Lunar Building Materials - Some Considerations on the Use of Inorganic Polymers
STUART M.LEE
The fabrication of lunar building materials and the assembly of structures requires the use of adhesives, binders, and sealants. These terrestrial materials are usually Produced from carbon, hydrogen, and nitrogen, elements scarce on the Moon. One possible approach would be to use inorganic polymer systems synthesized from the available lunar chemical elements, viz., silicon, aluminum, and oxygen. This paper considers inorganic polymer systems, their background, status, and shortcomings, and the use of network polymers as a possible approach to synthesis. In addition, other potentially useful considerations were advanced, including glassy metals for unusual structural strength, and the use of cold-mold materials as well as foam-sintered lunar silicates for lightweight shielding and structural building materials.
INTRODUCTION
Ideally, structural systems assembled on the lunar surface would be fabricated from available lunar materials. This process might include synthesis of adhesives for use in filament winding, sealants, binders for incorporating into lunar road-bed materials to minimize dust generation, as well as materials for insulation and radiation shielding. However, lunar materials do not contain the abundance of basic chemical elemental building blocks available in terrestrial minerals. The lunar composition is dominated by five minerals - pyroxenes, olivines, feldspars, ilmenite, and spinel - and consists mainly of metal silicates and titanates. Terrestrial technology concerned with adhesives, sealants, and binders revolves about the ubiquitous carbon atom. All current adhesive systems use macromolecular organic polymeric compounds which are produced from carbon containing small molecules that contain, in addition to carbon, hydrogen, nitrogen, and oxygen. Only oxygen, which is combined as oxides and silicates, is relatively abundant on the Moon, comprising more than 50 percent by weight of lunar materials (ref. 1), while the upper limit of indigenous organics in lunar samples is about 1 ppb (ref. 2).
In the absence of lunar materials for fabricating organic structures, several alternatives were considered:
STATUS OF INORGANIC POLYMERS
Nearly all synthetic plastics and elastomers in use today are organic polymers. These consist of very large molecules containing linear or branched chains with a backbone of carbon atoms. However, it has been long known that carbon is not a required component of materials that exhibit plastic behavior. For example, plastic sulfur, glassy selenium, and polyphosphonitrilic chloride are totally inorganic substances that demonstrate elastomeric properties at moderate temperatures, but are unstable and cannot retain their potentially useful properties for prolonged periods.
Inorganic polymers are defined here as materials composed of long linear chain molecules that do not contain carbon in the chain but may have carbon in the pendant groups or in the side chain. Polymer systems containing both inorganic elements and carbon in the backbone structure are considered to be serniorganic polymers.
Inorganic polymers have been in use a long time; for example, glass is an inorganic polymer comprised of rings and chains of repeating silicate units, as are most rocks, minerals (including sapphire and ruby), brick, concrete, and ceramics, which are three-dimensional inorganic polymers (cf. fig. 1). In pyroxene, one of the main lunar minerals, similar chains of silicate units are found. Also, ladder polymers or double chains are found in amphibole minerals such as one form of asbestos. In this latter type of structure, the negative charges of the oxygen atoms are neutralized by positively charged metal ions that tend to bind adjacent chains together to form a crosslinked matrix (ref. 3). However, these materials have limited use because of the difficulty in fabricating them into useful objects such as those requiring high temperature. These materials generally are not flexible, elastomeric, or impact-resistant. However, despite the difficulties with naturally occurring inorganic polymers, these facts have not precluded further investigation into synthetic inorganic polymer chemistry.
For years, chemists have sought a middle ground somewhere between organic polymers on one hand and numerological inorganic materials on the other. Generally, organic materials cannot resist heat, while inorganics cannot undergo strain. The better material should be between these extremes. The desired structure required is postulated to consist of a linearly structured polymer with nonionic substituent groups to favor flexibility. This is the basic concept behind the design of silicone polymers, the most solid claim to commercial success of an inorganic polymer. The most widely used silicone polydimethylsiloxane (fig. 2) - consists of a chain of alternating silicon and oxygen atoms with two methyl groups attached to each silicon heteroatom, materials similar to the atomic makeup (not structural) of inorganic materials that occur in nature (quartz, feldspars, and zeolites).
The chemistry required to produce long chains (thought necessary to develop "plastic properties"), based on elements other than carbon, is not so highly developed as that used to prepare carbon-based polymers. Many of the polymerization reactions that would be expected to provide high-molecular-weight inorganic polymers tend to produce only small cyclic molecules or low-molecular-weight products. Conversely, many inorganic polymer chains, including silicones (above about 350oC, show strong tendencies to rearrange to low-molecular-weight polymers or small ring compounds upon heating, due, in the case of silicones, to cleavage of the relatively strong sificon-oxygen bonds, while the supposedly less stable silicon-carbon bonds remain intact (fig. 3).
Up to now, inorganic polymer chemistry has generally been approached by investigating the two basic types: those consisting of chains of a single element (homoatom) chain (as in plastic sulfur or black phosphorus) and those having two or more dissimilar alternating elements (heteroatom) chains that utilize metals (such as the aluminoxanes). However, despite promising claims made by early workers (ref. 4), the aluminoxanes have been found to hydrolyze easily (which may not be a lunar problem).
PRESENT TRENDS IN INORGANIC POLYMER STUDIES
Metalosiloxanes
Few elements have been overlooked in the search for new inorganic polymers. One well-worn approach systematically replaces the silicon or oxygen atoms or both of the basic silicone structure by other elements of periodic groups II and VI. By replacing some silicon atoms with other metallic atoms such as aluminum or titanium, materials known as metalosiloxanes have been prepared (fig. 4).
The synthesis and investigation of polymers with siloxane-like structures built with Al, Ti, P, and O atoms have been the work of Andrianov and others (refs. 5-8).
They reported that double-chain metalosiloxane structures containing aluminum or titanium were stable up to 590o C, while aluminosiloxanes (chains with alternating aluminum and silicon atoms separated by oxygen atoms) were stable above 300o C (ref. 9). A priori, these polymers could be potentially useful because of the chemical elements available on the lunar surface.
These metalosiloxane polymers, even though basically inorganic, contain some carbon side chains. The carbon used to synthesize these materials remains in the final polymer and naturally is not available for reuse.
Metalosiloxanes may be prepared by hydrolyzing metal siloxanic compounds at room temperature. Thus, a polyorganic silylaluminoxane may be prepared from nonaethylaluminoxytrisilane according to:
The resulting polymer contains an Al-O-Al main chain protected by organic substituents (O-Si-R3) which also acts as a chain-stopping group. In this scheme, the aluminoxytrisilane is prepared by treating a trialkylsilanol with aluminum. The material can also be obtained by reacting trialkyl sodium silenoxide with aluminum trichloride (possibly available from proposed lunar aluminum process).
These methods of preparation are not entirely satisfactory since initially they lead to the formation of appreciable quantities of cyclic or low-polymerized products (as previously indicated in fig. 3) that can be polymerized further. A similar problem was observed when polymer chains containing Al-O-P linkages were prepared. In two recent patents, Dow workers (refs. 10, 11) succeeded in preparing relatively long-chain aluminum-oxygenphosphorus bond polymers by reacting an aluminum compound with an alkyl orthophosphonate or phosphonic and/or phosphinic acid.
In another study, inorganic polymers containing siliconoxygen-metal linkages were prepared over several years (ref. 12) and it was found that various techniques can be used to establish the silicon-oxygen-bond. However, Chamberlain (ref. 12) states that "nothing has been found during this study to indicate that the Si-O-metal linkage can be of great utility in the formation of chemically and thermally stable elastomers." In addition to single-chain polymers, the ladderlike double chains (fig. 5) were synthesized by General Electric. These phenylsilsesquioxanes, or phenyl-T polymers, are reported (ref. 13) to have molecular weights up to 200,000 and even up to 4X106 with branching. They do not melt on heating, they are stable up to 400o C, and they form good films by solvent deposition.
Coordination Polymers
Coordination or chelate polymers are unique to inorganic chemistry and are formed when certain electron donor atoms such as oxygen, nitrogen, or sulfur donate an unshared electron pair to a metal ion that can receive these electrons. The donor atoms may be part of an organic molecule, in which case the entire molecule is said to be coordinated to the metal.
Attempts to crosslink chelate polymers were undertaken by a Monsanto team to develop semi-organic structural adhesives (ref. 14). However, the products obtained were not homogeneous. It appears that the coordination process for preparing inorganic polymers is not fully understood. The problems include the lack of product solubility, infusibility, and depolymerization. of product.
Miscellaneous Polymers
Some inorganic polymers that have been synthesized show some promise. For example, polymers of titanium oxide form strong films and are used in the optical industry (ref. 15), while a mixed product containing silicon and titanium yields materials that can be cold-hardened to give strong films suitable as coatings for ceramic electrical insulators and as a water repellant.
SOME PROBLEMS WITH INORGANIC POLYMERS
The outstanding success of polymer chemists in developing organic polymers has been an inspiration to the inorganic polymer chemist. However, research has not yet revealed an entirely new inorganic polymer system other than silicones. The chemistry of inorganic polymers lags a half century behind the chemistry of organic polymers. Since the 1960s, no recent comprehensive texts on inorganic polymers have appeared (refs. 5, 16-18). The lag in technology is apparent not only in our lack of knowledge of the polymerization process, but also regarding bonding of inorganic polymers. For organic polymers, the general covalent s-p sigma bonds and sigma-pi multiple bonds are well characterized. However, the inorganic polymer systems involve d-orbital pi-bonding and other less familiar interactions of bonding electrons.
Bonding types range from wholly covalent at one extreme to ionic at the other. The ability of carbon to bond with another carbon or oxygen makes the production and study of monomers simple. These monomers can then be converted to polymers by reactions that open the double bonds to form polymer linking bonds, addition polymerization. However, outside of the first raw elements, carbon, nitrogen, and oxygen, multiple bonds of the p-pi type are rare. Multiple bonds involving p-pi-d-pi orbitals are fairly widespread, but do not react by addition polymerization. As a result, inorganic monomers generally cannot be isolated to study polymerization.
Molecular engineering as applied to inorganic polymer systems might involve the design of molecules with optimum thermochemical bond strengths. However, one considered factor alone cannot suffice; for example, bond energy data may lead one to conclude that polymers based on inorganic elements should be superior to organic polymers in thermal stability. To determine thermal stability, we must ponder other kinetic considerations, for example, the silicon-oxygen bond in the silicones has a value of 110 kcal/mole, which is 75 percent stronger than the silicon-carbon bond (64 kcal/mole). Yet linear polysiloxanes heated above 600 K yield low-molecular-weight cyclic siloxanes. This indicates that the silicon-carbon bonds remain intact while the silicon-oxygen bonds are broken, which involves a bond rearrangement that occurs through a transition state, lowering the energy barrier relative to that required for a direct thermal rupture.
Many inorganic systems tend to polymerize through two or more units. However, most tend to form oligomers and small cyclic molecules, viz., dimers, trimers, tetramers, and other low-molecular-weight polymers rather than desired high-molecular-weight materials. Similar results were encountered in the development of silicone polymers, viz., the formation of small cyclic polymers from two to six units (fig. 3). Present technology has circumvented some of the tendencies of silicones toward cyclization. However, much of the product is still obtained in this undesirable form. The formation of small cyclic units also presents a possible route for depolymerization. When higher silicone polymers are heated in the range of their upper temperature utility 350o C-450o C, they can depolymerize to these volatile, cyclic, low polymers. This phenomenon, observed occasionally in completed systems, is called reversion.
In retrospect, it appears that a lack of sufficient background information is the major problem. Previously, there has been a strong emphasis, both commercial and governmental, for high-temperature -resistant materials, with a resultant concentration on synthetic approaches to inorganic polymers, to the detriment of basic research. Additional studies would be most helpful in the areas of inorganic bonding and polymerization reactions and probably would require a joint effort of both inorganic chemists (who understand the inorganic syntheses) and organic polymer chemists.
POTENTIAL CONSIDERATIONS
Glassy Metals
For extremely high-strength composites, there appears to be a proclivity toward more exotic materials such as whiskers and wire reinforcements, which are natural candidates when high structural efficiency is desired. In the same vein of new and unusual materials are the glassy metals (ref. 19). It is feasible that high-strength structures could be fabricated from these metal support materials which could be embedded in an adhesive, preferably inorganic, or a metal matrix such as aluminum.
Normal metallics, which are malleable, ductile, and opaque and are good conductors of heat and electricity, have a crystalline structure in which atoms are regularly arranged in three-dimensional repeating structures. Glasses, because of their covalent bonding, are strong and directional. It has long been known that metals can exhibit properties other than those of their natural state if they are cooled so rapidly that a nonequilibrium state is frozen in. The quenching rates required to produce glassy metals are usually several orders of magnitude higher than those achievable in normal metallurgical practice, that is, by simply plunging the heated metal into water.
An ingenious system for producing continuous ribbons of metallic glasses ("metaglasses") was developed at the Allied Chemical Corporation. The process involves injecting the hot metal in liquid form between counter-rotating drums that are continuously cooled. Long lengths of threadlike ribbon of glassy metal, up to about 2000 m/min, have been produced. The production cost is quite low because, by producing the glassy alloys directly from metal, it is possible to bypass the otherwise expensive and energy-consuming stages (such as casting, rolling, and drawing) necessary when working with the material in solid form.
The most important property of new metallic glasses is their unusual combination of strength and plasticity. Common glass, even in the form of a thin filament, cannot be strained more than I percent. However, a metallic glass specimen similarly shaped can withstand a local plastic sheer strength well in excess of 50 percent.
A five-component alloy with a tensile fracture strength about 3 times that of stainless steel has been produced. Moreover, stainless steels cannot withstand strains similar to the 50 percent observed by metallic glasses. Strong and hard metals are expected to be rather brittle; therefore, a material that has both strength and ductibility is an unusual advance. Another unusual aspect of these materials is their reported magnetic properties (ref. 19).
Cold-Mold Materials
Primary constituents of many electrical and automotive components from 1910 to 1920 are still used in low-voltage switch gear for arc chutes arc-resistant areas. These are the "transite" cement asbestos composites similar to some cold-mold formulations (ref. 20).A typical cold-mold composition is a cement asbestos product that is volumetrically loaded in a compression mold, pressed to shape in a punch-press-like operation, and cured in a water or steam-bath oven. Calcium alumino silicate is used to make an asbestos-filled cold compression molded compound cured by steam. These cold-mold materials are silicates, not unlike some of the minerals found on the Moon. It would be worthwhile to study techniques for adapting lunar silicates coupled with a synthesized inorganic polymeric binder to form cold-mold building materials that could then be used to build space structures. In addition, the cold-formed materials could be used in sandwich multilayer construction, consisting of two outer aluminum thin face skins bonded to an intermediate thick layer of low-density, cold-formed material. The latter material could be a foam type, using available oxygen as the foaming gas.
An Approach to Inorganic Polymers - Network Type
Glass technologists deal with highly crosslinked network polymers, yet inorganic polymer chemistry is almost entirely concerned with linear polymers. Research into processible network polymers could result in major advances in inorganic polymer chemistry.
Many common inorganic substances are polymers (fig. 1). Each compound is composed of large numbers of identical structural units linked by covalent bonds, but in these polymers the units are not joined in long chains, but instead form three-dimensional networks. Over the last three decades, the synthesis of inorganic polymers has been based on the belief that the structural units must be linked together into long chain molecules similar to organic polymers.
Although many novel and scientifically interesting structures have been synthesized, no useful new materials other than silicones have appeared; as a result, interest in this area has greatly diminished in recent years. The silicones have truly achieved fruition and embrace an extensive variety of polymers based on chains or networks of alternating silicon and oxygen atoms that match many of the structures found in the mineral kingdom. But, whereas the polysiloxane skeletons in natural minerals are crosslinked by means of metaloxy groups, the corresponding backbones in silicones are isolated by substitution of organic groups at silicon sites. The siliceous minerals are crosslinked structures so involved that the whole specimen of a pure mineral may be but one continuous molecule, insoluble, infusible, and intractable within reasonable temperature limits. On the other hand, the silicone chains are laterally blocked by organic substituents and ran merely associate with each other by weak van der Waals forces. Several organosilicon polymers together with their structural counterparts in the mineral kingdom are shown in figure 6, (ref. 16).
If the inorganic polymeric network structure is an accepted fact in natural inorganic polymers and ceramics, should this type of structure be considered as a logical extension for the synthetic inorganics? This approach has been taken by N. H. Ray (ref. 21), who postulated that the problem was due to the fact that inorganic polymeric structures must be linear to have the required combination of properties. Strangely, this theory is expounded even though the vast majority of inorganic polymers do not contain long-chain molecules but are composed of two- and three-dimensional networks. In fact, long-chain molecules are rare except among carbon and silicon.
Why has research been restricted to a linear approach? The answer probably is processability: most easily processible polymers soften reversibly to viscous liquids and can be processed by extrusion, injection molding, and vacuum forming.
All polymers at low temperatures are rigid solids with their molecular units bound into a three-dimensional structure either by van der Waals forces or sometimes by hydrogen bonds (as for long-chain polymers), or by covalent linkages, or sometimes by ionic forces (as for network polymers). If a polymer is to undergo permanent deformation by viscous flow, it must move bodily. This requires that the polymer be a long-chain molecule. For a network polymer, a higher temperature is required to achieve viscous flow than for a similar chain polymer. Higher temperatures favor more complex reactions and possible degradation.
Inorganic network polymers differ from organic polymers in that the network atoms are not joined directly (as in graphite) but are united through intermediate connecting atoms that may be oxygen (as in the simple compounds of boric acid or silica). The bonds within these polymers are covalent between oxygen and boron or silicon.
The search for processible inorganic polymers need not be restructured to chainlike molecules. The use of linear polymers requires that one or more unreactive substituents be attached to each atom in the chain. These univalent groups are either unstable to oxidation or hydrolysis or are lost on heating. The silicones are exceptional because substituents such as methyl, when attached to silicon, make possible the formation of a linear polydimethylsiloxane chain.
For production of network polymers, a much smaller proportion of unreactive substituents is required because they serve to control the density of network crosslinking. This process is illustrated by the chemical durability of ordinary glass, which is tremendously resistant to hydrolytic attack because of its crosslinked network structure as compared to the corresponding linear polymer sodium metasilicate, which is soluble in water and decomposed by acids.
Factors to consider in design include the expected processing temperature and the crosslink density as defined by P. J. Flory (ref. 22).
The stability as well as desirability of network structures can be attested to by the fact that network linking structures are used to overcome depolymerization of silicones by use of phenylene structural units. The phenylene units in this semi-inorganic silicone polymer appear to suppress the ordinary depolymerization reaction. Some of these materials showed no visible decomposition up to 500o C, but measurements indicated some weight loss up to 20 percent at these higher temperatures.
Network structures of aluminum-oxygen-phosphorus have been prepared by reaction of orthophosphoric acid and aluminum oxide (refs. 23, 24). The products of this reaction from these components is a viscous fluid that can be dried to an amorphous material believed to be a three-dimensional network of Al-O-P chains. These phosphate materials have been used as insulating coatings and as binders (ref. 25).
CONCLUSIONS
