Radically different effects on materials by separated microwave electric and magnetic fields

Transcription

1 Mat Res Innovat (2002) 5: Springer-Verlag 2002 ORIGINAL ARTICLE Jiping Cheng Rustum Roy Dinesh Agrawal Radically different effects on materials by separated microwave electric and magnetic fields Received: 15 June 2001 / Accepted: 15 June 2001 Abstract Using a 2.45 GHz wave-guided cavity, in a single mode TE 103 excitation, we were able to physically locate compacted 5 mm pellets of samples separately at the H (magnetic) node (where the E field is nearly zero), or the E (electric) node (where H field is nearly zero). A preliminary survey of a variety of metals, (Cu, Fe, Co..) ceramics (ZnO, etc.), and composites, (WC-Co, ZnO-Co) showed remarkable differences in their heating behaviors. The results establish conclusively that the magnetic field interaction contributes greatly to microwave heating of common materials in a manner, previously neglected in most theories of microwave heating, albeit still to be understood. Keywords Microwaves E-field H-field ZnO WC Introduction and previous work The vast majority of papers dealing with microwave heating of solids ascribe the heating to energy loss mechanisms of the electric vector. Very recently our experimental findings have demonstrated that magnetic losses play an important role in microwave sintering of bulk materials for a wide range of conductor and semiconductor materials. In 1999 we established [1] experimentally (for the first time) that contrary to all previous practices and experiments that ordinary powdered metal samples of virtually any composition including very complex shaped (see Figure 1) and large size (100 mm diameter, 1 kilograms) could be fully sintered in 30 minutes or less in a 2.45 GHz multi-mode microwave cavity. Moreover, these samples had properties at least as good as, and usually better than, those sintered in conventional furnaces. This finding was outside the experience of a very large number of scientists whose extensive work claimed that J. Cheng ( ) R. Roy D. Agrawal Materials Research Laboratory, The Pennsylvania State University, University Park, PA 16802, USA jxc44@psu.edu Fig. 1a, b Commercial Powdered-metal gear parts made by microwave sintering in 30 minutes

2 powdered metals could not be sintered by microwave energy. These have been covered in many reviews [2 4]. Hence our achievement was puzzling to us and mainly unbelievable to colleagues (and NATURE s reviewers). Simple minded efforts to explain this by skin depth absorption etc did not work. The well known extensive theoretical treatment of microwave-material interaction by many workers (see e.g. Varadan and Varadan [5], Booske et al [6] and many others) have one feature in common: that they always treat the energy absorption mechanism as due to the dielectric loss factor. In 1994 Cherradi et al [7] published a paper in which they showed that the magnetic field must make substantial contributions to the heating of alumina (at high temperature) and semi-insulators, and metallic copper. But the possible significance of their work was missed by the vast majority of workers in the microwave sintering community. In that work, their experimental design of using samples of 120 mm length, was such that the sample was always heated in both H and E fields simultaneously, caused by a complicated interplay of the different absorption and conduction mechanisms. Moreover, their work did not involve microwave sintering of materials at all nor measuring the properties thereof. Our research in this study was based on our own unique success over the last several years with microwave sintering to full density in very short timesof all major ceramics, including oxides, nitrides, and carbides [8 9]. Various fully transparent and translucent ceramic samples, such as hydroxyapatite [10], mullite [11], alumina, spinel [12], aluminum nitride [13], and aluminum oxynitride [14], have also been successfully made by microwave sintering in minutes (Figure 2). Experimental In this work we changed from multimode to single mode cavities. A finely tuned (2.45 GHz) cavity with a cross section dimension of 86 mm by 43 mm which works in TE 103 single mode (Figure 3) was used. The distribution of the microwave field within the cavity is sketched in Figure 4. In the L/2 location along the length of the cavity, the maximum electric field is in the center of the cross section, where the magnetic field is at a minimum. The maximum magnetic field is near the wall, where the electric field is minimum. A quartz tube was introduced in this location to hold the sample and also to enable us to control the atmosphere around the sample. A 2.45 GHz, 1.2 kw microwave generator (Toshiba, Japan) with power monitor was used as microwave source. Small cylindrical samples (5 mm diameter and 5 mm high) was placed inside the quartz tube at the two different locations: the maximum electric field area (where the magnetic field is minimum), and the maximum magnetic field area (where the electric field is minimum), respectively. Sample temperatures were measured using an infrared pyrometer (Mikron Instrument Co., Model M90-BT, Temperature range 50 C 1000 C). During the experiments, atmospheric pressure nitrogen gas was passed through the quartz tube to avoid oxidation of metal samples at high temperature. Initially, we tried to use a fixed microwave power for all samples during heating, but the temperature increase was too fast and the highest temperature exceeded the measuring range of the pyrometer for some samples. And in some cases, discharging and arcing occurred at higher microwave power. So we set different microwave powers for different samples to get more stable heating results. Results and discussion 171 Fig. 2 Transparent ceramics samples made by microwave sintering processing Fig 3 The microwave setup used for the separated microwave heating Figure 5 shows the heating observed for a typical commercial powdered metal sample (Keystone Powderedmetal Company, Saint Marys, PA, USA). The composition of the powder is Fe+2%Cu+0.8%C. In the electric field, this kind of sample can hardly be heated up at all. At a power of 500 watts a maximum temperature of 180 C (after microwave heating for 8 minutes) was observed, and some arcing occurred around the edge of the sample. But in the magnetic field, under the same microwave power, the sample heated up very quickly and stably. The heating rate was higher than 300 C per minute in the first two minutes, then it slowed down. The final temperature reading was 780 C in 10 minutes, and a quite uniform heating result was observed. Since there is no insulation material around the sample, the thermal loss must be significant at high temperatures, and we think that is the main reason resulting in the lower heating rates at high temperature ranges, as compared to our usual work.

3 172 Fig. 4 The schematic of the microwave field distribution within the TE 103 single-mode microwave cavity Fig. 5 Comparison of the heating rate of powdered metal compact Some other compositions of pure metal powder-compact samples were also tested in this study, cobalt (Co), iron (Fe) and copper (Cu) (all came from Alfa Aesar, A Johnson Matthey Company, USA). The Co and Fe powder-compact samples exhibited the same behavior during Fig. 6 Comparison of the heating rate of cobalt (Co) powdercompact the microwave heating (Figures 6, 7). There were little heating in the electric field but high heating rates were observed in the magnetic field. The microwave heating of the Cu powder-compact sample was quite anomalous. The sample heated up very fast both in the electric and

4 173 Fig. 7 Comparison of the heating rate of iron (Fe) powdercompact Fig. 10 Comparison of the heating rate of alumina (Al 2 O 3 ) powder-compact Fig. 8 The heating rate of copper powder-compact sample in Fig. 11 Comparison of the heating rate of Zinc oxide (ZnO) powder-compact Fig. 9 Microwave heating of solid copper metal sample in magnetic fields. As shown in Figure 8, the sample s temperature rose to ~700 C in 1 2 minutes, then quickly dropped down to ~500 C and kept within that range during the continuous heating. For comparison, a pure solid Cu bar with the same shape and size was put in the microwave cavity to check out the energy absorption and heating behavior. It was found that there is no temperature rise for the solid Cu bar sample either in the electric field or in the magnetic field, even after being exposed in the microwave field for 10 minutes, the sample still remained at room temperature (Figure 9). Alumina (Al 2 O 3 ) is a proto-typical ceramic material with excellent dielectric properties. This material usually has very low dielectric loss, and it is not easy to heat up by microwave, especially at lower temperature. Since the dielectric loss of Al 2 O 3 increases with temperature, microwave heating of Al 2 O 3 becomes more efficient at high temperature. In our experiments, the microwave heating

5 174 Fig. 12 The heating rate of tungsten carbide (WC) powdercompact Fig. 14 The heating rate of Fe 2 O 3 powder-compact sample in Fig. 13 The heating rate of FeO powder-compact sample in Fig. 15 The heating rate of Fe 3 O 4 powder-compact sample in of high purity Al 2 O 3 samples (Baikowski International, Charlotte, NC) in the electric field went slowly in the beginning, the heating rate speeded up after the sample reached a temperature of ~500 C. But in the magnetic field, Al 2 O 3 cannot be heated at all under the same microwave power for same exposing time (Figure 10). Zinc oxide (ZnO) is another important dielectric material. It showed the same heating behavior as Al 2 O 3. In the pure magnetic field, there was almost no temperature raise. But in the pure electric field, the ZnO sample heated up very rapidly, the temperature reached 950 C in only 30 seconds (Figure 11). Tungsten carbide (WC) belongs to the larger family of semiconductor materials with some conductivity. We have studied the sintering and WC in depth and have developed commercial technologies (15) for microwave sintering of cemented WC hard metal products in the past few years. In this work, the pure WC (Telydyne) powder compact samples were microwave heated in E and H fields respectively. The results exposed that the magnetic loss factor, not the dielectric loss, is the principle source leading to microwave absorption (Figure 12). It was observed that there were some discharging or arcing occurred when the WC sample was put in pure E field, and in the pure magnetic field, the sample could be microwave heated to high temperature quite stably without any discharging phenomenon. The microwave heating behaviors of iron oxides in different fields are shown in Figures 13, 14 and 15. It is interesting that different iron oxide exhibited different heating results in the microwave fields. The FeO and Fe 2 O 3 could only be heated to high temperature in the

6 175 Fig. 16 The heating rate of tungsten carbide-cobalt (WC-Co) composite Fig. 18 The heating rate of Zinc oxide-cobalt (ZnO-Co) composite Fig. 19 The scheme of the temperature distributions within the ZnO-Co powder mixture samples in different microwave fields Fig. 17 The heating rate of alumina-powdered metal composite pure E field, and Fe 3 O 4 was heated up in both E and H fields. Figures 16, 17 and 18 show the heating data on some composite samples, alumina-powdered metal, tungsten carbide-cobalt (WC-Co) and ZnO-Co composite samples. Depending on the field, we obtained quite different results. The WC-Co sample can only be efficiently heated in the magnetic field, as the pure WC and Co samples also did. The alumina-powdered metal and ZnO-Co composite samples can be heated up in both electric and magnetic fields, since these composites contain two components, one of which is more sensitive to the E field (Al 2 O 3, ZnO), and the other more sensitive to the H field (powdered metal, Co). As shown in Figure 19, we assume that the temperature came from different contri- butions which depend on the sample located in different microwave fields. For example, when the ZnO-Co powder mixture sample was located in a pure H field, in the beginning only high magnetic loss Co powder absorbed microwave power and heated to high temperature. Meanwhile, there was no absorption occurring in the ZnO powder which remained at low temperature. The measured temperature resulted mainly from the Co absorption. As time passed, the ZnO powder got heated higher and higher due to conductive heat transfer (Figure 20). Conversely, when the sample was put in a pure E field, the temperature profile should be exactly reversed. Now the ZnO powder had higher temperatures and the heat transfer proceeded from ZnO to Co. Figure 21 and 22 show the X-ray diffraction patterns of the ZnO-Co powder mixture samples which were heated in microwave E and H fields for different time. The sample heated in a pure E field for one minute showed no change, but which heated in pure H field

7 176 Fig. 22 The X-ray diffraction patterns of the ZnO-Co powder mixture samples after microwave heating for 10 minute in E and H fields Fig. 20 The scheme of the temperature development within the ZnO-Co powder mixture samples in different microwave fields of the ZnO peaks became much lower, confirming that the ZnO had much higher reactivity with the E field (Figure 22). These results can support our assumption that the Co powder reacted much higher temperature when the ZnO-Co powder mixture sample was heated in pure H field, and ZnO powder reacted higher temperature when the sample was heated in pure E field. Fig. 21 The X-ray diffraction patterns of the ZnO-Co powder mixture samples after microwave heating for 1 minute in E and H fields showed that the metal Co phase pattern has almost disappeared (Figure 21). After ten minutes microwave heating, the samples exhibited remarkable differences. Some cobalt oxide phase appeared in the H field heated sample, but that cannot be detected at all in the E field heated sample. In the E field heated sample, the intensity Conclusions The results presented in this work have demonstrated that different materials exhibit greatly different heating behaviors in the E and H microwave fields respectively. In general, the conductive samples, such as metal powder sample and carbide sample, can be much more efficiently heated in the magnetic field. On the contrary, the pure ceramic samples (insulators with little conductivity), such as Al 2 O 3 and ZnO, showed much higher heating rates in the pure electric field. The structure state of the materials plays important roles in the microwave-materials interaction. For example, the powdered-compact copper sample absorbed lot of microwave energy in the microwave field, but the solid sample did not under the same condition. From this data, it is clear that for the general theory of energy loss in various materials when placed in a microwave field, it is no longer possible to ignore the affect of the magnetic component, especially for conductor and semiconductor materials. The contributions to the magnetic loss mechanism can be hysteresis, eddy currents, magnetic resonance, and domain wall oscillations. This set of empirical data is presented to re-open the matter of microwave-material interaction to incorporate more detailed consideration for the effects of the magnetic field.

8 177 Acknowledgements This work is supported by ARPA (-ONR) under Grant No. N References 1. R. Roy, D. Agrawal, J. Cheng, and S. Gedevanishvilli, Full sintering of powdered-metal bodies in a microwave field, Nature, 399, (1999) 2. D. Clark, and W.H. Sutton, Microwave processing of materials, Annu. Rev. Mater. Sci. 26, (1996) 3. J.D. Kaze, Microwave sintering of ceramics, Annu. Rev. Mater. Sci. 22, (1992) 4. W.H. Sutton, Microwave processing of ceramic materials, Am. Ceram. Soc. Bull. 68, (1989) 5. D.K. Ghodgaonkar, V.V. Varadan, and V.K. Vandan, Free-space measurement of complex permittivity and complex permeability of magnetic-materials at microwave-frequency, IEEE Transaction on instrumentation and measurement, 39: (2) (1990) 6. J.H. Booske, R.F. Cooper, S.A. Freeman, Microwave enhanced reaction kinetics in ceramics, Mater. Res. Innov. 1: (2) (1997) 7. A. Cherradi, G. Desgardin, J. Provost, and B. Raveau, Electric & magnetic field contributions to the microwave sintering of ceramics, Elelctroceramics IV, Vol. II, (eds. Wasner, R., Hoffmann, S., Bonnenberg, D., & Hoffmann, C.) RWTN, Aachen, (1994) 8. J. Cheng, D. Agrawal, S. Komarneni, M. Mathis, and R. Roy, Microwave processing of WC-Co composites and ferroic titanates, Mater. Res. Innov. 1, (1997) 9. R. Roy, D. Agrawal, J. Cheng, and M. Mathis, Microwave processing: triumph of applications-driven science in WCcomposites and ferroic titanates, Ceram. Trans. 80, 3 26 (1997) 10. Y. Fang, D. Agrawal, D.M. Roy, and R. Roy, Fabrication of transparent hydroxyapatite ceramics by microwave processing, Mater. Lett. 23, (1995) 11. Y. Fang, D. Agrawal, D.M. Roy, and R. Roy, Transparent mullite ceramics from diphasic aerogels by microwave and conventional processing, Mater. Lett. 28, (1996) 12. D. Agrawal, Fabrication of transparent ceramics by microwave sintering, presented at the First International Congress on Microwave Processing, Lake Buena Vista, FL, January 5 9 (1997) 13. J. Cheng, D. Agrawal, Y. Zhang, and R. Roy, Fabrication of translucent aluminum nitride using microwave method, presented at the 101 st American Ceramic Society Annual Meeting, Indianapolis, IN, April, (1999) 14. J. Cheng, D. Agrawal, Y. Zhang, and R. Roy, Microwave reactive sintering to fully transparent aluminum oxynitride (ALON) ceramics, presented at the Second World Congress on Microwave and Radio Frequency Processing, Orlando, Florida, April 2 6 (2000) 15. R. Roy, J.P. Cheng, D. Agrawal et al., U.S. Patents 6,004,505 (2000); 6,063,333(2000); 6,126,895(2000); 6,066,290(2000)