This document was prepared in conjunction with work accomplished under Contract No.

DE-AC09-96SR18500 with the U. S. Department of Energy.

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.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from: U.S. Department of Commerce, National Technical
Information Service, 5285 Port Royal Road, Springfield, VA 22161,
phone: (800) 553-6847,
fax: (703) 605-6900
email: orders@ntis.fedworld.gov
online ordering: http://www.ntis.gov/help/index.asp

Available electronically at http://www.osti.gov/bridge

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S.
Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN
37831-0062,
phone: (865)576-8401,
fax: (865)576-5728
email: reports@adonis.osti.gov
 WSRC-TR-2003-00500, Rev. 0

Uranium Metal Dissolution in the

Presence of Fluoride and Boron

KEYWORDS:
H-Canyon
Uranium
Metal
Dissolution
Nitric Acid
Fluoride

Robert A. Pierce
November 2003
 WSRC-TR-2003-00500, Rev. 0
Page 2 of 21

REVIEWS AND APPROVALS

_________________________________ ________________
Robert A. Pierce, Author, SRTC/SMTD/ATS Date:

___________________________________________________ _________________
Michael G. Bronikowski, Technical Reviewer, SRTC/SMTD/ATS Date

___________________________________________________ _________________
William H. Clifton, Jr., Technical Review, Technical Engineering Date

___________________________________________________ _________________
Alice M. Murray, Group Manager, SRTC/SMTD/ATS Date
 WSRC-TR-2003-00500, Rev. 0
Page 3 of 21
SUMMARY

H-Area Operations is planning to process plutonium-contaminated (Pu-contaminated)

uranium metal scrap in its efforts to de-inventory excess nuclear materials. The use of H-
Canyon and HB-Line was evaluated for feasibility in meeting processing targets and
requirements. SRTC performed flowsheet development to support the decision about how
and where the scrap will be processed. Specifically, experimental work was performed to
(1) gather reaction rate data at a range of processing conditions, (2) generate data for
calculating hydrogen and total gas generation rates, (3) propose a process flowsheet, and
(4) demonstrate that the proposed flowsheet does not pose a criticality hazard.

Uranium metal dissolution experiments have shown that acceptable dissolution rates can
be achieved for the Pu-contaminated scrap program using either nitric acid (HNO3)
concentrations above 7M or low HNO3 concentrations (1-4M) in the presence of fluoride
and boron. At low acid concentrations in the absence of fluoride, the reaction rates are
unacceptably slow for the Pu-contaminated scrap program. The observed behavior of
dissolution rates as a function of acid concentration and temperature are in general
agreement with what is expected based on the literature.

Gas generation tests have demonstrated that hydrogen generation is not an issue at the
conditions being proposed for plant operations. At HNO3 concentrations above 2M, the
hydrogen component of the offgas is less that 0.1% by volume. The total amount of gas
generation will be approximately 18.6 mL/hr per square centimeter of exposed metal
surface area.

Mixing studies have shown that criticality is not a likely event in the dissolver insert
either at room temperature or at 100oC. In 2M HNO3/0.025M potassium fluoride (KF)
and 2 g/L boron (B) at room temperature, a steady gas stream is generated from the
surface of uranium metal. The gas generation rate is sufficient to mix the contents of the
dissolver insert. In 4M HNO3/0.025M KF and 2 g/L boron, there is insufficient gas
generation to disperse the dissolved uranium. Instead of mixing, the denser uranium
solution drops down out of the dissolver insert and away from the metal being dissolved.
Variations in acid concentration away from 2M HNO3 should not be a problem since
dissolution in both 1M or 3M HNO3 (with KF and boron) at room temperature produces
gas generation comparable to that of 2M HNO3. When the temperature is raised to
100oC, the gas generation at 1-4M HNO3 in 0.025M KF and 2 g/L boron exceeds that of
2M HNO3 at room temperature, and thus will provide excellent mixing in the dissolver
insert.

Based on reactions at room temperature and boiling, and the rates of reaction at 100oC,
SRTC recommends the use of 2M HNO3/0.025M KF and 2 g/L boron at boiling in H-
Canyon to process the Pu-contaminated scrap material.
 WSRC-TR-2003-00500, Rev. 0
Page 4 of 21
INTRODUCTION

H-Area Operations is planning to process plutonium-contaminated (Pu-contaminated)

uranium metal scrap in its efforts to de-inventory excess nuclear materials. The use of H-
Canyon and HB-Line was evaluated for feasibility in meeting processing targets and
requirements. The Actinide Technology Section of the Savannah River Technology
Center (SRTC) performed flowsheet development to support the decision about how and
where the scrap will be processed.

Experimental work was performed with a piece of uranium sheet. The study had four
primary objectives. First, gather reaction rate data at a range of processing conditions to
compare against reaction rate data reported in earlier studies and recommend a flowsheet
for H-Canyon and/or for HB-Line. Second, develop new data for calculating hydrogen
and total generation rates during uranium metal dissolution. Third, the tests intend to
provide data to help demonstrate that the proposed flowsheet does not pose a criticality
hazard. Last, use the data to recommend a process flowsheet.

EXPERIMENTAL

The experimental program for the Pu-contaminated scrap campaign involved three basic
types of experiments. The first type was the collection of general dissolution rate data as
a function of temperature and solution concentrations. The second type of test was for
gas generation designed to measure offgas volume and concentration as a function of
solution concentrations. The third set determined if sufficient mixing occurs during
dissolution to prevent the creation of conditions that could lead to a criticality event. All
tests were performed with small pieces of enriched uranium metal all cut from a larger
sheet of metal. The sheet has a uniform thickness.

Dissolution Rate Measurements

Initial reaction rates were measured at six process conditions: 2M nitric acid (HNO3), 4M
HNO3, 7M HNO3, 10M HNO3, 2M HNO3/2M NaNO3, and 4M HNO3/3M sodium nitrate
(NaNO3). Subsequent studies evaluated reaction rates for 2M HNO3 and 4M HNO3 in
the presence of 2 g/L boron and 0.01-0.1M potassium fluoride (KF). Several
temperatures were studied for most of the solutions.

Dissolution rates were measured as follows. A small piece of uranium metal plate was
cut from a larger piece. The small piece was then pickled in 10M HNO3 at room
temperature for 3-4 minutes to remove the oxide coating; initial pickling typically
removed about 5% of the weight as oxide. The physical dimensions (width, length,
thickness, mass) of the sample were then measured. A balance was used to measure
weight; calipers were used to measure width, length, and thickness. It is important to
note that the balance in the glovebox measures only to 0.001 grams. Therefore, the
uncertainty on small weight changes can be quite large.
 WSRC-TR-2003-00500, Rev. 0
Page 5 of 21
Next, a beaker was filled with a known amount of dissolver solution and placed on a hot
plate/stirrer. The beaker also contained a sample holder to allow solution access to both
sides of the sample without interfering with the T Watch Glass
stirring (see Figure 1). The solution was heated
to a set temperature while the contents were
stirred. When at temperature, a pickled sample Sample holder w/
was placed in solution for a measured amount of U Metal perforated plate
time. At the end of the test, the metal sample Sample ( 0.070” holes;
42% open area)
was removed, rinsed with water, dried, and
weighed. Reaction rate was determined as
weight loss per unit surface area per unit time.

Gas Generation Tests Hot Plate

Tests were performed to measure both the With Stirrer
generation rates and composition of gases
generated during the dissolution of uranium
metal at different acid concentrations. Of
particular interest is the amount of hydrogen gas
Figure 1. Reaction Rate Setup
generated per mass of metal dissolved. Conditions tested include 10M HNO3, 7M HNO3
and 4M HNO3 without boron or fluoride; in presence of 2 g/L boron and 0.025M KF, 4M
HNO3, 2M HNO3, and 1M HNO3 solutions were tested.

A system was set up as shown in Figure 2. The system was a sealed glass vessel
designed to prevent the introduction or loss of gas. The vessel was filled with 350 mL of
liquid to fill the majority of the 400 mL of
Gas
space in the system. The liquid was heated
o Collect
to 100 C, the dissolution temperature Coolant Bag
selected for these tests. Once at T
temperature, the head space and lines of the
system were purged with an inert gas. For
10M HNO3 and 7M HNO3, nitrogen was Sample
used to purge the system; CO2 was used for U M etal Septum
Sample
the other four test conditions to avoid
interference with nitrogen analyses when
dissolving at lower acid concentrations.

After the system had been purged, a pickled

Hot Plate
and weighed uranium metal sample was With Stirrer
added to the vessel, and the vessel was
sealed. The sample sits on the bottom of
the vessel. As gases are generated, they
were collected in a gas-sample bag fitted Figure 2. Gas Generation Setup
with a valve and sample septum. After the metal sample had dissolved and gas
generation had ceased, 20 mL of gas sample was collected in a syringe and the residual
gas volume was measured using displacement of water.
 WSRC-TR-2003-00500, Rev. 0
Page 6 of 21

It should be noted that the first gas generation test (10M HNO3) was performed with an
inverted burette to capture the gas and measure its volume. However, it was found that
the water in the burette was absorbing the nitrogen dioxide (NO2) from the gas phase
faster than anticipated. As a result, the gas volume could not be measured. Nonetheless,
a gas sample could still be obtained from the dissolution vessel. Also, the gas generation
test at 1M HNO3/ 0.025M KF and 2 g/L boron was initiated at room temperature to
observe the room temperature gas generation and mixing in support of the mixing studies
discussed below.

Mixing Studies

Another system was set up to observe the mixing that occurs during dissolution of
uranium metal in a simulated dissolver insert at room temperature. No external mixing
was provided so that system mixing was solely a function of gas generation due to
reaction. The tests aim to determine whether dissolved uranium concentrated in a
stagnant layer around the piece that was dissolving, hence creating a criticality concern.

A picture of the system used for the mixing studies is provided in Figure 3. The outer
tube (LEFT) represents the dissolver that will contain the bulk dissolver solution. The
simulated dissolver dissolver insert is shown CENTER and RIGHT. The simulated
dissolver insert contains a perforated plate up from the bottom of the dissolver insert,
similar to the actual dissolver insert, to allow liquid to access the sample from above and

Figure 3. Mixing Test Assembly

 WSRC-TR-2003-00500, Rev. 0
Page 7 of 21
below. The sample will sit on the perforated plate as a half circle (RIGHT). The
simulated tube also contain holes above the sample to allow solution to flow between the
dissolver insert and bulk solution in a manner comparable to the actual dissolver insert.

Three tests were performed in the mixing test assembly – 2M HNO3, 3M HNO3, and 4M
HNO3, all in the presence of 2 g/L boron and 0.025M KF. The system was charged with
a pickled, weighed sample of uranium metal. The piece was bent to form a 180o arc
similar to Figure 3. Next, 500 mL of liquid was added to the system so that the liquid
level was above the level of the holes. Observations were then made of gas generation
and uranium partitioning in the system as a function of time. Pictures and videos of the
tests were taken for additional studies of the test results.

The first test was performed with 4M HNO3 and ran for 22 hours. The second test was
run in 2M HNO3 and operated for 5.5 hours. After 3.5 hours in the 2M HNO3 test, a
second metal sample was introduced to develop a better understanding of the reaction and
mixing characteristics. The test with 3M HNO3 was an abbreviated test (3.25 hours) that
was aimed at only observing the gas generation characteristics for comparison with the
first two tests. The 3M HNO3 solution was prepared from 50:50 mixture of used 2M and
4M HNO3 solutions used in the first two tests. In addition to the results obtained from
the above mixing studies, other observations of mixing during other studies were
obtained, as appropriate.

RESULTS

Dissolution Rate Measurements

The first set of tests focused on using only HNO3 at 100-105oC. The data are contained
in Table 1. A general increase of dissolution rate with increasing acid concentration is
observed. Also, the two data sets at 4M total nitrate and the two sets at 7M total nitrate
show good agreement even though the acidity varies within the data sets.

Table 1. Measured Dissolution Rates in Nitric Acid

Rate (mg/min-cm2) at Temperature (oC)
o
Solution 20-25 C 50oC 60oC 80oC 100oC
2M HNO3 --- --- nd* 0.007 0.051
nd = not detected
Rate (mg/min-cm2) at Temperature (oC)
o
Solution 20-25 C 50oC 65oC 85oC 105oC
2M HNO3/2M NaNO3 --- --- --- --- 0.254
4M HNO3 --- --- 0.003 0.041 0.309
4M HNO3/3M NaNO3 --- --- --- --- 1.258
7M HNO3 0.005 --- 0.074 0.841 1.285
10M HNO3 --- --- 1.637 3.496 4.057

During the tests with 2M and 4M HNO3, it was observed that the gas generation would
start and stop throughout the test. Closer inspection of the samples indicated that the
samples had a dull surface to them. Because it was speculated that there was an oxide
 WSRC-TR-2003-00500, Rev. 0
Page 8 of 21
coating on the samples even in the acid, additional experiments were performed with
varying amounts of KF and boron (2 g/L boron added as boric acid). Although there is
an initial oxide coating on the sample from exposure to air, empirical data suggests that a
coating persists (or reforms) in solution at low KF concentration.

The data are listed in Table 2 to show the

effect of varying KF concentrations. Except
where stated, all solutions in Table 2
included boron. After boron was added, an
increased shine on the surface of the samples
was observed with increasing KF
concentration. This shine was observed
when the samples were being weighed after
dissolution. A sample dissolved in 2M
HNO3/0.1M KF is shown in Figure 4. Figure 4. Stripped Uranium Metal
Table 2. Measured Dissolution Rates in Nitric Acid with Fluoride and Boron
Rate (mg/min-cm2) at Temperature (oC)
Solution 20-25oC 50oC 60oC 80oC 100oC
2M HNO3 (no boron) --- --- nd* 0.007 0.051
2M HNO3/0.01M KF 0.099 0.308 --- --- 1.066
2M HNO3/0.025M KF 0.137 0.54 --- --- 2.348
2M HNO3/0.05M KF --- --- --- --- 6.480
2M HNO3/0.1M KF --- --- --- --- 12.574
2M HNO3/0.1M KF 0.213 --- --- --- 18.491
(no boron)
nd = not detected
Rate (mg/min-cm2) at Temperature (oC)
Solution 20-25oC 50oC 65oC 85oC 105oC
4M HNO3 (no boron) --- --- 0.003 0.041 0.309
4M HNO3/0.01M KF 0.097 0.293 --- --- 1.307
4M HNO3/0.025M KF 0.097 0.37 --- --- 3.163
4M HNO3/0.05M KF --- --- --- --- 6.485
4M HNO3/0.1M KF 0.541 --- --- --- 20.464

For one test in 2M HNO3/0.025M KF and 2 g/L boron, the sample was dissolved at room
temperature without being pickled prior to coming in contact with the acid. No stirring
was provided. The dissolution rate at room temperature was 0.323 mg/min-sq.cm over 20
minutes. When compared to the pickled rate of about 0.137 mg/min-sq.cm, the data
indicates that the oxide coating dissolves faster than the base metal. Also, after the 20
minutes of dissolution, the sample no longer had a visible oxide coating; it had a dull
silver color instead of charcoal gray.

Gas Generation Tests

The rate of dissolution for each of the gas generation tests reflected what was observed in
the dissolution rate measurement experiments. The gas volume measurements are made
with the gas at ambient temperature and pressure. In each case, a significant amount of
 WSRC-TR-2003-00500, Rev. 0
Page 9 of 21
orange fumes was observed in the gas sample bag, indicating the presence of NO2. The
raw data are listed in Table 3. Based on the presence of oxygen, it appears as though
some air has leaked into the 7M HNO3 sample and may have affected the NO:NO2 ratio.
All data listed as “<0.1%” in Tables 3 and 4 are below the detection limit.

When the data are normalized to omit the presence of the purge gas and list the
“Undefined” as NO2, the data in Table 4 are obtained. As will be discussed later, there is
reason to expect that some of the “Undefined” gas of 1M HNO3/0.025M KF also contains
a significant fraction of N2O. The total gas generated (in mL/g U) is not affected by the
presence of the purge gas because the total displacement of gas into the sample bag is
solely a function of the amount of gas generated by the reaction.

Table 3. Gas Generation Data

Gas
Solution (mL/g U) H2 (%) N2 (%) O2 (%) CO2 (%) NO (%) Undef (%)*
1M HNO3/0.025M KF 131 4.4 <0.1 <0.1 18 3.2 74
and 2 g/L boron
2M HNO3/0.025M KF 64 0.07 <0.1 <0.1 13 77 10
and 2 g/L boron
4M HNO3/0.025M KF 132 <0.1 <0.1 <0.1 2.1 83 15
and 2 g/L boron
4M HNO3 103 <0.1 <0.1 <0.1 30 32 38
7M HNO3 111 <0.1 80 4.1 <0.1 0.7 15
10M HNO3 n/a <0.1 71 <0.1 <0.1 8.1 21
* Undefined component is likely NO2 and/or N2O

Table 4. Normalized Gas Generation Data

Gas
Solution (mL/g U) H2 (%) O2 (%) NO (%) NO2 (%)
1M HNO3/0.025M KF 131 5.4 <0.1 3.9 90.7*
2M HNO3/0.025M KF 64 0.08 <0.1 88.5 11.5
4M HNO3/0.025M KF 132 <0.1 <0.1 84.8 15.2
4M HNO3 103 <0.1 <0.1 45.7 54.3
7M HNO3 111 <0.1 <0.1 4.4 95.6
10M HNO3 n/a <0.1 <0.1 27.9 72.1
* May contain high fraction of N2O

Mixing Studies

The two primary mixing studies at room temperature (2M and 4M HNO3 with KF and
boron) yielded very different results, particularly in comparison with their behavior at
100oC (in gas generation and dissolution rate tests). At 100oC, both conditions create an
abundance of gas that creates mixing within the vessel. A picture of metal dissolution in
4M HNO3/0.025M KF and 2 g/L boron is shown as Figure 5.

When the mixing study was performed with 2M HNO3/0.025M KF and 2 g/L boron at
25oC, the behavior was consistent with what is observed at 100oC except that the gas
 WSRC-TR-2003-00500, Rev. 0
Page 10 of 21
generation rate is slower. After the sample comes in contact with the acid, it is only a
few minutes before very tiny gas bubbles begin to effervesce from the metal surface.
Because of the presence of fluoride, the uranium first forms a relatively stable blue U(IV)
complex.1 Over time in HNO3, or in the presence of heat, the U(IV) is converted to the
distinctive yellow U(VI) form as seen in Figure 5.

The generation of gas disperses the dissolving

uranium throughout the insert and outer vessel.
After about 90 minutes, the gas generation begins
to stop and start. On a few occasions, the vessel
was bumped and almost immediately the gas
generation was renewed. Late into the test, a
second piece of uranium was added. It was
observed that the two samples tended to react and
stop reacting at essentially the same time. An
estimated dissolution rate for the initial sample
was calculated at 0.108 mg/min-cm2 across the
5.5-hour test. Considering the differences
between the mixing test and the rate measurement Figure 5. Dissolution in 4M HNO3
tests, the rate shows good agreement with the 100oC
0.025M KF/ 2 g/L B at 100
value of Table 2.

Pictures of the test are included as Attachment 1. The pictures at 5, 170, and 240 minutes
clearly depict the abundance of fine gas bubbles being generated from the metal surface
and filling both the dissolver insert and outer vessel. The pictures at 75 and 170 minutes
show that a layer of dissolved uranium does not form as a stagnant layer near the metal
surface. At most, the presence of blue U(IV) is slightly more concentrated at the bottom
of the vessel than it is at the top.

When the mixing study was performed with 4M

HNO3/0.025M KF and 2 g/L boron at room
temperature, the behavior exhibits no resemblance to
what is observed in the tests at 100oC. After the
sample comes in contact with the acid, only a few
tiny bubbles are ever observed at one time. The size
of the bubbles is comparable to the 2M HNO3/
0.025M KF test, but the quantity is drastically
reduced such that they are barely visible to the
naked eye. As the test progresses, the gas
generation remains barely visible.

As with the previous test, the presence of fluoride

causes the uranium to form an initial blue U(IV)
complex. Because of the scarcity of bubbles, the
system is unmixed and the denser uranium solution
sinks. The uranium solution leaves the dissolver Figure 6. Dissolution in 3M HNO3
0.025M KF/ 2 g/L B at 25oC
 WSRC-TR-2003-00500, Rev. 0
Page 11 of 21
insert and is distributed fairly-evenly in the bottom section of the outer vessel. In spite of
the absence of gas generation, an estimated dissolution rate for the sample was calculated
at 0.135 mg/min-cm2 across the 22-hour test. Considering the differences between the
mixing test and the rate measurement tests, the value shows reasonable agreement with
the value of Table 2.

Pictures of the test are included as Attachment 2. All six pictures show the absence of
mixing caused by gas generation. The last five pictures show the gradual increase in
uranium concentration in the bottom of the test unit with the uranium level always
remaining below the level of the uranium metal sample. At the same time, the pictures
clearly indicate that no layer of dissolved uranium forms around the metal sample. The
picture at 1165 minutes depicts the gradual conversion of blue U(IV) to a mixture of
U(IV) and yellow uranyl (U(VI)) nitrate. The solution eventually changes completely to
yellow U(VI).

The brief test using 3M HNO3/0.025M KF and 2 g/L boron shows much resemblance to
the behavior of the 2M HNO3 test. Gas generation occurs as an effervescent stream of
tiny bubbles that mix the contents of the vessel. The gas generation rate is muted because
the solution was a mixture of solution used from the previous two mixing tests and
probably has a reduced level of uncomplexed fluoride. A picture of the gas generation is
shown as Figure 6. It should also be noted that the gas generation test at 1M
HNO3/0.025M KF and 2 g/L boron was initiated at room temperature in order to observe
the room temperature gas generation. Its gas generation characteristics at room
temperature are consistent with both the 2M and 3M tests.

DISCUSSION

Much of the background for this discussion is drawn from an earlier document discussing
the feasibility of dissolving this material in either H-Canyon or HB-Line.2

Chemistry Fundamentals: Several different acid-dependent chemical reactions can be

written for the dissolution of uranium metal in HNO3. Plant experience indicates that the
dissolution reaction is best represented by:3

U + 4HNO3 Æ UO2(NO3)2 + 2NO + 2H2O (1)

It has been proposed that the dissolution of aluminum in HNO3 might provide insight into
the dissolution chemistry of uranium.4

Al + 6HNO3 Æ Al(NO3)3 + 3NO2 + 3H2O (2)

[HNO3] Al + 4HNO3 Æ Al(NO3)3 + NO + 2H2O (3)
Increase 8Al + 30HNO3 Æ 8Al(NO3)3 + 3N2O + 15H2O (4)
10Al + 36HNO3 Æ 10Al(NO3)3 + 3N2 + 18H2O (5)
2Al + 6HNO3 Æ 2Al(NO3)3 + 3H2 (6)
 WSRC-TR-2003-00500, Rev. 0
Page 12 of 21
All of the reactions occur simultaneously during aluminum metal dissolution. However,
at higher HNO3 concentrations, the dissolution shifts towards Eq. (2). At lower acid
concentrations, the dissolution shifts towards Eq. (6). Experience suggests that Eq. (3) is
the dominant reaction in 6-8M nitric acid, which is analogous to Eq. (1).

Offgas Composition and Hydrogen Generation: It is known that the reaction of uranium
metal with HNO3 has the potential to form hydrogen gas. It is also known that the
presence of high concentrations of HNO3 during the reaction sets up competing reactions
that inhibit the formation of hydrogen gas. Lower hydrogen generation rates as a
function of higher HNO3 concentrations has been observed for uranium metal dissolution
and other analogous systems, which is in agreement with the dissolution of aluminum as
a function of HNO3 concentration.

The literature shows that in 55% HNO3 (11.6M) the percent hydrogen generated in the
offgas is 0.01% by volume.5 Similar behavior was observed in the cathode compartment
of equipment employing electrochemical dissolution techniques. In 2M HNO3, nominal
hydrogen generation was 2-3% by volume. In 4M HNO3, the hydrogen rate drops to
nominally 0.4%. In 6M HNO3 and higher, the hydrogen generation rate drops to less
than 0.1% by volume.6 As a result, literature data show that hydrogen evolution is very
low at concentrations above 1M or 2M HNO3.4 Conversely, below 2M HNO3, hydrogen
generation rates may exceed the lower flammability limit (LFL) of 4%.6 The best
explanation of this behavior is given with respect to the chemical reaction for aluminum
dissolution in HNO3 listed above. At lower acid concentrations, the dissolution shifts
towards Eq.(6) where hydrogen generation occurs.

The data in Table 4 show reasonable agreement with the parallel chemistry of aluminum
and earlier data with uranium. As acid concentration decreases, dissolution of uranium
shifts towards Eq. (6), and H2 and nitrogen oxide (NO) concentrations increase.
Hydrogen and NO are products of a more reductive environment when compared to NO2
(Eq. 2). The only exception to the trend is at 7M HNO3. Because of the presence of
oxygen in the sample, it is likely that air in the sample reacted with NO since NO reacts
with air to form NO2.5

Conversely, as acid concentration increases, dissolution shifts towards Eq. (2), and NO2
concentration increases since NO2 is the product of a more oxidative system. The only
exception is at 1M HNO3. The difference is probably due to the definition of unknowns
in Table 3 as purely NO2. As Eq. (4) shows, a decrease in acid might also produce an
increase in N2O. However, N2O is unidentifiable by the gas chromatography method
used and would be grouped with NO2. One flaw in the gas data of Table 4 relative to
Eqs. (2)-(6) is the absence of nitrogen in any of the lower acidity samples. Yet this
statement is predicated on the assumption that uranium dissolution chemistry will
perfectly mirror that of aluminum. That may be a faulty assumption. Other possibilities
are that the acid concentration range for Eq. (5) is very narrow and does not apply to the
conditions tested, or not enough N2 generation occurs to measure N2 reliably.
 WSRC-TR-2003-00500, Rev. 0
Page 13 of 21
It is interesting to note the significant difference between the gas composition in 4M
HNO3 compared to 4M HNO3/0.025M KF and 2 g/L boron. Based on the presence of
U(IV) in the mixing tests (blue color observed in Attachments 1 and 2) instead of the
typical yellow U(IV) (see Figure 5), there seems to be a shift from more oxidative
conditions (4M HNO3) to more reductive ones (4M HNO3/0.025 KF and 2 g/L boron). It
is possible that the shift is caused by participation of KF in the reaction. The more
reductive environment translates into reduced NO2 formation.

What is most important from an operations perspective is the fact that hydrogen
generation in 2M HNO3 and higher HNO3 concentrations is well below the LFL of 4%
for hydrogen. Consequently, as long as acid concentrations are maintained above 1.5M
HNO3, the LFL for H2 will not be exceeded even in the absence of a purge gas stream.
Even in 1.0M HNO3/0.025M KF and 2 g/L boron, the hydrogen concentration is only
marginally above the LFL such that the operation of the vessel sparge systems will
sufficiently dilute the reaction gases to below 4 volume percent.

Calculations indicate that if 25 kg of scrap is dissolved in a 7500-liter batch in H-Canyon,

reactions will reduce the acid concentration by about 0.05M HNO3. If carbon steel (to be
used in the dissolver insert) dissolves in manner analogous to Eq. (3), calculations show
that 47 kg of carbon steel could be dissolved before the HNO3 concentration in the
dissolver is reduced to 1.5M.

Gas Generation Rates: The reaction of uranium metal in nitric acid was expected to be
similar to that of the Pu-Al scrub alloy that was dissolved in 4M HNO3/0.15M calcium
fluoride. The Pu-Al scrub alloy “solids dissolve readily with light, effervescent gas
generation at the material surface.”7 This presumption turned out to be a reasonable
assessment of reaction as can be readily seen from the pictures in Attachment 1.

The amount of gas generation was initially estimated using Equation 1 that assumes two
moles of NO will be generated per mole of uranium. On a gram-of-uranium basis,
assuming no absorption into the solution, the estimated gas generation ratio is 188 mL/g
uranium. Table 4 shows an approximate gas generation rate of 107 mL/g uranium for
dissolution in 4-7M HNO3 and 132 mL/g uranium in 1M and 4M HNO3/0.025M KF and
2 g/L boron. It is likely that the data for 2M HNO3/0.025M KF and 2 g/L boron of 64
mL/g uranium represents a problem in the gas collection setup for that experiment.

Based on the observed gas generation quantities, it appears that some of the gases are
being absorbed before they can be released from the test vessel. In only HNO3, the
percent NO2 is above 50%, and below 15% for those tests with HNO3/0.025M KF and 2
g/L boron. It is known that NO2 can dissolve into water and other aqueous systems.4,5
The occurrence of higher NO2 absorption in HNO3 in the absence of KF (hence, overall
gas generation being lower) can be seen in Table 4.

When a gas generation rate of 130 mL/g uranium is multiplied by the expected
dissolution rate for 2M HNO3/0.025M KF and 2 g/L boron of 2.35 mg uranium/min-cm2,
this yields an expected gas generation rate of 0.31 mL/min-cm2. This value is a small
 WSRC-TR-2003-00500, Rev. 0
Page 14 of 21
amount of gas even under the worst-case conditions being considered for the processing
of the uranium scrap.

Reaction Rates in Nitric Acid: Earlier researchers measured the uranium dissolution rate
at different concentrations of HNO3. At HNO3 concentrations above 2M, the
instantaneous reaction rate at boiling was stated as:

R = (0.0115 mg/min-cm2-M3)* [NO3-]3 (7)

R is in mg/min-cm2 and [NO3-] is the total nitrate concentration in moles per liter.3
Above 2M HNO3 at boiling, the literature shows the instantaneous dissolution rate to be
independent of either the acid or the uranium concentration.3,5 The reaction rates
observed for these experiments are well below what the rates are estimated by Eq. (7). A
comparison of measured and calculated values is contained in Table 5.

Table 5. Experimental and Calculated Dissolution Rates

Expt. Rate Calc. Rate Percent
Solution (mg/min-cm2) (mg/min-cm2) Difference
2M HNO3 0.051 0.092 44.6
2M HNO3/2M NaNO3 0.254 0.736 65.5
4M HNO3 0.309 0.736 58.0
4M HNO3/3M NaNO3 1.258 3.944 68.1
7M HNO3 1.285 3.944 67.4
10M HNO3 4.057 11.500 64.7

Although rates seem to be a function of total nitrate concentration above 2M HNO3, the
difference between all measured and calculated values is very large, typically in the 58-
68% difference range. There is not enough detail in the literature to understand why the
data are different, but two reasons can be speculated. The first is that in the handling of
SRTC samples prior to being added to the test vessel, a thin oxide coating is able to build
up on the surface of the metal which can impede the reaction at the start of the tests, and
perhaps throughout the tests. It is also possible that since the sample is oriented primarily
in the horizontal position (see Figure 5), the presence of gases along the bottom surface
of the samples restricts solution access to the metal surface.

Effect of Fluoride and Boron: The effects of boron and fluoride on uranium metal
dissolution rate were not well documented in the available literature. The only relevant
understanding is analogous work performed with plutonium. To obtain acceptable
dissolution rates of plutonium metal in low concentrations of HNO3 (for H-Canyon or
HB-Line), the use of fluoride is required to dissolve the oxide coating that tends to form
on the metal surface. Without fluoride, the oxide coating remains at the metal surface
and impedes dissolution.7 Boron is frequently added to Savannah River Site canyon
dissolvers as a component of dissolution at 2 g/L to function as a neutron poison. It has
the added effect of complexing some of the available fluoride.8
 WSRC-TR-2003-00500, Rev. 0
Page 15 of 21
Comparable behavior was observed with uranium metal. The data from Table 2 are
plotted in Figure 7 to show the effect of fluoride. The data show that dissolution rate is
limited by the presence of the oxide coating for KF concentrations between 0 and 0.05M
KF. The effect of HNO3 concentration appears negligible until there is sufficient KF
(0.1M) to remove the oxide coating, especially when considering that the dissolution
temperature for the 2M HNO3 tests is only 5oC lower than the 4M HNO3 tests.

The addition of KF to the system accelerates the dissolution rates by removing the oxide
barrier to HNO3 dissolution. In effect, the data indicates that an oxide coating inhibits
HNO3 dissolution at lower KF concentrations. The data also suggest that the oxide
coating persists or continually reforms at the material surface in the presence of 0-0.05M
KF and 2 g/L boron. It is only at higher KF concentrations (>0.1M) that the oxide
coating has been sufficiently removed to see an effect from HNO3 concentration.

This behavior is consistent with both the comparative dissolution rates and empirical data
showing the metal surfaces to be dull when withdrawn from solutions at lower KF
concentrations and shiny when KF concentrations are higher (see Figure 4). Only
samples dissolved in 0.1M KF had a level of shine that suggests the absence oxide.
When a sample was not pickled to remove the oxide coating prior to rate testing, the
dissolution rate increased in comparison to the rate of a pickled sample (0.323 vs. 0.137
mg/min-cm2) because of rapid oxide dissolution. The change in rate should not be
confused with an increase rate in metal dissolution, rather the quick dissolution of oxide
from the sample surface, which is on the order of five weight percent. In addition, the
sample surface color changed from charcoal gray to dull silver. This indicates that
pickling is not required prior to dissolving samples in 2M HNO3/0.025M KF and 2 g/L
boron.

Figure 7. Effect of KF/Boron on Dissolution Rates

Uranium Dissolution in HNO3/KF/B

25

4M HNO3 at
Dissolution Rate (mg/min-cm2)

20 105 deg.C
2M HNO3 at
100 deg.C
15

10

0
0.00 0.02 0.04 0.06 0.08 0.10 0.12
KF Concentration (M)
 WSRC-TR-2003-00500, Rev. 0
Page 16 of 21

Effect of Temperature: The literature shows that increases in temperature produce

increases in dissolution rate. This behavior is as expected. Treatment of the data in the
literature indicates that the effect on reaction rate is essentially linear within the
temperature range of interest.3 This behavior seems questionable since an exponential
type relationship is typically expected.

Two plots of the data are shown in Attachment 3 for the tests containing only HNO3 and
those with HNO3/KF and boron. Since the rate data are plotted in a semi-log manner,
linearity of the data will indicate whether or not there is an exponential type relationship
with respect to temperature. The data plots indicate a high degree of linearity and, as a
result, an exponential relationship with respect to temperature consistent with what
should be expected.

Of the seven data sets, the trendlines for six of the data sets yield an R2 value of 0.97 or
greater, thereby indicating a high degree of linearity. Only one data set, 10M HNO3 (R2
= 0.87), exhibits less linearity. The data of Reference 3 at 10.4M HNO3 possesses a
linearity of 0.93 but has a slope comparable to that of the 10M HNO3 data. The decrease
in linearity in 10M and 10.4M HNO3 may be caused by a high gas generation rate at
105oC, which interferes with the ability of solution to access the surface of the metal.

The data also agree with what was observed in Figure 7. In 0-0.025M KF, the reaction
rate as a function of temperature is independent of whether the HNO3 concentration is
2M or 4M. The lines for 2M or 4M HNO3 in 0.01M KF and 2 g/L boron are essentially
identical; the behavior in 0.025M KF and 2 g/L boron shows little difference between 2M
and 4M HNO3.

Proposed Chemical Reactions:

The above data in conjunction with the literature3,4,5 make it possible to propose a series
of chemical reactions at 100-105oC to describe uranium metal dissolution under varying
acid conditions. The reactions are proposed (not explicitly in the literature except Eqs.
(10)9 and (11)3) to facilitate a better understand of the system chemistry as a function of
acid concentration. The reactions parallel those of Eqs.(2)-(6).

U + 6HNO3 Æ UO2(NO3)2 + 3NO2 + NO + 2H2O (8)

2U + 10HNO3 Æ 2UO2(NO3)2 + 3NO2 + 3NO + 5H2O (9)
[HNO3] U + 4.5HNO3 Æ UO2(NO3)2 + 1.57NO + 0.84NO2 + (10)
Increase 0.005N2O + 0.043N2 + 2.25 H2O
U + 4HNO3 Æ UO2(NO3)2 + 2NO + 2H2O (11)
5U + 14HNO3 Æ 5UO2(NO3)2 + 2N2O + 7H2 (12)
 WSRC-TR-2003-00500, Rev. 0
Page 17 of 21
System Mixing:

The mixing study data reveal the type of mixing behavior that can be expected at room
temperature or 100oC for either 2M or 4M HNO3 with 0.025M KF and 2 g/L boron. At
2M HNO3, dissolution at room temperature generated a visible gas stream of extremely
fine bubbles. The gas stream was able to mix the liquid in the dissolver insert and
prevent the formation of a uranium layer around the metal (Attachment 1). The gas
generation rate for 2M HNO3 at room temperature is only a fraction (~4-6% based on
Table 2) of what it will be at 100oC for either 2M or 4M HNO3 with KF and boron.
Therefore, thorough mixing will also be present in the dissolver insert at 100oC in either
2M or 4M HNO3.

In 4M HNO3/0.025M KF and 2 g/L boron at room temperature, very little gas generation
is observed at first and almost none is visible after about six hours. Nonetheless, uranium
dissolves at a rate comparable to 2M HNO3/0.025M KF and 2 g/L boron. As the
uranium dissolved, the more-dense liquid settled to the bottom of the vessel away from
the sample (Attachment 2). The uranium (blue due to U(IV)) spreads evenly throughout
the solution below the sample even though no mixing was present. Although this shows
that the use of 4M HNO3 in boron and KF does not present a criticality hazard at room
temperature, it produces a less desirable condition than that of 2M HNO3.

Use of an intermediate-acid case (3M HNO3) and a lower-acid case (1M HNO3) both
showed reaction characteristics comparable to the 2M HNO3 test. A steady stream of
small gas bubbles provides adequate mixing in the dissolver insert at room temperature.
Based on the reaction characteristics in conjunction with gas generation studies, 2M
HNO3 with 0.025M KF and 2 g/L boron will be recommended as the preferred solution
for H-Canyon processing.

CONCLUSIONS/RECOMMENDATIONS

Uranium metal dissolution experiments have shown that acceptable dissolution rates can
achieved for the Pu-contaminated scrap program using either high HNO3 concentrations
(>7M) or low HNO3 concentrations (1-4M) in the presence of fluoride (0.025M) and
boron (2 g/L). During dissolution in the presence of little or no fluoride, an oxide coating
forms on the surface of the uranium metal and impedes the reaction rate. At low acid
concentrations in the absence of fluoride, the reaction rates are unacceptably slow for the
Pu-contaminated scrap program. Tests have shown that samples with thick oxide
coatings dissolve readily in 2M HNO3/0.025M KF and 2 g/L boron. The observed
behavior of dissolution rates as a function of acid concentration and temperature are in
general agreement with what is expected based on the literature.

Gas generation tests have demonstrated that hydrogen generation is not an issue at the
conditions being proposed for plant operations. At HNO3 concentrations above 2M, the
hydrogen component of the offgas is less that 0.1% by volume. When the HNO3 is
reduced to 1M in the presence of KF and boron, the hydrogen concentration (corrected)
 WSRC-TR-2003-00500, Rev. 0
Page 18 of 21
rises above the LFL to 5.4% by volume. Sparging and the operation of the dissolver
offgas system will dilute this concentration below the LFL.

Mixing studies have shown that criticality is not a likely event in the dissolver insert
either at room temperature or at 100oC. In 2M HNO3/0.025M KF and 2 g/L boron at
room temperature, a steady gas stream is generated from the surface of uranium metal.
The gas generation rate is sufficient to mix the contents of the dissolver insert. In 4M
HNO3/0.025M KF and 2 g/L boron, minimal gas generation and mixing occurs. Instead
of mixing the uranium, the more-dense uranium solution drops down out of the dissolver
insert and away from the metal being dissolved. Variations in acid concentration away
from 2M HNO3 should not be a problem since dissolution in both 1M or 3M HNO3 (with
KF and boron) at room temperature produces gas generation comparable to that of 2M
HNO3. When the temperature is raised to 100oC, the gas generation at 1-4M HNO3
exceeds that of 2M HNO3 at room temperature and provides excellent mixing in the
dissolver insert.

Based on reactions at room temperature and boiling, and the rates of reaction at 100oC,
SRTC recommends the use of 2M HNO3/0.025M KF and 2 g/L boron at boiling in H-
Canyon to process the Pu-contaminated scrap material.

REFERENCES

1. Katz, J.J, Seaborg, G.T. and Morss, L.R. The Chemistry of the Actinide Elements
(2nd Edition), Chapman and Hall (New York), 1986.
2. R. A. Pierce, “Feasibility of Dissolving Uranium Metal in HB-Line and H-Canyon,”
WSRC-TR-2003-00359, August 2003.
3. Colven, T.J, Cline, C.W. and Wible, A.E. “Processing of Irradiated Natural Uranium
at Savannah River,” DP-500. August 1960.
4. Long, J.T. Engineering for Nuclear Fuel Reprocessing, American Nuclear Society,
1978.
5. Stoller, S.M. and Richards, R.B. Reactor Handbook, Fuel Reprocessing, Volume II
(2nd Edition), Interscience Publishers (New York), 1961.
6. Wheelwright, E.J., Bray, L.A. and Ryan, J.L. “Development of the CEPOD Process
for Dissolving Plutonium Oxide and Leaching Plutonium from Scrap or Wastes,”
PNL-6483. March 1988.
7. Pierce, R.A. “Dissolution of Plutonuim Scrub Alloy and Anode Heel Materials in H-
Canyon,” WSRC-TR-2003-00276. June 2003.
8. D. G. Karraker, et.al. “Flowsheet Modifications for Dissolution of Sand, Slag, and
Crucible Residues in the F-Canyon Dissolvers,” WSRC-TR-97-00395, December
1997.
9. D. L. Foster, J. E. Savolainen and R. G. Wymer. “Nuclear Reactor Fuel Dissolution.”
Proceedings of the International Conference on the Peaceful Uses of Atomic Energy,
Volume 9, Reactor Technology and Chemical Processing, United Nations (New
York, 1956), pp. 546-550.
 WSRC-TR-2003-00500, Rev. 0
Page 19 of 21

ATTACHMENT 1. 2M HNO3/0.025M KF and 2 g/L Boron Mixing Study Pictures

145 290
mins mins

75 240
mins mins

5 170
mins mins
 WSRC-TR-2003-00500, Rev. 0
Page 20 of 21

ATTACHMENT 2. 4M HNO3/0.025M KF and 2 g/L Boron Mixing Study Pictures

1165
175
mins
mins

75 445
mins mins

5 260
mins mins
 WSRC-TR-2003-00500, Rev. 0
Page 21 of 21
ATTACHMENT 3. Uranium Dissolution Rate Temperature Dependence

U Metal Dissolution Rates

100000
4M HNO3
7M HNO3
Dissolution Rate (ug/min-cm2)

10000 10M HNO3

2M HNO3
10.4M HNO3 (ref. 3)
1000

100

10

1
275 295 315 335 355 375 395
Temperature (K)

U Metal Dissolution Rates

10000
Dissolution Rate (ug/min-cm2)

2M HNO3/0.01M KF
2M HNO3/0.025M KF
4M HNO3/0.01M KF
1000 4M HNO3/0.025M KF

100

10
275 295 315 335 355 375 395
Temperature (K)