Open Access
Add to BibSonomyAbstract
The influence of post-sintering treatment by Hot Isostatic Pressing (post-HIP) of Nd:YAG ceramics on their sintering trajectories (grain size as a function of relative density or pore size) was evaluated and compared to trajectories obtained by solid-state reactive sintering (SSRS) under vacuum. This work shows that the separation phenomenon between pores and grain boundaries observed during SSRS can be efficiently avoided thanks to suitable post-HIP conditions. As a result, highly transparent Nd:YAG ceramics with small grain size (< 2 µm) were elaborated with limited sintering aid (SiO2) content.
© 2014 Optical Society of America
1. Introduction
The application of transparent Nd:YAG (Y3-xNdxAl5O12) ceramics as solid-state laser amplifier medium was firstly demonstrated in 1995 by Ikesue et al. [1]. Nowadays, polycrystalline Nd:YAG is commercially available as laser amplifier medium with various shapes and doping levels. It has also been demonstrated that transparent polycrystalline Nd:YAG ceramics can have similar laser efficiency and can deliver even higher laser output power than Nd:YAG single-crystals [2]. Polycrystalline Nd:YAG ceramics present key advantages over single-crystals such as higher luminescent dopant concentration [3] and the ability to be fabricated into large size with complicated architecture (i.e. “composite ceramics” like multilayer, clad-core, etc.) [4].
The main limitation of ceramic materials is linked to light scattering due to residual porosity. With the presence of pores, it has been evidenced that the transparency and thus laser efficiency rapidly shrink [5,6]. Different ways are favorable to eliminate the porosity in ceramics. The initial powder compact should present homogeneous and small porosity, sintering aids (e.g. SiO2 or MgO for YAG) can favor the densification stage during sintering [7,8], and post-sintering treatments under pressure like Hot Isostatic Pressing (post-HIP) can be applied to the sintered material to decrease its porosity content. For example, many studies have reported that the transparency of ceramics such as MgAl2O4 [9], Y2O3 [10], Al2O3 [11] and Nd:YAG [12,13] can be improved thanks to post-HIP treatment under suitable conditions. In all cases, the increasing of transparency was attributed to porosity decreasing. However, the exact mechanism of porosity elimination and the link between ceramics microstructure and their optical properties remains unclear.
The objective of this study is to investigate the effect of post-HIP treatment on the sintering process of Nd:YAG ceramics. Results were compared with that obtained for pressureless reaction sintering under vacuum. The evolution of density, grain size and pore size during sintering was analyzed to establish Nd:YAG sintering trajectories with associated microstructure maps. Finally, samples microstructure was correlated to their optical properties.
2. Experimental procedure
2.1. Elaboration of Nd:YAG ceramics
Submicrometer α-Al2O3 (purity > 99.9%; Baïkowski, Annecy, France), Y2O3 (purity>99.9%; Alfa Aesar, Schiltigheim, Germany) and Nd2O3 (purity > 99.9%, Alfa Aesar) powders were mixed together in stoichiometric proportions to form YAG phase by a solid-state reaction process during sintering. Powder mixing was carried out by ball milling in water with an organic dispersant in presence of different silica content ranging from 0.07wt.% to 0.14wt.% (SiO2, purity > 99.8%, Alfa Aesar) used as sintered aid. After cold uniaxial pressing and drying, green pellets were heated under air to remove organic residues.
Then, two sintering methods were explored as described in Fig. 1.The Conventional way (n°1 on Fig. 1) by solid state reactive sintering (denoted SSRS) under vacuum. SSRS was performed at temperatures between 1550°C and 1750°C for dwell time varying from 1 h to 3 h with heating and cooling rates of 5°C.min−1. SSRS was conducted in a tungsten mesh-heated furnace under vacuum (P < 10−2 Pa). Specimens were placed in alumina crucible. The Non-conventional way (n°2 on Fig. 1) by post-HIP treatment at a temperature of 1650°C for 1 h to 6 h. Before post-HIP, Nd:YAG samples were pre-sintered at 1550°C during 30 min under vacuum. Post-HIP was conducted in a graphite mesh-heated furnace under high pressure argon atmosphere (P = 150 MPa). Specimens were placed in alumina crucible.
2.2. Characterization of samples
As-obtained samples were then characterized in terms of density, microstructure (grain size, porosity) and optical properties. Relative density was measured by the immersion method in water. Before microstructural observations, sintered samples were previously polished and thermally etched under air during 1 h at temperatures lower than the sintering one by 100°C. The microstructure of all samples was observed by scanning electron microscopy (SEM, Philips XL30, Amsterdam, The Netherlands). Mean grains diameter was estimated by the linear intercept method over at least 300 grains of 5 different micrographs to ensure good measurement reliability.
For optical measurements, all specimens were polished on both surfaces to reach average roughness inferior to 0.2 nm, flatness near to λ/10 and parallelism < 10”. Optical transmittance of transparent ceramics in the wavelength range from 200 to 1200 nm was measured by using a UV-Vis-NIR spectrophotometer (Cary 5000, Varian, USA).
3. Results and discussion
3.1. Microstructural evolution during sintering of Nd:YAG ceramics
SEM micrographs of Nd:YAG ceramics sintered by SSRS or by SSRS + post-HIP are presented in Fig. 2.Pre-sintered Nd:YAG samples were obtained by SSRS at 1550°C for 30 min. With this treatment, the relative density is of 95% with closed intergranular porosity and grain size around 1 µm (Fig. 2a). This microstructural state corresponds to the final stage of sintering and it is worth to notice that no difference depending on silica content of samples was detected at this stage. All other samples were obtained from this pre-sintered state by further SSRS or post-HIP treatments with sintering conditions mentioned in Fig. 2.
Fig. 2 SEM micrographs for Nd:YAG ceramics pre-sintered by SSRS at 1550°C for 30 min (a), 0.14% SiO2-doped (b, c) and 0.07% SiO2-doped (d, e) Nd:YAG ceramics sintered by SSRS at temperatures from 1575°C to 1750°C; 0.14% SiO2-doped (f, g) and 0.07% SiO2-doped (h, i) Nd:YAG ceramics sintered by post-HIP under 150 MPa of argon for 1 h to 6 h at 1650°C.
First, if we look at Nd:YAG samples issued from the conventional SSRS way (Fig. 2(b-e)), relative density increases as a function of sintering temperature and silica content as already reported in a previous study [14]. For the highest sintering temperature used (i.e. 1750°C), grain size increases to values close to 5 µm and intragranular porosity was detected for different silica levels as mentioned by black arrows on Fig. 2(c) and 2(e).
Second, the microstructural evolution of Nd:YAG ceramics during post-HIP at 1650°C was analyzed by SEM observations reported in Fig. 2(f-i). It can be noticed that for Nd:YAG specimen with lower silica content (0.07wt.%), grain growth remains very limited with a constant grains diameter around 1.5 µm. Concerning porosity evolution, the most of pores were eliminated even for the shorter treatment (i.e. 1 h at 1650°C). On the contrary, rapid grain growth and intragranular porosity was observed after post-HIP treatment at 1650°C-6 h for the higher silica content (0.14wt.%). Thus, such microstructural defects can be efficiently avoided by decreasing silica content associated with suitable post-HIP conditions. This result is in accordance with previous work [14] showing that the formation of intragranular porosity is mainly due to rapid grain growth activated with silica.
3.2. Sintering behavior of Nd:YAG ceramics
The sintering behavior of Nd:YAG ceramics was emphasized during conventional SSRS under vacuum with various silica content and during post-HIP treatment with limited silica content (i.e. 0.07wt.%). Sintering trajectories corresponding to grain size G as a function of Nd:YAG relative density ρ were reported in Fig. 3.Sintering temperature and time were reported on each experimental point.
Fig. 3 Sintering trajectories from grain size G as a function of relative density ρ during SSRS under vacuum or post-HIP of Nd:YAG ceramics with different silica content.
In one hand, from trajectories obtained for SSRS without post-HIP, it is possible to notice that the grain size remains extremely limited (i.e. < 2 µm) until the relative density becomes higher than 98%. For higher relative density, rapid grain growth occurs during SSRS leading to coarser microstructures with grains diameter higher than 3 µm. Figure 3 also shows that silica addition increases grain growth compared with densification. Indeed, the ratio between dG/dt and dρ/dt is increased with silica doping as exposed in previous work [14]. Such behavior would not be favorable to obtain pore-free Nd:YAG ceramics because the driving force for densification decreases as the grain size increases. Thus porosity becomes hard to eliminate during SSRS even with more severe thermal treatment conditions. Under experimental conditions used for this study, Fig. 2 shows that the formation of intragranular pores during SSRS of Nd:YAG ceramics cannot be avoided even with low silica content (i.e. 0.07wt.%).
In the other hand, Fig. 3 shows the joint evolution of grain size and density during post-HIP treatment of Nd:YAG ceramics containing 0.07wt.% of silica. In that case, extremely limited grain growth occurs until the material reaches 100% of relative density. It is also interesting to notice that for similar sintering schedule (i.e. 1650°C for 5-6 h), the grain size of Nd:YAG ceramics doped with 0.07wt.% of silica remains very close to 2 µm whatever the applied pressure during sintering (i.e. P < 10−2 Pa or P = 150 MPa of argon). For this sintering schedule, relative density remains limited to 99.6% under vacuum whereas it becomes > 99.9% under an isostatic pressure of 150 MPa. At high temperatures, the applied pressure has thus a very limited impact on grain growth whereas it promotes significantly the densification process of Nd:YAG ceramics.
As observed in Fig. 2, the separation between pores and grain boundaries becomes active in Nd:YAG ceramics during SSRS at high sintering temperatures whatever silica content. Generally, it is considered that the separation between pores and grain boundaries is explained by a lower mobility of pores compared to that of grain boundaries. In that case, a critical pore size (rc) exists leading to the separation between pore and grain boundaries (for more details, see ref [14].). This separation domain can be usefully reported on Nd:YAG sintering trajectories to obtain corresponding sintering map. Thus, the sintering behavior of Nd:YAG ceramics during post-HIP was also emphasized by drawing the sintering map with G-rp sintering trajectories in Fig. 4.The separation domain between pores and grain boundaries was denoted Sep and colored in grey (upper and lower limits were arbitrary drawn by a dotted line) with corresponding critical pore radius rc. First, it is shown that grain size increases in the same time than the grain one during SSRS. The critical pore size value, i.e. when intragranular porosity starts to be observed, is in the order of rc = 0.16 µm. Second, if we look at the influence of post-HIP treatment on Nd:YAG sintering trajectory, post-HIP leads to pore size decreasing with limited grain growth. From Fig. 2, it was also observed that porosity remains intergranular even for relative densities superior to 99.9% with no formation of intragranular porosity. The sintering trajectory during post-HIP thus does not run into a separation domain on the contrary to SSRS. These results show that post-HIP is a very efficient method to obtain pore-free Nd:YAG ceramics with extremely limited grain size (< 2 µm).
Fig. 4 G-rp sintering trajectories of Nd:YAG ceramics (0.07wt.% SiO2-doped) during SSRS or post-HIP. Closed symbols and open symbols correspond to intergranular or intragranular pore radii, respectively. Separation domain between pores and grain boundaries was colored in grey and denoted Sep with arbitrary upper and lower limits.
From these results, it can be deduced that the applied isostatic pressure (150 MPa) during post-HIP at 1650°C increases dρ/dt without increasing dG/dt on the contrary to silica used as conventional sintering aid. Under these conditions, the main driving force for densification during post-HIP treatment should be the applied isostatic gas pressure rather than the temperature. According to Nabarro-Herring model [16], pressure activates densification kinetics according to Eq. (1):
where ρ is the relative density of the ceramic material, A is a geometric constant, k the Boltzmann’s constant, T the absolute temperature, G the mean grain size at an exponent n depending on grain growth mechanism, QD the activation energy of densification and σ is the applied isostatic stress. This equation shows that densification rate can be enhanced by different ways: (1) increasing the isostatic stress; (2) increasing the temperature; (3) minimizing the grain size G before and during post-HIP. At temperatures as low as 1650°C, grain growth rate remains very limited in 0.07wt.% SiO2-doped Nd:YAG ceramics as observed in Fig. 2, thus G value can be considered constant during post-HIP treatment at 1650°C. At a constant temperature and grain size, Eq. (1) shows that dρ/dt should be significantly increased by σ term increasing. According to ref [15], it is expected that grain-boundary sliding is the primary deformation mechanism of Nd:YAG specimens under post-HIP conditions used for this study (i.e. 1650°C and 150 MPa). The strain rate during post-HIP should be controlled by the diffusion necessary to accommodate the motion of the grains, i.e. the diffusion of rare-earths (Y3+ and Nd3+) which are the slowest species in Nd:YAG material [7,16].3.4. Optical properties
Figures 5(a)-5(d) shows pictures of polished Nd:YAG samples sintered by SSRS or by SSRS + post-HIP. Pre-sintered specimen (Fig. 5(a)) is opaque due to large amount of residual porosity (i.e. 5%). With further SSRS at a temperature of 1750°C, Nd:YAG specimens become transparent even with limited silica content (Fig. 5(b)). Transparency is however significantly higher with higher silica content (0.14wt.%) as observed in Fig. 5b. Post-HIP treatment also lead to transparent samples as shown in Fig. 5(d),5(e). With post-HIP, the better transparency seems to be obtained for the lower silica content (0.07wt.%). This result is in accordance with Fig. 2 showing residual porosity for other samples as pores are known to be very efficient light scattering centers [14].
Fig. 5 Nd:YAG ceramic pre-sintered by SSRS at 1550°C for 30 min (a); Nd:YAG ceramics sintered by SSRS at 1750°C for 3 h with 0.14wt.% (b) or 0.07wt.% SiO2 (c); Nd:YAG ceramics sintered by post-HIP at 1650°C for 6 h with 0.14wt.% (d) or 0.07wt.% SiO2 (e).
After polishing of optical quality, Nd:YAG samples were characterized in terms of optical transmittance. Results were reported in Fig. 6 and show that post-HIP treatment at lower temperature (1650°C for 6 h) allow obtaining Nd:YAG ceramics with significantly higher transmittance than samples obtained by classical SSRS under vacuum at 1750°C for 3 h. For both samples, only Nd3+ absorption peaks were detected in this wavelength range. The increasing of absorption observed for lower wavelengths can be explained by residual intragranular pores as reported in Fig. 2(c) for Nd:YAG specimen obtained by SSRS under vacuum. On the contrary, the same phenomenon observed for Nd:YAG specimen obtained by post-HIP still needs further investigations to be clearly understood. In fact, SEM observations reported in Fig. 2(i) have not revealed defects like pores or secondary phases.
Fig. 6 Optical transmittance in the visible range of Nd:YAG transparent ceramics elaborated by SSRS under vacuum at 1750°C for 3 h or post-HIP at 1650°C for 6 h.
4. Summary
This study has compared two different sintering routes to elaborate Nd:YAG transparent ceramics for laser applications: (1) conventional solid-state reaction sintering under vacuum (SSRS) and (2) non-conventional sintering by post-HIP treatment under an isostatic pressure (150 MPa) of argon gas.
During conventional sintering, pores and grain boundaries separate to each other at the final stage the vacuum sintering (ρ > 99.9%). This phenomenon is accompanied by the formation of intragranular porosity which cannot be eliminated by longer SSRS treatment. This work demonstrates that fine-grained (G < 2µm) transparent Nd:YAG can be elaborated with lower SiO2 levels by vacuum sintering with an additional post-HIP treatment. Post-HIP allows to reduce silica content thus drastically limiting grain growth in Nd:YAG ceramics. Under these conditions, the separation between pores and grain boundaries is no longer observed and higher optical quality ceramics can be elaborated.
From this study, post-HIP treatment appears to be a very promising way to elaborate defects-free, high optical quality Nd:YAG transparent ceramics. Future work will be performed on the optimization of post-HIP conditions (critical density before HIP treatment, HIPing temperature) to improve the optical quality of Nd:YAG ceramics. The laser efficiency of samples obtained by post-HIP will also be compared to that obtained for samples elaborated by conventional pressureless vacuum sintering.
References and links
1. A. Ikesue, I. Furusato, and K. Kamata, “Fabrication of polycrystalline, transparent YAG ceramics by a solid-state reaction method,” J. Am. Ceram. Soc. 78(1), 225–228 (1995). [CrossRef]
2. W. Liu, J. Li, B. Jiang, D. Zhang, and Y. Pan, “2.44 KW laser output of Nd:YAG ceramic slab fabricated by a solid-state reactive sintering,” J. Alloy. Comp. 538, 258–261 (2012). [CrossRef]
3. A. Ikesue, K. Kamata, and K. Yoshida, “Effects of neodymium concentration on optical characteristics of polycrystalline Nd:YAG laser materials,” J. Am. Ceram. Soc. 79(7), 1921–1926 (1996). [CrossRef]
4. A. Ikesue, Y. L. Aung, T. Taira, T. Kamimura, K. Yoshida, and G. L. Messing, “Progress in ceramic lasers,” Annu. Rev. Mater. Res. 36(1), 397–429 (2006). [CrossRef]
5. R. Boulesteix, A. Maître, J.-F. Baumard, Y. Rabinovitch, and F. Reynaud, “Light scattering by pores in transparent Nd:YAG ceramics for lasers: Correlations between microstructure and optical properties,” Opt. Express 18(14), 14992–15002 (2010). [CrossRef] [PubMed]
6. A. Ikesue and K. Yoshida, “Scattering in polycrystalline Nd:YAG lasers,” J. Am. Ceram. Soc. 81(8), 2194–2196 (1998). [CrossRef]
7. R. Boulesteix, A. Maître, J.-F. Baumard, Y. Rabinovitch, C. Sallé, S. Weber, and M. Kilo, “The effect of silica doping on neodymium diffusion in yttrium aluminum garnet ceramics: implications for sintering mechanisms,” J. Eur. Ceram. Soc. 29(12), 2517–2526 (2009). [CrossRef]
8. A. J. Stevenson, X. Li, M. A. Martinez, J. M. Anderson, D. L. Suchy, E. R. Kupp, E. C. Dickey, K. T. Mueller, and G. L. Messing, “Effect of SiO2 on densification and microstructure development in Nd:YAG transparent ceramics,” J. Am. Ceram. Soc. 94(5), 1380–1387 (2011). [CrossRef]
9. K. Tsukuma, “Transparent MgAl2O4 spinel ceramics produced by HIP post-sintering,” J. Ceram. Soc. Jpn. 114(1334), 802–806 (2006). [CrossRef]
10. A. Ikesue and K. Kamata, “Fabrication of transparent Ce:Y2O3 ceramics using a HIP,” J. Ceram. Soc. Jpn. 103(1203), 1155–1159 (1995). [CrossRef]
11. A. Krell, P. Blank, H. Ma, T. Hutzler, and M. Nebelung, “Processing of high-density submicrometer Al2O3 for new applications,” J. Am. Ceram. Soc. 86(4), 546–553 (2003). [CrossRef]
12. A. Ikesue and K. Kamata, “Microstructure and optical properties of hot isostatically pressed Nd:YAG ceramics,” J. Am. Ceram. Soc. 79(7), 1927–1933 (1996). [CrossRef]
13. S.-H. Lee, E. R. Kupp, A. J. Stevenson, J. M. Anderson, G. L. Messing, X. Li, E. C. Dickey, J. Q. Dumm, V. K. Simonaitis-Castillo, and G. J. Quarles, “Hot isostatic pressing of transparent Nd:YAG ceramics,” J. Am. Ceram. Soc. 92(7), 1456–1463 (2009). [CrossRef]
14. R. Boulesteix, A. Maître, L. Chrétien, Y. Rabinovitch, and C. Sallé, “Microstructural evolution during vacuum sintering of yttrium aluminum garnet transparent ceramics: Toward the origin of residual porosity affecting the transparency,” J. Am. Ceram. Soc. 96(6), 1724–1731 (2013). [CrossRef]
15. M. Jiménez-Melendo, H. Haneda, and H. Nozawa, “Ytterbium cation diffusion in yttrium aluminum garnet (YAG) - Implications for creep mechanisms,” J. Am. Ceram. Soc. 84(10), 2356–2360 (2001). [CrossRef]
16. H. Haneda, “Role of diffusion phenomena in the processing of ceramics,” J. Ceram. Soc. Jpn. 111(1295), 439–447 (2003). [CrossRef]
