Selective Precipitation of Rare Earth Chlorides from LiCl-KCl
Michael F. Simpson and Tae-Sic Yoo Idaho National Laboratory
Michael Shaltry and Supathorn Phongikaroon University of Idaho
Daniel LaBrier and Michael Lineberry Idaho State University
2012 International Pyroprocessing Research Conference Fontana, Wisconsin August 26-29, 2012 Actinide Drawdown as Key Component of ER Salt Waste Treatment
Fuel processing objectives – Recover high % of actinides – Minimize high level waste – Minimize processing cost – Support safeguards objectives Actinide drawdown unit operation functions – Remove high % of actinides from salt to prevent their loss into the waste process – Recover actinides in form suitable for recycle to the ER or use in new fuel fabrication – Produce salt suitable for processing in the FP separation unit operation or direct into the waste process Options for Achieving Drawdown
Electrolysis (-) (+) – Results in chlorine gas evolution U3+ + 3e- = U – Difficult to drive high extent of actinide Pu3+ + 3e- = Pu removal due to concentration overpotentials Chemical Reduction – Reduction via metals such as Li and K 3Cl- = 1.5 Cl + 3e- – Selectivity and product recovery viewed 2 as challenges – But scale up potential appears to be excellent Chemical oxide formation
– O2 gas bubbling 1 – Addition of Li O to molten salt UCl3 + O2 ®UOCl +Cl2 2 2 – Selectivity and product recovery viewed 3 as challenges UCl3 +O2 ®UO2 + Cl2 – Also appears to be suitable for scale-up 2
UCl3 + Li2O ®UOCl + 2LiCl 3 3 1 UCl + Li O ® UO + U + 3LiCl 3 2 2 4 2 4 Metal Precipitation from Li Reduction
o UCl3 +3Li ®U +3LiCl, DGrxn = -722 kJ / mole -10
Simple process known to work for Hg Mo reducing UCl3 concentration to prepare -20 W Ni for U/TRU recovery using a liquid H
cadmium cathode Fe -30 More readily scaled up than electrolytic Cd Zn, Cr processes (not constrained by electrode -40 current densities) V Zr Mn, Be
Free energy of formation values appear to )
e -50 n i r
be favorable with respect to being able to o l h
c U
f
separate actinides from rare earths. o
e Np l
o -60 Pu
m Cm, Am Experimental study needed due to r e p
l Y
uncertainty associated with non-ideal a
c Nd k (
Ce
characteristics of the molten salt/metal G -70 Pr, La Δ systems.
Initial experimental studies used only -80 Ca, Na Sr selectivity relative to different rare earths Li K for simplicity and cost effectiveness. Ba, Cs, Rb -90 Experimental Design for Rare Earth Reduction with Lithium
lithium addition Experimental Procedure
1. MgO crucible and salts loaded into inert atmosphere glove box (Ar w/
<10 ppm H2O&O2) stirrer 2. Quaternary salts (LiCl-KCl-XClx-MCly) heated to 500 C for ~12 hrs prior to each test 3. Stirrer rotated at 60 rpm molten Li metal 4. Lithium added to tube inserted in the salt (see diagram) 5. Salt samples taken every 2-4 hrs and analyzed via ICP-MS (stirring stopped during sampling) crucible Objective was to measure selectivity and reaction kinetics and understand process limitations. Rare earths used as surrogates for actinide/rare earth mixtures. Kinetics of Lithium Reduction
LaCl (2.5 mol%)-CsCl (2.5 mol%) Equilibrium was typically 3 reached in under 10 hrs 100% Concentration drift 90%
observed in some n o i
experiments t 80% c
– Attributed to poor glove box u d 70% e
atmosphere in early r
experiments e 60% d i
r Li equiv = 1.0
Complete lithium o
l 50%
h Li equiv = 0.52
consumption evidenced c 40% Li equiv = 0.55
by mole equivalents m u Li equiv = 0.28 reacted divided by moles n 30% a h of lithium added typically t
n 20% in the range of 1.0 to 1.2. a L 10%
0% 0 5 10 15 20 time (hr) Simultaneous Reduction of Different Rare Earth Chlorides
LaCl3 (0.59 mol%)-CeCl3 (0.59 mol%) 0.006
Note- only salt sampled. 0.005 metal precipitate observed but not sampled. 0.004 n o o LaCl DG =-293.8 kcal/mole Cl i 3 t c a
r X(La)
f 0.003 e l X(Ce) o m 0.002 o CeCl3 DG =-288.0 kcal/mole Cl 0.001
0 0 2 4 6 8 10 time (hrs)
Post drawdown salt showing Metal deposit layer Lithium Drawdown Model
Assume lithium completely reacts with the most reactive of the two rare earth chlorides. Extra Li then reacts with the other rare earth chloride. Then equilibrium is assumed to be established by reaction of the metals and metal chlorides such as in the case of La and Ce.
LaCl3 +Ce « CeCl3 + La
A simple equilibrium expression can be derived. x EQ yEQ K ' = Ce La x=mole fraction in salt EQ EQ y=mole fraction in metal xLa yCe
K’ can be expressed in terms of fundamental properties. But the activity coefficients are not known. So, the lumped parameter can be determined via fit to data
DGo g g - K ' = LaCl3 Ce e RT g LagCeCl3 Comparison of Equilibrium Results to Model
Primary Secondary K’ (fit) DG Species Species (kJ/mol Cl)
LaCl3 CsCl 0.9 73.7
CeCl3 LaCl3 0.2 5.8
NdCl3 LaCl3 0.3 9.4
1 0.9 0.8 0.7 0.6 '
K 0.5 0.4 0.3 0.2 0.1 0 0 20 40 60 80 DG (kJ/mole Cl) Selectivity for Reduction via Reaction with Metal
K drawdown experiments run by U.Idaho undergraduates. – Cs-La & La-Nd – Higher T (593 C) – Metal pieces added directly to the salt Based on limited data, it appears that the choice of reducing metal has little effect on selectivity. – Better selectivity of La in La-Cs – Slightly lower selectivity for Nd in La- Nd Reaction with Lithium Oxide
Proposed Implementation UCl + Li O ®UOCl + 2LiCl, DGo = -545 kJ / mole 3 2 rxn Of Li O-Based Drawdown 3 3 1 2 UCl + Li O ® UO + U + 3LiCl, DGo = -461 kJ / mole 3 2 2 4 2 4 rxn Actinide oxychlorides appear to be thermodynamically favored Avoids potential corrosion from bubbling oxygen gas into molten salt Lithium oxide will normally carryover from oxide reduction process unless reduced fuel is subjected to distillation Unknowns – Whether oxides or oxychlorides form – Selectivity of actinide removal versus rare earth removal – Maximum extent of actinide removal
without using excess Li2O Lithium Oxide Drawdown Tests with Rare Earth Chlorides
Same experimental Test Matrix apparatus as used for LiCl-KCl NdCl LaCl Li O Time Mixed? Li drawdown tests Test 3 3 2 Test conditions: (g) (g) (g) (g) (hr) LO-1 30 3.4 3.4 0.14 10 No – T=500 C LO-2 30 3.4 3.4 0.3 8 Yes – 8-10 hrs LO-3 30 3.4 3.4 0.9 8 Yes – Salt stirred except for LO-4a 30 3.4 3.4 0.60 10 No 10 minutes prior to LO-4b 30 3.4 3.4 0.60 10 Yes taking sample – “Mixed” refers to
whether the Li2O was held in a cup or mixed directly into the salt. Lithium Oxide Dissolution
0.025
0.024 x
, n o i 0.023 t c a r
F La 0.022 e l
o Nd
M 0.021
0.02 0 2 4 6 8 10 12 Time, t (hrs)
In LO-1 (results shown above), lithium oxide was loaded into sample cup and dipped into the salt (as shown on the right)
Very little reaction of LaCl3 or NdCl3 was observed. Reaction rate appears to be limited by dissolution of Li2O immersed into LO-4a Li2O in the salt, since its solubility is only about 0.65 mol% in LiCl-KCl at 500 C. Total Li2O immersed in the first experiment was equivalent to 0.89 mol% if completely dissolved in the salt. Reducing Mass Transfer Resistance for Lithium Oxide Reaction
In LO-2, LO-3, and LO-4b the Li2O was Mixed experiments show added directly to the salt and mixed complete Li2O reaction as a suspension. From results below, it can see that 1.4 O
this facilitated the reaction. 2 i L
1.2 v Extent of reaction is evidence of oxide i u q
e 1 formation. e l
o LO-2 (mixed)
m 0.8 /
d LO-4b (mixed)
80% e t
c 0.6
a LO-3 (mixed)
70% e r
v LO-4a (unmixed)
i 0.4 t l 60% u a q LO-1 (unmixed) s e
e 0.2 m 50% l o o r f m
La conv (unmixed)
d 40% 0 e
v La conv (mixed) 0 5 10 15 o 30% m Nd conv (unmixed) time (hr) e r Nd conv (mixed)
% 20% 10% Consistent Stoichiometry 0% 0 5 10 15 3Li2O + 2NdCl3 ® Nd2O3 + 6LiCl time (hr)
3Li2O + 2LaCl3 ® La2O3 + 6LiCl Oxide/Oxychloride Precipitate
Blue precipitate layer Fragment of Salt Showing analyzed via ICP-MS and Precipitate Layer white region high in Li/K found to be concentrated in La and Nd. XRD pattern of blue region more consistent with oxide phases than oxychlorides. All results are consistent with
Nd2O3 and La2O3 formation.
blue region high in RE Selectivity Comparison for Metal and Oxide Precipitation Approaches
100% 90% 80% l 70% a v o 60% Li reduced (data) m e r 50% Li reduced (model) 3 l
C 40% d K reduced (data) N 30% Li2O reacted (data) 20% 10% 0% 0% 20% 40% 60% 80% 100%
LaCl3 removal Conclusions
Reaction of metallic lithium with rare earth chlorides is fast and goes to completion in a very simple chemical reactor configuration. Potassium metal can likewise be used to reduce the rare earth chlorides. Reduction selectivity using lithium metal can be predicted with a model that is consistent with differences in free energy of formation of the rare earth chlorides. Limited lithium oxide solubility requires suspension of the powder in the molten salt to facilitate fast and complete reaction with rare earth chlorides. Nd and La chlorides primarily form oxides from reaction with lithium oxide at 500 C. All methods for removing rare earths have the drawback of low selectivity for specific species, but lithium metal reduction performs the best in this respect. While all of the examined approaches to drawdown are expected to be scaleable and economical, thermodynamic proximity between plutonium and rare earths is expected to require either some loss of plutonium into the waste stream or substantial carryover of rare earths into reduced product which would likely require post-processing to extract the rare earths.