NAMBE2016 Session MBE-WeA: Novel Antimonide-based Materials (1:30 pm-3:00 pm)/2D and Layered Materials (3:30 pm-5:00 pm)

Significant progress has been made in recent years in the areas of optics and electronics, including the miniaturization of the transistor and the proliferation of optical devices including lasers and LEDs. However, in order to develop new devices with new functionalities, we need to explore new material platforms. In this talk, I will discuss the MBE growth and characterization of two novel material classes: heavily-doped III/V semiconductors and chalcogenide-based topological insulators.

Within the area of optical devices, the field of plasmonics has exploded with the promise of enhanced light-matter interaction in subwavelength volumes with applications in sensing and detection. In order to create plasmonic devices in the infrared, we must use alternative optical materials like heavily-doped semiconductors. We have demonstrated the growth of silicon-doped InAs and InSb with carrier densities as large as ~1x10²⁰ cm^-3. By controlling the doping density, the optical properties of the semiconductor can be tuned across the mid-infrared, from 5.5μm to beyond 17μm. In order to efficiently incorporate the dopant atoms, we found that InAs films must be grown at the relatively fast growth rate of ~2μm/hour, while InSb had to be grown much colder than undoped InSb. Understanding the growth mechanisms of these materials enabled us to create infrared plasmonic nanoantennas for improved sensing as well as perfect IR plasmonic absorbers. We have recently used our growth techniques to demonstrate the first single-material hyperbolic metamaterial using doped and undoped InAs. This body of work indicates that doped semiconductors are excellent IR plasmonic materials.

Finally, I will discuss our work on the growth of topological insulators (TIs), materials which may have applications in spintronics and far-infrared optics. Although Tis are ideally bulk insulating, most films grown to date are bulk conducting and degrade when exposed to air. For Bi₂Se₃, this is thought to be due to selenium vacancies in the film which can be filled by oxygen. Unlike other groups, we have grown Bi₂Se₃ using a selenium cracker source. Our films exhibit electron densities independent of film thickness, indicating little to no bulk doping. We also observe very little change in the film electrical properties even after more than two months in air, again indicating few selenium vacancies. By improving the growth of TI films, we hope to begin to access their intrinsic properties.

[1] Ren, et al., Appl. Phys. Lett., 108, 081101 (2016)

[2] Woodson, et al., Appl. Phys. Lett., 108, 081102 (2016)

[3] Ren, et al., Appl. Phys. Lett., 108, 191108 (2016)

[4] For example, Vaughn, Ph.D. Dissertation (2006)

Given the known variation of composition with growth conditions in mixed group-V systems, we first grew a step-graded structure in which the As/(As+Sb) flux ratio was varied from 0.95 to 0.50 in 0.05 increments (Fig. 1a). This structure was then replicated in a total of 7 samples, with a variety of substrate temperatures (T_sub) and V/III ratios to determine the rate of Sb-incorporation as a function of growth conditions. For all conditions investigated, we were able to obtain alloys with at least 40% Sb, demonstrating promise for this technique as an avenue toward LWIR devices. Through ongoing energy dispersive spectroscopy, photoluminescence (PL), and x-ray diffraction (XRD) experiments we will determine the composition and relaxation of each of the graded layers, ultimately correlating these properties with growth conditions. While preliminary PL (Fig. 1b) and XRD have shown that we can obtain the desired Sb compositions, they also indicate the presence of significant phase separation. Since segregation and threading dislocations are key physical defects that typically limit the utility of a material system, transmission electron microscopy will be performed to monitor the nature and density of these features and provide critical feedback on growth conditions. Through this effort, we hope to provide guidance for producing high-quality InAs_1-xSb_x alloys for LWIR applications.

[1] Vurgaftman et al., J. Appl. Phys.89, 5815-5875 (2001).

InAsSb has the narrowest bandgap of any of the conventional III-V semiconductors: low enough for long wavelength infrared applications. Such devices are sensitive to point defects, which can be detrimental to performance. To control these defects, all aspects of synthesis must be considered, especially the atomic bonding at the surface. We use an statistical mechanics approach that combines density functional theory with a cluster expansion formalism to determine the stable surface reconstructions of Sb (As) on InAs (InSb) substrates. The surface phase diagram of Sb on InAs (Fig. 1) is dominated by Sb-dimer terminated α2(2x4), β2(2x4), and c(4x4) reconstructions. Smaller regions of mixed Sb-As dimers appear for high Sb chemical potentials and intermediate As chemical potential. We propose that InAsSb films could be grown on (2x4) surfaces, which maintain bulk-like stoichiometry, to eliminate the formation of typically observed n-type defects [1-3]. Scanning tunneling microscopy and reflection high energy electron diffraction confirm the calculated phase diagram. Additionally, depositing Sb on InAs causes an increase in the coverage of vacancies and islands in the As dimer rows. Based on these calculations and observations, we propose a new explanation for Sb-As intermixing in these materials.

Transition metal dichalcogenides (TMDs) are 2D, layered materials with composition and thickness dependent electronic properties. Due to the weak van der Waals interactions between adjacent layers, TMDs enable heterostructure growth with relaxed criterion for lattice matching, allowing the selection of materials based primarily on their electronic properties. In this work, we present our recent investigation into the physics and chemistry of W-based TMD growth on other inert van der Waals substrates. For WSe₂, we demonstrate high-quality, crystalline WSe₂ thin films and show that the VDW interactions are strong enough to cause rotational alignment between the epi-layer and the substrate, which plays a key role in the formation of grain boundaries. To suppress nucleation and grain boundaries and enhance larger area 2D growth, the complex interaction between the W flux, Se flux, and diffusion rates was investigated. W flux and substrate temperature primarily govern the competition between the attachment and diffusion rates and are critical to controlling the WSe₂ nucleation and grain shape as seen in Fig. 1. The growth mode is primarily affected by the Se:W flux ratio which, in conjunction with lower nucleation rates, allows for larger 2D grain growth. The impact of these growth parameters on the electrical characteristics of the grown films was also investigated and compared to purchased CVT grown WSe₂ films. WTe₂ growth was also successfully demonstrated with its equilibrium phase being the semimetallic, distorted octahedral (1T’) structure. WTe₂ is of interest for a variety of applications including contacts, topologically protected transport, and to alloy with WSe₂ to form a small bandgap semiconductor for tunnel FET applications.

This work is supported in part by the Center for Low Energy Systems Technology (LEAST), one of six centers supported by the STARnet phase of the Focus Center Research Program (FCRP), a SRC program sponsored by MARCO and DARPA. It is also supported by the SWAN Center, a SRC center sponsored by the NRI and NIST, and the Texas Higher Education Coordinating Board’s Norman Hackerman Advanced Research Program.

Graphene-like two-dimensional layered materials have been extensively studied in recent years due to their large surface area-to-volume ratio and distinct physical properties.¹ These materials consist of ionically and/or covalently bonded atoms in two-dimensional planes that are bonded by weak, out-of-plane Van der Waals forces. Gallium Selenide belongs to this family and is of particular interest due to its unique physics (i.e. single tetralayer to bulk bandstructure change) and many device applications (i.e. electronics and photonics, radiation sensing, non-linear optics).² In this work, we investigated the MBE growth of GaSe on GaAs(111)B and GaN(0001)/SiC substrates. The growth of GaSe was confirmed with Raman and XRD. We examined the surfaces of our growths with AFM and SEM, and the crystal quality was interrogated with TEM and electron diffraction. GaSe/GaAs(111)B was found to form coalesced films at growth temperatures between 500-550°C, with Se and Ga fluxes ~10^-7and ~10^-8 Torr respectively. We used ARPES to measure the electronic bandstructure of GaSe/GaAs(111)B and found that the effective hole mass was ~1.19m₀ as shown in Fig 1a. GaSe/GaN(0001) was grown between 500-575°C at the same range of Se and Ga fluxes. Triangular islands of GaSe were formed on the GaN Surface and examined by SEM (Fig 1b), AFM, and TEM.

The heteroepitaxial growth of III-V semiconductors on silicon (Si) has been a great interest to integrate optoelectronic devices on a silicon chip for optical interconnects for the past few decades. However, there are some issues to overcome, associated with the direct heteroepitaxial growth of high quality III/V compounds on Si. Here we demonstrate -for the first time- high quality and single crystal GaAs film on Si by using Quasi van der Waals epitaxy (QvdWE), which use layered materials as a buffer to relief the strain at the interface, making a remarkable step towards the integration of III/V compounds on Si ^[1].

In this study, two-fold epitaxial growth was employed to achieve 3D/2D/3D heterostructures, which is two-dimensional layered material (2D) growth on Si followed by a heteroepitaxy of GaAs thin film on it as shown into the inset-figure 1. GaAs layer was grown on layered material at temperatures around 600 °C and a growth rate of 0.7 Å/s. For the first few layers, GaAs forms widely separated islands originated form nucleation on layered material, and the islands then coalesce as the growth proceeds. X-ray diffraction (XRD) spectra showed the successful epitaxial growth of high quality single crystal GaAs on Si (111) (Figure 1). Details of this study including optoelectronic properties will be discussed in the talk.

Gallium nitride (GaN) is of tremendous importance for applications in UV optoelectronics and high power, high frequency electronic devices. Usage of large diameter Si substrates was proposed for GaN growth in order to enable the compatibility of the GaN technology with the existing and well-established Si processing lines and the associated cost reduction and market penetration. The growth of GaN on Si is challenging because of the large lattice and thermal expansion mismatch of the two materials that generates defects within the deposited layer, such as large numbers of dislocations and cracks, which are detrimental to device performance. During the last decade, significant progress has been achieved on defect reduction in GaN growth on Si. However the defect density remains high when compared to other III-V semiconductor materials. Typical strategies used to mitigate this issue involve the growth of thick, engineered buffer layers prior to the active GaN layer deposition, which can prove costly and time consuming [1].

In this work, we performed the growth of GaN on Graphene stabilized Porous Silicon (GPSi) compliant substrates and we characterized the resulting defect density. GPSi is a nanocomposite material in which the porous layer provides improved elastic flexibility, which is essential to reduce the stress within the system. Graphene coating on the high specific surface of the porous Si enhances its thermal stability so it can withstand the high temperature annealing cycles that are required for the growth of GaN-based structures.

+ Author for correspondence: Richard.Ares@usherbrooke.ca

[1] D. Zhu, D. J. Wallis, C. J. Humphreys, Rep. Prog. Phys. 76, 106501 (2013).

Mg_0.13Cd_0.87Te is a promising material for II-VI/Si (1.7 eV/1.1 eV) tandem solar cells, which can potentially reach over 30% efficiency while benefiting from low manufacturing costs. Undoped Mg_0.13Cd_0.87Te/Mg_0.5Cd_0.5Te double heterostructures (DHs) grown by MBE in our group revealed carrier lifetime of 0.56 µs, indicating that the material quality is high enough to be potentially applied to solar cells. This abstract reports the latest results of the photoluminescence (PL) study of n-type doped Mg_0.13Cd_0.87Te/Mg_0.5Cd_0.5Te DHs and their application in solar cells. It was found that indium-doped Mg_0.13Cd_0.87Te has lower radiative recombination efficiency than undoped, thus an undoped Mg_0.13Cd_0.87Te absorber layer is used in MgCdTe DH solar cell devices featuring an a-Si hole contact. An intrinsic top Mg_0.5Cd_0.5Te barrier layer acts as a surface passivation layer beneath the a-Si hole contact to prevent recombination at the edge of the absorber. In the PL spectrum of the solar cell, a very broad and flat defect-related spectrum appears below the bandgap (1.7 eV), which was identified to be a result of In diffusion from the CdTe buffer to the Mg_0.13Cd_0.87Te absorber. The external PL quantum efficiency is measured to be 0.6%, indicating an implied V_oc of 1.3 V. An initial demonstration of a Mg_0.13Cd_0.87Te solar cell resulted in a V_oc = 1.16 V and an efficiency of 11.2%. The short-circuit current density was calculated using the External Quantum Efficiency (EQE) spectra. Such a high V_oc indicates excellent material quality and further improvement in efficiency can be expected by optimizing the short-circuit current and fill factor.