Microscopic and Spectroscopic Characterization of Aluminosilicate Waste Form with Cs/Sr/Ba Loading Using Scanning Electron Microscopy, Transmission Electron Microscopy, and X-Ray Diffraction GARY CEREFICE, LONGZHOU MA, and MICHAEL KAMINSKI An aluminosilicate waste form has been proposed for the storage and disposal of cesium and strontium isolated from recycled nuclear fuel. To examine the impact of sintering temperature on the waste form product, thermal analysis (thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC)) was used to identify key transition temperature ranges. Samples were produced in each temperature range to examine the impact on phase formation and microstructure. Examination of the synthesized materials by X-ray diffraction (XRD) confirmed the formation of the expected Cs- and Sr-aluminosilicate crystalline phases. However, microscopic characterization by scanning electron microscopy (SEM) revealed a spongelike, glassy morphology with high porosity and no observed crystallinity. This discrepancy was investigated by transmission electron microscopy (TEM) and high-resolution TEM (HRTEM), which identified the presence of discrete, submicron, crystalline phases within the bulk amorphous matrix. Elemental analysis by energy-dispersive X-ray (EDX) indicated that the strontium and barium were incorporated into the crystalline phase, while the cesium was incorporated into the amorphous matrix. Further analysis of samples synthesized without barium or strontium allowed for the identification of submicron crystalline phases within the amorphous matrix, identifying the source of the cesium aluminosilicate crystal peaks in the XRD patterns, with elemental analysis showing that the cesium was present in both the crystalline inclusions and the amorphous bulk phase. DOI: 10.1007/s11661-009-0038-4 Ó The Minerals, Metals & Materials Society and ASM International 2009

I.

INTRODUCTION

ONE strategy to maximize the use of a high level waste repository is the separation of the cesium and strontium from the high level waste stream for decay storage prior to disposal, reducing the short-term heat load in a geological repository facility.[1,2] The isolated cesium and strontium waste stream would be immobilized prior to decay storage, with the decay storage form intended to eventually be the disposal form.[3,4] While decay storage is ideal for managing the radiological and thermal concerns regarding a separated cesium and strontium waste stream, it creates a new concern. Cesium decays to produce barium, a hazardous component under the Resource Conservation and Reclamation Act (RCRA), which would require that the waste GARY CEREFICE and LONGZHOU MA, Research Scientists, are with the Harry Reid Center for Environmental Studies, University of Nevada, Las Vegas, NV 89154. Contact e-mail: [email protected]. edu MICHAEL KAMINSKI, Staff Scientist, is with the Chemical Sciences and Engineering Division, Argonne National Laboratory, Argonne, IL 60439. This article is based on a presentation given in the symposium ‘‘Materials for the Nuclear Renaissance,’’ which occurred during the TMS Annual Meeting, February 15–19, 2009, in San Francisco, CA, under the auspices of Corrosion and Environmental Effects and the Nuclear Materials Committees of ASM-TMS. Article published online September 15, 2009 2876—VOLUME 40A, DECEMBER 2009

form be treated as a mixed waste at the time of disposal unless it meets the leaching requirements under RCRA at the time of disposal (300 years or more in the future).[1,5] Aluminosilicate waste forms have been proposed for the storage and ultimate disposal of an isolated cesium and strontium waste stream.[6,7] Initial studies have indicated that bentonite clay mixed with cesium chloride (CsCl) could form the crystalline compounds such as cesium aluminum silicate (pollucite, CsAlSiO4) with acceptable stability and low leach rate.[6] However, there is competition during synthesis of waste form using bentonite clay and Cs-included compounds. There is a tendency to form a glass phase and crystalline compound, which is beneficial to host the Cs waste. The previous study only investigated the Cs-loaded waste form synthesis.[6,7] The distribution of cesium, strontium, and barium within waste form is still unknown. The goal of this study is to identify the behavior of the projected waste stream components (Cs, Sr, and Ba) in the proposed aluminosilicate-based waste form. In this study, the aluminosilicate waste forms with and without Cs/Sr/Ba loading were synthesized from bentonite clay at different sintering temperatures to examine the impact of fabrication temperature on waste form. Microscopic and spectroscopic characterization of the synthesized METALLURGICAL AND MATERIALS TRANSACTIONS A

MATERIALS AND EXPERIMENTS

A. Materials

95.0

Onset: 1150°C

TGA DSC

85.0 0

200

400

B. Aluminosilicate Waste Form Synthesis Procedure To identify the sintering temperatures for this study, thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) measurements were conducted using the base bentonite clay as well as the synthetic waste stream loaded material. The results are exhibited in Figure 1. As seen in Figure 1(a), for the as-received bentonite clay, there were notable mass changes (7.19 and 5.75 pct) occurring at temperatures of around 400 °C and 1000 °C when the sample was heated from room temperature to 1200 °C, suggesting the phase transform and constituent water loss. In the accompanying DSC curve (Figure 1(a)), exothermal peaks could be observed at temperatures of 645 °C and 1150 °C, corresponding to the mass changes. Compared to the as-received material, the waste-loaded bentonite clay displayed significantly different thermal behaviors, as shown in Figure 1(b), suggesting that the addition of the waste to the matrix results in significant structural changes. When the waste-loaded bentonite clay was heated from room temperature to 1200 °C, a slight mass change of 0.28 pct was observed at a temperature of around 450 °C. Upon further heating of the sample, another tiny mass change of 0.03 pct at a temperature of 1050 °C could be observed. Transitional exothermal peaks around temperatures of 773 °C and 1068 °C were METALLURGICAL AND MATERIALS TRANSACTIONS A

600

800

1000

1200

Temperature, °C

(a) TAG, Mass Change (MC), %

mesh), was obtained from the American Colloid Company. The composition of the as-received materials was determined by X-ray fluorescence to be (wt pct) 55.2SiO2, 18.0Al2O3, 3.84Fe2O3, 2.68MgO, 1.84Na2O, 3.34CaO, 0.59K2O, 0.28S, and 0.15TiO2, with additional trace phases. The initial water content of the clay material was determined to be approximately 13 to 15 mass pct using loss-on-ignition analysis. The synthetic waste stream concentrate solution was prepared from cesium nitrate (Cs(NO3), Alfa Aesar, 99.99 pct), barium nitrate (Sr(NO3)2, Alfa Aesar, 99.999 pct), and strontium nitrate (Ba(NO3)2, Alfa Aesar, Ward Hill, MA, 99.9965 pct). The synthetic waste stream concentrate solution was prepared by dissolving the nitrate salts in deionized water, at a final concentration of (g/L): 75.26 Cs(NO3), 37.28 Sr(NO3)2, and 98.53 Ba(NO3)2 .

-100.0 -250.0 -400.0 -550.0

MC:5.75%

90.0

DSC, x10-3 uV

MC:7.19%

Commercial bentonite clay, VOCLAY*-325 (325 *VOCLAY is a trademark of American Colloid Company, Hoffman Estates, IL.

500.0 350.0 200.0 50.0

Onset: 645°C

100.0

500 400 300 200

100.5 Onset: 1068.4°C

MC:0.28%

Onset: 773.3°C

100.0

100 0 -100 -200 -300

99.5 TGA DSC

MC:-0.03%

99.0 0

200

400

600

800

1000

DSC, x10-3 uV

II.

TGA, Mass Change (MC), %

waste forms was conducted using scanning electron microscopy (SEM), transmission electron microscopy (TEM), and X-ray diffraction (XRD) to understand the in-depth interaction of waste from Cs/Sr/Ba loading during sintering as well as to detect the Cs-hosting site within the synthetic aluminosilicate waste.

1200

Temperature, °C

(b) Fig. 1—Measurements of TGA and DSC for as-received and wasteloaded samples: (a) TAG and DSC of as-received bentonite clay and (b) TAG and DSC of bentonite clay with Cs/Sr/Ba waste loading.

observed in the DSC curve, as shown in Figure 2(b). Based on the TGA and DSC measurements of both the as-received bentonite clay and the waste-loaded material, two potential sintering temperatures were identified for further investigation, 800 °C and 1000 °C. These temperatures were selected as potential sintering temperatures for this work. Both temperatures were used to examine the impact of sintering temperature on the final products. The general procedure to synthesize aluminosilicate waste forms followed the method proposed by Strachan and Schulz in study.[6] The as-received bentonite clay was mixed with the synthetic waste stream concentrate solution with a solid-to-liquid ratio of 3:4 (g:mL), ensuring over 20 pct waste loading. After mixing, the resulting materials were heated at 90 °C for 24 to 48 hours to remove excess water. The oven-dried material was ground by hand to a powder and pressed into green pellets using a 13-mm die at 35 MPa (approximately 2 to 4 grams per pellet). The green pellets were sintered at either 1000 °C (sample 3) or 800 °C (sample 4), depending on the experiment, and held at temperature for 48 hours. To explore the behavior of the base material without waste loading, the preceding procedure was modified by replacing the synthetic waste stream concentrate solution with deionized water. The resulting product was sintered at 1000 °C (sample 1) and 800 °C (sample 2) for 48 hours at temperature. To confirm the distribution of the cesium in the final waste form, a third batch of samples was prepared, replacing the synthetic waste stream concentrate solution by a cesium nitrate solution VOLUME 40A, DECEMBER 2009—2877

Fig. 2—SEM micrographs of samples 1 and 2: (a) SEM topographical image of sample 1, (b) SEM image of polished sample 1, and (c) SEM topographical image of sample 2.

Table I. Sample 1 2 3 4 5 6

Summary of Sample ID and Synthesis Parameters Sintering Temperature (°C)

Sintering Time (h)

Waste Loading

1000 800 1000 800 1000 800

48 48 48 48 48 48

no no Cs, Sr, and Ba Cs, Sr, and Ba Cs Cs

of the same cesium concentration. Sample 5 was prepared by sintering the cesium-only loaded matrix at 1000 °C, and sample 6 was the same material sintered at 800 °C. These samples were also held at temperature for 48 hours. The sample identification number and synthesis procedure are summarized in Table I. C. Characterization Methods 1. X-ray diffraction The synthesized materials were characterized by X-ray powder diffraction using a PANalytical X’Pert PRO X-ray diffractometer with a Cu Ka radiation (40 kV, 40 mA) and a multiple-strip solid-state detector (X’Celerator, PANalytical Company, Almelo, The Netherlands). The sample was prepared by suspending 2878—VOLUME 40A, DECEMBER 2009

the grounded sample in ethanol to make a slurry, which was spread in a thin layer on a low-background silicon sample holder. The patterns were recorded at room temperature with step sizes of 0.017 deg, 2h, and 46 seconds per step. The phase constitution was characterized using the International Center for Diffraction Database for powder diffraction data. 2. Microscopy The morphology of the synthesized samples was studied by SEM and TEM. The SEM imaging was performed on a JEOL** scanning electron microscope **JEOL is a trademark of Japan Electron Optics Ltd., Tokyo.

model JSM-5610 equipped with secondary electron (SE) and backscattered electron (BE) detectors and an Oxford ISIS EDS system (Oxford Instruments, Oxfordshire, UK). The acceleration voltage used in SEM was 15 kV. Using the SEM BE and SE modes, the samples were broken into halves, and the cross section of samples was observed using SEM. In order to improve the conductivity and image contrast, the samples were coated with a layer of gold. Gold-coated samples mounted on double-sided carbon tapes were used to investigate the topography of the samples. The other METALLURGICAL AND MATERIALS TRANSACTIONS A

half piece of one broken sample was mounted using epoxy into a mold to make the polished sample with a diameter of 30 mm and height of 40 mm. The mounted ∆

♦: Beidellite; ∆: Quartz; o: Sodium Aluminum Silicate ⊗: Potassium Aluminum Iron Phosphate; ×: Aluminum Phosphate

Intensity, Arbitrary Units

∗: Titanium Vanadium Oxide; +: Potassium Titanium Phosphate @: Sodium Aluminum Silicate Hydroxide @@ ∆

@

∆

∆ ×

o

×

×

+

Sample #1

×

×

∗ ♦

10

20

∆

⊗

♦

30

Sample #2

40

×

∆

50

+

∆

As-received ♦

∆

60

∆

70

2*Theta, degree Fig. 3—XRD patterns of as-received betonite clay powder: sample 1 (sintered at 1000 °C) and sample 2 (sintered at 800 °C).

sample was ground first and then polished using 1.0-lm alumina sandpaper, ensuring that the cross section of the sample was flat. These polished samples were observed using SEM. A TECNAI-G2-F30 transmission electron microscope with a 300 keV field emission gun was used to characterize the samples. Samples were analyzed using the conventional bright-field (BF) mode and HRTEM mode. All TEM images were recorded using a slow scan CCD camera attached to a Gatan GIF 2000 image filter (Gatan Inc., Pleasanton, CA). Localized fast Fourier transformed (FFT) micrographs were also used in the analysis of the selected area electron diffraction (SAED) patterns and HRTEM images. The elemental distribution of each sample was also determined using the corresponding X-ray energy dispersive spectrometry (EDX) of the scanning transmission electron microscopy (STEM) mode. For the STEM/EDX mode, the electron probe with a size of 0.2 nm was used to examine the dedicated area of the sample. The TEM samples were prepared by a solution-drop method. Two to five

Fig. 4—TEM BF and HRTEM micrographs of samples 1 and 2: (a) BF image of sample 1, (b) corresponding HRTEM image of (a), (c) BF image of sample 2, and (d) corresponding HRTEM image of (b). METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 40A, DECEMBER 2009—2879

Fig. 5—SEM micrographs of Cs/Sr/Ba-loaded samples: (a) sample 3 sintered at 1000 °C and (b) sample 4 sintered at 800 °C.

40% ♦: Beidellite; ∆: Quartz; o: Sodium Aluminum Silicate ⊗: Potassium Aluminum Iron Phosphate; ×: Aluminum Phosphate

30% 20%

SSample am ple#3 #3

∗: Titanium Vanadium Oxide; +: Potassium Titanium Phosphate

SSample am ple#2 #2

@: Sodium Aluminum Silicate Hydroxide; ◊: Cesium Aluminum Silicate ©: Barium Strontium Aluminum Silicate; ®: Barium Aluminum Silicate; #: Strontium silicate

SSample am ple#1 #1

10% 0% 0

1

2

3

4

5

Average Diameter of Particle, um Fig. 6—Normal distribution of measured pore size.

milligrams of the sample material was ground by hand in a mortar and pestle and added to 5.0 mL of reagent-grade methanol. This mixture was agitated in an ultrasonic water bath for 5 minutes to form a homogeneous colloidal suspension. One drop of the suspension was placed onto a 3-mm-diameter carbon-coated copper grid using a small-tipped transfer pipette. The solution was evaporated from the sample at room temperature, leaving the fine particulate sample deposited on the carbon film, which was then used in the TEM observation.

III.

RESULTS AND DISCUSSION

A. Synthesized Aluminosilicate without Waste Loading Samples of the as-received bentonite clay, without Cs/Sr/Ba loading, were sintered at 800 °C and 1000 °C. The resulting products are brown in color. The samples were divided by hand to allow the examination of the sample cross section by SEM. The SEM micrographs (Figures 2(a) and (c)) show that the synthesized materials are highly porous and appear to be without strongly defined grain structure. Increasing the sintering temperature appears to increase the fraction of porous structure, as shown in Figure 2(a). To closely examine the morphology of the porous structure and matrix, sample 1 was mounted and polished from cross section 2880—VOLUME 40A, DECEMBER 2009

Intensity, Arbitrary Units

Distribution, %

SSample am ple#4 #4

®

∆ ◊ ®◊

◊

# ◊ ®

◊ ◊ ◊ © © ©

©

∆

×

∆

◊

©

#

®

Sample #2

@

×

×

+

Sample #1

×

10

20

∆

Sample #3

◊

©

×

∆ ♦

⊗

♦

30

×

×

∗ ♦

Sample #4

◊

®

o@

∆

@

◊

40

+

∆

As- received ∆

50

60

∆

70

2*Theta, degree Fig. 7—XRD patterns of as-received betonite powder and synthesized samples with and without waste loading.

using 1.0- lm fine alumina polishing paper; the polished sample was observed using SEM with the SE mode. The SEM image of the polished sample shows numerous porous structures and the matrix appears glassy, and some small inclusions exist mixed with a ‘‘glassy’’ bulk phase. For the simplified synthetic procedure used in this work, given the water content of the starting clay (approximately 13 to 15 mass pct), the resulting porosity of the sintered product is not surprising.[8] Augmenting the low-temperature drying process with a calcination step prior to sintering should decrease the porosity of the final product.[8,9] Figure 3 presents the typical XRD patterns of as-received bentonite powder and the sintered samples. The XRD pattern of as-received materials shows the METALLURGICAL AND MATERIALS TRANSACTIONS A

well-defined sharp diffraction peaks, which were identified as the contribution of complex phases including bediellite (Na0.3Al2(Si, Al)4O10(OH)2Æ2H2O), montmorillonite ((Na0.3(Al, Mg)2Si4O10(OH)2 Æ8H2O), potassium aluminum iron phosphate ((KAlFe)PO4), quartz (SiO2), sodium aluminum silicate (Na1.15Al1.15Si0.85O4), and titanium vanadium oxide (Ti0.92V0.08O). The XRD pattern of samples 1 and 2 confirmed the TGA and DSC results, which indicated that sintering the as-received bentonite clay at 800 °C and 1000 °C would introduce the phase transformation. The major phase constituents are indexed in Figure 3. Compared to the XRD spectrum of as-received materials, the XRD patterns of samples 1 and 2 show the remarkably decreased intensity as well as a broadened background, which could be due to the amorphous glassy bulky phase shown in Figure 2. Comparing the XRD patterns from the sintered materials with the as-received materials shows significant differences corresponding to

phase transformations. It should be noted that the peak, identified as quartz, was detected in all three samples, suggesting that quartz has been preserved during sintering. Relative to sample 1, the XRD pattern of sample 2 shows the remarkable reduction in intensity and numbers of diffraction peak, broad Bragg peaks, and strong background, suggesting sample 2 sintered at 800 °C has a decreased degree of crystallinity. Based on XRD spectrum, sample 2 was identified including sodium aluminum silicate hydroxide (Na2Al2(Al2Si2)O10(OH)2) and quartz (SiO2) and sample 1 was identified containing aluminum phosphate (Al(PO4)), potassium titanium phosphate (KTi2(PO4)3), and quartz (SiO2). The XRD spectra of the synthesized materials indicate that the material is a multiphase system including crystalline and glassy amorphous phase, which produce the sharp diffraction peaks and background, respectively. Increasing sintering temperature should produce more crystalline phases.

Fig. 8—TEM micrographs of samples with waste loading: (a) TEM BF image of sample 3, (b) corresponding HREM of (a), (c) TEM BF image of sample 4, and (d) corresponding HRTEM image of (c). METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 40A, DECEMBER 2009—2881

The crystalline peaks in XRD spectrum are likely to be contributed from the tiny inclusions, as shown in the SEM images. Figure 4 presents the TEM observation for samples 1 and 2, wherein TEM BF and HRTEM images confirm that the synthesized materials consist of numerous inclusions and a glassy amorphous matrix. In Figure 4, TEM BF images (a) and (c) show numerous nanosized inclusions with average dimensions of 60 nm in length and 30 nm in width. The HRTEM and embedded FFT images show that these inclusions are well-defined crystalline and the matrix is amorphous, which would account for the degree of crystallinity and amorphous background in the XRD pattern.

image analysis to generate the distribution of porosity. Figure 6 plots the normal distribution of the measured pore size. Most pores have an average size of 1.0 lm, and the density increases with sintering temperature and Cs/Sr/Ba loading. These results are likely the result of residual water and nitrate from the loading process. Calcining the material at higher temperatures prior to pressing the green pellets would likely address the low porosity of the product.[9] Figure 7 presents the comparison of the XRD patterns for the as-received samples and synthesized samples with and without waste loading. As expected, comparing the XRD patterns of the waste-loaded and sintered as-received samples shows significantly different diffraction patterns, confirming the formation of different phases due to the incorporation of the waste stream components. However, both patterns show the same broad, amorphous background, suggesting similar behavior in terms of crystalline inclusions forming in a glassy substrate, which is confirmed by the SEM images. It is interesting to note that the quartz peak in XRD patterns of samples 3 and 4 disappeared and existed in patterns of samples 1 and 2, suggesting that quartz would interact with waste solution to modify the phase component of staring materials during sintering. Based on the XRD patterns, sample 3 sintered at 1000 °C contained barium strontium aluminum silicate (Ba0.75Sr0.25Al2Si2O8) and cesium aluminum silicate (or pollucite, CsAlSiO4). Sample 4 sintered at 800 °C contained cesium aluminum silicate, barium aluminum silicate, Ba(Al2Si2O8), and strontium silicate,

B. Synthesized Aluminosilicate with Cs/Sr/Ba Loading As-received bentonite clay with Cs/Sr/Ba loading was sintered at 800 °C and 1000 °C to form the waste-loaded aluminosilicate matrix. The resulting products are white in color, indicating that the final product is significantly different from the nonloaded samples. Based on the initial evaluation of this solidification process,[6] a crystalline product was expected after sintering at the target temperature. However, as shown in Figure 5, SEM micrographs of the synthetic Cs/Sr/Ba-loaded waste forms (samples 3 and 4) show that the resulting products are highly porous, amorphous materials with a spongelike morphology, again most likely due to the water content in the starting material. The SEM images for the porous structure of materials were subjected to

(a)

3000 O

C ounts

(b) (b )

Mg

2500 2000 1500 Al

1000

Ba

500

Sr

0 0

5

10

15

Energy, kev 1000

Si

(c) (c)

Counts

800 600

O Cs Al

400 200

Fe

Cs

0 0

5

10

15

20

25

30

35

Energy, kev Fig. 9—STEM imaging and EDX spectrum of sample 3: (a) STEM image showing the location of EDX examination, (b) EDX results of inclusion, and (c) EDX results of substrate. 2882—VOLUME 40A, DECEMBER 2009

METALLURGICAL AND MATERIALS TRANSACTIONS A

Sr4(Si4O12). The XRD results show that all waste element containing phases include silicon and oxygen, suggesting that quartz was the major phase from the starting material involved in the phase transformations. At 1000 °C, barium and strontium appear to form a mixed aluminosilicate phase (Ba0.75Sr0.25Al2Si2O8), which was not observed at 800 °C, where the barium and strontium partitioned into separate phases. Figure 8 presents TEM micrographs of the synthetic Cs/Sr/Ba-loaded waste form materials (samples 3 and

4). The TEM BF images and SAED patterns show that the waste-loaded samples contain numerous nanosized rectangular inclusions with an average length of 35 nm and width of 15 nm, as shown in Figures 8(a) and (c). The corresponding HRTEM imaging of the samples identified these nanocrystalline inclusions in the amorphous matrix, as seen in Figures 8(b) and (d), potentially explaining the crystalline peaks observed in the XRD patterns. The HRTEM images also demonstrate that the inclusion is crystalline, showing the well-defined 700

(a)

Si

600

Counts

500

(b)

O

400 300 Al

200

Ba Sr

100

Sr

0 0

5

10

15

20

Energy, kev 900 Si

800

(c)

700 O

Counts

600 500 400 300

Al

200

Cs

100

Fe

Cs

0 0

10

20

30

40

Energy, kev Fig. 10—STEM imaging and EDX spectrum of sample 4: (a) STEM image showing the location of EDX examination, (b) EDX results of inclusion, and (c) EDX results of substrate.

Fig. 11—SEM micrographs of samples 5 and 6: (a) sample 5 with only Cs loading and sintering at 1000 °C and (b) sample 6 with only Cs loading and sintering at 800 °C.

METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 40A, DECEMBER 2009—2883

lattice and symmetric spots in the FFT pattern. The HRTEM images as well as the accompanying FFT show that the matrix is amorphous.

Intensity, Arbitrary Units

∆

◊

To examine the elementary distribution of substrates and inclusions, STEM imaging coupled with EDX nanoprobe analysis was performed. Figure 9 shows the STEM images and EDX spectrums of sample 3. As shown in Figure 9(a), the STEM image shows a bright contrast of samples and particles and a dark contrast of the carbon film background. The particles including higher atomic number usually have stronger contrast due to the large cross section of electron scattering. During STEM imaging in this microscope, the electron beam has a probe size of 0.2 nm, allowing EDX examination of the individual inclusions with minimal interference from the amorphous background. As seen in Figures 9(b) and (c) of sample 3, the strontium and barium peaks were detected and identified in the nanocrystalline inclusions, but not in the amorphous matrix. Cesium was not identified in any of the nanoinclusions analyzed in these samples, but was found in the amorphous matrix. Figure 10 demonstrates the STEM image and EDX results of sample 4 with waste loading and sintering at 800 °C. Similarly, strontium and barium were found in the nanocrystalline inclusions and cesium was detected in the amorphous matrix, as seen in Figures 10(b) and (c).

∆: Quartz; @: Sodium Aluminum Silicate Hydroxide; ◊: Cesium Aluminum Silicate; ©: Barium Strontium Aluminum Silicate; ®: Calcium Aluminum Oxide Carbonate ◊ Sulfide Hydrate; #: Strontium Titanium Silicate; ⊗: Albite; ∗: Cesium Magnesium Aluminum Silicate ∆

◊ ◊

⊗

◊⊗

⊗ ⊗ ∆ ⊗

®

◊

®

20

∆

⊗ ◊

∆

Sample #5

∆

∆

∆

∆

∆

∗∗

Sample #6

⊗

◊ ∆

ר #

©◊ © ©◊ © ◊

©

10

∆

⊗

30

◊

#

© ◊

©

#

Sample #4

◊

Sample #3

◊

40

50

60

70

2*Theta, degree

Fig. 12—XRD patterns of synthesized samples.

180

(a)

(b) (b )

Al

160 140

Counts

120 Si

100 80 60

Cs -L Cu

40 Fe

Cs -K

20 0 1

11

21

31

Energy, kev

O

140

(c) (c)

120

Counts

100 Al

80

Cs -L

60 Si 40

Fe

20

Cs -K

0 2

7

12

17

22

27

32

Energy, kev

Fig. 13—STEM imaging and EDX spectrum of sample 5: (a) STEM image showing the location of EDX examination, (b) EDX results of inclusion, and (c) EDX results of substrate. 2884—VOLUME 40A, DECEMBER 2009

METALLURGICAL AND MATERIALS TRANSACTIONS A

In combination, these results, including the XRD pattern, STEM/EDX, and HREM data, suggest that the barium and strontium are segregated into submicron particles of separate structures, while cesium is at least partially incorporated into the glassy substrate matrix.

Crystalline inclusions incorporating cesium were not observed in the samples prepared from the Cs/Sr/Baloaded material, but they were identified in XRD patterns, suggesting that the cesium may be partitioning between the crystalline inclusions and glassy matrix 60

(a)

(b)

Fe Cs-L

50

Ca

Counts

40

Cu

30 20 10

Cs-K Fe

0 2

12

22

32

Energy, kev 80

(c) 60

Counts

Cs-L Fe

40 Ca

20 Cs-K

0 2

12

22

32

Energy, kev Fig. 14—STEM imaging and EDX spectrum of sample 6: (a) STEM image showing the location of EDX examination, (b) EDX results of inclusion, and (c) EDX results of substrate.

Fig. 15—TEM micrographs of sample 5 with Cs loading and sintering at 1000 °C: (a) TEM BF image and (b) corresponding HRTEM image and FFT. METALLURGICAL AND MATERIALS TRANSACTIONS A

VOLUME 40A, DECEMBER 2009—2885

phase. The quartz in the as-received bentonite clay was the major consumable component manifesting itself through interacting with loaded waste elements to form the new waste-containing phases during sintering. C. Synthesized Aluminosilicate Only with Cs Loading To confirm the observations of the cesium distribution in the waste form matrix, samples containing only cesium were synthesized. For these samples, the bentonite clay was mixed with a Cs(NO3) solution, dried, and sintered at 800 °C and 1000 °C. The resulting products are brown in color, and appear to be significantly less porous than the samples sintered from the Cs/Sr/Ba-loaded synthetic waste stream concentrate. Figure 11 shows SEM micrographs of samples with only Cs loading. Interestingly, the waste-formed products with cesium loading decreased the density of porous structure. At a sintering temperature of 1000 °C, the product appears to be primarily a glassy phase with minimal porosity (Figure 11(a)). Significant porosity was observed, however, in the sample sintered at 800 °C. The XRD patterns for the synthesized samples are presented in Figure 12. The Cs-loaded sample sintered at 800 °C (sample 6) shows few crystal peaks (only three major peaks) and a strong amorphous background, while the sample sintered at 1000 °C (sample 5) shows an increased number and intensity of crystalline peaks with a reduced background. For sample 6 sintered at 800 °C, sodium aluminum silicate (albite, NaAlSi3O8), quartz (SiO2), and cesium magnesium aluminum silicate (Cs-substituted indialite, CsMg2Al5Si4O18) were identified in the sample. For sample 5 sintered at 1000 °C, albite, quartz, and cesium aluminum silicate were identified. Quartz peaks were observed in XRD patterns of both sample 5 (strong intensity) and sample 6 (weak intensity), suggesting that the quartz phase in the starting material was not significantly consumed for the formation of the Cs-aluminosilicate phases, at least in comparison with the barium and strontium-loaded samples. The weak intensity of the quartz peak in sample 6 sintered at 800 °C is likely due to amorphization of materials in that temperature range. Based on these observations, higher sintering temperatures are suggested for producing more crystalline products. While the removal of barium and strontium from the waste-formed matrix appears to result in a more crystalline product, this result is more likely due to the reduced waste loading of the matrix for the cesium-only formulation relative to the Cs/Sr/Ba-loaded matrix. Additional work will be required to determine the exact cause for this observed increase in crystallinity. The STEM/EDX examination of samples 5 and 6 identified a number of crystalline inclusions. X-ray signals of Cs-K and Cs-L were detected in both the inclusion and the matrix, suggesting that cesium was distributed between both the crystalline inclusions and the matrix for the samples sintered at 1000 °C and 800 °C, as shown in Figures 13 and 14. The TEM BF image of sample 5 displays the rectangularly shaped nanoinclusions distributed throughout the substrate (Figure 15(a)). Examining the matrix for sample 5 by 2886—VOLUME 40A, DECEMBER 2009

Fig. 16—TEM micrographs of sample 5 with Cs loading and sintering at 800 °C: (a) HRTEM image and (b) FFT corresponding to (a).

HRTEM reveals a well-defined lattice structure (Figure 15(b)), with a ring-spot pattern evident by FFT, suggesting that the glassy phase identified by SEM is actually more crystalline than amorphous. At 800 °C, HRTEM and the corresponding FFT (Figure 16) do not show the same well-defined lattice structure observed for the substrate in the sample sintered at 1000 °C. The inclusions in both samples appear fully crystalline.

IV.

CONCLUSIONS

Aluminosilicate was synthesized from bentonite clay with and without Cs/Sr/Ba loading at 800 °C and 1000 °C. The materials were examined and analyzed using SEM, TEM, and XRD. The results suggest the following conclusions. 1. The synthesized aluminosilicate waste form is a multiphase system consisting of crystalline inclusions and amorphous aluminosilicate glass matrix. During synthesis of aluminosilicate waste form, quartz (SiO2) in the as-received bentonite clay was found to be the major consumable component interacting with loaded waste elements to form the new waste-containing phases. 2. The current bench-scale process results in a highly porous and low-density product. To increase the density of the final product, a calcinations step to METALLURGICAL AND MATERIALS TRANSACTIONS A

remove the nitrates and structural water from the loaded clay waste former is suggested prior to pressing the green pellets. 3. At the proposed 20 wt pct waste loading, strontium and barium precipitate as microcrystalline secondary inclusions, while cesium is distributed between microcrystalline inclusions and the amorphous bulk phase. 4. Removal of the strontium and barium from the waste matrix resulted in the formation of a multiphase crystalline matrix at a sintering temperature of 1000 °C. However, it is unclear at this time if this is due to the removal of the divalent ions or due to the reduced waste loading in the matrix. If a crystalline waste form is desired, additional experiments to explore crystallinity as a function of waste loading will be necessary.

ACKNOWLEDGMENTS This project is funded under the auspices of the United States Department of Energy (Grant Nos. DE-FG07-01AL67358 and DE-FC07-06ID14781). The authors also thank the UNLV TRP (Transmutation Research Program), administered by Dr. Tony Hechanova, Harry Reid Center for Environmental Studies, University of Nevada, Las Vegas, for supporting this work. Helpful discussions with Dr. Thomas

METALLURGICAL AND MATERIALS TRANSACTIONS A

Hartmann, Harry Reid Center for Environmental Studies, and Dr. Clay Crow, Department of Geoscience, University of Nevada, Las Vegas, regarding XRD analysis and interpretation are also appreciated.

REFERENCES 1. S. Clark and R.C. Ewing: Basic Research Needs for Advanced Nuclear Energy Systems, Office of Basic Energy Sciences, U.S. Department of Energy, Washington, DC, 2006, pp. 59–73. 2. M.T. Peters, R.C. Ewing, and C.I. Steefel: ‘‘GNEP Waste Form Campaign Science & Technology and Modeling & Simulation Program: Roadmap With Rationale & Recommendations,’’ prepared for the U.S. Department of Energy, Global Nuclear Energy Partnership, and the Waste Form Campaign, GNEP-M50-3040-303 and GNEP-M50-3030-101, Mar. 14, 2008. 3. I.W. Donald, B.L. Metcalfe, and R.N.J. Taylor: J. Mater. Sci., 1997, vol. 32, pp. 5851–87. 4. W.E. Lee, M.I. Ojovan, M.C. Stennett, and N.C. Hyatt: Adv. Appl. Ceram., 2006, vol. 105, pp. 3–20. 5. L.P. Hatch: Am. Scientist, 1953, vol. 41, p. 410. 6. D.M. Strachan and W.W. Schulz: ‘‘Glass and Ceramic Materials for the Immobilization of Megacurie Amount of Pure Cerium-137,’’ U.S. Department of Energy Research Report under Contract No. E(45-1)-2310, Apr. 1975. 7. G.S. Barney: ‘‘Immobilization of Aqueous Radioactive Cesium Wastes by Conversion to Aluminosilicate Minerals,’’ ARH-SA-218, Atlantic Richard Hanford Company Report, Richland, WA, May 1975. 8. V. Viswabaskaran, F.D. Gnanama, and M. Balasubramanian: Ceram. Int., 2003, vol. 29, pp. 561–71. 9. O.D. Velev, T.A. Jede, R.F. Lobo, and A.M. Lenhof: Chem. Mater., 1998, vol. 10, pp. 3597–3602.

VOLUME 40A, DECEMBER 2009—2887