Figure 1. Schematic of the graphite die setup in SPS.
Spark plasma sintering (SPS) can produce high density materials in minutes instead of hours or days by applying pressure and Joule heating simultaneously. Pyrochlore (Nd2Ti2O7) is one of the major phases targeted in multiphase ceramic waste forms. Single-phase Nd2Ti2O7 was synthesized by solid-state reaction prior to SPS. Effects of temperature and time of synthesis on the density and microstructure of the sintered product were studied. The reactive sintering of precursor oxides will be the topic of a future communication. The densification behavior was studied by halting the sintering at different temperatures and by holding for different amounts of time at 1140°C, chosen based on initial sintering curves at various temperatures. Sintered samples were characterized using scanning electron microscopy (SEM) and X-ray diffraction (XRD). Near theoretical density (99%) was achieved by sintering up to a temperature of 1225°C. Preferred orientation of the (001) and (012)-type planes was seen in all samples. The initial stage of sintering was governed by bulk deformation through neck formation and growth and the later stage involved a slower mass transport process. The average grain size of samples sintered at 1250°C with no dwell time was 4.67 μm with grains ranging from 1–15 μm. These results are presented and discussed in terms of optimization of the sintering of multiphase ceramic waste forms.
Development of ceramic multiphase waste forms is ongoing1,2. Currently, there is an ongoing collaborative project between Alfred University and Savannah River National Laboratory supported by the Department of Energy’s Nuclear Energy University Program (NEUP). High density and complete phase formation may be critical to the performance of the waste form3. The rapid sintering of materials by SPS is desired to limit the volatility of certain elements in the waste stream. The sintering behavior of individual phases must be understood to improve waste form design. This work is designed to be a preliminary study of the sintering behavior of Nd2Ti2O7, a pyrochlore phase of interest to the waste form community, by SPS under the NEUP project.
Pyrochlore is one of the ceramic phases of interest for the immobilization of nuclear waste due to its ability to incorporate actinides and rare-earth elements (REE) into its structure. Pyrochlore has the composition of A2B2X6X’, where A is a 2 or 3 + cation, B is a 4 or 5 + cation, and X and X’ are anions (O2-)4. The structure of pyrochlore is related to a defect fluorite and can be thought of as built up layers of perovskite sheets5.
Sintering of Nd2Ti2O7 is traditionally done by cold pressing and sintering in air at elevated temperatures between 1300 and 1500°C for multiple hours. The maximum reported density is ~96% theoretical density6,7. The sintering times and temperatures are dependent on synthesis route (sol-gel vs. solid-state, for example). Shorter times and lower temperatures are desired for the fabrication of ceramic multiphase waste forms to avoid volatilization of elements within the waste stream1, creating a drive to find an alternate sintering method.
Spark plasma sintering is an advantageous sintering technique due to the short sintering times that can be achieved. Pressure and a pulsed electric current are applied simultaneously. The current generates resistive heating of the graphite die (and the sample if the sample is electrically conductive), which creates fast heating rates of up to 1000°C/min8. SPS minimizes temperature gradients within the sample8, allowing for homogenous sintering. The shorter sintering times and lower temperatures limit grain growth that occurs with most forms sintering, which may affect properties such as aqueous chemical durability.
The manner in which a material densities when processed by SPS remains to be understood and varies by material. Many sintering mechanisms for SPS have been suggested, such as localized melting caused by plasma formation, plastic deformation and grain rotation9,10. The evolution of the microstructure of a material may contain the information needed to extract the mechanism of densification, although this may prove difficult due to the rapid densification of most materials via SPS.
Neodymium titanate powders used in this study were synthesized via solid-state reaction of mixed oxide precursors. Titanium oxide (99.99%, Alfa Aesar) and neodymium oxide (99.99%, Alfa Aesar) powders were dried at 600°C prior to weighing to remove any water on the surface. The oxides were then mixed together for thirty minutes using a mixer mill in a mixture of deionized water and ethanol with alumina media in a high-density polyethylene (HDPE) jar. The mixed powders were separated from the mixing media with a sieve and dried overnight at 100°C. The resulting powder was calcined at 600°C for 24 h to remove hydroxides that may have formed. Pellets of 20mm diameter were pressed using 100 MPa pressure and pre-reacted at 1350°C for 24 h. To complete the reaction, the pellets were crushed, repelletized and pre-reacted for 12 h at 1350°C then crushed into the resulting powder for SPS.
Samples were then sintered via SPS using a FCT HP D 25 (FCT Systeme GmbH, Rauenstein, Germany) furnace with graphite dies and punches. A schematic of the die setup is shown in Figure 1. All samples were sintered using a heating rate of 100°C/min and a pressure of 54 MPa. Microstructure studies were carried out on specimen sintered to a maximum temperature of 1250°C with no hold time. Densification mechanism studies were carried out on two sets of samples. The first group consisted of samples sintered to maximum temperatures of 1100, 1140, 1150, 1175 and 1225°C with no hold time. The second set contained samples sintered isothermally at 1140°C for 10, 20 and 30 minutes. The parameters of all SPS runs and the resulting density are summarized in Table I. No replicates were made.
Figure 1. Schematic of the graphite die setup in SPS.
Table I. List of samples used to analyze the sintering behavior by SPS.
Density of sintered samples was measured using Archimedes method and was compared to the theoretical density of Nd2Ti2O7 of 6.107 g/cm3. The samples were etched via thermal etching at 1350°C for 30 minutes to reveal the microstructures and the grain size and distribution was studied using ImageJ (National Institutes of Health, USA). Etched and fracture surfaces were examined using a FEITM Quanta 200 F scanning electron microscope (FEI, Hillsboro, Oregon, USA). XRD was performed on monolith samples resulting from SPS using a D-2 Phaser (Bruker, Massachusetts, USA) to verify phase purity.
The typical sintering behavior of Nd2Ti2O7 when processed using a heating rate of 100°C/min and a pressure of 54 MPa is displayed in Figure 2. The piston speed and temperature are shown as a function of time. A positive piston speed indicates when the sample is contracting. The piston speed up to ~3 minutes is due to the applied pressure packing the powder. There is a temperature variation at ~4 minutes due to a temperature ramp correction which creates a small bump in the piston speed and is an artifact. As the sample is heating up, a negative piston speed is seen, indicating thermal expansion. Sintering occurs from 1140°C to 1215°C as can be seen by the sharp peak, indicating compaction. The peak at ~12 minutes is due to contraction upon cooling. The entire sintering process (vacuum, pressing, temperature ramp and cooling) takes approximately 25 minutes.
Figure 2. Sintering curve of Nd2Ti2O7 sintered to a 1225°C.
Preferred orientation of the (001) and (012)-type planes is seen in all sintered samples and can be seen by comparing the indicated peaks in the XRD patterns in Figure 3. Other materials, such as TiB211 and Al2O312, have also shown preferred orientation after the SPS process. The presence of preferred orientation can have an effect on mechanical properties, so it is possible that it will impact properties of importance to waste form technology, such as aqueous chemical durability. The origin and possible effect of this will be studied and presented in a future communication.
Figure 3. XRD patterns of powder and SPS Nd2Ti2O7 samples. The increased relative intensity of the peaks at ~14, 21 and 28°2θ when compared to the peak at ~30°2θ indicates the presence of preferred orientation.
The densities of sample sintered via SPS (listed in Table I) are shown in Figure 4. Near theoretical density (99%) can be attained by heating to 1225°C at 100°C/min and 54 MPa pressure. The entire sintering process (vacuum, pressing, temperature ramp and cooling) takes approximately 25 minutes. Comparatively, conventional sintering methods requires hours at 1500°C6,7.
Figure 4. Relative density of Nd2Ti2O7 samples sintered at different temperatures and isothermally at 1140°C. The hold times only pertain to the samples sintered isothermally.
The density increases relatively linearly until the samples are 95% theoretical density, beyond which appreciable densification no longer occurs. Densification as a function of time was studied using isothermal sintering processing at 1140°C, the onset of sintering according to the piston speed. In the first 10 minutes of isothermal sintering, the density increased but attained a similar apparent limit of 95% dense within 30 minutes.
In order to determine the behavior of the sample during sintering, the process was halted at select temperatures below 1225°C. Examination of the microstructure shows that sintering occurs during the thermal expansion before the sintering peak in Figure 2. Figure 5 displays that no sintering is seen by piston movement at 1140°C, but an SEM micrograph of a sample sintered to 1140°C shows that larger particles have sintered together.
Figure 5. Piston travel vs. temperature in a) shows when the SPS detects sintering and b) shows sintering of large particles by 1140°C.
The sintering of large particles occurs before the piston speed indicates sintering. The stages of sintering are shown in Figure 6, with arrows pointing toward areas of interest. By 1100°C, some large particles have already fused together while others are still undergoing the sintering process of neck formation and growth. The smaller particles seemed to be thermally shielded by the larger particles and begin to form necks around 1140°C. Sintering is nearly completed by 1175°C, as can be seen by the lack of porosity. To reach 99% relative density by 1225°C, the slower sintering process of mass transport occurs13.
Figure 6. Fracture surfaces of Nd2Ti2O7 samples sintered to different temperatures.
During the isothermal sintering at 1140°C, the small particles have already fused together after 10 minutes. The sintering process is then governed by the slower mass transport process across these particles and pores, and reaches 95% relative density after 30 minutes of sintering. Fracture surfaces of isothermally sintered samples are shown in Figure 7.
Figure 7. Fracture surfaces of Nd2Ti2O7 samples isothermally sintered at 1140°C.
The average grain size and the grain size distribution were determined on samples sintered at 1250°C. The average grain size was 4.67 μm, and the grain sizes ranged from 1 to 15 μm. The revealed microstructures from two different areas on the sample are shown in Figure 8. Elongated grains show the effect of pressure applied. Pores are seen at some of the triple points.
Figure 8. Microstructure of Nd2Ti2O7 revealed by thermal etching.
Near theoretical density of Nd2Ti2O7 was achieved via SPS to 1225°C using a heating rate of 100°C/min and 54 MPa pressure. Preferred orientation of the (001) and (012)-type planes is seen in all samples. The densification behavior was studied by the microstructural evolution of samples sintered to different temperatures and isothermally at 1140°C. It was seen that large particles begin sintering prior to the outputs on the SPS indicate sintering. With increasing temperature, Nd2Ti2O7 sinters by bulk deformation through neck growth, then sinters by mass transport after reaching about 95% relative density. When sintered isothermally, it is mainly the mass transport that controls the sintering. The average grain size of samples sintered to 1250°C was 4.67 μm, with grain sizes ranged from 1 to 15 μm.
The authors acknowledge the support from DOE’s NEUP for our project. We also acknowledge the advice and feedback from Dr. Jim Marra of Savannah River National Laboratory and Dr. John Vienna from Pacific Northwest National Laboratory. SKS acknowledges the support from Kyocera Corporation for the support of the Inamori Professorship.
1. K. S. Brinkman, K. M. Fox, and M. Tang, “Ceramics for Immobilization of Advanced Fuel Cycle Reprocessing Wastes,” FCRD-SWF-2011-000310 SRNL-STI-2011-00516, September 22, 2011.
2. J. L. Woolfrey. K. D. Reeve, and D. J. Cassidy, “Accelerated Irradiation Testing of Synroc and Its Constituent Minerals Using Fast Neutrons,” J. Nuc. Mater., 108–109 739–47 (1982).
3. M. L. Carter, M. W. A. Stewart, S. H. F. Leung, and M. Colella, “Microstructures of Inactive Synroc-C Samples Produced by Different Hot-Consolidation Methods,” Ceramic Transactions, 71 491–504 (1996).
4. K. B. Helean, S. V. Ushakov, C. E. Brown, A. Navrotsky, J. Lian, R. C. Ewing, J. M. Farmer, and L. A. Boatner, “Formation Enthalpies of Rare Earth Titanate Pyrochlores,” J. Solid State Chem., 177 1858–66 (2004).
5. J. Sloan and R. J. D. Tilley, “Phase Formation in the Nd2Ti2O7-SrTiO3 System at 1350°C in the Presence of V2O5, CuV2O6, or SrCuO2,” J. Solid State Chem., 121 324–31 (1996).
6. R. V. Pai, S. K. Mukerjee, and V. Venugopal, “Synthesis, Thermo Physical and Electrical Properties of Nd2-XLixTi2O7-δ,” Solid State Ionics, 187 85–92 (2011).
7. G. Winfield, F. Azough, and R. Freer, “Dip224: Neodymium Titanate (Nd2Ti2O7)Ceramics,” Ferroelectrics, 133 [1] 181–6 (1992).
8. M. Nygren and Z. Shen, “Hot Pressing and Spark Plasma Sintering”; in Ceramics Science and Technology: Volume 3: Synthesis and Processing. Edited by R. Riedel and I.-W. Chen. Wiley-VCH Verlag GmbH & Co. KGaA, 2012.
9. Y. J. Wu, J. Li, X. M. Chen, and K. Kakegawa, “Densification and Microstructures of PbTiO3 Ceramics Prepared by Spark Plasma Sintering,” Mat. Sci. Eng. A, 527 5157–60 (2010).
10. R. Chaim, “Densification Mechanisms in Spark Plasma Sintering of Nanocrystalline Ceramics,” Mat. Sci. Eng. A, 443 25–32 (2007).
11. J. Schmidt, M. Boehling, U. Burkhardt, and Y. Grin, “Preparation of Titanium Diboride TiB2 by Spark Plasma Sintering at Slow Heating Rate,” Sci. Technol. Adv. Mat., 8 [5] 376–82 (2007).
12. D. Pravarthana, D. Chateigner, L. Lutterotti, M. Lacotte, S. Marinel, P. A. Dubos, I. Hervas, E. Hug, P. A. Salvador, and W. Prellier, “Growth and Texture of Spark Plasma Sintered Al2O3 Ceramics: A Combined Analysis of X-Rays and Electron Back Scatter Diffraction,” J. Appl. Phys., 113 [15] 153510–7 (2013).
13. S. Diouf and A. Molinari, “Densification Mechanisms in Spark Plasma Sintering: Effect of Particle Size and Pressure,” Powder Technol., 221 220–7 (2012).