Ammonia Gas Treatment for Uranium Immobilization at DOE Hanford Site

Abstract

Ammonia gas is a potential uranium (U) remediation technique for the vadose zone at the Department of Energy's (DOE) Hanford Site in Washington State via pH manipulation. The objective of this work was to investigate U removal from the aqueous phase and mineral dissolution with either NaOH or NH4OH in order to understand ammonia as a remediation technique for the vadose zone. Batch experiments were designed to investigate the fate of U and mineral dissolution upon treatment with either NaOH or NH4OH. These experiments investigated mineral dissolution and U partitioning in the presence of pure minerals (quartz, kaolinite, illite, montmorillonite, muscovite) and sediments relevant to the Hanford site in either NaCl or synthetic groundwater formulated based on the Hanford site groundwater components. Analysis of mineral dissolution via measurement of major cations has demonstrated that there is a significant increase in dissolution of minerals with basic treatment and likely greater secondary precipitation for the NH4OH as compared to NaOH treatment for clay minerals and Hanford sediments.

Full text

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
1
Ammonia gas Treatment for Uranium Immobilization at DOE Hanford Site –
17067
Silvina Di Pietro*, Hilary Emerson*, Yelena Katsenovich*
*Applied Research Center – FIU, email: hemerson@fiu.edu
ABSTRACT
Ammonia gas is a potential uranium (U) remediation technique for the vadose zone
at the Department of Energy’s (DOE) Hanford Site in Washington State via pH
manipulation. The objective of this work was to investigate U removal from the
aqueous phase and mineral dissolution with either NaOH or NH4OH in order to
understand ammonia as a remediation technique for the vadose zone. Batch
experiments were designed to investigate the fate of U and mineral dissolution
upon treatment with either NaOH or NH4OH. These experiments investigated
mineral dissolution and U partitioning in the presence of pure minerals (quartz,
kaolinite, illite, montmorillonite, muscovite) and sediments relevant to the Hanford
site in either NaCl or synthetic groundwater formulated based on the Hanford site
groundwater components. Analysis of mineral dissolution via measurement of major
cations has demonstrated that there is a significant increase in dissolution of
minerals with basic treatment and likely greater secondary precipitation for the
NH4OH as compared to NaOH treatment for clay minerals and Hanford sediments.
INTRODUCTION
The Department of Energy’s (DOE) Hanford site in Washington State has deposited
over 200,000 kg of uranium (U) into the vadose zone [1, 2]. This release occurred
largely as a result of improper disposal of waste from plutonium production during
World War II and the Cold War. Further, U is mobile within the site due to oxidizing
conditions and the presence of carbonate creating highly mobile uranyl carbonate
species. For example, the partitioning coefficient (Kd) for U was previously
measured in the range of 0.11 – 4 mL/g at pH 8 for Hanford sediments and
groundwater and the retardation factor was measured at 1.43 [2, 3].
Moreover, the Hanford vadose zone is 255+ feet thick with contamination measured
down to 170 feet below the ground surface [4]. Therefore, there is a desire to
create a remediation option that does not input additional liquid to the vadose zone
as this could increase flux of U to the groundwater below. Of the remediation
methods that the DOE is currently considering, ammonia gas injection appears to
be a favorable option. Gas injection has been previously described as a viable
remediation technique for inorganic radionuclides because they are highly affected
by solution chemistry [5, 6].
The goal of the remediation technique is to remove U from the aqueous phase by
raising the pH of the system leading to immobilization as insoluble precipitates or

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
2
strongly sorbed species. Basic injections, including the injection of the weak base
NH3, may lead to the slow dissolution of silica-containing minerals such as quartz,
montmorillonite, muscovite and kaolinite [3, 7-9]. This results in an increase in
dissolved Si4+ and Al3+ as well as small increases in Na+, K+, Fe2+/3+, Cl-, F- and
SO42- [3, 9]. Moreover, Ca2+ increases were reported in column experiments
following injection of U + 0.1 M NaOH + 1 M NaNO3 [3]. The dissolution of these
minerals will ultimately buffer the pH of the system [3, 10].
However, it must be noted that geochemical changes within the subsurface are
often temporary unless they are moving the system towards its natural equilibrium.
The injection of ammonia gas for remediation is designed to temporarily raise the
pH of the aqueous phase to dissolve natural aluminosilicate minerals. Based on
preliminary laboratory scale experiments, it is expected that the system may reach
a pH of 11 – 13 [9]. Then, as the system returns to a neutral pH as the ammonia
evaporates, U is expected to be immobilized as part of a co-precipitation process
with aluminosilicate minerals. As ammonia gas evaporates and the pH returns to
neutral, there are two phenomena that are expected to decrease the mobility of U
(1) U precipitation as solubility of Si, Al and similar ions decreases and (2) U
(co)precipitates are coated with non-U, low solubility precipitates. Some of the low
solubility precipitates that are expected to form include cancrinite, sodalite,
hydrobiotite, brucite and goethite [11-14].
It is important to understand the impact of mineral dissolution and secondary
precipitation processes on the fate of U during and after remediation. The objective
of this research is to investigate the partitioning of U and the mineral dissolution
upon pH manipulation with base treatment via addition of either NaOH or NH4OH.
Experiments focused on simplified batch experiments with pure minerals (quartz,
kaolinite, illite, montmorillonite, muscovite) and Hanford sediments in order to
determine the minerals controlling the dissolution and precipitation processes for
the Hanford sediments.
MATERIALS AND METHODS
Materials
Experiments were conducted with the minerals kaolinite (Alfa Aesar), illite (IMt-2,
Clay Minerals Society), montmorillonite (SWy-2, Clay Minerals Society), muscovite
(Ward Scientific, <2 mm size fraction) and quartz (Ottawa Sand Standard passed
through 20-30 mesh, Fisher) due to their significance at DOE’s Hanford site.
Montmorillonite, muscovite and kaolinite are common in the clay-sized fraction of
sediments at the site and were previously observed to undergo significant dissolution
under basic conditions [9]. Quartz represents the most significant fraction of the bulk
sediments [4]. Additional experiments were also conducted with clean sediments
collected by Dr. Jim Szecsody from the ERDF pit at a depth of 6.1 meters for
comparison. Further characterization of this sediment has been published previously
[3].

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
3
Minerals and sediments were washed and equilibrated with the either NaCl or
synthetic groundwater prior to experiments based on the methods outlined in Table
I below. Following the steps in Table I, the solids were dried at 35°C for ~3 days and
lightly crushed with a mortar and pestle to homogenize. Surface area for the
aforementioned minerals was measured in m2/g based on the BET method (Table II).
Two solutions were formulated to describe the Hanford groundwater, (1) a simplified
synthetic groundwater as described in Table II and (2) NaCl solution of similar ionic
strength (~3.2 mM) for comparison. The simplified synthetic groundwater in Table
III is based on correspondence with Dr. Szecsody and previous work [15].
TABLE I: Summary of Mineral Washing Methods
Mineral
Method
Reference
Quartz (Ottawa
Sand)
(1) Mix 100 g/L suspension in 0.01 M NaOH for 60 minutes, (2)
Centrifuge, decant, replace liquid with 0.01 M HCl, mix 60
minutes, (3) Centrifuge, decant, replace with Nanopure (>18
MΩ) H2O and mix 3 minutes, (4) repeat step three two more
times
[16]
Montmorillonite
(1) Mix 100 g/L suspension in 0.001 M HCl for 30 minutes, (2)
Add 0.5 mL H2O2 and mix an additional 30 minutes, (3)
Centrifuge 6 hours at 4500 rpm, decant aqueous and replace
with 0.01 M NaCl (or synthetic groundwater) and mix overnight,
(4) Repeat four times, (5) Centrifuge, decant and replace with
Nanopure H2O, (6) Repeat at least four times (until excess ions
are removed)
[16]
Kaolinite
(1) Mix 100 g/L suspension in 1 M NaCl (or synthetic
groundwater) for 30 minutes, (2) Centrifuge, decant and repeat
four more times, (3) Centrifuge, decant and replace with
Nanopure H2O, (4) repeat four more times
[17]
Illite
(1) Mix 100 g/L suspension with 1 M NaCl (or synthetic
groundwater) for three hours and allow to flocculate overnight,
(2) Decant and replace with 1 M NaCl (or synthetic groundwater)
and mix, (3) Repeat two more times, (4) Decant and replace
with Nanopure H2O, (5) Repeat until excess ions are removed
[18]

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
4
TABLE II: BET surface area for relevant minerals and Hanford sediments
Mineral ID
m2/g
Hanford Sediment
17.4
Quartz
0.046
Kaolinite
17.9
Muscovite
0.096
Illite
19.1
Montmorillonite
23.8
TABLE III: Synthetic groundwater composition with total ionic strength of 3.2 mM
Element
(mmol/L)
Na+
1.1
K+
0.22
Ca2+
1.4
Mg2+
0.6
HCO3-
1.32
Cl-
3.9
Batch Experimental Protocol
Batch experiments were conducted in triplicate at pH ~7.5 in the presence of 100
g/L quartz, 5 g/L kaolinite, illite, muscovite, montmorillonite, or 25 g/L Hanford
sediment and either synthetic groundwater (Table II) or NaCl at similar ionic
strength (3.2 mM). An aliquot of U (Spex Certiprep, New Jersey) was added
following equilibration of samples at pH ~7.5 to reach 500 ppb U. After equilibration
with U for three days on an end over end tube revolver at 40 rpm (Thermo
Scientific), a homogenous aliquot was removed for analysis for both controls
(without mineral) and samples.
Samples were centrifuged at 5000 rpm for 30 minutes (18100 rcf, Thermo
Scientific, Corvall ST 16R centrifuge) to remove particles >100 nm based on
Stoke’s law as described by Jackson [19]. Then, the supernatant was acidified in
1% HNO3 (Fisher, ACS Plus) for analysis by kinetic phosphorescence analyzer (KPA-
11, Chemchek) for U and inductively coupled plasma optical emission spectroscopy
(ICP-OES, Perkin Elmer, Optima 7300 DV) for major cations (Ca2+, Mg2+, Fe2+/3+,
H4SiO4+ as Si, Al3+, Na+, K+). Al and Si were analyzed to track dissolution of the
minerals throughout these experiments.
Following equilibration at pH ~7.5, the pH of each sample was raised with either
2.5 M NH4OH or 2.5 M NaCl + 0.025 M NaOH. It must be noted that NaCl is
included in the NaOH solution to maintain similar ionic strength for both solutions
allowing for a more representative comparison. Samples adjusted with NH4OH were
immediately capped and wrapped with parafilm following addition to reduce

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
5
volatilization of NH3 gas. Ammonia volatilization increases by an order of magnitude
for every unit above pH 6.0 and, therefore, is expected to be higher in alkaline soil
suspensions [20, 21]. The adjustment by either NH4OH or NaOH allows for
comparison of both options as a possible step to raise the pH (Thermo Scientific,
8175BNWP) during remediation of the subsurface. After adjustment, samples were
equilibrated for three days before analysis as described above for U and major
cations.
DISCUSSION
Mineral Dissolution
Fig. 1 represents dissolution of Hanford sediments as determined by aqueous Al
and Si measurements in 25 g/L batch reactors in the presence of synthetic
groundwater. Base treatment significantly increased both Si and Al concentrations
in the aqueous phase as compared to initial conditions at pH 7.5. It should be noted
that aqueous Al was below detection limits for ICP-OES for the initial conditions at
pH 7.5 (LOD 41 ppb for Al). However, significantly greater Si is present in the
aqueous phase than Al for all conditions. Previous work by Szecsody et al. noted
that less Al was measured in the aqueous phase than Si with base treatment [3].
Fig. 1: Aqueous Al (blue) and Si (gray) dissolved from Hanford sediment (25 g/L) in
synthetic groundwater with pH at ~11.5 via adjustment with either NaOH or NH4OH
or at ~7.5 to represent initial conditions prior to base treatment, Note: Al
measurements at pH 7.5 were below detection limits
Figs. 2-5 compare pure mineral dissolution with respect to base treatment where
Figs. 2-3 represent minerals in the presence of 3.2 mM synthetic groundwater and
Figs. 4-5 in the presence of 2.5 M NaCl + 0.025 M NaOH. Mineral dissolution is
estimated based on aqueous cation measurements by ICP-OES. Synthetic

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
6
groundwater is representative of the Hanford site groundwater, and NaCl
represents the simplest chemical system at a similar ionic strength to synthetic
groundwater. The NaCl solution allows for a better understanding of the fate of U
without the more complex divalent cations present in the actual and synthetic
groundwater.
In the presence of synthetic groundwater solution, the greatest fraction of Si
entered the aqueous phase at elevated pH for kaolinite, as shown in dark blue for
Fig. 2. In addition, aqueous Si fractions are similar for each of the other clays.
Based on Fig. 3, muscovite resulted in the highest fractions of Al in the aqueous
phase. However, in general, each of the minerals had similar aqueous Al
concentrations. Furthermore, there is not a clear difference between the aqueous
cation concentrations with the base treatments.
The greater aqueous Si as compared to Al for the clay minerals (kaolinite, illite,
muscovite, and montmorillonite) is similar to results for Hanford sediments (Fig. 1)
and confirms results from previous work by Szecsody et al [3]. Moreover, a
comparison of the ratios of Al to Si measured in the aqueous phase with the
theoretical ratios based on the minerals investigated (Table IV) suggests that
incongruent dissolution is occurring. Therefore, it is likely that a secondary
precipitate is forming following dissolution of the clay minerals.
Furthermore, Figs. 4 and 5 represent the results for mineral dissolution in the
presence of 3.2 mM NaCl solution for kaolinite, illite, quartz and montmorillonite. It
should be noted that Fig. 4 indicates an increase in Si in the aqueous phase for
NaOH-treated samples as compared to NH4OH. This effect can be explained by the
different impacts of the two treatments on mineral solubility. The addition of NaOH
adds singly charged ions (Na+) to solution. However, the addition of NH4OH adds
greater than 99% molecular species (NH3) at pH ~11.5 based on
ammonia/ammonium speciation. For the charged ions (NaOH), solubility increases
with ionic strength while molecular species (NH3) decreases [22]. Therefore, it is
expected that the increase in molecular species for the NH4OH treatment would
result in a significant decrease in solubility especially of Si as it is most likely to
dissolve as a molecular species (H4O4Si). The trend for Al in Fig. 5 is not as clear.

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
7
Fig. 2: Aqueous Si as a percentage of total initial Si based on initial minerals (layer
silicate clays and quartz) in the presence of synthetic groundwater with aqueous
measurements by ICP-OES
Fig. 3. Aqueous Al as a percentage of total initial Al based on initial minerals
(kaolinite, montmorillonite, illite and muscovite) in the presence of synthetic
groundwater based on aqueous measurements by ICP-OES

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
8
TABLE IV: Comparison of Al:Si molar ratios measured in the aqueous phase for
base treatment with NH4OH and NaOH in the presence of synthetic groundwater
with theoretical ratios, Note: ratio not included for illite-NH4OH because Si was
below detection limit
Mineral
NH4OH
NaOH
Theoretical
Kaolinite
0.049
0.098
1
Illite
-
0.053
0.5
Montmorillonite
0.089
0.019
0.5
Muscovite
0.532
0.646
1
Fig. 4: Aqueous Si as a percentage of total initial Si based on initial minerals (layer
silicate clays and quartz) in the presence of 3.2 mM NaCl with aqueous
measurements by ICP-OES

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
9
Fig. 5: Aqueous Al as a percentage of total initial Al based on initial minerals
(kaolinite, montmorillonite, illite and muscovite) in the presence of 3.2 mM NaCl
based on aqueous measurements by ICP-OES
Figs. 6 and 7 represent the fraction of Si and Al in the aqueous phase, respectively,
from 5 g/L kaolinite after three days of equilibration in synthetic groundwater (circles)
or NaCl (diamonds) after base addition of aliquots of 2.5 M NH4OH or 2.5 M NaCl +
0.025 M NaOH to reach variable basic pH values. The dissolution of kaolinite clearly
increases with pH as expected. Moreover, the initial electrolytes (either synthetic
groundwater or a composition of NaCl) do not significantly affect the dissolution. It
is important to note that both initial solutions are at a similar ionic strength (7.2
versus 3.2 mM), but the synthetic groundwater is more complex with a significant
contribution from divalent cations such as Mg and Ca (Table III).
However, the type of base treatment appears to have an effect on the concentration
of Si in the aqueous phase as compared to the initial electrolyte while the Al
concentration in the aqueous phase is similar for each treatment (Figs. 6 and 7,
respectively). Further, the 2.5 M NaCl + 0.025 M NaOH treatment led to a significant
increase in Si in the aqueous phase near pH 10.5-11 as compared to the NH4OH
treatment. The 2.5 M NH4OH treatment did not reach similar Si concentrations until
almost pH 12. Again, this can be attributed to the effects of molecular versus ionic
species on solubility as discussed above and likely indicates greater precipitation of
silicon-containing solids with NH4OH treatment as compared to NaOH.

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
10
Fig. 6. Comparison of aqueous Si as a fraction of the total based on the initial
kaolinite mineral concentration following three days of equilibration of 5 g/L
kaolinite in synthetic groundwater (circles) or 3.2 mM NaCl (diamonds) at variable
pH following treatment with either 2.5 M NH4OH (yellow) or 2.5 M NaCl + 0.025 M
NaOH (blue)

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
11
Fig. 7: Comparison of aqueous Al as a fraction of the total based on the initial
kaolinite mineral concentration following three days of equilibration of 5 g/L
kaolinite in synthetic groundwater (circles) or 3.2 mM NaCl (diamonds) at variable
pH following treatment with either 2.5 M NH4OH (red) or 2.5 M NaCl + 0.025 M
NaOH (black)
CONCLUSIONS
The work presented above is part of a larger, ongoing effort to understand the long
term fate of uranium at the Hanford site if ammonia gas injection is chosen as a
remediation technique. The data presented examines the dissolution behavior of
minerals and sediments with significant relevance to Washington State’s Hanford
site. Upon base treatment with either NaOH or NH4OH, quartz and aluminosilicate
clays dissolve. Although the NaCl versus synthetic groundwater for initial ionic
strength did not significantly affect mineral dissolution, some differences were
observed with base treatments. In the case of montmorillonite, illite and kaolinite,
significantly greater aqueous Si was observed for NH4OH treatment versus NaOH
treatment. However, aqueous Al measurements were similar for both treatments.
It is still unclear whether U mobility will be decreased long term, but it is expected
that these precipitation processes could remove additional U from the aqueous
phase and coat precipitated uranyl phases with lower solubility precipitates.

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
12
Investigations of these processes and characterization of solid phases to confirm
previous work and model predictions is the subject of ongoing and future work.
ACKNOWLEDGEMENTS
We would like to thank the Department of Energy Office of Environmental
Management for funding under Cooperative Agreement #DE-EM0000598.
REFERENCES
1. CORBIN, R.A., et al., "Hanford soil Inventory Model Rev.1", CH2M Hill
Hanford Group, Inc., Richland, WA (2005).
2. ZACHARA, J.M., et al., "A site-wide perspective on uranium geochemistry at
the Hanford Site", Pacific Northwest National Laboratory, Hanford, WA
(2007).
3. SZECSODY, J.E., et al., "Influence of acidic and alkaline waste solution
properties on uranium migration in subsurface sediments." Journal of
Contaminant Hydrology. 151, p. 155-175 (2013).
4. SERNE, R.J., et al., "Characterization of vadose zone sediment: Borehole
299-E33-45 near BX-102 in the B-BX-BY waste management area", Pacific
Northwest National Laboratory, Richland, WA (2008).
5. DENHAM, M.E. and B.B. LOONEY, "Gas: A Neglected Phase in Remediation of
Metals and Radionuclides", Savannah River National Laboratory, Aiken, SC.
p. 1-23 (2005).
6. DRESEL, P.E., et al., "Review: Technical and Policy Challenges in Deep
Vadose Zone Remediation of Metals and Radionuclides." Environmental
Science and Technology. 45,10, p. 4207-4216 (2011).
7. WAN, J.M., et al., "pH neutralization and zonation in alkaline-saline tank
waste plumes." Environmental Science & Technology. 38,5, p. 1321-1329
(2004).
8. WAN, J.M., et al., "Geochemical evolution of highly alkaline and saline tank
waste plumes during seepage through vadose zone sediments." Geochimica
et Cosmochimica Acta. 68,3, p. 491-502 (2004).
9. SZECSODY, J.E., et al., "Geochemical and Geophysical Changes during
Ammonia Gas Treatment of Vadose Zone Sediments for Uranium
Remediation." Vadose Zone Journal, p. 1-13 (2012).
10. QAFOKU, N.P., et al., "Aluminum effect on dissolution and precipitation under
hyperalkaline conditions: I. Liquid phase transformations." Journal of
Environmental Quality. 32,6, p. 2354-2363 (2003).
11. BICKMORE, B.R., et al., "Nitrate-cancrinite precipitation on quartz sand in
simulated Hanford tank solutions." Environmental Science and Technology.
35,22, p. 4481-4486 (2001).
12. QAFOKU, N.P., et al., "Transport-controlled kinetics of dissolution and
precipitation in the sediments under alkaline and saline conditions."
Geochimica Et Cosmochimica Acta. 68,14, p. 2981-2995 (2004).

WM2017 Conference, March 5 – 9, 2017, Phoenix, Arizona, USA
13
13. QAFOKU, N.P. and J.P. ICENHOWER, "Interactions of aqueous U (VI) with soil
minerals in slightly alkaline natural systems." Reviews in Environmental
Science and Biotechnology. 7,4, p. 355-380 (2008).
14. ZHAO, H., et al., "Alteration of Kaolinite to Cancrinite and Sodalite by
Simulated Hanford Tank Waste and its Impact on Cesium Retention." Clays
and Clay Minerals. 52,1, p. 1-13 (2004).
15. SZECSODY, J.E., et al., "Uranium Mobility During In Situ Redox Manipulation
of the 100 Areas of the Hanford Site", Pacific Northwest National Laboratory,
Richland, WA. p. 41 (1998).
16. POWELL, B.A., et al., "Development of a Composite Non-Electrostatic Surface
Complexation Model Describing Plutonium Sorption to Aluminosilicates."
Lawrence Livermore National Laboratory. LLNL-TR-408276 (2008).
17. HEIDMANN, I., I. CHRISTL, and R. KRETZSCHMAR, "Sorption of Cu and Pb to
kaolinite-fulvic acid colloids: Assessment of sorbent interactions." Geochimica
et Cosmochimica Acta. 69,7, p. 1675-1686 (2005).
18. BAEYENS, B. and M.H. BRADBURY, "Cation exchange capacity measurements
on illite using the sodium and cesium isotope dilution technique: Effects of
the index cation, electrolyte concentration and competition: Modeling." Clays
and Clay Minerals. 52,4, p. 421-431 (2004).
19. JACKSON, M.L., "Soil Chemical Analysis - Advanced Course, 2nd ed.", in Soil
Chem. Anal.-Adv. course, Parallel Press, Madison, WI (1985).
20. SINGANDHUPE, R.B. and R.K. RAJPUT, "Ammonia volatilization from rice
fields in alkaline soil as influenced by soil moisture and nitrogen." Journal of
Agricultural Science. 112, p. 185-190 (1989).
21. VLEK, P.L. and E.T. CRASWELL, "Ammonia volatilization from flooded soils."
Fertilizer Research. 2,4, p. 227-245 (1981).
22. LANGMUIR, D., Aqueous Environmental Geochemistry, ed. R. McConnin, et
al. Upper Saddle River, New Jersey: Prentice Hall. 600 (1997).