Development of Low-Toxicity Gelcasting Systems Mark A. Janney, * Ogbemi O. Omatete, * Claudia A. Walls, Stephen D. Nunn, * Randy J. Ogle, and Gary Westmoreland Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831 A series of low-toxicity gelcasting systems has been devel- oped. The reagents used in these systems have very low acute toxicity. The new systems perform at least as well as, and in some cases better than, the original acrylamide- based system. The development of these systems is de- scribed herein, including the search for new gel composi- tions, the study of suspensions made with the new gel precursor solutions, and pyrolysis of the dried gels and gelcast parts. Applications of the new gelcasting systems include complex silicon nitride parts, large-diameter rings, rapid prototyping by green machining, and metal-powder gel casting. I. Introduction G ELCASTING is an attractive new ceramic forming process for making high-quality complex-shaped ceramic parts. 1 A slurry made from ceramic powder and a water-based monomer solution is poured into a mold, polymerized in situ to immo- bilize the particles in a gelled part, removed from the mold while still wet, and then dried and fired. Gelcasting is a generic process. It is not limited to use with any particular ceramic powder, because the processing additives are all organic and leave no residual cation impurities in the fired part. Ceramic parts from more than a dozen different compositions (ranging from alumina-based (Al 2 O 3 -based) refractories to high-perfor- mance silicon nitride (Si 3 N 4 )) have been produced by gelcast- ing. It is a robust process and can be quickly adapted for use with new materials and new applications. A comparison of gelcasting with slip casting, injection mold- ing, and pressure casting is given in Table I. Gelcasting com- pares favorably with the other forming processes in all catego- ries and is the most desirable in many categories. Specifically, gelcasting provides a rapid forming cycle, good wet and dried strength, the option to use a range of mold materials, the ability to make large parts (maximum dimension of >1 m) that have thick and thin sections, and minimal molding defects. Industry has been reluctant to use gelcasting because the main component of the gel, acrylamide, is a neurotoxin. 2,3 To mitigate this problem, an effort was initiated to develop gel systems that have similar or superior properties to the acryl- amide-based system and yet were low in toxicity. The present paper chronicles the search for new gel compositions, the study of suspensions made with the new gel precursor solutions, and the pyrolysis of the dried gelcast parts. We also provide a brief summary of the environmental, safety, and health aspects of the best systems. II. Search for New Gel Compositions Developing an alternative to the acrylamide-based gel- casting system required the identification of other monomers that would form gels in aqueous solution. Monomers for gel- casting must have the following attributes. They must be water soluble (at least 20 wt% for monofunctional monomers; higher solubility is desirable). Monofunctional monomers (one vinyl bond) should form water-soluble homopolymers; this gives a high probability that the polymer will form a gel with water when it is crosslinked. Difunctional monomers should also be as compatible with water as possible. Solubility should be at least 2 wt%; as with the monofunctional monomers, higher solubility is desirable. The monomers should be low in toxicity. Finally, the monomers should be inexpensive. The targets for bulk chemical prices were <$25/kg for monofunctional mono- mers and <$40/kg for difunctional monomers. A search of the literature produced a list of potential mono- mers that might be useful for their gel-forming tendencies (Table II). All the monomers listed in Table II meet the mini- mum requirements that have been outlined previously. How- ever, some additional requirements have been added to the initial list. For example, some monomers, such as para-styrene sulfonic acid (sodium salt), are highly ionic; ceramic powders (e.g., alumina and silicon nitride) have been observed to be difficult to disperse in concentrated aqueous solutions (10–20 wt%) of this monomer, and, therefore, additional evaluation was not performed. Other monomers such as HEMA (see Table II) only form water-compatible gels at very high concentrations (>40 wt%); these concentrations were deemed too high to be economical. From the initial list, six monofunctional and four difunc- tional monomers were chosen for testing. The monomers were categorized into four functionality groups: acrylate, acryl- amide, allyl, and vinyl. In addition to the monomers, three free-radical initiator systems were identified as being po- tentially useful: ammonium persulfate/tetramethyl ethylene diamine (APS–TEMED, in a 1:1 weight ratio), azobis [2-(2- imidazolin-2-yl) propane] HCl (AZIP), and azobis (2- amidinopropane) HCl (AZAP). (1) Statistical Survey of Monomer Systems Evaluation of the numerous combinations of monofunctional and difunctional monomers, together with the free-radical ini- tiators, presented a significant experimental challenge. To re- duce the total number of experiments, a fractional factorial statistical study was undertaken. The Taguchi 4 statistical method is well suited to the study of a process that has many factors that need to be evaluated at several levels. This method was chosen for the selection pro- cess in the development of the new gel systems. The Taguchi approach is invaluable for identifying which factors are the most important in a complex system. The Taguchi approach does not necessarily identify the optimum combination of those elements; therefore, in the present context, it should be con- sidered to be a screening method rather than an optimization P. W. Brown—contributing editor Manuscript No. 191468. Received October 8, 1996; approved June 16, 1997. Research sponsored by the United States Department of Commerce, as part of the NIST Advanced Technology Program, under Cooperative Agreement No. 70NANB2H1262, and by the United States Department of Energy, Assistant Secre- tary for Energy Efficiency and Renewable Energy, Office of Transportation Tech- nologies, as part of the Advanced Automotive Materials Program, under Contract No. DE-AC05-960R22464 with Lockheed Martin Energy Research Corporation. * Member, American Ceramic Society. J. Am. Ceram. Soc., 81 [3] 581–91 (1998) J ournal 581