RALEIGH, N.C. The common view that amorphous materials are simply jumbled collections of atoms may give way to a more ordered theory of the materials' formation, according to experiments conducted at the University of North Carolina. Studies of an amorphous form of zinc chloride have revealed an unexpected order at nanoscale distances that may lead to new engineered materials in a wide number of industries, said project head James Martin.
"What I'm calling 'amorphous-materials engineering' will allow us to design nanostructures and then go in there and make them, [just as] we do with crystalline engineering today," said Martin, who was assisted by graduate students Stephen Goettler and Nathalle Fosse as well as Argonne National Laboratory researcher Lennox Iton. "Instead of using trial and error to discover new materials, amorphous-materials engineering will enable the properties of the bulk material to be fine-tuned by custom-tailoring the molecules used to manufacture it."
Since the 1980s, amorphous silicon has been finding its way into electronics applications such as photosensors and solar cells. Today it is being used to lay down low-cost, large-area thin-film transistor arrays for flat-panel displays.
Amorphous materials have been thought to be inherently disordered; thus, discovering new ones has mostly been a matter of experimentation. But Martin's instincts in chemistry made him suspect that the X-ray diffraction data he was looking at indicated there was a regular structure there, albeit at shorter distances than in crystalline materials.
"Right now if you want to change the characteristics of an amorphous material you have to resort to trial and error, just adding network modifiers without understanding what you are doing and without being able to specifically design what you want," said Martin.
But diffraction experiments seemed to show that the spacing between any patterns within liquids was at an atomic distance measured in angstroms (tenths of a nanometer). In silicon dioxide, as well as in a special "glass" form of zinc chloride he was studying, Martin suspected that the reason they form a transparent network (glass) is that the atoms have reciprocal affinities for each other in the 5- to 50-angstrom range.
"Zinc likes to be coordinated with four chlorine atoms, and chlorine likes to be coordinated with two zinc atoms no matter what the state. The only way they can do that is to share two chlorine atoms, forming a bridge between the two metal atoms. [That results] in a three-dimensional network," said Martin.
Martin became interested in zinc chloride glasses by accident, when he began to notice that every time he designed and synthesized a crystal using zinc and chlorine, he also ended up with a lot of glassy liquids as a by-product. Because the by-product material was produced in all instances, regardless of the crystal on which he was working, he began to wonder whether the liquids' glassy state was intermediate between a liquid and a crystal.
Since most solids undergo very little change in volume when melting from a solid into a glass or liquid, Martin reasoned that the affinities holding the molecules together was similar in liquids, glasses and crystals. The atoms in glasses are organized into a network that has no single-crystal structure in the range of 5 to 10 nanometers (the crystallinity range). But the individual bonds between atoms in a glass or a liquid occur in the range of 1 to 5 angstroms (10 times smaller).
Thus, Martin reasoned that an intermediate-range order between discrete chemical bonds and periodic crystalline lattices, namely between 5 angstroms and 5 nanometers (50 angstroms) governs the semiconducting and other bulk properties of these materials.
"As a chemist, I felt that the bonding between atoms could not change significantly between the crystalline and liquid or glass states of matter," said Martin. "Thus I should be able to control its characteristics by designing in the intermediate range." In glass the network structure of the liquid is frozen, making it easier to study. But in liquids, even though individual molecules drift around, the average density is similar to the glass form. This intermediate structure only "loosely" associates molecules according to their affinities. Nevertheless, controlling it could allow control of semiconducting and other properties for sensors, displays and even silicon circuitry.
Next step: widgets
"At the heart of every new technology is a new material. We've got the new material; now we need to get engineers interested in turning it into widgets," said Martin. "I can envision optical information-storage technologies based on molecules from 5 angstroms to 5 nanometers. I believe that in the near future we will be able to pattern control pathways of conductivity through amorphous semiconductors just like we know how to do with crystalline structures today."
To offer confirmation of his hypothesis that glasses have a nanoscale structure in the intermediate range, Martin designed a set of zinc chloride "template" molecules capable of displacing closely packed groups of one, four or 10 chlorine spheres in which zinc atoms filled a portion of the tetrahedral interstices. By choosing from several templates, the researchers designed network modifications to exhibit properties that were detectable when the structures were scrutinized with X-ray microscopy.
Martin took his amorphous material to Argonne National Laboratory to look at the material's atomic organization with a glass, liquids and amorphous materials diffractometer, which measures the interstitial distances between molecules. The results were eerily similar to the crystalline structure.
Currently, Martin is working on an undisclosed optical application. Next on the drawing board is a semiconducting electronics application.
An audio recording of reporter R. Colin Johnson's full interview with James Martin can be found online at AmpCast.com/RColinJohnson.