ArticlePDF Available

Removing Acetic Acid from a UREX+ Waste Stream: A Review of Technologies

Authors:

Abstract and Figures

This study explores different technologies for removing acetic acid from a UREX + waste stream. The waste stream contains both nitric and acetic acids, and the acetic acid must be removed from the waste stream to prevent potential problems in the downstream steps as well as affecting the recycle of nitric acid. The acetic acid is formed after the UREX step of the process as a result of hydrolytic degradation of acetohydroxamic acid used to suppress plutonium extraction. Of the available technologies, the two most attractive approaches are solvent extraction and distillation. In industry, solvent extraction is used for more dilute concentrations of acetic acid while distillation is used for concentrated acetic acid. If a liquid-liquid extraction is viable, this would be the best option with the addition of an extractant, like tributyl phosphate or tri-n-octyl amine, if needed. However, if acetic acid removal can be delayed until the end of the UREX + process when the nitric acid may be concentrated for recycle, distillation may remain an option, though not necessarily a better option than solvent extraction.
Content may be subject to copyright.
REPROCESSING
KEYWORDS: acetic acid removal,
UREX+, separation technologies
REMOVING ACETIC ACID FROM
A UREX+ WASTE STREAM:
A REVIEW OF TECHNOLOGIES
JESSICA A. MITCHELL, R. M. COUNCE, and J. S. WATSON
University of Tennessee, Department of Chemical and Biomolecular Engineering
1512 Middle Drive, Knoxville, Tennessee 37996-2200
B. B. SPENCER and G. D. DEL CUL Oak Ridge National Laboratory
Nuclear Science and Technology Division, 1 Bethel Valley Road
Oak Ridge, Tennessee 37831
Received June 6, 2008
Accepted for Publication August 26, 2008
This study explores different technologies for remov-
ing acetic acid from a UREX
waste stream. The waste
stream contains both nitric and acetic acids, and the
acetic acid must be removed from the waste stream to
prevent potential problems in the downstream steps as
well as affecting the recycle of nitric acid. The acetic
acid is formed after the UREX step of the process as a
result of hydrolytic degradation of acetohydroxamic acid
used to suppress plutonium extraction. Of the available
technologies, the two most attractive approaches are sol-
vent extraction and distillation. In industry, solvent ex-
traction is used for more dilute concentrations of acetic
acid while distillation is used for concentrated acetic
acid. If a liquid-liquid extraction is viable, this would be
the best option with the addition of an extractant, like
tributyl phosphate or tri-n-octyl amine, if needed. How-
ever, if acetic acid removal can be delayed until the end
of the UREX
process when the nitric acid may be con-
centrated for recycle, distillation may remain an option,
though not necessarily a better option than solvent
extraction.
I. INTRODUCTION
The objective of this study is to identify and evaluate
the most attractive approach to removing acetic acid from
the UREXprocess. The evaluation includes consider-
ation of acceptability of the approach in nuclear material
processing, the effects of the approach on the UREX
process and downstream process steps, as well as factors
that affect costs of the approach. Technologies including
solvent extraction, destruction, absorption, distillation,
and crystallization for the removal0destruction of acetic
acid are studied, based on available literature. The ob-
jectives of this study are ~a!to determine the most ap-
propriate method for effective removal of acetic acid
without removal of other key components in the process,
and ~b!the removal step must not interfere with other
downstream steps or the recycling of nitric acid. Once
chosen, this method will be experimentally verified at
UREXprocess conditions. The degree0percent of re-
moval remains a variable in this study since no specific
limit for residual acetic acid was available.
II. BACKGROUND
The UREXprocess, as presented in Fig. 1, is a
series of solvent extraction steps designed to treat spent
nuclear fuel by separating its various components for
reuse and disposal. Since this process is still under de-
velopment, there are many different flow sheet scenarios.
The flow sheet presented in Fig. 1 is used in this study.
The first step in Fig. 1 is the UREX step; nuclear
fuel dissolved in aqueous nitric acid is treated with an
organic solvent to remove uranium and technetium.
The solvent consists of tributyl phosphate ~TBP!dis-
solved in n-dodecane with acetohydroxamic acid ~AHA!
added to the aqueous stream to prevent the extraction of
plutonium. Both uranium and technetium are extracted
into this solvent.
1
The downstream steps of this process,
CCD-PEG ~or FPEX!, NPEX, TRUEX, and TALSPEAK,
*E-mail: Jmitch30@utk.edu
360 NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009
are discussed at length in papers on the UREX
process.
1– 4
The source of the acetic acid, which is the focus of
this study, is a degradation product of AHA. In the acid
environment, most of the AHA hydrolyzes to produce
acetic acid and hydroxylamine nitrate. Acetic acid and
residual AHAare assumed to leave the UREX segment of
the process in the raffinate. The complexant will be
completely destroyed prior to the plutonium removal
step ~NPEX!so that the plutonium will no longer be
suppressed. The acetic acid must be separated and0or
destroyed because it has potential to interfere with down-
stream steps and with the recycling of the nitric acid for
reuse in the UREXprocess.
3
The raffinate stream of interest was modeled as an
aqueous mixture of nitric acid and acetic acid in an ap-
proximate 10:1 concentration ratio ~0.5 M nitric acid and
0.05 M acetic acid!. This is believed to be close to the
acid concentrations in the raffinate streams, but, of course,
the concentrations will depend upon where the acetic
acid removal step is placed in the UREXprocess. There
will also be numerous salts in the raffinate stream, but
those are assumed to be at low concentrations and are not
expected to affect the acetic acid removal or destruction
step. However, the behavior of those salts during the
separation and destruction will be an important consid-
eration throughout this analysis of removal options since
many of these salts ~a!are highly radioactive, ~b!are
desired products to be recovered in downstream steps, or
~c!may affect the disposal of any wastes from the acetic
aicd removal step.
Relevant physical properties for acetic acid, nitric
acid, and water are shown in Table I. Additionally, nitric
acid and water form a maximum boiling azeotrope at
;68 wt% nitric acid. Section III summarizes the result of
the literature study of potential acetic acid removal
methods.
III. ACETIC ACID REMOVAL METHODS
III.A. Membrane Separation and Ion Exchange
Technologies such as membrane separation and ion
exchange were eliminated early in the literature review.
No membrane was found to be sufficiently selective to
remove acetic acid in a single pass, and multistage mem-
brane operation was not thought to be an attractive op-
tion. Also, organic membranes are not normally used in
high radiation fields because of potential radiation deg-
radation, so the membrane selection may be limited to
inorganic membranes.
Ion exchange was also eliminated as a possible tech-
nology because no ion exchange material was identified
with sufficient selectivity for acetate ions over nitrate
ions at conditions of this study. Ion exchange occurs
primarily through adsorption onto a resin with exchange-
able ions, anions in this case. For the raffinate stream of
interest, nitric acid depresses the dissociation of the ace-
tic acid, i.e., it lowers the acetic acid uptake and reduces
the selectivity of anion resins for acetate ions.
III.B. Crystallization
Crystallization is the formation of a solid phase from
a homogeneous liquid phase. It is of possible interest
because of the high freezing point of acetic acid com-
pared to water. Crystallization first requires a saturated
solution so that further changes in solution conditions
cause solid formation ~crystallization!. Currently, solu-
bility data are available only for binary systems, water–
Fig. 1. UREXprocess overview.
TABLE I
Physical Properties
Properties
Acetic Acid
~CH
3
COOH!
Nitric
Acid
~HNO
3
!
Water
~H
2
O!
Boiling point ~8C!118 122 100
Freezing point ~8C!16.6 42 0
Vapor pressure
~mm Hg at 208C!11 48 17.5
pKa 4.8 1.5 15.74
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009 361
nitric acid, water–acetic acid, and nitric acid–acetic acid.
Information on ternary systems of nitric acid–acetic acid–
water is not available, so evaluations were made on three
binary systems.
The acetic acid-water phase diagram ~Fig. 2!shows
for dilute acetic acid solutions ~to the left of the eutectic!
the crystallizing species is water. Substantial concentra-
tion of the aqueous acetic acid–water system must occur
before acetic acid becomes the crystallizing species ~to
the right of eutectic!. The freezing points for nitric acid
and water are quite a bit lower than for acetic acid and
water ~Fig. 3!. Since the solution of interest contains
substantial quantities of nitric acid and nitrate salts in
addition to acetic acid and water, the presence of this
species leads to additional changes in the freezing point.
In general, the addition of various salts as well as nitric
acid to aqueous acetic acid solutions results in an in-
crease in activity for the acetic acid and a decrease in
water activity.
7,8
Applying such information, the freezing
point curves will likely shift some for the solution of
interest from that shown for the aqueous acetic acid bi-
nary system. Other pertinent information comes from
Linke and Seidell,
6
who indicate for the acetic acid–
nitric acid binary system there is a nitric acid–acetic acid
species with an estimated freezing point of about 23.98C.
While this does not rule out the possibility of a useful
acetic acid–nitric acid compound for crystallization, it
indicates that considerable concentration of the acetic
acid and nitric acid is likely to be necessary, and the
eventual solid phase that removes acetic acid is likely to
contain some nitric acid. Also, the behavior of dissolved
salts during crystallization may be important if crystal-
lization is used for acetic acid removal. Multistage crys-
tallization equipment was noted to usually involve
considerable mechanical equipment for washing and re-
dissolving the crystals.
The primary conclusion from Figs. 2, 3, and 4 is that
considerable concentration of acetic acid appears to be
necessary before crystallization can be a viable candidate
technology for acetic acid removal. Thus crystallization
was not considered further in this study. It is noted, how-
ever, that in activities where nitric acid is concentrated
for recycle crystallization of acetic acid may again be of
interest.
III.C. Distillation
Acetic acid and water have been separated industri-
ally by simple, azeotropic, and extractive distillation.
9,10
The purpose of these industrial uses is to concentrate the
acetic acid starting with concentrations higher than 5
wt% of acetic acid not present in the UREXprocess. At
these concentrations, the most volatile component is water;
so distillation would initially remove water and concen-
trate the nitric and acetic acids. If distillation is employed
for another purpose like nitric acid concentration in the
recycle stream, then it may be useful to consider this
option more seriously for acetic acid removal. For this
reason, further discussion of distillation is needed.
Fig. 2. Phase diagram for acetic acid and water.
5
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
362 NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009
Acetic acid gas-liquid distribution coefficients at 13.3
kbar for the three-component system of nitric acid, water,
and acetic acid from Nagahama and Jiang
7
vary from
relative volatilities ~K
Acetic
0K
water
!of 0.594 at a temper-
ature of 329.6 K to 0.975 at a temperature of 343.8 K.
Using the UREX model stream of 0.5 M nitric acid with
0.05 M acetic acid, there is very little change in the vapor-
liquid equilibrium data from the binary acetic acid–water
Fig. 3. Phase diagram for nitric acid and water.
6
Fig. 4. Phase diagram for nitric acid and acetic acid.
6
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009 363
system to the ternary acetic acid–nitric acid–water sys-
tem.
7
Thus, distillation would continue to be attractive
for higher concentrations of acetic acid, but not so attrac-
tive for low concentrations.
III.D. Adsorption
Adsorption involves the transfer of a solute from a
fluid phase to a solid surface where it is bound by inter-
molecular forces. The solute being concentrated on a
surface is defined as the adsorbate, and the material on
which the adsorbate accumulates is defined as the adsor-
bent. The amount of material that can be accumulated on
a unit area of a surface is small, so useful adsorbents are
typically porous material such as activated carbon or other
materials with large internal surface areas. The term ad-
sorption as used in this review includes any form of bulk
uptake by solids, but only one term is used for the sake of
simplicity. The design of adsorption equipment requires
selection of an adsorbent and information on the equilib-
rium loading of the adsorbate on the adsorbent ~the iso-
therm!, the rate of transport of the adsorbate to the surface
during adsorption and away from the surface during re-
generation, and equilibrium loading of the adsorbate under
regeneration conditions. As part of the process for selec-
tion of the adsorbent, the following information is nec-
essary on the characteristics of the adsorbent: ~a!the
equilibrium capacity of the adsorbent, ~b!the selectivity
of the adsorbent, ~c!the physical and chemical charac-
teristics of the adsorbent, and ~d!the regeneration char-
acteristic of the adsorbent.
Generally, the adsorption of weak electrolytes from
aqueous solutions occurs through the association of the
undissociated molecule with the hydrophobic surface. At
conditions where the pH is lower than the pKa of the
ionizable solute, the equilibrium loading will be greater
than that expected at conditions where the pH is higher
than the pKa. Acetic acid in the stream of interest will be
largely associated, not ionized, at the expected pH, due to
the presence of nitric acid, and the pKa of the acetic acid
~see Table I!thereby favors the adsorption of acetic acid
in the conditions of this study.
Equilibrium adsorption capacity for acetic acid on
various carbons and polymeric adsorbents is widely re-
ported in the literature. In a study utilizing activated car-
bon with a surface area of 1080 m
2
0g, it was determined
that the equilibrium capacity for a 0.0333 molar aqueous
acetic acid is 0.081 g of acetic acid per gram of carbon.
11
In a similar study of commercially available activated
carbons with specific surface areas ranging from 390 to
2350 m
2
0g, capacities for 1 wt% aqueous solutions of
acetic acid were found to be 0.05 to 0.18 g of acetic acid
per gram of carbon.
12
Several different mathematical forms
for expressing the equilibrium data have been used such
as the Langmuir isotherm,
13
Frueundlich-type isotherms,
14
and the Radke0Praunsnitz-type isotherm.
15
The term “polymeric adsorbents” is used in this re-
port for synthetic organic adsorbents without functional
groups. Adsorption onto polymeric adsorbents without
functional groups is a surface-based phenomenon similar
to that of activated carbon; surface areas of 400 to
1000 m
2
0g are common. The uptake mechanism shifts
from surface-based phenomenon to that of bulk uptake at
surface areas of ;500 m
2
0g. Kuo et al.
12
found equilib-
rium capacities for 1 wt% acetic acid of up to 0.12 g of
acetic acid per gram of adsorbent for several commercial
nonfunctionalized adsorbents; this is somewhat lower than
similar equilibrium capacities of activated carbon but is
comparable when expressed on the basis of specific sur-
face area.
Nitric acid is not likely to be adsorbed to a signifi-
cant extent on activated carbon or other nonfunctional-
ized adsorbents. The possibility of nitration reactions of
nitric acid with polymeric adsorbents deserves careful
investigation. For activated carbon adsorbents, nitration
of carbon could occur after many cycles and possibly
lead to the formation of compounds, which not only would
be unfavorable but could also be a safety concern. The
low nitric acid concentration is a favorable factor in re-
ducing the likelihood of significant nitration. Neverthe-
less, it is likely to be desirable to restrict the useful life of
any carbon- or polymer-based adsorbent to reduce the
potential for accumulation of excessive nitration products.
The usual methods of adsorbent regeneration in-
clude stripping at a higher temperature ~usually with
steam!, desorption with a reactive solution ~such as an
aqueous base!, or leaching with an appropriate solvent
~such as acetone, various acetates, and methanol!. The
commercial nonfunctionalized adsorbents are generally
more easily regenerated by solvents than activate car-
bon.
16
Since acetic acid will be adsorbed as the un-
ionized molecule, it may be possible in this case to strip
with either a dilute basic ~caustic!solution that would
ionize the acetic acid or an organic solvent that has fa-
vorable acetic acid solubility.
III.E. Solvent Extraction
The removal of a solute from a liquid solution using
another immiscible liquid is referred to as liquid-liquid
or solvent extraction. Most investigations of solvent ex-
traction express the degree to which a solute is extracted
in terms of the distribution coefficient of the solute be-
tween the two liquids. The distribution coefficient is de-
fined as
K
D
@Solute#
Organic
@Solute#
Aqueous
,~1!
where @Solute#
Organic
and @Solute#
Aqueous
are equilibrium
concentrations of solute in the organic and aqueous phases,
respectively. Coefficients of this type are a strong func-
tion of the degree of ionization of the solute. Nonionized
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
364 NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009
solutes tend to produce higher distribution coefficients in
solvents with nonionized extractants ~nonion exchange
extractants!than in ionizing solutes since the extractant
removes the neutral solute molecule rather than an indi-
vidual ion, as in ion exchange.
A great deal of research has been done into the ex-
traction of dilute acetic acid from aqueous waste or prod-
uct streams. In industrial processes, the preferred method
of removal for dilute concentrations of acetic acid is sol-
vent extraction.
9
In many cases, the aqueous waste or
product streams are contacted with an organic diluent to
extract acetic acid. In some cases an extractant must be
added to the diluent to aid in the extraction. These ex-
tractants form weak complexes with the solute to be ex-
tracted that are insoluble in water but are soluble in the
organic phase.
17
Wardell and King
18
determined the dis-
tribution coefficients displayed in Tables II and III for
several types of organic solvents. Distribution coeffi-
cients were measured for acetic acid-water solutions only
~no nitric acid present!. Wardell and King
18
state that the
equilibrium distribution coefficients of phosphoryl com-
pounds as extractants in diluents follow the following
trend: phosphates ,phosphonate ,phosphine oxide.
Data presented by Wardell and King
18
examine the rela-
tionship between basicity, extractant concentration, and
molecular weight for the distribution of carboxylic acids
in various solvent systems. Some of the data indicate
acetic acid distribution coefficients for trioctylamines as
high as 9.9, while the distribution coefficient of TBP was
found to be typically 2.3 at 100% TBP with no diluent.
18
The degree of ionization for aqueous acetic acid so-
lutions can be estimated as follows:
K
c
1.77 10
5
10
~pK
a
!
@H
#@Ac
#
@HAc#~2a!
or
@Ac
#
@HAc#
1.77 10
5
@H
#.~2b!
Thus, using a concentration of ;0.05 M acetic acid, the
percent ionized ~degree of ionization multiplied by 100!
is 1.86% in a solution of only water and acetic acid. With
the addition of 0.5 M nitric acid, the percent ionized
decreased further. This also relates to the distribution
coefficient since with nitric acid present, most acetic acid
will be molecular and therefore more easily extracted
into the organic phase. This does not mean that the in-
crease in the distribution coefficient to the organic will
be directly related to the fraction of acetic acid to the total
acetic acid concentration, but it does give some insight
into the distribution of acetic acid in aqueous solutions of
interest. Apossible contradicting effect is the presence of
a mineral acid, which, if strong enough, could protonate
the acetic acid and cause it to be charged, therefore sup-
pressing extraction.
20
In discussing potential solvent extraction options for
this application, solvent options will be divided into two
main groups: solvents that are partially miscible in aque-
ous solutions and solvents that are essentially immiscible
in water. The first group includes the De Dietrich pro-
cess, which uses solvents such as ethyl acetate or methyl
isobutyl ketone to remove acetic acid during manufac-
ture of pharmaceutical products or cellulose acetate. Al-
though this is a relatively mature technology, it is not
likely to be attractive for removing acetic acid from nu-
clear spent fuel reprocessing streams because of the mis-
cibility of the solvent, which would leave some residual
solvent after the acetate is removed. Thus, an additional
process step is likely to be required to remove the re-
maining solvent. With a relatively volatile solvent, the
additional removal step may be a reasonable choice, but
in general, the use of miscible solvents makes solvent
extraction much less advantageous.
Immiscible solvents usually include a diluent and an
active extractant, much like the systems used in fuel pro-
cessing to remove actinides, fission products, and other
metals. The diluent is likely to be a hydrocarbon such as
kerosene or dodecane, and a variety of extractants can be
used. The extractant, as well as the diluent, should be
TABLE II
Ranges of Equilibrium Distribution Coefficients
for Dilute Acetic Acid between Classes of
Organic Solvents and Water
18
Solvents
Range of
Distribution Coefficients
Ethers ~C
4
through C
8
!0.63 to 0.14
Acetates ~C
4
through C
10
!0.89 to 0.17
Ketones ~C
4
through C
10
!1.20 to 0.61
Alcohols ~C
4
through C
8
!1.68 to 0.64
TABLE III
Equilibrium Distribution Coefficients of Acetic Acid
between Diluents and Water
18
Diluents
Distribution
Coefficient
Chevron Solvent 25
a
0.009
Chloroform 0.028
n-Hexanol 0.88
n-Heptanol0n-Hexanol ~2:1!0.30
Nitrobenzene 0.06
n-Heptane 0.02
a
Chevron Solvent 25 is mostly C-8 and C-9 alkylated
aromatics.
19
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009 365
essentially insoluble in the aqueous phase. At least three
extractants have been tested for removing acetic acid
from aqueous phase, trioctylphosphine oxide ~TOPO!,
tri-n-octylamine ~TnOA!, and TBP.The diluent-only case
was also considered.
The distribution coefficient using TBP is of particu-
lar interest for this investigation because TBP is already
used as a solvent during the UREX, NPEX, and TRUEX
segments of the UREXprocess. This makes it a leading
candidate for use as an extractant with an appropriate
organic diluent. TBP is a highly polar compound with its
phosphoryl group acting as a Lewis base. This allows
TBP to create an acid-base complex with acetic acid
and gives a high equilibrium distribution coefficient.
20
The system used in the UREXprocess is a 30% TBP-
dodecane organic solvent with acetic acid–nitric acid–
water. Data have already been measured for 30% TBP-
dodecane in a 1:1 organic to aqueous ratio with acetic
acid–water. These data are shown in Fig. 5, which shows
that as the initial concentration of the acid increases, the
distribution coefficient decreases.
21
When a strong min-
eral acid is added to an acetic acid–water–TBP system,
like nitric acid in this case, the acetic acid may be pro-
tonated and may be in an ionic state that will be difficult
to extract to the organic stream.
20
Much like TBP, TnOA will also have the same pro-
tonation result with the addition of a strong mineral acid
to an acetic acid–water mixture. The distribution coeffi-
cient may also decrease because of competition of the
strong mineral acid for TnOA since it has a higher affin-
ity for the TnOA than its weaker acetic acid counterpart.
TOPO can be used in the extraction of acetic acid.
TOPO’s ability to extract acetic acid at various concen-
trations is shown in Fig. 6. TOPO has been proposed as
a possible extractant to be used in nuclear processes;
however, the UREXprocesses do not include this sol-
vent. Also, TOPO is a very expensive extractant when
compared to TBP or TnOA. Therefore, TOPO will be
used only if these other extractants prove ineffective.
Another option that was considered was the diluent-
only case. That is, the use of the diluent used throughout
the fuel process system without the use of an extractant.
Although something of a long shot, this option can be
evaluated experimentally very quickly. If one can use the
same diluent as elsewhere in the fuel processing system,
there will be minimal risk of undesirable effects from
entrainment of diluent through the nitric acid product
into other process steps. Also, the diluent would not be
likely to extract any metal ions ~fission products, acti-
nides, etc.!, so the acetic acid should be relatively free of
radioactivity.The principal path for radioactivity to reach
the acetic acid probably would be via entrainment. Al-
though Table III shows that Chevron Solvent 25 does not
appear to extract acetic acid alone, it possibly will extract
some additional acetic acid in highly acidic solutions
where a substantial portion of the acetic acid will not be
ionized. There is also the rather remote possibility that
other diluents such as dodecane will give somewhat higher
distribution coefficients. However, the diluent-only op-
tion will be attractive only if the distribution coefficient
for dodecane or a similar satisfactory diluent is much
higher than that reported for Chevron Solvent 25. There
is no hard limit for the value of the distribution coeffi-
cient that would be necessary for the diluent-only option
to be viable, but we would like for the distribution coef-
ficient to be 0.1 or larger. Smaller distribution coeffi-
cients require increasingly larger solvent flow rates and
make operation of some liquid-liquid contactors diffi-
cult. Even if extraction could be practical with the diluent-
only option, it would also be necessary to strip the acetic
acid. The preferential approach would be to strip with
water alone since the pH of the strip could be close to
neutral, with hydrogen ion concentrations approaching
values as low as 10
7
molar. However, if one wanted to
improve the strip and0or concentrate the acetate, a slightly
alkaline strip could be used.
The best option for solvent extraction would be to
use an immiscible solvent. If the diluent-only case can
Fig. 5. K
D
as a function of acetic acid concentration using 30%
TBP-dodecane solvent with an organic-to-aqueous ratio
of1~Ref. 21!.
Fig. 6. K
D
as a function of acetic acid weight percent using
TOPO ~22 wt%!in Chevron Solvent 25 ~Ref. 17!.
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
366 NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009
yield sufficient extraction, it would be the most attractive
approach. If not, the addition of TBP to any diluent would
be second choice. The use of a TBP-dodecane system
would be most favorable for this choice because of its
expected use in UREXprocesses. For any choice, the
stripping of the acetic acid once it has been extracted into
the organic will be needed.
III.F. Destruction
This study assumed that there are limitations on the
nature of reagents that can be added to the UREXpro-
cess. So, there are limitations on how the acetic acid in
the raffinate can be destroyed. Some methods thought to
be potentially acceptable are supercritical water oxida-
tion, wet-air oxidation, using a corona discharge reactor,
and using hydrogen peroxide. Other oxidation methods
require adding a salt or other reagent that leaves a residue
in the solution, and those residues may complicate down-
stream process steps and hinder reuse of the nitric acid,
and0or could increase the amount of mass in the waste
streams.
Using supercritical water oxidation, under high tem-
perature and pressure, the water molecule will become a
nonpolar solvent. This makes it a very good environment
for oxidizing organic compounds. The process requires
oxygen or hydrogen peroxide at temperatures higher than
647.3 K and pressures higher than 22.12 MPa to maintain
supercritical conditions. Either of these two oxidizing
agents can completely break down the acetic acid into
carbon dioxide and water. This process occurs in four
main steps: ~a!raising the pressure of the oxidizing agents,
~b!the reaction, ~c!separating the salt, and ~d!depres-
surization and heat recovery.
22
While this method is ex-
pected to efficiently destroy acetic acid, the required high
temperature and pressure conditions seem too extreme
for a unit operation processing radioactive materials as
part of the UREXprocess.
Consideration of wet-air oxidation to destroy acetic
acid results in similar concerns as supercritical oxida-
tion. A high temperature and a high pressure—though
not as high as supercritical conditions—are required. This
is so that the oxygen molecules are sufficiently reactive
to interact with small carboxylic acids like acetic acid.
The small size of the acetic acid molecule is unfavorable
for oxidation under normal conditions. Also, wet-air ox-
idation is less efficient than supercritical oxidation at
destroying acetic acid. A favorable approach to oxidize
small acetic acid molecules may be to add a metal cata-
lyst. Even so, there is no guarantee for complete oxida-
tion of acetic acid unless the temperature is raised to
2008C or higher.
23
Hydrogen peroxide is a strong oxidant, and a mix-
ture of ozone and hydrogen peroxide makes even stronger
hydroxyl radicals. Hydroxyl radicals create acetate rad-
icals from acetic acid. These radicals will be very active
and will oxidize effectively to carbon dioxide and water.
However, high temperatures—unfavorable to the UREX
process—and a low pH are needed for this process to be
able to operate optimally.
24
Another possible destruction method is a corona dis-
charge reactor. With this method, oxygen radicals can be
produced from three different electron energy levels. Using
this process with oxygen radicals and raising the pH will
degrade the acetic acid to 75 ppm from 100 ppm at pH 1,
with the largest effect at pH 14 where it is degraded down
to 20 ppm ~Ref. 25!. A base would have to be added to
raise the pH to 14. Obviously, a higher pH would require
adding sodium ions or other neutralizing reagent, affect-
ing downstream processing and making nitric acid reuse
impractical. The higher pH is also undesirable because
plutonium is reduced at higher pH and may inappropri-
ately precipitate from solution. Therefore, the corona dis-
charge reactor is a nonviable option for acetic acid
destruction in the UREXprocess.
Upon considering each of these destruction technol-
ogies, three constraints are considered: ~a!nitric acid not
being destroyed, ~b!no new chemicals being added, and
~c!no extreme equipment specifications for radioactive
removal. It is desirable to recover and recycle nitric acid,
so technology which destroys nitric acid or does not se-
lectively destroy acetic acid over nitric acid is rejected.
Likewise, no new chemicals are to be added without
careful consideration. Also, the size of the equipment and
pressure requirements of supercritical and wet-air oxida-
tion tend to cause these options to be severely dis-
counted. No appropriate destruction technology is
identified.
IV. RESULTS
Each of the technologies discussed have some po-
tential for the UREXprocess, but some prove to be
more favorable than others. The potential advantages and
disadvantages for these technologies are summarized in
Table IV. Crystallization is a simple process but requires
a preliminary concentration step to create conditions where
the acetic acid will crystallize from solution; this tech-
nology would be considered if one needed to remove
acetic acid from the recovered and concentrated nitric
acid solution prior to its recycle. Distillation does not
seem to offer any advantage for acetic acid removal due
to the dilute concentration used unless coupled with crys-
tallization or another unit operation. Adsorption is a tech-
nically feasible option with similar concerns for additional
process complexity as those for solvent extraction. Sol-
vent extraction appears to be technically feasible, is a
similar technology to that used elsewhere in the UREX
system, and is a preferred method of removal for dilute
acetic acid–water systems in industry; regeneration of
the solvent and removal of the acetic acid adds complex-
ity to the process, however. No destruction technology
for acetic acid in the UREXprocess was identified.
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009 367
Overall, solvent extraction, probably using TBP ex-
tractant in dodecane ~materials already used in UREX!,
appears to be the most promising approach to acetic acid
removal. Distillation was not considered superior to sol-
vent extraction except for higher concentrations of acetic
acid than those expected in UREX. However, if acetic
acid removal can be delayed until the end of the UREX
process when the nitric acid may be concentrated for
recycle, distillation may remain an option, though not
necessarily a better option than solvent extraction. Dis-
tillation to remove acetic acid selectively from nitric acid
would be more complex than distillation simply to con-
centrate the nitric acid. As noted earlier, this would be a
ternary distillation system that is expected to involve an
acetic acid–nitric acid azeotrope.
V. CONCLUSIONS
The primary method that seems the most feasible is
solvent extraction, and the next efforts should focus on
solvent extraction options. With no assurance that a prac-
tical solvent extraction technology can be found, other
options should be considered if in the future solvent
extraction proves to be less attractive. These alternate
options are distillation, a combined distillation and crys-
tallization, or carbon- or polymer-based adsorption.
By careful selection of a solvent, acetic acid may be
preferentially extracted from the aqueous phase. With the
selection of the most promising technologies, further in-
vestigation through literature research and experimenta-
tion will lead to the most appropriate technology for the
UREXprocess. The studies will need to cover the be-
havior of selected metal ions ~metals used to simulate the
fission products and actinides in the UREX raffinate!as
well as acetic acid.
ACKNOWLEDGMENT
This work was supported by the U.S. Department of
Energy’s Nuclear Energy Research Initiative program, under
DOE contract DE-PS07-05ID14713 with Oak Ridge National
Laboratory.
REFERENCES
1. G. F. VANDEGRIFT, M. C. REGALBUTO, S. B. AASE,
H. A. ARAFAT, A. J. BAKEL, D. L. BOWERS, J. P. BYRNES,
M. A. CLARK, J. W. EMERY, J. R. FALKENBERG, A. V.
GELIS, L. D. HAFENRICHTER, R. A. LEONARD, C.
PEREIRA, K. J. QUIGLEY, Y. TSAI, M. H. VANDER POL,
and J. J. LAIDLER, “Lab-Scale Demonstration of the UREX
Process,” presented at 227th ACS National Meeting, Anaheim,
California, March 28–April 1, 2004, American Chemical So-
ciety ~2004!.
2. C. L. RIDDLE, J. D. BAKER, J. D. LAW, C.A. MCGRATH,
and D. H. MEIKRANTZ, Solvent Extr. Ion Exch.,23, 449
~2005!.
3. G. F. VANDEGRIFT, Personal Communication ~2007!.
4. E. D. COLLINS, D. E. BENKER, L. K. FELKER, R. D.
TAYLOR, G. D. DEL CUL, B. B. SPENCER, W. D. BOND,
and D. O. CAMPBELL, “Hot Test Evaluation of the Reverse
Talspeak Process,” presented at Global 2005 Int. Conf.,Tsukuba,
Japan, October 9–13, 2005.
TABLE IV
Technology Comparisons
Technology Pros Cons
Crystallization • Used with distillation
• Feasible
• Favors acetic acid over nitric acid
• Requires concentration
• Some mechanical operation needed
Distillation • Feasible
• Used with crystallization
• Removes water first, which concentrates nitric acid
• Multiple towers required to remove acetic acid
Adsorption0ion exchange • Uncertain feasibility • No highly attractive adsorbent found
Solvent extraction • Similar to other UREXoperations
• Could use extractant and diluent
already in process
• Data exist on similar systems
• Restricted to solvents in UREX
• Strip step required
• Possibly wash and0or solvent scrub step required
Destruction • Could possibly eliminate acetic acid
• Creation of hydroxyl radicals will
completely break down acetic acid
• Mainly need high temperature and pressure
• High pH needed for corona discharge
• Could also destroy nitric acid
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
368 NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009
5. R. S. BARR and D. M. T. NEWSHAM, Chem. Eng. J.,33,
79 ~1986!.
6. W. F. LINKE and A. SEIDELL, Solubilities: Inorganic
and Organic Compounds, 4th ed., pp. 1128–1132, Van Nos-
trand Reinhold Company, Washington, D.C. ~1965!.
7. K. NAGAHAMA and T. X. JIANG, Fluid Phase Equilib-
ria,53, 261 ~1989!.
8. D. J. SHAH and K. K. TIWARI, J. Sep. Process Technol.,
2,1~1981!.
9. H. CHEUNG, R. S. TANKE, and G. P. TORRENCE, Ul-
lmann’s Encyclopedia of Industrial Chemistry, 6th ed., Vol. 1,
pp. 149–178, Wiley-VCH, New York ~2003!.
10. Z. LEI, C. LI, Y. LI, and B. CHEN, Sep. Purif. Technol.,
36, 131 ~2004!.
11. S. K. GANGULY and A. N. GOSWAMI, Sep. Sci. Tech-
nol.,31, 1267 ~1996!.
12. Y. KUO, C. L. MUNSON, W. G. RIXEY, A. A. GARCIA,
M. FRIERMAN, and C. J. KING, Sep. Purif. Methods,16,31
~1987!.
13. I. ABE, K. HAYASHI, M. KITAGAWA, and T. URA-
HATA, Bull. Chem. Soc. Jpn.,52, 1899 ~1979!.
14. M. SUZUKI and T. FUJII, AIChE J.,28, 380 ~1982!.
15. S. W. WANG, A. L. HINES, and E. PEDRAM, Ind. Eng.
Chem. Fundam.,20, 350 ~1981!.
16. M. FRIERMAN, Y. KUO, D. JOSHI, A.A. GARCIA, and
C. J. KING, Sep. Purif. Methods,16,91~1987!.
17. G. MAURER and A. SCHUNK, Fluid Phase Equilibria,
239, 223 ~2006!.
18. J. M. WARDELLand C. J. KING, J. Chem. Eng. Data,23,
144 ~1978!.
19. N. L. RICKER, J. N. MICHAELS, and C. J. KING, J. Sep.
Process Technol.,1,36~1979!.
20. M. N. INGALE and V. V. MAHAJANI, Sep. Technol.,6,1
~1996!.
21. B. C. ROY, M. J. ALAM, and M. GOTO, J. Appl. Sci.,6,1
~2006!.
22. M. D. BERMEJO and M. J. COCERO, AIChE J.,52, 3933
~2006!.
23. V. S. MISHRA, V. V. MAHAJANI, and J. B. JOSHI, Ind.
Eng. Chem. Res.,34,2~1995!.
24. E. NEYENS and J. BAEYENS, J. Hazard Mater.,98,33
~2003!.
25. N. SANO, T. FUJIMOTO, T. KAWASHIMA, T. KANKI,
and A. TOYODA, Proc. 5th Asian Pacific Conf. Sustainable
Energy and Environmental Technologies, Wellington, New Zea-
land, May 8–11, 2005, p. 89, World Scientific Publishing Com-
pany ~2005!.
Mitchell et al. REMOVING ACETIC ACID FROM A UREX WASTE STREAM
NUCLEAR TECHNOLOGY VOL. 165 MAR. 2009 369
... Solvent extraction is used regularly in other industries to remove acetic acid from dilute solutions such as those expected in UREX+. Distillation is more likely to be used to remove acetic acid from more concentrated solutions.[4,5] ...
Article
Full-text available
In the UREX+ process, acetohydroxamic acid is added to the UREX step to suppress plutonium extraction. Hydrolysis of acetohydroxamic acid (AHA) produces acetic acid and hydroxyl ammonium nitrate (HAN). The capability to remove acetic acid from aqueous process solutions is desirable as UREX+ flowsheets continue to evolve. Reasons for acetic acid removal include enabling recycle of the nitric acid solution and prevention of problems in downstream process steps. After a literature review, solvent extraction was determined to be the most promising approach for removal of acetic acid at the low concentrations involved. The goal was to identify a solvent that is both insoluble in the aqueous phase and will selectively remove acetic acid. Solutions of tributylphosphate (TBP) have been shown to selectively extract acetic acid over nitric acid with sufficiently high distribution coefficients. Extractant solutions tested in this study include TBP in dodecane and TBP in phenyltrifluoromethyl sulfone (FS-13). The two solvents are both compatible with the UREX+ process flowsheets, so selection of the solvent will be based on the placement of the removal step. The location of the removal step within the process is dependent on many factors including avoiding raffinate stream solutions containing quantities of actinides that TBP would extract, interference of acetic acid with the process steps, and the ratio of nitric to acetic acids in each step so the distribution coefficient can be maximized.
... 9 In the PUREX process, TcO 4 − is extracted under various forms, in particular as mixed U−nitrate− pertechnetate complexes, requiring further steps (e.g., UREX + process) to separate Tc from the major actinides. 10,11 TcO 4 − has high mobility in the environment and is difficult to complex. It is thus important, from a practical and a fundamental perspective, to characterize its relationship with respect to uranyl (either a mere noncoordinated neutralizing counterion or an effective ligand) in different liquid environments. ...
Article
The complexation of perrhenate (ReO(4)(-)) anions by the uranyl (UO(2)(2+)) cation has been investigated by joint molecular dynamics simulations and spectroscopic (UV-vis, TRLFS, and EXAFS) studies in aqueous solution, acetonitrile, and three ionic liquids (ILs), namely, [Bmi][Tf(2)N], [Me(3)BuN][Tf(2)N], and [Bu(3)MeN][Tf(2)N] that are based on the same Tf(2)N(-) anion (bis(trifluoromethylsulfonyl)imide) and either Bmi(+) (1-butyl,3-methylimidazolium), Me(3)BuN(+), or Bu(3)MeN(+) cations. They show that ReO(4)(-) behaves as a weak ligand in aqueous solution and as a strong ligand in acetonitrile and in the ILs. According to MD simulations in aqueous solution, the UO(2)(ReO(4))(2) complex quickly dissociates to form UO(2)(H(2)O)(5)(2+), while in acetonitrile, a stable UO(2)(ReO(4))(5)(3-) species forms from dissociated ions. In the ILs, the UO(2)(ReO(4))(n)(2-n) complexes (n = 1 to 5) remained stable along the dynamics, and to assess their relative stabilities, we computed the free energy profiles for stepwise ReO(4)(-) complexation to uranyl. In the two studied ILs, complexation is favored, leading to the UO(2)(ReO(4))(5)(3-) species in [Bmi][Tf(2)N] and to UO(2)(ReO(4))(4)(2-) in [Bu(3)MeN][Tf(2)N]. Furthermore, in both acetonitrile and [Bmi][Tf(2)N] solutions, MD and PMF simulations support the formation of dimeric uranyl complexes [UO(2)(ReO(4))(4)](2)(4-) with two bridging ReO(4)(-) ligands. The simulation results are qualitatively consistent with spectroscopic observations in the different solvents, without firmly concluding, however, on the precise composition and structure of the complexes in the solutions.
Article
The ligand acetohydroxamic acid (AHA) suffers hydrolysis at acidic conditions. This reaction has been studied for a long time, due to its implications in different applications, by using indirect colorimetric methods. This work shows how Raman spectroscopy can be very useful as a direct technique for measuring the hydrolysis kinetics of AHA, faster, more versatile and easier compared with the indirect traditional UV–Vis method which needs a complex formation with Fe. Thereby, we present a detailed study of the qualitative and quantitative Raman spectra of 1 mol/L AHA and its hydrolysis products. These results enabled us to perform a complete kinetic study of this molecule at different pH ranging from 0.5 mol/L to 4 mol/L HNO3, i.e. not only at excess acidic conditions but also at limiting nitric acid conditions.
Article
Recycle of nitric acid in the UREX+ process requires removal of acetic acid. An analysis of the effects of acetic acid in each process step indicates no step will be significantly affected by the concentrations expected. Thus, acetic acid removal can be placed after the last salts are removed, just before the nitric acid is recycled. Two promising removal options have been considered, solvent extraction and distillation. Distillation requires removal of most of the water before large fractions of acetic acid are removed. The process would be energy intensive and would involve the handling of extremely concentrated nitric acid; therefore, solvent extraction appears to be more attractive.
Article
Ionic Liquids are of great interest in all fields of chemistry and material science. Apart from their "green" properties, their radioresistance and fantastic chemical versatility offer a new playground to nuclear chemists and others in view of liquid/liquid extraction of RE and actinides. Following a brief summary of RE and An liquid/liquid extraction by use of molecular solvents, the author reviews and discusses the main properties of ionic liquids that are of importance to liquid/liquid extraction. Then, the various ways ionic liquids are used for the extraction/separation of RE and An are critically reviewed and discussed. Focus is on mechanism and highlight is put on differences between ionic liquids and molecular solvents.
Article
Full-text available
The Advanced Fuel Cycle Initiative (AFCI), funded by the U.S. Department of Energy's Office of Nuclear Energy, Science and Technology, is developing advanced technologies to safely and economically reduce the volume and heat generation of material requiring geologic disposition, thereby extending the capacity of the Yucca Mountain repository and delaying or avoiding the need for a second repository. An important element of this initiative is the separation of key radionuclides followed by superior waste-disposal forms and/or transmutation of long-lived isotopes. To that end, the AFCI is developing advanced fuel reprocessing systems that separate key radionuclides from spent fuel. One of these systems is the UREX+ process.
Article
Freezing temperatures of water, methanoic acid, ethanoic acid, propanoic acid, butanoic acid, 2-methylpropanoic acid, their binary mixtures and some multicomponent mixtures have been determined using a kinetic technique. The liquid phase activity coefficients for the binary mixtures have been derived from the results and fitted to the equation of Renon and Prausnitz. The resulting parameters have been used to predict the freezing temperatures of the multicomponent mixtures. Good agreement with the experimental results is found.
Article
The radioisotope technique was used to study the competitive adsorption of organic acids from dilute aqueous solution onto activated carbon at 278 and 298 K. A competitive adsorption potential model was developed in this study and used to predict bi-solute equilibrium data. The calculated values were compared both with experimental data and with values predicted by the ideal adsorbed-solution theory. It was found in this study that the competitive adsorption potential model gives slightly better results than those calculated by the other model.
Article
Extraction of carboxylic acids, C2C6, was carried out with tributylphosphate (TBP) from an aqueous waste stream at pH = 2, 2.5, and at as such pH taken from an environmental process engineering point of view. The effect of the presence of H2SO4 and HCl on the distribution coefficients for C2C6 carboxylic acids was investigated. The measurements were also carried out in the presence of inorganic salts such as Na2SO4 and NaCl.
Article
For the recovery and purification of nitric acid from an aqueous solution containing a variety of dilute organic or inorganic acids, vapor-liquid equilibrium (VLE) data are needed. A search of the literature did not yield sufficient reliable VLE data for such mixtures, so an experimental study on VLE K-values (=y/x) of organic and inorganic acids at diluted concentration in nitric acid-water mixtures was carried out.Using the modified Rose-Williams still multicomponent VLE data were measured under reduced pressures. The concentration range of HNO3 is 0.0 to 0.35 mole fraction where appreciable decomposition does not occur. A range of 50–1000 ppm of HCl and various carboxylic acids (HCOOH, CH3COOH, C2H5COOH and C3H7COOH) were added to HNO3 aqueous solutions. KHCl changed from 0.01 to 100 and had positive logarithmic dependence on HNO3 content. K-values of organic acids were represented by a quadratic function of HNO3 concentration.
Article
A new separation method, complex extractive distillation, was put forward in this work for separating acetic acid and water. Tributylamine was selected as the separating agent. The reversible chemical interaction between acetic acid and tributylamine was verified through infrared spectra (IR) and mass chromatogram (MS) technique. The mathematics models of equilibrium (EQ) stage with and without incorporating chemical equilibrium equations were, respectively, established to simulate the extractive distillation column. From the comparison of simulated results with experimentally observed results, it was concluded that the EQ stage model was accurate whether chemical equilibrium equation was incorporated or not because the chemical equilibrium constant was small under the operation condition.
Article
The adsorption isotherms, from aqueous solution onto activated carbon, have been determined for the 22 aliphatic monofunctional compounds; alcohols, carboxylic acids, ketones, ethers, esters, and aldehydes. The free energy change for the adsorption process has been calculated from the adsorption equilibrium constant at infinite dilution; a linear relationship exists between the free energy change and the number of carbon atoms in the solute. The free energy of adsorption has been divided into two contributions: the hydrocarbon and the functional group contribution. The free energy change for the adsorption process has been compared with the free energy change for the precipitation process from solution into the pure liquid solute.
Article
A synergistic extraction solvent for the simultaneous removal of cesium and strontium from acidic solutions has been investigated. The extraction solvent consists of, 4,4',(5')-di-(t-butyldicyclo-hexano)-18-crown-6 (DtBuCH18C6), calix[4]arene-bis-(tert-octylbenzo-crown-6) (BOBCalixC6), and 1-(2,2,3,3-tetrafluoropropoxy)-3-(4-sec-butylphenoxy)-2-propanol (Cs-7SB modifier) in a branched aliphatic kerosene (Isopar® L). Extraction synergy for strontium was observed when DtBuCH18C6 was combined with the BOBCalixC6 cesium extractant solvent and Cs-7SB modifier or if the Cs-7SB modifier was substituted into the SREX (Strontium Extraction) solvent in place of TBP. The novel process extracted both cesium and strontium simultaneously from 1 M nitric acid solutions with distribution ratios of 8.8 and 7.7 for strontium and cesium, respectively, at ambient temperature. Distribution coefficients for cesium and strontium as a function of nitric acid concentration and temperature were also obtained with 0.5 M