Article pubs.acs.org/est
Separating and Stabilizing Phosphate from High-Level Radioactive Waste: Process Development and Spectroscopic Monitoring Gregg J. Lumetta,* Jenifer C. Braley, James M. Peterson, Samuel A. Bryan,* and Tatiana G. Levitskaia* Pacific Northwest National Laboratory, P.O. Box 999, MSIN P7-25, Richland, Washington 99352, United States S Supporting Information *
ABSTRACT: Removing phosphate from alkaline high-level waste sludges at the Department of Energy’s Hanford Site in Washington State is necessary to increase the waste loading in the borosilicate glass waste form that will be used to immobilize the highly radioactive fraction of these wastes. We are developing a process which first leaches phosphate from the high-level waste solids with aqueous sodium hydroxide, and then isolates the phosphate by precipitation with calcium oxide. Tests with actual tank waste confirmed that this process is an effective method of phosphate removal from the sludge and offers an additional option for managing the phosphorus in the Hanford tank waste solids. The presence of vibrationally active species, such as nitrate and phosphate ions, in the tank waste processing streams makes the phosphate removal process an ideal candidate for monitoring by Raman or infrared spectroscopic means. As a proof-of-principle demonstration, Raman and Fourier transform infrared (FTIR) spectra were acquired for all phases during a test of the process with actual tank waste. Quantitative determination of phosphate, nitrate, and sulfate in the liquid phases was achieved by Raman spectroscopy, demonstrating the applicability of Raman spectroscopy for the monitoring of these species in the tank waste process streams.
■
INTRODUCTION The U.S. Department of Energy (DOE) is responsible for environmental remediation at former nuclear weapons production sites. The Hanford Site in Washington State represents one of the most daunting challenges to the DOE in fulfilling its environmental remediation mission. Although there are a number of aspects to the Hanford remediation effort, retrieving, processing, immobilizing, and disposing of the 2.2 × 105 m3 of radioactive wastes stored in the Hanford underground storage tanks dominates the overall effort.1 These wastes were generated during plutonium production and other operations at Hanford during World War II and the Cold War.2 The cornerstone of the tank waste remediation effort is the Hanford Tank Waste Treatment and Immobilization Plant (WTP), which is currently under construction and is scheduled to begin full radioactive operations in 2019. As currently designed, the capability of the WTP to treat and immobilize the Hanford tank wastes in the expected lifetime of the plant is uncertain. For this reason, DOE has been pursuing supplemental treatment options for selected wastes. If implemented, these supplemental treatments will route certain waste components to processing and disposition pathways outside of WTP, thereby accelerating the overall Hanford tank waste remediation mission. The Hanford tank wastes are highly alkaline and consist of three general phasessupernatant liquid, salt cake, and sludge; however there is considerable variability in the composition depending on the specific separation process that generated the waste.3 The supernatant liquid and salt cake phases consist primarily of water-soluble sodium salts. The sludge phase © 2012 American Chemical Society
consists of water-insoluble metal hydroxides, including the transuranic elements. One waste component of particular concern is phosphorus. There is a relatively low tolerance for P in the high-level waste (HLW) melter in the WTP. For this reason, considerable effort has been devoted to investigating caustic leaching methods for removing P from the HLW tank sludge solids.4−7 The previous studies have shown that, in most cases, caustic leaching is an effective means to remove P as phosphate from the HLW solids. However, the previous work was done in the context of the WTP concept; that is, it was envisioned that the HLW solids would be leached with caustic, and the resulting leachate solution containing the phosphate would then be processed through ion exchange to remove 137Cs and immobilized in the low-activity waste (LAW) borosilicate glass waste form. On the other hand, an optimized flowsheet for processing high-phosphate wastes in the WTP has not been developed. It is reasonable to examine whether there are other options for managing the P from the Hanford tank sludge solids such that this element is routed to a disposition pathway that does not require it to flow through the WTP. We have been investigating a two-step process for removing phosphorus from high-phosphate Hanford tank sludge (see the Supporting Information for a more detailed description of the concept).8 The process involves removing the phosphate from the sludge solids, where it is contained primarily as FePO4,9 by Received: Revised: Accepted: Published: 6190
February 2, 2012 March 23, 2012 April 17, 2012 May 9, 2012 dx.doi.org/10.1021/es300443a | Environ. Sci. Technol. 2012, 46, 6190−6197
Environmental Science & Technology
Article
that the process can be monitored by these techniques. Although we describe here a specific chemical process, the spectroscopic online monitoring tools can be adapted to a wide range of process applications, and can be extended to monitoring of a variety of different analytes.14 Performing in-tank separation of phosphate, enabled by realtime analysis of the process solutions by vibrational spectroscopy, coupled to near-tank stabilization of the nonradioactive phosphate component of the waste, offers a novel method for pretreating high-level tank waste without investment in heavily shielded centralized processing facilities. The results described in this paper indicated that the combination of caustic leaching and calcium phosphate precipitation is a promising approach to economically processing and disposing of the phosphate material from the Hanford tank wastes; however a number of issues will need to be resolved before its implementation. These include determining the distribution of radionuclides between the various processing streams, determining the specific waste form into which the calcium product would be cast, where that material would be disposed, and its long-term performance in protecting the environment.
leaching with aqueous sodium hydroxide. This is believed to be achieved primarily through the following metathesis reaction: FePO4 (s) + 3NaOH(aq) → Fe(OH)3 (s) + Na3PO4 (aq) (1)
Following a solid/liquid separation, the HLW solids containing Fe(OH)3 are ultimately routed to the WTP for immobilization in borosilicate glass. The liquid phase is treated with CaO, which removes the phosphate from solution via precipitation. Earlier work suggests the phosphate is removed primarily as hydroxyapatite:8 3Na3PO4 (aq) + 5CaO(s) + 5H 2O → Ca5(OH)(PO4 )3 (s) + 9NaOH(aq)
(2)
The solid hydroxyapatite (and/or other calcium phosphate phases) is a very thermodynamically stable phase and could be readily mixed with cement to form a stable LAW waste form for onsite disposal. In this manner, the P can be routed for disposal without being processed through the WTP. This process should be generally applicable to any waste stream containing a metal phosphate phase which has a Ksp greater than the corresponding metal hydroxide. A similar caustic leaching/Ca precipitation scheme has been investigated for recovering P from wastewater treatment plant sludges.10 In parallel with the process development effort, we have been developing spectroscopic tools for online monitoring of radiochemical separations processes. The spectroscopic tools employed include vibrational techniques (infrared and Raman) and spectrophotometry (visible-near-infrared). The vibrational spectroscopic techniques are particularly well suited for the monitoring of processes involving Hanford tank wastes because many of the major waste components are Raman and infrared active, including nitrate, nitrite, carbonate, sulfate, oxalate, chromate, aluminate, hydroxide, and (most relevant to this work) phosphate. Construction of chemometric models, based on the spectral databases, allows rapid quantification of molecular oxyanions in these highly complex waste matrices. Indeed, systems based on Raman spectroscopy were designed, constructed, and qualified for monitoring the retrieval of liquid phases from Hanford HLW tanks,11 and for monitoring a continuous sludge leaching process as part of a near tank treatment system.12 In the off-line configuration, the Raman monitoring system was extensively tested using liquid solutions generated during engineering-scale testing of the WTP leaching and filtration systems.13 In this paper, we describe the results of testing of the phosphate removal and calcium stabilization process with both a tank waste simulant and with actual radioactive tank waste. Caustic leaching for removing phosphate from the Hanford tank waste solids has been previously demonstrated with actual tank waste material,4−7 but the subsequent treatment of actual tank waste leachate with CaO has not been previously demonstrated. The work reported here represents the first integrated demonstration of the concept. The specific objectives of the experiments were to demonstrate (a) the separation of the phosphate component from the waste solids, (b) the recovery of the phosphate from the leachate solution as a stable solid calcium phosphate phase, and (c) that Raman spectroscopy can be used to quantify phosphate, nitrate, and sulfate ions in the process solutions. The process solutions and solids were examined by Raman and Fourier transform infrared (FTIR) spectroscopies to demonstrate the proof-of-principle
■
EXPERIMENTAL SECTION Materials and Equipment. Phosphate-rich Hanford tank waste was acquired from archived samples at the Hanford Site.7 The following chemicals were obtained from commercial vendors and were used as-received: NaOH (Alfa Aesar), NaNO3 (Fisher Chemical), NaNO2 (EM Science), NaF (J.T. Baker Chemical), Na2SO4·10H2O (Aldrich), and CaO (J.T. Baker Chemical). Raman measurements were performed with an InPhotonics RS2000 high-resolution Raman spectrometer containing a thermoelectrically cooled charge-coupled device detector operating at −55 °C, a 300 mW 670 nm visible diode laser as the excitation source, and an InPhotonics RamanProbe focused fiber-optic probe operated in a 180° back reflection mode. For the Raman measurements, the samples or standard solutions were placed in a 2-Dram borosilicate glass vial and the vial was placed on top of the inverted Raman probe so that the laser entered the solution through the bottom of the vial. The laser beam focal point was 5 mm beyond the end of the laser probe tip within the sample phase, and the measured laser intensity at the sample was typically 50 mW. An integration time of 50 s was used for each acquisition, and 5 acquisitions were taken and averaged for each sample. Raman data analysis and partial least-squares (PLS) model calculation were performed using the PLS Toolbox (version 6.2.1) from eigenvector Research Inc. (Wenatchee, WA, USA) running in a MATLAB environment (version 7.9.0), from Mathworks Inc. (Natick, MA, USA). FTIR spectra were recorded on a Bruker Optics Alpha-P spectrometer equipped with a diamond attenuated total reflectance cell; samples were analyzed with 48 scans at resolution of 2 cm−1. All liquid samples were filtered prior to spectroscopic analysis. Elemental analysis was performed by inductively coupled argon plasma optical emission spectrometry (ICP-OES; Perkin-Elmer Optima 5300). The sludge solids were subjected to KOH/KNO3 fusion and then dissolved in nitric acid for chemical analysis;7 complete dissolution of the solids was achieved. Phosphate, sulfate, and nitrate anions were analyzed by a Dionex ICS-2500 modular ion chromatography (IC) instrument equipped with a conductivity detector using a Dionex AS18 column and hydroxide-based gradient eluant. 6191
dx.doi.org/10.1021/es300443a | Environ. Sci. Technol. 2012, 46, 6190−6197
Environmental Science & Technology
Article
Caustic Leaching of the Tank Waste Simulant. The simulant material was prepared as described in the Supporting Information. The simulant (19.20 g) was suspended in 104.27 g of DI water. With continuous stirring, 10.75 M NaOH (11.27 mL, 14.81 g) was added. The stirring was stopped after 30 min and the solids were settled by gravity. The supernatant leachate solution (108 g) was separated from the settled solids layer. The leached solids were washed three successive times with 105 mL each of DI water. Each washing step was conducted by mixing the washing water with the solids for 15 min, then letting the solids settle. The three washing solutions were collected separately. During the third washing step, it was necessary to raise the solution pH by adding 0.7 mL of 10.75 M NaOH to achieve complete settling of the solids. Lime Treatment of the Simulated Leaching Solution. The feed solution for the lime treatment portion of the simulant experiment was prepared by mixing proportionate amounts of the leachate and the three washing solutions. Calcium oxide (2.15 g) was added to 199 g of the feed solution and the mixture was stirred for 30 min. The supernatant liquid was separated after allowing the solids to settle by gravity. The solids were washed three successive times with 60 mL each of DI water. Each washing step was conducted as described above. Caustic Leaching of Actual Tank Waste. A portion of actual tank waste containing 2.6 g of water-insoluble solids (see Supporting Information) was suspended in 260 mL of DI water, then sufficient 10 M NaOH (18 mL) was added to provide a final NaOH concentration of ∼0.5 M. The total volume of leachate solution was adjusted to 280 mL with DI water. After stirring for 38 min, the leached solids were allowed to settle overnight. The leachate liquor was removed and the solids were subjected to three successive washes with DI water at 3× the settled solids volume. Each washing step was conducted as described above. As was the case with the simulant, the pH of the third washing liquid had to be raised by adding 1 mL of 10 M NaOH to achieve complete solids settling. Lime Treatment of the Actual Waste Leaching Solution. The phosphate-rich solution produced from leaching of the actual waste sample was combined in proportionate ratios with the three washing solutions to produce a total of 50 mL of solution for the lime treatment. Calcium oxide (4.64 g) was added to the combined leaching/washing solution with stirring. The amount of CaO added was estimated to be sufficient to yield a Ca:P molar ratio of ∼50, but subsequent analysis of the process solutions and solids indicated the Ca:P molar ratio to be ∼35. This might in part be explained by sorption of water by the CaO reagent. After stirring for 30 min, the solids were allowed to settle and the supernatant liquid was decanted. The Ca product solids were washed three times with DI water in a manner similar to that described above. In this case, it was not necessary to adjust the solution pH with NaOH to achieve complete settling of the solids in the final washing step.
Table 1. Composition of the Actual and Simulated Tank Waste actual tank waste component
conc., wt%
normalized to Fe, g/g Fe
Al Bi Fe Na P S Si U NO3−a NO2−a F−a PO43−a SO42−a
0.8 3.1 2.8 20.2 3.6 1.2 1.5 0.3 48.1 0.3 1.1 4.8 3.8
0.3 1.1 1.0 7.2 1.3 0.4 0.5 0.1 17.4 0.1 0.4 1.8 1.4
a
simulated tank waste conc., wt%
normalized to Fe, g/g Fe
10.3 22.1 5.3 2.1
1.00 2.15 0.52 0.21
Water-soluble fraction.
greater fractions of sodium phosphate and sodium sulfate salts. Ion chromatography analysis of a solution prepared by water leaching of the actual tank waste solids supports this supposition. The untreated solids contained 1.8 g of soluble phosphate (0.6 g soluble P) per gram of Fe, indicating 46% of the P in the waste was in a readily water-soluble form. By inference, the water-insoluble P fraction was 0.7 g/g Fe, which is more in line with the 0.5 g P/g Fe in the simulant material. Essentially all the S in the actual waste material was watersoluble with 1.4 g of soluble sulfate (0.5 g of soluble S) found per gram of Fe compared to 0.4 g of total S/g Fe found by ICPOES. Caustic Leaching. Table 2 presents the percent of each component found in the individual process solutions during the caustic leaching and washing of the simulated and actual tank waste. The distribution values were calculated based on the total amount of each component recovered in the leachate, washing solutions, and residual solids. An alternative method of determining the amount of P and S removed from the solids during the caustic leaching and washing steps is to compare the Fe-normalized concentrations in feed material to those in the residual solids. This method assumes that no significant amount of Fe is dissolved in the process, an assumption that is verified by the low Fe concentrations found in the simulant leachate and washing solutions (5.4 × 10−6 M Fe in the leachate; below detection in the washes). The results of both methods of analyzing the data indicate that the caustic leaching and washing steps effectively removed P and S from both the simulant and actual waste materials. The results of the actual waste and simulant tests are consistent. The caustic leaching and washing steps removed about 98% of the P from both the simulant and actual waste materials. One difference between the simulant and actual waste tests is the relative amount of P found in the leachate solution. In the case of the simulant test, only 31.4% of the P was found in the decanted leachate solution, whereas in the actual waste test 69.6% of the P was in the decanted leachate solution recovered. This can be attributed to the leaching step being performed under conditions oversaturated with respect to sodium phosphate during the simulant test, which was verified by solubility calculations. This is further evidenced by the enhanced amount of P in the first washing step in the simulant
■
RESULTS AND DISCUSSION Compositions of the Tank Waste Simulant and Actual Tank Waste. Table 1 compares the simulant and actual waste compositions. The actual waste matrix is more complicated than the simulant, with Al, Bi, Si, and U being present in significant quantities along with Fe, Na, P, and S. The Fenormalized concentrations for Na, P, and S in the actual waste are somewhat greater than in the simulant, perhaps reflecting 6192
dx.doi.org/10.1021/es300443a | Environ. Sci. Technol. 2012, 46, 6190−6197
Environmental Science & Technology
Article
Table 2. Behavior of Na, P, and S during the Caustic Leaching and Washing Steps of the Simulant (S) and Actual Tank Waste (TW) Tests component distribution, %a leachate
wash 1
component
S
TW
Na P S
68.5 31.4 81.1
69.3 69.6 72.9
S
wash 2 TW
wash 3
S
S
TW
S
TW
5.7 5.8 5.7
3.2 4.1 0.6
3.0 1.5 1.3
0.9 1.6 0.1
1.8 1.9 0.0
22.8 20.2 4.6 52.7 21.2 10.2 15.3 20.2 3.0 Fe-normalized conc., g/g Fe
initial sample
leached solids
TW
amount removed, %b
leached solids
component
S
TW
S
TW
S
TW
Na P S
2.15 0.52 0.21
7.31 1.31 0.43
0.086 0.014 0.00033
0.31 0.031 ND
96.0 97.3 99.8
95.7 97.6
a
Component distribution values were obtained by dividing the mass of the component in a given process stream by the sum of the masses for that component found across all the process streams, and multiplying by 100. bAmount removed determined as 100(Ci − Cf)/Ci, where Ci is the initial Fe-normalized concentration and Cf is the Fe-normalized concentration in the leached solids.
Table 3. Measured Concentrations of the Key Components in the Feed Solution for the Lime Treatment Portion of the Simulant (S) and Actual Tank Waste (TW) Tests and Distribution (%) of Ca, Na, P, and S during the Lime Precipitation and Washing Steps concn. in the feed solution, M
component distribution, %a P-depleted soln.
wash 4
wash 5
wash 6
Ca product solids
component
S
TW
S
TW
S
TW
S
TW
S
TW
S
TW
Ca Na P S
0.0 0.81 0.050 0.028
0.0 0.50 0.035 0.011
0.0 89.8 20.2 76.4
0.2 58.8 0.2 73.6
1.3 7.9 0.0 2.4
0.7 30.6 0.1 22.9
2.7 1.3 0.0 0.4
1.0 8.0 0.0 2.9
4.1 0.5 0.0 0.3
1.3 2.6 0.0 0.6
91.9 0.5 79.8 20.5
96.9 0.0 99.6 0.0
a
Component distribution values were obtained by dividing the mass of the component in a given process stream by the sum of the masses for that component found across all the process streams, and multiplying by 100.
simulant test contained only 80% of the P originally in the CaO treatment feed solution. This is considerably lower than the target P removal value of ≥95%. The Ca:P molar ratio during the CaO treatment was 4.0. Previous studies suggested that this should be sufficient to yield near quantitative removal of phosphate from solution.8 A likely explanation for the low removal of P is competition from other ions present in the simulant, especially fluoride and sulfate which form poorly soluble calcium salts in aqueous media. A fraction (20%) of the sulfate ion was removed from the feed solution by the CaO treatment. Furthermore, analysis of the solution for F− by IC before and after the CaO treatment indicated that 90% of the F− was removed during the CaO treatment of the simulant solution. Because of these apparent competing precipitation reactions, the Ca:P molar ratio was increased to ∼35 for the experiment with actual waste, resulting in greater than 99% recovery of the P in the Ca product solids. It was concluded that the additional Ca was needed to compensate for competing precipitation reactions, especially the precipitation of fluoride. Ion chromatography analysis of the feed solution before the CaO treatment indicated the fluoride concentration to be 0.024 M, but after treatment with CaO the F− concentration was 0.99 in all cases. The determined phosphate, sulfate, and nitrate concentrations are listed in Table 4, and also depicted in Figure 3 (the corresponding Raman spectra are presented in Figure S.4 in the Supporting Information). The analytical values derived from IC and ICP-
might suggest the incorporation of nitrate into the solids (ν3 and ν2, respectively). Comparison of the FTIR spectrum of the leached solids with that of the iron(III) phosphate phase introduced into the simulant showed the disappearance of the ν3(PO4) band at 1009 cm−1 in the leached solids (Figure S.3). This clearly indicates that the phosphate component of the simulant was removed from the solids by the leaching process, in agreement with the ICP-OES result. Raman and FTIR spectra of phases relevant to the NaOH leaching of the actual tank waste are shown in Figure 1a. Bands attributable to nitrate ion (ν3 at 1351 cm−1 and ν2 at 826 cm−1) are evident in the FTIR spectrum of the untreated tank waste sample; these bands are less pronounced after washing of the solids with water. A complex series of overlapping bands is evident in the PO stretching region of the FTIR spectrum of the untreated tank waste material, centered around 1000 cm−1. On the other hand, a very sharp nitrate ν1 symmetric stretch band15 is observed in the Raman spectrum of the untreated tank waste at 1050 cm−1 (Figure 1b). This band is completely missing in the Raman spectrum of the solids remaining after the NaOH leaching, indicating removal of nitrate from the solids, and is consistent with the FTIR analysis. The small band at 984 cm−1 in the Raman spectrum of the actual tank waste, attributed to the crystalline iron orthophosphate FePO4,16 was also absent after leaching. Likewise, the PO stretching bands in the FTIR spectrum are greatly diminished after the NaOH treatment, as are the nitrate bands. The latter observation is different from that for the simulant test in which there was some evidence for nitrate in the solids after leaching. The broad band at 810 cm−1 and the weak band at 602 cm−1 are attributed to the borosilicate glass vial in which the sample was held when making the measurement (refer to Figure S.2 in the Supporting Information). Phosphate Precipitation. As evident from Figure 2, there are significant differences between the spectra of the Ca solid products obtained in the simulant and actual waste tests. Comparison of the FTIR spectra of the Ca product solids from the simulant and hydroxyapatite suggests that the simulant product mostly consists of hydroxyapatite or perhaps fluoroapatite, indicated by the ν3(PO4) band at 1030 cm−1, (although this band is shifted somewhat from that observed for commercially procured hydroxyapatite at 1022 cm−1). In addition, there are bands at 1466, 1417, and 869 cm−1. These bands are most likely attributable to the presence of calcium carbonate, with the 1466 and 1417 cm−1 bands being associated with ν3(CO3) and the 869 cm−1 band assigned to the ν2(CO3) of CaCO3.17 Indeed, the FTIR spectrum of commercially procured CaCO3 displayed similar bands at 873 and 1392 cm−1, along with the ν4(CO3) band at 712 cm−1. FTIR examination of the starting CaO revealed similar spectral features with bands observed at 868, 1416, and 1462 cm−1, so this is a likely source of the calcium carbonate phase. The bands at 603 and 564 cm−1 in the FTIR spectrum of the calcium product solids can be assigned to splitting of the ν4(PO4) band in hydroxyapatite. In the Raman spectrum of the simulant Ca product solids the band observed at 963 cm−1 is identical to that for the ν1(PO4) band for commercially procured hydroxyapatite (Figure 2b). The very broad band at 1075 cm−1 was attributed to the ν1(CO3) in CaCO3 which was observed at 1087 cm−1 in the commercial CaCO3, in agreement with the previously reported value of 1087 cm−1.18 The broadness of the peak within the simulant suggests that this phase is of low crystallinity. The broad band at 810 cm−1 is attributed to the glass vial in which 6195
dx.doi.org/10.1021/es300443a | Environ. Sci. Technol. 2012, 46, 6190−6197
Environmental Science & Technology
Article
actual tank waste processing streams. The advantage of the Raman spectroscopic technique includes the ability to analyze the solutions with no sample pretreatment and the immediate analysis of data, compared to the time-consuming sample preparation and analysis required for the IC and ICP analytical methods.
■
ASSOCIATED CONTENT
S Supporting Information *
(a) A more extensive conceptual description of the phosphate removal and calcium stabilization process, (b) additional experimental descriptions regarding the simulant preparation and determination of the mass of water-insoluble solids used in the actual waste test, and (c) supplementary spectra. This information is available free of charge via the Internet at http:// pubs.acs.org/.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]; phone: (509)375-5696; fax: (509)375-5684 (G.J.L.); e-mail:
[email protected]; phone: (509)375-5646; fax: (509)375-5684 (T.G.L.); e-mail:
[email protected]; phone: (509)375-5648; fax: (509)3755684. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was funded by the U.S. Department of Energy through the Office of Environmental Management. Pacific Northwest National Laboratory is operated by Battelle Memorial Institute for the U.S. Department of Energy under contract DE-AC05-76RL01830.
■
REFERENCES
(1) Certa, P. J.; Wells, M. N. River Protection Project System Plan; ORP-11242, Rev. 5; Office of River Protection: Richland, WA, 2010. (2) National Research Council. Research Needs for High-Level Waste Stored in Tanks and Bins at U.S. Department of Energy Sites; National Academy Press: Washington, DC, 2001. (3) Swanson, J. L. Clean Option: An Alternative Strategy for Hanford Tank Waste Remediation; Detailed Description of First Example Flowsheet. In Chemical Pretreatment of Nuclear Waste for Disposal; Schulz, W. W., Horwitz, E. P., Eds.; Plenum Press: New York, 1994; pp 155. (4) Lumetta, G. J.; Rapko, B. M.; Liu, J.; Temer, D. J. Enhanced sludge washing for pretreating Hanford tank sludges. In Science and Technology for Disposal of Radioactive Tank Wastes; Schulz, W. W., Lombardo, N. J., Eds.; Plenum Press: New York, 1998; pp 203. (5) Lumetta, G. J.; Darnell, L. P.; Garza, P. A.; Greenwood, L. R.; Oliver, B. M.; Rinehart, D. E.; Sanders, D. R.; Soderquist, C. Z.; TrangLe, T.; Urie, M. W.; Wagner, J. J. Caustic Leaching of Hanford Tank T-
Figure 3. Comparison of phosphate, nitrate, and sulfate concentrations in the caustic leaching and washing solutions determined by Raman spectroscopy to those determined by ICP-OES and IC.
OES were plotted against the Raman-determined values in Figure 3, showing excellent agreement between the two analytical methods. As shown in this figure, concentration values obtained by the Raman method are in excellent agreement with those determined by ICP-OES and IC for the leachate and washing solutions. These results show the applicability of the Raman spectroscopic method for the quantitative determination and monitoring of the oxyanions in
Table 4. Comparison of Solution Phosphate, Nitrate, and Sulfate Concentrations Determined by Raman Spectroscopy with Those Determined by ICP-OES and IC phosphate concn., mol/L solution leaching solution leached solids, wash 1 leached solids, wash 2 leached solids, wash 3 a
Raman 9.5 2.9 5.5 2.4
× × × ×
10−2 10−2 10−3 10−3
ICP-OES 1.0 3.0 7.4 2.2
× × × ×
10−1 10−2 10−3 10−3
nitrate concn., mol/L IC
9.3 2.8 6.4 2.3
× × × ×
10−2 10−2 10−3 10−3
Raman 6.1 1.6 3.6 8.6
× × × ×
10−1 10−1 10−2 10−3
IC 6.0 1.6 3.7 1.9
× × × ×
10−1 10−1 10−2 10−3
sulfate concn., mol/L Raman 3.3 × 10−2 9.1 × 10−3 1.7 × 10−3 nda
ICP-OES 3.5 9.5 2.3 5.9
× × × ×
10−2 10−3 10−3 10−4
IC 3.0 7.6 1.9 4.8
× × × ×
10−2 10−3 10−3 10−4
Not detected. 6196
dx.doi.org/10.1021/es300443a | Environ. Sci. Technol. 2012, 46, 6190−6197
Environmental Science & Technology
Article
110 Sludge; PNNL-13956; Pacific Northwest National Laboratory: Richland, WA, 2002. (6) Lumetta, G. J. Mechanism of Phosphorus Removal from Hanford Tank Sludge by Caustic Leaching; PNNL-17257 (WTP-RPT-173); Pacific Northwest National Laboratory: Richland, WA, 2008. (7) Lumetta, G. J.; Buck, E. C.; Daniel, R. C.; Draper, K.; Edwards, M. K.; Fiskum, S. K.; Hallen, R. T.; Jagoda, L. K.; Jenson, E. D.; Kozelisky, A. E.; MacFarlan, P. J.; Peterson, R. A.; Shimskey, R. W.; Sinkov, S. I.; Snow, L. A. Characterization, Leaching, and Filtration Testing for Bismuth Phosphate Sludge (Group 1) and Bismuth Phosphate Saltcake (Group 2) Actual Waste Sample Composites; PNNL-17992 (WTP-RPT-166); Pacific Northwest National Laboratory: Richland, WA, 2009. (8) Lumetta, G. J.; Felmy, A. R.; Braley, J. C.; Carter, J. C.; Edwards, M. K.; MacFarlan, P. J.; Qafoku, O. Removing Phosphate from Hanford High-Phosphate Tank Wastes: FY 2010 Results; PNNL-19778; Pacific Northwest National Laboratory: Richland, WA, 2010. (9) Lumetta, G. J.; McNamara, B. K.; Buck, E. C.; Fiskum, S. K.; Snow, L. A. Characterization of high phosphate radioactive tank waste and simulant development. Environ. Sci. Technol. 2009, 43, 7843− 7848. (10) Stark, K.; Hultman, B. Phosphorus Recovery by One- or TwoStep Technology with Use of Acids and Bases. In Proceedings of the International Water Association (IWA) Specialist Conference, BIOSOLIDS 2003: Wastewater Sludge as a Resource; Trondheim, Norway, June 23−25, 2003, Department of Land and Water Resource Engineering, Royal Institute of Technology, Stockholm, Sweden, 2003; pp 281. (11) Bryan, S. A.; Levitskaia, T. G.; Sinkov, S. I. Process Monitor Development Project Acceptance Test Report; PNNL-15360; Pacific Northwest National Laboratory: Richland, WA, 2005. (12) Schonewill, P. P.; Edwards, M. K.; Smith, C.; Tranbarger, R.; Shimskey, R. W.; Peterson, R. A. Integrated Near Tank Treatment System Pilot Demonstration; PNWD-4298; BattellePacific Northwest Division: Richland, WA, 2011. (13) Kurath, D. E.; Hanson, B. D.; Minette, M. J.; Baldwin, D. L.; Rapko, B. M.; Mahoney, L. A.; Schonewill, P. P.; Daniel, R. C.; Eslinger, P. W.; Huckaby, J. L.; Billing, J. M.; Sundar, P. S.; Josephson, G. B.; Toth, J. J.; Yokuda, S. T.; Baer, E. B. K.; Barnes, S. M.; Golovich, E. C.; Rassat, S. D.; Brown, C. F.; Geeting, J. G. H.; Sevigny, G. J.; Casella, A. J.; Bontha, J. R.; Aaberg, R. L.; Aker, P. M.; Guzman-Leong, C. E.; Kimura, M. L.; Sundaram, S. K.; Pires, R. P.; Wells, B. E.; Bredt, O. P. Pretreatment Engineering Platform Phase 1 Final Test Report; PNNL-18894 (WTP-RPT-197 Rev 0); Pacific Northwest National Laboratory: Richland, WA, 2009. (14) Bryan, S.; Levitskaia, T.; Schlahta, S. Raman based process monitor for continuous real-time analysis of high level radioactive waste components. In Proceedings of the WM2008 Conference, February 24 −28, 2008: Phoenix, AZ, 2008. (15) Nakagawa, I.; Walter, J. L. Optically active crystal vibrations of the alkali-metal nitrates. J. Chem. Phys. 1969, 51, 1389−1397. (16) Zhang, L.; Brow, R. K. A Raman study of iron-phosphate crystalline compounds and glasses. J. Am. Ceram. Soc. 2011, 94, 3123− 3130. (17) Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordination Compounds, 5th ed.; John Wiley & Sons: New York, 1997. (18) Porto, S. P. S; Giordmaine, J. A.; Damen, T. C. Depolarization of Raman scattering in calcite. Phys. Rev. 1966, 147, 608−611. (19) Waterland, M. R.; Stockwell, D.; Kelley, A. M. Symmetry breaking effects in NO3−: Raman spectra of nitrate salts and ab initio resonance Raman spectra of nitrate-water complexes. J. Chem. Phys. 2001, 114, 6249−6258. (20) Jastrzebski, W.; Sitarz, M.; Rokita, M.; Bułat, K. Infrared spectroscopy of different phosphates structures. Spectrochim. Acta Part A 2011, 79, 722−727. (21) Beebe, K. R.; Pell, R. J.; Seasholtz, M. B. Chemometrics: A Practical Guide; John Wiley & Sons: New York, 1998.
(22) Sharaf, M. A.; Illman, D. L; Kowalski, B. R. Chemometrics; John Wiley & Sons: New York, 1986. (23) Long, G. L.; Winefordner, J. D. Limit of Detection. Anal. Chem. 1983, 55, A712−A724.
6197
dx.doi.org/10.1021/es300443a | Environ. Sci. Technol. 2012, 46, 6190−6197
