Separation and recovery study of uranium from spent NaF (fillers)

Article abstract of DOI:10.1016/j.jpcs.2004.07.019

The molten salt technique is considered to be a feasible technique applicable for the removal of uranium from the spent NaF (fillers). Simple spent NaF or NaF with NaCl, i.e. mixed salt, was used as an electrolyte and the uranium concentrations in the bath, pre-treatment condition, electrolytic temperature, current density, etc. were given as parameters in this experiment. Results from the fundamental experiments showed the potential applicability of the molten salt technique for the removal of uranium from spent NaF.

Full text of DOI:10.1016/j.jpcs.2004.07.019

Journal of Physics and Chemistry of Solids 66 (2005) 602–607
www.elsevier.com/locate/jpcs
Separation and recovery study of uranium from spent NaF (fillers)
a,b,*
a
, Takayuki Terai , Hiroshi Oobayashi , Reiko Fujita
c
d
Ippei Amamoto
a
Department of Quantum Engineering and Systems Science, School of Engineering, University of Tokyo, Tokyo 113-8656, Japan
b
Ningyo-toge Environmental Engineering Centre, Japan Nuclear Cycle Development Institute, Okayama 708-0698, Japan
c
Ooarai Engineering Centre, Japan Nuclear Cycle Development Institute, Ibaraki 311-1393, Japan
d
Nuclear Engineering Laboratory, Toshiba Corporation, Kawasaki-shi, Kanagawa 210-0862, Japan
Accepted 3 July 2004
Abstract
The molten salt technique is considered to be a feasible technique applicable for the removal of uranium from the spent NaF (fillers).
Simple spent NaF or NaF with NaCl, i.e. mixed salt, was used as an electrolyte and the uranium concentrations in the bath, pre-treatment
condition, electrolytic temperature, current density, etc. were given as parameters in this experiment. Results from the fundamental
experiments showed the potential applicability of the molten salt technique for the removal of uranium from spent NaF.
q 2004 Published by Elsevier Ltd.
1
. Introduction
electrolysis condition, composition of molten salt, etc. are
presented here.
Various kinds of chemical trap fillers such as sodium
fluoride (NaF), magnesium fluoride (MgF ), or activated
2
alumina (Al O ) are used for the removal of gaseous
2
3
2. Experimental
radionuclides, fluorine, etc. in the process off-gas of
uranium conversion facility [1] or similar facilities.
Gaseous uranium hexafloride (UF ) is removed with the
2
2
.1. Uranium compounds in spent NaF (fillers)
6
D
L
chemical trap charged with NaF (3 mm !3 mm ) as
adsorption fillers (adsorbent). As NaF has to be
recharged when its adsorption efficiency decreases, the
amount of radioactive waste increases during long-term
operation of the facility. Therefore, uranium adsorbed on
NaF should be separated and reused, wherever possible.
A wet treatment process [2], fluoride volatility process
.1.1. Determination of chemical compounds of uranium
in fillers with X-ray diffraction analysis
The spent NaF used in this experiment was from the
uranium conversion facility in Japan and was stored for
several years after its ‘breakthrough’ due to the adsorption of
gaseous UF6.
As it is important to identify the characteristics of the
uranium compounds adsorbed on NaF before experiment,
the composition of spent NaF was investigated by X-ray
diffraction analysis (XD-3AE, Shimadsu Co.). From the
result, the presence of NaF, NaUF , Na UF and Na UF
8
[3], etc. are considered as feasible techniques applicable
for this separation process.
The molten salt technique has been discussed as one of
the promising uranium separation methods [4,5]. Results of
fundamental experiments, the pre-treatment condition,
6
3
6
2
were confirmed but not uranyl fluoride (UO F ) which was
2
2
assumed to form due to reaction of UF with moisture.
6
2
.1.2. Conversion to involatile uranium compounds
NaF$UF (Na UF ), which is formed by the adsorption
*
Corresponding author. Department of Quantum Engineering and
2
Systems Science, School of Engineering, University of Tokyo, Tokyo
6
2
8
1
0
13-8656, Japan.
of gaseous UF on NaF, decompose and release gaseous UF
6 6
022-3697/$ - see front matter q 2004 Published by Elsevier Ltd.
doi:10.1016/j.jpcs.2004.07.019

I. Amamoto et al. / Journal of Physics and Chemistry of Solids 66 (2005) 602–607
603
Fig. 1. Schematic design of reduction apparatus.
Fig. 2. Schematic design of electrolytic separation and recovery cell.
again at high temperature. UF6 is the main substance
volatilized from NaF. Precaution should be taken to prevent
the release of gaseous uranium compounds into the
environment. One way to prevent uranium compounds
from being volatilized at high temperature would be to
3. Results and discussion
3.1. Uranium separation and recovery from simple
spent NaF
convert them into involatile substances such as UF , whose
4
3.1.1. Reduction experiment of adsorbed uranium
on simple spent NaF
vapour pressure is 650 Pa at melting point (1309 K) [6]. In
this experiment, the glass retort attached to the electric heater
was used as a reduction apparatus (refer to Fig. 1). The H gas
The experiment of uranium reduction by H was carried
2
2
out to prevent the volatility of adsorbed uranium on NaF.
100 g of the spent NaF as sample was filled into the
apparatus as shown in Fig. 1. Ar–7%H gas whose volume
was supplied through the glass filter from the bottom of the
retort and exhausted from the top to the off-gas line through
the buffer chamber and HEPA filter. UF could be reduced to
2
6
flow rate was 200 ml/min at 773 K (the maximum reduction
temperature) was used in this experiment. The chemical
UF in reaction with H . UF also forms involatile sodium
4
2
4
6
C
uranium fluorides [7] with NaF such as NaF$UF (NaUF ),
5
4
reaction between adsorbed uranium and H is: U
2
C
4
C
2
NaF$UF (Na UF ),3NaF$UF (Na UF ), etc.
4
H 0U C2HF.
2
2
6
4
3
7
As the white fume of hydrogen fluoride (HF) was
generated by this reduction, the progress of reaction could
be observed visually. The reaction temperature was raised at
2
.2. Experimental condition of molten salt technique
A uranium separation and recovery experiment by the
every 100 K with H gas being constantly and gradually fed
2
to the glass retort charged with NaF. Although the
generation of gaseous HF or UF could not be observed at
molten salt technique was carried out to obtain the recovery
rate of uranium and the electrolytic condition for its
realization. The schematic design of the apparatus used in
the experiment of Run-1, 4, 5 and 6 is shown in Fig. 2. To
prevent the cathode from dissolving, its material was
changed from the low-carbon steel to the high purity
graphite in Run-2. In Run-3, a crucible made of the high
purity graphite (O99.9997%, Nippon Carbon Co., Ltd),
which was ordinary graphite crucible treated at high
temperature, was employed as the anode due to its wider
surface area. The experimental parameters such as electro-
lytic temperature, current density, and uranium concen-
tration in the bath are shown in Tables 1 and 2.
6
373 K, reduction occurred at 473 K because white fume
regarded as HF could be observed. There was no change to
the NaF’s colour at this stage. Then, the reaction
temperature was raised to 573 K after 2 h of heating. At
this point, white fume was re-generated, whereupon, NaF’s
colour changed into yellow green. UF volatilization could
6
not be observed at further rise of temperature from 473 to
573 K. However, when the temperature was raised from 573
to 673 K, gaseous UF was generated. For the reaction
6
temperature from 573 to 773 K, it was raised at every 50 K
instead of the original 100 K, and held on for 2 h. Gaseous
UF could not be observed but NaF’s colour changed from
6

604
I. Amamoto et al. / Journal of Physics and Chemistry of Solids 66 (2005) 602–607
Table 1
Experimental conditions and results of uranium concentration in simple spent NaF at 1327 K
RUN no.
U concentration
before) (wt%)
Cathode size
(mm)
Cathode area
2
(cm )
Current density
2
(A/cm )
Cell voltage
(V)
U concentration
(after) (wt%)
(
OD
10 !30
L
L
L
1
2
3
5.92
7.5
9.4
9.4
3.1
0.2
0.2
6.4
approx. 1.3
approx. 2.8
approx. 4–12
4.63
7.5
OD
10 !30
OD
7.5
5
!20
7.5
(
unstable)
The measurement of current efficiency was unattainable caused by the dispropotionate reaction of uranium deposit.
Table 2
Experimental conditions and results of uranium concentration in spent NaF eutectic at 973 K
RUN no.
U concentration
before) (wt%)
Cathode size
(mm)
Cathode area
2
(cm )
Current density
2
(A/cm )
Cell voltage
(V)
U concentration
(after) (wt%)
Current
(
efficiency
OD
L
L
L
4
5
6
9.47
1.09
0.02
10 !30
9.4
14.1
14.1
0.2
approx. 6
approx. 7
approx. 7
6.55
0.007
61.8
15.2
0.3
OD
15 !30
0.02
0.02
OD
15 !30
!0.001
6C
yellow green to deep green. It was found that U adsorbed
3.1.3. Electrolytic separation and recovery experiment with
high purity graphite anode (RUN-2)
4
C
on NaF could be reduced to U by raising the reaction
temperature gradually (every 50 K) from 473 to 773 K.
The iron cathode was found to dissolve gradually in
RUN-1. Therefore, the cathode material was changed from
iron to high purity graphite for RUN-2, and the experiment
was carried out under the same condition as RUN-1. The
change of uranium concentration in the electrolyte and the
uranium concentration in the deposit on the cathode surface
are shown in Fig. 4. Here, the final uranium concentration in
the deposit is approximately 16 wt%, while the rest is salt as
electrolyte caught in the deposit. For all uranium existence
in the deposit, uranium concentration in the bath was almost
constant, i.e. approximately 8 wt%. This result was
probably due to the disproportionate reaction expressed as
3.1.2. Electrolytic separation and recovery experiment
from simple spent NaF as electrolyte (RUN-1)
6
C
The reduction of U
C
(UF and/or its double salts) to
6
4
U (UF and/or its double salts) would be an effective
4
method to prevent the generation of gaseous UF from
6
spent NaF during heating. The spent NaF treated by H2
was used in the experiment from RUN-1 to RUN-3. The
experimental temperature of the molten salt was 1323 K,
2
and 0.2 A/cm for the cathodic current density. The
electrolodes were inserted into the bath approximately
4
C
3C
4C
‘
3U CU04U ’ between deposited U and U in the
7
electrolytic decontamination was carried out with
.5 min before the start of electrolysis. Then, the
electrolyte but not due to the re-dissolution of the deposit
falling off from the cathode.
2
.88 A (0.2 A/cm of the cathodic current density). The
1
samplings of the bath were made every 1 h, and were
analysed by inductively coupled plasma-atomic emission
spectroscopy (ICP-AES: Seiko Instruments Inc. SPS-
3.1.4. Electrolytic separation and recovery experiment by
high electrolytic current (RUN-3)
7
700). From the analytical results shown in Fig. 3, iron
2
The high electrolytic current, i.e. 20 A (6.4 A/cm for
the cathodic current density), was employed to confirm
from the cathode (low-carbon steel) was found in the
electrolyte together with uranium. The initial uranium
concentration was approximately 6 wt% and it was
0
.12 wt% for the iron concentration. Both (U and Fe)
were likely to have their concentration levels reduced
with a lapse of time. The condition remained relatively
stable for 4 h with uranium concentration at approxi-
mately 5 wt% and approximately 0.01 wt% for iron.
The observation result of bath after experiment
showed a brown layer on the electrolyte’s surface. This
could be due to the dissolution of the iron cathode
caused by its reaction with fluorine in the electrolyte.
The electrolyte was also taken out for observation of the
inner crucible. It showed no corrosion which meant that
the high purity graphite crucible was effectively cor-
rosion-resistant.
Fig. 3. Change of uranium and iron concentration in electrolyte (Run-1).

I. Amamoto et al. / Journal of Physics and Chemistry of Solids 66 (2005) 602–607
605
C
possesses Na which is in common with NaF, and its
melting point (1073.4 K) is lower than that of NaF. The
lowest melting point of NaF–NaCl system is 945 K with
34 mol% NaF and 66 mol% NaCl [9]. As UF in this system
6
would form involatile compound which was found during
the melting test of NaF–NaCl, UF reduction by H as pre-
6
2
treatment was not necessary for the recovery of uranium
from NaF–NaCl system.
4.2. Separation and recovery experiment with high uranium
concentration range (Run-4)
The mixed salt which contained 9.47 wt% of uranium
was used as an electrolyte in order to confirm the
possibility of the separation and recovery process of
uranium using the molten salt technique. After placing
the mixed salt in a high purity graphite crucible, a eutectic
was formed at 973 K.
Fig. 4. Change of uranium concentration in electrolyte using high purity
graphite anode.
This experiment was carried out under the condition
shown in Table 2 (Run-4). Analysis showed the uranium
concentration reduced to 6.55 wt% after 4 h of the
experimental operation (refer to Table 2 and Fig. 6),
which meant that uranium recovery was feasible in a high
uranium concentration range in the eutectic, and its current
efficiency was 61.8%. Meanwhile, the thickness of the
alumina sheath for the anode became thinner and the
dendrite with eutectic was observed on the cathode.
4.3. Recovery experiment with medium uranium
concentration range (RUN-5)
Fig. 5. Change of uranium concentration in electrolyte by high electrolytic
current (Run-3).
The mixed salt which contained 1.09 wt% of uranium
was used as an electrolyte in order to confirm the
possibility of uranium separation and recovery from the
spent NaF under a medium uranium concentration range.
the recovery efficiency drop by the re-dissolution of the
deposit on cathode. Herein, as the vitreous carbon anode
worn off vigorously, a high purity graphite crucible was
made to act as the anode in order to create a stable condition
for the high current density. The uranium concentration in
the bath could not be reduced despite the application of
higher electrolytic current as shown in Fig. 5. Extensive
corrosion, however, was not obvious for the inner crucible
after experiment. The above-mentioned results showed that
high electrolytic current hardly contributed to the deconta-
mination of NaF, instead it worsened the anode’s condition.
However, the uranium concentration in deposit did increase
to approximately 25 wt% as expected.
OD
L
A cathode (10 mm !30 mm ) was applied to establish
0
2
.05 A/cm as a lower electrolytic current density. From
Fig. 7, the uranium concentration reduced to 0.007 wt%
after 5 h of the experimental operation, and its current
efficiency was 15.2%. There was no change for the
thickness of the anode and precipitates were observed on
the cathode.
4. Uranium separation and recovery experiments
from eutectic
4
.1. Eutectic point in NaF–NaCl system
The melting point of NaF (1269 K [8]) used as an
electrolyte can be reduced by the use of some eutectic.
Sodium chloride (NaCl) was selected because NaCl
Fig. 6. Result of uranium separation under high concentration of uranium
(Run-4).

606
I. Amamoto et al. / Journal of Physics and Chemistry of Solids 66 (2005) 602–607
conversion facility and obtained results in this experiments
are summarised as below:
(
a) From the results of X-ray diffraction analysis, NaUF6,
Na UF , and Na UF were identified as some of the
3
6
2
8
uranium compounds adsorbed on NaF;
(
b) Volatilitization of UF6 from spent NaF could be
effectively prevented by the reduction of UF to UF
using the reducing agent, H in this experiment;
6
4
2
(c) The spent NaF, after treatment by H reduction, can
2
have the uranium removed by electrolysis. There was,
however, difficulty in attaining the hypothetical clear-
ance level caused by the disproportionate reaction, i.e.
Fig. 7. Result of uranium separation under medium concentration of
uranium (Run-5).
3
C
U
re-dissolution into the electrolyte.
(
d) With the use of mixed salt, i.e. NaF–NaCl system as the
electrolyte, uranium in the bath would likely form some
involatile compounds at the experimental temperature
(973 K) without the need of reduction treatment. A
sufficient recovery rate of uranium could be obtained
using the NaF–NaCl eutectic due to the absence of the
disproportionate reaction of uranium deposit at a
relatively lower temperature. However, the current
efficiency decreased markedly caused by the electro-
lysis of NaCl after long-time operation. From the
viewpoint of waste reduction, not all the electric energy
are lost as some are used for the decomposition of NaCl.
The economical realization of the molten salt technique
should be further pursued.
Fig. 8. Result of uranium separation under high concentration of uranium
(Run-6).
(
e) After the removal of uranium from NaF, we should look
into the possible recycling of NaF into a new resource as
well as its safe disposal. The stable solidification of
4.4. Separation and recovery experiment with low uranium
concentration range (RUN-6)
fluorides i.e. Na AlF , etc. or oxidation (NaF0Na O)
3
6
2
is being considered as one of the disposal methods.
Conducted in succession Run-5, Run-6 were carried
out using the uranium concentration attained in Run 5,
i.e. 0.02 wt% for confirmation of the recovery possibility
of uranium from the spent NaF below a hypothetical
clearance level (0.001 wt%: this value is equal to the
clearance level of 0.3 Bq/g of uranium concentration by
IAEA Recommendation [10]), Fig. 8 shows the result of
the change of uranium concentration to the quantity of
electricity. Below the clearance level, could be achieved
after 8 h of experimental operation. Its decontamination
factor (DF) reached above 1000 in view of 1.09 wt% as
its initial concentration. However, its current efficiency of
Acknowledgements
The authors wish to thank Mr T. Hoshino (School of
Engineering, University of Tokyo, Tokyo, Japan) for his
useful and helpful advice on the X-ray diffraction
analysis.
References
0.3% was very low. This electric loss is mainly due to
the electrolysis of NaCl.
[
[
1] I. Amamoto, T. Terai, et al., Behaviour of impurities in recycled
uranium at uranium conversion process, J.Nucl. Sci. tchnol., Suppl. 3
(
2002).
2] Y. Yaita, I. Amamoto, et al., Wet process using hydrochloric acid for
treatment of uranium waste, Proceedings of 2003 Fall Meeting of the
Atomic Energy Society of Japan, Shizuoka, Japan, Sep 24–26, 2003,
pp. 548 [in Japanese].
5
. Conclusion
Fundamental experiment on the molten salt technique for
[
3] M. Takai, I. Amamoto, et al., Study on fluorination processing method
for spent uranium absorbent, Proceedings of 2001 Fall Meeting of the
Atomic Energy Society of Japan, Sapporo, Japan, Sep 9–21, 2001, pp.
772 [in Japanese].
uranium separation and recovery from the spent NaF was
carried out to investigate its possibility for the decrease in
the amount of radioactive waste from the uranium

I. Amamoto et al. / Journal of Physics and Chemistry of Solids 66 (2005) 602–607
607
[
[
[
4] R. fujita, I. Amamoto, et al., Study on U recovery process for chemical
trap materials by pyroprocess, Proceedings of 2000 Fall Meeting of
the Atomic Energy Society of Japan, Aomori, Japan, Sep 15–17, 2001,
pp. 593 [in Japanese].
[7] J.J. Katz, E. Rabinowitch, The Chemistry of Uranium, Dover, New
York, 1951, pp. 377–379.
[8] O. Knacke, O. Kubaschewski, et al., Thermochemical Properties
of Inorganic Substances, second ed., Springer, Berlin, 1991,
pp. 1319.
5] R. Fujita, I. Amamoto, et al., Study on U recovery process for
chemical trap materials by pyroprocess, Proceedings of 2001 Fall
Meeting of the Atomic Energy Society of Japan, Sapporo, Japan, Sep.
[9] L. Roth, et al., Gmelins Handbuch der Anorganischhen Chemie
0
Natrium Lieferung 7 , Verlag Chemie, Winheim, 1973, pp. 224.
9–21, 2001, pp. 774 [in Japanese].
[10] IAEA, Clearance Levels for Radionuclides in Solid Materials:
Application of Exemption Principles, Interim Report for Comment,
IAEA TECDOC- 855, Vienna, 1996.
6] M.H. Rand, O. Kubaschewski, The Thermochemical Properties of
Uranium Compounds, Oliver & Boyd, London, 1963, pp. 13–15.