SESSION 3
MATERIALS PROCESSING AND SYNTHESIS

Bruce M. Clemens

Stanford University

Stanford, California

Rapporteur's Report

INTRODUCTION

This workshop was organized to assist the Committee on Potential Applications of Concentrated Solar Photons in achieving the goals of its study. These goals, as listed in the Committee's statement of work, are

  • assess the knowledge base for photon/matter interaction phenomena underlying the use of concentrated solar flux for prospective new non-electric applications;

  • critically evaluate the merits of potential new applications for both near-and long-term use of concentrated solar energy, particularly those applications that address national needs and have the potential for being cost-effective through research and development; and

  • recommend research paths and priorities to enhance the scientific basis of, and increase the potential for, the development of successful applications.

Due to strong absorption, solar radiation will interact with the surface region of most materials. Thus, in Session 3, Materials Processing and Synthesis, the technology area judged to be most amenable to application of concentrated solar photons was surface treatment and modification. The first two speakers in Session 3 addressed this area, with descriptions of current surface modification and thin film growth technologies. A summary of the various current techniques was made, and the features which make each viable were accentuated. The third speaker discussed current research at the Solar Energy Research Institute aimed at exploring the application of concentrated solar photons to surface treatment. Proof of concept experiments have been performed in several application areas. The fourth speaker discussed the economic analysis of solar furnaces versus conventional intense light sources. Points raised during the discussion included the unique aspects of solar light which might make applications feasible and recommendations for future work.

This report is organized in a manner analogous to the presentations. The first section concerns current surface modification and thin film growth technologies. The second section is a discussion of the possible applications of concentrated solar photons. The third section summarizes the results of an economic analysis, and the fourth has suggestions for future research and organizational needs.

PRESENT SURFACE TREATMENT TECHNIQUES

Surface treatment is used to impart desirable properties to the surface of a sample or part without the added expense in materials and energy required



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



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 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy SESSION 3 MATERIALS PROCESSING AND SYNTHESIS Bruce M. Clemens Stanford University Stanford, California Rapporteur's Report INTRODUCTION This workshop was organized to assist the Committee on Potential Applications of Concentrated Solar Photons in achieving the goals of its study. These goals, as listed in the Committee's statement of work, are assess the knowledge base for photon/matter interaction phenomena underlying the use of concentrated solar flux for prospective new non-electric applications; critically evaluate the merits of potential new applications for both near-and long-term use of concentrated solar energy, particularly those applications that address national needs and have the potential for being cost-effective through research and development; and recommend research paths and priorities to enhance the scientific basis of, and increase the potential for, the development of successful applications. Due to strong absorption, solar radiation will interact with the surface region of most materials. Thus, in Session 3, Materials Processing and Synthesis, the technology area judged to be most amenable to application of concentrated solar photons was surface treatment and modification. The first two speakers in Session 3 addressed this area, with descriptions of current surface modification and thin film growth technologies. A summary of the various current techniques was made, and the features which make each viable were accentuated. The third speaker discussed current research at the Solar Energy Research Institute aimed at exploring the application of concentrated solar photons to surface treatment. Proof of concept experiments have been performed in several application areas. The fourth speaker discussed the economic analysis of solar furnaces versus conventional intense light sources. Points raised during the discussion included the unique aspects of solar light which might make applications feasible and recommendations for future work. This report is organized in a manner analogous to the presentations. The first section concerns current surface modification and thin film growth technologies. The second section is a discussion of the possible applications of concentrated solar photons. The third section summarizes the results of an economic analysis, and the fourth has suggestions for future research and organizational needs. PRESENT SURFACE TREATMENT TECHNIQUES Surface treatment is used to impart desirable properties to the surface of a sample or part without the added expense in materials and energy required

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy to alter the bulk. This gives the ability to independently optimize the bulk and surface properties; for instance, a surface region may need to be hard or corrosion resistant, while mechanical toughness is required of the bulk. In the first talk of Session 3, William D. Sproul of Northwestern University, discussed several current surface modification technologies. Surface Heating with Light Since visible light is absorbed in about the first 30 nm in most metals and opaque materials, intense light fluxes can be used to selectively heat the surface region. Techniques which use light to heat surfaces can be classified according to the interaction time and energy density. In general, the processes which have a short interaction time (10-4-10-10 sec.) have a high energy density (104 -1010 W/m2). Short interaction time processes can achieve extremely high heating and cooling rates (1012 K/sec.), and thus can be used to produce nonequilibrium phases such as metallic glasses. Laser glazing of a previously applied thin film can heal pinholes and defects as well as promote adhesion and densification. Slower interaction time processes can be used to anneal or harden the surface region or melt a previously applied powder or film. Powder melting can produce a dense well-adhered film. These films can be several mils to over a hundred mils thick. One of the primary disadvantages is that the surface tends to be quite rough and requires a postmachining process. Lasers can be used to ablate material to pattern the surface on a very fine scale. Printing roles are patterned in this manner to achieve optimum inking characteristics. Thermal Spray/Plasma Arc In the techniques of thermal spray or plasma arc, powder is fed into a heat source (either a flame or electric arc) where it is melted and then accelerated as molten droplets into the surface where it condenses into a film. This is a relatively inexpensive and rapid process for forming a surface film, but the films which result have a high density of several types of defects, including unmelted particles, voids, and oxidized particles. The process is extremely complex, with many parameters that affect the final film properties in a complicated manner. Nonetheless, this is a widely used process, particularly in the aerospace industry where a typical gas turbine engine has 15 pounds of thermal spray coatings. Chemical Vapor Deposition Chemical vapor deposition (CVD) uses thermal energy to decompose precursor molecules which results in deposition of a film. This technique is widely used in the semiconductor and metallurgical coatings industries to produce Si, SiO2, TiN, and Al2O3, among others. of particular interest is photo-assisted CVD where light is used to not only heat the surface region, but to photolytically assist in breaking the precursor molecule bonds. Principal disadvantages are the high temperatures required and the toxicity of the precursor materials. Advantages are nondirectionality of deposition, which results in uniform coatings on complex shaped parts, and low cost. Physical Vapor Deposition In physical vapor deposition, films are produced by physical transport of atoms from source to substrate. Evaporation techniques use resistance or electron beam heated hearths to evaporate the source atoms, while sputter deposition forms films from atoms ejected from the source by bombardment by energetic particles, usually inert gas ions. Both techniques need a vacuum environment, and the film properties are strongly affected by the energy and surface mobility of the arriving species.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Ion Beam Techniques George Fenske, of Argonne National Laboratory, discussed several techniques that utilize bombardment with energetic ions to alter the surface properties. These techniques can be classified as ion mixing, ion implantation, or ion-assisted deposition. In ion mixing, an ion beam is used to posttreat a film to achieve mixing at the interface resulting in improved adhesion. Ion implantation uses ion beams to directly enrich the surface region in the ion species. Both these techniques require expensive ion sources and are inherently slow. Ion-assisted deposition uses ion bombardment during vapor deposition to enhance mixing with the substrate and arriving species mobility, resulting in improved film qualities. In addition, ion bombardment can be used prior to film deposition to clean and roughen substrates and thus improve film adhesion. These later two uses of ion beams can utilize less expensive lower energy ion sources than ion beam mixing or implantation. MATERIAL USES OF CONCENTRATED SOLAR PHOTONS In order to assess the application of concentrated solar photons for materials, it is instructive to examine the properties obtainable in concentrated solar beams. J. Roland Pitts of the Solar Energy Research Institute discussed solar radiation, its concentration, and several possible applications. The solar spectra is (not coincidentally) mainly in the visible spectral range and has a penetration depth of about 20–50 nm in many metals and opaque materials. Thus highly concentrated solar beams are ideally suited for surface heating. Solar furnaces can obtain concentration factors of up to 10,000 resulting in fluxes of 1000 W/cm2. Solar light can be delivered at a fraction of the energy cost of conventional sources such as lasers or arc furnaces. During discussion, the relevant characteristics of solar light were proposed to be its radiative nature, allowing its use for surface heating; the ability to concentrate solar light to produce high flux-high area light sources; and the spectral characteristics of solar light. The possible applications which utilize these characteristics can be classified into thermal and photolytic. Thermal Applications This class of applications utilizes concentrated solar flux to heat sample surfaces. There are many technologies that can utilize this ability, and Pitts discussed many proof of concept experiments which have been performed at SERI. Included in these was powder melting or cladding, surface film glazing, including the initiating of self-propagating reactions of multilayer Ni/Al films, and hardening of steel surfaces. Concentrated solar photons were used to form the desired high-Tc, phase during rapid thermal annealing of a thin film in an oxygen atmosphere, as well as for thin film deposition by CVD. Films formed in this manner included diamond-like carbon and TiN, and SiC. In addition the high flux capability of a solar furnace could be used for rapid thermal chemical vapor deposition. Other applications suggested during the workshop included using solar energy as a heating source for evaporation. This could be used in environments where electron beams cannot be used, such as in evaporation in a partial vacuum to produce nanophase powders. Also suggested was the use of concentrated solar photons to heat a substrate either prior to or during deposition. This application is

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy attractive in cases where extremely high temperatures (~2000 K) are needed, as these are difficult or impossible to obtain in other manners. Photolytic Applications Photolytic applications would utilize solar light to directly break chemical bonds. Due to the broadband characteristics of the solar spectra, there is relatively little power at any given wavelength. Thus a photolytic application should take advantage of as much of the solar spectrum as possible. Furthermore, most absorption processes need shorter wavelength photons than are available in the solar spectrum, so some control of absorption processes is needed for these applications. This may be possible with absorption of molecules on surfaces. One area where these types of processes may be useful is in polymers, where bonds can be readily broken by solar radiation. No applications were suggested in the workshop, and a need for concepts in this area was expressed. ECONOMIC ANALYSIS Greg Kolb, of Sandia National Laboratories, presented the results of his study comparing the relative economic advantages of solar furnaces and high intensity arc lamps. Arc lamps are cheaper initially and are unaffected by weather. They perform better for assembly line continuous processes since they can be operated continuously. However, the energy costs of arc lamps are much higher, and as the cost of energy rises, the solar furnace will become more practical. In addition, for applications involving batch processing of parts which utilize 5000 suns or more, a solar furnace is more economical to run even with current energy prices. The solar furnace can also have a long distance between workpiece and optical elements, allowing for greater flexibility of process design, particularly in cases where the process generates dirt which can damage optical elements. The outline of a second study by Walter Short comparing solar furnaces with CO2 lasers was also circulated. CO2 lasers, which are widely used for surface annealing, have several advantages over solar furnaces. These include high intensity, availability, and control, as well as independence from the weather and the ability to use an existing building. However, CO2 lasers are even more energy inefficient than arc lamps, and thus solar furnaces have an even greater advantage. They can also deliver high power over larger areas and have better absorption characteristics than CO2 laser light. SUGGESTIONS FOR FUTURE RESEARCH AND DEVELOPMENT Future research should concentrate on areas that take advantage of the unique characteristics of concentrated solar photons. The ability to deliver high flux over a large area suggests that a solar furnace could be used for rapid heating and cooling of large areas. The limits on cooling and heating rates for solar furnaces should be explored. The surface absorption of solar light also implies that high gradients can be produced. In other words, the surface can be heated to high temperatures without affecting the bulk. Of course this also leads to high cooling rates, and applications exploring the formation of metastable phases should be explored. In addition solar furnaces can possibly be used to obtain extremely high temperatures (~2000 K). The limits on this ability should be explored and applications which need high temperatures should be sought. Applications in the areas of polymers should be researched. In general, it is felt that there needs to be a mechanism established to find applications. The researchers involved in solar methods need to connect with researchers and technologists working in the application areas. It was suggested that perhaps the solar furnaces could be made available to outside researchers to foster this interaction.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Also needed is more detailed economic analysis on specific applications to assess the practicality of the substantial investment required to fabricate a solar furnace.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy APPLICATIONS OF ENERGETIC PARTICLES IN IMPROVING COATING ADHESION PROPERTIES* George R. Fenske Argonne National Laboratory Argonne, Illinois Introduction In many engineering situations, material selection is often based on a compromise of bulk-mechanical properties and near-surface properties, with neither set of properties at its optimum value. Often the material of choice (e.g., high-Cr steel in corrosion applications, or TiN or WC in wear applications) cannot be fabricated into bulk components due to limitations based on cost, fabrication, and mechanical properties. In such cases, various processes are available to deposit coatings of the desired material on components fabricated from materials with desirable bulk properties. A key property of any coating process is adhesion of the coating to the substrate. Without adequate adhesion, the coating will be lost and thus can no longer protect the substrate. Numerous surface-modification processes can be used to modify the surface properties of a wide range of materials. These processes range from surface heat treatments (e.g., thermal hardening, carburizing, nitriding, carbonitriding, boriding, and metalliding) that rely on thermal processes (primarily diffusion) to produce the desired property in near-surface regions, to surface coating processes (e.g., electro-and electroless chemical deposition, chemical vapor deposition [CVD], physical vapor deposition [PVD], spraying processes, and welding processes) in which material is formed or deposited on the surface. Adhesion is usually not a major concern with surface heat treatments because thermal diffusion produces a gradual change in composition. In surface coating treatments, however, the transition from the bulk material to the coating material is much more abrupt and thus adhesion is a very important factor that must be addressed in selecting the deposition process. This paper addresses one approach (ion-beam-assisted deposition, or IBAD) that utilizes energetic ion beams to enhance the adhesion of metallic films to metallic and ceramic substrates. In one application, IBAD is used to improve the adhesion of silver films to ceramic substrates that are subjected to sliding wear conditions at elevated temperatures(1). In another application, the IBAD process is used to improve the adhesion and modify the microstructure of chromium films deposited on low-Cr steel[2]. Both applications suggest that adhesion can be increased by a number of mechanisms including (a) physical and chemical sputtering of surface contaminants (e.g., hydrocarbons and adsorbed water molecules); (b) preferential sputtering of a particular element of a compound, thus producing a surface enriched in a species that is chemically active with the depositing species; (c) activating chemical states; (d) mechanically toughening the surface, producing more surface area for bonding and sites to arrest surface cracks; and (e) recoil mixing during the initial stage of film deposition. *   Work supported by the U.S. DOE Office of Transportation Materials Tribology Project under Contract W-31-109-ENG-38.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Ion-Beam-Assisted Deposition A description of IBAD processes can be found elsewhere [3–5]. The main feature that separates IBAD from other PVD processes is that the film being deposited is also bombarded by energetic ions/atoms during deposition. This bombardment affects a number of phenomena (film nucleation and growth, film density, film crystallinity, and film orientation) that determine the adhesive strength of the film to the substrate. Figures 1A and 1B show the effect of sputter-cleaning on the adhesive strength of Ag films deposited on alumina and zirconia substrates, respectively[1,6]. The size of the error bars corresponds to one standard deviation in the measured values. In tests using an Ar beam only, the adhesion of Ag to Al2O3 increases with ion dose, rapidly at first and then reaching a steady-state value of approximately 55 MPa after presputtering for approximately 300 s, which corresponds to an ion dose of 5.6 × 1016/cm2. When an Ar + O beam was used, the adhesive strength increased at a faster rate as a function of time (compared to Ar ions only). Presputtering for 3 s was sufficient to raise adhesive strength to approximately 35 MPa. After 30 s of cleaning with Ar + O, the adhesive strength of the Ag to the Al2O3 exceeded the tensile strength of the epoxy bonding agent (approx. 60–70 MPa). In tests that used an Ar + O beam to sputter-clean and bombard the film during deposition[7], adhesion was higher than that for sputter-cleaning only with Ar and O. For Ar sputter-cleaning of ZrO2 (Figure 1B), adhesive strength increased with ion dose (time), rapidly at first and then peaking at approximately 50 MPa after presputtering for 300 s to an ion dose of 7.5 × 1016/cm 2. Beyond 300 s, adhesive strength decreased to 34 MPa after 2000 s (5 × 1017/cm2) and then increased above 65–70 MPa (failure in the epoxy rather than at the Ag/ZrO2 interface) for cleaning times of 3000 s (7.5 × 1017/cm2). When Ar + O was used to sputter-clean the zirconia, adhesion increased very rapidly to values in excess of 65–70 MPa. As seen in Figure 1B, only 30 s (7.5 × 1015/cm2) of cleaning with Ar + O was required to exceed the tensile strength of the epoxy. A number of mechanisms have been proposed to account for the increased adhesion observed in Figures 1A and 1B[8]. These include (a) physical and chemical sputtering of surface contaminants (e.g., hydrocarbons and adsorbed water molecules); (b) preferential sputtering of Al, Zr, or O that produces a surface enriched in a species that is chemically active with the depositing species; (c) activating chemical states; (d) mechanically roughening the surface, which produces more surface area for bonding and sites to arrest surface cracks; and (e) recoil mixing during the initial stage of film deposition. Sputter-cleaning, either by Ar alone or with Ar + O ions, is effective in removing surface contaminants, particularly adsorbed water and hydrocarbons[9]. The substrates used in these tests were cleaned with a series of three organic solvents before insertion into the IBAD system[10]. Thus, part of the improved adhesion observed in Figures 1A and 1B can be associated with the removal of organic residues on the substrate surfaces. The higher rates of increases in adhesive strength with sputter time (or dose) seen with the Ar + O beams relative to those for Ar alone can be attributed to a chemical (or reactive) sputtering process in which volatile compounds (such as CO) may have been formed by the reaction of O ions with organic contaminants. Preferential sputtering of one of the substrate species (such as O), leaving a surface layer enriched (and perhaps chemically active) with Al or Zr, is feasible. Monte Carlo TRIM calculations[11] of physical sputtering of Al2O3 by Ar + O indicate that O is preferentially sputtered. However, impingement of the sputtered surface by residual 02 in the vacuum chamber, particularly during the Ar + O sputter-cleaning (approximately 10-2 Pa of O2) probably negated the preferential sputtering effect, leaving a near-stoichiometric Al2O3 or ZrO2 surface.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Once a clean surface is established, continued bombardment of the surface could produce chemically active O atoms that react with depositing Ag atoms to form a stable compound. Silver is known to form stable oxides (such as Ag2O) at room temperature[12]; thus, it is feasible that the first monolayer of Ag reacts with O atoms to form a ternary Al-OAg compound across the interface similar to that found for Cu deposited on Al2O3[13]. Mechanical roughening due to sputtering is a viable process; however, the amount of material removed by sputtering is only about 40 nm, and measurements[14] of the center-line-average surface roughness before and after sputter-cleaning (from 17 nm before sputtering to 20 nm after sputtering) reveal that this is not a significant factor. Cross-sectional transmission electron microscopy (TEM) of IBAD silver-coated Al2O3 substrates[15], which shows a sharp interface between the deposited Ag and Al2O3 substrate, also indicates that dynamic mixing is not responsible for the improved adhesion of the Ag films tested. For Cr films deposited on metallic substrates[2] the situation is different. Here, the deposited Cr atoms can be mixed into the substrate through a combination of dynamic mixing and radiation-enhanced mixing processes. The radiation-enhanced processes are dominant in this case because one is dealing with diffusion of Cr in a metallic substrate that already contains Cr, rather than with diffusion of Ag in ceramic substrates. Chromium films deposited by pure PVD or IBAD with 100 eV Ar ions exhibited a gap between the film and the substrate (indicative of degraded adhesion), as observed in cross-sectional TEM micrographs. In contrast, Cr films deposited by IBAD with 300 or 1000 eV Ar ions exhibited no gap between the film and the substrate. The 300 and 1000 eV IBAD films also had an intermixed Cr-enriched layer (approximately 30 to 50 nm deep) in the substrate material; this layer can act as a 'glue' to enhance adhesion of the film to the substrate. Another difference between Ag and Cr deposition is the role of internal stresses on film adhesion. Cr films are known to delaminate from the substrate if large internal stresses exist. This situation is not as critical for Ag films because they are pliable and will easily undergo creep to alleviate high stress levels. Differences in the microstructures[2] and porosity of the IBAD films deposited with 100 to 1000 eV Ar ions suggest that internal stresses of the films can be controlled by proper selection of the Cr atom deposition rate, ion flux at the surface, and energy of the incident ions. Discussion The above examples illustrate the effects that energetic ions/atoms can have on the adhesion of coatings to substrates. The major effects observed were the (a) preparation or cleaning of the surface of the substrate prior to deposition, (b) promotion of strong chemical bonding at the interface, and (c) inducing mixing of deposited material into the substrate. Concentrated solar beams provide an alternative approach for obtaining comparable effects via pyro-and photolytic reactions. With pyrolytic approaches, intense solar beams could be used to heat substrates and vapor sources to elevated temperatures. Solar heating of the substrates to elevated temperatures before deposition could be used to desorb common surface contaminants (adsorbed water and organic cleaning compounds), thereby providing a clean or chemically active surface for improved chemical bonding to films deposited in subsequent steps. In conventional PVD processes, where substrate heating is commonly used to improve adhesion and nucleation and growth characteristics[16], this step is typically performed in a vacuum because of the requirements of the vapor sources. With solar heating, however, this step could be performed in an inert atmosphere or, alternatively, in a reactive environment (e.g., halide gases) to chemically assist the formation of volatile compounds from the surface contaminants. Concentrated solar beams could also be used to deposit films. In one approach, solar energy could directly heat and evaporate elements from a crucible. The power densities envisioned for intense solar beam systems (in

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy excess of several kW/cm2) exceed those typically found in electron beam evaporation sources that operate at power levels of 1 to 20 kW with a 5 cm diameter source (0. 05 to 1 kW/cm2) to evaporate pure elements. Alternatively, solar beam heating of the substrates could be used in CVD thin films in a manner similar to conventional CVD of coatings for electrical, decorative, and tribological applications. Such applications of solar beams the beam energy to produce pyrolytic phenomena—processes that require elevated temperatures. In many instances, elevated processing temperatures are undesirable. For example, steel components are often heat treated to achieve a certain hardness and toughness. Exposing these components to temperatures above their annealing point degrades these properties. It is therefore difficult to coat many steel components at elevated temperatures without some type of postdeposition heat treatment (which often distorts the component) to return the component to its original specifications. Consequently, processes that can deposit coatings at low substrate temperatures (such as ion-beam techniques) are often desirable. The energy associated with the incident ions can impart sufficient energy to local regions of the coating and near-surface regions of the substrate to produce effects typically achieved at elevated temperatures. Another example of a low-temperature process—and more germane to this workshop—is the use of energetic photons to photolytically activate processes for both cleaning surfaces and depositing coatings. A process known as photo-enhanced chemical vapor deposition (PHCVD) is receiving considerable attention in semiconductor applications as a low-temperature process for deposition of Si, SiO2, and GaAs[16]. The process is based on photodissociation of gas-phase molecules and typically uses UV light and/or excimer lasers to break the chemical bonds. PHCVD processes are strongly dependent on the absorption of photons with the appropriate wavelength to break the chemical bonds between atoms in the gaseous compound. A majority of the compounds used in PHCVD require UV light to dissociate the molecules, and thus a major obstacle to be overcome in implementing intense solar beams for PHCVD will be to identify compounds that photodissociate (either directly or through a catalytic reaction) at wavelengths that are more abundant in the solar spectrum. Summary Coatings are used in a wide number of applications where the properties of the underlying substrate will not suffice. Concentrated solar beams show potential as a viable energy source for a number of coating processes based on pyrolytic and photolytic reactions. References 1. Fenske, G.R., R.A. Erck, A. Erdemir, V.R. Mori, and F.A. Nichols. 1990. Ion-Beam-Assisted Deposition of Adherent, Lubricious Coatings on Ceramics. Paper presented at 17th Leeds-Lyon Symposium on Tribology (Vehicle Tribology), Leeds, England, Sept. 4–7, 1990. 2. Cheng, C.C., R.A. Erck, and G.R. Fenske. 1989. Microstructural Studies of IAD and PVD Cr Coatings by Cross Section Transmission Electron Microscopy. Mat. Res. Soc. Proc., 140:177. 3. Smidt, F.A. 1990. Use of Ion Beam Assisted Deposition to Modify Microstructure and Properties of Thin Films. Int. Mat Rv. 35 (2):61. 4. Handbook of Ion Beam Processing Technology. Cuomo, J.J., S.M. Rossnagel, and H.R. Kaufman, eds. 1989. Noyes Publication, Park Ridge, N.J.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy 5. Fenske, G.R., A. Erdemir, R.A. Erck, C.C. Cheng, D.E. Busch, R.H. Lee, and F.A. Nichols. 1989. Ion-Assisted Deposition of High-Temperature Lubricious Surfaces. Presented at 35th STLE/ASME Tribology Conference, Ft. Lauderdale, Fla., Oct. 16–19, 1989, STLE preprint 89-TC-2E-2. 6. Erck, R.A., and G.R. Fenske. 1990. Adhesion of Silver Films to Ion-Bombarded Alumina. Mat. Res. Soc. Symp. Proc. 157:85. 7. Erck, R.A., A. Erdemir, and G.R. Fenske. 1990. Effect of Film Adhesion on Tribological Properties of Silver-Coated Alumina. Proceedings of 17th International Conference on Metallurgical Coatings. April 2–6, 1990, San Diego, Calif. 8. Baglin, J. 1989. Interface Structure and Thin Film Adhesion. In Handbook of Ion Beam Processing Technology. Cuomo, J.J., S.M. Rossnagel, and H.R. Kaufman, eds. Noyes Publication, Park Ridge, N.J., p. 279. 9. Beglin, J.E.E. 1985. Adhesion at Metal-Ceramic Interfaces: Ion Beam Enhancement and the Role of Contaminants. Mat. Res. Soc. Symp. Proc. 47:3. 10. Erdemir, A., G.R. Fenske, F.A. F.A. Nichols, and R.A. Erck. 1989. Solid Lubrication of Ceramic Surfaces by IAD-Silver Coatings for Heat Engine Applications. Presented at 35th STLE/ASME Tribology Conference, Ft. Lauderdale, Fla., Oct. 16–19, 1989, STLE preprint 89-TC-2E-1. 11. TRIM-89. The TRansport of Ions in Matter. Computer program provided courtesy of J.F. Ziegler, IBM Research, Yorktown, N.Y. 12. CRC Handbook of Chemistry and Physics, 60th ed., CRC Press, Boca Raton, Fla. 13. Schrott, A.G., R.D. Thompson, and K.N. Tu. 1986. Interaction of Copper with Single Crystal Sapphire. Mat. Res. Soc. Symp. Proc. 60:331. 14. Erck, R.A., and G.R. Fenske. 1990. Adhesion of Silver Films to Ion-Bombarded Zirconia, to be presented at STLE/ASME Tribology Conference, Oct. 7–10, 1990, Toronto, Ontario. 15. Erdemir, A., G.R. Fenske, R.A. Erck, and C.C. Cheng. 1990. Ion-Assisted Deposition of Silver Films on Ceramics for Friction and Wear Control. Lubr. Eng. 46:23. 16. Handbook of Thin-Film Deposition Processes and Techniques. 1988. K.K. Schugraf, ed. Noyes Publication, Park Ridge, N.J.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Figure 1 Adhesive strength of silver films deposited on (A) alumina and (B) zirconia as a function of sputter-cleaning time(0.04 mA/cm 2). Reprinted with Permission from: Argonne National Laboratory

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy power of near 50 kW with a steady state power of 38 kW. This steady state power was chosen because it is in the same range as the 31 kW Vortek arc lamp and it was identified as the optimum that can be produced with a single large heliostat. To be comparable to the high quality of the arc lamp beam, we assumed that the flux level and shape of the beam from the furnace must remain relatively constant while in operation.1 This can be accomplished by adjusting the attenuator to maintain a constant power and by keeping the concentrator completely covered with light from the heliostat. If the concentrator is not completely covered, a skewed flux pattern could occur on the target. Because of these restrictions, a fraction of the solar power reflected early and late in the day is not usable because the power is too low or because the angle of the sun relative to the heliostat causes a heliostat image that is too small. Given the constraint of a high quality beam, we calculate that the commercial scale furnace would be able to operate an average of 6 hours per day over the course of a year. Definition of Case Studies We will compare the economics associated with materials processing tasks that require flux levels of either 1000 or 5000X with a total power of between 20 and 40 kW. An example of a task conducted at 1000X is rapid thermal annealing of semiconductors. We will also investigate assembly-line vs. batch processing tasks. In an assembly-line application, the materials processing task performed by the lamp or furnace is only one task of an assembly line that operates continuously. For example, initial tasks in an assembly line could be to manufacture a metal part. This part would travel down the line and subsequently undergo surface treatment by the arc lamp or furnace. The assembly line is assumed to operate 24 hours a day when using a lamp and during the daytime when using the furnace. An operating crew is employed to only perform actions related to the assembly line. The crew size is assumed to vary depending on the complexity of the assembly line process. If the lamp or furnace is unavailable due to equipment failures, or if the furnace is unavailable due to bad weather, the crew does not perform other tasks. When comparing the economics of arc lamps and furnaces within an assembly line, we included costs related to the size of the crew because one technology may have an advantage related to optimum utilization of crew time. For example, in a poor solar region the arc lamp should beat the economics of a furnace because the crew would waste time waiting for sunny conditions.2 In a batch processing application, the lamp or furnace would only be operated a fraction of the work day. In this analysis we investigate situations in which they are operated an average of 2, 4, or 6 hours per day. When comparing the economics of arc lamps and furnaces within a batch process, we did not include costs related to the size of the crew because they can perform other constructive tasks when the lamps and furnaces are unavailable. It should be noted that most of the lamps Vortek has sold are used in a batch mode. 1   This is a conservative assumption since certain materials processing tasks may be able to use variable power levels and flux shapes. 2   In a poor solar region the arc lamp would also beat the furnace because capital equipment located in the assembly line would be underutilized. As a first cut we have not included this effect and have implicitly assumed that the cost associated with the crew is more important than the capital expense.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Economic Models We compared the levelized energy cost (LEC) of delivering power to the target. Assuming that inflation and escalation do not occur during the life of the equipment (constant dollar analysis), the LEC can be calculated with the following equation: The annualized capital cost is the fraction of the installed capital costs that must be paid every year during the life of the equipment (assumed to be 30 years) to repay the principal and interest on the loan and other smaller items. This fraction is reduced to include depreciation allowances. It is calculated by multiplying the installed capital costs by a fixed charge rate. In this analysis we used a typical fixed charge rate of 10.5% [8]. The annual operating and maintenance (O&M) costs include items such as parts replacement, electricity to run the equipment, and general repair and maintenance. The annual energy is the summation of the optical power delivered to the target over the course of a year and is expressed in kilowatt-hours. In Table 1 and 2 we compare estimates for installed capital costs, O&M costs, annual energy, and LEC for the arc lamps and solar furnace for a batch mode operation in which they are operated 6 hours per day. Blank entries in the tables imply that the cost and performance item is identical to that listed in the column to the left. In Table 1 arc lamp costs are given for the 1000X and 5000X systems by assuming two different prices of electric power. Since the arc lamp systems use a significant amount of electric power, this parameter was varied to gauge its effect on LEC. The cheap electric power case assumes $0.05/kWh and the expensive, $0.094. The former case is typical in many parts of the United States, and the latter is what is charged for power in Albuquerque, New Mexico, home of Sandia National Laboratories and an excellent solar region. The anode and cathode within the lamp degrade after a few hundred hours of operation and must be replaced. In addition, we have assumed that the lamp and reflector must be changed once per lamp-year (8760 hours of operation). Since the target is only a few centimeters away from the lamp and reflector, it is plausible that a sputtering materials processing task or some other accident could damage these items. The costs related to the cooling tower and water are needed to cool the lamp, as described previously. The arc lamp system is also expected to have an electric hookup charge, beyond what would be needed by the furnace, because of the current and voltage ratings of the equipment. It can be seen in Table 1 that each arc lamp reflector delivers different amounts of optical power to the target; the 1000X model furnishes nearly 31 kW and is more efficient than the 5000X model which supplies 19.2 kW. This performance penalty associated with increased concentration is not expected with the solar furnace system. In Table 2 costs of solar furnaces are given for two different assumptions regarding the capital costs of the heliostat and concentrator. For the expensive case, the prices were estimated based on Sandia's experience with building our two experimental solar furnaces. These costs are high because they were specially built for us and we could not benefit from cost reductions due to mass production. For the cheap case, the prices were estimated by assuming a mass production scenario of approximately 2500 heliostats and 2500 concentrators per year. Such a scenario is predicted to

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy occur when solar central receive and Stirling dish power plants are built in the Southwest[8].3 Case Study Results Figure 3 displays the results for batch processing of materials. The LEC is reduced as operating hours per day increase because the systems deliver more annual energy to the target, given a fixed capital cost; i.e., the systems are utilized more fully. We have limited the chart to 6 hours per day because we calculate the commercial scale furnace could operate up to 6 hours a day, on the average, over the course of a year in the U.S. Southwest. It can be seen that the economics of the furnace and the 1000X lamp are similar for batch operations less than 4 hours per day. Beyond 4 hours the furnace has the advantage. It is also seen that the furnace beats the economics of the 5000X lamp for every case. Depending on the combination of assumptions depicted in the Figure 3, the LEC from the furnace is from 25 to 75% less than the 5000X arc lamp. The furnace has a greater advantage over the 5000X lamp because, as stated earlier, the 5000X reflector is less efficient than the 1000X reflector. Finally, it should be noted that if Figure 3 is modified by replacing LEC with the present value of the capital and annual O&M costs, given a 30 year life and a 6.5% discount rate, the conclusions do not change. Figure 4 displays the results for continuous processing of materials. These LEC curves are dominated by the wages paid to the crew members (the loaded salary was assumed to be $7000/man-years). The 1000X arc lamp is seen to take a significant advantage as the crew size associated with the process train becomes larger. The primary reason the furnace is not as good as the 1000X lamp is because the crew is not being used optimally. The process train and the crew are idle during cloudy periods, as well as during early morning and late afternoon when the insolation is poor. Operation of the lamp, on the other hand, is not affected by the weather. It can also be seen that the economics of the 5000X lamp is similar to the furnace. The 5000X lamp lost the advantage the 1000X lamp had over the furnace because of the efficiency drop described previously. Summary and Conclusions The arc lamp and solar furnace each have their own set of advantages. The advantages of each are summarized below. Arc Lamp Advantages Arc lamps are more convenient since they are not dependent on the weather. The materials processing business can precisely schedule their use. The materials processing industry does not have to be located in a sunny region such as the Southwest. Because of better utilization of the operating crew, the arc lamp is recommended for materials processing tasks located within a continuous process. The capital cost of an arc lamp is lower than that of a solar furnace. The arc lamp is much smaller than a solar furnace and would be easier to implement within an existing building. 3   We have increased the costs of the heliostat and concentrator beyond what is quoted in these studies because solar furnaces are small relative to power plants and do not enjoy the same economies of scale.

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Solar Furnace Advantages Since the focal point is much further from the reflector, the furnace will be able to operate in conjunction with dirtier materials processing tasks. The economics of the furnace is not dependent on the price of electricity. For Southwest locations, the furnace is significantly more economical than the lamp for batch processing tasks requiring concentrations near 5000X or greater. Operation of the furnace does not require cooling water. The solar furnace has lower operating and maintenance costs. The solar furnace will have less of an environmental impact than the are lamp because it uses an insignificant amount of electricity. In closing, it should be noted that for applications requiring significant exposure times at very high concentrations (< 10,000X), the solar furnace should significantly beat the economics of the are lamp. Currently, the lifetimes of arc lamp components are short at such high-intensity levels. Acknowledgements I would like to thank Jim Pacheco of Sandia for developing the heliostat image equations used in the optical calculation for the solar furnace. I also appreciate the efforts of Gary Albach of Vortek Industries Limited. He provided cost and performance information regarding the arc lamp and performed a detailed technical review of the analysis. References 1. Tan, R.K., H.L. Arrison, D.A. Parfeniuk, and D.M. Camm. 1988. Surface Treatment Using Powerful White Light Sources. Vortek Industries, Ltd., Vancouver, B.C. (October). 2. Pitts, J.R., J.T. Stanley, and C.L. Fields. 1990. Solar Induced Surface Transformation of Materials (SISTM). Solar Thermal Technology-Research, Development, and Applications. Proceedings of the Fourth International Symposium, B.M. Gupta, ed. Hemisphere Publishing Corporation. 3. Edgar, R.M., and J.T. Holmes. 1982. The Horizontal-Axis Solar Furnace at the Central Receiver Test Facility. High Temperature Technology . (November). 4. Cameron, C.P. 1991. 60 kW Horizontal-Axis Solar Furnace. (to be published). 5. Alpert, D.J., and R.M. Houser. 1989. Evaluation of the Optical Performance of a Prototype Stretched-Membrane Mirror Module for Solar Central Receivers. ASME Journal of Solar Energy Engineering. 111:37-43. (February). 6. Solar Kinetics, Inc. 1978. Development of a Stretched-Membrane Dish. SAND88-7031. (October). 7. Randall, C.M. 1978. Barstow Insolation and Meteorological Data Base. Aerospace Report No. ATR-78 (7695-05)-2. The Aerospace Corporation, El Segundo Calif. (March).

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy 8. Pacific Gas and Electric Company. 1988. Solar Central Receiver Technology Advancement for Electric Utility Applications, Phase 1 Topical Report, Report No. 007.2–88.2. San Francisco Calif. (September).

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Table 1 Cost/Performance of Arc Lamps Operated 6 Hours per Day   1000X Reflector 5000X Reflector   Cheap Electricity Expensive Electricity Cheap Electricity Expensive Electricity Capital cost ($)         Lamp system 136,500   147,000   Cooling tower 6,000   6,000   Electric hookup 5,500   5,500   Building 50,000   50,000   Total capital cost 198,000   208,500   Annual operating cost ($)         Electricity 11,700 22,000 11,700 22,000 Water 140   140   Lamp replacement 70   70   Reflector replacement 6,600   9,200   Argon gas 120   120   Anode replacement 9,500   9,500   Cathode replacement 3,300   3,300   General maintenance 5,000   5,000   Total O&M cost 36,400 46,700 39,000 49,300 Annual performance         Annual operating hours 2,190   2,190   Power to target (kW) 30.8   19.2   Annual energy to target (kWh) 67,450   42,100   Levelixed Energy Cost         ($/kWh) 0.86 1.01 1.47 1.72

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Table 2 Cost/Performance of Solar Furnaces operated 6 Hour per Day   Cheap Capital Cost Expensive Capital Cost Capital cost ($)     Building 130,000   Attenuator 50,000   Heliostat 22,000 ($150/m2) 100,000 ($670/m2) Concentrator 16,000 ($230/m2) 50,000 ($1700/m2) Test table 25,000   Computer 30,000   Clock 2,000   Installation 28,000 46,000 Total capital cost 303,000 503,000 Annual operating cost ($)     Electricity <100   General maintenance 5,000   Total O&M cost 5,100   Annual performance     Annual operating hours 2,190   Power to target (kW) 38   Annual energy to target ($/kWh) 83,000   Levelized energy cost (S/kWh) 0.44 0.70

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Figure 1 Arc lamp System Manufactured by Vortex Industries, Ltd. Reprinted with Permission from: Sandia National Laboratories

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Figure 2 The horizontal-axis solar furnace at the National Solar Thermal Test Facility. Reprinted with Permission from: Sandia National Laboratories

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Figure 3 Comparison of LECs for batch processing of materials. Reprinted with Permission from: Sandia National Laboratories

OCR for page 59
Proceedings of a Workshop: Potential Applications of Concentrated Solar Energy Figure 4 Comparison of LECs for assembly-line processing of materials. Reprinted with Permission from: Sandia National Laboratories