Chemistry Reference
In-Depth Information
(hexyl) compared with shorter ones (butyl) in the Am(III)-trisolvates extracted by
trialkyl phosphates.
Diamylamyl phosphonate (DAAP) has been evaluated as a substitute for TBP
to extract U(VI), Th(IV), Pu(IV), and Am(III) in
n
-dodecane (
89
). Some impor-
tant physical properties of this extractant (density, viscosity, surface tension, and
phase disengagement time) that should fulfill specific requirements for industrial
applications were measured and compared to those of TBP. However, although
higher than in the case of TBP,
D
Am
only reaches 0.16 at 303 K for [HNO
3
] = 1.5
M and [DAAP] = 1.1 M in
n
-dodecane. It remains poor compared with other tet-
ravalent and hexavalent actinides. The trend shows a steep rise of
D
Am
values with
[HNO
3
] from 0.5 to 1.5 M, followed by a gradual fall. It is, however, assumed
from this study that higher concentrations of DAAP, or even neat DAAP, could
lead to appreciable
D
Am
values, enabling DAAP application in trivalent actinide
extraction.
3.3.1.1.2 Trialkyl-phosphine Oxides: The TRPO Process
The TRPO process was developed at the Institute of Nuclear Energy Technology
(INET, Tsinghua University, Beijing, China) in the 1980s to recover the transura-
nium elements (TRU) from highly active defense waste (
90
). It uses a mixture of
trialkyl phosphine oxides (TRPO, with C
6
< R < C
8
or CYANEX 923 dissolved
at 30% in kerosene), which can extract An(III) and Ln(III) if the feed acidity is
decreased to less than 1 M HNO
3
. An(III) and Ln(III) are then selectively stripped
from the loaded solvent (containing other extracted actinides) in 5.5 M HNO
3
. Np
and Pu are recovered in 0.6 M oxalic acid and U in a 5% sodium carbonate solu-
tion. However, the presence of Fe(III) causes third-phase formation, which can be
avoided either by adding TBP or
n
-octanol to the TRPO solvent, or by diluting
the feed (
91
).
A 72-hour inactive test was performed to validate the whole process scheme in
five pulse columns and 24 stages of centrifugal contactors at the INET (
92
). Nd and
Zr simulated Am and Pu, respectively. DFs (expressed as the ratios: [M]
feed
φ
feed
/[M]
raff
φ
raff
, where φ represents the flow-rate of solute M) greater than 3000, 500, and
1000 were obtained for U, Nd, and Zr, respectively, thus demonstrating the feasibil-
ity and reliability of the TRPO process scheme.
Previous active countercurrent tests performed in compact contactors (10 mm in
diameter, 4−5 mL holdup per stage) on highly saline waste (~380 g/L of Na, Al, Fe,
Cr, and Ni) had demonstrated that 99.9% of the actinides could be recovered with the
rare earth elements (REEs) in six stages (
93
).
A new concept integrating both the PUREX and the TRPO processes is proposed
by the INET researchers. This simplified PUREX-TRPO process uses a binary mix-
ture of TBP (20%) and TRPO (20%) in kerosene to extract all actinides including
TPEs, which can be back-extracted together with the trivalent lanthanides in a 5.5 M
HNO
3
solution as in the TRPO process (
94
).
As part of a process comparison campaign carried out by the Institute for TPEs
(ITU, Karlsruhe, Germany), a TRPO flowsheet was implemented countercurrently
in 22 miniature centrifugal contactors. The process used six stages for extrac-
tion, two stages for the first scrub with 1 M HNO
3
, four stages for stripping the
