Synthesis of americium trichloride by the reaction of americium nitride with cadmium chloride

Article abstract of DOI:10.1016/j.jallcom.2007.02.011

Americium trichloride was synthesized by the reaction of americium nitride with cadmium chloride at 600-660 K in a dynamic vacuum. The product was hexagonal AmCl₃, of which lattice parameters were determined to be a₀ = 0.7390 and c₀ = 0.4215 nm. The results indicate that high purity AmCl₃ samples, in which the oxychloride was not found, were prepared without the use of corrosive reagents. The reaction of the nitrides with cadmium chloride is suitable for synthesis of high purity actinide and lanthanide chlorides.

Full text of DOI:10.1016/j.jallcom.2007.02.011

Journal of Alloys and Compounds 456 (2008) 243–246
Synthesis of americium trichloride by the reaction of
americium nitride with cadmium chloride
∗
Hirokazu Hayashi , Masahide Takano, Mitsuo Akabori, Kazuo Minato
Nuclear Science and Engineering Directorate, Japan Atomic Energy Agency, 2-4 Shirakata-shirane,
Tokai-mura, Naka-gun, Ibaraki-ken 319-1195, Japan
Received 22 September 2006; accepted 2 February 2007
Available online 8 February 2007
Abstract
Americium trichloride was synthesized by the reaction of americium nitride with cadmium chloride at 600–660 K in a dynamic vacuum. The
product was hexagonal AmCl , of which lattice parameters were determined to be a = 0.7390 and c = 0.4215 nm. The results indicate that high
purity AmCl samples, in which the oxychloride was not found, were prepared without the use of corrosive reagents. The reaction of the nitrides
with cadmium chloride is suitable for synthesis of high purity actinide and lanthanide chlorides.
2007 Elsevier B.V. All rights reserved.
3
0
0
3
©
Keywords: Actinides alloys and compounds; Chemical synthesis; Solid-state reactions; X-ray diffraction
1
. Introduction
of transuranium chlorides are limited because of difficulty of
handling the samples with high radioactivity and instability in
the atmosphere. The transuranium chlorides need to be synthe-
sized and handled in the hot facilities (glove boxes or hot cells)
maintained with inert atmosphere.
Americium trichloride, which is one of the transuranium
chlorides, has been synthesized in a small scale by the reaction
of the oxide with CCl4 [1] or HCl [2,3] and by the ammonium
chloride route [4–6]. In the ammonium chloride route, the anhy-
drous chloride was synthesized by the thermal decomposition of
Actinide chlorides are attractive materials in both basic chem-
istry and nuclear engineering. Some of the behaviors of the
actinide compounds are significantly different from those of the
lanthanide compounds though the elements in these two f-series
show similarity in many aspects. One of the differences is the
range of valence states of the ionic compounds including chlo-
rides. Investigation of the behaviors of the actinide compounds
is important to understand the role of 5f electrons.
Meanwhile, transuranium compounds are of particular inter-
ests in partitioning and transmutation strategy aiming to reduce
the radiotoxicity of high level nuclear wastes containing
transuranium elements. Pyrochemical separation process, one of
the candidates for the partitioning process, consists of electrore-
fining and molten salt/liquid metal reductive extraction using
molten salts (i.e., molten chlorides) as solvents. In the molten
chlorides bathes, a typical chemical form of the actinides is the
chloride. Therefore, thermodynamic data of actinide chlorides
are needed not only for basic chemistry but also for development
of the pyrochemical separation processes. However, the ther-
modynamic data (i.e., heat capacity and electrochemical data)
the complex chloride (NH4)3AmCl , which was synthesized by
6
heating hydrochloric acid solutions containing americium ions
and ammonium ions under HCl gas [5]. In all the literatures
shown above, corrosive reagents such as HCl were used, and the
samples were purified by sublimation to remove impurities such
as oxychlorides; they are easily formed by the reaction of the
chloride with moisture or oxygen in the atmosphere. The use of
corrosive reagents may be a problem because they can corrode
the materials of the hot facilities. A loss of the samples caused by
sublimation could also be a problem when using a large amount
of transuranium samples. A new method for synthesizing high
purity transuranium chlorides without using corrosive reagents
is needed.
On the other hand, to investigate the basic properties of Am in
LiCl–KCl eutectic melt, Am metal dissolved in liquid cadmium
were oxidized by CdCl2 in the molten salt to form the chloride
∗
Corresponding author. Tel.: +81 29 282 6633; fax: +81 29 282 5922.
E-mail address: hayashi.hirokazu55@jaea.go.jp (H. Hayashi).
0
925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2007.02.011

2
44
H. Hayashi et al. / Journal of Alloys and Compounds 456 (2008) 243–246
dissolved in the molten salt [7]. Furthermore, nitrides of some
lanthanides and actinides were oxidized by CdCl2 in LiCl–KCl
eutectic melt to form their chlorides dissolved in the melt [8,9].
Adopting the methods shown above, the mixture of the transura-
nium chloride and alkaline chlorides can be obtained. However,
it is difficult to separate the transuranium chloride from alka-
line chlorides. If the reactions can occur without a molten salt
bath as a solvent, they are adequate to prepare the transuranium
chlorides solely without using corrosive reagents.
In the present study, americium trichloride was synthesized
by the solid-state reaction of americium nitride with cadmium
chloride in several hundreds milligrams scale. This method does
not require the use of corrosive reagents and solvents.
Fig. 1. The evolution of the pressure while heating the mixture of the nitride
and CdCl2 on run #2.
2
. Experimental
AlltheprocedureswerecarriedoutinthemoduleforTRU-HighTemperature
Chemistry (TRU-HITEC) installed in Nuclear Fuel Cycle Safety Engineering
Research Facility (NUCEF), Japan Atomic Energy Agency (JAEA) [10]. The
TRU-HITEC consists of a glove box and three hot cells maintained with a puri-
fied argon gas atmosphere. Typical impurities in the argon gas were H2O < 1 ppm
and O2 < 1 ppm.
was monitored by a pressure gauge. The heating equipment has a water-cooled
cold finger set in the furnace to trap cadmium vapor produced by reaction (2).
Heating was stopped after the monitored pressure became low; it suggested the
termination of the release of nitrogen gas by reaction (2).
Starting material was 241AmO2 powder purchased from the Radiochemical
Each of the obtained samples was ground and mounted on a sample holder
made of platinum for X-ray diffraction analysis. The XRD profiles were obtained
2
41
Centre (UK) 35 years ago. Due to the decay of Am (T1/2 = 432 years) during
3
5 years of storage, the composition of the actinide elements was estimated to
◦
◦
◦
at room temperature typically by 0.02 step scanning between 5 and 80 in 2θ
2
41
237
be 95% of
Am and 5% of
Np. The oxide sample was heated at 923 K
+
using Mo K␣ radiation (Rigaku RINT Ultima ) to determine both the chemical
in an oxygen gas flow (0.5 L/min) for 1 h to remove organic contamination
and moisture adsorbed on the sample and to adjust the oxidation state of the
americium. The heated sample was identified as a single phase of the dioxide
with the fluorite structure by X-ray diffraction analysis.
form and the lattice parameters.
3. Results and discussion
Americium nitride was prepared from the oxide by the carbothermic reduc-
tion method [11]. The oxide sample was ground and mixed with carbon powder
at a molar ratio (C/oxide) of 3.2. The mixture was pressed to pellets; each pellet
had about 100 mg of the oxide. Several pellets were placed in a tungsten cru-
cible and heated at 1423–1823 K for about 8 h in a nitrogen gas flow (2.5 L/min).
The CO gas release was monitored continuously by an infrared spectrometer to
observe the conversion from the oxide to the nitride. The formation of the nitride
is expressed by
3
.1. Synthesis of the nitride from the oxide
The obtained nitride sample was determined as a sin-
gle cubic phase with a0 = 0.49973 ± 0.00001 nm. The lattice
parameter was close to the reported value a0(AmN) = 0.4998 ±
0
.0001 nm [11] and obviously larger than a0(NpN) = 0.48987
1
± 0.00001 nm [12]. The sample was considered to be a solid
AmO2 + N2 = AmN + 2CO ↑ .
(1)
2
solution (Am0 Np )N, of which lattice parameter was
.95
0.05
After the CO gas release subsided, the sample was additionally heated at 1823 K
for more than 8 h in N2 + 4% H2 gas flow (2.5 L/min) to remove residual carbon.
Americium trichloride was synthesized by the reaction of the nitride with
cadmium chloride; the amounts of the reactants used in two runs are shown in
Table 1. This reaction can be expressed by
expected to be 0.4993 nm adopting the Vegard’s law to the
reported values of AmN and NpN [11,12]. The observed lattice
parameter was 0.09% larger than the expected value. The lat-
tice expansion could be attributed to the self-irradiation damage
[13].
3
3
1
AmN + 2 CdCl2 = AmCl3 + Cd ↑ + 2 N2 ↑
(2)
2
3.2. Synthesis of the chloride from the nitride
The nitride was ground and mixed with CdCl2 (APL, 99.99%) powder at a
molar ratio (CdCl2/AmN) of 1.75. The mixture was pressed to pellets; each
pellet had about 100 mg of the nitride. Several pellets were placed in an alumina
crucible and heated up stepwise from room temperature to 723 K in a dynamic
vacuum kept with a turbo-molecular pump. The pressure in the vacuum system
Behaviors of the pressure monitored during heating of the
mixture of the nitride and CdCl2 were essentially same on two
runs. Fig. 1 shows the evolution of the pressure observed on
Table 1
The amounts of the reactants and the product, the yield of the reaction and the lattice parameters of the products (AmCl3) by the reaction of americium nitride with
cadmium chloride
Run#
Reactants
Product (mg)
Yield (%)
Lattice parameters
Nitride (mg)
CdCl2 (mg)
a0 (nm)
c0 (nm)
1
2
176.3
211.9
221.9
279.4
233.8
289.0
97
100
0.7391 ± 0.0002
0.7388 ± 0.0003
0.4214 ± 0.0003
0.4215 ± 0.0004

H. Hayashi et al. / Journal of Alloys and Compounds 456 (2008) 243–246
245
Fig. 3. The XRD pattern of the obtained chloride sample with the reported data
of AmCl3 [15]. The peaks assigned to the Pt sample holder are marked with (×).
Fig. 2. Temperature dependence of the pressure while heating the mixture of
the nitride and CdCl2.
in Table 1. The color of the cold finger became black; cadmium
metal from reaction (2) seems to be trapped. Residual CdCl2
could be vaporized and also trapped on the cold finger while
heating the sample around 700 K, at which vapor pressure of
CdCl2 reaches to 2 Pa [14].
run#2. The behavior of the pressure can be categorized into three
temperature regions. At temperatures lower than 600 K (region
−
4
I), the pressure increased up to about 10 Pa just after the tem-
perature was raised stepwise, and then decreased gradually at
each temperature. At 600 K < T < 660 K (region II), the pressure
reached constant at each temperature and the pressure increased
with temperature. At 660 K < T < 723 K (region III), the pressure
increased just after the temperature was raised stepwise and then
decreased gradually at each temperature.
Temperature dependence of the observed pressure is shown
in Fig. 2. Plotted data are the peak values for the regions I and
III, and the constant values for the region II. The behavior in
the region I can be attributed to the release of the gas adsorbed
on the sample or the apparatus. The pressure of the nitrogen gas
generated by reaction (2) would be very low in the region I. The
pressure in the region II increased as the temperature increased.
This result suggests that reaction (2) proceeds in the region II.
The temperature dependence of the pressure is expressed as
P = P0 exp(−Ea/RT), where Ea is the apparent activation energy
of the reaction; Ea was calculated to be 191 kJ/mol with the
results of run #2. The pressure decrease in the region III could
be due to the termination of reaction (2).
Fig. 3 shows both the XRD pattern of the obtained chloride
and the reported XRD peaks of AmCl3 [15]. All the observed
peaks except those assigned to the Pt sample holder were
assigned to the hexagonal trichloride. No peaks which can be
assigned to the oxides or the oxychlorides were found. Table 1
shows the lattice parameters of the hexagonal americium trichlo-
ride refined by the least square fitting with the software CellCalc
[16]. Table 2 compares the averaged lattice parameters for two
samples obtained in this study with the reported ones of AmCl3
[1–5,17] and NpCl3 [17–19]. The lattice parameters obtained in
this study are close to those reported for AmCl3 prepared by
other methods, and obviously smaller than those of NpCl3. The
sample was considered to be (Am0.95Np0.05)Cl3, of which lattice
parameters are expected to be a0 = 0.7384 and c0 = 0.4217 nm
adopting the Vegard’s law to the reported values of AmCl3 [3,15]
and NpCl3 [17,18]. The observed lattice parameters agree with
those expected for (Am0.95Np0.05)Cl3 within the experimental
uncertainties.
The results show that high purity transuranium chloride sam-
ples can be prepared without using corrosive reagents by the
method described in this paper. In addition, this method does
Dark yellow fine powder was obtained by the reaction. The
amounts of the product and the yield of the reaction are shown
Table 2
The averaged lattice parameters of two samples in this study and the reported lattice parameters of AmCl3 and NpCl3
Sample
Scale
Method of syntheses
a0 (nm)
c0 (nm)
Reference
241
241
241
243
243
243
237
AmCl3a
AmCl3
AmCl3
AmCl3
AmCl3
AmCl3
NpCl3
200–300 mg
∼␮g
Nitride + CdCl2
0.7390
0.4215
This work (average)
[1,17]
[2]
[3]
[4]
b
AmO2 + CCl4
0.7385 ± 0.001
0.7384 ± 0.0004
0.7382 ± 0.0001
0.7390 ± 0.0003
0.73880 ± 0.00005
0.7420 ± 0.001
0.4248 ± 0.001
0.4225 ± 0.0004
0.4214 ± 0.0001
0.4234 ± 0.0002
0.42235 ± 0.00005
0.42815 ± 0.0005
b
20–30 mg
–
AmO2 + HCl
AmO2 + HCl
b
b
b
multi-mg
∼20 mg
∼␮g
Ammonium chloride route
Ammonium chloride route
[5]
[17–19]
a
The composition of the sample is estimated to be (241Am0.95237Np0.05)Cl3.
The samples were purified by sublimation after the syntheses.
b

2
46
H. Hayashi et al. / Journal of Alloys and Compounds 456 (2008) 243–246
not require additional purification by sublimation because the
obtained sample did not contain impurities such as the oxy-
chlorides and the oxides. It could be due to the fact that the
reactants, the nitride and CdCl2, do not contain oxygen atoms
in principle. The reagents containing oxygen can easily react
with chlorides to form stable by-products, the oxychlorides and
the oxides. Therefore, this method is suitable for synthesizing
not only transuranium chlorides but also the chlorides of other
actinides and lanthanides.
References
[
[
[
[
1] S. Fried, J. Am. Chem. Soc. 73 (1951) 416.
2] J. Fuger, J. Inorg. Nucl. Chem. 28 (1966) 3066.
3] J.H. Burns, J.R. Peterson, Acta Cryst. B26 (1970) 1885.
4] L.B. Asprey, T.K. Keenan, F.H. Kruse, Inorg. Chem. 4 (1965) 985.
[5] T. Schleid, L.R. Morss, G. Meyer, J. Less-Common Met. 127 (1987)
183.
[
6] G. Meyer, in: G. Meyer, L.R. Morss (Eds.), Synthesis of Lanthanide
and Actinide Compounds, Kluwer Academic, Dordrecht/Boston/London,
1991, pp. 135–144.
[
[
7] J. Serp, R.J.M. Konings, R. Malmbeck, J. Rebizant, C. Scheppler, J.-P.
Glatz, Electrochem. Acta 51 (2006) 4024.
8] F. Kobayashi, T. Ogawa, M. Takano, M. Akabori, A. Itoh, K. Minato, S.
Takahashi, Proceedings of the International Conference on Future Nuclear
Systems (Global’99), Jackson Hole, USA, August 29–September 3, 1999,
CD-ROM.
4
. Conclusion
Americium trichloride has been successfully synthesized
by the reaction of americium nitride with cadmium chloride
at 600–660 K in a dynamic vacuum. The product is hexag-
onal AmCl3; its averaged lattice parameters for two samples
prepared with a nearly identical manner are a0 = 0.7390 and
c0 = 0.4215 nm. The results indicate that high purity AmCl3
samples, in which the oxychloride was not found, were pre-
pared without the use of corrosive reagents and purification by
sublimation.
[9] H. Hayashi, F. Kobayashi, T. Ogawa, K. Minato, J. Nucl. Sci. Tech. Sup. 3
(2002) 624.
[10] K. Minato, M. Akabori, T. Tsuboi, S. Kurobane, H. Hayashi, M. Takano, H.
Otobe, M. Misumi, T. Sakamoto, I. Kato, T. Hida, JAERI-Tech 2005-059,
2005, in Japanese.
[
11] M. Takano, A. Itoh, M. Akabori, T. Ogawa, S. Kikkawa, H. Okamoto,
Proceedings of the International Conference on Future Nuclear Systems
(Global’99), Jackson Hole, USA, August 29–September 3, 1999, CD-
ROM.
[
12] OECD Nuclear Energy Agency (Ed.), Chemical Thermodynamics of
Neptunium and Plutonium (Chemical Thermodynamics vol. 4), Elsevier,
Amsterdam, 2001, pp.187.
Acknowledgements
Part of the present study was carried out within the col-
laborative research program of TRU behavior in pyrochemical
processes with Tohoku Electric Power Company, Tokyo Electric
Power Company and The Japan Atomic Power Company. The
authors thank Dr. T. Kimura, Mr. M. Nakada, and Mr. H. Otobe
for providing the americium oxide sample. The authors thank
Messrs T. Tsuboi, M. Misumi, M. Kawasaki, and H. Kato for
technical supports.
[
13] E.V. Shimbarev, L.V. Sudakov, Radiochimiya 26 (5) (1984) 643.
[14] I. Barin, F. Sauert, E.S. Schultze-Rhorhof, W.S. Sheng, Thermochemical
Data of Pure Substances, VCH, Weinheim, Germany, 1989.
[
[
[
15] JCPDS 71-2102, Calculated from Ref. [3].
16] H. Miura, J. Cryst. Soc. Jpn. 45 (2003) 145.
17] W.H. Zachariasen, Acta Cryst. 1 (1948) 265;
W.H. Zachariasen, J. Chem. Phys. 16 (1948) 254.
[18] JCPDS 75-1887, Calculated from Ref. [17].
19] S. Fried, N. Davidson, J. Am. Chem. Soc. 70 (1948) 3539.
[