Sn-containing group-IV semiconductors for energy applications in photovoltaics and thermoelectricity

Project: Research project

Project Details


During the past decade silicon technology has maintained its dominance of the semiconductor markets by incorporating new compatible materials that extend the capabilities of Si and its oxide for micro- and optoelectronic applications. The group-IV alloy Ge1-xSix has played a key role in these efforts. The latest generations of commercial microprocessors, starting with the 90 nm node, use GeSi alloys as stressors,1 and these alloys also play a key role in the burgeoning field of silicon photonics.2, 3 The GeSi system closely follows the so-called virtual crystal approximation, which predicts a very smoothly evolving electronic structure from pure Si to pure Ge. This leads to alloy electrical properties, such as conductivity, that are comparable to those of the end compounds. In contrast, the very different atomic masses of Si and Ge represent a very large perturbation to the vibrational structure, and the thermal conductivity of the alloy is substantially lower than that of the end compounds. Thus GeSi alloys are good materials for thermoelectricity applications,4 and in fact they dominate some niche markets such as radioisotope thermoelectric generators for space applications.5 The large lattice mismatch between Ge and Si enables the strain engineering applications of GeSi alloys, but is also the source of serious strain management problems in alloys grown on Si substrates. In applications that require thick layers, such as photovoltaics, the incorporation of active GeSi layers has not been possible due to the generation of strain-relieving dislocations. Moreover, strain and band structure cannot be decoupled, since they are controlled by a single compositional parameter. In view of these limitations, the interest in ternary group-IV semiconductors has grown over the past decade. The introduction of SiGeC alloys represented a first step in that direction.6 Unfortunately, the solubility of C in Ge1-xSix is very limited, and the perturbation represented by the C atoms is so large that the electronic structure is no longer amenable to a virtual crystal description. More recently, the chemical vapor deposition (CVD) growth of ternary Ge1-x-ySixSny alloys with large Sn atomic fractions was demonstrated by our group for the first time.7 These alloys represent the first realistic opportunity for independent strain and band structure manipulation. This can be seen quite dramatically in recent work showing an absorption edge tunable over a broad range while maintaining the lattice constant exactly matched to that of Ge. Combined with novel CVD approaches to the growth of binary GeSn alloys8 as well as superior quality Ge films on Si substrates,9 the Ge1-x-ySixSny system represents an entire new family of materials with the potential to revolutionize the field of group-IV alloys. The purpose of this proposal is to explore the potential for energy applications of the novel Sncontaining group-IV alloys developed at Arizona State University. The two-dimensional compositional space of the ternary alloy as well as the many possible lattice matched and un-matched combinations of ternary, binary, and elemental group-IV semiconductors provides unprecedented flexibility to further develop and optimize the basic properties of these materials for photovoltaics and thermoelectricity. In particular, the following fundamental questions will be addressed: Is it possible to achieve four-junction solar cells with an ideal combination of band gaps by combining GeSn (or pure Ge), GeSiSn, and III-V alloys which are perfectly lattice matched? Is the Si-Ge-Sn system suitable as an intermediate band photovoltaic material? Can the electronic and vibrational structure of ternary GeSiSn alloys be engineered to yield large impact ionization rates for solar cells with internal quantum efficiencies above unity? Is it possible to develop Si-Ge-Sn systems with thermal conductivities much lower than those available today
Effective start/end date9/1/098/31/13


  • National Science Foundation (NSF): $884,357.00

Fingerprint Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.