- Email: [email protected]
Radiation damage of hollandite in multiphase ceramic waste forms
Journal of Nuclear Materials 494 (2017) 61e66
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Radiation damage of hollandite in multiphase ceramic waste forms Braeden M. Clark a, *, Priyatham Tumurgoti a, S.K. Sundaram a, Jake W. Amoroso b, James C. Marra b, Vaithiyalingam Shutthanandan c, Ming Tang d a
Kazuo Inamori School of Engineering, Alfred University, Alfred, NY, USA Savannah River National Laboratory, Aiken, SC, USA c Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, WA, USA d Los Alamos National Laboratory, Los Alamos, NM, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 March 2017 Received in revised form 22 June 2017 Accepted 6 July 2017 Available online 8 July 2017
Radiation damage was simulated in multiphase titanate-based ceramic waste forms using an ion accelerator to generate high energy alpha particles (Heþ) and an ion implanter to generate 7 MeV gold (Au3þ) particles. X-ray diffraction and transmission electron microscopy were used to characterize the damaged surfaces and nearby regions. Simulated multiphase ceramic waste forms were prepared using two processing methods: spark plasma sintering and melt-processing. Both processing methods produced ceramics with similar phase assemblages consisting of hollandite-, zirconolite/pyrochlore-, and perovskite-type phases. The measured heavy ion (Au3þ) penetration depth was less in spark plasma sintered samples than in melt-processed samples. Structural breakdown of the hollandite phase occurred under Heþ irradiation indicated by the presence of x-ray diffraction peaks belonging to TiO2, BaTiO5, and other hollandite related phases (Ba2Ti9O20). The composition of the constituent hollandite phase affected the extent of damage induced by Au3þ ions. © 2017 Elsevier B.V. All rights reserved.
Keywords: Hollandite Ion beam irradiation Nuclear waste storage ceramics
1. Introduction Ceramic waste forms are candidates for the immobilization of high-level waste (HLW) generated by reprocessing of commercial used nuclear fuel (UNF). Ceramics have been investigated for their potential advantages over glass waste forms such as higher waste loading of elements that have low solubility in glasses and improved chemical durability [1,2]. Titanate-based ceramic materials, specifically those in the SYNROC family, have been studied for the HLW containment [3e5] as these multiphase assemblages can accommodate elements with a wide range of valences and ionic sizes in the crystal lattices [5] within the waste form. SYNROC materials are traditionally prepared via hot isostatic pressing (HIP), but alternative fabrication methods such meltprocessing and spark plasma sintering (SPS) offer potential advantages over HIP. Melt-processing of ceramics would be simplified compared to HIP methods and melter technologies have been successfully demonstrated for several decades as the preferred option to vitrify of HLW throughout the world. A distinct advantage
* Corresponding author. E-mail address: [email protected] (B.M. Clark). http://dx.doi.org/10.1016/j.jnucmat.2017.07.013 0022-3115/© 2017 Elsevier B.V. All rights reserved.
of SPS compared to HIPing is the potential to reduce volatilization owing to short processing times (can be < 30 min) that can be achieved. Ceramics have been shown to be susceptible to damage induced by a-radiation [6e8]. Specifically, a-radiation can induce atomic displacements that lead to structural rearrangements affecting the physical and chemical properties of the material. An a-decay event consists of the release of an a-particle (light particle) and a recoil nucleus (heavy particle). Actinides and their daughter products have long half-lives and therefore a-decay becomes dominant over long time scales. This necessitates the need to use ion implantation techniques to study the effects of a-decay on a laboratory time scale. One method to study these differing damage mechanisms is by using Heþ ions to simulate a-particles and heavy ions, such as Au3þ or Kr3þ, to simulate a-recoils. Charged particle implantation has been widely utilized to study radiation damage in crystalline ceramics [9e13]. However, relatively few studies have been performed on multiphase crystalline ceramics for nuclear waste immobilization of a combined waste to include a Cs/Sr separated waste stream, the Trivalent Actinide Lanthanide Separation by Phosphorous reagent Extraction from Aqueous Komplexes (TALSPEAK) waste stream, the transition metal
62
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
fission product waste stream resulting from the transuranic extraction (TRUEX) process, MoO3, and noble metals [14e17]. Although the materials modeled were inert matrix fuel and not waste forms, Men et al. studied the radiation response of a fourphase oxide ceramic composite using Au and Xe ions [18]. The Al2O3, Y2O3 stabilized ZrO2, and MgAl2O4 phases exhibited high amorphization resistance when irradiated with 1020 ions/m2 10 MeV Au ions and 1016-1017 ions/m2 Xe ions. In the present paper, we report results on the effects of radiation damage in multiphase titanate-based simulated waste forms containing hollandite, perovskite, and pyrochlore/zirconolite/ zirconium-rich phases produced by SPS and melt-processing. Single phase hollandite material produced by solid state sintering is also presented for comparison. The radiation resistance was compared via irradiation with light (Heþ) and heavy (Au3þ or Kr3þ) ions and characterized using grazing incidence x-ray diffraction (GIXRD) and transmission electron microscopy (TEM).
2. Experimental The ceramic waste form compositions were developed by Savannah River National Laboratory (SRNL) and are listed in Table 1. The compositions targeted hollandite phase with metal oxide additions of Cr, Al, and Fe (designated CAF-MP) and hollandite phase with only Cr additions (designated Cr-MP). Samples were prepared by melt-processing and SPS methods that have been documented previously [19]. The compositions prepared by SPS were heated to a maximum temperature of 1125 C (Cr-MP) or 1000 C (CAF-MP), as read by an optical pyrometer focused on the outside of the die and held at temperature for 3 min. Melt-processed samples were ramped to 1500 C and held for 30 min (both compositions). Single phase Cr-hollandite (Ba1.15Cr2.3Ti5.7O16, designated Cr-HOL) was produced via solid state sintering stoichiometric amounts of commercially available oxides/carbonates that were mixed, ballmilled and dried at ~ 120 C to form the precursor batch. The resulting powders were compacted into pellets and subsequently sintered at 1100 C for 4 h. Solid-state synthesis (SSS) was performed, as opposed to SPS, because composition was found to have a greater impact on amorphization behavior than processing.
Samples for ion irradiation experiments were prepared by polishing a surface to 1 mm using a diamond suspension. The radiation dose in these materials was estimated using the Stopping and Range of Ions in Matter (SRIM) program [20]. The displacements per atom (dpa) as a function of depth for CAF-MP (similar results are obtained for Cr-MP) using 7 MeV Au3þ and 200 keV Heþ is shown in Fig. 1. The peak in dpa (5, 0.5 or 0.05) is located at 870 nm (Au3þ) and 700 nm (Heþ). The dpa is adjusted by changing the fluence of the incoming gold ions (ions/m2). The experiments using Au3þ ions were carried out at room temperature at the Environmental Molecular Sciences Laboratory (EMSL), Pacific Northwest National Laboratory (PNNL) and the experiments using Heþ ions at the Ion Beam Materials Lab (IBML), Los Alamos National Laboratory (LANL) for the multiphase materials. 7 MeV Au3þ implantations were performed at EMSL on a National Electrostatics Corporation (NEC) model 9SDH-2 3.4 MV tandem accelerator [21] simulating three damage levels in SPS samples: 0.05, 0.5 and 5 dpa and one damage level, 5 dpa, in melt-processed samples. These values were chosen to simulate varying storage times, from approximately 102-106 years [22]. The implanted area was 6 6 mm for both melt-processed and SPS samples. 200 keV Heþ implantations were performed at IBML using a 200 kV Danfysik high current ion implanter at a dose of 5 dpa to simulate aparticle damage after long storage times. 380 keV Heþ and 3 MeV Kr3þ implantations were performed on single phase hollandite using the Ion Beam Laboratory at the University at Albany. Single phase hollandites received doses of 0.5 dpa of either Heþ or Kr3þ ions. The details of implantation experiments are summarized in Table 2. Powder x-ray diffraction (XRD) was carried out using a D-2 Phaser (Bruker, Massachusetts, USA) for phase identification. GIXRD was performed on implanted samples using a grazing angle of 2 on the EMSL (Au3þ) samples and 1 for the LANL (Heþ) samples with a Bruker AXS D8 Advance (Bruker, Massachusetts, USA) instrument using Cu Ka radiation in theta e 2 theta geometry. A plot of x-ray penetration depth vs grazing angle is shown in Fig. 2. The unirradiated areas of the Au3þ samples were visible and ground off prior to measurement. Damaged cross-section specimens were prepared for transmission electron microscopy (TEM) by using the lift-out method with a focused ion beam (FIB). The damaged areas were examined using a FEI Tecnai F20 TEM operating at 200 kV.
Table 1 Simulated waste form compositions. Oxide
CAF-MP (wt%)
Cr-MP (wt%)
Cr-HOL (wt%)
Al2O3 BaO CaO Cr2O3 CdO Ce2O3 Cs2O Eu2O3 Fe2O3 Gd2O3 La2O3 MoO3 Nd2O3 Pr2O3 Rb2O SeO2 Sm2O3 SnO2 SrO TeO2 TiO2 Y2O3 ZrO2
1.27 12.76 1.39 6.33 0.11 3.10 2.88 0.17 6.65 0.16 1.58 0.85 5.23 1.45 0.42 0.08 1.08 0.07 0.98 0.66 49.16 0.63 2.99
0.00 12.72 1.38 14.5 0.11 3.09 2.87 0.17 0.00 0.16 1.58 0.84 5.22 1.44 0.42 0.08 1.07 0.07 0.98 0.65 49.01 0.63 2.98
0.00 28.70 0.00 21.73 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 49.57 0.00 0.00
Fig. 1. Damage profile in CAF-MP using 7 MeV Au3þ and 200 keV Heþ ions.
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
63
Table 2 Summary of ion implantation experiments. Composition
Processing method
Ion (Energy)
Fluence/dpa (ions/m2)
CAF-MP CAF-MP CAF-MP CAF-MP CAF-MP
Melt SPS SPS SPS SPS
Au Au Au Au He
(7 MeV) (7 MeV) (7 MeV) (7 MeV) (200 keV)
1.4 1.4 1.4 1.4 2.0
1019/5 1019/5 1018/0.5 1017/0.05 1021/5
Cr-MP Cr-MP Cr-MP Cr-MP Cr-MP
Melt SPS SPS SPS SPS
Au Au Au Au He
(7 MeV) (7 MeV) (7 MeV) (7 MeV) (200 keV)
1.3 1.3 1.3 1.3 2.0
1019/5 1019/5 1018/0.5 1017/0.05 1021/5
Cr-HOL Cr-HOL
SSS SSS
He (380 keV) Kr (3 MeV)
2.8 1020/0.5 2.0 1018/0.5
Fig. 2. X-ray penetration depth vs grazing angle for the materials studied.
3. Results and discussion Fig. 3 shows XRD patterns of the two compositions prepared using SPS and melt-processing. Similar phase assemblages consisting of hollandite (~70%), perovskite (~10e30%), and pyrochlore/ zirconolite/zirconium-rich (~5e10%) were identified. The meltprocessed samples contained a small amount of unreacted TiO2 (~3e8%). The microstructures of both SPS and melt-processed samples of CAF-MP have been described previously [19]. No residual glassy phases were identified, and samples prepared by both processes were dense (91.4% theoretical density (TD) and 98.7% TD for melt-processed and SPS respectively). However, in general samples processed using SPS exhibit a fine grain structure (<1 mm) while the melt-processed samples contain large grains (10s of mm). Fig. 4 shows the GIXRD patterns of CAF- and Cr- MP samples prepared using SPS and melt-processing subjected to 7 MeV Au3þ at a fluence corresponding to 5 dpa. All samples exhibited amorphization evidenced by a decrease in intensity of the diffraction peaks and the presence of diffuse scattering between 26 and 35 2q. CAF-MP samples produced by both processing routes show complete amorphization at 5 dpa, as demonstrated by the disappearance of all diffraction peaks. There is no new phase formation identified within the detectable limits (>5 wt%) of the technique. The highest intensity peaks of the major phases in Cr-MP samples are labeled in Fig. 4. By comparing the intensity of these peaks, the TiO2 and perovskite phases in the Cr-MP melt-processed are the least affected by the damage, indicating that the hollandite phase is more prone to heavy ion radiation damage. Previous results have shown that hollandite materials are susceptible to damage via heavy ion irradiation [23]. Bright field TEM images of damaged cross-sections of both meltprocessed and SPS CAF-MP samples irradiated with 7 MeV Au3þ to 5 dpa are shown in Fig. 5. The corresponding selected area diffraction (SAD) patterns confirm that amorphization occurs in samples fabricated by both methods. Due to the large grain sizes in
Fig. 3. XRD patterns of CAF- and Cr-MP samples produced by SPS and melt-processing.
the melt-processed samples, only a hollandite grain was extracted with FIB, confirmed through indexing of the SAD pattern (space group C2/m, grain oriented in the [010] direction) and EDS (Fig. 5). Cu present in the EDS spectra was attributed to the sample grid holder. The amophized region in the hollandite grain was approximately 2.1 mm in depth. In contrast to the melt-processed samples, multiple grains are contained within the FIB sample of CAF-SPS. This was confirmed by examining the SAD pattern in the crystalline region. The pattern was unable to be indexed properly due to the interaction of the beam with multiple grains. The amorphized region of the SPS sample extended approximately 1.8 mm from the surface. The penetration depth of the Au3þ ions was greater in the meltprocessed than that in the SPS fabricated sample. This was attributed to the interaction of the ions with multiple phases of much smaller (<1 mm) grain sizes. Grain boundaries act as sinks for radiation damage induced defects, therefore inhibiting ion penetration in SPS materials [24]. The ion penetration depth in both samples was greater (~0.4 mm) than that predicted by SRIM however, underestimation is not unprecedented such as of Au ion penetration investigated in SiC [25]. GIXRD patterns for the samples fabricated by SPS and subjected to 3 different fluences of 7 MeV Au3þ ions corresponding to 0.05, 0.5, and 5 dpa are shown in Fig. 6. Amorphization increased with increase in dpa as would be expected and only minor changes in the diffraction patterns were observed in samples exposed to a fluence corresponding to 0.05 dpa, suggesting a threshold fluence that lies above this level. At 0.5 dpa a marked decrease in peak intensity and diffuse scattering can be seen. Complete amorphization is observed in CAF-MP compositions at 5 dpa, while Cr-MP displayed an
64
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
Fig. 4. GIXRD of CAF- and Cr-MP samples fabricated by SPS and melt-processing exposed to 5 dpa of Au3þ ions.
Fig. 5. TEM images and corresponding SAD patterns of the damage cross-sections in CAF-MP materials prepared by melt-processing and SPS after exposure to 5 dpa of Au3þ ions.
increase in diffuse scattering. GIXRD patterns for the SPS samples exposed to 200 keV Heþ ions corresponding to 5 dpa are shown in Fig. 7. In these patterns, diffuse scattering is seen from 26 to 35 2q indicating amorphization, and diffraction peaks not present in the pristine sample were identified and could be indexed to Ba2Ti9O20. This Ba2Ti9O20 phase is related to the hollandite phase in which some of the Ba ions are located on O sites, disrupting the tunnel structure of hollandite [26,27]. A recent report [28] also identifies the presence of this new phase in multiphase ceramics. The presence of these new
diffraction peaks, along with the disappearance of the diffraction peaks belonging to the hollandite phase, suggests that structural rearrangement of the hollandite phase occurs along with amorphization under these irradiation conditions. The perovskite phase could be identified after irradiation and did not appear to amorphize as severely as the hollandite phase when irradiated with Heþ ions, indicating radiation stability under light ion irradiation up to 5 dpa. The zirconolite/pyrochlore (zirconium-rich) phase exhibited a relatively low-intensity diffraction pattern that could not readily be identified from the diffuse scattering.
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
65
Fig. 6. GIXRD of CAF- and Cr-SPS samples irradiated to 3 different damage levels with Au3þ ions.
To further study the response of hollandite in the multiphase assemblages, single phase hollandite (Cr-HOL) was irradiated with Kr3þ (heavy) and Heþ (light) to doses equivalent to 0.5 dpa. GIXRD patterns from the pristine and irradiated surfaces of single phase hollandite are shown in Fig. 8. The single phase hollandite exhibited similar behavior to the hollandite within the multiphase assemblages. 0.5 dpa He and Kr both resulted in amorphization similar to multiphase samples as evidenced by diffuse scattering. Heþ irradiations also induced phase evolution as evidenced by the emergence of diffraction peaks corresponding to TiO2, BaTiO5, and
Fig. 8. GIXRD patterns of single phase hollandite materials irradiated with 0.5 dpa Kr3þ and Heþ ions.
Ba2Ti9O20. The presence of the TiO2 and BaTi2O5 diffraction peaks suggests that the hollandite phase experiences structural breakdown when irradiated with light ions. The intensity of the Ba2Ti9O20 diffraction peaks is reduced in Cr-HOL compared to that
Fig. 7. GIXRD of CAF- and Cr-MP SPS samples irradiated with 5 dpa Heþ ions.
66
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
in the multiphase samples, indicating that the transition from hollandite structure to Ba2Ti9O20 is incomplete at 0.5 dpa Heþ ions. Further TEM studies detailing the radiation response of the other multiphase assemblages are needed.
Department 08SR22470.
4. Conclusions
[1] I.W. Donald, Waste Immobilization in Glass and Ceramic Based Hosts: Radioactive, Toxic and Hazardous Wastes, first ed., Wiley, 2010. [2] J.N. Cachia, X. Deschanels, C. Den Auwer, O. Pinet, J. Phalippou, C. Hennig, A. Scheinost, J. Nucl. Mater 352 (1e3) (2006) 182e189. [3] G.R. Lumpkin, K.L. Smith, M.G. Blackford, J. Nucl. Mater 224 (1995) 31e42. [4] T. Murakami, Chem. Waste Man. 5 (1985) 269e278. [5] A.E. Ringwood, S.E. Kesson, N.G. Ware, W. Hibberson, A. Major, Nature 278 (1979) 219e223. [6] S.X. Wang, G.R. Lumpkin, L.M. Wang, R.C. Ewing, Nucl. Instrum. Meth. B 166e167 (2000) 293e298. [7] V. Picot, X. Deschanels, S. Peuget, B. Glorieux, A.M. Seydoux-Guillaume, R. Wirth, J. Nucl. Mater 381 (3) (2008) 290e296. [8] M. Lang, F. Zhang, W. Li, D. Severin, M. Bender, S. Klaumünzer, C. Trautmann, R.C. Ewing, Nucl. Instrum. Meth B 286 (2012) 271e276. [9] C. Davoisne, M.C. Stennett, N.C. Hyatt, N. Peng, C. Jeynes, W.E. Lee, J. Nucl. Mater 415 (1) (2011) 67e73. [10] J. Lian, L.M. Wang, R.C. Ewing, L.A. Boatner, Nucl. Instrum. Meth. B 241 (1e4) (2005) 365e371. [11] S.X. Wang, L.M. Wang, R.C. Ewing, G.S. Was, G.R. Lumpkin, Nucl. Instrum. Meth. B 148 (1e4) (1999) 704e709. , G. Balakrishnan, D.M. Paul, [12] Y. Zhang, J. Jagielski, I.-T. Bae, X. Xiang, L. Thome W.J. Weber, Nucl. Instrum. Meth. B 268 (19) (2010) 3009e3013. [13] K.E. Sickafus, R.W. Grimes, J.A. Valdez, A. Cleave, M. Tang, M. Ishimaru, S.M. Corish, C.R. Stanek, B.P. Uberuaga, Nat. Mater 6 (3) (2007) 217e223. [14] M. Tang, A. Kossoy, G. Jarvinen, J. Crum, L. Turo, B. Riley, K. Brinkman, K. Fox, J. Amoroso, J. Marra, Nucl. Instrum. Meth. B 326 (2014) 293e297. [15] A.E. Ringwood, V.M. Oversby, W. Sinclair, The effects of radiation damage on synroc, in: C.J.M. Northrup (Ed.), Scientific Basis for Nuclear Waste Management, Springer US, Boston, MA, 1980, pp. 273e280. [16] J.D. Vienna, Closed Fuel Cycle Waste Treatment Strategy, Fuel Cycle Research & Development, 2015. U.S. Department of Energy Report, FCRD-MRWFD2015e20000674. [17] S.E. Kesson, C.J. Ball, J. Aust. Ceram. Soc. 24 (1) (1988) 89e99. [18] D. Men, M.K. Patel, I.O. Usov, M. Toiammou, I. Monnet, J.C. Pivin, J.R. Porter, M.L. Mecartney, J. Nucl. Mater 443 (1) (2013) 120e127. [19] B.M. Clark, P. Tumurugoti, S.K. Sundaram, J.W. Amoroso, J.C. Marra, K.S. Brinkman, Metall. Mater. Trans. E 1 (4) (2014) 341e348. [20] J.F. Ziegler, J.P. Biersack, U. Littmark, The Stopping and Range of Ions in Matter, Pergamon Press, New York, 1985. [21] S. Thevuthasan, C.H.F. Peden, M.H. Engelhard, D.R. Baer, G.S. Herman, W. Jiang, Y. Liang, W.J. Weber, Nucl. Instrum. Meth. A 420 (1e2) (1999) 81e89. [22] R.C. Ewing, W.J. Weber, F.W. Clinard, Prog. Nucl. Energ 29 (1995) 63e127. [23] M. Tang, P. Tumurugoti, B. Clark, S.K. Sundaram, J. Amoroso, J. Marra, C. Sun, P. Lu, Y. Wang, Y.B. Jiang, J. Solid State Chem. 239 (2016) 58e63. [24] T.D. Shen, Nucl. Instrum. Meth. B 266 (6) (2008) 921e925. [25] K. Jin, Y. Zhang, H. Xue, Z. Zhu, W.J. Weber, Nucl. Instrum. Meth. B 307 (2013) 65e70. [26] G. Grzinic, L.A. Bursill, D.J. Smith, J. Solid State Chem. 47 (2) (1983) 151e163. [27] G.D. Fallon, B.M. Gatehouse, J. Solid State Chem. 49 (1) (1983) 59e64. [28] M. Tang, Characterization and Tesing of Crystalline Ceramic Waste Forms, 2014. U.S. Department of Energy Report, M3FT-14LA0308011.
Heavy and light ion implantation was used to induce damage in simulated multiphase ceramic waste form materials produced from SPS and a melt-process. The multiphase waste forms responded differently under the two different irradiation conditions. Under heavy ion (Au3þ) irradiation, the displacements initiated by the ballistic processes cause bulk material to amorphize. Under light ion (Heþ) irradiation, the hollandite phase appeared to break down while the other phases remained crystalline as evidenced by emergence of diffuse scattering and new crystalline phases at the expense of hollandite. Samples fabricated by both melt-processing and SPS exhibited similar amorphization behavior when irradiated with Au3þ ions although the penetration depth of Au3þ ions was less material fabricated by SPS compared to melt-processing. This is attributed to the smaller grain sizes of the phases in the SPS materials; the grain boundaries act as sinks for the radiation induced defects. The lower dose of Heþ ions was demonstrated as a reliable way to render structural changes in the hollandite phase to be captured in the GIXRD patterns. Acknowledgements The authors would like to thank the Department of Energy's Nuclear Energy University Program (NEUP_12-3809) for supporting this project. The authors would also like to thank J. Vienna (Pacific Northwest National Laboratory) for project oversight and guidance. BMC acknowledges support from NEUP fellowship. SKS is grateful to the generous support of Inamori Professorship by Kyocera Corporation. This work made use of the electron microscopy facility of the Cornell Center for Materials Research (CCMR) with support from the National Science Foundation Materials Research Science and Engineering Centers (MRSEC) program (DMR 1120296). The authors gratefully acknowledge the financial support of the US DOE Fuel Cycle R&D program and Materials Recovery and Waste Form Development (MRWFD) Campaign under the direction of Terry Todd, National Technical Director, John Vienna, Deputy Manager, and Kimberly Gray, DOE Federal Manager. Work conducted at Savannah River National Laboratory was supported by the U.S.
of
Energy
under
contract
number
DE-AC09-
References
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Radiation damage of hollandite in multiphase ceramic waste forms Braeden M. Clark a, *, Priyatham Tumurgoti a, S.K. Sundaram a, Jake W. Amoroso b, James C. Marra b, Vaithiyalingam Shutthanandan c, Ming Tang d a
Kazuo Inamori School of Engineering, Alfred University, Alfred, NY, USA Savannah River National Laboratory, Aiken, SC, USA c Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, WA, USA d Los Alamos National Laboratory, Los Alamos, NM, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 March 2017 Received in revised form 22 June 2017 Accepted 6 July 2017 Available online 8 July 2017
Radiation damage was simulated in multiphase titanate-based ceramic waste forms using an ion accelerator to generate high energy alpha particles (Heþ) and an ion implanter to generate 7 MeV gold (Au3þ) particles. X-ray diffraction and transmission electron microscopy were used to characterize the damaged surfaces and nearby regions. Simulated multiphase ceramic waste forms were prepared using two processing methods: spark plasma sintering and melt-processing. Both processing methods produced ceramics with similar phase assemblages consisting of hollandite-, zirconolite/pyrochlore-, and perovskite-type phases. The measured heavy ion (Au3þ) penetration depth was less in spark plasma sintered samples than in melt-processed samples. Structural breakdown of the hollandite phase occurred under Heþ irradiation indicated by the presence of x-ray diffraction peaks belonging to TiO2, BaTiO5, and other hollandite related phases (Ba2Ti9O20). The composition of the constituent hollandite phase affected the extent of damage induced by Au3þ ions. © 2017 Elsevier B.V. All rights reserved.
Keywords: Hollandite Ion beam irradiation Nuclear waste storage ceramics
1. Introduction Ceramic waste forms are candidates for the immobilization of high-level waste (HLW) generated by reprocessing of commercial used nuclear fuel (UNF). Ceramics have been investigated for their potential advantages over glass waste forms such as higher waste loading of elements that have low solubility in glasses and improved chemical durability [1,2]. Titanate-based ceramic materials, specifically those in the SYNROC family, have been studied for the HLW containment [3e5] as these multiphase assemblages can accommodate elements with a wide range of valences and ionic sizes in the crystal lattices [5] within the waste form. SYNROC materials are traditionally prepared via hot isostatic pressing (HIP), but alternative fabrication methods such meltprocessing and spark plasma sintering (SPS) offer potential advantages over HIP. Melt-processing of ceramics would be simplified compared to HIP methods and melter technologies have been successfully demonstrated for several decades as the preferred option to vitrify of HLW throughout the world. A distinct advantage
* Corresponding author. E-mail address: [email protected] (B.M. Clark). http://dx.doi.org/10.1016/j.jnucmat.2017.07.013 0022-3115/© 2017 Elsevier B.V. All rights reserved.
of SPS compared to HIPing is the potential to reduce volatilization owing to short processing times (can be < 30 min) that can be achieved. Ceramics have been shown to be susceptible to damage induced by a-radiation [6e8]. Specifically, a-radiation can induce atomic displacements that lead to structural rearrangements affecting the physical and chemical properties of the material. An a-decay event consists of the release of an a-particle (light particle) and a recoil nucleus (heavy particle). Actinides and their daughter products have long half-lives and therefore a-decay becomes dominant over long time scales. This necessitates the need to use ion implantation techniques to study the effects of a-decay on a laboratory time scale. One method to study these differing damage mechanisms is by using Heþ ions to simulate a-particles and heavy ions, such as Au3þ or Kr3þ, to simulate a-recoils. Charged particle implantation has been widely utilized to study radiation damage in crystalline ceramics [9e13]. However, relatively few studies have been performed on multiphase crystalline ceramics for nuclear waste immobilization of a combined waste to include a Cs/Sr separated waste stream, the Trivalent Actinide Lanthanide Separation by Phosphorous reagent Extraction from Aqueous Komplexes (TALSPEAK) waste stream, the transition metal
62
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
fission product waste stream resulting from the transuranic extraction (TRUEX) process, MoO3, and noble metals [14e17]. Although the materials modeled were inert matrix fuel and not waste forms, Men et al. studied the radiation response of a fourphase oxide ceramic composite using Au and Xe ions [18]. The Al2O3, Y2O3 stabilized ZrO2, and MgAl2O4 phases exhibited high amorphization resistance when irradiated with 1020 ions/m2 10 MeV Au ions and 1016-1017 ions/m2 Xe ions. In the present paper, we report results on the effects of radiation damage in multiphase titanate-based simulated waste forms containing hollandite, perovskite, and pyrochlore/zirconolite/ zirconium-rich phases produced by SPS and melt-processing. Single phase hollandite material produced by solid state sintering is also presented for comparison. The radiation resistance was compared via irradiation with light (Heþ) and heavy (Au3þ or Kr3þ) ions and characterized using grazing incidence x-ray diffraction (GIXRD) and transmission electron microscopy (TEM).
2. Experimental The ceramic waste form compositions were developed by Savannah River National Laboratory (SRNL) and are listed in Table 1. The compositions targeted hollandite phase with metal oxide additions of Cr, Al, and Fe (designated CAF-MP) and hollandite phase with only Cr additions (designated Cr-MP). Samples were prepared by melt-processing and SPS methods that have been documented previously [19]. The compositions prepared by SPS were heated to a maximum temperature of 1125 C (Cr-MP) or 1000 C (CAF-MP), as read by an optical pyrometer focused on the outside of the die and held at temperature for 3 min. Melt-processed samples were ramped to 1500 C and held for 30 min (both compositions). Single phase Cr-hollandite (Ba1.15Cr2.3Ti5.7O16, designated Cr-HOL) was produced via solid state sintering stoichiometric amounts of commercially available oxides/carbonates that were mixed, ballmilled and dried at ~ 120 C to form the precursor batch. The resulting powders were compacted into pellets and subsequently sintered at 1100 C for 4 h. Solid-state synthesis (SSS) was performed, as opposed to SPS, because composition was found to have a greater impact on amorphization behavior than processing.
Samples for ion irradiation experiments were prepared by polishing a surface to 1 mm using a diamond suspension. The radiation dose in these materials was estimated using the Stopping and Range of Ions in Matter (SRIM) program [20]. The displacements per atom (dpa) as a function of depth for CAF-MP (similar results are obtained for Cr-MP) using 7 MeV Au3þ and 200 keV Heþ is shown in Fig. 1. The peak in dpa (5, 0.5 or 0.05) is located at 870 nm (Au3þ) and 700 nm (Heþ). The dpa is adjusted by changing the fluence of the incoming gold ions (ions/m2). The experiments using Au3þ ions were carried out at room temperature at the Environmental Molecular Sciences Laboratory (EMSL), Pacific Northwest National Laboratory (PNNL) and the experiments using Heþ ions at the Ion Beam Materials Lab (IBML), Los Alamos National Laboratory (LANL) for the multiphase materials. 7 MeV Au3þ implantations were performed at EMSL on a National Electrostatics Corporation (NEC) model 9SDH-2 3.4 MV tandem accelerator [21] simulating three damage levels in SPS samples: 0.05, 0.5 and 5 dpa and one damage level, 5 dpa, in melt-processed samples. These values were chosen to simulate varying storage times, from approximately 102-106 years [22]. The implanted area was 6 6 mm for both melt-processed and SPS samples. 200 keV Heþ implantations were performed at IBML using a 200 kV Danfysik high current ion implanter at a dose of 5 dpa to simulate aparticle damage after long storage times. 380 keV Heþ and 3 MeV Kr3þ implantations were performed on single phase hollandite using the Ion Beam Laboratory at the University at Albany. Single phase hollandites received doses of 0.5 dpa of either Heþ or Kr3þ ions. The details of implantation experiments are summarized in Table 2. Powder x-ray diffraction (XRD) was carried out using a D-2 Phaser (Bruker, Massachusetts, USA) for phase identification. GIXRD was performed on implanted samples using a grazing angle of 2 on the EMSL (Au3þ) samples and 1 for the LANL (Heþ) samples with a Bruker AXS D8 Advance (Bruker, Massachusetts, USA) instrument using Cu Ka radiation in theta e 2 theta geometry. A plot of x-ray penetration depth vs grazing angle is shown in Fig. 2. The unirradiated areas of the Au3þ samples were visible and ground off prior to measurement. Damaged cross-section specimens were prepared for transmission electron microscopy (TEM) by using the lift-out method with a focused ion beam (FIB). The damaged areas were examined using a FEI Tecnai F20 TEM operating at 200 kV.
Table 1 Simulated waste form compositions. Oxide
CAF-MP (wt%)
Cr-MP (wt%)
Cr-HOL (wt%)
Al2O3 BaO CaO Cr2O3 CdO Ce2O3 Cs2O Eu2O3 Fe2O3 Gd2O3 La2O3 MoO3 Nd2O3 Pr2O3 Rb2O SeO2 Sm2O3 SnO2 SrO TeO2 TiO2 Y2O3 ZrO2
1.27 12.76 1.39 6.33 0.11 3.10 2.88 0.17 6.65 0.16 1.58 0.85 5.23 1.45 0.42 0.08 1.08 0.07 0.98 0.66 49.16 0.63 2.99
0.00 12.72 1.38 14.5 0.11 3.09 2.87 0.17 0.00 0.16 1.58 0.84 5.22 1.44 0.42 0.08 1.07 0.07 0.98 0.65 49.01 0.63 2.98
0.00 28.70 0.00 21.73 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 49.57 0.00 0.00
Fig. 1. Damage profile in CAF-MP using 7 MeV Au3þ and 200 keV Heþ ions.
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
63
Table 2 Summary of ion implantation experiments. Composition
Processing method
Ion (Energy)
Fluence/dpa (ions/m2)
CAF-MP CAF-MP CAF-MP CAF-MP CAF-MP
Melt SPS SPS SPS SPS
Au Au Au Au He
(7 MeV) (7 MeV) (7 MeV) (7 MeV) (200 keV)
1.4 1.4 1.4 1.4 2.0
1019/5 1019/5 1018/0.5 1017/0.05 1021/5
Cr-MP Cr-MP Cr-MP Cr-MP Cr-MP
Melt SPS SPS SPS SPS
Au Au Au Au He
(7 MeV) (7 MeV) (7 MeV) (7 MeV) (200 keV)
1.3 1.3 1.3 1.3 2.0
1019/5 1019/5 1018/0.5 1017/0.05 1021/5
Cr-HOL Cr-HOL
SSS SSS
He (380 keV) Kr (3 MeV)
2.8 1020/0.5 2.0 1018/0.5
Fig. 2. X-ray penetration depth vs grazing angle for the materials studied.
3. Results and discussion Fig. 3 shows XRD patterns of the two compositions prepared using SPS and melt-processing. Similar phase assemblages consisting of hollandite (~70%), perovskite (~10e30%), and pyrochlore/ zirconolite/zirconium-rich (~5e10%) were identified. The meltprocessed samples contained a small amount of unreacted TiO2 (~3e8%). The microstructures of both SPS and melt-processed samples of CAF-MP have been described previously [19]. No residual glassy phases were identified, and samples prepared by both processes were dense (91.4% theoretical density (TD) and 98.7% TD for melt-processed and SPS respectively). However, in general samples processed using SPS exhibit a fine grain structure (<1 mm) while the melt-processed samples contain large grains (10s of mm). Fig. 4 shows the GIXRD patterns of CAF- and Cr- MP samples prepared using SPS and melt-processing subjected to 7 MeV Au3þ at a fluence corresponding to 5 dpa. All samples exhibited amorphization evidenced by a decrease in intensity of the diffraction peaks and the presence of diffuse scattering between 26 and 35 2q. CAF-MP samples produced by both processing routes show complete amorphization at 5 dpa, as demonstrated by the disappearance of all diffraction peaks. There is no new phase formation identified within the detectable limits (>5 wt%) of the technique. The highest intensity peaks of the major phases in Cr-MP samples are labeled in Fig. 4. By comparing the intensity of these peaks, the TiO2 and perovskite phases in the Cr-MP melt-processed are the least affected by the damage, indicating that the hollandite phase is more prone to heavy ion radiation damage. Previous results have shown that hollandite materials are susceptible to damage via heavy ion irradiation [23]. Bright field TEM images of damaged cross-sections of both meltprocessed and SPS CAF-MP samples irradiated with 7 MeV Au3þ to 5 dpa are shown in Fig. 5. The corresponding selected area diffraction (SAD) patterns confirm that amorphization occurs in samples fabricated by both methods. Due to the large grain sizes in
Fig. 3. XRD patterns of CAF- and Cr-MP samples produced by SPS and melt-processing.
the melt-processed samples, only a hollandite grain was extracted with FIB, confirmed through indexing of the SAD pattern (space group C2/m, grain oriented in the [010] direction) and EDS (Fig. 5). Cu present in the EDS spectra was attributed to the sample grid holder. The amophized region in the hollandite grain was approximately 2.1 mm in depth. In contrast to the melt-processed samples, multiple grains are contained within the FIB sample of CAF-SPS. This was confirmed by examining the SAD pattern in the crystalline region. The pattern was unable to be indexed properly due to the interaction of the beam with multiple grains. The amorphized region of the SPS sample extended approximately 1.8 mm from the surface. The penetration depth of the Au3þ ions was greater in the meltprocessed than that in the SPS fabricated sample. This was attributed to the interaction of the ions with multiple phases of much smaller (<1 mm) grain sizes. Grain boundaries act as sinks for radiation damage induced defects, therefore inhibiting ion penetration in SPS materials [24]. The ion penetration depth in both samples was greater (~0.4 mm) than that predicted by SRIM however, underestimation is not unprecedented such as of Au ion penetration investigated in SiC [25]. GIXRD patterns for the samples fabricated by SPS and subjected to 3 different fluences of 7 MeV Au3þ ions corresponding to 0.05, 0.5, and 5 dpa are shown in Fig. 6. Amorphization increased with increase in dpa as would be expected and only minor changes in the diffraction patterns were observed in samples exposed to a fluence corresponding to 0.05 dpa, suggesting a threshold fluence that lies above this level. At 0.5 dpa a marked decrease in peak intensity and diffuse scattering can be seen. Complete amorphization is observed in CAF-MP compositions at 5 dpa, while Cr-MP displayed an
64
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
Fig. 4. GIXRD of CAF- and Cr-MP samples fabricated by SPS and melt-processing exposed to 5 dpa of Au3þ ions.
Fig. 5. TEM images and corresponding SAD patterns of the damage cross-sections in CAF-MP materials prepared by melt-processing and SPS after exposure to 5 dpa of Au3þ ions.
increase in diffuse scattering. GIXRD patterns for the SPS samples exposed to 200 keV Heþ ions corresponding to 5 dpa are shown in Fig. 7. In these patterns, diffuse scattering is seen from 26 to 35 2q indicating amorphization, and diffraction peaks not present in the pristine sample were identified and could be indexed to Ba2Ti9O20. This Ba2Ti9O20 phase is related to the hollandite phase in which some of the Ba ions are located on O sites, disrupting the tunnel structure of hollandite [26,27]. A recent report [28] also identifies the presence of this new phase in multiphase ceramics. The presence of these new
diffraction peaks, along with the disappearance of the diffraction peaks belonging to the hollandite phase, suggests that structural rearrangement of the hollandite phase occurs along with amorphization under these irradiation conditions. The perovskite phase could be identified after irradiation and did not appear to amorphize as severely as the hollandite phase when irradiated with Heþ ions, indicating radiation stability under light ion irradiation up to 5 dpa. The zirconolite/pyrochlore (zirconium-rich) phase exhibited a relatively low-intensity diffraction pattern that could not readily be identified from the diffuse scattering.
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
65
Fig. 6. GIXRD of CAF- and Cr-SPS samples irradiated to 3 different damage levels with Au3þ ions.
To further study the response of hollandite in the multiphase assemblages, single phase hollandite (Cr-HOL) was irradiated with Kr3þ (heavy) and Heþ (light) to doses equivalent to 0.5 dpa. GIXRD patterns from the pristine and irradiated surfaces of single phase hollandite are shown in Fig. 8. The single phase hollandite exhibited similar behavior to the hollandite within the multiphase assemblages. 0.5 dpa He and Kr both resulted in amorphization similar to multiphase samples as evidenced by diffuse scattering. Heþ irradiations also induced phase evolution as evidenced by the emergence of diffraction peaks corresponding to TiO2, BaTiO5, and
Fig. 8. GIXRD patterns of single phase hollandite materials irradiated with 0.5 dpa Kr3þ and Heþ ions.
Ba2Ti9O20. The presence of the TiO2 and BaTi2O5 diffraction peaks suggests that the hollandite phase experiences structural breakdown when irradiated with light ions. The intensity of the Ba2Ti9O20 diffraction peaks is reduced in Cr-HOL compared to that
Fig. 7. GIXRD of CAF- and Cr-MP SPS samples irradiated with 5 dpa Heþ ions.
66
B.M. Clark et al. / Journal of Nuclear Materials 494 (2017) 61e66
in the multiphase samples, indicating that the transition from hollandite structure to Ba2Ti9O20 is incomplete at 0.5 dpa Heþ ions. Further TEM studies detailing the radiation response of the other multiphase assemblages are needed.
Department 08SR22470.
4. Conclusions
[1] I.W. Donald, Waste Immobilization in Glass and Ceramic Based Hosts: Radioactive, Toxic and Hazardous Wastes, first ed., Wiley, 2010. [2] J.N. Cachia, X. Deschanels, C. Den Auwer, O. Pinet, J. Phalippou, C. Hennig, A. Scheinost, J. Nucl. Mater 352 (1e3) (2006) 182e189. [3] G.R. Lumpkin, K.L. Smith, M.G. Blackford, J. Nucl. Mater 224 (1995) 31e42. [4] T. Murakami, Chem. Waste Man. 5 (1985) 269e278. [5] A.E. Ringwood, S.E. Kesson, N.G. Ware, W. Hibberson, A. Major, Nature 278 (1979) 219e223. [6] S.X. Wang, G.R. Lumpkin, L.M. Wang, R.C. Ewing, Nucl. Instrum. Meth. B 166e167 (2000) 293e298. [7] V. Picot, X. Deschanels, S. Peuget, B. Glorieux, A.M. Seydoux-Guillaume, R. Wirth, J. Nucl. Mater 381 (3) (2008) 290e296. [8] M. Lang, F. Zhang, W. Li, D. Severin, M. Bender, S. Klaumünzer, C. Trautmann, R.C. Ewing, Nucl. Instrum. Meth B 286 (2012) 271e276. [9] C. Davoisne, M.C. Stennett, N.C. Hyatt, N. Peng, C. Jeynes, W.E. Lee, J. Nucl. Mater 415 (1) (2011) 67e73. [10] J. Lian, L.M. Wang, R.C. Ewing, L.A. Boatner, Nucl. Instrum. Meth. B 241 (1e4) (2005) 365e371. [11] S.X. Wang, L.M. Wang, R.C. Ewing, G.S. Was, G.R. Lumpkin, Nucl. Instrum. Meth. B 148 (1e4) (1999) 704e709. , G. Balakrishnan, D.M. Paul, [12] Y. Zhang, J. Jagielski, I.-T. Bae, X. Xiang, L. Thome W.J. Weber, Nucl. Instrum. Meth. B 268 (19) (2010) 3009e3013. [13] K.E. Sickafus, R.W. Grimes, J.A. Valdez, A. Cleave, M. Tang, M. Ishimaru, S.M. Corish, C.R. Stanek, B.P. Uberuaga, Nat. Mater 6 (3) (2007) 217e223. [14] M. Tang, A. Kossoy, G. Jarvinen, J. Crum, L. Turo, B. Riley, K. Brinkman, K. Fox, J. Amoroso, J. Marra, Nucl. Instrum. Meth. B 326 (2014) 293e297. [15] A.E. Ringwood, V.M. Oversby, W. Sinclair, The effects of radiation damage on synroc, in: C.J.M. Northrup (Ed.), Scientific Basis for Nuclear Waste Management, Springer US, Boston, MA, 1980, pp. 273e280. [16] J.D. Vienna, Closed Fuel Cycle Waste Treatment Strategy, Fuel Cycle Research & Development, 2015. U.S. Department of Energy Report, FCRD-MRWFD2015e20000674. [17] S.E. Kesson, C.J. Ball, J. Aust. Ceram. Soc. 24 (1) (1988) 89e99. [18] D. Men, M.K. Patel, I.O. Usov, M. Toiammou, I. Monnet, J.C. Pivin, J.R. Porter, M.L. Mecartney, J. Nucl. Mater 443 (1) (2013) 120e127. [19] B.M. Clark, P. Tumurugoti, S.K. Sundaram, J.W. Amoroso, J.C. Marra, K.S. Brinkman, Metall. Mater. Trans. E 1 (4) (2014) 341e348. [20] J.F. Ziegler, J.P. Biersack, U. Littmark, The Stopping and Range of Ions in Matter, Pergamon Press, New York, 1985. [21] S. Thevuthasan, C.H.F. Peden, M.H. Engelhard, D.R. Baer, G.S. Herman, W. Jiang, Y. Liang, W.J. Weber, Nucl. Instrum. Meth. A 420 (1e2) (1999) 81e89. [22] R.C. Ewing, W.J. Weber, F.W. Clinard, Prog. Nucl. Energ 29 (1995) 63e127. [23] M. Tang, P. Tumurugoti, B. Clark, S.K. Sundaram, J. Amoroso, J. Marra, C. Sun, P. Lu, Y. Wang, Y.B. Jiang, J. Solid State Chem. 239 (2016) 58e63. [24] T.D. Shen, Nucl. Instrum. Meth. B 266 (6) (2008) 921e925. [25] K. Jin, Y. Zhang, H. Xue, Z. Zhu, W.J. Weber, Nucl. Instrum. Meth. B 307 (2013) 65e70. [26] G. Grzinic, L.A. Bursill, D.J. Smith, J. Solid State Chem. 47 (2) (1983) 151e163. [27] G.D. Fallon, B.M. Gatehouse, J. Solid State Chem. 49 (1) (1983) 59e64. [28] M. Tang, Characterization and Tesing of Crystalline Ceramic Waste Forms, 2014. U.S. Department of Energy Report, M3FT-14LA0308011.
Heavy and light ion implantation was used to induce damage in simulated multiphase ceramic waste form materials produced from SPS and a melt-process. The multiphase waste forms responded differently under the two different irradiation conditions. Under heavy ion (Au3þ) irradiation, the displacements initiated by the ballistic processes cause bulk material to amorphize. Under light ion (Heþ) irradiation, the hollandite phase appeared to break down while the other phases remained crystalline as evidenced by emergence of diffuse scattering and new crystalline phases at the expense of hollandite. Samples fabricated by both melt-processing and SPS exhibited similar amorphization behavior when irradiated with Au3þ ions although the penetration depth of Au3þ ions was less material fabricated by SPS compared to melt-processing. This is attributed to the smaller grain sizes of the phases in the SPS materials; the grain boundaries act as sinks for the radiation induced defects. The lower dose of Heþ ions was demonstrated as a reliable way to render structural changes in the hollandite phase to be captured in the GIXRD patterns. Acknowledgements The authors would like to thank the Department of Energy's Nuclear Energy University Program (NEUP_12-3809) for supporting this project. The authors would also like to thank J. Vienna (Pacific Northwest National Laboratory) for project oversight and guidance. BMC acknowledges support from NEUP fellowship. SKS is grateful to the generous support of Inamori Professorship by Kyocera Corporation. This work made use of the electron microscopy facility of the Cornell Center for Materials Research (CCMR) with support from the National Science Foundation Materials Research Science and Engineering Centers (MRSEC) program (DMR 1120296). The authors gratefully acknowledge the financial support of the US DOE Fuel Cycle R&D program and Materials Recovery and Waste Form Development (MRWFD) Campaign under the direction of Terry Todd, National Technical Director, John Vienna, Deputy Manager, and Kimberly Gray, DOE Federal Manager. Work conducted at Savannah River National Laboratory was supported by the U.S.
of
Energy
under
contract
number
DE-AC09-
References