NAMBE 2024 Session NAMBE-MoP: NAMBE Poster Session

Dilute bismuthides are a class of highly mismatched alloys consisting of small amounts of bismuth incorporation in III-V semiconductors.¹This incorporation of bismuth, due to its larger size compared to other elements within the host matrix, induces valence band anticrossing, thereby reducing the bandgap.² For this reason, dilute bismuthides are explored as a class of materials in optoelectronic devices such as infrared photodetectors.^2,3,4

Previous studies have shown that InGaAs based photodetectors have a wavelength of 1.7μm. Bi incorporation in InGaAs reduces bandgap by 56meV/%Bi.³ In our work, we aim at optimizing InGaBiAs to establish it as a promising material in the intrinsic region of shortwave infrared photodetectors and extend the wavelength beyond 1.7μm. Growth of dilute bismuthides is a challenging process. Hence, we are using molecular beam epitaxy to grow InGaBiAs at low temperatures of 265^oC (as measured by band-edge thermometry) and near-stoichiometric conditions. Be:InGaAs as a capping p-type layer and Si:InGaAs as a n-type layer are grown at 490^oC for infrared photodetectors. In our work, we show that 1.5%Bi incorporation in InGaBiAs extends the wavelength up to 1.8μm. We are also studying the recombination lifetime of the carriers in InGaBiAs using optical pump terahertz probe spectroscopy. The optimization of Bi in InGaBiAs to give longer wavelengths and higher recombination lifetimes is currently ongoing research.

[1] Crystals 17,7,63 (2017) [2] J. Vac. Sci. Technol. A 40, 042702 (2022) [3] Appl. Phys. Lett. 99, 031110 (2011) [4] Appl. Phys. Lett. 100, 112110 (2012)

Efficient, controlled absorption of thermal radiation from the MBE system substrate heater into the growth substrate continues to be a technological challenge [1]. This problem is especially difficult for the case of wide- and ultra-wide bandgap (WBG and UWBG) semiconductors where near IR absorption is low and growth temperatures exceeding 1000°C may be required for high quality epitaxy. UWBG semiconductors are attracting increased attention for high power density RF electronics and other applications, so addressing the practical MBE growth temperature issues is important. In addition, a commonly-used method to increase IR absorption by using thick, opaque, specular backside metal layers can be detrimental to substrate temperature monitoring with diffuse reflectance [1,2], so a different approach is needed.

Woltersdorff [3] showed that in the far IR where the optical constants n and k are both large for metals, the maximum absorption is 50% for a thin metal film that has a sheet resistance of half that of the characteristic impedance of free space, namely 377/2 = 188.5 ohms/sq.Thus, one would expect that the ideal thickness of a thermal absorbing layer to depend on the wavelength range and the absorbing layer properties, and that thick layers may be counter-productive for efficient heating.

Here we report our recent work in modeling and experiments using thin stacks of silicon (Si) on titanium nitride (TiN) to greatly increase the thermal absorption of WBG SiC substrates during MBE growth.Modeling with the Python software package WPTherml [4] indicates that a 100 nm thick Si layer on a 9 nm thick TiN layer increases the peak absorbance of IR from a 1000°C blackbody substrate heater to over 58%.We used a similar thin layer stack on a 4H-SiC substrate and a substrate heater temperature of 1100°C to efficiently heat the substrate to a real temperature of ~ 1000 °C (hundreds of degrees hotter than bare SiC [1]) as revealed by a bright Ö3 x Ö3 R30° RHEED pattern [5], showing the value of this approach.

This work was funded by the Office of Naval Research.

[1] D. S. Katzer et al., J. Vac. Sci. Technol. B 38 032204 (2020).
[2] S. R. Johnson et al., J. Vac. Sci. Technol. B 11 1007 (1993).
[3] Z. Woltersdorff, Z. Phys. 91 230 (1934).
[4] J. F. Varner et al, J. Open Res. Softw. 7 28 (2019).
[5] X. N. Xie and K. P. Loh, Appl. Phys. Lett 77 3361 (2000).
⁺ Author for correspondence: douglas.s.katzer.civ@us.navy.mil

DISTRIBUTION STATEMENT A. Approved for public release: distribution unlimited.

Quantum spin Hall insulators (QSHIs) are advanced materials with topologically protected surface states that allow electronic transport without scattering from defects. QSHIs with a large energy gap in the topological phase are of great interest for development of both quantum and electronic devices. The ability to control the topological phase transition in a QSHI would represent a functional improvement for applications that require in-situ tunability of the material’s topological properties such as spintronics and fault-tolerant quantum computing.[1]

It is possible to induce topological states in III-V semiconductor quantum wells with a brokengap band alignment and use quantum well width to adjust the size of the hybridization energy gap. We are therefore interested in investigating the topological phase transition in InAs/GaInSb quantum well heterostructures varying different structural parameters. By simulating band topology and topological invariants we can observe changes in wavefunction in k-space and control the quantized responses of the material.[2] Two key parameters that shape a material’s band structure are symmetry and strain. We compare (001)- and (111)-oriented III-V semiconductor quantum well heterostructures under both compressive and tensile strain to explore their effects on the emergent topological states.

QSHIs consisting of InAs/GaInSb quantum well heterostructures are considered since by changing the well widths and gate voltage we can control the degree of band inversion and corresponding topological properties.[3] The ability to tune the strain in these structures by varying the composition of the ternary GaInSb quantum well can enhance the size of the hybridization gap energy between the topological states compared to QSHIs based on binary InAs/GaSb QWs.[3]

We have therefore carried out computational simulation to investigate InAs/GaInSb double quantum well heterostructures with both (001) and (111) crystalline orientations.[4] We will present our results that show the influence of tunable parameters such as quantum well width and strain on topological band structure.

Furthermore, given the result of our computational work, we can create these structures with molecular beam epitaxy (MBE) down to monolayer precision. This precise control enables us to adjust the quantum well widths and strain levels, to influence topological properties of the InAs/GaInSb quantum wells. We anticipate that MBE's high degree of control over the deposition of our simulation parameters will lead to the realization of QSHIs with optimized performance for applications in spintronics and quantum computing.

Hybrid Superconductor/Semiconductor (SP/SE) structures play an important role in condensed matter and mesoscopic physics, particularly in topological quantum computing aimed at realizing Majorana Zero Modes. Aluminum growth on InAs or InSb quantum wells using Molecular Beam Epitaxy emerges as the most promising direction due to the high electron mobility and strong spin-orbit interactions within these wells. Nevertheless, achieving a thin, continuous Al layer (~10 nm) poses challenges in standard MBE systems due to Al's high surface mobility in ultra-high vacuum environments and its tendency for 3D nucleation.

We have demonstrated that a thin, continuous Al layer can be successfully grown on an InGaAs surface at temperatures above room temperature using a high Al-growth rate of 3 Å/s [1]. In the current study, we compare this process with the growth of Al on a homomorphic epitaxial InSb surface. In-situ RHEED studies reveal similar Al behaviors on both surfaces. Regardless of the semiconductor surface used, 3D nucleation occurs after the initial Al monolayers for both growth rates tested (0.1 and 3 Å/s). With a growth rate of 0.1 Å/s, the 3D growth mode dominates throughout the deposition of the nominal 10 nm of Al. In contrast, at the Al growth rate of 3 Å/s, the initial 3D nucleation RHEED pattern quickly transitions to a streaky pattern, indicating the coalescence of Al islands and a shift to 2D growth for both InSb and InGaAs surfaces.

Surface morphology inspections with SEM and AFM of the samples grown at 0.1 Å/s reveal a similar discontinuous layer of Al islands, regardless of the semiconductor surface type. For InGaAs surface, faster deposition rates consistently result in a 2D Al layer. However, for the InSb surface, 2D morphology is preserved only for Al deposition on InSb with Sb-rich surface reconstruction, while growth on Sb-depleted reconstruction, despite showing 2D growth mode during deposition, was found to have a distinct 3D morphology when inspected with SEM. Most likely, the dewetting process took place at some point after finishing Al deposition, before moving the wafer out of the MBE system.

For both InSb and InGaAs surfaces, the 3D to 2D growth mode transition is abrupt and is governed by the interplay between the wafer’s thermal trajectory and the Al deposition rate.We will discuss the temperature evolution during and after Al deposition, as monitored with band-edge thermometry. Detailed surface morphology and interface characterization using SEM, AFM, and STEM will also be addressed during the presentation.

[1] A. Elbaroudy et al. “Observation of an Abrupt 3D-2D Morphological Transition in Thin Al Layers Grown by MBE on InGaAs surface” DOI: https://doi.org/10.1116/6.0003459.

MnTe is one of the 3D semiconductors that can exhibit anomalous Hall effect. The potential edge states correlated with the alter-magnet properties in the α-phase MnTe is also under study these days. The epitaxial growth becomes one method to tune the electronic structure. In this work we have grown both α-phase and β-phase MnTe by Molecular Beam Epitaxy on GaAs (111) and sapphire (0001) substrates with different buffer layers.

While in bulk crystal MnTe α-phase is the most stable state at room temperature, in the epitaxial structure β-phase MnTe can also be achieved in the as-grown thin films without post-growth annealing. On GaAs (111) substrates α-phase MnTe are naturally favored without any buffer layer. When using Bi₂Te₃ series TI (topological insulators) the buffer layers on sapphire (0001) substrates, we found out that β-phase MnTe are favored over pure Bi₂Se₃ due to the smaller lattice mismatch. However, when we add some alloy effect to the buffer layer, though the lattice mismatch is still smaller in β-phase, α-phase is actually grown. This unveil the role of the entropy effect and the changed in the surface potential. The nanorods structure in MnTe α-phase can also be controlled by buffer layer control and a CrSe_x layer under it.

Conventional optoelectronic devices face limitations when serving as a single detector for broadband and narrow-band applications due to their one-way photocurrent, restricting their potential uses in photodetection. However, a promising solution is emerging: bi-directional photocurrent switching. This advancement opens doors to unique functions like optical logic operations. We design an epitaxial GaN nanowall network-based ultraviolet bidirectional photocurrent photodetector. By introducing the different surface potential-induced photocurrent switching effects, the photocurrent direction can be switched in response to the wavelength of incident light at 0V bias. In particular, the photocurrent direction exhibits negative when the irradiation wavelength is less than 285 nm, but positive when the wavelength is longer than 285 nm. Special logic gates in response to different dual UV light inputs are proposed via a single bipolar PD, which may be beneficial for future multifunctional UV photonic integrated devices and systems.

Tin telluride (SnTe) is a narrow bandgap semiconductor which has many attractive properties, such as in-plane ferroelectricity, good thermoelectric performance, and a topological crystalline insulator (TCI) band structure.¹ In addition to these properties, SnTe also serves as an important buffer layer for developing Te-based heterostructures for magneto-transport devices.² Despite such promises, there have been few investigations focused on understanding the optimal epitaxial growth conditions for SnTe layers, which has led to poor surface morphologies and/or low crystallinity in synthesized films.³ A major challenge in the growth of thin film SnTe is the lack of suitable substrates, owing to the rock-salt crystal structure of SnTe and the large lattice mismatch with common substrates.

In this study, we report on the molecular beam epitaxy (MBE) growths of SnTe(111) layers on InP(111)A substrates. We have conducted a detailed investigation of the surface morphology and film crystal quality based on growth parameters including substrate temperature, Te/Sn flux ratio, and film growth rate. Despite the 7.4% lattice mismatch between the film and substrate, we found that a narrow substrate temperature range from 300 °C to 340 °C, a Te/Sn flux ratio of ~3, and a growth rate of 0.48 Å/s yield both smooth and single-crystalline SnTe(111) layers with thicknesses ranging from 10 nm up to 800 nm. Using these conditions, fully coalesced and smooth SnTe layers with root-mean-square (rms) roughness as low as 0.3 nm (Fig. 1a) can be obtained. The as-grown SnTe layer is free of rotational twin domains (Fig. 1b) and has exceptional crystal quality, including a full-width-at-half-maximum (FWHM) value as narrow as 0.06° from x-ray diffraction (XRD) rocking curves (Fig. 1c). Reciprocal space mapping confirms that thin (15 nm) SnTe layers are fully relaxed. Detailed transmission electron microscopy (TEM) imaging suggests that formations of In-Sn-Te nanoclusters during substrate annealing in a Te environment prior to growth aids strain relaxation as well as improves crystalline quality (Fig. 1d). Finally, preliminary angle-resolved photoelectron spectroscopy further indicates the three-fold symmetry of SnTe (111) layer at the Γ point, as well as Fermi level located 0.19 eV below the Dirac point. These promising results lay the foundation for employing SnTe on InP as a platform for developing all-telluride heterostructures integrated with III-V semiconductor devices.

[1]DOI:10.1063/5.0012300.

[2]DOI:10.1038/ncomms11623.

[3]DOI:10.1002/pssa.202200555.

As a long-known high performance epitaxial technique, Molecular Beam Epitaxy has experienced several major innovation across the past 50 years. Epitaxy recipes complexity has gradually increased to obtain advanced structures and devices thanks to the unique MBE capabilities. To cover this complexity, numerous monitoring probes and communicating devices (pumps, valves, gauges, thermocouples,…) have gradually been used on the reactors.

Nowadays, the MBE process is integrating additional class of optical in situ real time measurement techniques, including curvature / bowing monitoring.

Based on Magnification Inferred Curvature principle, Riber developed a dedicated curvature instrument on the basis of a CNRS-LAAS Toulouse patent. Named EZ-CURVE®, this instrument aims at giving insight in real-time about the growth process: it enables investigations at the growth mechanisms level for fundamental and applied research purposes, or alternatively, it can help evaluating process repeatability and stability without requiring detailed knowledge of all growth parameters.

This talk will give recent examples of EZ-CURVE uses’ cases for different class of materials (arsenides, antimonides, phosphides, others,…), and highlight the capability of such an approach to quickly converge on process optimization and monitoring.

Lanthanide-based nitride compounds are an understudied group of materials compared to lanthanide oxides. Their 4f electron shell gives rise to a variety of interesting physics such as unconventional superconductivity. Samarium nitride (SmN) has been recently identified as a material where ferromagnetic order and p-type superconductivity coexist. Our team will present results on the growth conditions of pure and doped SmN using molecular beam epitaxy, and its electronic transport properties as a function of temperature and magnetic field. The team used an yttria-stabilized zirconia (YSZ) substrate in the (001) crystallographic direction. After outgassing and removal of the native oxide, substrates were heated to 600°C and Sm and N were deposited for 60 minutes. Sm films deposited under similar conditions oxidized immediately upon contact with the air, while optimized SmN films avoided these samarium oxide peaks. Substrate temperature during growth of SmN films heavily influences the level of oxidation upon removal from the vacuum environment. The films have so far not been confirmed to be single crystalline due to the presence of SmN(111) peaks. AC magnetic susceptibility studies have shown the oxidized sample to be mostly paramagnetic, with a potential superconductive transition around ~10 K. Doping effects on crystal structure and electronic properties are characterized.

The atomic site-to-site correlations of a binary alloy can determine its optical and electronic properties. A direct bandgap is theoretically expected to occur in group IV alloys composed of silicon, germanium, and tin. However, an indirect bandgap is observed experimentally, which is thought to be caused by short range order (SRO). The atomic organization of thin films is determined by a combination of deposition rates, temperature, and templating effects of the substrate. The characterization needed for visual representation is made further arduous due to the atomic-scale nature of these factors. Here, using scanning tunneling microscopy (STM), we visualize step-by-step growth of Sn on Si in situ by separately controlling sub-monolayer growth and sample heating.

This work aims to understand the impact that SRO has on a material’s electronic proprieties through systematic characterization. While long-range ordering of monolayers of Sn has been observed on Si and Ge substrates in both 100 and 111 orientations, the impact of temperature and film thickness on SRO remain unknown. Additionally, surface segregation of Sn depends strongly on annealing, which is expected to influence SRO. We use in situ STM to analyze growth in a series of steps that include Sn deposition, Si deposition, and various annealing stages to probe the dynamics of thin film growth at the substrate interface. By discretizing the growth process, we create regions of metastable ordering that can be dynamically tuned throughout investigations, improving our ability to relate growth conditions to SRO and, correspondingly, to a material’s optical and electronic properties.

SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525

Extending the emission wavelengths of monolithically integrated nanowire (NW) lasers to longer wavelengths and even the mid-infrared (MIR) spectral range, shows great promise for optical on-chip communication and sensing applications using the mature silicon photonic circuit platform. Strong absorption within the visible range in silicon waveguides necessitates the integration of lasers with wavelengths in the near- or mid-IR for low-loss optical transmission. While examples of NW lasers have been shown with emission in the near [1] to mid-infrared [2], few operate under continuous wave operation necessary for on-chip processors.

In this work, we report on the mid-infrared continuous wave lasing of individual InAs nanowires at cryogenic temperatures. Figure 1a shows finite difference time domain simulations of the diameter dependent threshold gain, that allow us to determine the optimal nanowire geometry. Catalyst-free InAs NWs are grown site-selectively and with high homogeneity on SiO₂-templated Si(111) substrates via molecular beam epitaxy, whereby the diameter and length is tuned from 160-745nm and 6-28μm by varying the pitch and growth durations (Fig 1c). Under optical pumping with a 976nm continuous wave laser, stimulated emission is demonstrated for individual NWs transferred on a sapphire substrate with diameters of 745±55 nm and nanowire resonator lengths between 10-30 µm (Fig 1b). Typical lasing thresholds are found to range from 2-30 kW/cm² with emission wavelengths of 2.4-2.7μm (0.455-0.515eV).

⁺Corresponding Author Email: Steffen.Meder@wsi.tum.de

[1] P. Schmiedeke, et al., Appl. Phys. Lett. 118, 221103 (2021)

[2] H. Sumikura, et al., Nano Lett. 19, 8059 (2019)

[3] S. Meder, et al., in preparation (2024)

We have previously shown that AlN/Al_xGa_1-xN short period superlattices (SPSL’s) can be formed through the introduction of a constant Ga overpressure during the metal modulated epitaxy (MME) growth of AlN. It was found that when a growth temperature of 830 ^oC was used, the incorporation of Ga into the layer structure was limited due to the final stages of Al consumption due to Ga adatom desorption. As a result, the lowest Al composition in the Al_xGa_1-xN layers obtained was 72% for a Ga BEP of 1x10^-6 Torr. For this work, the impact of Ga adatom desorption on the formation of the AlGaN layer was investigated in order to realize the growth of AlN/GaN SPSL’s. In this study, a series of 3 growths was done where substrate temperatures of 800 ^oC, 760 ^oC or 730 ^oC were used. Outside of the variations in substrate temperature, Al, Ga and N fluxes were kept constant for each sample. Shutter conditions were chosen such that the Ga shutter opened only after complete consumption of Al on the surface had occurred. This removes any potential influence that Al adatoms have on Ga. Comparing XRD coupled scans of each sample showed that the 0^th order peak shifted to lower ω-2ϴ angle as the growth temperature was lowered. However, the position of the higher order peaks remained mostly unchanged. Therefore, while there was a reduction in the Al composition of the Al_xGa_1-xN layers, the overall SPSL period remained mostly unchanged. This was confirmed by STEM imaging, which found that for all 3 samples, the AlN had a thickness of ~ 5 nm while the Al_xGa_1-xN had a thickness of ~3 ML. By inserting the thickness values determined by STEM into XRD simulations it is possible to determine the Al content in the Al_xGa_1-xN layers. For the growth temperatures of 800 ^oC, 760 ^oC, and 730 ^oC, the resulting Al composition was calculated to be 84.3%, 66% and 52.5% respectively. These results represent the lowest Al content we have achieved using MME to date. However, the fact that no GaN formed serves to highlight that the current model explaining the formation of Al_xGa_1-xN is incomplete. Fully fleshing out this model will be the focus for continuation of this work.

GaSe is a layered semiconductor with potential applications for single photon emission and for ultrathin field effect transistors. The quality of epitaxially grown GaSe critically depends on the roughness and chemical termination of the starting surface. Previous work reports improved GaSe film morphology on GaAs (111)B using a pre-growth surface selenization treatment of the de-oxidized GaAs (111)B (1). Here I will show that an epitaxially grown buffer layer of GaAs (111) B in conjunction with the surface selenium treatment results improved GaSe film quality, as determined by X-ray diffraction, Raman spectroscopy, and cross sectional TEM.

This work was supported by NSF QLCI HQAN and ARL

(1)https://doi.org/10.48550/arXiv.2401.10425

Our study focuses on the growth of ErAs:InGaAlBiAs thin films using a digital alloy approach to achieve a bandgap of 0.8 eV suitable for telecom wavelength excitation (1550 nm). This semiconductor thin film is the active layer within a photoconductive switch (PCS) designed for terahertz (THz) pulse generation and detection. Key to the performance of such PCSs are several intrinsic semiconductor properties, such as dark resistivity, carrier lifetime, and carrier mobility. The incorporation of ErAs nanoparticles within the InGaAlBiAs matrix significantly influences these properties. Beyond the solubility limit, erbium incorporation leads to the formation of ErAs nanoparticles, resulting in a decrease in the material's carrier lifetime. Additionally, ErAs nanoparticles exert a pronounced pinning effect on the effective Fermi level of the material, thereby impacting carrier concentration and, consequently, dark resistivity. The size of these nanoparticles impacts the position of the Fermi level, a parameter controlled through interrupted growth and migration-enhanced epitaxy due to constraints posed by the bismuthide matrix, necessitating lower growth temperatures.

Despite the requirement for relatively low growth temperatures (~280°C) for the bismuthide matrix, this system offers enhanced flexibility in tuning bandgap and band alignment while ensuring the growth of high-quality lattice-matched films on an InP substrate. The ability to manipulate band alignment facilitates positioning the Fermi level deep in the bandgap, thereby reducing carrier concentration and elevating dark resistivity. Moreover, the positioning of the Fermi level with respect to the band edges plays a pivotal role in carrier lifetime, affecting the efficacy of nanoparticles in trapping free carriers.

Different characterization techniques are employed to assess these pertinent properties comprehensively. Dark resistance is determined through the Hall effect and Van der Pauw measurements, while carrier lifetime is evaluated via optical pump THz probe spectroscopy. Optical bandgap measurements are conducted using spectrophotometry, and material quality is scrutinized through high-resolution X-ray diffraction analysis.

This research was primarily supported by NSF through the University of Delaware Materials Research Science and Engineering Center, DMR-2011824.

Diluted magnetic III-N semiconductors (DMSs) are promising for spintronic devices. Incorporating Mn atoms induces ferromagnetic behavior in III-nitride materials, notably AlN. Doping atoms during molecular beam epitaxy (MBE) growth process can significantly impact film properties. This study delves into the growth technique of AlN:Mn films using alternating Al and Mn atom fluxes, a technique known as metal modulated epitaxy (MME).

Samples were grown on Si (111) substrates utilizing a 200 nm thick AlN buffer layer grown at 850°C (Figure 1(a)) by MBE. Al and Mn atom fluxes were supplied from Knudsen-type effusion cells. Active Nitrogen (N) flux was provided by a rf-plasma source operating at 150 W with a N₂ flow of 0.25 sccm. The first set of samples employed a continuous growth method, simultaneously supplying fluxes of Al and Mn atoms along with Nitrogen. The second set of samples was grown using an alternating method, with the Al shutter opened for 5 s followed by Mn shutter for 1 s, with a 3 s pause in between, while the Nitrogen flux was constant. The alternated shutters sequence during the growth process are depicted in Figure 1(b). The continuous growth utilized a growth temperature of 750°C, while the alternating method employed 720°C. In both samples, the Al flux was set at BEP_Al=2×10^-7 Torr and the Mn flux at BEP_Mn=5×10^-9 Torr.

Reflection High-Energy Electron Diffraction (RHEED) patterns revealed that continuous growth process yielded a surface with flat regions with some 3D features, evident from the combination of linear and spotty patterns (Fig. 2(a)). In contrast, the alternating growth process resulted in a completely spotty RHEED pattern, indicating nanostructure formation, as depicted in Fig 2(b). This observation was corroborated by AFM micrographs (Fig. 3), showcasing flat regions in the continuous growth method, and nanostructures formation for the alternating growth process. From AFM measurements we estimated the size of the AlN:Mn nanostructures of 30 nm in height with a base of 80 nm. Moreover, we confirmed that Mn atoms induced ferromagnetic behavior in the material. As illustrated in Fig. 4, magnetization curves exhibited clear hysteresis at 300K, with no diamagnetic signal in either case, and stronger magnetic moments observed in nanostructured samples.

In conclusion, supplying alternating fluxes of Al and Mn atoms facilitates the formation of AlN:Mn nanostructures exhibiting ferromagnetic behavior at room temperature, with higher magnetic moment intensity compared to samples prepared by the continuous growth process. Thus, an alternative approach for forming ferromagnetic nanostructures is presented.

Surface emitting lasers have become indispensable in our everyday life, with applications in a wide range of fields from optical communications and data transmission to medical diagnostics and therapeutics. While surface emitting lasers in the near-infrared (NIR) range are relatively mature, progress in the development of GaN based surface emitting lasers operating in the ultraviolet (UV) range has been sluggish. To mitigate various issues related to conventional vertical cavity surface emitting lasers (VCSELs) such as material quality, mirror reflectivity, and so on, nanowire photonic crystals (NPCs) have received increasing attention over the past few years for surface emitting lasers. NPCs offer a promising approach by simplifying the design of the lasing cavity in which utilizing the photonic band edge mode at the Γ point in a photonic crystal structure can produce optical gain through the formation of slow light and further achieve light emission form top surface when proper diffraction condition is satisfied. However, for these lasers, the formation of NPCs often requires selective area epitaxy (SAE) at high substrate temperatures on patterned substrates. Such a high substrate temperature could affect the NPC formation and lasing performance.

Herein, we demonstrate ultralow threshold surface emitting lasing in the UV range by exploiting molecular beam epitaxy (MBE) grown GaN NPCs using low-temperature SAE regime, which offers an unprecedented controllability of nanowire formation on patterned substrates (Figure S1a). Utilizing GaN NPC arranging in a square lattice with a center-to-center spacing (a) of 200 nm and a nanowire diameter (d_NW) of 173 nm, UV SE lasing at ~367 nm is measured with a threshold of merely 7 kW/cm² (Figure S1b), a 100× reduction compared to the conventional AlGaN UV VCSELs at a similar wavelength (Figure S1c). Further developments on such GaN based surface emitting UV lasers will also be reported in the conference.

We investigated the tradeoff between Hall sensitivity and frequency limit of an AlGaN/GaN two-dimensional electron gas Hall effect sensor. For this study, we utilized three different heterostructure designs that had variations in sheet carrier density, carrier mobility, sheet resistance, and capacitance. The heterostructure designs used for fabricating Hall sensors are capable of operating at high temperatures. The efficiency of Hall sensor quantified in terms of supply voltage-related sensitivity (SVRS) is 0.024, 0.044, and 0.051 T^-1, while the supply current-related sensitivity (SCRS) is 101, 48, and 67 VA^-1 T^-1, similarly supply power related sensitivity (SPRS) is measured to be 57, 92, and 103 VW^-1T^-1, at room temperature respectively. By varying the Hall device carrier velocity, carrier density, and inbuilt capacitance, we investigate in detail the tradeoff between Hall sensitivity and frequency limit in terms of Hall signal rise time and phase shift.In addition, we have proposed a method to address the frequency limitation that arises from the current spinning technique. This method involves measuring the induced voltage at the Hall measurement terminal, which results from the time-variable magnetic field, without applying any external bias to the Hall sensor.

Lasers based on self-assembled quantum dot (QD) gain media have attracted considerable attention due to their low sensitivity to operating temperature and record-low threshold current densities. InAs QD based edge emitting lasers (EELs), vertical cavity surface emitting lasers (VCSELs) or vertical external cavity surface emitting lasers (VECSELs) have been demonstrated with excellent performance. In this study, we demonstrate a QD based optically-pumped photonic crystal surface emitting laser (QD-PCSEL). The PCSEL fabrication process includes epitaxial regrowth, which enables the photonic crystal (PC) to be buried in the laser’s waveguide near the upper clad layers. The structure is grown using elemental source molecular beam epitaxy (MBE). In the first epitaxial step the bottom AlGaAs cladding layer and the GaAs waveguide including the QD active region are grown. In the next step, the wafer is removed from the MBE reactor and the PC layer is fabricated by electron beam lithography (EBL) patterning and etching into the GaAs waveguide using inductively coupled plasma (ICP) dry etching. Subsequently the top AlGaAs cladding layer and a top contact layer are grown on the sample. One of the most crucial steps is the removal of the native oxide before growing the top clad layer. This is accomplished by an acid etch prior to loading the sample into the vacuum chamber and a thermal surface treatment with arsenic supply prior to growth. The thermal step at >600^oC can alter the QD gain medium, i.e. cause a blue shift and narrowing of the emission spectrum of self-assembled QDs [1-3]. This study employs a dot-in-a-well (DWELL) design as the active region, where the InAs QDs are embedded in a InGaAs quantum well (QW). Our experiments indicate that the DWELL active region is stable during the regrowth process, i.e. there is no significant change to the emission wavelength. This is key to the realization of this laser. The PCSEL is tested by optical pumping and we present measurements of both light-input light-output (LL) and optical spectrum.

1. Photonics2018, 5(3), 27

2. Journal of Lightwave Technology, vol. 35, no. 20, 4547-4552, 2017

3. Crystal Growth & Design 2021, 21, 6,3521-3527

The development of sensors that operate in the mid (3-5 μm) and long (8-14 μm) wavelengths are essential for space-based sensing applications. Many Sb-based, III-V material solutions exist ranging from random alloys to superlattices. Furthermore, the growth of high-quality coherently-strained materials is required to achieve the desired long (1-10 μs) lifetimes for sensor applications. Yet, performance is limited by alloy and interface disorder and also band structure. The impact of disorder on the performance and band structure of MBE-grown InAs, InAsSb, InAs/InAsSb, and InGaAs/InAsSb is examined using the temperature and excitation dependent photoluminescence lineshape.

The excitation and temperature-dependent measurements are analyzed using a lineshape model that is sensitive to the band structure parameters, including bandgap energy, characteristic energy of Urbach tail states, and the electron-hole effective mass ratio and Coulomb interaction. The model also captures the quasi-Fermi level separation, which is the chemical potential of the photoexcited electron-hole population. As a result, the model predicts an excitation induced blue-shift in the lineshape peak position as the chemical potential approaches the bandgap energy. Furthermore, the lineshape model predicts the so-called S-shape behavior of the peak position as a function of temperature and excitation for materials with large Urbach tails. The S-shape behavior typically occurs at low temperatures, where in disordered materials, the width of the electronic Urbach tail states exceeds the width of the thermal Fermi occupation tail. This causes the lineshape to red shift as the electron-hole chemical potential is reduced and the occupied tail states dominate emission. As the temperature is reduced further, the chemical potential for a given excitation density increases, thereby blue-shifting the lineshape peak back towards the bandgap, thus completing the S-shape.

In the experiment, the various materials listed above are grown by MBE and examined using excitation and temperature dependent photoluminescence spectroscopy. The measurement temperatures range from 12 to 295 K and the pump powers range from 0.4 to 200 mW. The lineshape model is fit to the resulting families of spectra from each material. From the results, i) the lineshape peak position and the chemical potential of the photoexcited electron-hole population is determined as a function of both pump power and temperature, ii) the bandgap and Urbach tail width are determined as a function of temperature, and iii) the effective mass ratio is determined as a single global parameter for all temperatures and excitations measured.

In k-space, the iso-frequency surface of an electromagnetic wave in a hyperbolic material is a hyperboloid, in contrast to the spherical or elliptical surface for an ordinary dielectric material. This unusual property arises when one principal component of the dielectric tensor has the opposite sign to the other two principal components. Hyperbolic materials are expected to exhibit a range of distinctive behaviors, including negative refraction and enhanced superlensing effects [1]. Ionic crystals with anisotropic optical-phonon frequencies provide natural low-loss infrared hyperbolic resonances through the excitation of phonon polaritons. However, the operational bandwidth of these materials is limited to a few hundred wavenumbers (cm^-1) or tens of millielectronvolts. A promising route to a wider infrared bandwidth is the excitation of plasmon polaritons in homoepitaxial material engineered to have alternating layers of high- and low-carrier concentration [2,3].

In this work, we implement a low-loss Type-II hyperbolic metamaterial covering a wide spectral bandwidth of 2000 cm^-1 for wavelengths above 5.3 μm. We produced the hyperbolic metamaterial with a stack of intercalated heavily-doped InAs and undoped InAs epilayers grown by molecular beam epitaxy. Electron concentrations of 7×10¹⁹ cm^-3 were obtained by Tellurium doping of InAs epilayers and the optical properties of this stack were measured by infrared ellipsometry. These materials were then dry etched to form one-dimensional square gratings (with periods from 2 to 10 μm) and modeled by finite-element electromagnetic calculations (COMSOL). The models agree with measurements, showing the formation of hyperbolic plasmon polaritons at the same frequencies where experimental features were observed. Additionally, we have identified an Epsilon Near Zero mode associated with long-range surface plasmon polaritons contained in the dielectric layers. This work demonstrates that highly subdiffractional light confinement can be achieved with a III-V metamaterial that can be integrated with III-V semiconductor infrared devices such as photodetectors and emitters at a large scale.

This material is based upon work supported by the Office of the Undersecretary of Defense for Research and Engineering Basic Research Office STTR under Contract No. W911NF-21-P-0024. Disclaimer: The content of the information does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.

1. A. Poddubny et al., Nature Photonics7, 948 (2013).
2. D. Wei et al., Optics Express24, 8735 (2016).
3. Angela J. Cleri et al., Advanced Optical Materials11, 2202137 (2023).

There is a demand for mid-wave infrared (MWIR) photodetectors capable of detecting single photons for advanced industrial, military, and biomedical applications. This goal can be realized with cooled avalanche photodiodes (APD) operating in Geiger-mode for single photon counting. The development of the APDs with InAsSb absorption and AlGaAsSb multiplication regions grown lattice matched to GaSb was pursued. Al_0.9Ga_0.1As_0.08Sb_0.92 alloys are reported to have high hole impact ionization coefficients compared to electrons in these materials as well as greater than the corresponding coefficients in other semiconductors^[1]. Following the procedure in Ref. 1 we determined the hole and electron multiplication coefficients in Al_0.9Ga_0.1As_0.08Sb_0.92separately using p-i-n and n-i-p diodes within the field range of 140 to 600 kVcm^-1. The structures were grown in Veeco GEN-930 solid-source MBE system. The p-i-n diode structures for measurements of the hole impact ionization coefficient were grown on p-type GaSb substrates consisting of a 500-nm-thick Al_0.9Ga_0.1As_0.08Sb_0.92, Be-doped contact layer, a 300-nm-thick nominally undoped Al_0.9Ga_0.1As_0.08Sb_0.92hole multiplication region, and a 1.2-mm thick Al_0.9Ga_0.1As_0.08Sb_0.92Te-doped top contact layer capped with a 50-nm-thick Te-doped GaSb. The n-i-p diode structures for electron impact ionization measurements were grown on n-type GaSb with Te-doped, undoped and Be-doped Al_0.9Ga_0.1As_0.08Sb_0.92 layers of similar thicknesses. The wafers were processed into diodes with mesa diameters in the range from 80 mm to 320 mm. CV measurements confirmed that the 300-nm-undoped multiplication regions were fully depleted. Temperature dependences of I-V characteristics showed that above 140 K the dark current was limited by thermal generation of excess carriers in the depletion region with the activation energy of 0.3 eV. The carrier multiplication coefficients were determined by measuring photocurrent versus bias voltage. For a nearly complete absorption of photons in the top layer and multiplication dominated by one type of carriers the excess carriers were generated at 532 nm. Preliminary experimental data show that the hole ionization coefficients are up to 100 times greater than the electron ionization coefficients.

[1] Collins, X., et al. 2018. “Impact Ionization in Al_0.9Ga_0.1As_0.08Sb_0.92 for Sb-Based Avalanche Photodiodes,” Applied Physics Letters 112(2), 021103. DOI:10.1063/1.5006883.

Thermoelectric energy conversion offers widespread applications in waste heat recovery for power generation but requires suitable materials with large thermoelectric conversion efficiency. To aim at high efficiency with large figure-of-merit ZT=σS²T/κ, relevant parameters such as electrical conductivity (σ), thermoelectric power factor (σS²), and thermal conductivity (κ) need to be optimized for any given Seebeck coefficient S and operating temperature T. Due to the interdependence between thermal and electrical properties, modification of parameters for high ZT is limited in bulk materials, but in nanostructures, the properties can be decoupled^[1].

The aim of our research is to demonstrate independent control of these parameters by exploiting the 1D density of states in very high mobility III-V nanowires (NWs). First proof-of-principle studies were performed on modulation-doped GaAs-AlGaAs core-shell NWs grown by MBE, where we found enhanced thermopower in 1D-subbands at substantially lowered κ^[2]. While these initial studies were limited to cryogenic temperatures, studying more suitable low-bandgap III-V materials with lower electron effective mass are more appealing for enhanced 1D-thermoelectric transport at higher temperatures. Here, we show our recent progress towards InAs/AlAsSb core-shell NW heterostructures^[3], where both contact formation establishment and electrical characterization were performed up to elevated temperatures. Using NW-field effect transistor (NW-FET) test structures, the DC output characteristics were confirmed and transfer characteristics up to temperatures >100 K demonstrated clear 1D sub-band quantization, with transmission probabilities of ~0.4 for 700-nm long channel devices.

To further introduce n-type modulation-doping (Si-δ doping) into the InAs/AlAsSb core-shell NW-system, two strategies are illustrated: first, δ-doping of a very thin (<2nm-thin) InAs quantum well (QW) embedded in the AlAsSb shell to overcome the otherwise p-type nature of amphoteric Si-dopants in Al(As)Sb^[4]. Secondly, we also developed quaternary InAlAsSb shell layers to enable direct Si-δ doping without the need of InAs QWs. By guiding the design via band-profile calculations (nextnano-3), first results of the MBE growth and structural characterization of the desired InAs-InAlAsSb core-shell NW heterostructures are demonstrated^[5].

1. C. J. Vineis, et al., Adv. Mater. 22, 3970 (2010)
2. S. Fust, et al., Adv. Mater. 32, 1905458 (2020)
3. F. del Giudice, et al., Appl. Phys. Lett. 119, 193102 (2021)
4. C. R. Bolognesi, et al., IEEE Dev. Lett. 19, 83 (1998)
5. G. B. Hirpessa, et al., under preparation (2024).

Semiconductor nanowires (NWs) are promising materials for nanoscale devices with diameters of less than several hundred nanometers. GaAs has a high electron mobility and direct bandgap, and is used in lasers, solar cells, and transistors through heterojunctions with related compounds. In addition, the integration of III-V semiconductor NWs on Si can integrate superior electrical and optical properties to mature Si technology, which is expected to be applied to the next generation optical and electronic devices. However, the surface recombination velocity of GaAs is several orders of magnitude higher than that of other III-V semiconductors, and the effect is further exacerbated by the large surface-to-volume ratio in NWs. To suppress the effect of non-radiative recombination, surface passivation on the GaAs core with an AlGaAs shell is known to be effective.[1] Besides, preserved material properties at temperatures higher than room temperature is desirable for the practical device applications.

We here report wafer scale growth of GaAs/Al_0.8Ga_0.2As core–shell NWs on the 2-inch Si(111) wafer by constituent Ga-induced vapor–liquid–solid growth using molecular beam epitaxy (MBE).^[2]The sample is investigated by the optical reflectance, photoluminescence (PL) and time-resolved PL measurements from 300 to 400 K. The nature of the light scattering induced by the complex NWs structure on the Si substrate enables the measurement of diffuse reflectance, making the Kubelka-Munk (K-M) transformation^[3] to be applicable, obtaining the absorption characteristics of the sample. The sample shows a low optical reflectance less than 2% at energies higher than GaAs absorption edge in the visible to near-infrared region at 300 K, resulting in the dark observation of the sample wafer. That band gap of the sample is estimated to be 1.41 eV from the K-M plot, and it’s deviation between the PL peak energy suggests that the Stokes shift is close to be negligible. The carrier lifetime obtained from the time-resolved PL measurement is longer than 1 ns between 300 and 400 K. That indicates the effective passivation of the GaAs core by the AlGaAs shell, suppressing the surface non-radiative recombination. From these results, we obtain wafer scale GaAs/AlGaAs core-shell nanowires on 2-inch Si(111) substrate showing efficient light emission/absorption with high thermal stability.

Quantum optical metasurfaces (QOMs) hold promise for generating entangled photons and engineering complex quantum states. To generate entangled photon pairs, QOMs leverage spontaneous parametric downconversion (SPDC). Despite SPDC being the leading method for generating entangled photon pairs in quantum optics, QOMs dependent on SPDC face limitations due to low conversion efficiency. This project aims to enhance SPDC efficiency in QOMs by utilizing GaAs (111)B substrates for the growth of the bulk GaAs material from which the QOMs are fabricated.

GaAs is appropriate for QOMs because of its very high second-order susceptibility. It is predicted that the nonlinear response of GaAs in the (111) orientation is even stronger than it is in the typically used (001) orientation. Calculations indicate a potential 1-3 orders of magnitude increase (pump frequency dependent) in QOM conversion efficiency for GaAs with (111) surface orientation as compared to (001).

Growth on (111) oriented GaAs substrates is challenging and requires a very different set of growth parameters for optimized deposition than growth on (001) surface; therefore, precise in-situ monitoring and control are necessary.

In the pursuit of optimized growth parameters, we have grown GaAs and AlGaAs on GaAs (111)B substrates with an intentional 2° surface misorientation towards [11-2]. In-situ reflection high-energy electron diffraction (RHEED) is monitored to observe changes in surface reconstruction. Desorption mass spectrometry is used to monitor the group V/III ratio. A laser light scattering system has been implemented to measure the diffuse scatter from the wafer during the epitaxial process. The latter is a sensitive probe of the onset of roughening of the growth front.

Correlated metals with perovskite structure have generated much interest due to their intriguing combination of electrical and optical properties rendering them as alternative transparent conductor material to degenerately doped wide band gap oxide semiconductors, such as ITO. While SrVO₃ and CaVO₃ have been reported to outcompete ITO, SrNbO₃ has been shown unparalleled performance of electrical conductivity while maintaining a high optical transparency deep into the UV range to show unparalleled performance in the UV range. SrMoO₃ is yet another candidate material, which meets the requirement of the new transparent conductor design paradigm.

In this talk we will discuss the growth of SrMoO3 thin films by suboxide molecular beam epitaxy. We will show that optically transparent and electrically conductive SrMoO₃ films can be grown by supplying elemental strontium via a conventional effusion cell and thermally evaporating MoO₃ pellets as a molybdenum source. The direct supply of a molecular oxygen flux to the MoO₃ charge was found critical to prevent Mo reduction to lower oxidation states and to ensure congruent evaporation. Optimal growth conditions were found by varying the Sr to MoO₃ flux ratio resulting in SrMoO₃ films being optically transparent with transmission between 75 toand 91% throughout the visible spectral range, and electrically conducting with a room temperature resistivity of 5.0 × 10-5 Ω· cm.

Vertical-cavity surface-emitting lasers (VCSELs) have applications in facialrecognition in smart phones, 3D sensing for augmented reality/virtual reality, etc. The current VCSELs emitting at 940 nmfor cell phones are manufactured on GaAs substrates, utilizing the well-established GaAs/AlGaAs distributed Bragg reflectors (DBRs). However, long-wavelengths, 1300 nm to 1550nm,are desirablefor eye-safety and transparency to cell phone screens.

Here we explore strain-compensated GaAsP/InGaAs/GaAsSb W quantum wells (W QWs), grown by MBE, for long-wavelength emission on GaAs substrates. The tensile-strained GaAsP compensates for the compressive-strained InGaAs and GaAsSb layers. Three cycles (samples A-D) and one cycle (samples E and F) of GaAs_1-zP_z/In_yGa_1-yAs/GaAs_1-xSb_x/In_yGa_1-yAs/GaAs_1-zP_z ‘W’ QWs were grown withy = 0.3 and z = 0.3. The Sb composition (x = 0.06-0.21) of GaAs_1-xSb_x QW in all samples was controlled by the QW growth temperature in the range of 410-520 ºC. X-ray rocking curve (XRC) analysis and 20K photoluminescence (PL) measurement are performed.

The W QW with different Sb contents are modelled by the energy band-diagram simulations. The input parameters, QW thicknesses and material compositions, are extracted by the XRCanalysis. The one-cycle of 15 nm-GaAs_1-zP_z/3.5 nm-In_yGa_1-yAs/16 nm-GaAs_1-xSb_x/3.5nm-InyGa_1-yAs/15nm-GaAs_1-zP_zWQWstructureis designed as shown in Fig. 1. The strained bandgap energies of QWs are obtained from the bulk bandgap using various bowing parameters as the fitting parameters and energy shifts due to strain using various deformation potentials.

where b_i is the bowing parameters, and i = g, c, v_{hh, lh} for bandgap, conduction and heavy hole/light hole valence bands, respectively. The material parameters are taken from the literaturefor the end compounds of InAs, GaSb, GaP and GaAs. In our calculations, the bandgap bowing parameters b_g of GaAs_1-xSb_x and InGa_1-yAs_y are set as the fitting parameters to our experimental results. The experimental PL peak positions can be fit well by b_g of -1.58 and -1.1 for the strained GaAs_1-xSb_x and In_yGa_1-yAs QWs, respectively.The one-band time-independent Schrödinger equation with Hamiltonian in the effective mass approximation is solved by COMSOL Multiphysics, enabling us to obtain the electron and hole wavefunctions, the energy levels of electrons in conduction band and holes in valence band, and the estimated emission wavelengths.