MICROENCAPSULATED POWDER APPROACH FOR ADVANCED MATERIALS Animesh Bose, Andrew J. Sherman, Asit Biswas Powdermet, Inc., Sun Valley, California ABSTRACT Processing two-phase composite materials using the powder metallurgy approach is rapidly becoming an important fabrication route for advanced materials production. Processing such composites using conventional powder metallurgy methods generally relies on the blended powder approach. Microencapsulated powders provide an exceptionally flexible technology, whereby almost any powder can be coated with different material(s) to produce unique composite powders with exceptional property combinations. The coated powder approach gives the powder metallurgy industry an unique raw materials pool that can be used for processing numerous advanced materials. This paper discusses the benefits of the microencapsulated powder approach obtained through the chemical vapor deposition (CVD) process, and illustrates applications for such powders. INTRODUCTION The search for new and advanced materials capable of meeting expected improved performance demands has been prompted by the rapid pace of technological development. Particulate materials are increasingly used in the materials arena. Particulate materials enjoy wide use in advanced materials processing, per their great flexibility. Processing controlled porosity materials, the ability to use two different materials insoluble in one another, dispersion of relatively high volume fraction of reinforcing phase, significantly lowering processing temperatures for high temperature material consolidation, and tailoring composite materials properties are some advantages unique to particulate materials. Some classic P/M products include porous self-lubricating bronze bearings, tungsten carbide-cobalt based cutting tools, metal and alloy foams, and composites such as tungsten heavy alloys, Mo-Cu, W-Cu, and TiC-Fe. Though P/M processing has come a long way since its early days, several problem areas exist in which using innovative technology could provide solutions. Microencapsulated powders represent one technological innovation that can be used by the P/M industry in a variety of applications. Using encapsulated powders (which eliminate the need to mill powder components together) could significantly benefit one of the earliest P/M applications – WC-Co based cutting tools. Similarly, the process for producing metal bonded diamonds could be significantly improved by using coated diamond powders. In several other cases (including tungsten heavy alloys) using the microencapsulated powder approach would allow a low contiguity alloy microstructure to be obtained, which in turn would translate into improved properties. There are numerous examples of the advantages and improvements made possible using P/M processed materials created through the microencapsulated powder approach. Coated or microencapsulated powders provide the P/M industry with a new powder source that will prove highly useful in several industry segments. Microencapsulated powder is a particulate material in which, if properly processed, each powder particle is coated with a layer of different material. Microencapsulated powders become especially advantageous when the minor phase volume fraction is extremely low compared to that of the major phase. In W-10Cu based composite powder, the conventional approach would entail mixing or milling elemental tungsten and copper powders, followed by cold pressing and liquid phase sintering. In comparison, the microencapsulated powder approach would result in a thin copper coating layer being applied to the tungsten powder (coating thickness depends on the copper weight fraction to be used). Thus, the microencapsulated powder approach creates a composite powder where the copper no longer must be mixed or milled separately with the tungsten powder, almost eliminating segregation and inhomogeneous copper distribution. MICROENCAPSULATED POWDERS As previously discussed, using microencapsulated powders can provide the P/M industry with numerous advantages. It is possible to obtain improved materials and processes if fine powders are produced with controlled surface properties, or if multiphase powders are produced with controlled phase distribution, such as through CVD coating. The CVD process gives powder producers the flexibility to coat a large number of materials on widely different material substrates. Extremely high chemical purity is attainable using the CVD process. One major consideration when coating particulate based materials using CVD is the particulate material’s ability to fluidize. Fluidized beds are the ideal tool for using CVD to produce or modify particulate materials in an economical manner. Unfortunately the most desirable powders for producing advanced engineering materials are very fine; typically below 25 um. In fact, most powders of interest in advanced materials usually range from 0.5 to 15um and often have large aspect ratios. These materials, which fall into Geldart’s Class C category of particulate material [1], are very cohesive and difficult to handle, fluidize, and process in a non-agglomerated form. Some fine powders can be fluidized by agglomerate formation; however uniformly coating most particles using this process becomes extremely difficult. This problem becomes acute when coatings are applied at elevated temperatures, as undesirable hard agglomerates are often formed [2]. The fluidized bed is a highly efficient solid-fluid contacting device well suited to coating particles and enabling more precise microstructural and compositional control over materials. This technique was originally adopted for coating large particles such as uranium fuel kernels. In conventional fluidized bed approaches, the fines concentration results in considerable elutriation. A substantial percentage of material is therefore lost through entrainment, and carryover. Earlier attempts to fluidize fine powders relied on particle agglomeration, which allowed the powders to fluidize yet resulted in substantial elutriation [3]. Thus, conventional fluidized bed CVD coating reactors are primarily suited for larger particles that do not have the tendency to form agglomerates. However for materials of the Geldart’s Class C type, a different type of fluidized bed reactor becomes necessary. Surface phenomena tend to dominate fine powders, primarily due to their high surface-to-volume ratio. In such cases van der Waals, electrostatic, and surface tension forces often have dominating effects on properties. Handling these particles is extremely difficult, and high aspect ratios (as in fine whiskers) further complicate the handling problem. One can imagine the difficulty encountered attempting to modify these particles by encapsulation with a different material through the CVD approach. One way to overcome this problem is to fluidize these fine particles in turbulent and fast-transport regimes, where the high gas shear forces and massive turbulence greatly nullify cohesive effects on fine particulates. Another advantage of operating in turbulent and fast-transport regimes is that they allow high gas shear and particle collision forces to continually break up agglomerates as they form. At this point it should be realized that the high superficial gas velocity necessary to attain the fast-transportation regime would also lead to considerable entrainment and elutriation. If sufficient freeboard height is allowed (the column height above the bed surface), larger particles and agglomerates are returned to the fluidizing bed. In a turbulent fluidized bed, the bed upper surface is still present. The turbulent regime then extends to the transport velocity, where there is a sharp increase in particle carryover rate. Typical fluidized bed solid loading in this operating region is well below five percent. Only recycling the entrained solids allows solid loading to be increased above transport velocity and moves the operation into the fast-transport regime. Above the transport velocity solids fed to the bottom of the column traverse the column in fully entrained transport flow, and the concentration of the resulting suspension depends on the gas velocity and on the solid flow rate. Dilute-phase flow will result with low solid flow rates. However if the solid flow rate is kept sufficiently high by circulating solids carried over from the column through external cyclones, filters, and a standpipe, it is possible to maintain the column at relatively large solid loading fractions. Adding a cyclone separator at the top of the column allows most of the particles to be recirculated or returned to the bed. Recirculation efficiency (E) is defined by the mass of powder recirculated per pass divided by total powder mass. By using cyclones, recirculation efficiencies greater than 95% have been achieved, and adding pulse-cleaned filters results in greater than 99% recirculation efficiencies. It has been determined that operating a fluidized bed in the fast-transport operating regime (above transport velocity), or in the turbulent fluidization regime allows fine particles including fine powders, whiskers, chopped fibers, etc. to be fluidized with high product yields [4]. A Recirculating Fast-Fluidized Bed (RFFB) pilot plant reactor is designed, constructed, and put into operation to achieve this.  The RFFB reactor is used to develop CVD coatings on fine powders, which will provide improved starting materials for a variety of P/M industries. The CVD technique for microencapsulating powders relies upon chemical reactions to produce the desired material coating. It is possible to utilize different chemical precursors to coat the same material (for example, nickel). Chemical precursor choice is predicated on factors including ease of precursor introduction into the reaction chamber, by-product(s) of the chemical reaction, safety of by-product removal, possibilities for contamination, and finally precursor cost and availability. Sometimes pre-treating the powder surface with a different material is necessary prior to applying the actual coating. It therefore becomes imperative that the specialty powder producer have good knowledge of precursors available for depositing a particular material, of the ensuing chemical reactions, and of the suitability of the precursor to the process. As previously stated, numerous chemical precursors are available for most materials: for example, aluminum precursors include triethylaluminum (TEA), trimethylaluminum (TMA), dimethylaluminum hydride, diethylaluminum chloride, dimethylethylamine alane, and triisobutal aluminum (TIBAL). There also exists one class of chemical compounds that can be used to deposit different metals. For example, metal carbonyls can be used as chemical precursors for coating different metals including commercially important ones such as nickel, iron, and cobalt. This process can result in pure coatings for certain metals. A typical chemical reaction for coating iron is given below: Fe(CO)5=Fe + 5CO In this process iron carbonyl at the proper reaction temperature can disassociate into iron and carbon monoxide; the latter is subsequently removed from the reaction chamber. Several coatings can become contaminated with carbides, oxides, and even graphite (carbon) from carbon monoxide disassociation and subsequent adsorption of the products into the coating. Microencapsulated particulate materials technology is expected to find applications in numerous areas including the thermal spray, syntactic foam, cutting tool and wear parts, thermal management, superabrasives, electronics, and ordnance industries. A further application is in metal matrix composites for industries including aerospace and sporting goods. Key improvements are made possible using such powders, such as: Compositions with tailored grain boundaries Good homogeneity and extremely low segregation, even in multiphase powders where the concentration of one phase is very low Very low contiguity in composites such as tungsten heavy alloys and WC-Co Allowing net shape processing techniques to be used (such as powder injection molding) for multiphase materials, without the fear of segregation Allowing the possibility of using higher volume fraction reinforcing phase in metal matrix composites Reduced processing steps Reduced handling and safety concerns Producing a consistent, uniform, and reliable product

This technology opens up new horizons in the particulate materials industry.

Tungsten heavy alloys illustrate the advantages of microencapsulated powders. A brief background of this alloy system follows.

Tungsten heavy alloys generally are two-phase composites consisting of nearly pure bcc tungsten grains embedded in a soft and ductile fcc matrix alloy [5,6]. Tungsten content in conventional heavy alloys varies from 90 to 98 weight percent. Commonly, the remaining alloy constituent contains nickel, iron, cobalt, and copper. (Usually a minimum of two elements are used. For example, Fe:Ni, or Ni:Cu.) Nickel-iron is the most popular additive, in a ratio of 7Ni:3Fe or 8Ni:2Fe (weight ratio). The conventional processing route for tungsten heavy alloys includes mixing the desired amount of elemental powders, followed by cold pressing and liquid phase sintering to almost full density. The matrix alloy melts and takes some tungsten into solution during liquid phase processing, resulting in a microstructure through which large tungsten grains (20–60um) are dispersed in the matrix alloy. The as-sintered material often is subjected to thermomechanical processing by swaging and aging, which results in increased strength and hardness in the heavy alloys.

Conventional heavy tungsten alloys exhibit an unique property combination. Properly processed materials show a combination of high density, high strength, high ductility, good corrosion resistance, high radiation adsorption capability, and reasonably high toughness. This property combination has made this alloy a candidate for defense and civilian applications. Some of these applications include radiation shields, counter weights, kinetic energy penetrators, vibration dampening devices, medical devices for radioactive isotope containment, heavy duty electrical contact materials, balancing crankshafts for race car engines, and gyroscopes.

Early research and development of this alloy system focused on improving tensile ductility, and subsequently concentrated on developing higher toughness heavy alloys. During this phase of work, considerable effort was directed at understanding compositional, processing, and microstructural effects on heavy alloy properties. Microstructure was found to play a key role in determining alloy properties, with tungsten-tungsten contacts discovered to be one of the microstructure’s weakest links (as measured by contiguity). In the conventionally processed alloy, ductility drops off sharply with increasing tungsten content (especially above 95wt% tungsten), with a concomitant increase in contiguity. The coated powder approach - wherein tungsten powder is coated with nickel and iron rather than being mixed with nickel and iron powder – allows for these alloys to be processed with very low contiguity (very little direct tungsten-tungsten contact). Thus, the material can be processed to near full density in the solid state with the resultant microstructure exhibiting very low contiguity. Initial properties obtained in such alloy systems show that good property combination is attainable by using microencapsulated powder as the starting material. A major use for tungsten heavy alloys is in kinetic energy penetrators, where they are in direct competition with depleted uranium (DU). Recent investigations conducted at the Army Research Laboratory show that DU’s superior properties resulted from its ability to localize shear during ballistic penetration events. It was therefore argued that if localized shear can be imparted to tungsten heavy alloys, these alloys would exhibit penetration performance matching that of depleted uranium (which had become an environmental problem).

Localized shear has been reported in materials that have low strain hardening rate, low strain rate sensitivity, low thermal conductivity, and a high thermal softening rate. Hafnium and titanium have some material characteristics that make them ideal candidates for matrix materials in heavy tungsten alloys. Of these, hafnium is the definite material of choice, due to its high density (compared to titanium). However, fine hafnium powder that can be mixed with elemental tungsten powder is usually very difficult to obtain. Furthermore, alloy processing must be conducted in the solid state, yet processing this alloy in the solid state leads to very high contiguity. Investigations done on a W-Hf system show that improper size distribution results in tungsten surrounding the hafnium instead of the other way around (the desired structure) [7]. It is postulated that the above microstructural problems can be eliminated using the microencapsulated powder approach. This leads to the possibility of fabricating tungsten heavy alloys in the solid state, with improve property combinations.

Microencapsulated powders are expected to have applications in powder injection molding (PIM), a near net-shape process. PIM is an unique processing technique capable of producing complex shapes from numerous high performance materials [8,9]. This relatively new technology has come of age over the last quarter century. The process provides the shaping advantages of traditional plastic injection molding, but expands applications to many advanced materials such as metals, alloys, ceramics, cermets, intermetallic compounds, and composites [8,10].

The generic PIM process consists of mixing a small amount of organic binder material (typically polymers, waxes, oils, etc.) with the desired inorganic powder (metal, alloy, intermetallic compounds, composites, etc.) The powder and organic binder mixture is cooled and granulated or pelletized, resulting in small, discreet pieces that can be fed into the hopper of an injection molding machine. This is known as the PIM feedstock. The feedstock is introduced into the injection molding machine’s hopper, where it is re-melted in the barrel and pushed into an oversized die cavity. This molded shape (commonly referred to as the green part) is usually an oversized replica of the final product. Generally, the green part is subjected to debinding – removal of the part’s organic portion. Debinding is used to gradually and completely remove the organic phase without leaving residual contamination. After the binder is removed, the part is subjected to a thermal treatment known as sintering. This results in part densification to the desired level. The PIM process, therefore, provides designers and engineers with a powerful technique capable of shaping materials like plastics, while not confining them to plain organic polymers (thermoplastic or thermosetting type). Though the process described above is quite generic, there are many variations on the process.

The PIM process offers tremendous advantages over other processes when manufacturing complex shaped parts and parts made from expensive or difficult to machine materials. Often high temperature materials such as ceramics or refractory metals and alloys are powder injection molded to yield complex shaped parts. Advanced composite materials are normally expensive and difficult to machine; therefore machining them into complex shapes is not considered a practical manufacturing method. Powder injection molding is one of the only shaping processes available for fabricating complex shaped components from such materials. Some advanced composite materials ideally suited for the PIM process are:

	AlN-Al composites
	W-Cu and Mo-Cu
	Tungsten heavy alloys
	WC-Co or TiC-Fe-Ni based hardmetals

Availability of suitable powders presents a problem when processing these materials using the PIM route.

Interestingly, large powder producers have been reluctant to enter the PIM market, as PIM powders are used in small quantities and have diverse and often conflicting property requirements. Characteristics of an ideal PIM powder include:

Almost spherical shape with a slight aspect ratio, as this allows better packing and flow

	Fine powder particle size (around 5m )
	High tap density, preferably greater than 40% of theoretical
	Compacted angle of repose between 50 and 60°
	Powder should not react with the binder
	Powder should be free of internal voids
	Powders should have low agglomeration tendency
	Mixed or milled powders should not extensively segregate

A homogeneous feedstock is extremely important to the PIM process. When dealing with mixed powders, it becomes important to homogeneously disperse the different phases in the feedstock. Powder separation problems are accentuated when two different powders of varying sizes are being used (where the smaller powders do not fit into the interstitial gaps and increase packing density), when the powders have very wide density differences, and when the volume fraction of one powder is small compared to the other. In all of these cases, coated powder use is expected to provide significant advantages, as they have the ability to uniformly coat the minor phase onto the major phase.

One alloy system that could benefit from using microencapsulated powders is the popular PIM alloy nickel-steel (Fe-2Ni, Fe-2Ni-0.5Mo, Fe-7Ni). These alloys are extensively used as structural components in the as-sintered, carburized, or carbonitrided form, in a number of industries such as household goods, gun components, and business equipment. The material is well suited for use as structural components, except where there are compelling reasons for almost eliminating residual austenite phase in the microstructure. Using nickel – an austenite stabilizer – is a problem if proper homogenization does not occur during sintering. During low temperature exposure, retained austenite undergoes transformation and creates stresses within the structure, resulting in premature part failure.

Retained austenite in a martensitic microstructure is especially undesirable in carbonitrided parts used in close-fitting assemblies. In these cases, the retained austenite not only lowers hardness, but also could actually bind or "seize up" during operation due to the volume increase during transformation. It is possible to reduce retained austenite by using subzero treatments, but this could lead to microcracking on the case structure, especially if the grain size is coarse.

The conventional route of blending elemental iron and nickel powders is apt to produce areas where the nickel has not been completely homogenized. This could lead to a chance that retained austenite could be formed after carbonitriding. If however nickel is deposited as a coated layer onto the iron powder surface, uniform nickel distribution on each powder particle will significantly reduce the diffusion distances and therefore homogenization time.

Consequently coated powders could provide significant advantages in PIM nickel-steels processing, in spite of the powders’ slightly higher cost.

Ferrous-based powders coated with less than 1wt% boron provide another interesting area wherein microencapsulated powders could benefit the PIM industry. It has been demonstrated that an excellent combination of high mechanical properties and good erosion resistance can be obtained in austenitic stainless steel with a boron content of around 0.8wt% [11], and that boron is a potent additive to tool steels [12]. Boron is added as either the elemental form or as a compound. Boron addition to iron-based materials would result in forming an Fe2B intermetallic, which in turn forms an iron-rich eutectic at about 1175° C [13]. Uniform boron distribution would difficult to obtain using the conventional PIM process, as an exceedingly low amount of boron is required. Conceptually, if a small amount of boron can be coated onto high value added iron-based alloy powder (such as austenitic stainless steels) uniform boron segregation at particle boundaries will result in rapid material densification. In such a case it may be possible to use coarser, less expensive water atomized 316L stainless steel powder, and still obtain the high density desired in PIM processed samples.

Other material systems that could benefit from the coated powder approach include W-Cu or Mo-Cu-based thermal management materials. Typically, copper is a minor phase and can be coated onto tungsten particles, creating an easily sintered composite powder. Hardmetals and other composites also can be easily processed using the PIM route, again taking advantage of microencapsulated powders.

It should be noted that the advantages afforded the PIM industry by the coated powder approach extend beyond the ability of this process to provide uniform dispersion of various phases. The CVD coating process results in a subtle change to the powders’ surface properties, which can be used to some advantage in the PIM process. The uncoated powder has a smooth surface and would generally be unsuitable for PIM applications. The coated powder particle, though still spherical in nature, has significantly higher surface roughness. This enhanced powder surface roughness will increase interparticle friction, allowing better shape retention [14]. This allows the twofold advantage of being able to incorporate a grain boundary phase while using the coarse spherical powder, which is far less expensive than its finer counterpart.

In conclusion, the coated powder approach can benefit the PIM process through its ability to produce powders with practically no segregation between constituents, and through subtle shape modifications that result in the possibility of using coarser and less expensive powders. As other P/M industries embrace this process, it is expected that the PIM industry will also follow.

The two above examples mainly deal with coating fine particulate materials. However, the CVD coating process also is suitable for producing coatings on coarse particulate materials. This section discuss another potential application for such microencapsulated powders.

In the aerospace industry, structural efficiency – namely stiffness-to-weight ratio – is the dominant design criterion for fabricating many structural elements including aircraft wing, fuselage, and engine nozzle structures. Using a well designed skin/stringer concept can improve structural efficiencies by a factor of approximately 25 over solid material designs, while using sandwich panels offers improvements of up to several hundredfold. A structural sandwich consists of two thin, stiff, strong facings separated by a lightweight core, and produces extremely stiff and strong structures that maximize the material’s efficiency [15].

Using syntactic foams instead of a honeycomb structure provides several advantages. The most outstanding characteristics of syntactic foam are its resistance to hydrostatic pressure, and its high specific shear strength and modulus. Spherical inclusions such as microspheres also show lower stress concentrations than alternate microstructures, and therefore should lead to improved fatigue and impact resistance relative to honeycomb, fiber reinforced materials, or foams with irregular cell geometry. Thus, hollow ceramic microspheres coated with titanium (and its alloy) have been chosen for developing syntactic foam material.

The current process under development consists of encapsulating hollow ceramic microspheres with lightmetals such as aluminum or titanium. Aluminum would be deposited using organometallic chemical vapor deposition (OMCVD) onto high strength, low cost silica microspheres in a fluidized bed deposition system. Titanium can be deposited on the microspheres using hydrogen reduction of titanium bromide in a high temperature fluidized bed deposition system. The hollow spheres would be coated with 20-40% titanium by volume. These coated microspheres would be used as starting material for processing sandwich panel structures. Some key difficulties anticipated during processing include: potential for impurity dissolution in the titanium from the ceramic microspheres, and the ability to keep the microspheres intact during processing. This approach has much promise and its success will revolutionize the way in which lightweight foam structures are processed.

Microencapsulated particulate materials provide the powder metallurgy and particulate material industries with a new raw material source that can provide numerous advantages in advanced materials processing. Using microencapsulated powders is quite generic and applicable to several completely diverse industries such as ordnance, aerospace, electronics, electrical, cutting tools, superabrasives, and more. It is expected that coated materials use will continue to expand as the materials community becomes aware of their tremendous potential.

1.  A.R. Abrahamson and D. Geldart "Behavior of Gas-Fluidized Beds of Fine Powders," Part  Homogeneous Expansion," Powder Tech.,Vol. 26, 1980.

2.    D. Kunii and O. Levenspiel Fluidization Engineering, John Wiley & Sons, New York, New  York, 1966.

3.   A.A. Avidan and J. Yerushalmi "Bed Expansion in High-Velocity Fluidization," Powder Tech., Vol. 32, 1982.

4. A.J. Sherman and V.M. Arrieta, 1994 "Fluidized Bed Surface Modification of Fine            Powders," Final Report (ULT/TR-94-6677), Grant III-9362136, National Science Foundation, Washington, DC.

5. N.M. Parikh, Report ARF 2182-12, WAL 372/32, March 23, 1961 Armor Research Foundation Technical Report, Watertown Arsenal, Watertown, MA.

6. B.C. Muddle "Interphase Boundary Precipitation in Liquid Phase Sintered W-Ni-Fe and W-Ni-Cu Alloys," Metall. Trans., Vol. 15A, 1984, pp. 1089-1098.

7. A. Bose, H.R.A. Couque, and J. Lankford, Jr. "Development and Properties of New Tungsten-Based Composites for Penetrators," Inter. J. Powder Metall., Vol. 28, No. 4,1992, pp. 383-394.

8.  Randall M. German and Animesh Bose, Injection Molding of Metals and Ceramics, Metal Powder Industries Federation, Princeton, NJ, 1997.

9.  Beebhas C. Mutsuddy and Renee G. Ford, Ceramic Injection Molding, Chapman & Hall, London, UK, 1995.

10. A. Bose "Some Critical Aspects of Commercial Powder Injection Molding Process, "Proceeding of the 1994 World Congress on Powder Metallurgy, Vol. 2, Les Editions de Physique Les Ulis, France, 1994, pp. 1079-1086.

11. R. Tandon and R.M. German "Sintering and Mechanical Properties of Boron-Doped Austenitic Stainless Steel," Inter. J. Powder Metall., Vol. 34, No. 1, 1998, pp. 40-49.

12. K.M. Kulkarni, A. Ashurst, and M. Svilar, "Role of Additives in Full Dense Sintering of Tool Steel," Modern Developments in Powder Metallurgy, Vol. 13, Edited by H.H. Hausner, H.W. Antes and G.D. Smith, Metal Powder Industries Federation, Princeton, NJ, 1980, pp. 93-120.

13. D.S. Madan "Enhanced Sintering and Property Improvement in Ferrous P/M Compacts" Inter. J. powder Metall., Vol. 27, No. 4, 1991, pp. 339-345.

14. Animesh Bose Advances in Particulate Materials, Butterworth/Heinemann, Newton, MA, 1995.

15. A. Marshall "Sandwich Construction," Handbook of Composites, Reinhold, New York, 1982.