Technical: This project aims at the fundamental study of the growth of near-single crystal Ge on non-crystalline substrates such as glass. A glass substrate does not have a well ordered atomic lattice and therefore cannot induce the growth of epitaxial semiconductors. A key challenge is whether it is possible to grow a single crystal film on glass at all. In this project it is proposed to use a biaxial CaF2 buffer layer on glass to induce an epitaxial growth of near-single crystal Ge films. The biaxial CaF2 layer, which possesses strong out-of-plane and in-plane texture orientations, will be grown using an oblique angle deposition technique. The aims of the proposed research are to (a) understand the fundamental mechanism of the heteroepitaxial growth of near-single crystal Ge films on biaxial CaF2 buffer layer, and (b) to characterize the structural defects and electronic properties of the epitaxial Ge films, and (c) to understand the fundamental nature of the defects in the Ge films, in particular, the small angle grain boundaries, using classical and quantum mechanical atomistic methods. The expectation is that these near-single crystal Ge films may possess superior device quality properties compared to those polycrystalline films fabricated by conventional means on glass where the grains are randomly oriented. The strategy proposed in this research is robust and may stimulate more study on the growth of other semiconductors on amorphous substrates.
To date, high performance semiconductor devices are made out of single crystal semiconductor films grown on single crystal substrates. However, single crystal substrates are too expensive for large area applications such as display and solar cells. Most low-cost commercial large area devices are made of either polycrystalline or amorphous semiconductors on glass or metal substrates with less than ideal efficiency and long term stability problems. In this project the PIs are aiming to create a near-single crystal semiconductor films such as Ge grown on glass. The knowledge gained from this study can aid researchers to design and functionalize textures to obtain the desirable physical properties of thin films. It is projected that these near-single crystal semiconductor films should possess substantially better performance than that of the films deposited by conventional means on glass for large area electronics applications. This interdisciplinary work, research will be integrated with education and outreach activities. In addition to training graduate students, students supported through the NSF-REU and Rensselaer-sponsored URP (Undergraduate Research Participation) programs will also participate in the research. Outreach activities will include open houses and summer programs on campus for high school juniors/seniors. These efforts will enhance and stimulate students' interest in pursuing science or engineering careers.
To date, high performance semiconductor devices are made out of single crystal semiconductor films grown on single crystal substrates. However, single crystal substrates are too expensive for large area applications such as display and solar cells. Most low-cost commercial large area devices are made of either polycrystalline or amorphous semiconductors on glass or metal substrates with less than ideal efficiency and long term stability problems. This is because a glass substrate does not have a well ordered atomic lattice and therefore cannot induce the growth of epitaxial semiconductors. It was suggested that thin films with small angle grain boundaries may possess electrical characteristics close to that of single crystals that can give rise to very efficient devices. Theoretical work based on first principles calculations conducted in the present project confirmed that although the small angle grain boundaries are extended defects they are indeed benign and are not detrimental to device performance. Remarkable agreement with experiment is obtained, which enables us to identify the physical origin for the disappearance of hole transport barriers as a transition to electrically benign grain boundaries consisting partial dislocations. Figure 1 shows a comparison between experimental and theoretical carrier energy barrier as a function of total mosaic grain boundary angular spread. Guided by this theoretical work, near-single crystal germanium (Ge) films were experimentally grown on glass substrate through a calcium fluoride (CaF2) buffer layer. The CaF2 layer was deposited using a thermal evaporation technique with an oblique incidence flux to create crystallographically oriented films with small angle grain boundaries. Both Ge and CaF2 possess a preferred (111) crystal orientation as well as a preferred orientation along the plane of the surface. Figure 2 shows a scanning electron microscopy (SEM) cross section image of a Ge(111)/CaF2(111)/glass sample. Furthermore, the Ge(111)/CaF2(111)/glass structure was used as a substrate to grow a near-single crystal cadmium telluride (CdTe) film (a solar material) successfully. The method employed to grow this film is called metallorganic chemical vapor deposition (MOCVD). Figure 3 shows a home-made MOCVD system that was used to grow this film. Extensive characterizations of the film using x-ray diffraction, reflection high-energy electron diffraction, and SEM revealed that the CdTe film was epitaxially grown on the Ge(111)/CaF2(111)/glass substrate. The CdTe film also has a preferred crystal orientation in the  direction as well as a preferred orientation along the plane of the surface and can be classified as a near-single crystal film. Figure 4 in shows an SEM cross section image of a CdTe(111)/Ge(111)/CaF2(111)/glass sample. The successful growth of these near-single crystal films suggested that high quality, large area semiconductors on inexpensive substrates are possible compared to that of conventional polycrystalline films. The knowledge gained is this research can aid researchers to design and control the crystal orientation of semiconductor films on non-single crystal surfaces. In the interdisciplinary work, research was integrated with education and outreach activities. In addition to training four graduate students, six undergraduate students (including one minority student) supported through the NSF REU (Research Experiences for Undergraduates) and Rensselaer-sponsored URP (Undergraduate Research Participation) programs also participated in the research. Figure 5 in shows an undergraduate student learning solar cell research in the lab with the help of a graduate student. Our faculty also participated in local outreach programs such as Edison Tech Center education series: http://youtu.be/CAIuC6ZwtHc (Why engineering?) and http://youtu.be/xTurvYt-C78 (Copper in our electrical world). These efforts had stimulated studentsâ€™ interest in pursuing science or engineering careers.