Technical. This project addresses synthesis/processing/fabrication research to explore quantum-cascade (QC) mid- and far-infrared light-emitting materials based on SiGe quantum wells (QWs). Since QC lasers are semiconductor light sources based on intersubband (ISB) transitions (i.e., electronic transitions between quantized states within the same energy band), their operation is essentially unaffected by the nature of the energy band gap of the underlying materials. Hence such considerations provide an approach for the demonstration of laser action in silicon ? a goal complicated by the indirect band gap of (Si)(Ge). Strain, however, associated with Si/SiGe QWs appears as a significant challenge to progress. This project takes a new approach, in which elasti-cally relaxed SiGe nanomembranes are used as the growth substrates and/or the active material. Such nanomembranes will be grown epitaxially on Si(Ge)-on-insulator substrates and then re-leased from the handle wafer by removing the underlying oxide layer via wet etching. The de-sired result would be a free-standing heterostructure where strain relaxation occurs via elastic strain sharing among the constituent epilayers without the formation of defects, and thus be vir-tually free of dislocations. The released nanomembranes could then be transferred onto other substrates. Further challenges complicating the demonstration of silicon-based QC lasers are provided by design issues specific to the SiGe QW materials system, which are also being ad-dressed on this project. An approach based on electronic ISB transitions in the L valleys of Ge/SiGe QWs will be explored. Recent calculations indicate that this approach may have advan-tages over the p-type structures that have been investigated so far, including longer nonradiative lifetimes, larger oscillator strengths, and more efficient tunneling transport. Non-Technical. The project addresses fundamental research issues in a topical area of elec-tronic/photonic materials science having technological relevance. The proposed use of nanomembrane technology for the fabrication of complex semiconductor quantum structures has the potential for broad technological impact beyond the SiGe materials system and the QC-laser device application described above. Moreover, the project activities will promote education through the training of students across disciplines, ranging from semiconductor epitaxial growth to nanomembrane synthesis and processing, bandgap engineering, and THz (terahertz) photonics. To increase the effectiveness and scope of the program, the involvement of undergraduates and high-school interns will be emphasized, by leveraging existing programs with a strong focus on under-represented minorities.
This research program has been focused on the development and characterization of strain-engineered group-IV semiconductor nanomembranes (NMs) for optoelectronic device applications. The use of strain to tailor materials properties in a controlled and potentially useful fashion has a long history within the general field of materials science. Semiconductor NMs (i.e., single-crystal sheets with a thickness of only a few ten nanometers) offer novel and unique opportunities for strain engineering, both through spontaneous strain relaxation upon NM release and through the application of external mechanical stress. In the present research project, this basic idea has been applied to a grand challenge of semiconductor optoelectronics, namely the demonstration of efficient group-IV light emitting materials. Group-IV semiconductors (i.e., Si, Ge, and related alloys) provide the leading materials platform for microelectronic devices and integrated circuits. At the same time, these materials are not suitable to the development of light emitting diodes (LEDs) and lasers based on traditional approaches, due to their indirect energy bandgap which results in exceedingly low interband radiative efficiency. This issue represents the key factor currently limiting the large-scale on-chip integration of electronic and photonic devices, as required to extend the performance and functionality of existing microelectronic systems. A possible solution to this grand challenge is the use of intersubband transitions in quantum cascade structures, i.e., transitions between quantized states derived from the same energy band (so that the presence of an indirect fundamental energy bandgap is irrelevant). However, the development of such structures based on SiGe quantum wells (QWs) has so far been hindered by limited material quality, due to the large lattice mismatch between Si and Ge and the resulting strain accumulation that typically relaxes through the formation of misfit dislocations. In this research program, we have shown that SiGe QWs of superior structural quality can be developed using NMs as the growth substrate and active material. Specifically, we have demonstrated the strain-compensated growth of thick multiple-QW structures on previously released NMs, prepared by growing a few periods of the same QWs on a Si-on-insulator (SOI) substrate. Because of the elastic relaxation that takes place upon NM release, the NM in-plane lattice constant perfectly matches that of the desired QW structure. As a result, an arbitrarily large number of additional QWs can be subsequently grown on the NM without any strain accumulation [Fig. 1(a)]. Strong far-infrared intersubband absorption features with record narrow linewidths have been measured with samples based on this novel approach [Fig. 1(b)]. The same approach can be extended to arbitrary SiGe QW structures such as the active materials of quantum cascade lasers. Furthermore, we have investigated the use of mechanically stressed, tensilely strained NMs [Fig. 2(a)] as a means to enable interband optical gain in Ge. Biaxial tensile strain in Ge has been theoretically predicted to lower the conduction-band edge at the direct point relative to the L-valley minima (thereby increasing the efficiency for interband light emission and enabling the formation of population inversion), while the overall bandgap energy correspondingly decreases. If the strain can be made to exceed 1.9%, the fundamental bandgap even becomes direct. These ideas have been experimentally demonstrated in this project for the first time, through the measurement of straining beyond the threshold for direct-bandgap behavior, together with strong strain-enhanced mid-infrared photoluminescence (PL) [Fig. 2(b)] and evidence of population inversion under optical pumping. Furthermore, we have developed an optical-cavity geometry compatible with the flexibility requirements of mechanically stressed ultrathin active layers. Altogether these advances comprise the key ingredients required for the demonstration of strain-enabled laser action in Ge. These activities have been carried out by six Ph.D. students at Boston University or the University of Wisconsin – Madison (including two students from traditionally underrepresented groups), thereby contributing to their professional development and progress towards graduation. Several undergraduate students have participated in related research endeavors each year, and in the process have gained valuable exposure to state-of-the-art semiconductor nanoscience and nanotechnology. Many of these students have subsequently enrolled in graduate school.
