
of the input signals); temperature instability of solu-
tions during photometric measurements; and influence
of chemical impurities.
2
2
= 2 (1
r),
(1)
R
where
is the relative rms deviation (RRMSD) of
the results of direct intensity measurements;
RRMSD of the intensity ratios R; and r, correlation
,
R
coefficient.
Expression (1) shows that RRMSD can decrease at
the correlation coefficient r > 0.5 only. Previously [1],
we estimated the correlation coefficient at 0.97.
The number of the measurement cycles n can be
preset at 1 250, which allows decreasing the random
error by a factor of n . The measurement results are
Solving the problems involved in application of the
IRSP method to determination of Pu in mixed solu-
tion, reducing the fixed and random errors of the
method, and widening the scope of its application
called for further improvement of the analytical in-
strumentation and technique.
1/2
summed up, averaged, and computer-processed. Fur-
ther increase in the number of cycles has no sense, as
this significantly increases the measurement time,
which can be responsible for nonidentical conditions
of the photometric measurement for the calibration
and test solutions. The concentration of the element in
solution, measured by the two-reference method, is
calculated by the formula
IMPROVEMENT OF INSTRUMENTATION
AND SOFTWARE
To decrease the random component of the error of
the method, we undertook improvement of the detec-
tion system and its software. This improved the time
(C1
C2)(logRx
logR1)
Cx = C1
+
,
(2)
logR1 logR2
RADIOCHEMISTRY Vol. 48 No. 2 2006

200
STEPANOV et al.
1
0.8% deg . Evidently, such a temperature dependence
is due not only to physical expansion of aqueous solu-
1
tions, whose coefficient is only 0.03% deg , but also
to a decrease in the population of the ground f levels
involved in the corresponding electronic transitions of
Pu and Nd.
In view of a substantial contribution of the tem-
perature to the fixed error of the method, we had to
take fairly hard measures to stabilize the temperature
conditions of measurement of the Pu concentration.
In the room where the two-channel spectrophotometer
is located, an automatic conditioner was set to main-
tain a constant air temperature. Also, the optical cell
holder is designed in such a way that the temperature
of the cell can be maintained constant using a water
ultrathermostat. These measures, combined with a
brief time taken by one measurement series ( 20 min),
allowed decreasing the temperature component of the
relative fixed error to 0.04%.
Fig. 1. Temperature dependence of the optical density of
the Pu(VI) solution at 830 nm.
resolution of the system in recording one-electron
pulses and made the operation of the instrument more
stable via decreasing the interferences.
We created a new signal amplifier with a photo-
multiplier, which decreased the dead time from 380 to
80 ns and increased the signal/noise ratio from 150 to
1000. We chose the optimal discrimination threshold
of the comparator and ensured its additional stabiliza-
tion. To decrease the electric and other interferences,
we chose an appropriate connection point for the
analog ground and the instrument case. We revealed
an additional source of error, magnetic interferences
from electric motor rotating the channel switching
disk. In this design, setting a magnetic shield is trou-
blesome. Therefore, we modified the software of the
instrument, namely, chose the observation time strict-
ly divisible by the interference period. The resistance
to interferences was enhanced via changing the
scheme of formation of signals from the sensors of
the position of the rotating disk for channel switching.
We developed a microcontroller-based scheme for
controlling the instrument operation, collecting the
data obtained, and sending them to a PC. We devel-
oped an appropriate controlling program for the mi-
crocontroller. Also, a computer program was devel-
oped that allows the operator to control the instrument
in the Windows environment, acquire information and
carry out its statistical processing, print the results,
and store them in a database.
EXPERIMENTAL
Sample preparation. The initial sample of Pu
1
with a concentration close to 10 mg g was prepared
by dissolving a reference material of metallic Pu
[OSO (Branch Reference Material) 95.086 89] in hy-
drochloric acid which was further removed by adding
HNO and heating the solution on a water bath. The
3
solutions for photometric measurements with a Pu
1
concentration of 200 220 g g in 1 M HNO were
3
prepared from the initial solution of Pu by two-stage
dilution by weight.
For quantitative oxidation of Pu(IV) to Pu(VI) by
the reaction
Pu(IV) + 2Ce(IV)
Pu(VI) + 2Ce(III)
(3)
we used a 0.05 M solution of Ce(IV) in 1 M HNO ,
3
4 2
prepared from analytically pure grade (NH )
Ce(NO
.
3)6
Requirements to the spectral width of the mono-
chromator slit. The measured half-width of the Pu
absorption band is 3.38 nm at the Pu concentration of
Nitric acid solutions of U were prepared from a
U O [GSO (State Reference Material) 2396-83P]
3
8
reference material.
3
0.94 10 M. The optimal spectral width of the
monochromator slit at R 1 was estimated at 1.7 nm.
A reference material of the Np solution composi-
tion [SOP (Reference Material of Enterprise) RI 007-
83P] with the Np concentration (corrected for evapora-
tion) of 1.05 mg g served as the working Np solu-
tion.
Role of temperature. A study of the temperature
dependence of the optical density of the Pu(VI) solu-
tion at 830 nm within 20 30 C showed (Fig. 1) that
the temperature coefficient of the optical density is
1
1
0.4% deg . The experiments with a Nd solution
Solutions simulating corrosive impurities were pre-
pared by the following procedures: Cr(VI) solution,
showed that, for Nd(III), the temperature coefficient
for the band at 794 nm is even greater, namely,
from pure-grade CrO ; Cr(III) solution, by reducing
3
RADIOCHEMISTRY Vol. 48 No. 2 2006

IMPROVEMENT OF PRECISION SPECTROPHOTOMETRIC METHOD
201
Cr(VI) with hydrazine; Fe(III) solution, by dissolving
does not differ substantially from unity; the deviation
is smaller than RRMSD.
Fe (pure grade) metallic powder in 4 M HNO ; and
3
Ni(II) solution, from chemically pure grade Ni(NO )
3 2
The influence of Np was elucidated at the mass
1
6H O. The working solutions of the metals had the
2
fraction of Pu in solution of 210 g g and the
1
following concentrations, mg ml : Fe 2.64, Cr(III)
Np/Pu ratio of 2%. Given below are the Pu concentra-
tions calculated by Eq. (2).
2.7, Cr(VI) 2.15, and Ni 2.41.
The solution of Al(III) was prepared by dissolving
8.75 g of ultrapure-grade Al(NO ) 9H O in 50 ml of
distilled water; 1-day settling yielded a transparent
solution.
Pu/U ratio, Number of RRMSD [(Cx C0)/C0]
3 3
2
%
observations of Cx, %
deviation, %
2
3
0.095
0.11
The deviation of the average found from the taken
Pu concentration was 0.11%, which is comparable to
RRMSD. This suggests that, when in amount of 2%
relative to Pu, the Np impurity does not significantly
interfere with its determination.
INFLUENCE OF IMPURITIES
Determination of Pu in mixed solutions with U
and Np. Considering that the instruments used for
implementation of the IRSP method are fairly simple
and inexpensive and that their use, especially with
Pu-containing materials, allows the mass of Pu in the
samples analyzed to be reduced by approximately an
order of magnitude (compared to electrochemical
methods under comparable precision of the analysis),
it seems promising to extend the application field of
the method from pure Pu solutions to mixed solutions
containing, along with Pu, above all, U.
Influence of corrosion products. To a weighed
portion of Pu containing 2.5 mg of Pu, aliquots of
four solutions [of Cr(VI), Cr(III), Fe, and Ni] were
added simultaneously in amounts such that the con-
centration of each of the impurities was 2% relative to
Pu and all the concentrations made 8% in total. The
solution of Pu with the mixture of corrosion products
was kept for 1 day. Simultaneously with preparation
of the test solution, we prepared (by weight) two
calibration solutions of Pu, free form impurities.
The mass fraction of Pu in the test solution was
The U to Pu concentration ratio in these solutions
was close to that in MOX fuel.
1
In determination of Pu in mixed nitric acid solu-
tions, Pu was not treated to remove U. In preparation
of such solution to analysis, we adjusted the HNO
220 g g . Five observations yielded the following
results: C RRMSD 0.24, and [(C
tion, 0.12%.
C )/C ] devia-
x
x
0 0
3
concentration to 1 M, and the concentration of the
oxidant, Ce(IV), was maintained at 0.05 M.
When in amount of 8%, the sum of the Ni, Fe, and
Cr impurities does not affect substantially the fixed
error in determination of Pu by the IRSP method. In
real products from the plutonium cycle, the content of
the corrosion products is significantly lower.
As calibration solutions (references) we used Pu
solutions free form U. The mass fraction of Pu in the
1
test and calibration solutions was 200 250 g g ;
the references differ in the Pu concentration by under
5%. In view of the difference in the densities of the
calibration and test solutions, due to the presence in
the latter of U, the mass fraction of Pu, calculated by
formula (2), was multiplied by the d /d ratio, where
Influence of aluminum and HF. The reference
materials of the mass fraction of Pu, certified in the
framework of the Program for Development of Rus-
sian Reference Materials, for Accounting and Control
0
x
of Nuclear Materials are represented by PuO , whose
2
d and d are the densities of the calibration and test
0
x
stability in time is ensured through calcinations at
a high temperature (close to 1200 C). However, such
treatment severely complicates the subsequent dissolu-
solutions, respectively. Given below are the data for
Pu in mixed uranium plutonium solutions.
tion of PuO . Obviously, when reference materials of
2
Pu/U ratio, Number of RRMSD [(Cx C0)/C0]
the mass fraction are used in destructive methods,
they should be dissolved quantitatively. We used the
IRSP method for studying the degree of dissolution
%
observations of Cx, %
deviation, %
11
5
2.5
9
9
9
0.2
0.23
0.2
+0.14
0.06
0.02
of high-calcined PuO intended for certification of
2
the RMs.
It is seen that, in all the experiments, the ratio of
Most techniques for transferring high-calcined
the introduced, C , and found, C , Pu concentrations
PuO to a nitric acid solution imply addition of minor
0
x
2
RADIOCHEMISTRY Vol. 48 No. 2 2006

202
STEPANOV et al.
Results of determination of the mass fraction of Pu in a solution of high-calcined PuO2
Number of re-
plicate meas-
urements
Average mass
fraction of Pu,
%
Aliquot
no.
RRMSD,
%
Preparation of aliquots for analysis
1
5
3
4
0.1
0.2
0.1
87.97
88.04
88.12
Without preliminary treatment
Evaporation of F with 3 M HNO3 on a water bath
Evaporation and introduction of 0.25 ml of 1 M Al(NO3)3 per
10 ml of the measured solution in 1 M HNO3
2
3
3
7
0.1
0.1
88.12
88.05
Introduction of 0.25 ml of 1 M Al(NO3)3 per 10 ml of the meas-
ured solution
Introduction of 1 ml of 1 M Al(NO3)3 to 30 ml of the intermediate
solution
amounts of HF. Even after removal of excess HF the
solution can contain the fluoride ion interfering with
the spectrophotometric determination of Pu via hinder-
ing its conversion to the hexavalent state because of
formation of complexes with Pu(IV) and Ce(IV).
HNO ), we evaporated the weighed portions of the
solution on a water bath before introducing Al. The
2
table presents the mass fractions of Pu in PuO , cal-
2
culated from the measured results.
To estimate how the sample preparation affects the
results of determination of Pu by the IRSP method,
we carried out one-factor analysis of the variance of
the tabulated data. Additional operations of binding of
free F ions or their removal from the solution by
evaporation proved to affect the spectrophotometric
results insignificantly. The lack of influence of the
sample preparation allows us to consider all the tabu-
lated data as uniform and belonging to the same popu-
lation.
3+
The logarithms of the stability constants of PuF
2+
and PuF in the perchlorate medium were estimated
at logK = 7.87 and logK = 6.67, respectively [5].
2
1
2
The fluoride complexes of Ce(IV) are comparable in
stability with the Pu(IV) complexes. However, con-
sidering that in the analytical technique the Ce con-
centration is 50 times that of Pu, we can expect pre-
dominant complexing with the F for Ce(IV) rather
3+
than for Pu(IV). However, formation of the CeF
4+
Next, we studied in detail how HF in concentra-
tions of up to 0.1 M affects reaction (3) at the concen-
complex decreases the ionic potential of Ce , thereby
3+
3+
decreasing the redox potential of the CeF /Ce
3
2
4+
3+
trations, M, of Pu of 0.8 10 , Ce, of 5 10 , and
couple relative to the Ce /Ce couple, which is
responsible for a shift of equilibrium (3) to the left.
HNO , of 1.
3
We took weighed portions of Pu and introduced the
required volume of solution of 0.48 M HF, whereupon
the resulting solutions were kept for 3 15 h. Next, a
Ce(IV) solution was added, the volume was brought
to 10 ml, the solutions were weighed, and the molar
concentration of Pu was calculated. The solutions
were subjected to photometric measurements both
immediately after their preparation and in regular
intervals; in further calculations by formula (4) we
used the average values R
For binding free fluoride ions by the reaction
3+
3
Al + 6F
AlF , we introduced a 10 30-fold
6
excess of Al into aliquots of the solutions with the Pu
concentration of ca. 0.001 M before photometric
measurements. Since prolonged storage of Pu solu-
tions in 4.5 M HNO can involve accumulation of
3
the HNO decomposition products (in particular, of
3
D830/720 = log(R /R0) = Cl,
(4)
where R is the average ratio of the light intensities
0
for air, equal to 0.341; R, average ratio of the light
intensities for the solution; l, cell length; and , molar
extinction coefficient.
Fig. 2. Optical density of Pu solution at 830 nm, measured
with respect to the transmission band (720 nm), as a func-
tion of the HF concentration in solution of the composition,
Figure 2 presents the optical density of the solution
at 830 nm, measured with respect to the transmission
3
M: 0.8 10 Pu(VI) 0.05 Ce(IV) 1 HNO .
3
RADIOCHEMISTRY Vol. 48 No. 2 2006

IMPROVEMENT OF PRECISION SPECTROPHOTOMETRIC METHOD
203
band (720 nm), as a function of the HF concentration.
tope dilution mass spectrometry (IDMS) with OSO
A clear inflection is observed at C /C = 1, which
95.836 2002, characterized by a certified value of
HF Ce
239
3+
the mass fraction of
Pu in solution, as the label.
suggests complete binding of Ce into CeF . As C
HF
in solution further increases, a precipitate appears in
The results obtained by two different analytical
methods are totally consistent.
the solution, and the latter loses color; at C
=
HF
0.24 M, CeF precipitates completely.
4
Thus, we improved the instrumentation and tech-
nique for spectrophotometry with internal reference.
The IRSP method was also applied to certification of
the reference Pu materials in studies of the solubility
Thus, our studies showed that the presence of HF
in the solution with the composition of interest (see
above) does not prevent oxidation of Pu and its deter-
mination up to C
= 0.05 M.
HF
of high-calcined PuO . The relative random error
2
achieved was 0.04%. The relative fixed error is deter-
mined by the error for the reference materials used.
Comparison of the results obtained by IRSP and
IDMS revealed no bias.
Also, we examined the molar extinction coefficient
of Pu(VI) as dependent on the time of keeping of the
solution of Pu with C = 0.024 M. Within 2 h, the
HF
molar extinction coefficient remained unchanged,
1
1
488.0 0.6 l mol cm (RRMSD 0.11%).
The influence of U, Np, Fe, Ni, Cr, and Al impuri-
ties and the fluoride anions on the fixed and random
errors of the method was studied. The suitability of
the precision IRSP method was demonstrated, and its
application limits (with RRMSD of 0.1 0.2%) for
analysis of mixed Pu solutions were established.
APPLICATION OF THE IRSP METHOD
TO CERTIFICATION OF THE REFERENCE
MATERIALS
Earlier, we more than once applied the IRSP meth-
od to measuring the mass fraction of Pu in solutions
free from impurities. The reference materials certified
to the mass fraction of Pu using the IRSP method are
as follows: Pu metal composition, GSO 8567 2004;
REFERENCES
1. Stepanov, A.V., Nikitina, S.A., Karasev, V.T., et al.,
Radiokhimiya, 2002, vol. 44, no. 2, pp. 165 169.
2. Stepanov, A.V., Nikitina, S.A., Karasev, V.T., et al.,
Tr. Radievogo Inst. im V.G. Khlopina, 2003, vol. 10,
pp. 50 58.
3. Doerffel, K. und Geyer, R., Analytikum. Methoden der
analytischen Chemie und ihre theoretischen Grund-
lagen, Leipzig: Deutscher Verlag fur Grundstoffin-
dustrie, 1971. Translated under the title Rukovodstvo po
analiticheskoi khimii, Moscow: Mir, 1975, p. 17.
239
240
RM of the composition of solution:
Pu/ Pu, OSO
242
95.836 2002;
tope labels
Pu, GSO 8568 2004; and the iso-
235 238 240
240
U/ U/ Pu and
Pu prepared by
the Radium Institute for the Analytical Laboratory for
Guarantees, IAEA, [6] and Institute of Reference
Materials and Measurements (IRMM) [7].
Given below are the results of the certification
242
analysis of the
Pu solution for the mass fraction of
the total Pu, illustrating the application of the IRSP
method.
4. Grigor’eva, M.G., Nikitina, S.A., Pevtsova, E.V., et al.,
Radiokhimiya, 2003, vol. 45, no. 4, pp. 351 356.
5. The Chemistry of the Actinide Elements, Katz, J.J.,
Seaborg, G.T., and Morss, L.R., Eds., London: Chap-
man and Hall, 1986. Translated under the title Khimya
aktinidov, Katz, J., Seaborg, G., and Morss, L., Eds.,
Moscow: Mir, 1999; vol. 2, p. 333; vol. 3, pp. 407.
For calibration we used solutions of the reference
material of the composition of Pu metal GSO 8567
2004 with the relative error of the mass fraction of
Pu of 0.05%.
Method
), mg g
Number of determina-
tions
IRSP
IDMS
6. Penkin, M.V., Stepanov, A.V., Belyaev, B.N., et al.,
Khlopin Radium Institute Activity in the Field of De-
velopment of the Nuclear Reference Materials for
Safeguard Purposes, Proc. INMM 40th Annual Meet.,
Phoenix (Arizona), July 1999.
7. Stepanov, A.V. et al., Prigotovlenie i analiz rastvora
plutoniya-240 (Preparation and Analysis of a Plutoni-
um-240 Solution1), Report of the Khlopin Radium
Inst., 1995, St. Petersburg, no. 2485.
1
(CPu
0.9155 0.0006 0.9157 0.0019
48
0.06
4
Relative total error, %
0.21
For certification measurements, we took three am-
pules with the
ments were carried out; the relative random error of
242
Pu solution. A total of 48 measure-
the result was 0.04% for the confidence level p = 0.95.
1
240
This
Pu solution was further certified by the Institute for
For control measurement of the mass fraction of
Reference Materials and Measurements (Geel, Belgium) as
Spike Isotopic Reference Material IRMM-083.
242
the total Pu in the
Pu solution, we applied iso-
RADIOCHEMISTRY Vol. 48 No. 2 2006