Processing of Ceramics. Группа авторов
is that “exploration on the truth of natural science and prediction of the future is the mission of a scientist” (at that time I was a refractory engineer and not holding a Ph.D.), and I could not simply accept their opinion. So, I decided to work on “potential of ceramics for future exploration in the optical field.” The idea “laser oscillation by ceramics” started in the summer of 1991 and confirmed the success of production in December of that year, but since the author belonged to the private company, it could be published in 1995.
In 1995, the author demonstrated highly efficient laser oscillation by using polycrystalline ceramic materials with performance that could match or surpass high‐quality single crystal [2], but materials and laser experts at the time were highly skeptical about our report. One of the reasons is that I was not an expert in ceramics and lasers, and the invention was by a person from a different field (i.e., a refractory‐related engineer working for steel smelting). Generally, a large number of scattering sources (such as residual pores and heterogeneous phases) are present in the common ceramics, causing significant Mie scattering. Especially from the technical point of view, no one proposed to remove those residual pores completely in ceramics. The density of the transparent ceramics is much higher than the opaque ceramics used for other applications, and it shows good transparency, but even in this case, numbers of pores of more than 1000 ppm are remained inside the ceramic material. Another hurdle to overcome is even if we were able to remove those residual pores completely, nobody had answers to “the problem of Rayleigh scattering from grain boundary phase and grain boundary,” which is the biggest technical problem in terms of technology. There were many reports from ceramists at that time describing that a lot of residual pores (scattering sources) are present in the ceramics material, and also, structure defects certainly generated when the granulated raw materials are pressed in molding process, and these defects are definitely remained even in the final sintered body. They have reported those descriptions with photographs of microstructure of ceramics as evidence. However, the reports only described the observation method for defects in ceramics and observed results, and there were no discussion on the cause of defect formation and how to solve this problem. The authors thought that structure defects in materials were generated mainly by artificial factors, and we should clarify the problem sources and modify the microstructures; finally, we will be able to eliminate those defects including residual pores. Once Mie scattering can be removed from the inside of the material, the remaining problem will be only Rayleigh scattering. When I attended a conference in ceramic society, I saw some researchers reported that “This ceramic has a clean grain boundary,” but the transparency of their ceramics was not as good as single crystals although they showed a clean grain boundary. That was a basic contradiction point noticed to me, and I simply interpreted their “clean grain boundary” that during observation with an electron microscope they could observe only the clean area where secondary phase does not exist. Actually, we should pursue what is the fundamental nature of Rayleigh scattering. Also, we should further pursue that even if we can form a really clean grain boundary, whether or not it will be possible to obtain ceramic material with ultra‐low scattering or without scattering. To acquire such evidences are the essence of good research work, and hence, there can be progress in science and technology and there are roots that can create the next innovation. At that time, I was an amateur (refractory engineer) who knew nothing about lasers and ceramics, but when I reconsidered now probably, I had a challenging spirit pursuing the essence of material science because of lack of knowledge about laser and ceramics. In this chapter, we will describe how the ceramic material, which is a guideline for the development of various optical ceramics, changed from the translucency level [3] to the optical grade material, based on the development of ceramic laser materials.
1.2 Technical Problem of Conventional Single Crystal
Since single crystal materials are widely applied in various industrial fields and it is difficult to describe the technical problems of these single crystals together at one time, firstly we will focus on laser crystals that require the highest quality. In 1960, laser oscillation was firstly demonstrated by Cr doped Sapphire by Maiman, and then, laser oscillation at room temperature using Nd‐doped YAG (Y3Al5O12) single crystal by Guesic in 1964 was a trigger for the birth of solid‐state laser. After laser oscillation by YAG single crystal, unique laser performance of various types of laser gain media, such as YVO4, Cr:Forsterite (Mg2SiO4) KGW (KGd(WO4)2), Ti:Sapphire (Al2O3), Cr:ZnSe, and Cr:ZnS, has been reported. However, new laser gain materials, which can exceed the YAG, have not been found in total performance including the quality of materials, and even now, the mainstream of solid‐state lasers is YAG and it is still unchanged. YAG materials are applied in most of solid‐state laser as gain medium, and almost all of them are single crystals grown by the Czochralski (CZ) method. A transition metal element such as Cr, Ti, or a lanthanide element such as Nd, Yb, Er, Tm, and Ho is added to the YAG single crystal as a laser (active) element. Among these active elements, Nd is belonging to the four‐level system and it is easy to create a population inversion (a state in which number of electrons in the upper state is higher than the lower state) at the f‐f electron transition of Nd by external excitation. The laser oscillation is relatively simple in this system, and it has a fluorescent line with a narrow spectral line width and high quantum efficiency; hence, it is considered to be the most important laser active element in YAG host crystal. In recent years, however, LD (laser diode) excitation system became typical in these days and strong excitation became possible as well; therefore, Yb with three‐level system with higher quantum efficiency of 91% has also been used as laser gain material. The CZ method shown in Figure 1.2 is generally used to grow the YAG single crystal. The starting materials are Y2O3, Al2O3, and Nd2O3 powders with high purity above 4 N (99.99 mass%) grade, and these raw materials are weighed in YAG composition (not strict stoichiometric composition) and mixed. Then the mixture is molded and calcined. A relatively dense sintered body obtained by sintering is used as a raw material for melting. It is filled in an iridium (Ir) crucible. The Ir crucible is heated by high frequency induction and melted at temperature higher than the melting point of YAG (1950 °C).
Figure 1.2 Schematic diagram of YAG single crystal grown by CZ method.
Generally, YAG seed crystals with <110> orientation which has the smallest surface‐free energy is used but <110> or <100> oriented seed crystals are also applied in some cases. This seed crystal is immersed in the molten YAG, and then, crystal growth is continuously carried out at a rotation speed of 10 ~ 30 rpm and a pulling rate of about 0.2 mm/hour. In the case of Nd‐doped YAG crystal, Nd ions substitute Y sites in YAG lattice. The ionic radius of Nd ion is too large compared with that of the Y ion, so it is well known that it is not easy to dissolve the Nd ions in the YAG host crystal as a solid solution. The segregation coefficient of Nd ion to YAG host crystal is very small (that is, the concentration ratio of Nd in crystal to Nd in the melt is very small). According to the literature, since the segregation coefficient of Nd to YAG crystal is about 0.2 [4], normally the concentration of Nd ions in the melt is set to be several times higher than the target Nd concentration (that is, the concentration of Nd doping in the YAG sinter body which is used as starting material for melting is prepared with higher than the target composition). Even if the concentration of Nd in the melt is prepared with very high, the concentration of Nd in the YAG crystal may not be automatically increased homogeneously and simply in proportion to its concentration. When the concentration of Nd in the YAG crystal is increased to higher than 1 at.%, many precipitates (scatterers) are generated in the crystal and it is difficult to utilize it as a laser gain medium. Even in the case of the commonly used 1 at.% Nd:YAG, (211) facets tend to be formed from the pulling axis of the crystal (ingot) toward the outer periphery [4], and thus, only the outer periphery of the columnar crystals can be used as a laser gain medium. Also, when pulling