Special Issue on Creation of Materials by Superthermal Field
Direct Laser Sintering of Bulk Alumina Using 1070 nm Fiber Laser
2023 Volume 64 Issue 6 Pages 1183-1187
Details
2023 Volume 64 Issue 6 Pages 1183-1187
Pellets made of α-alumina were sintered by fiber laser irradiation (1070 nm) for 1 min. The optical measurements showed that the laser selectively applied heat around the voids and gaps between grains in the green pellets. Even though the laser can penetrate deep inside the pellets, the heating efficiency is low because alumina poorly absorbs the laser energy. The graphite layer effectively improved the heating efficiency on the irradiated surface at the low-temperature region to “ignite” laser sintering. The microstructure of the sintered pellets was controlled by the microstructure of the green bodies. Porous pellets were obtained from coarse powders and transparent dense pellets from mixtures of coarse and fine powders. The porous pellets showed excellent bending strength owing to the joining of grains by the selective heating of gaps between them. The transparent pellets comprised large alumina single crystals anisotropically formed by spontaneous melt growth under laser irradiation.
Fig. 6 Microstructures of transparent alumina body of AA-03_18 pellet. (a) optical image (b) EBSD image of area inside grain (c) EBSD image of area around grain boundaries and (d) cross-sectional optical image of pellet after 30 s of irradiation.
Almost all ceramic parts are manufactured by firing. The process usually employs an electric or gas furnace, with the firing process taking a long time, typically 20 to 40 h. Large furnaces are suitable for mass production because hundreds or thousands of parts can be fired at a time. However, the demand for small-lot (or small quantity) ceramic production of a wide variety of shapes has been increasing in recent years. For example, biological ceramic components, such as artificial teeth and joints, need to be shaped to suit each patient, and aircraft ceramic components have various shapes depending on the use. However, even for such small-quantity production, the firing time cannot be shortened if a conventional furnace is used. There are various approaches to developing a process suitable for small quantity production. One of them involves local heating rather than the overall heating employed in furnaces. A laser is a potential source of local heating. In the machining industry, high-intensity lasers are widely used to cut or weld steel plates because of their ability to localize heat rapidly and precisely.1,2)
Since the development of high-intensity lasers in the 1990s, laser sintering of ceramics has attracted attention and become a “dream technology” because it does not require traditional furnace heating. Early studies3) explored laser sintering techniques for materials, combining ceramic powders with a binder phase having a relatively low melting point, such as SiC with polyamide resin,4) ZrO2 with polymethylmethacrylate-butyl methacrylate (PMMA-BPA) copolymer,5) and silica sand with phenol resin.6) After the 2000s, laser sintering was in the spotlight again with the development and commercialization of additive manufacturing (or 3D printing) of plastics and metals, and ambitious efforts to improve direct laser sintering have been made since then. Kruth et al. reported the CO2 laser sintering of YSZ with 0.25% alumina as a sintering aid.7) They painted a 0.1 mm thick slurry layer on substrates, sintered the layer by CO2 laser irradiation, and repeated the painting–sintering process 25 times. The surface of the resulting sintered YSZ body became smooth at a high laser scanning speed, but the sintering density decreased. Cruz et al. reported the CO2 laser sintering of hydroxyapatite (HAp) for applications in bone implants.8) They used a slurry with 66% solid content with optimized viscosity as a source. 100 µm thick HAp layers containing large pores of 70–100 µm were sintered, and HAp was partially decomposed into tricalcium phosphate (TCP) and amorphous calcium phosphate.
As mentioned above, laser sintering research has been focused on the sintering of thin layers (approximately several 10 s–100 µm) for application in 3D printing;9–18) however, little consideration has been given to sintering bulk ceramics although the direct laser sintering is a key technology for other additive manufacturing processes such as welding and overfill welding. In this study, we have attempted laser sintering of bulk alumina using a 1070 nm fiber laser and described the optical properties of the green bodies and microstructure and properties of the sintered bodies.
As raw powders for laser sintering, commercial high-purity (≧99.99) α-alumina powders were used with different average grain sizes: “AKP-3000” with the average grain size of 0.67 µm, “AA-03_18”, a mixture of AA-18 and AA-03 (8:2 in weight) with the average grain size of 20.3 and 0.40 µm, respectively. These alumina powders were obtained from Sumitomo Chemical, Japan. 0.2 g of raw powder was uniaxially pressed into pellets (10 mm in diameter, ∼1 mm thick) at 10 MPa. The optical properties of a sapphire wafer ((11–20) surface; Shinkosha, Japan; 1 mm thick) and sintered alumina plates (purity: >99.99%; relative density: >98%; SSA-9999W, Nikkato, Japan; 1 mm thick) were compared. A graphite layer was sprayed onto the pellets using a spray-can (graphite spray, DGF, Japan); thus, the layer thickness could not be precisely controlled and was 5–10 µm.
A continuous wave (CW) fiber laser (maximum output: 400 W; red power-SP-400C, SPI Lasers UK Ltd., UK) operating at 1070 nm was used as a heating source. The laser beam was expanded using an optical lens, and the entire surface of the pellet was irradiated. A schematic of the laser sintering setup used in this study is shown in Fig. 1. The green pellet was placed on a stainless-steel sample stage, and the laser was positioned above. The irradiated laser light had a Gaussian intensity distribution, and the center of the laser spot was adjusted to the sample center. The sintering occurred in air, and the irradiation time was fixed at 60 s. The sample temperature was measured during laser irradiation using a two-color pyrometer (Thermera-Phantom, Nobby Tech., Ltd., Japan) and analyzed using software (Thermera HSS, Mitsui Photonics, Ltd., Japan).
Schematic of sample setting.
The microstructure and crystalline orientations were investigated using a scanning electron microscope (SU-8000, Hitachi, Japan) equipped with an electron backscatter diffraction (EBSD) probe. The optical properties were evaluated using a UV-VIS-IR spectrometer (Lambda950, Perkin Elmer, UK). Further, 3-point bending strength and Vickers hardness were measured using the 5582 universal testing system (Instron, USA) and the Fischerscope HM2000 (Helmut Fischer GmbH, Germany), respectively. The Archimedes and mercury injection methods were used to measure the density and pore distribution of the pellets.
Laser heating is caused by the absorption of laser light by matter. The optical properties were measured using a UV-VIS spectrometer with an integrating sphere in two measurement modes to estimate the laser absorption of alumina bodies with different microstructures: “transmission measurement mode” for the forward-scattered light intensity and “reflection measurement mode” for the backward-scattered light intensity. The summation of these intensities gives the overall outgoing light intensity; the difference between the incoming (or incident) and the overall outgoing (or emitted) light gives the absorbed light intensity. Three types of samples were used for the measurements: (a) sapphire wafer, (b) commercial sintered alumina, and (c) green alumina pellet (AKP-3000). No surface polishing was performed before the measurements.
The experimental UV-VIS spectra and calculated absorption spectra of the samples are shown in Fig. 2 and 3, respectively. The forward and backward-scattered light intensities ratios to the incident light intensity (I/I0) at a laser wavelength of 1070 nm were 77.82% and 17.67% for the sapphire wafer, 4.19% and 70.16% for the sintered body, and 3.90% and 60.26% for the green body, respectively. The calculated absorbed light intensities were 4.51%, 25.65%, and 35.84% for sapphire wafer, sintered, and green bodies, respectively. The difference between the sapphire wafer and the sintered body is in the presence of the grain boundaries; the grain boundaries in the sintered body absorbed approximately 20% of the incident light. Absorption by the green bodies was remarkable. Although the relative density of the green body was 60%, the absorption was 15 times higher than that of the (a) sapphire wafer. These results indicate that light is absorbed not only by alumina grains but also by microstructures such as grain boundaries and voids. Thus, when the laser irradiates the green body, heat is generated predominantly at/around these microstructures.
Experimental optical spectra of (a) sapphire wafer, (b) sintered body and (c) green body.
Calculated absorption spectra of (a) sapphire wafer, (b) sintered body and (c) green body.
In Fig. 3, the calculated absorption increases in the following order: the sapphire wafer, sintered body, and green body. The absorption of the sintered body and the green body in the 1100–2500 nm region simply increased compared to that of the single crystal; the alumina-specific absorption peaks (∼1400 nm, 1950 nm, and 2200 nm) were at approximately the same wavelengths. The increase in the intensity of absorption peaks in this region could have been caused by an increase in the optical path length resulting from reflection at the grain boundary and the grain surface inside the sample. In the 500–1100 nm region, the intensity of the absorption bands of the green body is quite high compared to that of the other samples. The high absorption in this region could have been caused by the microstructure of the green body, such as voids and gaps between grains. However, analysis of the microstructures of the green body is challenging because of its mechanical weakness; therefore, a quantitative discussion on the origin of the high absorption is currently impossible.
3.2 Enhancement of laser heating using a graphite surface layerLaser sintering of alumina pellets was preliminarily examined by direct laser irradiation, but the sample temperature was not sufficient for sintering even at the highest laser power output of our equipment (400 W) because of the poor laser absorption by alumina. However, it could be enhanced through the changes in the equilibrium electron occupation following the Fermi equation.19) Thus, we tried “preheating” the pellet surface using a graphite coating. Graphite is known for its excellent light-absorption properties over a wide range of wavelengths. The thin graphite layer heats strongly and quickly and disappears under laser irradiation because it burns in air and bursts in an inert atmosphere owing to the stress caused by the thermal expansion difference between alumina and graphite. The surface of the alumina pellet would be heated by heat transfer from the laser-heated graphite coating, and after the graphite layer disappeared, the laser would be directly incident to the “preheated” pellet surface.
Figure 4 shows the temperature profiles of the alumina pellet with the graphite surface layer measured using a pyrometer at laser powers of 217 and 318 W/cm2. In both cases, the temperature increased rapidly to 1500°C owing to the strong laser absorption of the graphite layer and continued slowly increasing up to ∼1750°C, where graphite oxidized (burned). After the graphite layer removal, the temperature change depended on the laser power. At a laser power of 318 W/cm2, the temperature increased up to 2300°C due to laser absorption of preheated alumina and then became constant. In contrast, at 217 W/cm2, the temperature dropped rapidly after the graphite layer was removed, although the temperature reached the same level. These results indicate that even after improving the absorption of alumina by preheating, specific laser power is required for sintering because absorption (or heat generation) is determined by the absorption coefficient multiplied by the laser intensity. It should be noted that the 2500°C measured by the pyrometer does not represent the temperature of the entire pellet but its maximum value in the region where the laser is strongly absorbed, namely, the grain boundary and the grain surface inside the sample.
Temperature profile of alumina green pellet under laser irradiation at laser power of 217 and 318 W/cm2.
As discussed in Section 3.1, the laser heating behavior was strongly affected by the microstructure of the green pellets. Laser sintering experiments were carried out using two types of pellets with different microstructures.
(1) AKP-3000 pelletsFigure 5 shows the cross-sectional SEM image of the AKP-3000 pellet with a graphite surface layer after irradiation at 383 W/cm2 for 60 s. The laser-sintered pellet had a porous microstructure, and the grains were connected to each other to form a labyrinth microstructure comprising elongated particles. The width of the elongated particles was almost the same as the grain size of the initial powder (∼0.7 µm). This microstructure suggests that the grains fused significantly; however, further grain growth rarely occurred under laser irradiation.
Cross-sectional SEM image of porous alumina body of AKP-3000 pellet.
The porosity and the average pore size of the laser-sintered alumina were 41.1% and 0.17 µm, respectively. It is important to note that its bending strength was 192 MPa, which is 30% higher than that of the porous alumina (149.02 MPa) made from a raw alumina pellet containing 5 mass% of additives by a furnace heating at 1350°C.20)
Such a high bending strength of the laser-sintered porous alumina results from selective heating under laser irradiation, as discussed in Section 3.1. With conventional furnace heating, the mechanical properties also improved; however, porosity decreased because of surface and/or volumetric diffusion with increasing the sintering temperature. With laser heating, the alumina grain was heated to a relatively low temperature, and the grain surface was selectively heated around voids (or gaps between particles); thus, the grains were strongly combined as though sintered at high temperature.
Details on this high-strength porous alumina will be reported in future, including the effect of laser power on the mechanical strength and microstructure.
(2) AA-03_18 pelletsFigure 6(a) shows an optical image of the AA-03_18 pellet with a graphite surface layer after laser irradiation at 508 W/cm2 for 60 s. The pellet was transparent and comprised large transparent grains of 1 mm in size. EBSD analysis of the cross-section (Fig. 6(b)) indicates that each grain was single crystal alumina with no cracks or dislocations. The “lines” observed between the large single crystals were not cracks but grain boundaries. As shown in Fig. 6(c), no amorphous phase was formed between the grains, and the grains were directly fused. Vickers hardness tests were performed at the grain (the indenter edge was placed on the grain). The average results (N = 5) were 1890 HV, which is the comparable to that of single crystal alumina. Grain growth was anisotropic, and crystals grew preferentially in a direction parallel to that of the incident laser.
Microstructures of transparent alumina body of AA-03_18 pellet. (a) optical image (b) EBSD image of area inside grain (c) EBSD image of area around grain boundaries and (d) cross-sectional optical image of pellet after 30 s of irradiation.
The AA-03_18 powder was a mixture of larger particles (AA-18) surrounded by small particles (AA-03). The small particle layer strongly absorbed the laser because the numerical density of its voids was high. When a laser irradiated the green pellet at a certain intensity, the small particle layer near the pellet surface heated and melted. The temperature of the large particle was relatively low; thus, alumina melt crystallized on the surface of the large grains leading to their growth. As the melt growth proceeded, voids disappeared near the pellet surface, and the laser could penetrate deeper through the grown large crystal near the pellet surface, which is a poor absorber of the laser light. Thus, the alumina melt zone moved from the pellet surface to the inner part and finally to the pellet backside. The upper half of the pellet became transparent when the laser irradiation was stopped at half time (30 s) (Fig. 6(d)). This mechanism is similar to the single-crystal growth in the floating zone (FZ) method, and the formation of transparent alumina is a spontaneous melt growth in the pellet under laser irradiation.
Figure 7 shows the transmission spectra of the laser-sintered transparent alumina and sapphire wafers. Their shapes are similar, but the transmittance of the laser-sintered sample was approximately 80% of that of the single crystal. As shown in Fig. 6(a), the structure was disordered at the periphery of the pellet owing to the Gaussian distribution of the laser intensity, which could reduce transmittance. Overall, the structure that easily absorbed the laser was substituted by one formed by melt growth under laser irradiation, and it efficiently transmitting the laser.
Transmittance spectra of (a) sapphire wafer and (b) transparent alumina obtained from AA-03_18 pellet.
Laser sintering is a short process. If the sample is small (10 mm φ × 1 mm in this study), 1 min laser irradiation is enough to complete the sintering. Laser scanning is necessary for a larger sample, which is easy to implement since this technology has already been developed (but scanning parameters, such as scanning speed and scanning pattern, should be optimized). The laser heating behavior and resulting microstructure strongly depend on the microstructure of the green pellets, which can be controlled by the source powders. In this study, we demonstrated porous and transparent alumina pellets prepared by laser sintering using a graphite surface layer as a preheating medium. Here, we qualitatively assessed the mechanisms of laser heating and sintering. However, the optical properties of the green bodies should be examined for a quantitative discussion on the mechanisms, especially optical properties at a high temperature and laser–material interactions in the inhomogeneous microstructure of green bodies.
This work was partly supported by JSPS KAKENHI (Grant Number JP21H05199), by New Energy and Industrial Technology Development Organization (NEDO) (Project Number JPNP22005), by Council for Science, Technology, and Innovation (CSTI), Cross-ministerial Strategic Innovation Promotion Program (SIP), “Innovative Design/Manufacturing Technologies” (Funding agency: NEDO), by Japan Fine Ceramics Center, Advanced Research & Technology Incubation Program, and by Japan Fine Ceramics Center - Sumitomo Chemical Co., Ltd. Collaboration Research Program.
