ORNL-2661.txt

MALRTIN MARIETTA ENERGY SYSTEME LIBRARIES

DTN

5k 031350 b

 

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 

 
ORNI~2661

Contract No. W-TLOS-eng-26
CHEMICAL TECHNOLOGY DIVISION

Chemical Dévelapment Section A

THE FUSED SATT-FLUORIDE VOIATILITY PROCESS

FOR RECOVERING URANIUM

¢. T. Cathers, M. R. Bennett, R. L. Jolley

DATE ISSUED

APR 11959

OAK RIDGE NATIONAL LABORATORY
Osk Ridge, Tennessee
Operated by
UNION CARBIDE CORPORATION
for the
U.S. ATOMIC ENERGY COMMISSION

DA

3 445k 0361350 &
ABSTRACT

A fluoride volatility process for recovering enriched uranium
from UF)-NaF-ZrF) melts, produced by the dissolution of reactor
fuel elements in molten NaF-ZrF), through hydrofluorination, is
described. The UF) is fluorinated to UFg, which volstilizes, and
is separated from volatile fission products in a NaF absorption-
desorption cycle. In laboratory studies gross bets and gamms
decontamination factors of 10” with a urenium loss of less than
0.1% were secured. The product UFE may dbe conveniently returned
to UFYy or metal for further use as reactor fuel.

Chemical studies, supplementary to process development work,
showed that impurities, possibly oxides, have a much grester effect
in the fused salt-fluorination step than such factors as use of
nitrogen with the fluorine, or the method of introducing the
fluorine into the fluoride melt. The colloidal behavior of NiFo,
a slightly soluble corrosion product formed in the fluorination,
indicated that this materisl would not interfere in molten salt
handling if thbe fused salt was not allowed to stand without
agitation for prolonged periods. The absorption of UFg in NaF
was found to be due to the formation of a UFg-NaF complex. The
egquilibrium between gasecus UFg and sclid complex was established
for the temperature range of 80 to 320°C. Decomposition of the
UFg~-NalF complex to a UF5wNaF complex dees not lesd to appreciable
uranium loss in the process 1f specified process conditions are
maintained.
CONTENTS
Page
1.0 Immmcmm L
2.0 DESCRIPTION OF PROCESS h
3.0 PROCESS DEVELOPMENT STUDLES 6
3.1 Fused Salt-Fluorination Work 6
3.1.1 Fluorine Efficiency 6
3.1.2 Corrosion Studies 9
3.1.3 Recovery Yield of UFg | 1L
3.1.4 Behavior of NiF, in Corrosion | 15
3.2 NaF Decontemination Step | 19
3.2.1 Operation iu = Temperature Range of 100-400°¢C 19
3.2.2 Operation st a ‘I‘émperature Above 600°C 23
4.0 BASIC CHEMICAL STUDIES | 26
L.1 NaF Cepacity for UFg é6
4.2 Pressure of UFg Above UFg-+3Na¥ Complex 26
4.3 Decomposition of UFge3NaF Complex 29
5.0 RECOMMENDATIONS 32

6.0 REFERENCES | 33
1.0 INTROLUCTION

A new approach to processing of enriched uranium civiliau
power reactor fuels by a flvoride volatility method bas been
reportedfilfg The mathod consists in three steps:; dissolution
of the metal or elloy in a fluoride melt by hydroflueorination,
volatilization of UFg from the molten selt through fluorination,
and fipal purification of the UFg from volatile or eptrained
fission product fluorides by absorption or distillation. A
typical salt composition is 50-50 mole % NeF-ZrF),, with a melting
point of about 510°C. The second step eppears feasible with use
of either elementel fluorine or bremine pentafluoride. The ORNL
development program on this process has been directed toward use
of elemental fluorine in the second step and absorption of the
UFg on Ne¥ as a mesns of completely deconteminating the URg
product from fission product sctivity.

A practiceble process flowsheet for the fused selt-fluorination
and NaF decontemination steps is described in this report along with
the results of laboratory process test studies. 1In addition, the
status of some of the more basic development work carried on con~
currently with the process studies is presented. This supporting
work bhas included further study of the fluorination step, some of
the chemistry involved in the WeF decontamination process, apd
exploratory work on the corrosion problem.

2.0 DESCRIFTION CF PROCESS

The reccmmended flowsheet for the fused salt-volatility process
(see Fig. 2.1) hes the following festures:

Fused Salt-Fluorination Step

 

1. A Fo/UF), mole ratio of 6/1 results in essentially complete
UFg volatilization at 600°C. The wvolatilization is over 90% complete
at a mole ratio of 3/1. If the fused salt contains an vwnusiually bigh
oxygen content (oxymcompounds) a ccupensatory emount of Fo is needed.

2. The optimum fluorination pericd is probably about 2 hr. For
8. 10-kg batch of uranivm a Fo rate of 1.7 scfm is required. A bhigher
flow rate over a shorter period would incresse the Fo efficiency at
the expense of encountering excesgive mechanical entrairment of the
salt and a lower decontamination effectiveness. A lower Tlow rate
would only leasd to magnification of the corrosion problem.
UNCLASSIFIED
ORNL-LR-DWG 34376

 

F.P COLD TRAP
< 80 % Ru activity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
   

 

 

 
 

 

 

   

 

 

2
o
i
c F o CJJ FLUORINE
FEED 2 x| DISPOSAL
< Bwt % inw 50-50 mole % NaF-ZrFy4 @ ©
<7
\ lst NaF BED 2nd NaF BED
FL UORINATION L
- S om | UF~+F, Absorption: .
Fp/UFq mole mg‘;’oifrf” 570-630°C 4 SF mig2 | 160°C, 2hr g Desorption: UFs COLD TRAP
F - =
E fi I00-400°C BD.F ~ 105
| x 3 f hour » D.F. ~s103
¥ Desorption: @ O
WASTE 100-400°C e
< 0005wt % U i hour
> 99 % gross beta and | > UFg STORAGE

 

 

 

 

 

 

 

gamma activity

 

 

 

 

 

— —

 

 

WASTE
D.F~Decontaomination Factor <002 wt % U
F.P-Fission Products >80 % Nb activity

 

 

Fig. 2.1, Flowsheet for Fused Salt-Fluoride Volatility Process.
NaF Absorption~Desorption Cycle

1. A N&F/U‘Weight ratio (U being the total uranium being pro-
cessed) of 2/1 to 3/1 is needed in each of the two NaF beds. The
recommended grede of NeF is 12-20 mesh, prepared from pelletized
material (Harshaw Chemical Company). The NaF beds should be
preconditioned with a slow flow of Fo for 1 hr at 400°C before
pProcess use.

2: The absorption cycle, using only the first bed, is carried
out at gpproximately 100°C. Overstion on a large scale will result
in the bed temperature rising 50° or more due to heat of abscrption.
There is no reason why the temperabure could not be closer to the
triple point of UFg (65°C) initially to partly compensate for this
effect,

2. Descrption of UFg from the first bed through the second
bed requires about 1 hr, using the same ¥, flow rate employed in
the absorption cycle. Ths desorption cycle consisted, in labora-
tory tests, of raising both NaF beds simultaneously from 100 to
LOOBC inm about 0.5 hr, at which time the transfer of UFs to the
cold trap system was essentially complete.

3.0 PROCESS DEVELOFMENT STUDIES

3.1 Fused Sali-Flucrination Work

 

3.1.1 Fluorine Efficlency

 

When fluorine is introduced into fused Na¥F-ZrFi-UF), the uranium
content drops sharply as the Fo/U mole ratio incresses (Fig. 3.1).
Fluorine utilization efficiency is highest when 90% or.more of the
UFg has been volatilized {Fig. 3.2). The efficiency decrasses
thereafter, with the Fo acting essentially as a sweep gas. In an
ideal case, the amount of UFg volatilized would be stoichiocmetrically
equivalent to the amount of Fp introduced up to a Fo/U mole ratioc of
1, thereafter decreasing hypserbolically.

The Fg/U mole ratio required for volstilization of more than 99%
of the UFg was decreased by the elimination of impurities, but it
was not significantly affected by the concentrstion of uranium in
the initial fused salt (Fig. 3.3).

Use of No with the Fp, the method of gas intrcduction into the
melt, and the rate of gas flow hes some effect on the Fg/U’lee
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

UNCLASSIFIED UNCLASSIFIED
QRNL.A R-DWG. 1C827R1 ORNL-L2fW5. 10B2BRT
$00 1.0
50 . \_ — S
0.8 “ ------- - —te
20 8 \
5 2
3 2 N
0 = \ =
= =z \
. o ? ,
; w06 fe- \ S CESV I
S s e \ |
-
: 0 \
2 o —t—
W a \
z 2 N \
z Joa | \ _
= B N
Z a \.
g > \
a " e N
2 o \
0.5 o ®
|_
é 02 L . . .“, [ __k e
2 o
0.2 = - /
/‘
04 * 0 ,——.—Jw——. l

 

0 1 2 3 4 5 0 1 2 3 4 5 6
MOLE RATIO OF FLUDRINE TO TOTAL. URAN'UM

MOLE RATIO OF Fp INTRODUCED TO TOTAL URANIUM

Fig. 3.1. Amount of UF4 Remaining in
375 g of NaF~ZrF4-UF4 (50-46-4 mole %)
Fluorinated at 600°C at a Rate of 100 ml/min Fig. 3.2. Efficiency of UFg Volatilization
as a Function of Amount of Fluorine Intro- as a Function of Amount of Fluorine Introduced.

duced.
UNCLASSIFIED
ORNL-LR-DWG., 12611

 

 

 

{ / ! r i
/ | ’
80 S . |
| |
£ | |
0 60 i ‘
™ |
b |
'L<'£ ;
2
Q 40 | ‘
© l |
W !
|
20 x \ |
[ ’
O l. e e
1 2 3 4
Fo /U MOLE RATIO
RUN U CONTENT
NO. MOLTEN SALT (mcle %)
1 AS-RECEIVED NafF-7rF,~UF, (50-46-4 mole %) 4
6 NaF - ZrF, - UF, (50-46-4 mole %), HYDROFLUJORINATED
4 hr AT 800°C 4
7 PREFLUORINATED NofF-ZrF, (50/46 mole RATIO) PLUS UF,
(76.29%% URANIUM) 4
8 PREFILUCRINATED NaF-Zrk, (50/46 mole RATIO) PLUS UF,
(76.2 %% URANIUM), AIR-SPARGED FOR 2hr BEFORE
FLUORINATION 4
o PREFLUORINATED NaF-Zrf, {(50/46 mole RATIO) PLUS UF,
(72.99%% URANIUM) 4
3 PREFLUORINATED NaF - ZrF, (50/46 mole RATIO) PLUS UF,
(72.9 %" URANIUM) 4
4 PREFLUORINATED NarF-ZrF, {50/46 mole RATIO) PLUS Uk,
(72.9%" URANIUM) o
5 PREFIUCORINATED NoF~ZrF, (50/46 mole RATIO) PLUS UF,

(72.9 %" URANIUM) 1
TTHEORETICAL = 75.8 %

Fig. 3.3. Effect of Uranium Concentration and Impurities of the Fused Salt Fuel Mixtures
on the Fluorine~to-Uranium Mole Ratio Required for UF, Volatilization., Conditions: 100 m]
of F, per min; 1/16~in,~dia sieve plate on dip tube thaéinfroduced fluorine to melt.
ratio, but the results could not be correlated with the known
variables (Fig. 3.4).

These experiments were performed at 600°C in a 2-in.-dis
nickel reactor with a 375-g charge of NaF-ZrF,-UF,. In some of
the tests the salt was made by the addition of UFy to Na¥ -ZrF),
(50/46 mole ratio) that was believed to be relatively free of
oxide impurities as a result of previous use in a fluorination
run. Uranium tetrafluoride concentrations of 1, 2, and 4 mole %
were used to study the effect of concentration. In other tests
NaF-ZrF)-UFy, (50-46-U4 mole %) was used as received. Data were
obtained by direct sampling of the salt at intervals during the
fluorination., The curves were extrapolated to the 100% volatiliza-
tion point for comparison, but usually a sharp break was observed
in the curve between 95 and 100%, which extended the curve to higher
Tluorine~to~uranium mole ratios for volatilization of the last traces
of UF¢.

Volatilization of more than 99% of the UFg from as-received
NaF-2rF),-UF), required a fluorine-to-uranium mole ratio of about
3.1/1,'which was reduced to about 2.2/1 by sparging with HEF for
I hr before fluorination. In two tests with the fuel mixture
synthesized by adding UF}, with a uranium content of only 72.9%
(theoretical, 75.8%) to prefluorinated NaF-ZrF), the fluorine-to-
uranium mole ratio required for more than 99% UFg volatilization
was about 2.4/1. When very pure UF|, uranium assay of 76.2%, was
used, the fluorine-to-uranium mole ratio was 1.4/1, which represents
a fluorine utilization efficiency of about T0%.

A quick~Treeze sampling technique was used in all of the fluori-
nation work. A comparison of sampling methods showed that agreement
to within 3% was obtained in uranium analyses of samples taken ,during
the course of fluorination experiments with the use of a dip ladle,
immersion of a solid rod into the fused salt to obtain a quick~freeze
sample, and samples taken after fluorination by grinding and sampling
the entire batch of salt. This study was made with three different
uranium concentratlions in the Nal’-ZrF) salt, 8, 2, and 0.5 wt %.

3.1.2 Corrosion Studies

 

The corrosion of nickel test couponsg and of a nickel vessel was
fairly low after 20 fused salt-fluorination runs at 650°C, confirming
previous work.” Since conditions changed continually during the
runs and since the various components of the vessel were attacked
to different degrees, a calculated over-all corrosion rate would
have no significance. However, it appears that a large nuiber of
fluorination runs can be made in one reaction vessel before the
100

UF, VOLATILIZED (%)

-10~-

UNCLASSIFIED
ORNL-LR-DWG, 12612

 

 

80 e

 

 

 

 

 

 

 

 

 

F, /U MOLE RATIC

RUN FLOW RATE (ml/min)
NO. Fo No
10 100 0

5 40 200
4 100 G
2 {00 200
3 150 0
4 150 150
& 300 o
7 100 0
8 100 200
9 150 O
1 100 Q
12 100 O

GAS DISPERSION DEVICE ON END OF
/,-in.~DIA DIP TUBE

NONE

SIEVE PLATE, 3/g, -in.-DIA HOLES
SIEVE PLATE, %4 -in -DI& HOLES
SIEVE PLATE, ¥4 -in -DIA HOLES
SIEVE PLATE, 3/ga-in.-DIA HOLES
SIEVE PLATE, 34 -in-D!A HOLES
SIEVE PLATE, 3ga-in-DIA HOLES
SIEVE PLATE, /g ~in.-DIA HOLES
SIEVE PLATE,'/yg -in.-DIA HOLES
SIEVE PLATE, ¥/4g -in.—DIA HOLES

THREE SIEVE PLATES, 3/gq-in.-DIA
HOLES, '/ in. APART

PERCCLATOR DRAFT TUBE

Fig. 3.4. Effect of Sparge Gas Flow Rate and Method of Intreduction of Fluorine

into the Melt on UF

6 Volatilization.
-11-

corrosion is 00 severs. A summary of the resistance of various
metals to fluorination conditions st high temperatures has besn
reported elsevhere.

The "A" nickel reaction vessel was 2 in. in diameter. The
three fest coupons were mounted in an upright position at the
bottom of the reaction vessel, as shown in Fig. 3.5, in suck & way
that one-third the surface area of each coupon extended from the
liquid into the gas phase. The coupons were 3 in. long, 3/ in,
wide, apd 1/4 in. thick. Two of the coupons were "A" nickel
(nominel composition: 99.4% Ni, 0.05% C), and one of them was
cat longitudinally and welded. The third coupon, which was "L"
nickel (nominal composition: 99.4% Ni, 0.01% C), was also cut
longitudinally and welded.

Each run was made with 200 g of NaF-ZiF)-UF) (50-46-% mole %).
The time for a run varied from 4.58 to 0.83 hr, the resction vessel
and the coupons belng exposed to process copditions for a totel of
30 br. The fluorine flow rate varied from 50 to 300 mi/min and was
regulated so that 9.4 moles of fluorine was used per mole of uranium
in each run.

Corvosion of the welded coupons (both "A" and "L" nickel) was
scmewhat grester than that of unwelded ones, but in bolh cases the
corrosion was of the solution type (Fig. 3.6), and there was fairly
uniform surface removal. Dimensional and weight-change analyses
also showed that corrosion may have been slightly grester in welded
than in urwelded coupons (Tsble 3.1). The most severe attsck was
on the outer surface of the flucorine gas inlet tube in the vapor
zone (Fig. 3.7). It is very likely that this attack, about 2 in.
above the salt surface, was due to the freguent admission of atmos-
pheric molsture and oxygen Intec the reactor possibly producing an
aqueous HF and oxidation attack when if was ab an elevated tempera-
ture. The same type of attack 4did not occur at the resctor wall.
Corrosion on the Fo Inlet tube in the liquid zore was more uniform
and varied from 4.0 to 7.5 mils in depth.  The reaction vessel
showed nonuniform attack of 5 to0 9 mils in both the liquid and gas
zones; in the region in contact with molten salt the atiack was of
a solution nature {Fig. 3.8). |

 

3+.1.3 Recovery Yleld ofVUFé

Uraniuwm hexafluoride recovery was more than 99,0% in the 20-run
corrosion series (Sec. 3.1.2). The recovery of uranium was nigh in
all runs (Table 3.2). The wrenium loss in the waste salt wes con-
sistently lowest in the 50-min runs at the highest fluorine flow rate.
This result was possidbly due to & smeller loss of fluorine in corrosion
-12-

UNCLASSIFIED
Y-14847

 

 

Fig. 3.5. Cross Section of Assembled ''A" Nicke! Reaction Vessel. The pitting
on the fluorine gas inlet tube may be seen at point A,

B UNCLASSIFIED !
Y-15140 |

 

v
*
*

 

 
   

Fig. 3.6. Cross Section of Welded "'L"" Nicke! Test Coupon Exposed to Melten
Salt in a Nickel Reaction Vessel. Note uniformity of attack. Etched with KCN

plus (NH,),$,0g. 12X.
Table 3.1.

Welght Loss of Nickel Corrcosion Coupons Tested
in Laborataryw8¢ale Flucrination Runs

 

 

 

Original Final Weight Change
Weight Welght -
Type of Coupon (g) (g) (2) (%)
Welded "L" nicksl 83.9878 80.3760 3.6118 L.3
Welded “A" nickel 86,3445 82,7515 3. 5930 Yoo
Unwelded "A" nlckel 82,6071 80.2160 £.3911 2.9

 
~14-

UNCILASSIFIED
Y-14850

 

Fig. 3.7. Outer Surface Attack of Fluorine Gas Inlet Tube in Vapor Zone of
Reaction Vessel, Section taken at point A of Fig., 3.5, Etched with KCN plus
(NH,),5,0g. 20X,

e UNCLASSIFIED
P : Y-15139

  

Fig. 3.8. Inner Surface of Specimen of ""A'' Nickel Reaction Vesse! Taken From
Region Exposed to Na F—ZrF4"UF4 Fuel. Note nonuniform surface attack. Ftched
with KCN plus (N H4)25208' 250X,
-15-

in the short runs than in the long runs. Out of 3935 g of salt,
341 g of uranium was recovered as UFg, which corresponds to an
initlal uranium content of 8.66%. Analyses of this batch of fuel
ranged from 8.30 to 8.76% uranium. Even if the higher velue is
assumed, recovery was 99.0%.

 

3.1.4 Behavior of NiFs in C@frosion

The behavior of NiFy in molten NeF-ZrF) and Ne¥F-ZrF|-UF) systems
was studied to determine if the presence of this corrosion product
would form sludges which would interfere with salt transfers in the
fluoride volatility process. Although WNiFo has been reported to be
fairly insoluble in this type of salt (approximately 0.2 wi % as Ni
at 600°C), at higher concentrations NiFo readily forms a viscous
dispersion which settles slowly. Based on this observation, it was
concluded that NiF, concentrations up to 2 wt % would not interfere
with salt transfers unless the molten salt were permitted to stand,
unagitated, for long periocds of time.

Anhydrous NiFp was added to molten NaF-2ZrF) (50«50 mole %); the
mix was heated until a clear sclution was obtalned and then cooled
until turbidity reappeared. Solubility values (Table 3.3) estimated
by the disappearance of turbidity were in fairly good agreement with
those determined electrochemically on 53 mole % NaF-47% ZrF) salt.

Additicn of as much as 6 wbt % NiF, to molten NaF-ZrF) (50-50
mole %) at 600°C resulted in the formation of a viscous dispersion
which was falrly stable. Although some settling of NiFo was evident
after only 0.5 hr with an initial nlckel concentration of 2 wt %,
complete settling had not occurred even after 72 hr (Table 3.4).
With 1 wt %, settling was more nearly complete at T2 hr since the
Ni¥o concenbration and viscosity at the bottom could not increase
as much. The results for the Nalf-ZrF|~UF) system appear very
similar to those for the uranium-free system (Table 3.5). However,
with 2 wt % Ni and with uranium present the settling was less after
2 hr than in the test with no uranium. Since the solubllity of NiFo
is represented by the lower 1limit of nlckel concentration encountered
in the settling tests, it appears to be approximately the same in
uranium~-bearing snd uranius-free salt.

The tests were made by dry mixing the required smount of salt
(~ 30 g), and melting in a 1/2-in.-i.d. nlckel tube. MNitrogen was
used initially for agitation and then as a blanket while the material
was kept at 600°C for varicus times. The tube was quickly guenched
with cold water at the end of the test to fix the NiFo concentration
at varicus heights in the tubs. The tube was then cut into 1/2m1n0~
long sections and the salt was analyzed for nickel.
~16-

Table 3.2. Uranium Losses in laboratory-Scale Fluorination Runs

 

 

Uranium Loss

 

Number of Duration Fluorine Flow Rate in Waste
Runs (nr) (m1/min) (% of Total)
1 .58 55 0,11
5 2,50 100 0.02 to 0,16
5 1.25 200 0.06 to 0.23
9 0.83 300 0.01 to 0.0k

 

Table 3.3. Visual Determinations of Solubility of NiFs. in
NaF-ZrF) (50-50 mole %)

 

 

 

 

Temperature Solubility
(°c) (wt % NiFy)
640 0.7
670 1.0
685 1.3

 
-17-

Table 3.4. Sedimentation of NiF, in Molten NaZrF s

 

Temperature: 600°C

il PR

Ni Concentration (wt %)
Relatlive Position

 

of ' Initial  After After  After  After
Ssxple Content 0.5 hr 2 hr 8 hr T2 hr

 

Initial Ni Content® - 2 wt %

 

 

 

1 (top) 1.60 - 0.28 0,20 -
2 1.74 0.73 0.30 0.22 0.20
3 1.72 1.72 2,20 1.40 0.21
L 1.78 1.86 2.62 3.48 1.23
5 (bottom) 2.13 2,02  3.02 3.54 2.99
Initial Ni Content® - 1 wt %
1 (top) --- --- ~—- - 0.22
2 ——— ——— 0.30 - 0,16
3 - - 0.31 —— 0.17
b - — 1.71 - 0.18
5 (bottom) - - 3.06 - -

 

aThe nickel was added as NiFeo
18~

Table 3.5. Sedimentation of NiFo in Molten NaF-ZrF)-UF),
(48-48-4 mole %)

 

Temperature:

600°¢

 

Ni Concentration (wt %)-

 

Relative Initial Ni, 2 wt %

Initial Ni, 1 wt %

Initial Ni, 0.5 wt %

 

 

Pogition
of After After After After After
Sample 2 hr 48 hr 2 hr 48 hr 48 hr
1 (top) 1.26 - ——— - e -
2 1.23 0.25 0.36 0.18 0,24
3 2,12 0.34 0.40 0,20 0.33
) 2.78 2,75 2,53 0. 47 0.22
5 {bottom) 2.81 6.94 3.08 - 0.85

 
-19-

3.2 HNaf¥ Dacontaminafiion Step

 

3.2.1 Qperation in a Temperature Range of 100-k00°¢

The flowsheet (”ig, 2.1) for the volastiiity process provides
for volatilizing UFg from molien fluoride salt with fluorine,
shsorbing the UFg on NeF at 1009C, then desorbing with fluorine
at 100-400°C and pessing the desarbea UFg through & second WaF
bed to the finel cold trap. In the case of long-decayed uranlum
resctor fuel, most of the volaiile achbivity 1n the UWFg stresm
from the molten salt step is due to ruthenium and nicbium, both
of which form wvolatile pentafluorides. Laboratory tests have
demonstrated that ruthenium is not absorbed very much on NaF at
1009C, but effectlvely passes throvgh the first NaF bed with the
excess Tluorine used in the molten salt step. Nioblum, on the
other hand, is absorbed on the NaF with the UFg. This sbsorpiion
is predominantly irrveversible siunce the nicblum remains for the
most part with the Na¥ during U’5 desorption. Use of the second
NaF bed appears essential to prevent cross contamination and to
achieve effective decontamination of the UFg (particulerly from
any ruthenium revolatilized from the end of the first bed) in
processing consecutlve batches. Much of the rubthenium remelining
in the first Na? bed is "plated out” over all of the metal surface;
including that wuear the outlet., This preblem has also been
encountered in dlstillistion work. The effective absorpiion of
ruthenius activity on NaF et high temperatures (see Sec. 3.2.2) is
perheaps partly responsirle for the efficiency of the secan& bed.

in six tests of the double-bed procedure (using 2040 g uranium),
the mcbtlivity of the product Ufg was less than the Ui)~-UXp activity
normel in nsturel uranium. Four of the tests were mede consecubively
with the same 60-ml Naf beds end showed that the decontemination
effectiveness of the system does not decrease with use (able 3.6).
The over-ell beta- or gamma-decontamination factor in each of the
six runs was no less than 105 with 102 veing sttributeble to the
fused salt-fluorination stepl and 103 to the gbsorption-desorption
process. The low product activity made calcilation of specific
decontamination factors imp“actlcal. The effectivensss of the
double bed system wss shown in the six tests by the distribution of
the volatilized activity (Teble 3.7).

An, over-all beta- or gamma»dficonuamination factor of gresler
than lQ% was obbtained by the asbsorption-desorption procedure using
200 ml WeF 1n s single bed (Table 3.8). Calcuwlatlon of speeclfic
decontamination factors was possible because the producht UFg was
radicactive in excess of the Uk -Up level. The typlesl behavior of
the ruthenimn and rdobium activitles was also observed iam this 1un.
D0 -

Table 3.6. Summary of Four Consecutive Runs in Two-Bed Fused Salt
Fluoride-Volatility Process

 

 

Conditions: 128 g of uranium in NaF-Zr¥)-UF), (52-44-U4 mole %) with
gross beta activity per milligram of uranium of
5 x 10° counts/min. Each run fluorinated with
1/1 Fp-Ny mixture for 1.5 hr and then with pure Fs
for 0.5 hr; UFg in Fo-No gas stream absorbed on NaF;
UFg desorbed at 100-400°C through second NaF bed
into a cold trap

Average Fp/U mole ratio in absorption period: L4/1
Average Fo/U mole ratio in desorption period: 2/1
Absorbent beds: 60 ml of 12~ to 4O-mesh NaF in l-in.-dia tubes

NaF/U weight ratio after four runs: 1/1

 

Uranium Retention (%)

 

Product Gamms Activity

 

Product First First Second per Milligram
Yield Cold NaF NaF Waste of Uranium®
Run (%) Trap Bed Bed  Salt (cts/min)
1 83 0.01 0.02 3.6
2 35.2 0,10 0.05 3.1
3 151 0.07 0.08 1.0
Ly 3.8 0.08 0.02 2ol
Over-all 70.1  0.06 0.5 5.1  0.0L 2.5

 

& Gamme, activity per milligram of natural uranium is 8 cts/min.
Table 3.7. Distribution of Volatilized Activity in the Two-Bed NaF Procedure

Runs 1, 2; 3, and b: Consecutive tests with two 60-ml NaF beds (12-40 mesh) in
1-in.-dis tubes. Total NeF/U welght ratio for both beds
in each run: 4/1. NeF/U ratio over four rums: 1/1

Runs 5 and 6: Single-batch runs with two 90-ml NaF beds (12-40 mesh) in l-in.-dis
tubes. Total NaF/U weight ratlo for both beds: 6/1

 

FPercent of Total Volatilized Activity

 

 

 

Runs 1, 2, 3, snd b Run 5 Run 6
Fission First S@c@nd' Figsion First Second Figsion Plrst Becond
Product NaF Na¥ Product NaF NaF - Product NaP NeF i
Activity Cold Trap Bed Bed Cold Trap  Bed Bed Cold Trap  Bed Bed §?

drogs bats, 81 158 0.85 51 W 0.8 59 26 17
Gross gamms, 11 89 Dells | 3 G7 J.07 7 93 0.02
Ru gamms, 97 1.6 1.1 81 18 0,9 86 1 Very low
Zr-Nb gamms, 2.2 98 0.0k Dk w100 0,0h 0.8 99 0,02

Tybal rars-earth beta 4.3 o2 3.6 3 57 0ol 3 97  Very low

 
Table 3.8. Decontamination of UFg in the Single-Bed Fused Salt
Fluoride-~Volatility Process

 

UFG in Fo-Np gas stream from fluorination of NaF-ZrF)-UF), (gross
beta activity per milligram of uranium in salt = 5 x 105 cts/min)
at 600°C; absorbed on NaF at 100°C, desorbed with excess Fo by
increasing the temperature from 100 to L00OC

Absorbent: 200 ml of 12- to LO-mesh NeF in 1l-in.-dia bed

Total Fo/U mole ratio: =~ §

NaF/U weight ratio: =~ 6

Product yield: 87%

 

Decontamination Factors

 

 

Over-all,
Including
Activity Absorption® Desorption® Fluorination

Gross beta 2.1 40 1.2 x 104
Gross gamma 1.2 310 1.k x 10%
Ru gamma 2.k L6 1100
Zr-Nb gamms, 1,0 1600 5.9 x 10%
Total rare-earth beta 1.0 5

 

®Based on activity not absorbed with UFg on NaF but passed into
cold trap.

bBased on activity remaining on NaF after desorption of UFg.
-23-

The uranium retention on the two Nal’ beds In the four-run
series was respectively 0.5 and 5.1% (Table 3.6)., This excessive
loss was due to some back-pressure buildup, which was evlident in
all runs, as the result of either partial plugging of the NaF beds
or a stoppage in the UFg cold trap. It is essential to avold any
prolonged restriction of the gas flow, which would result in a
major part of the UFg remsining in the bed as it approaches 400°C
(see discussion of decomposition effect, Sec. %4.3). For example,
three successive complete cycles with UFg in a two-bed NaF system,
where little plugging occurred, resulted in retentions of 0.04 and
0.01%, respectively, in the two beds. In a single-cycle test with
the same equipment, the uranium retentions were 0.05 and 0.01%
respectively. The per cent of loss, therefore, does not appear to
depend on the number of cycles,

 

3.2,2 Operation at a Temgérature Above 600°C

Fair decontamination was obtained with NaF in either a 9- or
an« 18-in.-long bed at 650°C when the NeF/U welght ratio was 3/1
(Table 3.9, Runs 1, 2, 3) or 6/1 (Table 3.10, Runs 1 and 3). When
the 9-in.-long bed was re-used (Table 3.9; Runs 3 and 4), so that
the over-all NaF/U weight ratio was 1.5/1 for both runs, deconbami-
nation factors for gross beta, gross ganme, and ruthenium dbeta
decreased sharply. With the 18-in.~-long bed the same effect was
observed, although here the over-all NaF/U weight ratio for the
two runs was 3/1 (Table 310, Runs 1-2 and 3-4). The imperfect
decontamlnation of the UFg 1n any single run and the ruthenium
absorbed in one run leads to cross-contamination cver consecutive
runs (illustrated,by sharp decrease in ruthenium gamms-decontami-
netion factors in subsequent runs). However, the high effectiveness
of NaF in removing rutheniwm st > 600°C was in contrast to the much
smaller ef?ect observed in a UFg absorption cycle at 100°C (see
Secs 3.2.1).

Use of NaP at 600-650°C was discontinued because of the excessive
corrosion, the diffieulty with which the tempersture was maintained,
and the great variation of uranium retention on the bed. The best
results were secured with a high flow rate for the UFg-Fo gas stream
Trom the molten salt step. When the flow rate was decreased 1n order
to secure more decontamination, the uraniwas retentlion became very
high (see Sec. 4.3)e Actual sinbering of the NaF bed usually occurred
at high uranium retentions, thus accounting for some of the excessive
corrosion.
Table 3.9. Decontasmination in a 9~in.~long NaF Absorbent Bed

 

UFg-Fo gas stream from fluorination of NaF-ZrF)-UF) (52-44-l mole %) fuel
at 600 to 650°C passed through l~in.-dias NaF bed (12-40 mesh) with
temperature of 650°C in hottest portion; same bed used in Runs 3 and b

Fo/U mole ratio: Run 1 - 3.7 Run 3 - 8.2
Run 2 - 3.6 Run b - 4.8

NaF/U weight ratio in ebsorber: 3/1 for Runs 1, 2, and 3; over-all ratio
for Runs 3 and 4 = 1.5/1

 

Decontamination Factors

 

 

 

Activity Run 1 Run 2 Run 3 Run I
Over-all
Gross beta 1.3 x 10% 5800 1.0 x 103 2900
Gross gomma, 3.2 x 10%  2.17x 10 2.4 x 10 14500
Ru ganma 1700 1600 1.0 x 10% 200
r-Nb gamma 3.2 x 107  T.h x10% 7.0 x10% 1.0 x 107
Total rare-earth beta 5.2 x 105 5.0 x 10% 4.9 x 10% 1.1 x 107

 

Across Absorbent?

 

Gross beta

Gross gamma,

Ru gamms

Zr-Nb gamma

Total rare-earth beta

340 200 35
930 4100 220
9lo 1400 62
900 1.2 x 10% 2000

8 17 8

 

aCa,lculated on basis of

activity found in absorbent and final product.
~25-

Table 3.10. Decontamination in an 18-in.-long NaF Absorbent Bed

UFg-Fo gas stream from fluorinstion of NaF-ZrFL-UFy (52-44-4 mole %) fuel
at 800°C passed through l-in.-dis NaF bed (12-%0 mesh) with hottest
polnt at 670°C; same bed used in Buns 1 and 2 end in Runs 3 and L

Fo/U mole ratio:

NaF/U welght ratio in absorber:

Run 1 - 8.1 Run 3 - 8.4
Run 2 = 9.9 Run b - 10,2

for Runs 1 and 2 and Runs 3 and 4 was 3/1

6/1 for Runs 1 and 3; over-all ratioc

 

Decontamination Factors

 

 

 

 

 

Activity Run 1 Run 2 Run 3 Run k4
Over~all
Gross beta 3900 1600 4300 2400
Gross gamma 9700 2700 2,0 x 104 1000
Ru gamma 1500 140 6700 450 |
Zr-No gauma 3.5 x 10% 2.9 x10% 9.8 x 1ot 5.2 x 10
Total rare-earth beta 3.7 x 10“ 1.0 x 102
Across Absorbent®
~ Gross beta 2 1k
Gross gamma 4o 8g
Ru géumna 5 31
Zr-Nb gamms, 250 TOO

 

%Calculated on basis of activity found in absorbent and finsl product.
~26-

4.0 BASIC CHEMICAIL STUDIES

4.1 NaF Capacity for UFg

 

The absorption capacity of Nal for UFg was about 0.9 g of
uranium per gram of salt for one lot of Harshaw Chemical Company
meterial under a variety of conditions (Table 4.1). The capacity
of Bsker and Adamson Company reagent-zrade NaF was 1.89 g of
uranium per gram of salt. Absorption of UFE on Naf under an
initial vacuum indicated a capacity range of 0.80 to 0.90 for
different mesh sizes, while absorption values obtained when the
excess UFg was removed with fluorine as a sweep gas varied from
0.72 to 0.86. Practically the same capacity values were obtained
at 70, 100, and 150°C and at two different pressures. The capacity
of 0.9 for the Harshew Chemical Company materisl is independent of
the particle size to which it is degraded. The 1.89 wvalue for the
capacity of the Baker and Adamson Company material corresponds
closely to the molecular ratio in the complex UFS+3Nalf'. A complex
of this nature has been reported by other investigators.

As indicated, the original 1/8-in. pellets (Harshaw Chemical
Company) had practically the same capacity as the same material
degraded to 12-40 mesh. Examination showed that the lemon-yellow
color, characteristic of the complex, was uniform inside the
pellet. Although the material was therefore fairly porous, the
limited capacity of approximastely 474 of theoretical and a surface
area of only about 1 mc per gram indicated that the crystallite
size was sbout 103 sngstroms. Severe conditioning of the NaF;
such ag heating to 100-400°C with either vacuum or an atmosphere
of fluerine to eliminate HF or moisture, did not affect the
capacity measurements. Welght-loss tests after prolonged heating
showed thet the HF or molisture content was no more than 0305%0

4.2 Pressure of UFg Above UPg-3NaF Complex

 

Vapor-pressure data for the UFgG-NaF complex were obtained by
the transpiration method over the temperature range 80 to 320°0¢
(see Fig. 4.1). The remction involved is described by the equation
UFg*3NeF (s) — UFs(g) + 3NaF(s). (1)

The data are fitted with the analytical expression

log By = 10.88 - (5.09 x 103/1),
Tavle 4.1. Absorption Capscity of NeF for UFg

-27-

 

Material

Conditions

 

Fg Pressure
(psia)

&

Removal of
Excess @FG

Capaalty
(g of U per g
of NelF)

 

Harshaew Chemical o,

 

 

 

12 to 40 mesh 15% 160 Not removed 0.86
15% 100 By vacuum 0,56

To58 100 By vacuam 0,80

158 TC By vacuum 0,88

is58 150 By wvecuun 0.90

155 100 By fluorine 0.86

15P 100 By fluorine 0,86

8 to 12 mesh 150 100 By fluorine 0.81
1/8«1n, pellets 150 100 By fluorime 0.72
159 100 By vacumm 0,61

Baker and Adamson Co.
Powder 152 100 By vacwum 1.89

 

aUF6 introduced o N&F under vacuum.

hUFs introduced st atmospheric pressure 1o displace fluorine.
UFg VAPOR PRESSURE {(mm Hg)

   
    

1 O I ._.'."__ B A R

10

~28-

UNCLASSIFIED

ORNL-LR-DWG. 12613

 

 

 

 

 

 

 

 

 

 

 

0 100 mi/min
& 50 ml/min
& 20 mi/min

 

 

 

T T Ny FLOW RATE:

 

T T T Tl 47 - T T
T T r R i* i e 1" -

 

 

 

 

 

 

 

" log A, =10.88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

%

 

- 5.09

 

 

 

 

X 103/ 7— e

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.6 1.9 2.2
A

Tk

(X4103)

Fig. 4.1. Dependence of UFé-NdF Complex Vapor Pressure on Temperature,
-29..

wvhere T is the absolute temperature, Use of the Clausius~Clapeyron
formula with this equation gave a value of +23.2 kcal per mole of
UF, for the enthalpy change of Eq. 1.

The data were obtained by passing nitrogen at a flow rate of
100 ml/min, or less, through a prepared bed of the UFg-NaF complex
at any desired temperature, trapping out the UFg in the nitrogen
stream in a dilute A1(NO3)3 solution, and measuring the total
volume of nitrogen with a wet-test meter. The UFg hydrolysis
samples were analyzed by colorimetric or fluorimetric methods +o
an accuracy of better than +5%. Temperature control of the bed
was maintained always to within +0,29C. The UFg-NaF' complex was
prepared by saturating a 30-g bed of 12-20 mesh Harshaw NaF in a
l-in.-dia vertical nickel reactor with UFg at 100°C. Over 90% of
bed saturation was maintained throughout the tests. Check runs,
made at various No flow rates, showed no flow-rate effect.

Crude adiabatic experiments were made with 100-ml batches of
NaF to show that the reaction heat of 23.2 keal per mole of UFg
produces a large temperature rise, approximately 130°C, if total
saturation with UFg (preheated to about 100°C) is carried out
guickly in a period of a few minutes.

4.3 Decomposition of UFg-3NaF Complex

 

A study of the decorwosition of the UFg+3NaF complex at temper-
atures of 245°C and higher has confirmed the belief that uranium
retention on the NaF bed will be excessive in the NaF desorption
step 1f the temperature and sweep-gas flow rate are not properly
controlled. The retention results from decomposition of the UF o= 3Nal’
complex to a complex of NaF with UFs5, which is not volatile.”? A
maxirum decomposition rate of about 0,01, 0.09, and 0.5% per minute
is incurred at 250, 300, and 350°C, respectively, in the absence of
fluorine if all the uranium is assumed to be 1n the form of the
solid complex UFg-3NalF. Under optimum condltions, UFg desorption
from the NaF bed (see Secs. 3.1.1 and 4,2) competes favorably with
the decomposition effect, resuliting in a small uranium loss.

Fluorine appears essential to inhibit the decomposition reaction,
and possibly to promote refluorination of the nonvolatile U{V) com-
pound, formed in the decompesition. The possibility of any uranium
retention by the decomposition mechanlsm in the absorption step at
100°C appears to be insignificant, even over extended periods.

The temperature-dependence of the rate of decomposition of
UFge3NalF was determined in a series of runs over the temperature
-30-

range 245-355°C (Fig. 4.2). The probable reaction involved is
UFge3NaF (s0lid) ——» UFg«xNaF (solid) + O.5F; (gas).
The dependence of decomposition rate on temperature is
log r = 6.09 - 5.22 x 103/T

where r is the fractional decomposition rate in reciprocal ninutes

and T is the absolute temperature. The rate was calculated on the
basis of an absorption capacity of 1.33 g of UFg per gram of NaF.

The energy of activation was calculated as +23.9 kcal/mole of UFg-«3Nal
complex. It is possibly significant that this energy change is
approximately the same as the enthalpy change of +23.2 kcal per mole
involved in the volatilization of UFE from the U¥Fg-3NaF complex.

The decomposition data were obtained with 4~ to 5-g samples of
Na¥ (Harshaew Chemical Company pellets classified to 12-20 mesh) held
in a U tube (1/2-in.-dia stainless steel tubing), through which
gaseous UFg was passed at atmospheric pressure. An oll bath was
used Tor manually controlling the temperature 1o iBOC during the
course of each experiment. The runs were ended by removing the
cil vath and rapidly cooling the sample. The UFge-3NalF was formed
at the beglnning by saturating the NaF at a high UFg flow rate,
after which the flow rate was decreased to 0.1-1 g/min for the
remaining time, The length of the runs at various temperatures
was adjusted so as to obtain a U(V) content in the final product
of 1 to 10%. The excess UFg still absorbed on the NaF at the end
of each test was not desorbed because of the difficulty of achieving
this without increasing the U(V) content. The amount of UFg«3NaF
complex affected by the reaction was determined from U(V) and U(VI)
analyses. A temperature sbove 355°C was not used since it would be
difficult to maintain saturation of the NaF with UFg without use of
a pressurized system.

In preliminary work on the decomposition reaction, a U(V)
content of 20-26% represented a limit which could not be exceeded
in one cycle of saturation of the NaF with UFg at 100°C followed
by heating as a closed system to 350-400°C. Generally, in these
runs, the NaF weight increase verified the asssumption that the
decomposition product is a complex of UFs with Na¥. X-ray crystal-
lography data indicated that the UF5’ ag complex in the decomposi-
tion product has an orthorhombic strycture with cell dimensions
a0 = 4.90 &, bg = 5.47 &, ¢g = 3.87 A. The x-ray pattern of y-
or B-UF5 was not observed in the material.
-31-

UNCLASSIFIED
ORNL-LR-DWG. 156894

 

 

 

 

 

 

 

 

 

 

2
o)

~ l |

- log r=6.09-5.22x103/T

- e '

UFg" 3NaF ~——= UF5- x NaF +, Fp
c
£
}_ -
o .
l& »
o
=
o
- *
&L 1073
% | ®
O | .
&
L)
i) |
.|
2 N
=
) N
'_
&
<L
m o
L
$
®
®
o4
|5 1.6 1.7 1.8 1.9 2.0 (X1073)
A
Tk

Fig. 4.2. Dependence of Rate of UF6'3NGF Complex Decomposition on Temperature,
-32-

5.0 RECOMMENDATIONS

Both the fused salt-~fluorination and NeF-absorption steps of
the volatility process require further research and development.
The presented laboratory studies have demonstrated the chemical
feasibility of the process. Much further work will be needed,
however, in sdapting the process for use st the pilot-plant level.
Chemical-engineering regulrements will doubtless be a major factor
in future modification of the process for use with various types
of reactor fuel. Continued research is alsoc essentisl to securing
a better understanding of the basic chemistry invclved in the two
steps of the wvolatility process.

In the NaFf absorption step, there is particular need of research
on the behavior of NaF with vericus volatile fission-product fluorides.
It seems quite likely; for example, that Na¥ has s high capaclty with
respact to the fluorides of zirconium and niobium due to the formation
of defTinite chemlcal complexes. This would make renewal of the NaF
beds possibly more economical than regeneration. On the other hand,
the vclatile fluorides of other fisslen products, such as Mo, Te,
and 1, may behave similarly to RuFs which apparently does not form
a complex except at much higher temperatures.

The NaF method of decontaminating UFg from fissiocn-product
activity is potentially useful in many diverse forms. Although
the decontsmination factor of 107 (for the entire process) probably
dees not represent the top limit obtsinable; use of a third bed
should be considered if the double-bed system does not give the
decontamination necegsary with high-burn-up reactor fusls. It
seems gquite possible also thet the multiple-bed system could be
replaced with a single bed, either moving or fixed, used in con-
Junection with a thermal gradient. Much development work on the
optimum physical structure of the NaF is needed. A suggested
modification is the use of NaF structurally supported in some
manner, either with metal (Ni) or a diluent material such as
CaFn or A1F3.
1,

-33-

6.0 REFERENCES

G. Y. Cathers and R, B, leuze, "A Volatilizsiion Process for
Ursnium Recovery,"” Paper 278, presepted at Nuclear Bnginesring
and Science Congress, Cleveland, Dec. 12-16, 1955. Published
in Reactor Operastional Problems, Vol. II, pp. 157-164, Pergamon
Press, London, 1957.

 

H. H. Hyman, R. C. Vogel and J., J. Katz, "Decontamination of
Irradiated Reactor Fuels by Fractionsal Distillation Processes
Using Uranium Hexefluoride,” Vol. IX, Proceedings cf the
International Conference on the Pesaceful Uses of Atomic Energy,
Aug. 1955,

W. R. Myers and W. B. Delong, "Fluorine Corrosiocn,” Chemical
Engineering Progress, May, 1948.

 

R. A. Gustison, S. 5, Kirslis, T. S, McMillan and H. A. Berchardt,
"Separation of Ruthenium from Uranium Hexafluoride, " K~586,
April 26, 1950.

H. Martin, A. Albers and H, P. Dusit, "Double Fluorides of Uraniun
Hexafluoride," Z Anorg. Allg. Chemie 265, 108-138 (1951).
ORNL-2661
Chemistry-Separation Processes
for Plutonium and Uranium
TID-4500 (14th ed.)

INTERNAL DISTRIBUTION

 

1. C. E. Center 53. F. R. Bruce
2. Biology Library S5h. D. BE. Ferguson
3. Health Physies Library 55. R. B. Lindauer
4-5, Central Research Library 56. H. B. Goeller (
6. Reactor Experimental 57+ C. W, Hancher
Englneering Library 58. R. A. Charpie
7-26. Laboratpry Records Department 59. J. A. Lane
27. Laboratory Records, ORNL R.C. 60. M. J. Skinner
28. A. M. Weinberg 61. R. E. Blanco
29. L. B. Emlet (K~-25) 62. G. E. Boyd
30. J. P. Murray (Y-12) 63. W. E. Unger
31. J. A. Swartout 6h. R. R. Dickison
32. E. H. Teylor 65. A. T, Gresky
33. E. D. Shipley 66, E. D. Arnold
34-35, F. L. Culler 67. C. B, Guthrie
36. M. L. Nelson 68. J. W. Ullmann
37. W. H. Jordan 69. K. B. Brown
38. C. P. Kein 70. K. 0. Johneson
39. J. E. Frye, Jr. T7L. d. C. Bresee
40, 8. ¢. Lind 2. M. R. Bennett
41. A, H. Snell 73. G. I. Cathers
2. A. Hollaender T4. R. L. Jolley
43. K. Z. Morgan 75. P. M. Reyling
Wi, M, T. Kelley 76. M. Benedict (comsultant)
L5, T. A. Lincoln 77. D. L. Katz (consultant)
46. R. 8. Livingston 76. C. BE. Larson {consultant)
L7. A. 8. Householder 79. I. Perlman (consultant)
8. ¢. 8. Harrill 80. J. H. Rushton (consultant)
4G, C. BE. Winters 81. Hood Worthington (consultant)
50. H. B. Seagren 2. ORNL - Y-12 Technical Library,
51. D, Phillips Document Reference Section
52. W. K. Eister

EXTERNAL DISTRIBUTTION

 

83, Division of Research and Development, AEC, ORO
84-600. Given distribution as shown in TID-4500 (14th ed.) under Chemistry-
Separation Processes for Plutonium and Uranium (75 copies - OTS)