2
-
1
.
....
.
y
:
!
!
'
-
5
twicininizinin
US
#t
TOF |
ORNL P
2731
.
•
.
..
1
20
!
.
:
.
:
.
:
'
og
6
-
.
-
.
.
.
.
-
.
-
-
.
.
.
.
.
.
inn
1:45
!!! So
56
101 3 6
+
VO
S
.
11.25 |1.4 1.6
.
MICROCOPY RESOLUTION TIST CHART
NATIONAL BUREAU OF STANDARDS - 1963
ORNU-P-2031.000
DEC 6 1966
. CONF-660907-4
A GAS SORPTION-DESORPTION RECOVERY METHOD FOR CESTI PRICES
PLUTONIUM HEXAFLUORIDE*
Sidney Katz and G. I. Cathers 10.732.00; 12.65
MASTER

RELEASED FOR ANNOUNCEMENT
IR BUCLEAR SCIENCE ABSTRACTS
LEGAL NOTICE
This report was prepared as an accourt of Government sponsored work. Ncither the United
States, nor the Commission, nor any person acting on behalf of the Commission:
A. Makes any warranty or representation, expressed or implied, with respect to the accu-
racy, completeness, or usefulness of the information contained in this reports or that the u89
of any information, apprratis, method, or process disclosed in this report may not infringe
privately owned rigata; or
B. Assumes any liabilities with respect to the use of, or for damages resulting from the
use of any information, apparatus, method, or process disclosed in this report.
As used in the above, person acting on beliall of the Commission" Includes any am-
ployee or contractor of the Commission, or employce of such contractor, to the extent that
such employee or contractor of the Commission, or employee of such contractor prepares,
disseminates, or provides acce88 to, any information pursuant to his employmeni ui contract
with the Commission, or his employment with ouch controctor.
I'. Warto
*To be published in Advances in Chemistry Series, which is a compilation of papers
given at the 152nd National Meeting, American Chemical Society, New York
City, New York, September 11-16, 1966.
Research sponsored by the U.S. Atomic Energy Commission under contract with the
Union Carbide Corporation.
Fluoride-volatility processing, a potentiolly important method of treating nuclear
reactor fuel, is based on the high volatility of the hexafluorides of uranium and plutonium.
in processing low-enrichment UO, power-reactor fuel, it is desirable to recover
PUF, and UF, separately, with both adequately decontaminated from fission products.
We will outline the develpment of a sorption process for recovering Puf, that may be
applicable to fluoride-volatility processing. In particular, the interaction between Puf
and LiF will be described as a selective and reversible method of collecting Puf,
posibly decontaminated from highly radioactive fission products. This Lif method
appears to be an analog, in some respects, of the Nof method for recovering UF , and
it is expected tha: it will be as effective as the NaF system in achieving fission product
decontamination. However, as yet no work on decontamination has been done.
The scope of our efforts is given in Table 1. It must be emphasized that our
results represent preliminary development work; nevertheless they appear to have
sufficient validity to justify optimism about the eventual usefulness of a sorption-
desorption process for PuF
First, the choice of Lif as the optimum sorbent will be justified. Next, some
possible process flowsheets will be presented, one of which illustrates the separation
of Puf, from UF, Also, the chemical basis of what, for purposes of simplicity, is
referred to as sorption-desorption will be discussed to a limited extent.
We chose LiF as the optimum sorbent simply because Puf, can be completely
regenerated (or desorbed) from this material more easily thar. from other fluorides
that act as effective sorbents, for example, the fluorides of group IA and group IIA
elements. Two types of fluoride-volatility process flowsheets showing PuFc recovery
are presented to illustrate the end-purpose to which a LiF sorption step might be
applied (Fig. 1). In one, the UF, and Puf, are evolved consecutively through the
use, first, of interhalogen fluoride and, then, of fluorine; in the other, the UJF, and
Puf, are volatilized toge her in the primary step by using elemental fluorine. We
believe that the fuF, olone, or the UF, and PUF, together, will be held up and
separated from excess fluorine in a cold trap before being introduced into the LiF
sorption-desorption step. Since Pufe does, but UF, does not, form a complex with
LiF, a separation of Puf, und UF, is obtained. In either flowsheet, it is expected
that Puf, will be sufficiently decontaminated (from fission products) in the LiF step
before being reconverted to Puo, for fuel refabrication.
We arrived at our choice of LiF as a sorbent in the following manner. First,
sinall samples of 31 metal fluorides were statically exposed to an excess of Puf, in
fluorine at 250°C, with the system temperature slowly being lowered to 100°C and
with continuous flushing of the system with fluorine. The results in Table 2 show
the relative retention of plutonium; these data indicate that only the fluorides of
the IA and IIA metals retain sufficient plutonium to be of interest as sorbents.
However, the data on the remaining fluorides are useful since some of these com-
pounds will be encountered in process equipment as corrosion products from
materials of construction, fission products, and rearjent impurities.
In the second stage of screening sorbents, ten group IA und group IIA fluorides
were tested with an apparatus of the type shown in Fig. 2. The test sorbent at the
desired temperature was placed in this reactor in series with an NaF backup trap
at 200 to 500°C to collect Puf, not retained on the test sorbent. The Pufo in a
stream of fluorine was produced in the generator. (Variations of this apparatus were
used in other tests that will be discussed later; for desorption tests, only fluoride flowed
through the train; for some soiption tests, larger reactors or several reactors in series
were substituted for the small test sorbent reactor.) About 1 to 10 mg of Puf, was passed
into the train for 20 min for each of the temperatures shown in Table 3. The most
significant conclusion from these data is that there is no sorption-desorption system
with PUF, similar to that known for UF, and NaF. If such a system did exist, there
would be a temperature above which little or no retention of PeF, would be expected.
Another conclusion is that Puf, is so strongly sorbed by some of the fluorides that its
desorption probably would not be possible under any conditions.
With a few of the fluorides, however, there was some evidence of a weak inter-
action. Additional tests were made with NaF and Caf, since they had potential process
value. Potassium fluoride was not evaluated further because of its hygroscopicity.
Work with MgF, is planned since it is a possible alternative to LiF. In the case of
LiF, we now know that at temperatures below 300°C the reaction rate was too small
for complete retention and that above 300°C partial desorption was occurring.
In the third stage of testing to select a sorbent, desorption tests of longer duration
wera undertaken (Table 4). First, NoF and Caf, were evaluated at temperatures from
100 to 600°C, with some tests lasting 10 hr; essentially no desorption was noted from
NaF under the most strenuous conditions, and only at 600°C was 7 to 15% of the
sorbed PUF, desorbed in 10 hr from Cafz. However, with LiF, significant desorption
was noted at temperatures from 350 to 600°C.
The results of the desorption of Puf, froni LiF at different temperatures are given in
Table 4. Most of these tests were made with about 40 mg of plutonium sorbed on 2.0 g bed.
Fluorine flow rates of about 30 ml/min were used in both the sorption and desorption
cycles. It is apparent that ihe rate of desorption increased rapidly from 400 to 500°C.
Beginning at 550°C, there may be some decrease in the rate of desorption, possibly the
result of microsintering effects. However, recent work denies this effect. In ten-
hour Lif desorpii on tests, it was possible to recover about 98% of the plutonium.
Passage of additional fluorine recovers additional plutonium. Large volumes of fluorine
would be available in a processing plant where fluorine is recycled.
The physical form of the fluoride material that we used in most of the tests is
important. The UF, serption-clesorption process was developed by using semi-activated
NaF, which was generated by the decomposition of NaHF.. In cur early tests with
Puf, finely divided powders were used, which resulted in channeling in the dynamic
tests or sorption predominantly at the bed surface in static tests. In the later work,
only semiactivated forms of LiF, Cafq, or NaF were used. The LiF has generally becn
prepared by fusing Li,coz (mp 723°C), grinding and sizing to 20 to 40 mesh, and
then carefully fluorinating at 400°C. The resul:ing material has generally been quite
similar to semiactivated NaF, with surface areas of 1.5 to 3 m g and a high-porosity
fraction.
Data from an extended series of tests with 1.7-g beds indicated that 300°C was
the optimum temperature for the sorption of Puf, on LiF. However, it is now rec-
ognized that these results were obtained partially because of decomposition of
PuF to PuFd, which was attributed to a low F/Pu mole ratio in the gas coming from the
generator. Since the generator was operated at 600°C, the Fy/Pu ratio may have been
24
25
5
as low as 67. The required F/Pu ratio at 300°C is 396; nevertheless, since the
F/Pu ratio for the LiF--PuF:--F, equilibrium is believed to be about 3000, these
sorption results are not totally without significance.
Figure 3 shows the sorption effect at 300°C and also indicates that a bed loading
of about 10 wt% is possible. After an extended number of additions of Puf, the loss of
plutonium fron the bed during each addition seemed to level out at about 4 mg. Ap-
proximately this amount of loss (in absolute weight) is usually encountered in 1.7-g beds
with a single 30 to 40 ng loading. Why there was practically no loss on the first two
increinental additions is not understood. Perhaps a more-stable complex is involved.
The probable chemical mechanism involved in PuF, sorption-desorption is shown
in Tuble 5. The first equation illustrates that sorption is actually a reduction reaction
with formation of a quadrivalent-plutonium complex. The equation, as given, can be
cornpared with the Puf --F, --Puf, equilibrium, but the equilibrium is approximately
eight times more favorable to quadrivalent plutonium when LiF is present. Re-
volatilization of Puf, is accomplished by reflucrination (and not by dissociation
of a PuF,--MF complex as in the UF,--NaF sorption-desorption process).
Our limited information on the plutonium sorption complex is reviewed in Table 6.
The absence of iodine evolution when a solution of (LiF), PUF, was titrated with KI
solution, together with X-ray diffraction studies on the material (a pink solid), indicate
that it is isomorphous with 4LiF.UF2. The valence of the plutonium in the sorption
complexes with NaF and Caf, depends on the sorption temperature, with the penta-
valeni state probable below 400°C. That this was not a case of Puf, adduct formation
or surface adsorption was supported by the identification of Puot in infrared absorption
studies of solutions of the complexes.
Actual separation of plutonium and uranium is demonstrated by the results presented
in Table 7. Here, Puf, and UF , mixed together, were fed in a stream of fluorine into
the separation train of two small 1.7-g LiF sorbent beds at 300°C end an NaF backup
bed at 150°C. These data show that no detectable uranium was associated with the
plutonium on the Lif sorbent and that only the small amount of plutonium expected to
leak through the Lif was found with the uranium on the Naf bed.
In summary, we have done the following: (1) presented results indicating the possible
usefulness of activated LiF as a means of recovering Puf, in the fluoride-volatility
processing of nuclear reactor fuel; (2) suggested two typical flowsheets, including one
in which LiF is used to achieve separation of Puf, from UF (3) presented sorption
and desorption results; (4) suggested that an oxidation-reduction equilibrium between
the solid (LiF), PUF4 complex and Puf 6-F2 gas is the probable mechanism involved;
and (5) presented chemical evidence to support the suggestion in (4).
In future work we will consider measuring the kinetics of both the sorption and
desorption steps, in addition to evaluating the actual equilibrium or equilibria
involved. This will recessarily be accompanied by developnient work directed toward
practical application of the process.
Table 1. PUF, Recovery by Sorption-Desorption
1. Lif as the optimum sorbent.
2. Process flowsheets.
3. Sorption and desorption results.
4. Characterization of sorbed plutonium; probable sorption mechanism.
5. Separation tests for UF, and Pufc.
.
.
. . .
-
w
.
.
...
.
.
.
.
.
.
-
..
.
..
...
..
.
...
.
.
..
.
1:..
,
W
T
..
Table 2. Sorption of Puf, on 31 Metal Fluorides
Pu/metal fluoride weight ratio.
Li 0.8
Na 1.4
K 2.2
Rb 0.6
Cs 0.6
Be
Mg
Ca
Sr
Ba
1
0.03
0.3
0.5
0.4
Ag 0.02
AI 0.02
cd 0.002
Ce 0.02
Co 0.001
Cr 0.01
Cu 0.005
Fe 0.001
HF 0.015
in 0.0?
La 0.02
Mn 0.01
Ni 0.005
Pb 0.005
T
Sc 0.01
Sn 0.005
Th 0.001
! 0.002
Y 0.05
Zn 0.005
Zr 0.01
Table 3. Retertion of Puf, on Groups IA and IIA Metal Fluorides.
Sorbent
100
Percentage of Plutonium Retained at Temperature (°C)
200 300 400 500 600
600
700
LiF
87.5
88.7
96.9
89.7
29.4
98.6
NaF
46.8
99.8
98.5
99.6
99.92
99.94
99,97
KF
97.1
97.6
94.4
RbF
90.4
98.2
98.0
98.5
95.8
87.1
83.6
90.4
97.5
99.9
97.9
70.9
93.8
94.2
99.9
99.8
CsF
99.6
98.6
BeF.
29.1
20.1
76.6
71.0
73.4
MgF.
98.8
98.2
96.3
96.9
98.6
98.6
94.5
CaF2
99.98
Sifa
99.99
99.95
99.4
99.995
99.97
99,95
99.997
98.9
99.99
99,7
99.95
99.97
99.96
99.93
99.996
99.97
99.99
Batz
99.97
L
L
L.
.
..
.
Table 4. Puf, Desorption Tests
30 ml/min F, flow through a 2-g bed containing 40 mg of Pu.
Results
Weight Desorbed
(mg)
Time
Temperature
(°C)
Sorbent
(hr)
600
<600
0.12
<0.12
600
3.0
600
5.2
500
1.0
<500
<1.0
400
7.4
450
10.6
475
17.5
500
22
500° tesi continued
500° test continued
20
12
3
2
2 nd hour
3 rd hour
4 th hour
5 th hour
500° test continued
500° test continued
Residue of 1.3 mg
Table 5. Mechanism of Lif Sorption-Desorption Process
PuFG + 4LIF (LiF)APUFA+F2
K300°c ~3000
Put → PUF 4 + F2
K3000c = 396
Not comparable to
UF G + 2NaF
(NaF),UFG
..
www.com
ara
C
.
.
.
.
.
Table 6. Characterization of Plutonium Sorption Complex.
Temperature
(°C)
Oxidation Number
NaF
Lif
100
4.0
4.7
4.7
400
4.0
4.2
4.2
600
4,0
4.0
4.0
Probable complex
(600°C)
(LiF), PUF4
(NaF),PuF4
Table 7. Separation of UF, and PuFc
Conditions: 30 ml/min F2 flow through approximately 2-g beds of sorbent for 30
min. Two Lif beds at 300°C in series with an Naf bed at 150°C.
Bed Analysis
Pu (mg)
U (mg)
Test 1
LiF - 1
66 (U DF >240.)
3.0
LiF - 2
<0.02
<0.02
4.9 (Pu DF 240.)
NaF
0.28
Test 2
LiF-1
67 (U DF >820.)
--
LiF - 2
4.7
<0.02
<0.02
17.6 (Pu DF 176.)
--
-
-
.
NaF
0.41
.
. .
.
.
.
.
.
.
.
. .
.
. .
.
.
.
.
.
- ..
-
-
ORNL-Dwg: 66-9160
Puf go UFG
Eye F.P's
- UFRECOVERY
F-
FLUORIDE
VOLATILITY
PROCESS
COLD
TRAP
Lif
SORPTION
MI DESORPTION
> Puso RECOVERY
F
BED DISCARD
WITH F.P'S
RECYCLE
-
-
AF (1)
FLUORIDE
VOLATILITY
PROCESS
UFG RECOMERY
F2 (2)—
(2) Pufy F2,
COLD
2-64
LiF
SORPTION
DESORPTION
> PuFG RECOVERY
TRAP
BED DISCARD
WITH F.P'S


Fig. I. Process Applications of Lif Sorption-Desorption for Puf, Recovery.
ORNL DWG. 66-8961-R
TO DISPOSAL TRAPS

B
2oC
Fa-o
o
A - NaF backup trap. Effective sorbent for P.JF, fram
100 to 600°C. 2 g of 20- to 40-mesh material.
B - Test sorbent trop. Holds about 2 g of sorbent.
In soine cases two traps were used in series; in
others, an 8-g bed was used.
C - For desorption tests, the fluorine bypassed the
Pufo generator. Normal flow was 30 ml/min.
D - Pufo generator. PuOn added before each test.
Normal F, flow was 30 ml/min.
Fig. 2. Sorption und Desorption Test System.

ORNL-Dwy. 66-9159
5 mg
4 mg
ooo ©
3 mg
Weight Loss of Plutonium from the Bed (mg) During Each Addition.
2 mg
Ž
12
13
14
15
3 4 5 6 7 8 9 10 11
INCREMENTAL 50 mg ADDITIONS OF PUF
A TOTAL OF 760 mg OF PU WAS PASSED INTO THE 7.8-9 BED OF LIF AT 300°C,
WITH 710 mg RETAINED.
Fig. 3. Plutonium Loss Through LiF (Collected on Backup NaF).
- ...
*
4
.
*.
.
.
.
..
.FO
.
-'*
: .
.
.
:
.
.
''.
'.
T
A
.
.
.
.
END



DATE FILMED
1 / 3 / 167







RA'