Strained Relationships
September 14, 2010
Whereas
chemistry is generally concerned only with the composition of matter at a molecular level,
materials science looks at how both composition and microstructure affect the macroscopic properties of materials. One important material property is
strain, and it's strain that transforms
polyethylene terephthalate into
Mylar (a.k.a., biaxially-oriented polyethylene terephthalate), thereby increasing the
tensile strength from 80
MPa to about 225 MPa.
Many decades ago, it was found that strained
semiconductors have significantly different properties than their unstrained cousins. For example,
stretching silicon by epitaxially growing it atop a substrate of
silicon-
germanium increases the
electron mobility and the speed of the silicon
transistors. Both silicon and germanium have the
diamond cubic crystal structure, but the
lattice constant of germanium (0.565791 nm) is greater than that of silicon (0.54311 nm). Advanced techniques of selective doping to achieve local stress at certain portions of the transistor structure allow an even greater affect.[1]
Similar tricks might be applied in a more dramatic way on an atomic level to increase the
infrared absorption characteristics of
photovoltaics and also produce
thermoelectrics of higher efficiency. An international team of physicists and materials scientists from the
University of Michigan (Ann Arbor, Michigan) and the
Tyndall National Institute of the
University College Cork (Cork, Ireland), have been investigating highly mismatched semiconductor alloys for solar energy applications.[2-3] Lattice constant mismatched compounds will solidify into more than one
crystal phase on cooling because a single phase is not
thermodynamically favored. If the atoms of these compounds are laid down using
molecular beam epitaxy (MBE), they're forced into congress with each other, and the resulting material can exhibit unique properties.
The absorption characteristics of a photovoltaic depend on the
electronic band structure of the semiconductor. If energy levels are not present that can absorb a photon of a particular wavelength, the conversion of light into electricity is not possible. Alloying different semiconductors (for example, making
copper indium gallium selenide) improves photovoltaic efficiency over a simpler compound. Efficient thermoelectric materials need to have the unique combination of low
thermal conductivity and high
electrical conductivity. As any physicist knows, the conductance of both heat and electricity in simple metals is by electrons, a principle that's expressed by the
Wiedemann-Franz law that the ratio of these conductivities at a given temperature is the same among the metals. It's only when you use semiconductors instead of metals that you break away from this restriction and find materials for which the ratio of conductivities is different, and better thermoelectric efficiency can be achieved. Even then, your material palette doesn't allow for much variation.
Not a good solar energy day (Photo by the author).
The Michigan-Ireland team, lead by University of Michigan materials scientist,
Rachel Goldman,[4] made gallium arsenide nitride, a mismatched alloy of
gallium arsenide and
gallium nitride, using molecular beam epitaxy. Gallium arsenide (GaAs) and gallium nitride (GaN) each have the
zincblende crystal structure, but the lattice constant of GaAs (0.56533 nm) is much larger than that of GaN (0.452nm). Gallium arsenide is a good photovoltaic material, but the resulting mixed alloy semiconductor absorbed more efficiently in the infrared.
What's interesting about this research is that the gallium arsenide nitride was found to be very inhomogeneous. At a 10 ppm
nitrogen level, many of the nitrogen atoms were clustered together and not distributed in the lattice. You have, in effect, a phase decomposition at a nano level, which in retrospect might have been expected. There's still much to be learned from such non-equilibrium alloys. The most important question is whether semiconductor alloys of this type will decompose when used in their intended photovoltaic application, where temperatures run rather hot. This research was funded by the
National Science Foundation and the
U.S. Department of Energy.
References:
- Katherine Derbyshire, "Strain Engineering: Helping Transistors Scale Beyond 90 NM," Semiconductor Manufacturing, vol. 6, no. 3 (March, 2005).
- Nicole Casal Moore, "Forcing mismatched elements together could yield better solar cells," University of Michigan Press Release, September 8, 2010.
- T. Dannecker, Y. Jin, H. Cheng, C. F. Gorman, J. Buckeridge, C. Uher, S. Fahy, C. Kurdak, and R. S. Goldman, "Nitrogen composition dependence of electron effective mass in GaAs1-xNx," Phys. Rev. B 82, 125203 (2010).
- Rachel S. Goldman Home Page.
Permanent Link to this article
Linked Keywords: Chemistry; materials science; strain; polyethylene terephthalate; Mylar; tensile strength; MPa; semiconductor; strained silicon; silicon; germanium; electron mobility; transistor; diamond cubic; lattice constant; infrared; photovoltaics; thermoelectrics; University of Michigan (Ann Arbor, Michigan); Tyndall National Institute; University College Cork (Cork, Ireland); crystal phase; Gibbs Free Energy; molecular beam epitaxy (MBE); electronic band structure; copper indium gallium selenide; thermal conductivity; electrical conductivity; Wiedemann-Franz law; Rachel Goldman; gallium arsenide; gallium nitride; zincblende; nitrogen; National Science Foundation; U.S. Department of Energy.