Open Access
Add to BibSonomyAbstract
Femtosecond laser of pulse width ~40 fs and 800 nm wavelength with a pulse energy output ~500nJ is used to modify the microstructures of ceramic wasteforms. Different compositions prepared by different processes show surface melting and solidification on exposure to laser irradiation. Structural changes are analyzed using x-ray diffraction (XRD) and scanning electron microscopy (SEM) techniques. SPS processed samples have higher density as compared to the melt-processed counterparts. No new phase is formed in any of the materials upon surface resolidification. Microstructures of the laser-modified area of all samples are similar despite the variation in the original microstructures from one another. XRD phase analysis and comparison of elemental maps show that homogenization and partial amorphisation of the surface occurred with hollandite phase retaining its crystallinity in an amorphous matrix.
© 2015 Optical Society of America
1. Introduction
Immobilization of radioactive waste streams from nuclear industry as well as various cold war era fuel reprocessing facilities in suitable host materials is one of the most challenging environmental problems of the modern day. Safe disposal of radioactive elements such as minor actinides with long half-life (Np, Am, etc.) and other fission products (Cs, Tc, etc.) is a major concern. Design of waste forms with the desired properties has been a major area of research and development [1]. The most popular and traditional method of using borosilicate glasses for waste immobilization has some drawbacks, e.g., limited solubility of some of the radionuclides in the glasses that limits the waste loading [2]. SYNROC (SYNthetic ROCk) - a multiphase composition containing with major phases, namely, hollandite, perovskite, zirconolite, and pyrochlore - has been proposed and being studied as an alternate material for effective containment of nuclear wastes [3,4]. Hollandite, with a general formula [BaxCsy][M3+2x + yTi8-2x-y]O16 (M: trivalent cation) [5], is the preferred host for Cs, Ba, and other large alkali ions, perovskite (CaTiO3) is the host lattice for alkaline-earth and rare-earth elements; and zirconolite (CaZrTi2O7)/pyrochlore (Nd2Ti2O7) accommodates minor actinides. Wide range of processes such as conventional sintering [6], hot isostatic pressing (HIP) [7], cold crucible induction melting (CCIM) [8], and spark plasma sintering (SPS) [9], have been used to prepare the ceramic waste forms. Each process, depending on the composition and process variables, would potentially result in a distinctly different microstructure, which controls the waste form performance, chemical durability, and radiation resistance. Morphological features like porosity and secondary parasitic phases can adversely affect the chemical durability of waste forms produced.
High-energy laser sources have been used for different processes such as laser beam welding, surface cladding, surface alloying, ablation, surface melting, and quenching [10]. Material processing using both continuous laser and pulsed laser sources has been long recognized as a powerful tool for rapid processing of materials. Laser-material interaction is mainly defined by the energy absorbed by the material, which in turn is determined by beam intensity and exposure time. Precision processing is possible via high-intensity pulses with very short pulse duration. These ultra-short pulses help achieve extremely high heating and cooling rates in a tightly focused region with minimum heat-affected zones beyond the region [11].
Electrons are initially excited into higher unoccupied states when laser pulses start interacting with a material, particularly a semiconductor. Energy is then transferred into the lattice when the excited electrons relax to the ground state. Different mechanisms or regimes are possible by which energy is transferred to the lattice from the electrons, each of them having a characteristic time scale [12–14]. After phonon absorption, during the initial timescale of 10 −100 femtoseconds (fs), carrier-carrier scattering occurs, followed by thermalization of carriers that occurs in picosecond time regime (10−13 – 10−12 s). Once the excited carriers reach equilibrium with the lattice, the temperature of the solid is defined, resulting in all subsequent thermal events. Detailed discussions on energy distribution mechanisms during electron relaxation can be found elsewhere [12, 13, 15, 16]. Bloembergen has pointed out that the heating rates could reach as high as 1014 K/s with the use of nanosecond pulses [17]. Hence, structural changes resulting by the use of laser pulses longer than a picosecond are thermal in nature. Several models, based on well-defined thermophysical properties of the material, have been developed and validated with experimental data [18–20].
In sub-picosecond regime, carrier excitation occurs by laser pulses that are shorter than the characteristic phonon emission time. Once the free carrier density reaches high enough, lattice becomes unstable. These structural changes are non-thermal since the lattice would still remain cold. However, it has been reported that both thermal and non-thermal melting can occur during femtosecond laser processing depending on the laser fluence [21]. Sokolowski-Tinten and associates have demonstrated the effects of fluence on laser-material interaction on GaAs semiconductors experimentally [22]. Depending on how high the fluence is above the melting threshold, different events such as heating and cooling, ultrafast melting and quenching, ablation would occur.
In this study, we report the interaction of femtosecond laser pulses with different multiphase ceramic waste forms. We have attempted to modify the surface microstructure, which inherently had multiple phases/flaws that would potentially deteriorate the waste form performance. Since this work is first-of-its kind, no data on thermophysial parameters or melting threshold of the materials used in the study is available. Hence this work is exploratory in nature, mainly focused on characterizing the laser-modified microstructures in an attempt to enhance the durability of waste forms.
2. Experimental
2.1. Material composition and processing
Compositions of interest for laser exposure are multiphase based compositions mainly targeted at phases – hollandite, perovskite, pyrochlore, zirconolite, and rutile. Two different compositions synthesized by two different processes, melt-processing and SPS, were used in this study. Different materials and corresponding notations used are described in Table 1. In melt-processing, the stoichiometric mixtures of carbonates/oxides are taken in an Al2O3 or Pt crucible, heated up to 1500°C, held for 0.5 h and cooled back room temperature. SPS process is of interest because of the high density of the end product and shorter processing time (~20 min). Starting material is taken in a graphite die and on application of uniaxial pressure and DC current, resistive heating of the die occurs resulting in heating rates up to 1000 K/min [23]. The two compositions, Cr - Multiphase and Cr/Al/Fe - Multiphase, were designed by Savannah River National Laboratory (SRNL) to simulate waste stream compositions. To reduce the melting temperature, Cr3+ in Cr-Multiphase precursor material is partially substituted by Fe3+ and Al3+ to form Cr/Al/Fe - Multiphase composition. Consequently, distinct microstructural features are anticipated during melt-processing.
Table 1. Different materials used in this study and their corresponding notations
2.2. Laser irradiation experiments
For our laser processing experiments, a commercial femtosecond laser system, FEMTOSOURCE Scientific XL 500 (FEMTOLASERS Produktions GmbH, Austria) was used. It is a Ti:Sapphire based high energy oscillator system that provides pulses with energies of 500 nJ, a pulse duration of less than 50 fs and at a repetition rate of 5 MHz. The XL 500 system has three major components: a laser head with a pump source emitting continuous green light of 532 nm, a Ti:Sapphire oscillator chamber, a laser cavity including a multi-pass Herriot cell and a prism compressor to compress the pulses into femtosecond timescale. The pump source is a solid-state diode laser operating in continuous wave (CW) mode. Light beam from the pump hits the Al2O3 crystal doped with Ti3+ ions (Ti:Sapphire) in the crystal chamber resulting in the excitation of Ti3+ ions which on relaxing to the ground state emit radiation corresponding to longer wavelengths. In our system, the output wavelength was tuned to 800 nm. All the lasing experiments were done with the laser beam in mode-locked mode. One of the end mirrors in the optical cavity is a saturable Bragg refelector (SBR) crystal that is responsible for mode-locking mechanism. The mode-locked laser beam passes through the prism compressor resulting in ultra-short pulses that are directed to the sample fixed to a XYZ translation stage (Thorlabs, USA). A plano-convex focusing lens, of focal length 100 mm, was used to focus the beam on the flat surface of the sample. Diameter of the beam focused on the sample, as measured by CCD camera, was observed to be 55- 60 μm. Average power of the laser beam, before the focusing lens, measured by a power meter was maintained close to 2.3 W. The peak power of individual pulses and fluence were calculated based on the geometry of the optics and the instrument specifications. Pulse peak power was 8.5 MW and the fluence calculated was 0.025 J/cm2.
Irradiation experiments were done on polished surfaces of the samples held at a normal incidence to the laser beam. The translation stage was set so that the surface to be modified was at focal point and the stage was moved in a two-dimensional plane normal to the beam. In this way the sample was rastered by laser beam in parallel lines, 0.2 mm distant apart, across an area of 1 cm2 at a speed of 2 mm/s.
X-ray diffraction (XRD) using D-2 Phaser (Bruker AXS, USA) was used to analyze the phase transformations. Laser irradiated surface were scanned for diffraction peaks over 2θ range, 10 – 80°, at a rate of 0.015°/s. Samples were cut through the cross section, perpendicular to the direction of laser rastering lines, mounted, and polished. Microstructural transformations from laser-modified surface to the base material were observed by a FEI Quanta 200 F scanning electron microscope (FEI, Hillsboro, Oregon, USA). Phase changes and elemental redistribution were investigated by collecting elemental maps using JEOL JXA-8200, WD/ED Combined Electron Probe Microanalyzer (EPMA) (JEOL Ltd., Tokyo, Japan). Elemental scans were performed at an accelerating potential of 15kV, over an area of 225 × 225 μm2, step size of 0.5 μm in x- and y- directions, and a dwell time of 80 ms at each point.
3. Results and discussions
Two different preparation processes yielded similar phase assemblages but very different microstructures. Figure 1 compares the back-scattered electron (BSE) images of all materials used. Different phases and corresponding labels are indicated in the images. Hollandite (H) was observed to be the major phase in all the samples followed by perovskite (P) and a Zr- rich phase (Z) that could not be identified from XRD. For either composition, SPS process produced dense microstructures with little porosity. The phase distribution was similar in Cr-MP-SPS and CAF-MP-SPS with smaller grain size (< 0.5 – 10 μm). Melt-processed samples were porous, had larger grains (5 – 30 μm), and showed definite grain boundaries within and phase boundaries between different phases when compared to respective SPS samples. In both cases, Cr-MP-Melt and CAF-MP-Melt, the dark grey hollandite phase is randomly distributed in a matrix of other bright phases, perovsite, and Zr-rich phase. Visual observation of as-prepared Cr-MP-Melt indicated minimum bulk melting, as supported by the BSE image. This suggested that the maximum temperature during the melt-processing of Cr-MP-Melt could have been below the liquidus of the system resulting in such a microstructure. CAF-MP-Melt sample flowed better at the processing temperatures due to lower melting point, resulting in a relatively dense and coarse-grained microstructure.
Fig. 1 Microstructures of different waste forms indicating the phase distribution – hollandite (H), perovskite (P), Zr-rich phase (Z) and TiO2 (T).
During the rastering process, when the laser pulse hit the sample, a bright white light emission due to scattering of the laser light was observed at the earliest point of exposure. Figure 2 shows a representative of the consequent drastic microstructural change on the surface. Surface melting and solidification occurred resulting in a homogeneous surface layer. The movement of the molten material along the beam during rastering and formation of quench cracks due to the high cooling rates can be observed in Fig. 2 that resembles a weld-like microstructure. Our initial experiments showed that the extent of heat-affected zone (HAZ) increased with slower rastering speeds. Different parameters that are affected by rastering speed – heat conductivity of liquid and solid phases, their relative proportions and the rate of solidification – could determine the extent of HAZ. In this study, to obtain a homogeneous surface after laser processing, a relatively lower speed (2 mm/s) was used. Width of HAZ resulted from these experiments is in the range of 100 – 150 μm, depending on the material as discussed further in more detail. Figure 3 presents more details on structural changes that took place during surface re-solidification. Polished cross sections of CAF-MP-Melt, CAF-MP-SPS, Cr-MP-Melt and Cr-MP-SPS are compared in Fig. 3. The extent of laser-modified region was seen to be dependent on microstructural features. It ranged from 50 to 100 μm in CAF-MP-Melt and is close to 50 μm in CAF-MP-SPS. Laser-induced micro cracks were observed to be 1 - 3 μm wide and extended deep down the affected region. In Cr-MP-Melt, the paths of individual laser beam line scans can be seen indicating a minimum volume of heat-affected region. Energy intensity on the regions away from the center of the beam would be expected to be lower than at the center. However, this pattern was not observed throughout the cross section of the samples. Heat conductivity, apart from being a material property, is also strongly dependent on microstructural features like grain boundaries, phase boundaries, porosity etc. All these microstructural features, coarse grains and porosity among others, associated with melt-processed samples affect the laser processing behavior. Since CAF-MP-Melt melts at a lower temperature than Cr-MP-Melt, molten surface would have higher fluidity resulting in a relatively homogeneous surface. These compositional effects are not seen in dense SPS samples.
Fig. 2 Microstructure (top view) showing the effect of laser irradiation on the surface of a melt-processed sample (CAF-MP-Melt).
3.1. Microstructural effects
Although the volume of laser-modified region was dependent on the original microstructure, the modified microstructures were found to be similar for all the four materials used in this study. It can be seen from Fig. 4 that the re-solidified surface mainly consisted of needle-like long dark gray phases in a light gray matrix. On laser exposure, the originally distinctly different microstructures were subjected to extremely high heating and quenching rates resulting in similar morphology of a solidified material in all cases. X-ray phase analysis of the laser-modified show no additional phases compared to pristine samples. Figure 5 compares the XRD plots of laser treated and untreated surfaces. Different peaks are labeled. Relative changes in intensities on modified surfaces can be observed for each of the materials. A common trend in all materials is a measurable drop in the intensities of peaks corresponding to the phases other than hollandite (H). Considering this along with the microstructural information of laser-modified region, we attribute these changes to possible amorphization of the material on the surface.
Fig. 4 Microstructures showing similar phase assemblages of resolidified surface on different samples.
Fig. 5 Comparison of XRD patterns collected from original untreated surface and laser-modified surface. Different phases are labeled to the corresponding peaks – hollandite (H), perovskite (P), Zr-rich phase (Z) and TiO2 (T).
In Fig. 6, wavelength-dispersive spectroscopy (WDS) elemental maps across the laser-modified and base material show the elemental redistribution for CAF-MP-Melt. Similar maps for CAF-MP-SPS (Fig. 7), Cr-MP-Melt (Fig. 8) and Cr-MP-SPS (Fig. 9) are also presented. Although each of these materials had more than 20 elements, key major relevant elements were chosen to identify different phases. Elemental distributions observed on laser-modified surfaces were identical in all the materials. Dark gray needle-like phases were found to be rich in Ba and Cr indicating that it is hollandite phase. Ba- and Cr-rich hollandite areas are obviously seen in the melt-processed samples (Fig. 6 and Fig. 8) when compared to corresponding SPS samples (Fig. 7 and Fig. 9). Ca- and Zr- rich areas corresponding to perovskite and Zr-rich phase respectively can also be seen from the elemental distribution maps of untreated areas in each of the samples. These phases disappeared on laser melting and quenching. The light gray area of the modified surface was observed to have random distribution of different elements. Except for Ba- and Cr-rich areas (hollandite phase), no preferential segregation of other elements is seen in all these samples. Hence, based on the elemental distribution maps and XRD analysis, it is evident that partial amorphization of the surface occurs when waste forms are irradiated by femtosecond laser pulses.
Fig. 6 WDS maps showing the distributions of selected elements in Cr/Al/Fe-Multiphase prepared by melt-processing (CAF-MP-Melt). BSE indicates the corresponding back-scattered image of the scanned area. Cs rich ‘bright’ areas were evident along the cracks of resolidified material. These were not seen in the corresponding BSE image of CAF-MP-Melt (Fig. 4). This Cs segregation was an undesired effect, observed to develop with time only for epoxy-mounted samples of this composition.
Fig. 7 WDS maps showing the distributions of selected elements in Cr/Al/Fe-Multiphase prepared by SPS (CAF-MP-SPS). BSE indicates the corresponding back-scattered image.
Fig. 8 WDS maps showing the distributions of selected elements in Cr-Multiphase prepared by melt-processing (Cr-MP-Melt). BSE indicates the corresponding back-scattered image.
Fig. 9 WDS maps showing the distributions of selected elements in Cr-Multiphase prepared by SPS (Cr-MP-SPS). BSE indicates the corresponding back-scattered image.
4. Conclusion
Our results show that the surface morphology and chemistry of multiphase ceramic waste forms can be significantly modified on exposure to femtosecond laser pulses. A dense and uniform phase assemblage (SPS processed) result in a homogeneous surface layer after laser irradiation. XRD analysis, SEM micrographs and WDS maps strongly suggest that during melting and quenching of the surface by laser pulses, hollandite phase retain its crystal structure, whereas the other phases – perovskite and Zr-rich phase – transform into an amorphous phase. These findings show promise for microstructural modification and control of phase assemblage of the waste forms for long-term storage in geological media.
Acknowledgments
The authors acknowledge the support from the Department of Energy (DOE), Nuclear Energy University Program (NEUP). SKS acknowledges support from the Kyocera Corporation in the form of Inamori Professorship.
References and links
1. S. V. Stefanovsky, S. V. Yudintsev, R. Giere, and G. R. Lumpkin, “Nuclear Waste Forms,” Geol. Soc. Lond. Spec. Publ. 236, 37–63 (2004).
2. D. Caurant, Glasses, Glass-Ceramics and Ceramics for Immobilization of Highly Radioactive Nuclear Waste (Nova Science Publishers, 2009).
3. A. E. Ringwood, S. E. Kesson, N. G. Ware, W. O. Hibberson, and A. Major, “SYNROC Processes – Geochemical Approach to Nuclear Waste Immobilization,” Geochem. J. 13(4), 141–165 (1979). [CrossRef]
4. A. E. Ringwood, S. Kesson, N. Ware, W. Hibberson, and A. Major, “Immobilisation of high level nuclear reactor wastes in SYNROC,” Nature 278(5701), 219–223 (1979). [CrossRef]
5. S. E. Kesson and T. White, “[BaxCsy][(Ti,Al)3+2x+yTi4+8-2x-y]O16 Synroc-type hollandites: I. Phase chemistry,” Proc. R. Soc. Lond. A. Mat. 405, 73–101 (1986).
6. V. Aubin-Chevaldonnet, D. Caurant, A. Dannoux, D. Gourier, T. Charpentier, L. Mazerolles, and T. Advocat, “Preparation and characterization of (Ba,Cs)(M,Ti)8O16 (M = Al3+, Fe3+, Ga3+, Cr3+, Sc3+, Mg2+) hollandite ceramics developed for radioactive cesium immobilization,” J. Nucl. Mater. 366(1–2), 137–160 (2007). [CrossRef]
7. Y. Zhang, H. Li, and S. Moricca, “Pyrochlore-structured titanate ceramics for immobilisation of actinides: Hot isostatic pressing (HIPing) and stainless steel/waste form interactions,” J. Nucl. Mater. 377(3), 470–475 (2008). [CrossRef]
8. S. Stefanovsky, A. Ptashkin, O. Knyazev, S. Dmitriev, S. Yudintsev, and B. Nikonov, “Inductive cold crucible melting of actinide-bearing murataite-based ceramics,” J. Alloys Compd. 444–445, 438–442 (2007). [CrossRef]
9. B. M. Clark, P. Tumurugoti, S. K. Sundaram, J. W. Amoroso, J. C. Marra, and K. S. Brinkman, “Microstructures of Melt-Processed and Spark Plasma Sintered Ceramic Waste Forms,” Metall. Mater. Trans. E 1(4), 341–348 (2014).
10. C. Suryanarayana, Permagon Materials Series, Volume 2: Non-equilibrium (Elsevier, 1999), Chap. 5.
11. C. Momma, S. Nolte, B. N. Chichkov, F. Alvensleben, and A. Tunnermann, “Precise laser ablation with ultrashort pulses,” Appl. Surf. Sci. 109–110, 15–19 (1997). [CrossRef]
12. D. von der Linde and K. Sokolowski-Tinten, “The physical mechanisms of short-pulse laser ablation,” Appl. Surf. Sci. 154–155, 1–10 (2000). [CrossRef]
13. D. von der Linde, K. Sokolowski-Tinten, and J. Bialkowski, “Laser-solid interaction in the femtosecond time regime,” Appl. Surf. Sci. 109–110, 1–10 (1997). [CrossRef]
14. S. K. Sundaram and E. Mazur, “Inducing and probing non-thermal transitions in semiconductors using femtosecond laser pulses,” Nat. Mater. 1(4), 217–224 (2002). [CrossRef] [PubMed]
15. K. Sokolowski-Tinten, J. Bialkowski, and D. von der Linde, “Ultrafast laser-induced order-disorder transitions in semiconductors,” Phys. Rev. B Condens. Matter 51(20), 14186–14198 (1995). [CrossRef] [PubMed]
16. J. Siegel, PhD Thesis, Instituto de Óptica “Daza de Valdés” Consejo Superior de Investigaciones Científicas Madrid, Spain, November, 1998.
17. N. Bloembergen, “Pulsed Laser Interactions With Condensed Matter,” Proc. MRS 51, 3 (1985). [CrossRef]
18. P. A. Atanasov, E. D. Eugenieva, and N. N. Nedialkov, “Laser drilling of silicon nitride and alumina ceramics: A numerical and experimental study,” J. Appl. Phys. 89(4), 2013–2016 (2001). [CrossRef]
19. C. Y. Ho and J. K. Lu, “A closed form solution for laser drilling of silicon nitride and alumina ceramics,” J. Mater. Process. Technol. 140(1–3), 260–263 (2003). [CrossRef]
20. P. P. Pronko, S. K. Dutta, D. Du, and R. K. Singh, “Thermophysical effects in laser processing of materials with picosecond and femtosecond pulses,” J. Appl. Phys. 78(10), 6233–6240 (1995). [CrossRef]
21. C. Wu, PhD Thesis, Harvard University, USA, August, 2000.
22. K. Sokolowski-Tinten, H. Schulz, J. Bialkowski, and D. von der Linde, “Two distinct transitions in ultrafast solid-liquid phase transformations of GaAs,” Appl. Phys., A Mater. Sci. Process. 53(3), 227–234 (1991). [CrossRef]
23. R. Riedel, I.-W. Chen, Ceramics Science and Technology, Volume 3, Synthesis and Processing (Wiley-VCH, 2012), Chap. 8.
