NANO LETTERS
Ultralow-Temperature Superplasticity in Nanoceramic Composites
2005 Vol. 5, No. 12 2593-2597
Guo-Dong Zhan* Department of Chemical and Biological Engineering, the UniVersity of Colorado at Boulder, Colorado 80309
Javier E. Garay Department of Mechanical Engineering, the UniVersity of California, RiVerside, California 92521
Amiya K. Mukherjee Department of Chemical Engineering and Materials Science, The UniVersity of California, DaVis, California 95616 Received October 12, 2005; Revised Manuscript Received November 11, 2005
ABSTRACT We report the successful demonstration for low-temperature and high-strain-rate superplastic forming of nanoceramic composites for the first time. Porous preforms of nanoceramic composites that were partially densifed at low temperatures were superplastically defomed by SPS at the record low temperatures of ∼1000 to 1050 °C, which are comparable to those of Ni-based superalloys. The maximum strain rate achieved is over 10-2 s-1, and a compressive strain over 200% can be obtained without cracking. The final products have nanosized grains with excellent optical properties. The present findings present a new strategy for nanoceramic superplasticity, demonstrating that a more practical application of nanoceramic superplasticity is not in the shaping of already-dense materials but in the near-net-shape forming of partially dense parts.
Superplasticity has evolved as an attractive forming method for ceramic materials, which are hard to shape by machining.1-4 However, superplasticity in ceramics is characterized by rather low forming rates and the forming time is of the order of 3-6 h.5-7 This makes the process commercially unattractive. Superplasticity can be defined as the ability of a polycrystalline material to undergo large elongations prior to failure. It can be described by the phenomenological equation: ´ )
() ()
A b kBT d
p
( )
Q σ n D0 exp E kBT
(1)
In this equation, ´ is the strain rate, A is a dimensionless constant that depends on the mechanisms, kB is Boltzmann’s constant, b is the Burgers vector, d is the grain size, σ is the applied stress, E is Young’s modulus, D0 is the preexponential factor for diffusion, Q is the activation energy for superplastic flow, and T is the temperature in degrees Kelvin. The constant n is the stress exponent, and p is the grain-size exponent. The phenomenon usually occurs in fine-grained * Corresponding author. E-mail:
[email protected]. 10.1021/nl0520314 CCC: $30.25 Published on Web 11/19/2005
© 2005 American Chemical Society
materials under conditions of moderate temperatures (T > 0.5 Tm) and moderate-to-slow strain rates (10-6 to 10-2 s-1).1-3 On the basis of this equation and a typical p value of 2-3, a 10-fold reduction in grain size can theoretically give rise to a factor of 100-1000 increase in strain rates under an identical applied stress and temperature. Alternatively, the same grain size reduction could give rise to a lower deformation temperature or a decrease in flow stress (at the same strain rate). Therefore, it is expected that nanocrystalline ceramics, which are often characterized by having grain sizes below 100 nm, can be made to deform in superplasticity at lower temperatures and higher strain rates as compared to their microcrystalline counterparts. However, evidence for room-temperature4 superplasticity in nanostructured materials is not very convincing. To date, compressive superplasticity has been observed in several studies on nanocrystalline ceramics, but elevated temperatures are required for this phenomenon to manifest itself.5-7 In addition, the low strain rates (typically, 10-5 to 10-4 s-1) encountered in superplasticity of fine-grained ceramics have limited their applicability. Recently, high-strain-rate superplasticity (HSRS), which is usually referred to as the demonstration of high ductility at strain rates around 10-2 s-1or greater, has stimulated much
scientific and technological interest.8-14 The first HSRS report in ceramic materials is Kim et al.,15 who reported a composite ceramic material with a submicrometer grain size that exhibits superplasticity at strain rates up to 1 s-1 at 1650 °C. Recently, the current investigators produced a nanocrystalline ceramic composite with the same composition as that used by Kim et al. and found that the deformation temperature can be reduced to 1400 °C.16 However, the deformation temperature is still rather high and makes it very difficult for commercial utilization of superplasticity as a forming method. The realization of nanoceramic superplasticity into practice remains a challenging issue. In the present study, we strive to bring the HSRS deformation temperature for nanoceramic composites down to 1050 °C or lower by utilizing spark-plasma-sintering (SPS) apparatus as a consolidation as well as a forming avenue. As pointed out by Mayo,2 minimizing grain growth during consolidation is essential to producing a nanocrystalline ceramic that will exhibit enhanced strain rates during the initial stages of a forming operation. Maintaining this superior strain rate capability will, however, require a minimization of the dynamic grain growth that occurs during deformation. Because of slow heating rates, the existing conventional sintering and forming methods could not meet these requirements. SPS, a relatively newly developed sintering process based on the theory of high-temperature plasma (spark plasma) momentarily generated in the gaps between powder materials by electrical discharge during ON-OFF DC pulsing, has provided an overwhelming advantage over these conventional methods.17 The process is similar to conventional hot pressing, where the powders are loaded in a die and a uniaxial pressure is applied during the sintering. However, instead of using an external heating source, a pulsed direct current is allowed to pass through the electrically conducting pressure die that allows the die to act as a heating source as well so that the sample is heated from both outside and inside. Principally, there are three factors that can contribute to the rapid densification process: (i) the application of a mechanical pressure; (ii) the use of rapid heating rates; and (iii) the use of pulsed direct current, implying that the samples are also exposed to an electrical field. The generated spark discharge and/or plasma is said to clean the surfaces of powders from adsorbed species, such as CO2, H2O, and OH-. The cleaning process is expected to enhance the grain-boundary diffusion processes that, together with the proposed spark discharges and/or plasma processes, can promote transfer of material and also enhance the densification by enhancing diffusion along particle surfaces. Thus, the unique features and the combined effects of the process provide SPS with very fast heating rates and very short holding times to consolidate powders to near theoretical densities. SPS has been demonstrated to be not only an effective sintering process for fabricating fully dense nanocrystalline ceramics and composites18-19 but also a new forming method for enhancing ceramic ductility.20-21 The first attempt to apply the SPS approach to speed up superplastic forming was by Shen et al.,20 who started with fully dense ceramics 2594
that sinter via either transient or permanent liquid-phase modes. The observed enhanced ductility is thought to be associated with the enhanced grain sliding at the boundary of the glassy/liquid phase resulting from the electric-fieldinduced motion of charged species.20-21 In the present study, the combination of sinter-forging strategies and noVel spark-plasma-sintering effects was utilized for the first time to lower deformation temperatures and enhance the deformation rates in ceramic nanocomposites. Here we started with a porous preform instead of fully dense blanks and simultaneously consolidated and superplastically formed the specimens in the SPS equipment. The use of concurrent deformation and consolidation provides nanoceramic superplasticity with many advantages. First, porous preforms are easy to produce (75-90% TD) by different methods such as pressureless sintering, plasma spray, SPS, and so forth. Second, in terms of the effects of pore-pinning, the presence of open porosity networks in the materials at densities below 90% TD was found to seriously inhibit static grain growth during sintering. However, at densities over 90% TD density, as the open pores become closed, the closed pores are not as effective as open pores in pinning grain boundaries, and thus beyond this point the grain size begins to increase dramatically.22 It has been found that the open porosity can also limit dynamic grain growth during the deformation.23 Finally, the microstructure with nanocrystalline grain size can lead to much lower deformation temperatures and higher strain rates. The starting materials used in the present study include a high-purity alumina powder doped with 500-ppm MgO and 300-ppm Y2O3 obtained from Baikowski International and cubic MgO nanopowder with a particle size of 40 nm. Alumina (86.7 wt %) and MgO (13.3 wt %) were mixed by the conventional ball-milling method with ethanol alcohol and zirconia balls in a plastic vessel for 24 h to prepare a composite powder. The composite mixture was consolidated by SPS to produce a composite consisting of 50 vol % Al2O3 and 50 vol % MgAl2O4 (through the solid reaction of Al2O3 and MgO during sintering). Microstructural observation was carried out using an FEI XL30-SFEG high-resolution scanning electron microscope with a resolution better than 2 nm. Grain sizes were estimated from high-resolution SEM of fracture surfaces. The final densities of the sintered compacts were determined by the Archimedes’ method with deionized water as the immersion medium. The phases present were determined by X-ray diffraction (XRD) using Cu KR radiation. SPS was carried out under vacuum in a Dr. Sinter 1050 SPS apparatus (Sumitomo Coal Mining Co., Japan) with a pulse duration of 3.3 ms. The pulse sequence consists of 12 pulses followed by 2 periods (6.6 ms) of zero current. The temperature was monitored with a thermocouple inserted into a “nonthrough” hole (2 mm in diameter and 5 mm in depth) in the graphite die. The compressive deformation tests of SPS-prepared partially dense nanoceramic composites were performed via the same SPS apparatus by loading one-fourth disk shape with a diameter of ∼20 mm and a thickness up to 4 mm in a graphite die of 20-mm inner diameter under Nano Lett., Vol. 5, No. 12, 2005
an uniaxial compressive stress that was applied via the graphite punches of the pressure die. Before the desired temperature was reached, a constant load of about 3-4 kN was applied to the sample, which corresponds to an initial stress of 40-50 MPa. When the desired temperature was reached, constant load rates of 45-125 N/s were applied. The compressive strain is defined here as ln(1 - ∆L/L0), where ∆L and L0 represent the shrinkage of sample height and the original height of the sample before deformation, respectively. The compressive strain rate is defined as the time derivative of the measured compressive strain. As discussed earlier, the use of pulsed direct current may generate spark discharges between the powder particles and plasma generation may also occur. Although the generation of a plasma has not been confirmed yet by direct experiment, it is likely that the electric field generates internal localized heating, impact pressure, and ionization, which promote mass transfer and accelerate localized reactions, and as a result lead to enhanced densification. It should be pointed out that a discharge process can play a major role only in the initial part of a sintering process but the other electrical field effects are operative over the entire sintering cycle, the reason being that the discharge process cannot be operative in fully dense samples. The field-enhanced mass transfer effects such as grain-boundary diffusion and grain-boundary migration processes can be effective in dense materials. To fully take advantage of these effects, we prepared partially dense materials instead of fully dense materials (the conventional approach) for subsequent superplastic deformation to take advantage of both the discharge process and possible spark plasma effect on the whole deformation process. Here, we demonstrate spark-plasma-sintering enhanced low-temperature-superplasticity in partially dense nanoceramic composites. Figure 1a is a typical microstructure of 75% theoretical density (TD) in the alumina-spinel composite produced by SPS at 980 °C. The materials have nanosized grains with a bimodal microstructure containing a certain amount of porosity. It is expected that these open porosities will not only limit static grain growth but also may be effective second-phase barriers to dynamic grain growth. In terms of these combined effects, it may be easily understood why such high deformation temperatures were observed in Shen’s work20 because the only electrical field effect could be operative for the fully dense materials during the deformation. Figure 2a and b shows the spark-plasmaenhanced superplastically formed samples at 1000 °C and 1050 °C, respectively. It can be seen that the strain rates up to 10-2/s could be achieved. A compressive strain of -0.8, equivalent to a reduction to half the original height, can be achieved without cracking. Surprisingly, the process deformation temperature is as low as 1000 °C, remarkably lower than those found for conventional superplastic ceramics. These record-low temperatures are comparable to those of Ni-based superalloys (typically, 950 °C), suggesting that the forming tools used currently for superplastic metal forming can also be used for shaping nanoceramic composites. Importantly, the present study demonstrated that the more practical use of ceramic superplasticity is not in the shapeNano Lett., Vol. 5, No. 12, 2005
Figure 1. Scanning electron microscope images of two compacts before and after being deformed by SPS. (a) Microstructure of the alumina-spinel composite produced by SPS at 980 °C before it is deformed. The material has a 75% TD. (b) Microstructures of alumina-spinel nanocomposite after deformation at 1050 °C. XRD results indicated that MgO and Al2O3 reaction to form spinel and alumina two-phase materials without any residual MgO in the final composites.
forming of already-dense materials but in the near-net-shape forming of dense parts starting from less-than-fully-dense preforms. Figure 3 shows an example of materials before and after deformation. It should be noted that the same composites did not exhibit superplasticity at these low temperatures by conventional deformation methods because both static grain growth during the slow heating and dynamic grain growth during high-temperature deformation took place. It is generally accepted that application of mechanical pressure is helpful in removing pores from compacts and enhancing densification. The increasing applied pressure during deformation is expected to provide extra driving force to promote rapid densification and grain-boundary sliding. Therefore, applying pressure at low temperature that allows the grain-boundary sliding to become kinetically favorable can enhance deformation rates. This is consistent with our findings where the strain rate increases with increasing loading rate. When a constant load instead of a constant stress is applied, the strain rates are decreased significantly. The latter is similar to the observations of Shen et al.,20 where they applied a constant load during the deformation, indicating a decreasing applied stress during deformation. It can also be noted that the strain rates are increased significantly when a little higher deformation temperature is applied (Figure 2b) even though the loading rates are slightly lower, 2595
Figure 4. Infrared spectrum of spark-plasma-enhanced superplastic formed sample at 1050 °C. The infrared spectra of the aluminaspinel composite samples were collected on an FTIR instrument (Mattson Galaxy Series FTIR 3000). The spectrometer was set to collect 16 scans in transmittance mode with a resolution of 4 cm-1 over a range from 650 nm to 16 µm. Over 40% transmittances have been obtained in the wavelength range of 3-6 µm, showing a maximum value of 65% at 5 µm. Note that there is an absorbing peak in the 4.4 µm because of the water-bending effect that comes from the environmental conditions.
Figure 2. Spark-plasma-enhanced superplastic deformation behavior of 50 vol % Al2O3/MgAl2O4 composite. The compressive strain data obtained plotted vs time for the composite at 1000 °C (a) and at 1050 °C (b).
way for fabricating truly nanocrystalline ceramic composites with novel properties. Moreover, nearly fully dense materials can be achieved in the final products through SPS superplastic deformation. Importantly, the deformed nanoceramic composites exhibit excellent infrared transparency (Figure 4). To our knowledge, infrared transparent nanoceramic composites instead of polycrystalline monolithic ceramic materials have never been reported in the literature.24 In conclusion, this new SPS approach, starting with partially dense materials and concurrent deformation and densification in the SPS apparatus, provides a new route for lowtemperature and high-strain-rate superplasticity for nanostructured materials and should impact and interest a broad range of scientists in materials research and superplastic forming technology. Acknowledgment. This work was supported by the U.S. Army Research Office (grant no. W911NF-04-1-0348) with Dr. William Mullins as the Program Manager and Office of Naval Research (grant no. N00014-03-1-0148) with Dr. Lawrence Kabacoff as the Program Manager. References
Figure 3. Nanoceramic composites before (left) and after (right) superplastic deformation at 1000 °C.
suggesting the temperature has profound effect on deformation, because of the exponential dependence of strain rate on temperature (eq 1). Finally, it is very interesting to note that a nearly nanosized microstructure has been obtained in the deformed composites (Figure 1b), whereas much grain growth can be seen in the nondeformed samples even at the same SPS temperature. These results suggest that the SPS approach may be a new 2596
(1) Gleiter, H. In Mechanical Properties and Deformation BehaVior of Materials HaVing Ultra-Fine Microstructures; Nastasi, M., Parkin, D. M., Gleiter, H., Eds.; Kluwer Academic Publishers: Boston, MA, 1993; pp 3-35. (2) Mayo, M. J. In Mechanical Properties and Deformation BehaVior of Materials HaVing Ultra-Fine Microstructures; Nastasi, M., Parkin, D. M., Gleiter, H., Eds.; Kluwer Academic Publishers: Boston, MA, 1993, pp 361-80. (3) Mukherjee, A. K. Treatise on Materials Science and Technology; Arsenault, R. J., Ed.; Academic Press: New York, 1975; Vol. 6, pp 163-224. (4) Karch, J.; Birringer, R.; Gleiter, H. Nature 1987, 330, 556. (5) Hahn, H.; Averback, R. S. J. Am. Ceram. Soc. 1991, 74, 2918. (6) Hahn, H.; Averback, R. S.; Hofler, H. J.; Logas, J. In Clusters and Cluster-Assembled Materials (Mater. Res. Soc. Symp. Proc., 206); Averback, R. S., Bernholc, J., Nelson, D. L., Eds.; MRS: Pittsburgh, PA, 1991; pp 569-580. (7) Mayo, M. J. In Superplasticity in AdVanced Materials; Hori, S., Tokizane, M., Furushiro, N., Eds.; Japan Society for Research on Superplasticity: Osaka, Japan, 1991; pp 541-550.
Nano Lett., Vol. 5, No. 12, 2005
(8) Sherby, O. D.; Nieh, T. G.; Wadsworth, J. Mater. Sci. Forum 1997, 243-245, 11. (9) Langdon, T. G. Mater. Sci. Forum 1999, 304-306, 13-20. (10) Nieh, T. G.; Wadsworth, J. Mater. Sci. Eng., A 1991, 147, 229. (11) Mukherjee, A. K.; Mishra, R. S.; Bieler, T. R. Mater. Sci. Forum. 1997, 233-234, 217. (12) Mishra, R. S.; Bieler, T. R.; Mukherjee, A. K. Acta Mater. 1997, 45, 561-68. (13) Bieler, T. R.; Mukherjee, A. K. Mater. Sci. Eng. 1990, A128, 171182. (14) Mukherjee, A. K. Mater. Sci. Eng. 2002, A322, 1-22. (15) Kim, B. N.; Hiraga, K.; Morita, K.; Sakka, Y. Nature 2001, 413, 288. (16) Zhan, G.-D.; Kuntz, J. D.; Wan, J.; Mukherjee, A. K. Mater. Res. Soc. Symp. Proc. 2003, 740, 41. (17) Omori, M. Mater. Sci. Eng. 2000, A287, 183.
Nano Lett., Vol. 5, No. 12, 2005
(18) Zhan, G.-D.; Kuntz, J. D.; Wan, J.; Mukherjee, A. K. Nat. Mater. 2003, 2, 38. (19) Zhan, G.-D.; Kuntz, J. D.; Wan, J.; Garay, J.; Mukherjee, A. K. J. Am. Ceram. Soc. 2003, 86, 200. (20) Shen, Z. J.; Peng, H.; Nygren, M. AdV. Mater. 2003, 15, 1006. (21) Shen, Z. J.; Peng, H.; Nygren, M. J. Am. Ceram. Soc. 2004, 87, 727. (22) Burke, J. E. J. Am. Ceram. Soc. 1957, 40, 80. (23) Uchic, M.; Hofler, H. J.; Flick, W. J.; Tao, R.; Kurath, P.; Averback, R. S. Scr. Metall. Mater. 1992, 26, 791. (24) Harris, D. C. Materials for Infrared Windows and Domes: Properties and Performance; SPIE, the International Society for Optical Engineering: Bellingham, WA.
NL0520314
2597
