ORNL-3544.txt

ORNL-3544
UC-4 — Chemistry
TID-4500 (25th ed.)

REDUCTION OF URANIUM HEXAFLUORIDE

2N

/ ”

[ - o ol iR
/ Cind i :
\s - -

S P —

RETENTION ON BEDS OF MAGNESIUM " 1L 2}
FLUORIDE USED FOR REMOVAL OF
TECHNETIUM HEXAFLUORIDE

Sidney Katz

OAK RIDGE NATIONAL LABORATORY
operated by

UNION CARBIDE CORPORATION
for the

U.S. ATOMIC ENERGY COMMISSION
DISCLAIMER

This report was prepared as an account of work sponsored by an
agency of the United States Government. Neither the United States
Government nor any agency Thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product,
process, or service by trade name, trademark, manufacturer, or
otherwise does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any
agency thereof. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States
Government or any agency thereof.
DISCLAIMER

Portions of this document may be illegible In
electronic image products. Images are produced
from the best available original document.
Printed in USA. Price: $0.50 Available from the
Office of Technical Services
U. S. Department of Commerce
Washington 25, D. C.

LEGAL NOTICE

This report was prepared as an account of Government sponsored work. Neither 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 accuracy,
completeness, or usefulness of the information contained in this report, or that the use of
any information, apparatus, method, or process disclosed in this report may not infringe
privately owned rights; 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 behalf of the Commission'' includes any employee or

contractor of the Commission, or employee of such contractor, to the extent that such employee

or contractor of the Cemmission, or employee of such contractor prepares, disseminates, ur
provides access to, any information pursuant to his employment or contract with the Commission,

or his employment with such contractor.

ORNL- 3544

Contract No. W-7405-eng-26
CHEMICAL TECHNOLOGY DIVISION

Chemical Development Section B

REDUCTION OF URANIUM HEXAFLUORIDE RETENTION ON BEDS OF
MAGNESIUM FLUORIDE USED FOR REMOVAL OF TECHNETIUM
HEXAFLUORIDE

Sidney Katz

DATE ISSUED
JAN 3 1 1964

OAK RIDGE NATIONAL LABORATORY
Qak Ridge, Tennessee
operated by
UNION CARBIDE CORPORATION
for the
U.S. ATOMIC ENMERGY COMMISSION
THIS PAGE
WAS INTENTIONALLY
~ LEFT BLANK
CONTENTS

Abstract
I»nfroducfion
Materials
Experimental Work

Test 1: Deleterious Effect of Grossly Inadequate Pretreatment of
Magnesium Fluoride Pellets

Test 2: Favorable Effect of Extensive Pretreatment of Magnesium
Fluoride Pellets

Test 3:. Importance of the Prefluorination Step in the Pretreatment

of Magnesium Fluoride Pellets

Test 4: Desorption of Uranium Hexafluoride from Well-Stabilized
Magnesium Fluoride

Test 5: Lack of Effect of Hydrogen Fluoride on Well-Stabilized
Magnesium Fluoride Pellets

Discussion

References
REDUCTION OF URANIUM HEXAFLUORIDE RETENTION ON BEDS OF
MAGNESIUM FLUORIDE USED FOR REMOVAL OF TECHNETIUM
HEXAFLUORIDE

Sidney Katz

ABSTRACT

The excessive loss of uranium incurred when discarding magnesium
fluoride, (the adsorber used to selectively remove technetium hexa-
fluoride from uranium hexafluoride streams) is a problem common to all
volatility processes for recovering enriched uranium fuels. As a result
of the work described, two schemes for the release of the uranium hexa-
fluoride from the magnesium fluoride and its separation from the tech-
netium hexafluoride are proposed. One scheme depends on preferential
thermal desorption of the uranium hexafluoride at 350°C and the other
on selective adsorption of the uranium hexafluoride on sodium fluoride
pellets following the codesorption of the two hexafluorides with fluorine
at 500°C from the magnesium fluoride pellets. These proposals are aim-
ed at reducing the amount of retained uranium to less than 1 g per
1000 g of discardable magnesium fluoride.

!

In the work reported here, the deposition of uranium on magnesium
fluoride as a function of heating, fluorination, and hydrogen fluoride
pretreatment of the magnesium fluoride pellets prior to exposure to
uranium hexafluoride was characterized in a series of gasometric studies.
The dependence of the quantity of uranium hexafluoride adsorbed on
pressure and temperature was also determined.

The data show that physical adsorption is the mechanism far the
deposition of most of the uranium hexafluoride on well-stabilized
magnesium fluoride pellets. More than 90% of the adsorbate can be
removed by heating to 350°C. Chemisorption (formation of a double
salt) is probably not involved because of the small 0.05) mole ratio
of UFg/MgF7 observed. .

INTRODUCTION

This report describes a gasometric study of the mechanisms of the undesirable
deposition of uranium hexafluoride on magnesium fluoride and suggests two methods to
reduce to acceptable amiounts the uranium loss on the discarded magnesium fluoride.
The codeposition of uranium on magnesium fluoride beds that are used to
selectively remove technetium hexafluoride from uranium hexafluoride streams is
a problem common to all volatility processes for recovering enriched uranium from
spent fuel elements. The magnitude of this codeposition is indicated from the ex~
perience in the Oak Ridge National Laboratory (ORNL) Volatility Pilot Plan’r,] in
which 14 g of uranium was deposited on 1000 g of magnesium fluoride out of the 600 g
of uranium passed through the bed as uranium hexafluoride. The extent of codeposition
was somewhat less in a large-scale operation at the Paducah Gaseous Diffusion Plant,2
where massive quantities of uranium hexafluoride are passed through magnesium fluoride
beds; 3.25 kg of uranium was recovered from 500 kg of the used magnesium fluoride.

In the previous application of magnesium fluoride beds for the separation of
technetium from uranium hexafluoride at the Paducah Caseous Diffusion Plant, the
codeposition of uranium on the bed was of small concern because (1) the uranium
was of low isotopic enrichment and represented only a small fraction of that which
passed through the bed, and (2) the technetium recovery process dlso provided eco-
nomical uranium recovery. However, in the ORNL volatility application, the iso-
topic enrichment is high; the fraction of the throughput codeposited is greater; and
the reprocessing costs are higher because of the fission product activity involved.
Since in volatility applications, it is desirable to merely discard the used magnesium
fluoride, the uranium that accompanies it must be held to an economic maximum
(less than 1 g of uranium per 1000 g of magnesium fluoride).

In the work reported here, the quantity and form of uranium deposited was studied
as a function of a variety of pretreatments of the magnesium fluoride pellets. The
pressure and temperature dependence of the amount of adsorbed uranium hexafluoride
was also observed. The data showed that the uranium hexafluorlde is physically ad
sorbed when well-stabilized magnesium fluoride is used. Also, the uranium hexa-
fluoride can be desorbed to such an extent that the used magnesium fluoride can be
economically discarded.

MATERIALS
Magnesium Flucride Pellets

The "as-received" pellets, taken from the same batch used in the ORNL Volatility
Pilot Plant, contained 10.7% water. They had been manufactured at the Paducah
Gaseous Diffusion Plant to meet the requirements of their technetium trapping program.
Similar pellets were reported to have a surface area of 111 m2/g after heating and purg-
ing with fluorine.2 :

In a preliminary examination of the pellets, the weight loss and surface area were
determined for a number of possible pretreatments. The effect of heating the pellets
for half an hour was tested at four temperatures until only 0.07% water remained. The
data follows:
Temperature (°C) Cumulative Wt Loss (%) Surface Area (m2/g)
160 10.0 102
260 13.2 80
360 16.6 35
460 - 17,5 20

From the original water content (10.7%) and the cumulative weight loss (17.5%), «
calculation indicates that 52.3% of the water was converted to hydrogen fluoride
during the heat treatment, “

The effect of a combination of heating at 160°C for a half hour followed by
treating with fluorine at atmospheric pressure for 2 hr at 100°C resulted in a cumu-
lative weight loss of 11.1% and a surface area of 89 m2/g.

These data permit an estimate of the physical and chemical properties of the
magnesium fluoride pellets as used in the tests that follow.

EXPERIMENTAL WORK

A gasometric sysfem3 was used in a series of five tests to determine (1) if in-
adequate pretreatment of the magnesium fluoride could result in gasometrically
measurable adsorption of uranium hexafluoride, (2) how much uranium hexafluoride
would be adsorbed on well-stabilized magnesium fluoride, (3) the importance of the
fluorination step in the pretreatment of magnesium fluoride, {4) the temperature de-
pendence of the desorption of uranium hexafluoride from magnesium fluoride, and
(5) whether hydrogen fluoride pretreatment of the magnesium fluoride influenced
subsequent uranium hexafluoride adsorption.

In each of the tests, after some specific pretreatment of the magnesium fluoride
pellets, a gasometric measurement of uranium hexafluoride adsorption was made under
the following conditions: 200 mm Hg pressure of uranium hexafluoride with the mag-
nesium fluoride pellets at 100°C (deviations from these conditions are noted in specific
cases). After the adsorption, the chemical form of the retained uranium was determined
by chemical analysis and by gas evolution methods. The definitive chemical makeup
of the magnesium fluoride pellet, itself, was deduced from chemical analysis and
gasometric measurements.

The data are presented with the description of each of the five tests and are sum-
marized in Table 1.

Test 1: Deleterious Effect of Grossly Inadequate Pretreatment of
Magnesium Fluoride Pellets

Part A: Pretreatment by Heating at 150°C

The conditions and observations are listed below:
Table 1. Adsorption of Uranium Hexafluoride on Magnesium Fluoride: Effects of Various Pretreatments

Magnesium Fluoride Pellets

Uranium Hexafluoride

Magaesium Fluoride Pellet Residue

Pretreatment Retained (millimoles) Wt % Uranium Final Ny Surface
Test Wt (g) Heat Fo HF Gasometric®  Anal.¢  Total U(Vl) Wt (g) Area (m2/g)
1A 0.631  150°C No No 0.1 125 to
2 hr 25°C
1B No No Yes 0.7 at 25°C 18.7 18.4 0.728
2 12.526  400°C 300°C No 0.95 0.64 1.40 1.37 10.877 16.5
reached 1 atm
slowly 18 hr
3A 12.594  500°C No No 10.564
reached
slowly
3B 400°C No No 1.23 0.68 1.44 1.01 10.805 15.2
1/2 hr
4 42,651 450°C 350°C No 3.92 0.12 0.05 35.625 17.0
' 2 hr 1. atm
2 hr
5 253159 No 350°C  Yes  2.20 0.23  2.05 25.320 17.6
1 atm
1 hr

a . . . .
Hydrogen fluoride treatment as used to activate sodium fluoride.3

Gasometric measurement with pressure of 250 mm Hg UFg in reactor; at 100°C unless nated otherwise.

c » . s . M - .
Remaining on the pellet residue after evacucting reactor at 100°C; calculated from urarium analysis.

d_, . . I .
This starting material is part of the pellet residue from run 4,
Magnesium fluoride: 0.631 g of "as-received" pellets

Pretreatment: Heated at 150°C for 2 hr, with pumping to about
1 mm Hg

UF¢ adsorption:’ None detected gasometrically at 125°C to 25°C

It was concluded that the limit of detection for the gasometric system (0.1 millimole)
was too large to permit the measurement of the adsorption of uranium hexafluoride on
a small sample to magnesium fluoride (10 millimoles) under these conditions.

Part B: Effect of Excess Hydrogen Fluoride on Adsorption by Inadequately Pretreated
Magnesium Fluoride Pellets

The conditions and observations follow:
Magnesium fluoride: Residue from part A

Pretreatment: Exposed to hydrogen fluoride at atmospheric
pressure at room temperature; removed excess
gases by pumping to less than 1 mm Hg

UF, adsorption: 0.7 millimole at 25°C, by gasometric measurement

Desorption: Heated the pellets to 320°C, resulting in evolution
of 1.2 millimoles of gases which were not UFy,
as determined from condensation characteristics

Solid residue: 0.728 g containing 18.7 wt % total U [18.4 wt % U(VI)]}

The implications are that the adsorbed uranium hexafluoride had been converted to
a nonvolatile oxyfluoride by reaction with water. Also, treating magnesium fluoride
that contains water with hydrogen fluoride makes the water more readily available for
reaction with adsorbed uranium hexafluoride. (It will be shown in test 5 that excess
hydrogen fluoride does nol similarly offect.adsorption of uranium hexafluoride on well-
stabilized magnesium fluoride.)

Test 2: Favorable Effect of Extensive Pretreatment of Magnesium Fluoride Pellets

Conditions and obscrvations were:

Magnesium fluoride: 12.526 g of "as-received" pellets; larger sample
taken to improve gasometric sensitivity

Pretreatment: Heated slowly to 400°C; copious qudntifies of
gas evolved,mostly below 200°C: fluorination
for 18 hr at 300°C; fluorine pressure, 1 atm

Solid residue: 10.877"g containing 1.40 wt % U [1.37 wt % U(VI)];
surface area, 16.5 m%/g
Converting the results to a weight basis, about 14 g of uranium was retained as hexa-
valent uranium per 1000 g of magnesium fluoride. Another 7 g uranium per 1000 g of
magnesium fluoride had been adsorbed at 200 mm Hg pressure and desorbed upon pump-
ing down to about 1 mm Hg pressure.

Test 3: Importance of the Prefluorination Step in the Pretreatment of
Magnesium Fluoride Pellets

Conditions and observations for this test are shown below.
Magnesium fluoride: 12,594 g of "as-received" pellets

Pretreatment: ' Heated to 500°C slowly; 105 millimoles of gas
evolved; the 105 millimoles of gas are estimated
to weigh 2.03 g, assuming 52.3% of held water
was converted to hydrogen fluoride; that weight
agrees well with a measured weight loss of 2.03 g
during pretreatment; sample was removed for that
weight measurement

UF¢ adsorption: Reheated to 400°C for half an hour, starting part B
_ of this test; 1.23 millimoles measured gasometri-
cally; after removing uranium hexafluoride in
gas phase from reactor by pumping, only 0.68
millimole remained, as moasured by analysis of
residue

Residue: . 10.805 g containing 1.44 wt % total °
U[1.01 wt % U(VI)] surface area, 15.2 m2/g

Only 4 g of uranium per 1000 g of magnesium fluoride was retained in a chemically re-
duced form when prefluorination was omitted, that quantity may be lower if the adsorp-
tion is performed in the presence of fluorine, as is done in the Volatility Pilot Plant at
ORNL. This suggests that prefluorination of the magnesium fluoride may not be necessary.

Test 4: Desorption of Uranium Hexafluoride from Well-=Stabilized
Magnesium Fluoride

In the desorption test, the conditions and cbservations were:

ng.nesium fluoride: 42,651 g of "as-received" pellets

Pretreatment: Heated at 400°C for 2 hr followed by fluorination
for 2 hr at 350°C under fluorine at 1 atm

UF ¢ adsorption: 3.92 millimoles by gasometric measurement; 2,25
millimoles estimated to have remained after
removing uranium hexafluoride in gas phase
from reactor by pumping
UF¢ desorption: " The temperature was raised stepwise, holding each
new temperature for half an hour

Cumulative Desorption

Temperature (°C) (millimoles)
160 0.40
220 1.36
345 2.89
420 3.28
480 >4.28
Residue: 35.625 g containing 0.12 wt % U[0.05 wt % U(VI)];

surface area, 17.0 m2/g

It is significant that, of the uranium adsorbed on well-stabilized magnesium fluoride,
most of the hexavalent uranium is readily desorbed; the chemically reduced uranium
remaining as a residue represents less than 1 g of uranium per 1000 g of magnesium
fluoride. Assuming that uranium hexafluoride was desorbed first in this test, a temper-
ature of less than 350°C should be adequate for removing adsorbed uranium hexa-
fluoride down to acceptable concentrations. The volatiles desorbed in excess of the
uranium hexafluoride must have been residual compounds not previously removed, for
example, water. .

Test 5: Lack of Effect of Hydrogen Fluoride on Well~Stabilized
Magnesium Fluoride Pellets
The conditions and remarks are listed below.
Magnesium fluoride: 25.315 g or residue from previous test

Pretreatment: Refluorination for 1 hr at 350°C under 1 atm of Fgp;
expesing to 1 atm of HF followed by pumping
off excess, all at room temperature

UF 4 adsorption: 2.20 millimoles, measured gasometrically

UF, desorption: Residue raised to 350°C and evolved gases removed
: by pumping

Residue: 25.320 g containing 0.23 wt % U,[0.05 wt % U6+]

and measuring 17.6 m2/g

No appreciable retention of uranium was noted when well-stabilized magnesium fluo-
ride was pretreated with excess hydrogen fluoride, in contrast to the results obtained
in test 2b.
DISCUSSION

The uranium adsorbed after the exposure of rigorously pretreated magnesium
fluoride to uranium hexafluoride at 100°C is largely hexavalent and can be removed
by heating or pumping (see tests 2, 3, 4, and 5 in Table 1); therefore, the adsorbed
uranium must be present as the hexafluoride, either adsorbed physically or in the form
of a complex. Physical adsorption is the most probable mechanism, since the maximum
quantity of uranium held is insufficient to yield a reasonable complex with the mag-
nesium fluoride. Significantly, at 350°C, less than 1 g of the uranium per 1000 g of
magnesium fluoride remains adsorbed.

The drastic loss of surface area of the magnesium fluoride pellets (down to 15.2 ta
17.6 m2/g for the pellets in tests 2, 3, 4, and 5) represents primarily the cumulative
sintering effects of exposure to heat. The quantities of uranium hexafluoride adsorbed
or recovered in these tests and in ORNL pilot plant run R-8 and at Paducah3 are in
sufficiently good agreement to indicate that the magnesium fluoride in the larger=-

scale operations also undergo surface area reductions.

Some of the volatile material associated with the pellets remains trapped even
after heating them to over 400°C and after extensive fluorine treatment at 300°C
(see test 4). The occluded volatile material, presumably a mixture of hydrogen
fluoride and water, must be unavailable to the uranium hexafluoride since otherwise
the water would react with the hexafluoride and prevent subsequent desorption of the
uranium. ‘

Little uranium in a reduced valence state was found on the magnesium fluoride
residues except where prefluorination had been omitted; in each case (tests 1 and 3)
about 0.3 to 0.4% quadrivalent uranium waos present. This reduction may be accounted
for by an equivalent fluorination of the nickel reactor or the tray upon which the
pellets rested.

CONCLUSIONS AND RECOMMENDATIONS

Physical adsorption is responsible for most of the uranium adsorbed on well-
stabilized magnesium fluoride pellets, and the uranium hexafluoride can be removed
down to less than 1 g of uranium per 1000 g of magnesium fluoride by heating to 350°C.
These two facts lead to two schemes for the release of the physically adsorbed uranium
hexaflucride and its separation from technetium hexafluoride and provide a means of
economically discarding used magnesium fluoride pellets.

" The first scheme, which appears simplest to try and put into pilot-plant practice,
is to heat the loaded pellet bed to about 350°C in order to preferentially release the
uranium hexafluoride. According to the data of Golliher and co-workers,2 the
technetium compound is poorly desorbed (18% at 1000°F in nitrogen).
The alternative scheme is to release both the uranium and technetium hexa-
fluorides from the loaded pellet bed by heating to 500°C in fluorine and thén to
selectively adsorb the uranium hexafluoride on sodium fluoride at 100°C; Golliher
~ and co-workers2 found that only 4% of the technetium that passed through a sodium
fluoride trap at 200°F was retained.

Simplifying the pretreatment of the magnesium fluoride pellets might be considered
also. A more rigorous preheating treatment may permit omission of the fluorination step.

REFERENCES

1. Chemical Technology Division, Annual Progress Report, Period Ending May 31, 1963,
ORNL-3452, p 26-50 (Sept. 20, 1963).

2. W. R. Golliher, R. A. LeDoux, S. Bernstein, and V. A. Smith, Separation of
Technetium=99 from Uranium Hexafluoride, TID-18290 (1960).

3. S. Katz, A Gasometric Study of Solid-Gas Reactions, Sodium Fluoride with
Hydrogen Fluoride and Uranium Hexafluoride, ORNL=-3497 (Oct. 15, 1963).

THIS PAGE
WAS INTENTIONALLY
LEFT BLANK
1

ORNL-3544
UC-4 — Chemistry
TID-4500 (25th ed.)

INTERNAL DISTRIBUTION

1. Biology Library 51-52, Sidney Katz
2-4. Central Research Library 53. L. J. King
5. Reactor Division Library 54. C. E. Larson
6-7. ORNL — Y=12 Technical Library 55. R, B. Lindaver
Document Reference Section 56, M. J. Skinner
8-42. Laboratory Records Department 57. S. H. Smiley (K=25)
43. Laboratory Records, ORNL R.C. 58. J. A. Swartout
44, R, E. Blanco 59. A. M. Weinberg
45. G. E. Boyd 60. M. E. Whatley
46. J. C. Bresee 61. P. H. Emmett (consultant)
47. W. H. Carr 62. J. J. Katz (consultant)
48. F. L. Culler 63. T. H. Pigford (consultant)
49. C. E. Guthrie 64, C. E. Winters (consultant)

50. H. L. Hemphill

65.

66.

67.

68,
69.
/0.
71,
72,
73.
74,
75.
76.
77-665,

EXTERNAL DISTRIBUTION

E. L. Anderson, Atomic Energy Commission, Washington, D.C.
H. Schneider, Atomic Energy Commission, Washington, D.C.
H. M. Roth, Atomic Energy Commission, ORO

L. P. Hatch, Brookhaven National Laboratory

G. Strickland, Brookhaven National Laboratory

O. E. Dwyer, Brookhaven National Laboratory

R. H. Wiswall, Brookhaven National Laboratory

R. C. Vogel, Argonne National Laboratory

A. Jonke, Argonne National Laboratory

J. Fischer, Argonne National Laboratory

J. Schmets, CEN, Belgium

Research and Development Division, AEC, ORO

Given distribution as shown in TID-4500 (25th ed.) under Chemistry
category (75 copies — OTY)