Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 79
Microwave Processing of Materials 5 MICROWAVE APPLICATIONS INTRODUCTION Due in large part to the overwhelming success of microwave ovens for home use, microwave processing is seen by the unwary as a panacea for all heating applications. Microwave energy is perceived to provide a means for rapid, even heating, improved processing efficiencies, and heretofore unobtainable materials properties. However, as previous sections of this report have shown, not all materials and processes are amenable to microwave processing. Even for materials and processes where microwave heating is technically an option, additional technical and economic considerations must be evaluated, on a case-by-case basis, to determine whether it is the best alternative. This chapter provides examples of work accomplished in applications of microwaves in materials processing. The observations made in previous chapters on equipment selection, process design and evaluation, and application criteria will be amplified through the examples given. Microwave energy has found general, commercial application in very few areas. These include food processing, analytical chemistry, and heating and vulcanization of rubber. Food processing and rubber manufacture involve relatively high-volume, continuous processing. Analytical chemistry applications are broad in scope and involve high-volume, repetitive, batch processing, often with long intermediate drying and reaction steps that can be shortened using microwave heating. Much work has been undertaken to investigate the use of microwaves for the processing of a wide range of materials, including ceramics, polymers, composites (ceramic and polymer matrix), powders, and minerals. Microwaves have also been investigated in a broad range of plasma processes (surface modification, chemical vapor infiltration, powder processing), chemical synthesis and processing, and waste remediation. Despite the considerable effort that has been expended in microwave process development, there has been little industrial application to date, with most of the effort still in the laboratory stage. Some of the more significant problems that have inhibited industrial application of microwave processing include: the cost of equipment; limited applicability; variation in dielectric properties with temperature; and the inherent inefficiency of electric power.
OCR for page 80
Microwave Processing of Materials Much of this work has been undertaken without the initial cross-disciplinary evaluation and processing system design approach emphasized in this report. Discussion of these results in light of this multidisciplinary approach will serve to highlight the limitations in terms of capabilities and scaling and will lead to identification of promising processes and needed research. A broad range of applications will be discussed. Much work has been accomplished on ceramic, polymer, and plasma processing, and the lessons that can be learned from this work will help to identify promising applications for future development and will help processors avoid possible pitfalls. Emerging and innovative applications in microwave chemistry, minerals processing, and waste remediation are also reviewed. CERAMICS/CERAMIC MATRIX COMPOSITES The use of microwave energy for processing ceramics and ceramic matrix composites has been the subject of a large amount of exploratory research. The range of materials and processes that have been investigated is shown in Table 5-1. TABLE 5-1 Examples of Ceramic Microwave Processing Research and Development
OCR for page 81
Microwave Processing of Materials The potential advantages of microwave processes over conventional processes for ceramic processing include reduced processing time, improved product uniformity and yields, improved or unique microstructure, and the ability to synthesize new materials (Sutton, 1989). A number of review articles on ceramic microwave applications have been published (Sutton, 1989, 1993). Rather than providing a broad review of ceramic processes, this section will examine two important processing areas——sintering and powder processing——in light of the perceived advantages described above, comment on lessons to be learned from previous work in these areas, and suggest promising applications for the future. Microwave sintering is of interest because of the extensive exploratory work accomplished and because of the broad range of ceramic materials that have been investigated. Microwave processing of ceramic powders is a relatively new area with promise of broad applications in synthesis and processing. Sintering Microwave heating has been touted as a means of sintering ceramics since the early 1970s. Microwave sintering of a number of oxides and nonoxide ceramics ranging from low-loss materials like Al2O3 to relatively high-loss materials such as SiC, TiB2, and B4C has been reported. The perceived advantages of microwave sintering over conventional sintering include expectations for more-uniform heating, better properties of the product, greater throughput with resulting smaller plant size, and greater energy efficiency. It is generally assumed that since microwave energy is deposited in the bulk, significantly less time is required to heat the part to the sintering temperature than would be required to diffuse the heat from the exterior, particularly for large parts or large batches of small parts. The resulting rapid sintering may lead to smaller grain size at a given density, with consequently better mechanical properties. Although some advantageous application of microwave processing in sintering has been demonstrated, the perceived potential of the technology has gone largely unrealized on a production scale. Oxides Berteaud and Badot (1976) investigated the sintering of alumina and zirconia and the melting of silica at 2.45 GHz in a rectangular single-mode cavity. They recognized many of the potential advantages of microwave sintering, including high thermal efficiency as well as rapid processing, and also discovered many of the problems that have plagued the process, including difficulty in temperature measurement due to temperature gradients and the propensity for thermal runaway. Colomban and Badot (1978, 1979) investigated the sintering of β-alumina, again in a single-mode cavity, where they observed rapid sintering but not the expected small grain size. Microwave sintering of alumina sparkplug insulators was investigated with goals to replace large (50-m long), gas-fired line kilns with significantly smaller equipment and to reduce process cycle-time (Schubring, 1983). Microwave sintering was found to be feasible, with cycle time reduction from 24 hours to 3—6 hours. The energy consumption was half that for gas
OCR for page 82
Microwave Processing of Materials firing, but the energy costs were higher for microwave heating because of the relative costs of gas versus electric energy. Although part to part density variations were observed in the 186-part sagger, acceptable properties were obtained. A similar study of the sintering of ferrites resulted in similar conclusions regarding feasibility (Krage, 1981). However, neither process was carried to practice. Large castable refractory crucibles have been successfully sintered in a microwave cavity (Sutton, 1988). Firing times were significantly reduced compared with conventional heating, because the penetration depth of microwaves allowed even heating throughout the thickness of the rather large sections involved. Ultra-rapid sintering of β-alumina, Al2O3, and Al2O3/TiC rods and thin-walled tubes using a rapid pass-through, zone-sintering process in single-mode applicators, has been investigated (Johnson, 1991). Isostatically pressed rods, 4 mm in diameter, of β-alumina powder were sintered at specimen translation rates of up to 40 mm/min, with the final density independent of translation rate. The time from onset of heating to final density was on the order of 30 s at the highest translation rate. Attempts to sinter thin-wall β-alumina with a diameter of 15 mm tubes failed, because a small region of the tube would become hot and remain hot to the exclusion of the rest of the specimen, even though the tube was rotated in the cavity. The size of the spot, on the order of several millimeters in diameter, was sensitive to the power applied, was stable in time, and did not propagate around the circumference of the tube. A single-mode cavity was used to sinter α-alumina rods with a diameter of 4 mm to high density and fine grain size (99.8 percent dense and 2 μm, respectively; Tian et al., 1988a). To avoid thermal runaway, applied power had to be carefully controlled as the sintering temperature was approached. Stable heating, again with high densities and fine grain size, was also observed in sintering Al2O3/TiC rods with a diameter of 4 mm (Tian et al., 1988b). Thermal runaway was avoided if the concentration of TiC was greater than about 20 percent by weight. Of several reported attempts to sinter Al2O3 in multimode cavities, the experiments of Patterson et al. (1991) were among the most successful. Single and multiple specimens (19 mm diameter by 16 mm long) were sintered to greater than 98 percent density with three different alumina powders. If the heating rate was too high, nonuniform grain sizes resulted, with the largest grains in the center of the specimen. A 60-minute firing cycle that resulted in uniform grain sizes was developed. Few details about procedures, thermal insulation, or the oven configuration were given. Sintering of a few other oxide materials with varying degrees of success has been reported. Some of these reports are listed in Table 5-2. In most cases, the procedures were not described well enough for the committee to determine whether sintering enhancement was observed. Nonoxides B4C and TiB2 have been successfully heated to very high temperatures using granular Y2O3 as the microwave-transparent primary insulation system (Holcombe and Dykes, 1991a, b). Increased density and improved mechanical properties of microwave-sintered B4C were reported
OCR for page 83
Microwave Processing of Materials TABLE 5-2 Selected Microwave Sintering reports. Material Insulation Coupling Reference Al2O3 None Self (Tian et al., 1988a) Al2O3 Al2O3 fiber Hybrid (insulation + self) (Patil et al., 1991) Al2O3 Un-named fiber Hybrid (SiC liner) (Dé et al., 1991b, c) Al2O3 Not given Not given (Patterson et al., 1991) Al2O3 Un-named fiber Hybrid (tubular receptors) (Brandon et al., 1992) Al2O3/MgO Al2O3-SiO2 fiber Not given (Cheng et al., 1992) Al2O3-TiC None Self (Tian et al., 1958b) Al2O3-ZrO4 Al2O3 and/or ZrO2 Hybrid (picket fence, 2.45 GHz); self at 28 GHz (Kimrey et al., 1991) Al2O3-ZrO4 Un-named fiber Hybrid (tubular susceptors) (Brandon et al., 1992) ß-alumina None Self (Johnson, 1991) B4C Y2O3 grain Self (Holcombe and Dykes, 1991a) BaTiO3 Y2O3 fiber and powder Probably self (Lauf et al., 1992) Hydroxyapatite Zr2O3 fiber Not given (Agrawal et al., 1992) LaCrO3 Self (Janney and Kimrey, 1992) Si3N4 ZrO2 and Safil fibers Hybrid (powder bed) (Patterson et al., 1992a) Si3N4 Not given Not given (Zhang et al., 1992) TiB2 Y grain Self (Holcombe and Dykes, 1991b) TiO2 nanophase ZrO2 fiber Not given (Eastman et al., 1991) YBa2Cu3Ox Al2O3-SiO2 Hybrid (SiC liner) (Ozzi et al., 1991) ZnO varistor Not given Probably self (Levinson et al., 1992; McMahon et al., 1991) ZrO2/12% CeO2 ZrO2 fiber Hybrid (insulation) (Janney et al., 1992b) ZrO2/8% Y2O3 ZrO2 fiber next to specimen Hybrid (insulation, picket fence; self at 28 GHz (Janney et al., 1991b, 1992b)
OCR for page 84
Microwave Processing of Materials compared with the conventionally sintered material. A barrier layer was required to preclude Y2O3 contamination of the TiB2. Batches of Si3N4 cutting tools, with 90 parts per batch, were sintered using a cylindrical multimode cavity (Patterson et al., 1992a). Parts were arranged in six layers embedded and isolated from each other within a packing powder and were enclosed in a cylindrical alumina crucible. The packing powder, consisting of 40 percent SiC, 30 percent BN, and 30 percent Si3N4 by weight, served multiple purposes——providing a source for N2, providing high thermal conductivity, acting as a getter for O2, and acting as a microwave absorber. Three conductive rings were placed around the alumina crucible to shape the microwave field. The temperature increased with some degree of nonuniformity during a slow increase in microwave power. A locally high-temperature area would commence at one end of the load and gradually spread throughout the entire load. Thus, after 50 minutes the surface temperatures ranged from 536—1190 ºC, whereas after 140 minutes the range was from 1540—1610 ºC. After optimizing the process, uniform density among the parts was obtained. Energy consumption was estimated to be on the order of 80 percent less than experienced with conventional heating. In this case, electric heat is mandated for conventional processing because of the tendency of the material to oxidize in combustion gases. Issues in Microwave Sintering Microwave Enhancement Effects There have been numerous reports of enhancement of sintering kinetics when using microwave processing. Probably the most startling is a report of as much as a 400 ºC reduction in sintering temperature along with a dramatically reduced activation energy for Al2O3 processed in a 28-GHz microwave cavity (Janney and Kimrey, 1988, 1990). As discussed in Chapter 3 of this report, significant errors in temperature measurement can lead to misleading processing results. Shielded and grounded thermocouples, as discussed in Chapter 3, as well as optical pyrometers, were used to minimize temperature-measurement errors, and carefully designed insulation systems were used in the studies referenced above to minimize temperature gradients. By switching the microwave power off and on, and observing thermocouple response, it was demonstrated that the microwave field did not bias the thermocouple output (Janney et al., 1991a). While temperature measurement may yet be a problem, it is difficult to imagine a 400 ºC error. Significant reductions in sintering temperatures or enhancements in the diffusion coefficient for sintering have also been reported for Al2O3 (Patil et al., 1991; Cheng et al., 1992) and Al2O3 doped with MgO (Cheng et al., 1992). Reductions in sintering temperature and activation energy have been reported for the sintering of zirconia-toughened alumina (Kimrey et al., 1991) and zirconia (Janney et al., 1991b, 1992a). A variety of other ceramics were similarly sintered using microwaves, including B4C (Holcombe and Dykes, 1991a), LaCrO3 (Janney and Kimrey, 1992), and Si3N4 (Tiegs et al., 1991; Kiggans et al., 1991; Kiggans and Tiegs, 1992). These results, for a broad range of materials, indicated that the reduction in sintering temperature was observed in insulators and ionically conducting materials but not in
OCR for page 85
Microwave Processing of Materials electronically conducting materials. Reductions in sintering temperature and activation energy were greater at 28 GHz than they were at 2.45 GHz. Enhanced microwave plasma sintering of alumina and a few other oxides has been observed, but only the data for alumina were presented quantitatively (Bennett et al., 1968). A 200 ºC reduction in the sintering temperature of Linde A alumina was reported compared with the same temperature and time in a conventional furnace. Rapid pass-through sintering of thin-wall tubes and rods in the radio frequency (RF) induction coupled plasma and microwave plasmas has been investigated (Sweeney and Johnson, 1991). Densification times in thin-wall tubes in the RF induction coupled plasma were as low as 10 seconds from onset to completion of densification, with final densities as high as 99.7 percent for MgO-doped alumina. Similar sintering speeds were obtained with rods in a microwave plasma. Although the microwave plasma process has not yet been thoroughly characterized, a clear enhancement of sintering was observed in the 5-MHz induction coupled plasma sintering of alumina (with pains taken to correct for temperature measurement errors through extensive calibration procedures). Microwave enhancement effects have not been observed universally. Patterson et al. (1991) saw a slight increase in the sintering rate of three alumina powders——the sintering time was cut in half at 1600 ºC relative to conventional sintering, with comparable densities and elastic modulus. Levinson et al. (1992) found no significant difference in the sintering of ZnO varistor materials in microwaves relative to that in conventional firing, and there was no difference in properties. The interdiffusion of Cr2O3 and Al2O3 under microwave heating was studied to determine if there was enhanced diffusion in ceramics heated by microwaves (Katz et al., 1991). Although a slight apparent enhancement was observed, it was concluded that this could be accounted for without resorting to a rate enhancement by microwaves. Part of the controversy surrounding the ''microwave effect'' is that satisfactory physical explanations are missing. Booske et al. (1992) proposed a theory in which the enhanced sintering is attributed to enhancement of the phonon energy distribution in the high end of the distribution. The same research group later reported that further calculations showed the proposed effect was of insufficient magnitude to explain the observations (Booske et al., 1993). A satisfactory physical explanation of microwave effects must show why electronically insulating materials have shown the effect while conducting materials have not. A series of careful experiments is needed to eliminate the doubts that remain about the "microwave effect." Since temperature measurement is often problematical, some method of internal calibration of the temperature is imperative. Hybrid Heating Electrically transparent (low-loss) materials, such as SiO2 and Al2O3, are difficult to heat at room temperature. Additionally, many materials that are hard to heat at room temperature possess electrical conductivity or dielectric loss factors that rise rapidly in magnitude as the temperature rises. Thus these materials will absorb microwave energy if they can be preheated to a suitable temperature using another source of heat. This has led to the development of passive hybrid heating using higher dielectric loss susceptors, insulation, or coatings that absorb incident microwave power more readily at low temperature.
OCR for page 86
Microwave Processing of Materials The sintering of ZrO2-toughened Al2O3, ZrO2/8% Y2O3 and zirconia/12% CeO2 at 2.45 GHz provides an example of how a hybrid heating process can improve unstable heating (Janney et al., 1992b). Although these materials could be sintered readily at 28 GHz, attempts at 2.45 GHz, where equipment costs are more attractive, were frustrated until SiC rods were inserted into the insulation that surrounded the specimens in what was referred to as the "picket fence" arrangement. The microwave energy initially heated the SiC rods, which, in turn, transferred heat to the insulation and eventually to the specimens. As the specimen temperature increased, it more effectively coupled with the microwave energy and began to heat directly. Figure 5-1, part a shows the calculated field distribution (using finite-difference time-domain modeling) in a simulation of four ceramic samples surrounded by insulation inside the cavity, showing a concentration of fields at the sides of the insulation. Although the inclusion of four SiC rods, as shown in Figure 5-1, part b, reduced the overall field strength, it helped to concentrate the fields in areas of interest near the samples rather than in the insulation. With this arrangement, these materials were successfully sintered, albeit at fairly low heating rates (2 ºC/min). Another hybrid microwave heating scheme involved applying a thin layer of SiC powder to the interior of the thermal insulation chamber that is placed within the microwave oven (Dé et al., 1991a, b, c). As in the picket fence arrangement, the silicon carbide is initially heated by the microwaves, transferring heat to the specimen. The silicon carbide layer is thin enough that significant penetration of the microwaves occurs. With this arrangement, a number of ceramic materials have been successfully sintered. In yet another hybrid heating process, tubular susceptors of various sizes were inserted over relatively large sized compacts of alumina and zirconia toughened alumina (50 mm diameter by 60 mm long) which were then sintered to 1500 ºC at heating rates of about 10 ºC/min (Brandon et al., 1992). Compacts of this size could not be sintered using conventional processes at 10 or even 5 ºC/min without cracking. Simulations have shown that increasing the ambient temperature through some form of hybrid heating can increase the critical temperature to as high as required for sintering. These results explain the success achieved with hybrid heating processes that is reported in the literature. The simulations also explain, qualitatively at least, the observed difficulties in sintering low-loss oxide materials by microwave heating (Spotz, et al., 1993). Although some impressive results have been reported in the hybrid heating of alumina, controlled rapid heating of oxides with both low initial dielectric loss factors and high temperature dependence of dielectric loss factors is difficult to achieve. Success generally has been limited to single specimens of simple geometry in carefully designed sintering chambers. Insulation Unstable heating due to changing permittivity or thermal gradients caused by heat loss from part surfaces can be minimized using effective insulation. Almost all cases in which successful microwave sintering has been reported have necessitated carefully designed insulation systems. In many cases, packing powders are required, which would only be acceptable for products of very high value. Development of a workable insulating system has been identified
OCR for page 87
Microwave Processing of Materials FIGURE 5-1 Calculated electric field distribution in a multimode cavity when four ceramic samples ( , σ = 64 × 10-6 S/m) are surrounded by insulation. (a) samples alone; (b) samples plus SiC rods.
OCR for page 88
Microwave Processing of Materials as "one of the most challenging tasks in the high temperature microwave processing of ceramics" (Janney et al., 1992b). While temperature gradients within the specimen can be reduced by the presence of insulation, they can be eliminated only if all of the microwave energy is absorbed by the insulation, or a susceptor, and subsequently transferred to the specimen. In fact, the various hybrid heating approaches move in this direction. However, temperature gradients of a certain magnitude may be acceptable. If the microwave power absorption increases sharply with increasing temperature, there always will be heating problems because of this volumetric heating and surface cooling of even well-insulated specimens. In some cases, surface thermal gradients may be overcome by slow heating, but, in other cases, microwave heating will be unstable, making uniform sintering impossible. This difficulty is particularly exacerbated in the case of multiple-part sintering, where a hotter part may preferentially absorb more microwave energy at the expense of cooler parts. Materials with well-behaved heating behavior, such as the carbides, usually require very high sintering temperatures. This presents problems with regard to setters and insulation. There are no truly microwave-transparent insulation materials capable of operation in the 2000 ºC range, although granular Y2O3 has shown some promise. The need for carefully controlled insulation has forced microwave sintering to be basically a batch process, often with only a single part being sintered at a time. Successful demonstration of large batches has only rarely been successful. Even in successful cases, part-to-part variations were common. In the extreme cases, some parts have undergone thermal runaway and others were not fully sintered. Unfortunately, these effects are exacerbated by large batch size and rapid heating, both of which are desirable from a manufacturing point of view. Further investigation is needed to discover the regimes of microwave-power absorption characteristics, batch size, heating rate, and other variables in which microwave sintering can be reproducible and uniform. Thermal Runaway As discussed in Chapter 2, the rapid rise in dielectric loss factor with temperature is the major issue in thermal runaway and temperature nonuniformity. Therefore, although microwave heating frequently is touted as providing more uniform heating, nonuniform heating is a reality in many oxides, often at nominal heating rates. The situation is worse when a multitude of parts are heated together, or for other than simple specimen geometry. Some general observations can be made about factors relating to thermal runaway. First, if the temperature dependence of the power absorption is less than the temperature dependence of the heat dissipation at the surface of the specimen plus insulation system, stable heating should be observed. Second, hybrid heating using either lossy insulation or other susceptors that absorb a significant fraction of the microwave power and transmit it to the specimen by conduction or radiation is important in stable heating. Finally multiple specimens of differing size, or specimens with varying cross section or complex shapes, will be particularly difficult
OCR for page 89
Microwave Processing of Materials to heat uniformly. Materials with lower temperature dependence of dielectric loss factor may be heated stably. However, the uniformity issues for complex shapes or differing sizes within a batch will persist. Further work is required to determine more fully the conditions under which stable heating of various materials can be achieved. Property Enhancement The final issue is the question of whether there are fundamental differences in the properties achievable by microwave sintering and those achievable by other methods. Microwave sintered and hot isostatically pressed Si3N4 cutting tools showed significantly improved performance compared with commercially available cutting tools (Patterson, 1992b). Dé et al. (1991b, c) investigated the effects of heating rate on the densification and microstructure of conventionally and microwave sintered materials in a hybrid system. They observed that higher heating rates result in higher density and smaller grains, just as with conventional fast firing. However, higher heating rates were achieved in the hybrid system than was possible with the same specimen size in a conventional furnace. It may be significant that the microwave sintered specimens had a smaller grain size at any given density during the densification process than did conventionally fast-fired specimens. Unfortunately, other researchers have not reported relationships of grain size versus density that would make it possible to determine whether this effect is widely realized with other materials. Powder Processing The synthesis and processing of powders is a key technology area affecting the future development of advanced ceramic materials. The application of microwaves to powder processing technology is relatively new and will be discussed briefly. Table 5-3 summarizes some of the areas where microwaves have been applied to ceramic powder processing. Powder Synthesis The characteristics of a starting powder (composition, size, structure, shape, etc.) strongly affect the control over the sintering behavior, microstructural development, improved properties, and reliability of the final product (Johnson, 1987). For this reason, there continues to be a significant effort to develop improved and tailorable powders to meet the increasing demands for a wide range of future, advanced ceramic products (Messing et al., 1987, 1988a, b). The application of microwaves to the synthesis of ceramic (oxide and nonoxide) powders is a recent and emerging development and offers some unique benefits, especially with respect to producing particles of submicron (nano) size with controlled compositions. Microwave synthesis of ceramic powders offers greater process flexibility by taking advantage of several
OCR for page 106
Microwave Processing of Materials plasma processing. Depending on the process, the differences may be an issue in deciding whether to use a microwave plasma. An excellent review of microwave plasmas has appeared recently (Moisan and Pelletier, 1992). The current literature on microwave plasma processing is heavily dominated by reports on diamond film formation. The growth of diamond films requires an abundance of atomic hydrogen, which etches graphitic nuclei in the deposit and leaves the diamond-like nuclei to grow. Plasmas generated by any means are, in general, good sources of this species. There are certain perceived advantages of microwave plasmas over other diamond film-forming methods. Cited examples include stability and reproducibility of the plasma, energy efficiency, availability of inexpensive magnetrons, and potential for scaling to larger sizes (NRC, 1990). A further advantage is that the microwave plasma can heat the substrate to the temperature required for good deposition conditions (greater than 500 ºC). Microwave plasma processing has had a major impact in microelectronics device processing, where it is a mature art. A state-of-the-art review listed microwave plasma processing as a key technology that was sufficiently developed for imminent implementation in industry (NRC, 1986). The two major applications are plasma-enhanced chemical vapor deposition and etching, which includes the possibility of high-resolution etching of silicon (Moisan and Pelletier, 1992). Deposition Microwave plasma deposited materials include silicon films, which are amorphous or polycrystalline depending upon the substrate temperature, and silicon oxide and nitride. In addition, silicon can be oxidized to form silicon oxide films. The primary advantage of microwave plasma-enhanced chemical vapor deposition is reduction in radiation damage compared with conventional RF plasma chemical vapor deposition. This is because the microwave discharge results in a lower acceleration potential between the plasma and the substrate. The electron cyclotron resonance plasma technique is particularly useful in depositing silicon oxide and silicon nitride films on silicon for device processing. Films deposited at temperatures less than 150 ºC have chemical and physical properties equivalent to films deposited at 900 ºC using conventional chemical vapor deposition processes, and the low-energy ion bombardment does not damage the substrate. Similarly, silicon oxide films grown on silicon appear comparable to those grown by conventional thermal oxidation at 1000 ºC. Proper design of equipment, including positioning of feedstock injection, is important to avoid unwanted depositions on the walls of the reactor or other places. The second major application of microwave plasmas is etching in electronic device processing. The principal advantage is that the microwave plasmas are more selective between photoresist and the underlying material. The second advantage is the lower intensity of radiation damage in reactive ion etching compared with conventional plasma etching because of the lower acceleration potential for ions. Finally, microwave plasma etching is reported to give highly anisotropic etching, although an RF bias is usually required to achieve the desired level of anisotropic etching (Moisan and Pelletier, 1992).
OCR for page 107
Microwave Processing of Materials An RF bias to a microwave plasma not only increases the directionality of the etching, it also increases the rate of etching. Thus the microwave plasma is more selective than the RF plasma, whereas the RF plasma provides better directionality, so a combination of the two is required to obtain the desirable degree of both selectivity and directionality. Surface Treatment A third area of use of microwave plasmas is in surface treatment, where it has been applied to polymer fibers, as well as in the microelectronics industry. Chemical modification of the surface can be achieved with or without adding reactive components in the plasma. It has been demonstrated that treatment of polyamide fibers in a large microwave plasma system improves the bonding between the fiber and the matrix in composites (Wertheimer and Schreiber, 1981). This results in a dramatically different response to mechanical loads, providing for higher strength but at the same time a lower ballistic strength. Fiber mechanical properties can be degraded by the microwave plasma treatment. Microwave plasmas are used also to promote adhesion of films in microelectronics processing. Advantage is taken here of the lower degree of radiation damage that is achievable with the microwave plasma than with other plasmas. By using a combination of microwave excitation and RF biasing, it is possible to independently control the relative contribution of the chemical component and the physical component (energetic ions, electrons, and photons). In addition, microwave plasma sources have been used to passivate the surface of GaAs, resulting in superior device properties. The avoidance of direct ion bombardment of the surface was key to the success of this application. The interactions among the physical and chemical components of a microwave plasma system are numerous and not well understood. Further work of a basic nature is required to better elucidate these interactions. Until then the industry and art probably will be dominated by solutions arrived at by trial and error. One of the aspects that should be explored in more detail is the effect of variable frequency on the chemical and physical processes occurring in the microwave plasma and on interactions with the substrate during deposition, etching, and surface modification. MINERALS PROCESSING The minerals and extractive metallurgy industry is a major consumer of energy and contributor to environmental degradation. For instance, about 4 percent of the carbon dioxide emitted to the atmosphere comes from the worldwide extractive metallurgy industry (Forrest and Szekely, 1991). Microwave processing may provide substantial benefits in reducing energy consumption and environmental impact by this industry. In mineral processing, the extraction of values in an ore from the waste or gangue is an energy intensive and energy inefficient process. According to Walkiewicz et al., (1991) approximately 50—70 percent of the energy used for minerals extraction is consumed during
OCR for page 108
Microwave Processing of Materials comminution (grinding) and separation. The energy efficiency of conventional grinding is about 1 percent, and most of the energy is wasted in heat generated in the material and equipment. Microwave processing of ores provides a possible mechanism to induce fractures between the values in the ore and the waste material surrounding it, due to the differential in absorption of microwaves and the differences in thermal expansion among various materials. These differentials induce tensile fractures in the material (Figure 5-5), and as a consequence, substantially reduce the energy required in grinding to separate the values from the waste material. A limited amount of work has been done in this area, and it is clear that microwaving of certain sulfide and oxide ores does result in fractures along the interface between the values and the waste material. Grindability tests show improved grindability (less energy is required to achieve a given mesh size for the ore) for a series of iron ores. However, it is not clear that the reduction in energy required in grinding will balance or exceed the energy expended in the microwave treatment of the ore. In iron ores the preliminary results indicate a deficit in the energy balance. To justify using microwave processing, it is also necessary to consider wear on grinding mills, cleaner liberation of the values in the ore, and lower chemical emission during the pyrometallurgy and hydrometallurgy processing steps. Additional research needs to be done to determine the efficiency of coupling the microwave energy to the ore, the effect of particle size on susceptibility to cracking, and the effect on the cracking efficiency of using high power sources (the present work has been limited to maximum of about 3 kW). MICROWAVE CHEMISTRY Microwave chemistry is a rapidly growing field that has been gaining attention recently (IMPI, 1992; EPRI, 1993). The effect of microwave processing on chemical reactions or processes touches on most of the application areas emphasized elsewhere in this report. Some of these include ceramic sintering and synthesis, polymer curing, plasma processing, and waste remediation. In this section, applications in analytical and synthetic chemistry and extensions of these applications to the chemical industry are considered. The most widespread use of microwaves in chemistry is in analytical laboratories. Microwave energy has been used in analytical chemistry since the mid-1970s, primarily for sample preparation. In that time, microwave ovens have become generally accepted tools in the modern analytical laboratory, increasing from a couple hundred units in 1975 to close to 10,000 units in 1992, while the annual expenditures for laboratory microwave systems increased from under $1 million in 1975 to close to $50 million (Neas, 1992a). Applications span a wide range of sample preparation methods including drying, extractions, acid dissolution, decomposition, and hydrolysis. In these applications, microwave heating has been used as a replacement for conventional heating techniques. In general, analytical chemistry involves time-consuming sample preparation steps to get the samples in a suitable form for analysis.
OCR for page 109
Microwave Processing of Materials FIGURE 5-5 Photomicrograph of pyrite ore (a) nonmicrowaved and (b) microwaved, showing stress cracking. The light phase is pyrite; and the dark phase is quartz (magnified 100×). (Courtesy of J. Walkiewicz, U.S. Bureau of Mines) Microwave digestion of materials, such as minerals, oxides, glasses, and alloys, is used in laboratories worldwide to prepare samples for chemical analysis. The decomposition rate of many difficult-to-dissolve materials in closed-reaction vessels is greatly enhanced by using microwave energy; often only a few minutes are required as opposed to the several hours needed for conventional means (Kingston and Jassie, 1988a, b). In addition, volatile elements such as selenium and phosphorous can be quantitatively retained in a sealed vessel using microwave decomposition prior to instrumental analysis (Patterson et al., 1988).
OCR for page 110
Microwave Processing of Materials Applications in the chemical laboratory generally use relatively simple ovens and controls. Most of the development work in equipment for these applications has involved improving the existing equipment to extend the operating range and to improve safety and reproducibility. Examples include improved turntables and sample fixtures, pressure vessels fabricated from glass and quartz to allow higher reaction temperatures and pressures than teflon vessels, and optimized pressure relief valves (Baghurst and Mingos, 1992b). The result of these advances has been the development of testing standards that are simple, reproducible, and automatable (Kingston, 1992). A wide variety of organic synthetic reactions have been shown to be enhanced by microwave processing (Bose et al., 1993; Majetich and Hicks, 1993). Using microwave processing, a number of fundamental organic reactions have shown accelerated reaction rates and increased yields over conventional techniques. While these processes have not yet been scaled to production, important advantages have been realized in education, where reactions that took too long to accomplish in a laboratory session using conventional heating can now be completed using microwave heating. The primary motivation for use of microwave heating has been time savings through rapid heating, rather than any nonthermal effects. Penetrating radiation (and reverse thermal gradients), the ability to superheat polar solvents, and the ability to selectively heat reactive or catalytic compounds were responsible for time savings realized in chemical processes. Microwave energy penetrates into the interior of the sample without relying on conduction from the surface required in conventional heating methods. This allows the entire sample temperature to be raised rapidly without overheating, and possibly degrading, the surface. Convective heat losses from the surface to the cooler surroundings allow processors to take advantage of reverse thermal gradients. The reaction temperature of solvent diluents can be raised above the ambient boiling points of the diluents in both closed-and open-reaction vessels (Baghurst and Mingos, 1992a). This allows for significant increases in reaction rates in a variety of applications (Mingos, 1993). Reaction rate enhancements were attributable to Arrhenius rate effects due to increased reaction temperature or selective heating of reactants over diluents. There are no persuasive arguments to support nonthermal reaction enhancements attributable to the use of microwaves (Majetich, 1992; Mingos, 1992). In closed vessels, the increased vapor pressure over the liquid suppresses further boiling. Microwave super-heating of volatile solvents can lead to significant acceleration of chemical processes compared with conventional reflux conditions. The development of microwave transparent glass and quartz reaction vessels and improved pressure-relief valves has been critical in allowing attainment of higher temperatures and pressures than was possible with low-loss teflon vessels (Baghurst and Mingos, 1992b). In open vessels, most polar solvents have an inherent ability to be heated above their conventional boiling points. This effect has been observed by several researchers (Mingos, 1993; Baghurst and Mingos, 1992a; Neas, 1992b; Majetich, 1992). The phenomenon has been explained using a model of nucleation-limited boiling point (Baghurst and Mingos, 1992a). During a boiling process, bubbles nucleate preferentially at sites (cavities, pits, scratches) on the vessel wall, allowing growth of the vapor phase. With conventional heating, the vessel wall and liquid surface are generally hotter than the bulk. In microwave processes, however, the vessel wall is cooler than the bulk solution due to convective heat losses from the surface, allowing the
OCR for page 111
Microwave Processing of Materials bulk to attain temperatures above the conventional boiling point before the boiling process commences. As shown in Table 5-7, the nucleation-limited boiling point varies with the solvent and with the ability of the solvent to wet the vessel surface, that is, the more effectively the solvent wets the vessel, the more difficult bubble nucleation becomes. TABLE 5-7 Nucleation-Limited Boiling Points for a Range of Solvents Solvent B.p./(ºc) Nucleation-Limited Boiling Point (ºC) Boiling Point Change (ºC) Water 100 104 4 Ethanol 79 103 24 Methanol 65 84 19 Dichloromethane 40 55 15 Tetrahydrofuran 66 81 15 Acetonitrile 81 107 26 Propan-2-ol 82 100 18 Acetone 56 81 25 Butanol 118 132 14 1.2-Dimethoxyethane 85 106 21 Diglyme 162 175 13 Ethyl acetate 78 95 17 Acetic anhydride 140 155 15 iso-Pentyl alcohol 130 149 19 Butan-2-one 80 97 17 Chlorobenzene 132 150 18 Trichloroethylene 87 108 21 Dimethylformamide 153 170 17 Chlorobutane 78 100 22 iso-Propyl ether 69 84 16 Source: Baghurst and Mingos, 1992a. By using microwave heating, the processor is able to target compounds with high dielectric loss over less-lossy compounds. This characteristic has been shown to enhance a number of chemical processes, including catalytic reactions utilizing metallic or dielectric catalysts, gas-phase synthesis of metal halides and nitrides, and metal reduction processes (Bond et al., 1992). The promising future of microwave chemistry to the chemical industry is just beginning to be realized. Advantages in the form of time savings, increased reaction yields, and new processes have been demonstrated in the laboratory using simple multimode ovens. Scaling to
OCR for page 112
Microwave Processing of Materials widespread production applications requires development of applicators and handling systems to account for increased product throughput, automation, and large reactors. Microwave processing for small-scale, custom organic synthesis looks promising due to the relatively modest equipment investment, broad applicability to a variety of reactions, and significantly reduced processing times (Bose et al., 1993). The availability of equipment and the long history of use in the laboratory makes these near-term applications low risk. Another area where microwaves can show an advantage is in producing products or intermediates that are needed in small quantities and may be hazardous and expensive to ship, store, and handle (Wan and Koch, 1993). WASTE PROCESSING AND RECYCLING The processing of industrial wastes is an area of tremendous promise for the application of microwave energy. The types of industrial waste that have been shown to be amenable to microwave processing, at least at laboratory scale, include hazardous waste (including toxic and radioactive) with high disposal, storage, or treatment cost and nonhazardous waste where the recovery or reuse of a raw material represents a significant cost or energy savings. Waste processing includes treatment or remediation of process wastes, detoxification or consolidation of stored waste, or cleanup of storage or disposal sites. The application of microwave energy in the processing of industrial waste has seen significant progress in terms of process development and demonstration but limited commercial application. In varying degrees, applications in this area take advantage of unique features of microwave heating: rapid heating, selective coupling with lossy constituents, and reaction steps not possible or practical with other methods. Process Waste Treatment Potential applications of process waste treatment include microwave plasma hydrogen sulfide dissociation, detoxification of trichloroethane (TCE) through microwave plasma assisted oxidation, and microwave plasma regeneration of activated carbon. The dissociation of hydrogen sulfide (H2S) in a microwave plasma was first described in Soviet literature (Balebanov et al., 1985). The potential for this process is in the refuting industry for the treatment of the sour gas resulting from hydrodesulfurization of hydrocarbon feedstocks. Subsequent work at Argonne National Laboratory has validated the applicability (Harkness et al., 1990) and economic viability (Daniels et al., 1992) of this process in the treatment of refining wastes. A schematic of the microwave dissociation process is shown in Figure 5-6. The attainable conversions, between 40 and 90 percent, were most sensitive to gas flow rate and power. Conversions up to 99 percent are achievable by cycling the residual H2S back through the process in multiple passes. Work is continuing to scale the process. The economic viability of the process depends on the sale of recovered hydrogen and is sensitive to required dissociation power.
OCR for page 113
Microwave Processing of Materials FIGURE 5-6 Hydrogen sulfide waste-treatment process utilizing microwave plasma dissociation. (Courtesy of E. Daniels, Argonne National Laboratory) TCE oxidation and activated carbon regeneration are accomplished through selective heating of lossy components (SiC or carbon) in a fluidized bed. These processes cause degradation of hazardous organic compounds at significantly lower overall temperatures than conventional heating methods. Additionally, severe corrosion of furnace components caused by the gases released in conventional high-temperature oxidation of chlorinated hydrocarbons is eliminated in the microwave process. Stored Waste Treatment Examples of microwave processing of stored waste include ''in-can'' evaporation of water and consolidation of low-level radioactive waste (Oda et al., 1991; White et al., 1991a). These processes heat low-level sludge wastes in the final storage containers by applying a microwave field using a slotted waveguide applicator. In-can processes use the rapid, selective heating of
OCR for page 114
Microwave Processing of Materials water possible with microwaves and the portability of microwave equipment to reduce transfer and handling of hazardous materials that is necessary when using conventional drying approaches. This approach need not be limited to radioactive sludge but can be applied to any evaporative treatment of stored materials. Even though, as mentioned earlier, the efficiency of bulk drying using microwaves is questionable, the cost avoidance realized by reducing handling steps may justify the increased energy costs. Waste-Site Cleanup The cleanup of contaminated industrial, disposal, and storage sites is a formidable task due to the large number of waste sites and the complex chemistries and remediation requirements involved. Currently, the majority of site cleanup efforts require removal and incineration of the waste. Since removal and transfer of contaminated materials for incineration may represent an unacceptable risk, innovative processes to cleanup contaminated storage and disposal sites are being investigated. Microwave processing shows great promise for site cleanup applications, since microwaves can be applied in situ, avoiding costly and risky excavation and transportation, and can target compounds with high dielectric loss for selective heating, for example, moisture in soils (Dauerman, 1992). Potential applications of microwave processes for cleanup of contaminated sites include removal of volatile organic compounds from soil (George et al., 1991; Windgasse and Dauerman, 1992) and remediation of soils contaminated with nonvolatile organic compounds, by causing reaction with bound indigenous organics (Zhu et al., 1992), and chromium, by causing conversion from the toxic hexavalent form to the nontoxic trivalent form (Sedhom et al., 1992). The feasibility of these processes has been demonstrated on a bench scale in a multimode oven. However, the challenge of bringing applicators and sufficient power to waste sites is formidable. If the promise of these applications is to be realized, additional work needs to be done to develop applicators for in situ processing and to show applicability and cost effectiveness on a larger scale in the field. The cleanup of surface layers (0.5—5 cm depth) of concrete structures contaminated with radioactive or hazardous materials is a potentially costly problem affecting research laboratories, power plants, and processing and storage facilities. Mechanical removal methods create potentially hazardous dust, may drive contamination into the interior, or may create a large waste stream of contaminated water generated in dust amelioration efforts. Microwaves have been shown, in experiments in Japan (Yasunaka et al., 1987), Britain (Hills, 1989), and the United States (White et al., 1991b), to be effective in rapidly removing the outer layer of concrete in a dry process with reduced dust generation. Work is underway at Oak Ridge National Laboratory to scale-up the microwave concrete-removal process and to develop, build, and test a full-scale prototype. It is believed that removal rates exceeding those attainable through mechanical techniques are possible with optimized power, frequency, and applicator design. A schematic of the prototype apparatus is shown in Figure 5-7.
OCR for page 115
Microwave Processing of Materials FIGURE 5-7 Schematic of prototype concrete-scabbing apparatus. (Courtesy of Dr. Terry L. White, Fusion Energy Division, Oak Ridge National Laboratory, Department of Energy). SUMMARY While a wide variety of materials have been processed using microwaves, including rubber, polymers, ceramics, composites, minerals, soils, wastes, chemicals, and powders, there are characteristics that make some materials very difficult to process. First, materials with significant ionic or metallic conductivity cannot be effectively processed due to inadequate penetration of the microwave energy. Second, insulators with low dielectric loss, including oxide ceramics and thermoplastic polymers, are difficult to heat from room temperature due to their low absorption of the incident energy. Since permittivity and loss factors often increase with temperature, hybrid heating may be used to process these types of materials by using alternate or indirect heating to raise the temperature of the parts to where they can be more effectively heated with microwaves. Finally, materials with permittivity or loss factors that increase rapidly during processing, such as alumina, can exhibit hot spots and thermal runaway. Although insulation or hybrid heating can improve the situation, stable microwave heating of these types of materials is problematic.
OCR for page 116
Microwave Processing of Materials Enhanced apparent process kinetics due to microwave processing have been claimed for a range of materials, most notably ceramic sintering and polymer curing. However, in most cases, insufficient care was taken in temperature control and measurement and in measurement of critical process variables and material physical properties. A series of careful experiments with an internal calibration of the temperature is needed to eliminate the doubts that remain about the microwave enhancement effects. Further investigation is needed to develop maps of the regimes of microwave-power absorption characteristics, batch size, heating rate, and other variables where microwave processing can be reproducible and uniform. This would allow processors to make informed decisions concerning microwave applications and process and equipment selection, while avoiding inefficient heating, uneven heating, and thermal runaway problems that have plagued earlier attempts. Specific processes that show promise for future development include: ceramic processes including drying, chemical vapor infiltration, reaction bonding of silicon nitride, powder synthesis, and joining; polymeric composite pultrusion, ultradrawing of polymeric fibers, and adhesive bonding with intrinsically conducting organic polymers; chemical processes, including custom organic synthesis, hazardous materials processing, solvent extraction, and drying; and industrial waste processing, including treatment or remediation of process wastes, detoxification or consolidation of stored waste, and cleanup of storage or disposal sites. In general, the elements required for successful application of microwave processing to industrial materials include selection of materials amenable to microwave processing; an understanding of the process requirements; an understanding of the process economics; characterization of material thermochemical properties; selection of equipment and design of applicators suitable for the application; an understanding of how the parts to be processed will interact with the microwave field; and adequate measurement and control of process variables such as incident power, part temperature, and field strength.
Representative terms from entire chapter: