Symposium EL04 : Emerging Chalcogenide Electronic Materials—From Theory to Applications

The search for reaction paths to new inorganic materials is a fundamental goal of chemistry. In contrast to solid-state methods, inorganic syntheses in liquid fluxes permit bond formation, framework assembly and crystallization at lower temperatures due to facile diffusion and chemical reactions with and within the flux itself. The liquid fluxes are bona-fide solvents similar to conventional organic or aqueous solvents. These reactions can produce a wide range of materials, often metastable, from oxides to intermetallics, but typically the formation mechanisms are poorly understood. In this talk I will describe how we design, approach, perform, observe, understand, and engineer the formation of compounds from inorganic melts. I will focus on how novel chalcogenides can form using the fluxes but also design concepts such as the “dimensional reduction” and “panoramic synthesis”. For example, in systems such as K-Cu-S and K-Sn-S compounds that span metallic and insulating behavior can be isolated. Common structural motifs within these materials systems belie structural precursors in the melt that may be controlled by tuning reaction conditions and composition. Using complementary techniques of in-situ x-ray diffraction we can create time-dependent maps of reaction space and probe the mobile species present in melts. An important link in our chemistry is the concept of a ‘functional group,’ a fragment of a few atoms that behave predictably when combined with other functional groups or reagents. When well-defined building blocks are present and stable in the reaction, prospects for increased structural diversity and product control increase substantially. In such tunable and dynamic fluxes, synthesis can be directed towards new materials with an astonishing variety of properties ranging from nonlinear optics to energy conversion.

The lacunar spinels AB₄Q₈ are compounds whose properties are dominated by tetrahedral clusters of B atoms and are exemplified by compounds such as GaV₄Se₈. Thi compound is known to undergo a polar distortion to a ground state structure in the R3m space group, and orders ferromagnetically with a relatively small magnetic moment. We develop an understanding into the relationship between crystal structure and magnetic order in this material, and the influence of electron correlations in establishing the observed ground state using first-principles density functional theory (DFT) electronic structure calculations. Because electrons are delocalized within V₄ clusters but are localized between them, the usual approaches to simulate electron correlations — such as the use of the Hubbard U in DFT + U schemes — do not adequately recreate the experimental ground state. Additionally, we find that magnetism and crystal structure are strongly coupled in this material, and only certain arrangements of magnetic moment within a V₄ cluster can stabilize the observed structural distortion. Once distorted, these compounds are capable of exhibiting skyrmionic ground states, and these are described in GaV₄Se₈ and some related compounds.

The chalcogenide lacunar spinel structures with composition AM₄Q₈ can host novel electronic responses, including metal-insulator transitions (MITs) for resistive random-access memory platforms and skyrmionic spin textures for novel spin-based logic. Indeed, many of the known compounds within the family are narrow-gap semiconductors that undergo structural, electrical, and magnetic phase transitions with external stimuli, e.g. pressure, electric pulse. Based on the current understanding of how local structure affects these functional responses, one may ask the naïve question of how to rationally select the chemistries for quaternary spinels to achieve improved performance and to what extent Vegard’s principle holds in ordered compositional variants. Here we utilize Bayesian optimization with Gaussian processes (BOGP) to identify quaternary lacunar spinels superior to their ternary counterparts. Our model successfully discovers most of the compounds on the Pareto front by exploring only less than 20% of the entire design space, leading to accelerated materials discovery via BOGP approach. We then perform density functional theory simulations to both validate the MITs in the predicated compounds and assess synthesizability, but also to expand our materials physics understanding. Our model presents an effective way for accelerated materials design and discovery in chemical composition space without using extra features (e.g. bond length, electronegativity). This method may be readily applied to other challenges in materials sciences and chemistry, especially where existing data is limited, and new data acquisition is costly.

High-entropy materials are formed by mixing typically five or more principal components into a single crystal structure, and show improved thermodynamic stability due to the large configurational disorder. This simple design principle has led to the discovery of a series of novel crystalline materials stabilized by the configurational entropy, from high-entropy metallic alloys, to entropy-stabilized oxides, carbides, and borides. While significant progress has been made to synthesize entropy-stabilized metals and ceramics for structural applications, little attention has been paid to the discovery of new semiconducting materials using the design principle of entropy stabilization.

Here, we present a new class of entropy-stabilized semiconducting alloys based on the IV-VI binary chalcogenides, namely Ge_xSn_yPb_1–x–yS_zSe_tTe_1–z–t high-entropy chalcogenides (HECs). By utilizing high-throughput first-principles calculations, we investigate the thermodynamic stability of HECs over their entire composition space, and show that more than 50% of the investigated compositions are stable with respect to phase segregation to the competing binary ingredients at the experimental synthesis temperature. We further studied the enthalpic effect of the individual elements via machine learning with Gaussian Kernel-Ridge Regression on the high-throughput data. We show that Sn and Se lower the enthalpy of mixing, while S is detrimental to the phase stability of HECs. We validate our theory prediction by synthesizing the equimolar GeSnPbSSeTe HEC using solid-state synthesis techniques. We find that the HEC crystalizes in a single-phase rocksalt structure upon fast quenching in liquid nitrogen, and shows a reversible phase transition in both DSC and temperature-dependent XRD, driven by the configurational entropy. In terms of their functional properties, equimolar HEC shows promising ambipolar dopability with Bi donors and Na acceptors, which opens up a wide range of possibilities for semiconducting electronic and energy applications. Our work demonstrates the potential of entropy stabilization in the discovery of novel multicomponent semiconductor alloys.

Engineering interfaces of topological materials with other classes of functional materials, such as oxides, ferromagnets, or superconductors, serves as the basis for designing novel electronic devices based on topological materials. Topological crystalline insulators (TCIs), such as SnTe, are a particularly promising class of materials, possessing the same properties as conventional topological insulators and controllable by electric fields when integrated with oxide field effect devices. This work presents growth and characterization of SnTe films thin enough for incorporation with other functional materials. X-ray diffraction and atomic force microscopy confirm we achieve single-domain and continuous SnTe film growth on SrTiO₃ substrates by integrating molecular beam epitaxy with net SnTe sublimation at raised substrate temperatures. By analyzing the hole carrier density, extracted from Hall measurements over a range of film thicknesses, we observe two-dimensional carriers confined to the SnTe/SrTiO₃ interface. Magnetoconductance measurements confirm the observation of two-dimensional transport and furthermore reveal that conduction is consistent with topological states. The growth of ultrathin TCI films on an oxide substrate presented here provides a foundation for designing additional TCI interfaces and engineering novel topological devices.

Unraveling structure-property relationships is often limited by the inability to test proposed relationship by synthesizing either isostructural compounds or homologous series of compounds. Major limitations to discovering these "missing" compounds with targeted structures include the lack of synthetic routes to compounds that are metastable and the challenge of predicting the energy landscape around local free energy minima to assess potential metastability. Traditional solid-state reactions are typically diffusion limited, producing thermodynamic products as a result of high reaction temperatures and long reaction times. Fluid phase approaches are nucleation limited with high diffusion rates enabling much of the energy landscape to be explored, but limited means to control what crystalizes. We will discuss a third approach, based on controlling the composition of an amorphous intermediate on the nanoscale. Nucleation is typically the rate-limiting step, but slow diffusion rates limit the extent that the energy landscape is explored. Hence the nanoarchitecture of the precursor is preserved, enabling many new compounds to be synthesized that are intergrowths of two targeted constituents. The periodic structure of the precursor enables a variety of analytical approaches to be used to determine the reaction pathway, total and local compositions, and the structure of interlayer stabilized structures. In parallel, we have developed a theoretical approach to probe the energy landscape around proposed target compounds. We create islands of different known structures of a particular stoichiometry between slabs of the other constituent structure and allow the system to relax. One of three limiting situations typically occurs. The interlayer atoms may react with the known structure, rearrange into a disordered "island of disaster" or they can maintain the initial structure with surface and edge distortions. These last cases, where the structures are preserved, become potential synthetic targets. This approach lets islands with structures or compositions not known as bulk compounds to be tested for stability. Several examples will be presented showing that predicted structures could be synthetically prepared.

Chalcogenide semiconductors constitute an important class of electronic materials with a broad range of crystal structures and a plethora of useful properties. Independent control of chalcogenide composition and structure holds a promise to design these properties for practical applications. Recently, we demonstrated synthesis of low-density polymorphs of chalcogenide semiconductors by heterostructural alloying in the MnTe_1-xSe_x materials system [1]. The resulting “negative pressure” MnTe_0.5Se_0.5 alloy has noncentrosymmetric wurtzite structure and piezoelectric response that is not present in either of the MnSe or MnTe parent compounds.
This presentation will focus on two additional methods to synthesize metastable wurtzite structure of manganese telluride (MnTe), as a prototypical chalcogenide material with known polymorphism. The first approach is alloying MnTe with a small amount (10-20 %) of ZnTe, which drives the structural phase transition similar to that in yttria-stabilized zirconia (YSZ) [2,3]. The second approach is templated growth on MnTe on 5 nm thin ZnTe seed layers deposited on amorphous glass substrates.[3] Both methods lead to metastable wurtzite polymorph of MnTe with very different properties comparted to the ground state nickeline structure, including 1 eV wider band gap and 1000−10 000 times lower electrical conductivity.
Overall, these results demonstrate new methods to synthesize metastable polymorphs of chalcogenide semiconductors, broadening the range of materials design methods for this class of compounds.
[1] S. Siol et al. Science advances, 4 eaaq1442 (2018)
[2] S. Siol et al. Journal of Materials Chemistry C, 6, 6297 (2018)
[3] Y.Han et al. The Journal of Physical Chemistry C, 122, 18769 (2018)
[4] Y.Han et al. under review

Optical phase change materials (O-PCMs) have found widespread adoption in photonic switches, reconfigurable meta-optics, reflective display, and optical neuromorphic computers. Current phase change materials, such as Ge-Sb-Te (GST), exhibit large contrast of both refractive index (Δn) and optical loss (Δk), simultaneously. The coupling of both optical properties fundamentally limits the function and performance of many potential applications. We introduce a new class of O-PCMs based on Ge-Sb-Se-Te (GSST) which breaks this traditional coupling, as demonstrated with an optical figure of merit improvement of more than two orders of magnitude. The optimized alloy, Ge₂Sb₂Se₄Te₁, combines broadband, low loss transparency (1 – 18.5 μm), large optical contrast (Δn = 2.0), and significantly improved glass forming ability, enabling an entirely new range of infrared and thermal photonic devices. In this talk, we will also review an array of reconfigurable photonic devices enabled by the low-loss O-PCM, including free-space light modulators, electrically switchable and nonvolatile reconfigurable metasurfaces, and transient couplers facilitating wafer-scale device probing and characterizations.

For the last 30 years, researchers in Southampton have been developing a novel family of chalcogenide glasses, driven initially by infrared optical applications. These materials, which include gallium and lanthanum sulphides and selenides, offer improved thermal and mechanical properties over commercially available chalcogenides. Today, the application space has shifted considerable and these same materials are increasingly being explored for electronic applications.

In this talk, we describe our work in this field, giving an overview of a glass fabrication and characterization before focusing on three specific devices under development; photonic crystals, flexible thermoelectric generators and electrochemical cells leading to new research in chalcogenide based solid-state batteries.

Three-dimensional complete photonic bandgap materials or photonic crystals block light propagation in all directions. The rod-connected diamond structure exhibits the largest photonic bandgap known to date and supports a complete bandgap for the lowest refractive index contrast ratio down to ∼ 1.9. We have confirmed this threshold by measuring a complete photonic bandgap in the infrared region in Sn–S–O (n ∼ 1.9) and Ge–Sb–S–O (n ∼ 2) inverse rod-connected diamond structures. The structures were fabricated using a low-temperature chemical vapor deposition process via a single-inversion technique.
Flexible thermoelectric generators (TEGs) can provide uninterrupted, green energy from body-heat, overcoming bulky battery configurations that limit the wearable-technologies market today. High-throughput production of flexible TEGs is currently dominated by printing techniques, limiting material choices and performance. Here we report on the properties of roll to roll sputtered Bi2Te3 films are reported and we demonstrate the ability to tune the power factor by lowering run times, lending itself to a high-speed low-cost process. To further illustrate our route to high speed printing, we fabricate a thermoelectric device using Virtual Cathode Deposition, a novel high deposition rate PVD tool, for the first time. This Bi2Te3/Bi0.5Sb1.5Te3 TEG exhibits S = 250 μV/K per pair and P = 0.2 nW per pair for a 20 °C temperature difference.
Finally we have recently fabricated electrochemical metallization cells using a GaLaSO solid electrolyte focussing on endurance testing. Using ITO and silver electrodes, up to 4000 cycles were measured, (limited by the measurement equipment) with some deterioration in the on/off ratio seen after 1000 cycles and we are currently optimizing performance through doping with Se.

Perovskite Chalcogenides are a new class of semiconductors, which have large chemical and structural tunability that translates to tunable band gap in the visible to infrared part of the electromagnetic spectrum. Besides this band gap tunability, they offer a unique opportunity to realize large density of states semiconductors with high carrier mobility. In this talk, I will discuss some of the experimental advances made both in my research group and in the research community on the theory, synthesis of these materials and understanding their optoelectronic properties.
Perovskite structure is composed of an octahedrally coordinated transition metal or main group element with anions such as oxygen, chalcogen or halogens. The octahedra is typically connected in the corners and the voids are filled by alkali, alkaline or rare earth elements. The valence and the size of the cations and anions can lead to different connectivity of these octahedra, which offers a knob to control both the chemical composition and the dimensionality of these materials. Moreover, the large number of elements in the periodic table can be accommodated in these extended perovskite and related structures, which allows us finer knobs to control the physical and chemical properties, in our case, we tailor light-matter interaction precisely over a broad energy range spanning the visible to infrared spectrum. We leverage this effect in early transition metal based perovskite chalcogenides and related phases to achieve properties such as highly anisotropic absorption and refraction (BaTiS₃, Sr_1+xTiS₃), unconventional band gap evolution (BaZrS₃ and Ba_n+1Zr_nS_3n+1 for n ≥ 1). Finally, I will provide a general outlook for future studies on these exciting new class of materials.

References:
S. Niu et al. Nature Photonics, 12, 392-396 (2018).
S. Niu et al. Advanced Materials 29, 1604733 (2017).
S. Niu et al. Chemistry of Materials, 30 (15), 4897-4901 (2018).
S. Niu et al. Chemistry of Materials, 30 (15), 4882-4886 (2018).

In today’s computing systems based on the conventional von Neumann architecture, there are distinct memory and processing units. Performing computations results in a significant amount of data being moved back and forth between the physically separated memory and processing units. This costs time and energy and constitutes an inherent performance bottleneck. It is becoming increasingly clear that for AI application areas, we need to transition to computing architectures in which memory and logic coexist in some form. Brain-inspired neuromorphic computing and the fascinating new area of in-memory computing or computational memory are two key non-von Neumann approaches being researched. A critical requirement in these novel computing paradigms is a very-high-density, low-power, variable-state, programmable and non-volatile nanoscale memory device. There are many examples of such nanoscale memory devices in which the information is stored either as charge or as resistance. However, one particular example are phase-change-memory (PCM) devices, which are very well suited to address this need, owing to their multi-level storage capability and potential scalability.
In in-memory computing, the physics of the nanoscale memory devices, as well as the organization of such devices in cross-bar arrays, are exploited to perform certain computational tasks within the memory unit. For instance, crossbar arrays of PCM or other memristive devices can be used to store a matrix and perform analog matrix-vector multiplications at constant O(1) time complexity without intermediate movements of data. I will present how computational memories accelerate AI applications and will show small- and large-scale experimental demonstrations that perform high-level computational primitives, such as ultra-low-power inference engines, optimization solvers including compressed sensing and sparse coding, linear solvers and temporal correlation detection [1-3]. Moreover, the efficacy of this approach to efficiently address not only inferencing but also training of deep neural networks is discussed.
However, enhancing the performance of the devices is crucial for increasing the range of targeted applications and building the next-generation of computational memory-based systems. For example, PCM devices exhibit a limited dynamic range and the conductance response is highly nonlinear and stochastic. Moreover, additional technology-specific device behavior such as the conductance response asymmetry and temporal evolution of conductance values pose significant challenges for PCM. One approach to address some of these challenges is advances in synaptic-cell architectures. We will show how by using multiple PCM devices as a single synaptic unit we can improve the conductance change granularity, as well as the effects of nonlinearity, asymmetry, stochasticity and drift of PCM cells[4]. A second approach is the so-called projected PCM concept that aims to decouple the device read out from the electrical properties of the amorphous phase, thus significantly reducing the drift, noise and temperature-sensitivity [5-6]. We will also elucidate implementation and functionality of this class of devices.
The results show that this co-existence of computation and storage at the nanometer scale could be the enabler for new, ultra-dense, low-power, and massively parallel computing systems. Thus, by augmenting conventional computing systems, in-memory computing could help achieve orders of magnitude improvement in performance and efficiency.
[1] A. Sebastian et al.,Nature Communications1115, 2017.
[2] M. Le Gallo et al., Nature Electronics, 246–253, 2018.
[3] S. R. Nandakumar, et al., ISCAS 2018.
[4] I. Boybat, et al.,Nature Communications, 2514, 2018.
[5] I. Giannopoulos, et. al.,IEDM, 2018.

Interfacial phase-change memory (iPCM) based on chalcogenide superlattices have lower switching energy requirements than conventional phase-change memory [1]. The reason for this is believed to be associated with the minimization of thermal losses in the superlattice during the switching process [1-4]. Recently, it was reported that iPCM devices based upon Ge-Te/Sb-Te superlattices exhibit differences from GeSbTe-alloy phase-change memory devices, such as resistance dynamics at elevated temperatures [5]. Furthermore, bipolar switching [6] and the dependence of the switching properties of iPCM devices on the presence of an external magnetic field was found [7]. The latter suggests that the presence of an external magnetic field can affect the structure of Ge-Te/Sb-Te superlattices. In the current work, we used pre-magnetized TbFeCo bottom contacts in iPCM devices in order to explore the effects of magnetic fields applied to Ge-Te/Sb-Te superlattice films during their growth with regards to the switching properties of the fabricated iPCM devices.
The iPCM devices fabrication process have been reported in ref. [8, 9]. The bottom contact was formed by a 50 nm TbFeCo layer grown at room temperature. The contact was magnetized by applying an external magnetic field (1.5 T) after deposition at room temperature. Lower temperatures were used (150°C) for the superlattice growth in order to prevent demagnetization of the TbFeCo contacts. Thus, the Ge-Te/Sb-Te superlattices were grown in the presence of the external magnetic field induced by the bottom contact and this condition was maintained during device characterization: R-V, R-I, and I-V curves were obtained.
Under normal conditions, Ge-Te/Sb-Te superlattice structures used as an active (programmable) medium in iPCM devices do not contain elements sensitive to magnetic fields. However, according to recent reports [7], an iPCM device in the presence of an external magnetic field during the switching process at elevated temperatures, could lead to the localization of spin-polarized electrons, which would in turn result in a significant change in the resistance of the device. Therefore, a superlattice film grown in the presence of a magnetic field could exhibit enhanced magnetoresistance effects, especially when the switching process is performed at elevated temperature. In this work, it is shown that iPCM devices grown in the presence of a magnetic field have switching characteristics different from conventional iPCM devices. For example, bipolar switching was achieved in DC mode, similar to previously reported results [6]. But in the short pulse switching regime, it was found that the RESET process shows a lower switching voltage than it is required for the SET process, which is opposite to the typical iPCM switching behavior.
In conclusion, we achieved enhancement of magnetoresistance effects in Ge-Te/Sb-Te superlattices by use of a novel iPCM device structure with pre-magnetized bottom contacts.
R. E. Simpson et al., Nat. Nanotech., 6 (2011) 501-505.
J. Tominaga et al., Adv. Mat. Interfaces, 1, (2014) 1300027.
A. V. Kolobov et al., APL Materials, 2 (2014) 066101.
X. Yu, J. Robertson, Sci. Rep., 5 (2015) 12612.
K. V. Mitrofanov et. al., Jap. J. of Appl. Phys. 57 (2018) 04FE06.
K. V. Mitrofanov et. al., Phys. Stat. Sol. RRL (2019).
J. Tominaga et al., Adv. Funct. Mater., 27 (40) (2017) 1702243.
Y. Saito et al., Phys. Stat. Sol. (b) 252 (2015) 2151–2158.
Y. Saito et al., AIP Advances, 6, (2016) 045220.

Phase change memory (PCM) is an emerging high speed, high density, scalable non-volatile memory that can be switched electrically between amorphous (high resistivity) and crystalline (low resistivity) phases of chalcogenides materials such as Ge₂Sb₂Te₅ (GST) [1]. The resistance of GST drifts spontaneously after amorphization following a power law [2]. This behavior limits the utilization of intermediate states for multi-bit storage due to potential overlap of different resistance levels. Multiple processes may be contributing to resistance drift, including structural relaxation, crystallization and charge trapping. It is possible to get a better understanding of the relative contributions of these various phenomena by performing measurements at cryogenic temperatures.
In this work, we have electrically characterized resistance drift of melt-quenched amorphous Ge₂Sb₂Te₅ (GST) line cells [3] in the 125K – 300 K temperature window with and without optical illumination using high resolution I-V sweeps. Each device is amorphized using a 500 ns pulse and drift is monitored at the same temperature up to 10⁴ seconds. We have found the drift coefficients to be approximately 0.1 in dark, with slightly decreasing trend with decreasing temperature. The experiments under light show a significant decrease in cell resistance and a distinctly different drift behavior from dark. Periodic switching of light shows upward resistance drift in dark and recovery under light with long time constants (> 10² s at 150K). The presence of resistance drift at cryogenic temperatures with similar drift coefficients as room temperature, as well as different behaviors with and without light, suggest that charge trapping is the dominant mechanism that contributes to resistance drift in amorphous GST.

References:
[1] H. P. Wong et al., “Phase change memory,” Proc. IEEE, vol. 98, no. 12, pp. 2201–2227, Oct. 2010.
[2] A. Pirovano, A. L. Lacaita, F. Pellizzer, S. A. Kostylev, A. Benvenuti, and R. Bez, “Low-field amorphous state resistance and threshold voltage drift in chalcogenide materials,” IEEE Trans. Electron Devices, vol. 51, no. 5, pp. 714–719, 2004.
[3] F. Dirisaglik et al., “High speed, high temperature electrical characterization of phase change materials: Metastable phases, crystallization dynamics, and resistance drift,” Nanoscale, vol. 7, no. 40, pp. 16625–16630, 2015.

Because of its fast operation speed, resistive random-access memory (RRAM) is widely anticipated to dominate future’s non-volatile memory (NVM) device technologies. However, the often high operating voltage window, the complex and hence expensive fabrication process are some of the major technological and economic obstacles for its commercialization. Here, we describe the development of solution processable copper sulfide materials and their application in high performance two-terminal NVM devices. We show that carefully controlled composition of copper sulfides (i.e. Cu₂S, Cu_2-xS and CuS) yields fast responsive memory device with remarkably low operating voltage window and highly reprogrammable operating characteristics. In particular, we demonstrate NVM devices with highly reprogrammable memory window of <300 mV, write/erase channel current ratio of >10³, excellent data retention and superior memory endurance in excess of 1500 write/erase cycles. Finally, we demonstrate successful resistive switching in planar two-terminal sub-20 nm scale NVM devices. Using complementary suite of characterization techniques in combination with density functional theory calculations, we are able to elucidate the nanoscopic origin of the resistive switching mechanism.

Phase change memories (PCMs) are based on the bad glass forming ability and metastability of the thermodynamic and kinetic transition in chalcogenide materials. This relies on the electrical and optical properties changing substantially when the atomic structure of the materials is altered. This transition, between a significant electrical resistance in the amorphous phase and a highly conductive state in the crystalline phase, has lent itself to numerous applications that include optical storage (i.e. blue ray and CDs) to electronic devices (i.e. Intel x-point technology). A novel subset of these materials uses the superlattice structure in order to greatly reduce the switching current and total energy required, thereby overcoming the joule heating constraint common to conventional PCMs. These are known as interfacial phase change materials (iPCM). Though currently unsettled as to the origins of the mechanism, they have shown promise for use in microwave devices based on interlayer switching by reducing the thermal loads required. In this presentation, we investigate the growth of interfacial GeTe-Sb₂Te₃ structures via Molecular Beam Epitaxy (MBE) with differing orientations and various substrates (GaAs, Si, Al₂O₃) and report on the electro-optical properties associated with the morphological and structural changes in this material system. By varying the elemental flux and novel heating method, we are able to stabilize the superlattice structure in a 2D growth regime. The ability to grow via MBE on transparent substrates allows us to incorporate the iPCMs into next-generation electronic and optical devices that may benefit applications such as computing, sensing and communications.

Chalcogenide alloys are the most common phase change materials used in phase change memory (PCM), which has entered the market as a high-speed non-volatile memory [1,2]. One of the main challenges for PCM is the large power required to heat the active region above crystallization or melting temperature. Lower energy and higher speed operations have been demonstrated with thin film superlattice stacks of phase change materials known as interfacial phase change memory (iPCM) [3-6]. The mechanisms behind the improved performance of iPCM are still under investigation but recent work indicates similar crystallization and melt-quench operation of these devices [5].

In this work, we perform electrothermal finite element simulations of reset and set operations on iPCM structures consisting of alternately stacked Ge₂Sb₂Te₅ (GST) and GeTe layers using COMSOL multiphysics [7-10]. Electrical pulses are applied for reset and set processes utilizing an internal circuit model where a transistor is used as an access device. Coupled electric current and heat transfer physics are employed to incorporate Joule heating and thermoelectric effects (Thomson heat within a single material and Peltier heat at material interfaces) with temperature dependent Seebeck coefficients, thermal conductivities, electrical resistivities, heat capacities and thermal boundary resistances (TBR) for each material / material pairs. Latent heat of fusion is included in the amorphous-crystalline and solid-liquid transitions [10], giving rise to heat release at the crystal-amorphous boundaries during crystal growth and heat absorption at the grain boundaries during amorphization. Grain boundaries and material interfaces have high energy sites, making them easier to melt, described as heterogeneous melting [10].

iPCM [5] structures utilize engineered interfaces formed between nanometer scale thin-film stacks, promoting amorphization through increased number of material interfaces and reduced thermal conduction due to TBR. Furthermore, the melting temperature, electrical conductivity and Seebeck coefficient of the different materials within an iPCM device differ. Hence, such layered structures may have the advantage of melting of only one of the alternating layers assisted by local heating or cooling due to Peltier effect at the interfaces.
Our results on iPCM and conventional PCM structures of same dimensions and geometry (20 nm wide, 150 nm high pore-cells) show ~ 50% reduction in reset times and more consistent set times for iPCM cells due to lesser variations in grain sizes and location of boundaries.

Acknowledgment: This work is partially supported by the National Science Foundation under award DMR-1710468.

References:
[1] S. W. Fong et al., IEEE Trans. Electron Devices 64, 4374 (2017).
[2] G. W. Burr et al., IEEE J. Emerg. Sel. Topics Circuits Syst. 6, 146 (2016).
[3] J. Tominaga et al., physica status solidi (RRL) 13, 1800539 (2019).
[4], K. V. Mitrofanov et al., Japanese Journal of Applied Physics 57, 04FE06 (2018).
[5] K. L. Okabe et al., Journal of Applied Physics 125, 184501 (2019).
[6] T. Ohyanagi et al., AIP Advances 6, 105104 (2016).
[7] Z. Woods et al., IEEE Trans. Electron Devices 64, 4466 (2017).
[8] Z. Woods et al., IEEE Trans. Electron Devices 64, 4472 (2017).
[9] J. Scoggin et al., Applied Physics Letters 112, 193502 (2018).
[10] J. Scoggin et al., Applied Physics Letters 114, 043502 (2019).

Material platform based on chalcogenide alloys, such as the recently emerged Ge-Sb-Se-Te, is uniquely poised for enabling new class of active meta-optical devices. Ge-Sb-Se-Te exhibits unique optical and structural properties, including high refractive index (3.3 - 4.7) in mid-infrared, low losses (k < 0.01), and the non-volatile phase-change ability with a significantly larger switchable thickness than the traditional Ge-Sb-Te alloys. Based on this material platform, we demonstrated all-dielectric phase-change metalens with a switchable focusing capability. This research finding opens up new avenues for further development of ultra-compact reconfigurable optical devices for thermal imaging and chemical sensing applications.

CdS exhibits large and persistent photoconductivity due to lattice relaxations associated with sulfur vacancies [1]. Sulfur vacancies are deep donors at equilibrium in the dark, but under photoexcitation they convert to shallow donors in a metastable state. This donor-level switching mechanism suggests a new way to control conductivity in electronic devices.

Here we demonstrate two-terminal electrical devices that exhibit donor-level switching. We fabricate and test vertical thin-film devices consisting of Ag/CdS/MoO₃/Au layer. MoO₃ is a hole-injection layer, due to its large electron affinity and nearly type-III alignment with CdS. We hypothesize that MoO₃ extracts electrons from the deep donor levels in CdS (possibly via the valence band), and thereby switches CdS into a highly-conductive state in a thin layer near the interface.

The hypothesis of electrically-stimulated donor-level switching (“photo”conductivity without photons) leads to several testable predictions, all of which are confirmed by device measurements. Our devices are in a high-conductivity state in equilibrium, and exhibit repeatable resistive switching with no electro-forming step. The active n-type layer is switched into a low-conductivity state by injecting electrons, and a high-conductivity state by injecting holes. These results support the hypothesis that the observed electrical switching behavior is due to donor-level switching, rather than mass transport (e.g. filament formation). We further support the donor-level switching hypothesis with numerical device simulations, pulsed electrical measurements, capacitance-voltage profiling, electron beam induced current (EBIC) measurements, X-ray absorption spectroscopy (XAS) mapping, and Raman microscopy. Spatially- and chemically-sensitive techniques suggest that no filaments are formed during switching.

Donor level switching represents a new mechanism for the design of electronic circuit elements such as variable resistors (e.g. for analog computing), ovonic switches, oscillators, and compliance limiters. We suggest that this mechanism is not unique to CdS, and can be found in any material with metastable charged defect states, including ZnO, CuInS₂, and AlGaAs. We end by discussing proposals to optimize defect level switching for particular applications by material selection and device design.

[1]. Han Yin, Austin Akey, R. Jaramillo, Phys. Rev. Materials 2, 084602 (2018).

Although the combination of RAM chips and Flash memory have been leveraged widely in big data and artificial intelligence tasks, they usually fail to meet the commercial demands of either speed or scalability. Phase-change memory materials may overcome these limitations, yet the typical crystallization time is many times slower compared to the melting time, and as a result is insufficient for any real-world application. Here we demonstrate how the crystallization kinetics of a wide range of phase-change systems can be controlled by using a single-shot treatment via “initial crystallization” effects. Ultra-rapid and highly stable phase-change structures viz. conventional and sub-10 nm sized cells, stackable cells, and multilevel configurations have been demonstrated. Material measurements and thermal calculations also reveal the origin of the pretreatment-assisted increase in nucleation and growth time and the thermal diffusion in chalcogenide structures, respectively. This insight may accelerate the technological deployment of new types of phase-change memory.

References
1) Wuttig, M., 2005. Phase-change materials: Towards a universal memory?. Nature materials, 4(4), p.265.
2) Wong, H.S.P., Raoux, S., Kim, S., Liang, J., Reifenberg, J.P., Rajendran, B., Asheghi, M. and Goodson, K.E., 2010. Phase change memory. Proceedings of the IEEE, 98(12), pp.2201-2227.
3) Loke, D. K., Lee, T.H., Wang, W.J., Shi, L.P., Zhao, R., Yeo, Y.C., Chong, T.C. and Elliott, S.R., 2012. Breaking the speed limits of phase-change memory. Science, 336(6088), pp.1566-1569.
4) Loke, D. K., Skelton, J.M., Wang, W.J., Lee, T.H., Zhao, R., Chong, T.C. and Elliott, S.R., 2014. Ultrafast phase-change logic device driven by melting processes. Proceedings of the National Academy of Sciences, 111(37), pp.13272-13277.
5) Loke, D. K., Shi, L., Wang, W., Zhao, R., Yang, H., Ng, L.T., Lim, K.G., Chong, T.C. and Yeo, Y.C., 2011. Ultrafast switching in nanoscale phase-change random access memory with superlattice-like structures. Nanotechnology, 22(25), p.254019.

In the switching of the Phase Change Memory based on the telluride chalcogenides, eg GST, the following two mechanism can be recognized [1,2]:
- reversible switching mechanism between an electrical low conducting state and high conductive one, named as threshold switching or Ovonic Threshold Switching (OTS). This mechanism is associated with an S-Shaped Negative Differential Conductance (SNDC);
- thermally activated phase transition between high conductive crystalline and low conductive amorphous states – named as Ovonic Memory Switching (OMS).

OTS is a fundamental mechanism for PCM devices . Despite intensive research into the role of the OTS mechanism, the nature of the physical principle responsible for the conductance switching is still not fully understood [1,4].
One of the reasons for the lack of understanding of the OTS mechanism are deficiencies in the understanding of SNDC mechanism.
The critical analysis of the current state of knowledge in the field of SNDC, in particular the impact of SNDC-related effects on the OTS process will be presented based on the developed phase change algorithm.

1.S.R. Ovshinsky, Phys.Rev.Lett. v. 21, p. 1450
2. A. Redaelli,et al., J.Appl.Phys.,2008, v.103, p.111101
3. A.Redaelli – ed., Phase Change Memory Device Physics, Reliability and Applications, Springer, 2018
4. V.G. Karpov et al., Appl. Phys. Lett. 92, 173501 (2008)

Phase change random access memory (PCRAM) is a promising candidate for next-generation non-volatile memory applications, which rely on reversible electrical resistivity change in telluride-based phase change materials between the crystalline and amorphous phases. Ge-Sb-Te alloys have been the most studied phase change materials since they were discovered for rewritable optical disc applications in the 1990’s. Essentially, the same material has been used in conventional PCRAM applications, but it suffers from both thermal stability problems and a limited cycling due to the large changes in density (volume) occurring during the phase change processes. These were not serious issues for optical discs applications, but are for PCRAM. In this work, using an empirical relation linking the physical characteristics of phase change materials, we predict novel materials that are expected to exhibit small density changes upon the phase transition.
The optical properties of 18 different transition-metal-based ternary tellurides have been systematically examined by density functional theory simulations. On the basis of the aforementioned empirical law, which correlates density change with the optical contrast between the amorphous and crystalline phases of chalcogenides, the density changes occurring upon crystallization for hypothetical materials have been predicted. Cr₂Ge₂Te₆ was found theoretically to be one of the best materials for phase-change memory applications owing to its small density change. This novel criterion will be useful in the screening of materials yet to be experimentally realized, and will help target promising materials for high-endurance phase-change memory applications [1].

[1] Y. Saito et al. Appl. Phys. Express, 12, 051008 (2019).

Phase change memories (PCMs) are based on the bad glass forming ability and metastability of the thermodynamic and kinetic transition in chalcogenide materials. This relies on the electrical and optical properties changing substantially when the atomic structure of the materials is altered. This transition, between a significant electrical resistance in the amorphous phase and a highly conductive state in the crystalline phase, has lent itself to numerous applications that include optical storage (i.e. blue ray and CDs) to electronic devices (i.e. Intel x-point technology). A novel subset of these materials uses the superlattice structure in order to greatly reduce the switching current and total energy required, thereby overcoming the joule heating constraint common to conventional PCMs. These are known as interfacial phase change materials (iPCM). Though currently unsettled as to the origins of the mechanism, they have shown promise for use in microwave devices based on interlayer switching by reducing the thermal loads required. In this presentation, we investigate the growth of interfacial GeTe-Sb₂Te₃ structures via Molecular Beam Epitaxy (MBE) with differing orientations and various substrates (GaAs, Si, Al₂O₃) and report on the electro-optical properties associated with the morphological and structural changes in this material system. By varying the elemental flux and novel heating method, we are able to stabilize the superlattice structure in a 2D growth regime. The ability to grow via MBE on transparent substrates allows us to incorporate the iPCMs into next-generation electronic and optical devices that may benefit applications such as computing, sensing and communications.

Ovonic Threshold Switch (OTS) is a critical component for suppressing the sneak current in the cross-bar array (CBA) phase change random access memory (PcRAM), and GeSe is a feasible functional material of the OTS. The conventional method to grow GeSe OTS film is sputtering[1,2]. However, the sputtering is not suitable for the vertically integrated PcRAM (V-PcRAM), which must be the ultimate integration structure of the CBA-PcRAM. For this V-PcRAM fabrication, a growth method which guarantees the thickness and composition uniformities within the deep hole structure is necessary. Atomic layer deposition (ALD), therefore, must be the process of choice for such a purpose, and the authors’ group already reported high functionality of ALD-GeSe film as a feasible OTS material [3]. However, the previous work has a limitation of the growth process; the ALD-GeSe film must be grown at a low temperature of 70 oC due to the serious desorption of the precursors at higher temperatures. This might also hinder the achievement of the ultimate functionality of the same material.
In this work, therefore, the ALD process was improved to allow the growth of the ALD-GeSe films at much higher temperatures (up to 180 oC) by changing the Ge(II)-precursor from HGeCl₃ to (Ge(iPrN)₂CNMe₂)NMe₂ and co-injecting NH₃ with the new Ge(II)-precursor. Under this condition, the (Ge(iPrN)₂CNMe₂)NMe₂ was activated by accepting the lone pair electrons of NH₃, which allowed the growth of the GeSe film at temperatures as high as 180 oC. The activated Ge(II)-precursor molecules feasibly reacted with the Se-precursor, (Si(CH₃)₃)₂Se, resulting in the stoichiometric GeSe thin film deposition. The ALD process showed self-limiting growth behavior and produced highly uniform and stoichiometric GeSe films with the saturation growth rate around 39 ng*cm²*cy^-1 at 70 oC substrate temperature (Figure 1). X-ray photoelectron spectroscopy analysis revealed that the composition ratio of Ge:Se was ~1:1 for all the ALD temperature from 70 to 180 oC. However, the N impurity level in the film varied depending on the substrate temperature; film grown at higher temperature contained lower N- impurity concentration (Figure 2 (a), (b)). This may cause discrepancies in the density and morphology of the deposited film. When the film was grown at 180oC, a high film density (4.23g*cm-3) and low N impurity content could be obtained. The presentation will also report the high electrical performance of the film as the OTS using an integrated cell structure of mushroom type, where the 10nm thick GeSe film was interposed between the planar-TiN top electrode and W-plug (diameter of 2um) bottom electrode.

A chalcogenide material, Ge₂Sb₂Te₅ (GST), can undergo a fast nanosecond-transition between the amorphous and crystalline phases that have a resistivity difference of 1,000 times or more. This phase change can be controlled electrically and is used as phase change memory (PCM). However, amorphous and crystalline phases have a density difference of about 10% and cause a mechanical stress in the device due to the volume change during the switching operation. Repetitive switching operations between two phases lead to an open circuit failure, more specifically, the breakdown of the interface between GST and its contact electrode by the fatigue, which is called stuck reset. In this study, finite element analysis (FEA) was performed for several PCM cell architectures by combining electrothermal and solid mechanics models using COMSOL Multiphysics, a commercial FEA software. Simulation techniques and comparative study on the cell architectures will be presented.

Chalcogenide alloys are the most common phase change materials used in phase change memory (PCM), which has entered the market as a high-speed non-volatile memory [1,2]. One of the main challenges for PCM is the large power required to heat the active region above crystallization or melting temperature. Lower energy and higher speed operations have been demonstrated with thin film superlattice stacks of phase change materials known as interfacial phase change memory (iPCM) [3-6]. The mechanisms behind the improved performance of iPCM are still under investigation but recent work indicates similar crystallization and melt-quench operation of these devices [5].

In this work, we perform electrothermal finite element simulations of reset and set operations on iPCM structures consisting of alternately stacked Ge₂Sb₂Te₅ (GST) and GeTe layers using COMSOL multiphysics [7-10]. Electrical pulses are applied for reset and set processes utilizing an internal circuit model where a transistor is used as an access device. Coupled electric current and heat transfer physics are employed to incorporate Joule heating and thermoelectric effects (Thomson heat within a single material and Peltier heat at material interfaces) with temperature dependent Seebeck coefficients, thermal conductivities, electrical resistivities, heat capacities and thermal boundary resistances (TBR) for each material / material pairs. Latent heat of fusion is included in the amorphous-crystalline and solid-liquid transitions [10], giving rise to heat release at the crystal-amorphous boundaries during crystal growth and heat absorption at the grain boundaries during amorphization. Grain boundaries and material interfaces have high energy sites, making them easier to melt, described as heterogeneous melting [10].

iPCM [5] structures utilize engineered interfaces formed between nanometer scale thin-film stacks, promoting amorphization through increased number of material interfaces and reduced thermal conduction due to TBR. Furthermore, the melting temperature, electrical conductivity and Seebeck coefficient of the different materials within an iPCM device differ. Hence, such layered structures may have the advantage of melting of only one of the alternating layers assisted by local heating or cooling due to Peltier effect at the interfaces.
Our results on iPCM and conventional PCM structures of same dimensions and geometry (20 nm wide, 150 nm high pore-cells) show ~ 50% reduction in reset times and more consistent set times for iPCM cells due to lesser variations in grain sizes and location of boundaries.

Acknowledgment: This work is partially supported by the National Science Foundation under award DMR-1710468.

References:
[1] S. W. Fong et al., IEEE Trans. Electron Devices 64, 4374 (2017).
[2] G. W. Burr et al., IEEE J. Emerg. Sel. Topics Circuits Syst. 6, 146 (2016).
[3] J. Tominaga et al., physica status solidi (RRL) 13, 1800539 (2019).
[4], K. V. Mitrofanov et al., Japanese Journal of Applied Physics 57, 04FE06 (2018).
[5] K. L. Okabe et al., Journal of Applied Physics 125, 184501 (2019).
[6] T. Ohyanagi et al., AIP Advances 6, 105104 (2016).
[7] Z. Woods et al., IEEE Trans. Electron Devices 64, 4466 (2017).
[8] Z. Woods et al., IEEE Trans. Electron Devices 64, 4472 (2017).
[9] J. Scoggin et al., Applied Physics Letters 112, 193502 (2018).
[10] J. Scoggin et al., Applied Physics Letters 114, 043502 (2019).

Non-volatile memories are pivotal components for emerging opto- and nanoelectronic applications such as smart displays and brain-inspired neuromorphic computing. Flash memories are currently the key player in the field, yet they fail to meet the commercial demands of scalability and speed. Phase-change materials (PCM), based on reversible switching between the crystalline state and amorphous state of a chalcogenide material, showing fast switching on 10 ns time scale, are promising alternatives explored by the industry. However, segregating-binary-alloy (SBA)-based PCM tends to show high consumption of energies and segregation of elemental components. Nanowire-like PCM has so far evolved as the most ideal candidate to achieve low-energy consumption PCM, however they are often synthesized by vapor-liquid-solid methods above 720 K, which would cause irreversible corruption of SBA-based PCMs. Here we leverage M13 bacteriophage as biological template to control the structure and assembly of SBA PCMs. We demonstrate a biologically templated, low temperature, wire-like PCM simply by leveraging the binding affinity of the negatively-charged amino acids on M13 bacteriophage surface to the precursor components of SBA-type germanium-tin-oxide systems. This study may open intriguing perspective for realizing genetically-engineered PCM with tunable, reliable and fast switching characteristics.

1) Wuttig, M. (2005). "Towards a universal memory?" Nature Materials 4(4): 265-266.
2) Loke, D., T. Lee, W. Wang, L. Shi, R. Zhao, Y. Yeo, T. Chong and S. Elliott (2012). "Breaking the speed limits of phase-change memory." Science 336(6088): 1566-1569.
3) Loke, D., J. M. Skelton, W.-J. Wang, T.-H. Lee, R. Zhao, T.-C. Chong and S. R. Elliott (2014). "Ultrafast phase-change logic device driven by melting processes." Proceedings of the National Academy of Sciences 111(37): 13272-13277.
4) Dang, X., H. Yi, M.-H. Ham, J. Qi, D. S. Yun, R. Ladewski, M. S. Strano, P. T. Hammond and A. M. Belcher (2011). "Virus-templated self-assembled single-walled carbon nanotubes for highly efficient electron collection in photovoltaic devices." Nature Nanotechnology 6: 377.
5) Nam, Y. S., A. P. Magyar, D. Lee, J.-W. Kim, D. S. Yun, H. Park, T. S. Pollom Jr, D. A. Weitz and A. M. Belcher (2010). "Biologically templated photocatalytic nanostructures for sustained light-driven water oxidation." Nature Nanotechnology 5: 340.

The lattice engineering of the sulfide based transition metal dichalcogenide (TMDs) is difficult compared to selenide and telluride materials.1 Here, we report on the continuous phase transition of the few layers to bulk molybdenum disulfide (MoS2) system by realizing the hetero-structures with the low work function (2.6 eV) two dimensional electride (2D-Ca2N) that contain highly mobile electrons2, where the large work function differences (>2 eV) between them is the key of the degenerate electrons doping that is distributed over a distance of few tens of nanometer from the contact interface in contrast to the other surface limited (~1 nm) chemical functionalizations or ionic gating approaches.3 An electron doping density of ~1014 cm-2 was estimated on the MoS2 layers that performed exceptional layer thickness dependent lattice symmetry change from the 2H to 1T′ phases until a few-layer (~10 nm) and then strong doping effect to the bulk samples resulting a giant band gap re-normalization by ~200 meV along with the softening of the commonly observed Raman modes by the Δω=10 cm-1. As the MoS2 thickness was decreased, well defined Raman peaks of 2H crystal were gradually disappearing along with the emergence of other 1T' phase Raman modes, thus elucidating the symmetry change feature due to concentrated charge density in a thinner film. Additionally, a hetero-structure of the few-layer MoS2 demonstrated multiple in-plane anisotropic Raman modes, including those of metallic phases Raman modes along with the presence of the hugely shrunk (250 meV) optical gap and enhanced PL intensity than a pristine monolayer MoS2; a realization of the few-layer 2H-MoS2 crystal with distorted symmetry and direct band gap structure when placed on the top of the Ca2N. This is attributed to the long-range electron doping induced structural change across the K-Γ line, hence provided an opportunity to discover various kinds of 2D materials using a single 2D-dopant suitable for the future optoelectronic application.
KEYWORDS: MoS2, electride, doping, long range, phase-transition, direct band gap structure

References
Duerloo et al., Nat. Commun., 2014, 5, 4214.
Lee et al., Nature, 2013, 494, 336-340.
Kim et al., Nano Lett., 2017, 17, 3363-3368.

Improvements in actual computing performance, which is based on von Neumann architecture, requires designing novel scalable, fast and energy-efficient technologies with both processing and storage capabilities. A typical example of systems with such capabilities is the human brain. Indeed, it has major advantages over supercomputers in terms of energy efficiency and adaptability due to its complex processing capability, made possible by the interconnection between ≈ 10¹¹ neurons and 10¹⁵ synapses. The current ''holy grail'' of neuromorphic computing technology is thus to mimic brain performance for the development of superior computational systems. Among the proposed next generation device candidates to reproduce neuromorphic core architectures for pattern recognition and machine learning, redox-based resistive switching (RS) random access memories (RRAM) are particularly highly regarded due to their predicted high memory density, energy efficiency and speed. Within their metal−insulator−metal architecture, these devices store binary code information using the electric field-induced resistance change of the insulating oxide layer by conductive filament (CF) formation and rupture at chemical and structural defects. Despite such very attractive properties, a lack of control on the location and spacing of CFs formation, as well as on their properties, results in stochasticity and randomness of CFs formation in these devices, which in turns leads to detrimental variation in device performance and lack of reliability in training processes. Unfortunately, there is actually a lack of fundamental knowledge connecting these defects to their nucleation and evolution. Within this context, we initiated a study on the effect of strain on the microstructure, chemistry and RS properties of TiO₂ thin films to resolve the thermodynamic and kinetic relations of heterogeneities to their intrinsic and local RS properties.

Epitaxial TiO₂ films with varying thicknesses were pulsed laser deposited on Nb-SrTiO₃ and Nb-TiO₂ single-crystal substrates to investigate the effect of phase, strain, crystallinity and defects concentration on the nature of the CFs and the local RS properties of the films. The fabrication and characterization of TiO₂-based two-terminal devices was also carried out to get information about their RS properties, mainly in terms of switching consistency and reproducibility. Such studies will not only allow gathering insights into defects location of formation but will further serve as a guideline while looking to introduce chemical disorder at specific locations into the TiO₂ matrix. Such a selective doping along defects is expected to locally decrease the reduction enthalpy and provide preferential sites for CF formation. The objective here is to achieve proper control of CFs formation and properties to accordingly reduce the variability among RRAM devices and improve their switching symmetry.

There has been a growing interest in using specialized neuromorphic hardware to implement artificial neural networks for use in applications such as image and speech processing. These neuromorphic devices show promise for meeting the significant parallel computing demands of such applications with higher speed and lower power consumption than software-based implementations. One approach to achieving this goal is through oxide thin film resistive switching devices arranged in a crossbar array configuration. Resistive switching can mimic several aspects of neural networks, such as short- and long-term plasticity, via the dynamics of switching between multiple analog conductance states--dominated by the creation, annihilation and movement of ionic species within the film (such as oxygen vacancies). These processes can be stochastic in nature and contribute significantly to device variability, both within and between individual devices.
This study focuses on reducing the variability of the set/reset voltages using model systems of HfO₂ grown on Nb:SrTiO₃ and Si/TiN substrates. We control the oxide microstructure via growth paramters and thermal treatment, to compare amorphous HfO₂ films with nanocrystalline columnar-grained films with both single (-111) texture and multiple orientations and selectively dope grain boundaries using thermal diffusion. By statistically comparing the electrical characteristics of at least 10 devices from each processing condition, film microstructure and dopants may be optimized for maximum resistive switching repeatability. Because the device requirements for practical resistive switching arrays are significant, controlling the variability of individual devices will likely be a consideration for every fabrication and processing step. This work provides a significant step towards understanding the mechanisms behind device variability and achieving devices that meet the strict requirements of neuromorphic computing.

Analog resistive switching processor arrays are promising as in-memory computing hardware and to construct physical neural networks capable of implementing machine learning algorithms at higher operating speed and lower energy cost than conventional von-Neumann architectures. State-of-the-art analog resistive switches rely on the mechanism of either forming conductive filaments or inducing a phase change. These processes suffer from poor repeatability or high energy consumption, respectively. A fundamentally different mechanism is desired to address these challenges.

Herein we demonstrate proton intercalation induced resistive switching, that can be applied on a variety of layered or channeled materials including chalcogenides and metal oxides. A prototypical three-terminal resistive switching device was designed, with a channel of active material probed by two terminals (source and drain), a proton reservoir layer as gate terminal, and a proton conducting solid electrolyte separating the two. The electrical conductivity of the active material can be modulated by the precise control of proton intercalation, providing a highly reproducible resistive switching behavior. As the lightest cation, the shuffling of protons between the reservoir and the active material requires minimal energy.

A proof-of-concept device has been fabricated. 50 nm WO₃ was deposited by reactive sputtering as the active material, with a channel size of 100 μm x 500 μm. A 400 nm Nafion polymer electrolyte layer was deposited by spin coating. A 50 nm Pd electrode serving as both gate and solid proton reservoir was sputtered on top and later hydrogenated to palladium hydride. The energy consumption is only 3.5 fJ/μm² per state per potentiation and depression operation, comparable with state-of-the-art demonstrations. Extreme on/off ratio of more than 10⁷ in conductance was observed and more than 1000 distinct conductance states were demonstrated. Most important of all, a symmetric potentiation and depression behavior was achieved in a constant current pulse mode. Further, insight into the switching mechanism involving protons has been obtained through a combined examination of crystal and electronic structure using synchrotron X-ray absorption spectroscopy and in-situ X-ray diffraction. It is found that protons serve as donors to reduce the metal oxidation states in the active material and create gap states. The continuous filling of electrons into the empty d-orbital of W metal increase the charge carrier density and conductivity. Additionally, proton intercalation induced lattice distortion of the active material is found to result in higher crystal symmetry and better alignment of W-O₆ octahedrals. This also contributes to the enhanced conductivity after proton intercalation.

In summary, we successfully proved the feasibility of utilizing proton intercalation chemistry to induce the modulation of resistivity and prototyped an all-solid three-terminal analog resistive switch. The advantages of this switching mechanism, namely, low energy dissipation, good operating symmetry and multi-state capability, are demonstrated.

The charge transfer and spin coupling effects are explored at the interface of two-dimensional (2D) superconducting FeSe nanosheets and one-dimensional (1D) molecular photochromic potassium-7,7,8,8-tetracyanoquinodimethane (KTCNQ).^[1] 2D superconducting FeSe layer is obtained in a large scale by liquid exfoliation and maintains the characteristics of its bulk counterpart.^[2] Light-induced current density in 2D FeSe nanosheets device is enhanced by the electron doping from KTCNQ by the destabilized spin-Peierls phase^[3] through their interface. The spin coupling at the interface of FeSe and KTCNQ shifts the dimerization transition temperature of KTCNQ from 389 K to 395 K. Our results suggest 2D exfoliated FeSe nanosheet as a versatile strongly correlated platform for the electron doping and inorganic-organic interface studies.


[1] X.-L. Mo, G.-R. Chen, Q.-J. Cai, Z.-Y. Fan, H.-H. Xu, Y. Yao, J. Yang, H.-H. Gu, Z.-Y. Hua, Thin Solid Films 2003, 436, 259-263.
[2] F.-C. Hsu, J.-Y. Luo, K.-W. Yeh, T.-K. Chen, T.-W. Huang, P. M. Wu, Y.-C. Lee, Y.-L. Huang, Y.-Y. Chu, D.-C. Yan, M.-K. Wu, Proc. Natl. Acad. Sci. USA 2008, 105, 14262-14264.
[3] J. G. Vegter, T. Hibma, J. Kommandeur, Chem Phys Lett 1969, 3, 427-429.

Chalcogenide perovskite-type materials have recently been investigated as alternative solar absorbers to the familiar oxide ferroelectric and halide perovskite families. In particular, Ba₃Zr₂S₇ (BZS) is a Ruddlesden-Popper chalcogenide, which was shown to exhibit a band gap of approximatley 1.3 eV [1] that is ideally suited for solar absorption, and thermal stability up to at least 550 °C [2]. Knowledge of the thermal expansion of Ba₃Zr₂S₇ is also important during operation and fabrication of a photovoltaic device based on this material. Here we compute the temperature-dependent structural, lattice dynamical, and thermal expansion properties of Ba₃Zr₂S₇ using the self-consistent quasi-harmonic approximation [3] within the framework of density functional theory. We explain the mechanism of thermal expansion, report its effect on the band gap, and compare this behavior to other perovskite-structured oxides and chalcogenides materials.

[1] Niu, S.; Sarkar, D.; Williams, K.; Zhou, Y.; Li, Y.; Bianco, E.; Huyan, H.; Cronin, S. B.; McConney, M. E.; Haiges, R.; Jaramillo, R.; Singh, D. J.; Tisdale, W. A.; Kapadia, R.; Ravaichandran, J. Chem. Mater. 2018, 30 (15), 4882-4886
[2] Niu, S.; Milam-Guerrero, J. A.; Zhou, Y.; Ye, K.; Zhao, B.; Melot, B. C.; Ravichandran, J. Journal of Materials Research 2018, 33 (24), 4135.
[3] Huang, L-F.; Lu, XZ; Rondinelli, J. M. Computational Materials Science 2016, 120, 84-93.



N.Z.K. and J.M.R. were supported by U.S. Department of Energy (DOE) under Grant No. DE-SC0012375 and the National Science Foundation’s MRSEC program (DMR-1720319) at the Materials Research Center of Northwestern University, respectively. R.A.K. was also supported by the National Science Foundation’s MRSEC program (DMR-1720319) at the Materials Research Center of Northwestern University. A.B.A. was supported by AFOSR in the form of a PECASE (FA9550-17-1-0247).

Complex oxides in perovskite and Ruddlesden-Popper crystal structures often have large, interesting, and useful dielectric response. Complex sulfides in similar structures may feature similarly large dielectric response while also featuring band gap in the visible and infrared, and likely have higher charge transport mobility than complex oxides. Complex chalcogenides may therefore be an uncommon class of highly-polarizable semiconductors.

We report measurements of the dielectric susceptibility of single crystal samples of BaZrS₃ (BZS-113) and Ba₃Zr₂S₇ (BZS-327). BZS-113 has a distorted-perovskite structure and a band gap of 1.8 eV. BZS-327 is a Ruddlesden-Popper variant and has a band gap of 1.3 eV. We use impedance spectroscopy to measure the complex dielectric response of single crystal and cold-pressed pellets over the frequency range 0.1 Hz – 1 MHz. We find that both phases are highly-polarizable, with room-temperature static dielectric constant (ε₀) over 50 and dielectric loss tangent below 0.05. We also report temperature-dependent susceptibility and electrical resistivity and discuss the effects of incipient ferroelectricity, predicted for chemically-adjacent Sr-Ba-Zr-S compounds. The combination of strong and low-loss dielectric response and band gap in the VIS-NIR spectral range may be unique for inorganic semiconductors and is interesting for applications including photonics and solar energy conversion.

Cubic pi-phase monochalcogenides (MX, M = Sn, Ge; X = S, Se) are an emerging new class of materials that have recently been discovered. Here, their thermodynamic stability, progress in synthetic routes, properties, and prospective applications are reviewed. The thermodynamic stability is demonstrated through density functional theory total energy and phonon spectra calculations, which show that the pi-phase polytype is stable across the monochalcogenide family. To date, only pi-phase tin monochalcogenides have been observed experimentally while pi-phase Ge-monochalcogenides are predicted to be stable but are yet to be experimentally realized. Various synthetic preparation protocols of pi-SnS and pi-SnSe are described, focusing on surfactant-assisted nanoparticle synthesis and chemical deposition of thin films from aqueous-bath compositions. These techniques provide materials with different surface energies, which are likely to play a major role in stabilizing the pi-phase in nanoscale materials. The properties of this newly discovered family of semiconducting materials are discussed in comparison with their conventional orthorhombic polymorphs. These could benefit a number of photovoltaic and optoelectronic applications since, apart from being cubic, they also possess characteristic advantages, such as moderately low toxicity and natural abundance</div>

The discovery of superconductivity in the two-leg ladder compound BaFe₂S₃ has established the 123-type iron chalcogenides as a novel and interesting subgroup of the iron-based superconductor family. However, in this 123 series, BaFe₂Se₃ is an exceptional member, with a magnetic order and crystalline structure different from all others. Recently, an exciting experiment reported the emergence of superconductivity in BaFe₂Se₃ at high pressure [J. Ying et al., Phys. Rev. B 95, 241109(R) (2017)]. Here, our analysis unveils a variety of qualitative differences between BaFe₂S₃ and BaFe₂Se₃, including in the latter an unexpected chain of transitions with increasing pressure. First, by gradually reducing the tilting angle of iron ladders, the crystalline structure smoothly transforms from Pnma to Cmcm at ∼6 GPa. Second, the system becomes metallic at 10.4 GPa. Third, its unique ambient-pressure Block antiferromagnetic ground state is replaced by the more common stripe (so-called CX-type) antiferromagnetic order at ∼12 GPa, the same magnetic state as the 123-S ladder. This transition is found at a pressure very similar to the experimental superconducting transition. Finally, all magnetic moments vanish at 30 GPa. The information obtained in our calculations suggests different characteristics for superconductivity in BaFe2Se₃ and BaFe2S₃: in 123-S pairing occurs when magnetic moments vanish, while in 123-Se the transition region from Block- to CX-type magnetism appears to catalyze superconductivity.

Recent experimental results indicate that palladium chalcogenides present interesting electronic behavior. In particular, superconductivity appears at 4.5 K in PdTe at ambient pressure. Similarly, PdS becomes a superconductor (T_C = 1.8 K) for pressures starting at 19.5 GPa [1]. As the pressure increases so does the T_C reaching a value of 7 K at a pressure of 41.3 GPa. Recently we have predicted the superconducting transition temperature (T_C) of the Wyckoff phase of bismuth as 1.5 mK or below [2], and this was later experimentally corroborated. Afterwards we calculated T_C for all bismuth solid phases under pressure and our results agree with experiment [3]. Here, we report an ab initio computational study of the effects of pressure on palladium sulfide (PdS), maintaining the original crystalline structure and compressing a 108-atom crystalline supercell computationally, as previously conducted for bismuth [4]. We calculate the electronic and vibrational density of states for the PdS system under pressure and relate them with its superconducting properties. We compare our results with the experimental ones reported in the literature.

REFERENCES:

[1] L.C. Chen, et al. J. Phys-Condens. Mat. 30(15):155703 (2018).
[2] Z. Mata-Pinzón, A.A. Valladares, R. M. Valladares and A. Valladares, PloS One, 11(1):e0147645 (2016).
[3] I. Rodríguez, D. Hinojosa-Romero, A. Valladares, R. M. Valladares, A. A. Valladares, Sci. Rep. 9(1):5256 (2019).
[4] D. Hinojosa-Romero, I. Rodríguez, Z. Mata-Pinzón, A. Valladares, R. M. Valladres and A. A. Valladares, MRS Adv. 2(9):499-506 (2016).

Bismuth, as a pure monatomic material, is a semimetal which possesses several interesting physical properties, not all of them well understood. The prediction of a critical superconducting temperature for its crystalline form at ambient pressure by our group [1] and its experimental verification several months later, have led us to ab-initio study bismuth in different conditions: amorphous [1], bi-layered [2], under-pressure phases [3] and mixed with copper [4]. In all of these systems, bismuth seems to contribute to their superconductive behavior and the modification of their critical temperature. Here we study the bismuth-sulphur chalcogenide which is known to contribute to superconductivity in the layered compound LaO_1-xF_xBiS₂ which has a T_C as high as 10.6 K. Using ab initio molecular dynamics and the undermelt-quench method, we generate layered and amorphous structures of the BiS₂ system to study its electronic and vibrational properties and then relate them to possible superconducting properties. We shall report these results and study similarities and differences for the amorphous and layered structures and infer its relevance for superconductivity. We shall compare these results with those found for the bulk amorphous bismuth, the bi-layered, the under-pressure phases and mixed with copper, that we have recently calculated and reported.
REFERENCES:
[1] Z. Mata-Pinzón, A. A. Valladares, R. M. Valladares, A. Valladares. Superconductivity in Bismuth. A New Look at an Old Problem. PLoS ONE 11 (2016), e0147645. DOI:10.1371/journal.pone.0147645.
[2] D. Hinojosa-Romero, I. Rodriguez, A. Valladares, R. M. Valladares, A. A. Valladares. Possible superconductivity in Bismuth (111) bilayers. Their electronic and vibrational properties from first principles. MRS Advances 3 (2018), pp. 313-319. DOI:10.1557/adv.2018.119.
[3] I. Rodríguez, D. Hinojosa-Romero, A. Valladares, R. M. Valladares, A. A. Valladares. A facile approach to calculating superconducting transition temperatures in the bismuth solid phases. Scientific Reports 9 (2019) 5256 DOI:10.1038/s41598-019-41401-z
[4] D. Hinojosa-Romero, I. Rodriguez, A. Valladares, R. M. Valladares, A. A. Valladares. Ab initio Study of the Amorphous Cu-Bi System. MRS Advances 4 (2019), pp. 81-86. DOI:10.1557/adv.2019.83.

The electronic and optical properties of CuM³⁺O₂ delafossites have been thoroughly investigated for numerous applications. A notable feature of these materials is a symmetry disallowed VBM to CBM transition often resulting in an optical band gap significantly larger than the smallest direct band gap [1]. CuGaO₂ has application potential as a hole transport layer for perovskite solar cells [2], and as part of hybrid electrodes for various photoelectrochemical cells [3][4]. CuInO₂ is known for its unusual bipolar dopability considering its wide optical band gap [1].

R. Nagarajan et al. successfully synthesized a set of 2+/5+ mixed B-site antimony based analogues [5] opening up a new materials space free of the relatively scarce and increasingly in demand gallium and indium. Of these analogues two earth abundant variants, Cu₃Mg₂SbO₆ and Cu₃Zn₂SbO₆ , were found to have ordered crystal structures. In this study hybrid density functional theory with the HSE06 functional is used to predict the electronic structure, optical properties, and band alignments of the Mg and Zn variants. We compare and contrast these results with those of the known Ga and In based delafossites, and speculate on their potential applications.

[1] Nie, X.; Wei, S.-H.; Zhang, S. B. Phys. Rev. Lett. 2002, 88 (6).
[2] Zhang, H.; Wang, H.; Chen, W.; Jen, A. K.-Y. Adv. Mater. 2016, 29 (8), 1604984.
[3] Kumagai, H.; Sahara, G.; Maeda, K.; Higashi, M.; Abe, R.; Ishitani, O. Chem. Sci. 2017, 8 (6), 4242–4249.
[4] Choi, M. U.; Hayakawa, T. Mater. Res. Bull. 2019, 113, 84–89.
[5] Nagarajan, R.; Uma, S.; Jayaraj, M. K.; Tate, J.; Sleight, A. W. Solid State Sci. 2002, 4 (6), 787–792.

Manufacturing of paper, which started two thousand years ago, simplified all aspects of information technology: generation, processing, communication, delivery and storage. Similarly powerful changes have been seen in the past century through the development of integrated circuits based on silicon. In this talk, I will discuss how we can realize these integrated circuits thin and free-standing, just like paper, using two-dimensional materials based on transition metal dichalcogenides and how they can impact the modern information technology.
In order to build these atomically thin circuits, we developed a series of chemistry-based approaches that are scalable and precise. They include wafer-scale synthesis of three atom thick semiconductors and heterojunctions (Nature, 2015; Science 2018), a wafer-scale patterning method for one-atom-thick lateral heterojunctions (Nature, 2012), and most recently, atomically thin films and devices that are vertically stacked to form more complicated circuitry (Nature, 2017). Once realized, these atomically thin circuits will be foldable and actuatable, which will further increase the device density and functionality.

Two-dimensional architectures are considered excellent candidates for electronics and electrochemical applications. Some of the most commonly researched 2D layered materials are: graphene, boron nitride, black phosphorous, and transition metal chalcogenides (TMCs). TMCs are promising high-performance materials for next-generation energy storage systems because of their high theoretical capacity, high energy density, and high voltage. Among them, copper-based chalcogenide nanocrystals and related alloys have been widely investigated due to their non-toxicity, low cost, and ability to achieve band-gap energies of 1.0–1.5 eV [1-2].
Copper selenide has attracted a lot of attention because of its abundant applications in photoelectric devices, medical treatments, gas sensors, and catalysis. Copper selenides exist as a variety of stoichiometric and nonstoichiometric phases such as CuSe, Cu₃Se₂, Cu₅Se₄, Cu_1.8Se, Cu₂Se, CuSe₂, and Cu_2−xSe [4]. Among them, Cu_2−xSe shows unique structures and properties. Cu_2−xSe has Se atoms in a simple face-centered cubic (FCC) structure with the space group Fm-3m, but the superionic Cu ions are kinetically disordered throughout the structure, resulting in a quite high electrical conductivity. Thus, researchers concentrated their synthesis of Cu_2−xSe with different sizes and shapes through various techniques.
Here, we report the synthesis of bulk 2D layered copper selenide using atmospheric pressure chemical vapor deposition (AP-CVD), with elemental selenium as a precursor. Using top-down approaches (drop casting, etc.), we deposited mono/few-layer flakes on substrates. The layered material was characterized using XRD, HRSEM, TEM, UV-VIS, and AFM to confirm its morphology and stoichiometry. We identified our material as a modulated structure based on the FCC basic structure (Fm-3m), similar to reported in[5]. It should be noted that due to structural modulation, structure characterization was possible here through electron diffraction only which allowed to conclude modulation vector. AFM measurements demonstrated that the single flakes had a lateral size of approximately 3-6 µm. HRSEM images of the dropcasted material on Si/SiO₂wafer showed copper selenide flakes on top of each other and the presence of layers has also been confirmed by TEM. We will show that this material is a potential candidate for electronic and electrochemical applications.
References:
[1] Wang et.al, Adv. Mater. 2018, 1801993
[2] Chen et.al, Cryst. Eng. Comm., 2014, 16, 2810
[3] Chen et.al, J. Mater. Chem., 2011, 21, 3053
[4] Lee et.al, J. Phys. Chem. C, 2017, 1219, 5436-5443
[5] Vucic et.al, Phys. Rev. B, 1981, 24, 5398

2D metal chalcogenide semiconductors have received considerable attention due to their compelling properties and layered crystal structure. Much of this work has focused on transition metal dichalcogenides (TMDs, MX₂ where M=Mo, W and X=S, Se, Te), but there is growing interest in group III (Ga, In) and group IV (Sn, Ge) chalcogenides to expand the suite of layered materials. Much of the research to date has been carried out using flakes exfoliated from bulk crystals but techniques for wafer-scale epitaxial growth of single crystal monolayer and few layers films are rapidly developing.

Our work has focused on the epitaxial growth of 2D layered chalcogenides using gas source chemical vapor deposition (CVD) also referred to as metalorganic CVD (MOCVD). The process is carried out in cold wall reactor geometries at moderate pressures (100-700 Torr) using hydrides (H₂Se, H₂S) as the chalcogen source in a H₂ carrier gas. In the case of TMDs, a multi-step method involving modulation of the metal hexacarbonyl precursor source flow rate was developed to independently control nucleation and lateral growth of domains at elevated temperature (>700^oC) which is beneficial to enhance the surface diffusivity of transition metal adatoms. Using this approach, uniform, coalesced monolayer and few-layer TMD films (MoS₂, WS₂, MoSe₂ and WSe₂) were obtained on 2” sapphire substrates at growth rates on the order of ~1 monolayer in 10-60 minutes. In-plane X-ray diffraction demonstrates that the films are epitaxially oriented with respect to the sapphire. Post-growth dark-field transmission electron microscopy carried out on monolayers transferred from the sapphire reveals that the films consist of highly ordered micron-sized single crystal regions bounded by low angle grain boundaries. Epitaxial growth of indium selenide was also investigated using trimethyl indium (TMIn) and H₂Se in H₂. In this case, lower growth temperatures (<500^oC) and lower reactor pressures (100 Torr) were necessary to reduce the extent of gas phase pre-reaction of precursors and obtain a stoichiometric film. Under these conditions, epitaxial growth of beta-In₂Se₃ was demonstrated on c-plane sapphire and (111) Si substrates with both island growth and step-flow growth modes observed. Prospects for utilizing MOCVD for the growth of epitaxial 2D heterostructures will also be discussed.

Two-dimensional (2D) transition metal dichalcogenides (TMDs) have demonstrated enormous potential in optoelectronic, catalytic, and device studies. Although “top-down” approaches can be used to define crystal morphologies and dimensions, the ability to tailor the physical properties of TMD crystals through explicit synthetic control is a major challenge. We introduce a gas-phase synthesis method that significantly transforms the structure and dimensionality of MoS₂ crystals without lithography. Synthesis of MoS₂ on Si (001) surfaces pre-treated with phosphine (PH₃) furnishes high aspect ratio (~1:20) nanoribbons of crystalline 2H phase MoS₂. The widths of the MoS₂ nanoribbons can be systematically controlled in the range of 70 – 500 nm by varying the concentration of PH₃ gas introduced during the Si (001) surface treatment step. Detailed kinetic studies along with adsorption energy calculations indicate that the nature and concentration of the underlying Si–phosphide moieties are instrumental in transforming growth of MoS₂ from a conventional 2D triangular to quasi-1D morphology. Notably, the room temperature photoluminescence (PL) of 1D MoS₂ crystals exhibits an emission peak which is blue-shifted by 50 meV relative to that of 2D MoS₂ nanocrystals. Extensive structural, electronic, and optical characterization of these new materials was carried out through a combination of electron microscopies (ac-STEM, HAADF, EELS), scanning probe microscopies (AFM, NSOM), optical spectroscopies (confocal Raman, PL), and electrical transport measurements.

Two dimensional layered materials offer remarkable properties that are not found in their three-dimensional counterpart. Given its simplicity mechanical exfoliation is conventionally applied to prepare these material systems in the monolayer limit. However, since the materials are not very robust, but rather mechanically soft and chemically reactive in ambient atmosphere – making them prone to a sizeable defect formation during the top-down preparation method – the utilization of a bottom-up approach is highly desirable to study the intrinsic properties in the monolayer limit. Molecular beam epitaxy (MBE) is an ideal method of choice; however the nucleation and growth of layered chalcogenide materials encompasses a specific set of challenges.
In this talk the growth kinetics specific to chalcogenide thin films is discussed and contrasted to the growth of other materials by MBE. Emphasis is placed on experiments to determine the temperature dependent sticking coefficient of the volatile chalcogenide element using a heated quartz crystal monitor to derive growth conditions to access self-regulated growth. Strategies to suppress the formation of twin domains during the layer nucleation stage and strategies to promote the growth of large single domains to achieve high quality monolayer and few layer materials are discussed using the specific examples of PtSe₂ on sapphire, Te on TiO₂ and FeSe on SrTiO₃. Emphasis will be placed, how in-situ spectroscopic ellipsometry can be used as complimentary diagnostic tool to RHEED to optimize growth conditions and achieve high quality layered chalcogenide-based materials.

The lack of large-area synthesis methods for emerging 2D materials presents a significant challenge for nanoelectronic devices and systems. In particular, InSe is a semiconducting van der Waals (vdW) material that possesses exceptional band-gap tunability as a function of thickness in the ultrathin limit. Despite its desirable electronic properties, InSe is a relatively under-investigated 2D material, primarily due to the fact that high-quality samples have only been achieved via mechanical exfoliation as opposed to large-area ultrathin-film growth. While many of the extensively studied and synthesized vdW 2D material systems have relatively simple phase diagrams (e.g., MoS₂), the complex phase diagram for InSe has hindered the development of large-area films, thus motivating more fundamental studies aimed at characterizing and understanding growth mechanisms.

Towards that end, we use diverse surface characterization techniques (e.g., X-ray photoelectron spectroscopy, Raman spectroscopy, atomic force microscopy, transmission electron microscopy, and selected area electron diffraction) and in operando X-ray diffraction to study the structural and compositional evolution of ultrathin InSe films grown by pulsed laser deposition with subsequent vacuum thermal annealing. By monitoring the post-deposition annealing temperatures, we rationally determine the synthesis conditions to realize ultrathin InSe films with high uniformity over large areas, controlled thickness, and no detectable impurities. Using this optimized method, ultrathin InSe films are patterned for the fabrication of top-gated field-effect transistors that demonstrate homogenous device behavior. Thus, our work provides a pathway to large-area ultrathin InSe films with high crystalline quality, thickness tunability, and generalizability to a wide range of substrates suitable for nanoelectronic applications.

Gallium sulfide is a III-VI semiconductor with multiple stable phases including (layered) hexagonal and monoclinic, GaS and Ga₂S₃, respectively. The layered GaS crystal structure is composed of Ga-Ga and Ga-S covalent bonds that extend in two dimensions while interlayer bonding occurs by van der Waals interaction. With a room-temperature band gap of 2.6 eV, GaS has been used for a variety of applications requiring ultrathin layers including transistors and photodetectors and is an attractive candidate for flexible electronics. On the other hand, Ga₂S₃ exists mainly in the monoclinic form with covalent bonds in three dimensions (i.e., not layered) and a reported band gap of ~3.0 eV.

Thin films of GaS and Ga₂S₃ have been grown by a variety of methods including chemical vapor deposition, chemical vapor transport, chemical bath deposition. Each method has advantages and challenges. Until now, pulsed-laser deposition (PLD) has been considerably less explored, yet possesses characteristics that are attractive for the synthesis of layered materials. Specifically, PLD is a physical vapor process that enables the “digital” deposition of target materials due to the use of a pulsed laser as an energy source for the evaporation of source materials.

We demonstrate the selective, pulsed-layer deposition of hexagonal GaS and monoclinic Ga₂S₃ epitaxial films on sapphire substrates from a single Ga₂S₃ target under high vacuum conditions. Growth with substrate temperatures between 400 °C and 550 °C results in the formation of GaS films, indicating non-stoichiometric transfer from target to film (i.e., loss of sulfur presumably due to its high vapor pressure). Surprisingly, the stoichiometric transfer occurs for substrate temperatures above 650 °C but with a lower growth rate under otherwise the same growth conditions as those used to grow GaS. In the temperature window between 550 °C and 650 °C, films displayed a mixture of phases which is consistent with a monotonic decrease in the Ga:S ratio in films from 1:1 to 2:3. By changing the substrate temperature in this narrow growth range, we are able to synthesize heterostructures of GaS and Ga₂S₃, thereby providing the unprecedented opportunities to realize heterojunctions from a single materials system in a single growth process and to investigate the heteroepitaxy of two-dimensional crystals on three-dimensional crystals and vice versa.

Materials synthesis and characterization were supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Materials Sciences and Engineering Division, under contract DE-AC02-05CH11231 within the Electronic Materials Program (KC1201). K.E. acknowledges support from SUMCO Corp., Japan.

The large elastic strains that can be sustained by van der Waals bonded layered materials (e.g., TMDs) cause significant modification of the electronic band structure through modulation of bond distances and crystal symmetries. This manifests as measureable variations in the electronic and optical properties [1,2]. In addition to static (i.e., time-invariant) strain, ultrafast transient strains can be induced in such materials via femtosecond (fs) photoexcitation. For example, in-plane atomic motions and out-of-plane interlayer coupling in MoS₂ have been studied with ultrafast reciprocal-space electron and X-ray diffraction [3,4]. Similarly, bright-field imaging in ultrafast electron microscopy (UEM) has been used to study the nucleation, propagation, and relaxation of coherent elastic strain waves in a number of TMDs arising from the release of impulsive local strain mediated by defects [5-7].

Here, we report the direct imaging with UEM of photoexcited localized anisotropic coherent acoustic-phonon dynamics in single, freestanding, multilayer MoS₂ flakes. Combining picosecond temporal resolution and nanometer spatial resolution in bright-field imaging, we observe and quantify both the in-plane nucleation and propagation of acoustic phonons and an interlayer low-frequency breathing mode spatially isolated and localized within two regions of different layer number and along two distinct crystallographic zone axes. Transient structural information induced via fs photoexcitation was extracted by monitoring Bragg diffraction-contrast motion within discrete, nanoscale regions. Propagating and stationary coherent contrast motion occurring within differently-oriented crystal regions occurs via excitation of both intralayer and interlayer phonon modes. The distinct dynamics are resolved within a single field of view and along a single incident electron wavevector. Isolation of the strongly-scattering coherent dynamics reveal a response delay of a few ps for the intralayer mode, indicating a coupling to, and excitation of, the low-frequency interlayer breathing mode. Further, the sensitivity of the measurement is such that a rapid dephasing of the breathing mode relative to an adjacent nanoscale crystal region is observed, indicative of a difference of one to three MoS₂ layers and a lag in energy transmission between the crystal boundaries. These results provide new insights into the microscopic nature of intra and interlayer acoustic-phonon dynamics and mode coupling in MoS₂ and layered materials in general, especially with respect to the dramatic influence of nanoscale morphology on photoinduced strain [8].

[1] T. Shen, A. V. Penumatcha, and J. Appenzeller, ACS Nano 10 (2016), 4712.
[2] D. Jariwala, et al., ACS Nano 8 (2014), 1102.
[3] E. M. Mannebach, et al., Nano Lett. 15 (2015), 6889.
[4] E. M. Mannebach, et al., Nano Lett. 17 (2017), 7761.
[5] A. J. McKenna, J. K. Eliason, and D. J. Flannigan, Nano Lett. 17 (2017), 3952.
[6] D. R. Cremons, D. A. Plemmons, and D. J. Flannigan, Nat. Commun. 7 (2016), 11230.
[7] D. R. Cremons, D. A. Plemmons, and D. J. Flannigan, Struct. Dyn. 4 (2017), 044019.
[8] This material is based upon work supported by the National Science Foundation under Grant No. DMR-1654318. Partial support was provided by the Arnold and Mabel Beckman Foundation in the form of a Beckman Young Investigator Award.

MoS₂ is a two-dimensional transition metal dichalcogenide that has gathered much attention with the shift in the research community’s focus from metallic graphene to alternative two-dimensional materials with semiconducting properties. With the lifting of inversion symmetry, monolayer MoS₂ has a direct bandgap of 1.8 eV leading to strong light-matter interaction. In addition, the 2D character of the monolayer leads to reduced dielectric screening and the resulting enhancement in Coulomb interaction leading to large excitonic effects. The high melting point of MoS₂ of more than 1450 K, however, has led to challenges in the growth of thin films. In particular, the large vapor pressure of Sulphur in combination with the high growth temperatures required for film growth have been problematic underscoring the need for a non-conventional approach to growth. In this presentation, ab-initio molecular dynamics is used to show that crystalline MoS₂ can be grown from the amorphous phase on sub-nanosecond time scales with the van der Waals (vdW) gaps serving to strongly orient the resulting structure. The plane wave Vienna ab-initio simulation package (VASP) was used to first form a melt-quenched amorphous phase. The amorphous phase was then heated to just below the melting point and the dynamics of the system were studied. The formation of ABAB rings of Mo-S bonds was observed to precede the emergence of vdW gaps leading to the formation of well defined lattice planes. In the presentation, we will explore details of growth process and probe critical interactions that allow the formation of the crystalline phase on sub-nanosecond time scales using confined thin film amorphous opening up new avenues for large scale growth of 2D MoS₂.

In this talk, I will discuss our recent work in studying the electronic and photonic properties of emerging low-dimensional chalcogenides materials, and in developing them for novel semiconductor device applications. The first part of the talk will focus on discussing our recent research on the phonon-spin and phonon-electron coupling in chalcogenides ferromagnetic monolayer materials and its electronic device applications. In the second part of the talk, I will discuss our work on studying the unique optical properties of chalcogenides perovskite BaTiS₃ over a broad wavelength range from mid-IR to visible, and the new non-linear photonic device applications it can enable. I will conclude with remarks on promising future research directions of low-dimensional chalcogenides materials, and how these newly developed low dimensional materials may enable new functional electronic and photonic devices for sensing, communication, and computing applications.

The band gap is of paramount importance to almost all of the electronic and optical properties of semiconducting materials. By controlling the size of the band gap and their electronic structure, we can control both the transport properties and the optical interactions of such materials. The ability to energetically control electrons in solid-state devices is pivotal in the fields of sensing, renewable energy [1], information processing and communications technology [2]. Here we present the rules of 2D band gap engineering in 2D heterostructures. Our insights offer the potential of engineering not just the band gap, but the electronic dispersion itself; making it far more versatile than strain engineering of band gaps.

The field of 2D semiconductors has been of growing research interest in recent years, but 2D heterostructures have only recently become experimentally viable [3]. We have performed a large scale first principles study of many transition metal dichalcogenides and other 2D semiconductors using density functional theory. These 2D layered heterostructures demonstrate weak inter-layer interactions. Due to this, the band structures of individual layers of a heterostructure have a high fidelity to their isolated counterparts. We show that a layered heterostructure will, therefore, have a dispersion which consists of an overlay of its components' band structures with one key change due to interlayer interaction which we discuss at length. We determine that the electronic dispersions of 2D layered heterostructures can be tailored by layer composition, and demonstrate a wide range of potentially attainable tunable band structures.

1. De Sanctis A., Amit I., Hepplestone S.P., Craciun M.F., Russo S., (2018), Nat. Commun. 9, 1652.
2. Raja A., et al., (2017), Nat. Commun. 8, 15251.
3. Jung Y., Shen J., Sun Y., Cha J.J., (2014), ACS Nano, 8, 9550.

Silicon Telluride (Si₂Te₃) is layered semiconductor material with a unique crystal structure where Si atoms form dimers to fill the sites between the hexagonally close-packed Te atoms.¹ Theoretical investigations showed that Si dimers can rotate between four possible orientations within the Si₂Te₃ layer, giving rise to a unique structural variability that could have potential new applications in electronic and optoelectronics devices.² Recently, two-dimensional Si₂Te₃ multilayers, along with nanoplate and nanowires have been reported.³ Combining the reduced dielectric screening from low dimensionality and the structural variability from Si dimer rotation, the optical properties of 2D Si₂Te₃ is particularly interesting.
In this study, we report a combined computational and experimental investigation of the optical properties of 2D Si₂Te₃. Using many body GW approximation and Bethe-Salpeter equation (BSE), we obtain the dielectric constants of bulk and monolayer of Si₂Te₃. A strong optical anisotropy is discovered. The imaginary dielectric constant in the direction parallel to the Si-Si dimers is half of the value perpendicular to the dimer. We show this effect originates from the particular compositions of the wavefunctions in the valence and conduction bands. The optical measurement of the absorption spectra of 2D Si₂Te₃ nanoplates shows modulation of the absorption coefficient under 90-degree rotation, confirming the computational results. We also found that the BSE calculations reduce GW quasiparticle band gap by 0.3 eV in bulk and 0.9 eV in the case of a monolayer, indicating a large excitonic effect in Si₂Te₃. Furthermore, including the excitonic effect in bulk calculations significantly reduces the imaginary dielectric constant in the out-of-plane direction, indicating a strong effect of Coulombic interaction of the out-of-plane excitons in the case of Si₂Te₃ multilayers. The results show rich features and tunability in the optical properties of 2D Si₂Te₃.

(1) Gregoriades, G.; Bleris, G. L.; Stoemenos, J. Acta Cryst. 1983, B39, 421-426.
(2) Keuleyan, S.; Wang, M.; Chung, F. R.; Commons, J.; Koski, K. J. Nano Lett. 2015, 15, 2285.
(3) Shen, X.; Puzyrev, Y. S.; Combs, C.; Pantelides, S. T. Appl. Phys. Lett. 2016, 109, 113104.

ACKNOWLEDGMENTS
This work was supported by National Science Foundation (DMR-1709528), Computational resources were provided by the NSF XSEDE under grants TG-DMR 170064 and 170076, and by the High-Performance Computing Facility at the University of Memphis.

The field of nanoelectronic chemical sensors faces a gap between theory and observation in applied sensor demonstrations. For example, theoretical predictions of the operating point at which maximal signal-to-noise ratio (SNR) will be obtained are at odds with trends of measured SNR versus gate voltage. This work seeks to bridge this gap for the case of molybdenum disulfide (MoS₂) chemical sensors by studying a MoS₂ device that is influenced by a controlled local charge, which simulates an analyte. The extensive control available in this physical sensor simulation provides pathways to explore and resolve discrepancies between computational sensor simulations and demonstrated sensors. Realizing this level of control with sub-nanometer and pico-amp precision required the development of an advanced characterization system (a customized scanning gate microscopy setup), which was then utilized to test the response of MoS₂ field-effect transistors (FETs) in a variety of configurations. Sensitivity of the device to a local charge was mapped across locations near the device surface, revealing localized sensitive “hot spots” within the 2D channel. Similar maps were taken across a range of operating points and charge-to-surface distances, allowing for the extraction of sensor metrics. For instance, the SNR was shown to peak in the transistor subthreshold regime, reaching over 4 times its value in the linear regime. The comparison of these controlled measurements with electrostatic and 2D current flow models leads to increased understanding of the behavior of MoS₂ FETs in sensors, indicating directions for the optimization of future sensor design and operation.

Can one control photoresponse in materials via patterning optical intensity distribution and polarization? In this talk, we explore this idea on a new class of topological material, Weyl semimetals (WSMs). WSMs are a family of gapless topological materials with broken inversion and/or time reversal symmetry in which the conduction and valence bands cross at single points in momentum space, which are called the Weyl nodes. The electronic bands are topologically nontrivial and display linear dispersion around the Weyl nodes, hosting unique quasiparticles called the Weyl fermions with interesting properties. The optical excitation properties of WSMs are only starting to be explored. We will introduce a new second order nonlinear optoelectronic probe, which we call spatially dispersive circular photogalvanic effect (s-CPGE) and demonstrate its sensitivity to detailed band features owing to its unique excitation mechanism. In WSMs such as MoTe2 and MoxW1-xTe2 (x=0.3, 0.9) in the inversion symmetry broken Td phase, we observe a strong circulating photocurrent when illuminated with circularly polarized light at normal incidence. We will explain these observations from both a simple phenomenological perspective and our efforts to develop a new microscopic response function within a nonlinear susceptibility framework by taking into account the optical field gradients. Our theoretical analysis shows that s-CPGE is controlled by a unique symmetry selection rule combined with asymmetric carrier excitation and relaxation dynamics. By evaluating this response for a minimal model of a Weyl semimetal, we obtained the frequency-dependent scaling behavior of the photocurrent for experiments below the Lifshitz scale where the Weyl dispersion is well-developed. Our recent results of s-CPGE performed over a wide spectral range shows the effect of band inversion and asymmetric electron relaxation effects on the spectral response, from where important materials properties can be extracted. These results also provide a new approach to controlling photoresponse by patterning optical fields on Weyl semimetals and other broken-symmetry systems to store, manipulate and transmit information over a wide spectral range. The implication of our results for assembling chiral on-chip photodetectors for quantum photonic applications will be discussed.

Two-dimensional (2D) dilute magnetic semiconductors (DMS) have attracted immense attention in the recent past for their possible use in future, energy-efficient, spintronic devices. A class of 2D-DMS candidates are transition-metal dichalcogenides (TMDs), substitutionally doped with magnetic-transition-metals (Cr, Fe, Mn, etc.). Especially promising are TMDs based on heavy elements like tungsten (W) and selenium (Se). Thanks to their strong spin-orbit interaction, they easily mediate magnetic interactions among the dopant transition-metal atoms.
In this work, we investigate magnetism in magnetically doped WSe₂ monolayers using Density Functional Theory (DFT). We investigate various doping concentrations and various configurations with different relative positions of the dopant atoms. Specifically, we investigate the effect of clustering of the dopant atoms on the magnetic properties of doped monolayer WSe₂. In our computational model, we first determine the magnetic ground state of the doped WSe₂ monolayers by comparing the total energies of various magnetic configurations calculated using DFT. Next, we simulate their magnetic phase change using lattice Monte-Carlo simulations. Finally, we extract the Curie temperature for the configurations that yield a ferromagnetic ground state.
For the total energy calculations using DFT, we use the Hubbard U model within DFT+U to take into account the enhanced electron-correlation energy of the dopant atoms. The electron-correlation energy in the d orbitals of the magnetic-transition-metals significantly impacts the magnetic order of the material system. To correctly account for the electron-correlation energy of the dopant magnetic transition-metal atoms, we first calculate the Hubbard U parameter for each configuration using the linear-response method. Next, we use the calculated Hubbard U parameters for subsequent total energy calculations using DFT+U.
For structures with a ferromagnetic ground state, we quantify the magnetic order of doped WSe₂ with different doping configurations by calculating their Curie temperature using a two-step approach. In the first step, we build the classical Heisenberg Hamiltonian by obtaining the exchange interactions among the dopant transition-metal atoms. To account for the long-range interactions, we use a parameterized function to define the exchange interactions with asymptotic decaying exponential behavior. We optimize the parameters of the functional form representing the exchange interactions by least square fitting the Heisenberg Hamiltonian to the total energy difference between the magnetic and the non-magnetic configurations obtained from the DFT+U calculations. In the next step, we simulate the magnetic phase change of the classical Heisenberg Hamiltonian using the lattice Monte-Carlo method to obtain the Curie temperature.
From our DFT+U calculations, we additionally extract formation energies, showing that the clustering of the dopant atoms is energetically favorable in monolayer WSe₂. Further, we find that the magnetic order strongly depends on the type of dopant transition-metal atoms and their relative position in the monolayer WSe₂. The strong positional dependence of the dopant atoms on the magnetic order is the result of a trade-off between the minimization of the kinetic energy and the exchange energy. Finally, we conclude by calculating the Curie temperature for various doped structures with ferromagnetic ground states. We predict the optimum doping conditions necessary for obtaining a Curie temperature which exceeds room temperature in transition-metal doped monolayer WSe₂. Among our studied magnetic configurations, we find a Curie temperature of iron (Fe) doped WSe₂ as high as 307 K and manganese (Mn) doped WSe₂ as high as 232 K.

We utilize first-principles density functional theory (DFT) and many-body perturbation theory (MBPT) to study the optoelectronic properties of monolayer germanium selenide (GeSe), emphasizing the role of point defects and electron-phonon interactions; two phenomena that will be present in and can dominate the properties of real materials. We systematically study a series of charged vacancies, their trap state energies, and their impact on optical absorption. Additionally, by approximating the role of electron-phonon interactions, we determine how the defect-induced trap states are modified by the presence of phonons at finite temperature. We determine that the excitonic properties of the material are significantly affected by the presence of defects and phonons, with implications for devices fabricated using this material system.

Fe₉₀Ta₁₀ (Fe-Ta) thin films, deposited using a pulsed laser deposition (PLD) method, have been found to exhibit characteristics of a soft ferromagnetic material with very low coercivities (1-10 mT) and saturation magnetization ~ 16 × 10⁵ A/m. Our field dependent magnetization data (MH) at various temperatures (20 K - 250 K) were used for the indirect measurement of magnetocaloric effect (MCE) to determine the entropy change in the material. An MCE is reported for the first time in rare-earth free Fe₉₀Ta₁₀ (Fe-Ta) thin films. This material system in thin film has shown a crystallographic transition from regular body centered cubic (BCC) crystal structure to an iron rich hexagonal laves phase structure which is isomorphous with the structure type, MgZn₂ (C14). Applying the Maxwell relation to the MH curves at various temperatures, we have numerically calculated δM/δT vs H the integration of which provides quantitative information about the isothermal entropy change. We have observed a positive MCE with maximum entropy change of 6.9 J/K-m³ for the magnetic field changing from 0.05-0.5 T in Fe-Ta system. Although the maximum of the mass specific entropy change is rather small in comparison with existing magnetocaloric material, the entropy change saturates for moderate applied magnetic fields around 0.15 T. This is more than an order of magnitude lower than the magnetic fields generally used to realize a large MCE effect. Efforts have been made to improve the mass specific entropy change by optimizing the deposition parameters and by varying the levels and types of dopants in the FE host.

Monolayer MoS₂ is a promising two-dimensional material for electronic and optoelectronic devices. As-grown MoS₂ is an n-type semiconductor, however, the origin of this unintentional doping is still not clear. Here, using hybrid density functional theory, we carried out an extensive study of the often observed native point defects, i.e., V_S, V_Mo, V_S2, V_MoS3, V_MoS6, Mo_S2, and S2_Mo, and found that none of them cause n-type doping. Specifically, the S vacancy (V_S), which has been widely attributed to n-type conductivity, turns out to be an electron compensating center. We report that hydrogen, which is almost always present in the growth environments, is most stable in its interstitial (H_i) and H-S adatom forms in MoS₂ and acts as a shallow donor, provided the sample is grown under S-rich condition. Furthermore, they have high migration barriers (in excess of 1 eV), which would ensure their stability even at higher temperatures, and hence lead to n-type conductivity.

Very small variations in atomic displacements due to point defects as well as in van der Waals forces between atomic layers in epitaxial films were detected in this work, revealing a scenery of drastic changes in film properties due to subtle variation in defects formation. Advanced X-ray methods were combined to provide the tools necessary to understand the defects dynamic of epitaxial thin films ruled by van der Waals interactions. The results impact in the comprehension of bismuth chalcogenide topological insulator materials.

Van der Waals (vdW) epitaxy of bismuth chalcogenides has recently regained significant interest due to the possibility of suppressing bulk conduction in archetypes of 3D topological insulators. Spin polarized currents flowing exclusively through surface states in high quality thin films can provide a basic platform for novel physics and devices. Although weak vdW interactions play a fundamental role in emerging fields of 2D materials and materials-by-design, the combination of atomic layers of distinct materials without lattice matching suffer from the absence of strong interlayer forces to dictate lateral order in the heterostructures. As a consequence, the overall quality of the epitaxial films are difficult to control. Even in the case of systems with excellent lattice matching, such as bismuth telluride on BaF2 substrate, structural defects are easily introduced with significant impact on the electronic phase diagram of the epitaxial films. Systematic investigation on film characteristic versus growth parameters can provide a general road map for obtaining films with the desired bulk insulating behaviour. However, it is extremely important to understand the key features behind such subtle sources of structural defects. Distinguishing features related to the weakness of vdW forces, and hence relevant to most systems based on vdW epitaxy, and features specifically related to some particularities of the system under investigation.

Here, several x-ray diffraction tools are combined to investigate defects in bismuth telluride films grown by molecular beam epitaxy on BaF2 (111) substrates. Intrinsic topological surface states have already been observed in such films [1], making them good candidates for applications in spintronics or quantum computing. Compositional fluctuation, twinned domains, and bulk free carriers are among the main problems reported in these films [2]. Desorption of tellurium is the root of compositional fluctuation. But, from tellurium vacancies to bismuth rich phases there are a whole gamut of possible processes involved. Antisite occupation can follow the introduction of vacancies in the vdW gaps, promoting films full of point defects detrimental to lattice perfection. At some instant, lattice energy is minimized by turning point defects into bismuth bilayers (BLs). This is a critical instant in terms of free carriers since films with only point defects, only BLs, or coexisting point defects and BLs may have completely different behaviors. Twinned domains can be a simple consequence of low degree of lateral ordering in vdW epitaxy, or of any other source of lateral inhomogeneities still to be identified. Moreover, lattice misfit can be slightly different in bismuth rich films [3], giving the opportunity to study the actual impact of lattice misfit in vdW epitaxy.

[1] C. I. Fornari, P. H. O. Rappl, S. L. Morelhão, T. R. F. Peixoto, H. Bentmann, F. Reinert, E. Abramof. APL Materials 4, 106107 (2016).
[2] C. I. Fornari, P. H. O. Rappl, S. L. Morelhão, G. Fornari, J. S. Travelho, S. de Castro, M. J. P. Pirralho, F. S. Pena, M. L. Peres, E. Abramof. Mater. Res. Express 5, 116410 (2018).
[3] S. L. Morelhão, S. Kycia, C. I. Fornari, P. H. O. Rappl, E. Abramof. Appl. Phys. Lett. 112, 101903 (2018).

Due to strong light-matter interactions and exciton stability, transition-metal dichalcogenides (TMD) provide a platform for optoelectronic applications at room temperatures. By considering coupled dynamics of cavity photons and TMD excitons, we studied exciton-polariton propagation in an optical microcavity with an embedded TMD layer. Specifically, we considered the case where the TMD layer is non-uniform and consists of a set of separate, topologically disconnected microflakes. Via numerical simulations we demonstrated that polaritonic modes can propagate in the system if the TMD coverage is high enough. However, the polaritonic excitation spectrum is broadened compared to that in the case of an ideal homogenous embedded TMD layer. We also found that although the TMD flakes are disconnected, the phases of the excitonic wave functions in different flakes are strongly correlated. We attribute the latter to the exchange by cavity photons emitted and re-absorbed by the TMD excitons. Finally, we discuss the possibility to use the TMD polaritonic system in optoelectronic applications. This work was supported in part by the Department of Defense under grant No. W911NF1810433.

Two-dimensional (2D) materials such as graphene and transition metal dichalcogenides (TMDs) exhibit many interesting properties that make them promising materials for novel electronic and optical applications. The application space could be further broadened by having the ability to combine these materials into more complex heterostructures. Conventional patterning methods rely on coating a polymer-based resist that is then exposed (by light or electron beam), developed, and selectively removed. Each of these steps adds to the process complexity, and in the case of 2D materials, potential materials damage. A resistless patterning technique is therefore desirable, due to fewer potentially damaging processing steps, which naturally improves yield and is more cost-efficient.

In this work, we demonstrate two distinct direct-write patterning methods for 2D TMD heterostructures grown by chemical vapor deposition (CVD). The first approach is based on laser-induced method for controlling the location of MoS₂ nucleation within patterned WS₂. We investigate WS₂ defect formation as a function of the laser power and wavelength and demonstrate the site selectivity of subsequent MoS₂ growth. In an effort to push the patterning resolution, the second resistless technique uses electron beam to directly pattern MoS₂ on Si/SiO₂ substrates. By exposing the substrate by electron beam, we show that CVD-grown MoS₂ preferentially nucleates at the exposed patterns. We will discuss the underlying mechanisms that govern both patterning methods, properties of the resulting TMD materials, resolution limits, as well as the application for more complex TMD heterostructures. These direct-write technique will simplify the process of heterostructure patterning and enable the routine fabrication of complex device architectures.

Properties of transition metal dichalcogenides (TMDs) are sensitively dependent upon a number of structural and morphological features within the few to monolayer limit. For example, the electronic and optical properties of MoS₂ can be tuned via variation in single layer number and applied elastic strain [1,2]. Owing to this sensitive dependence at the single-layer level, development of methods for precisely and accurately determining local layer number is important for unambiguously determining the relationship between function and structure. Among other viable methods (e.g., Raman spectroscopy and atomic force microscopy), transmission electron microscopy (TEM) is an especially attractive platform for such measurements, as it enables combined comprehensive characterization and correlative in situ measurements of the electronic and structural properties, with atomic- to millimeter-scale spatial resolution. For layered materials, a practical approach in TEM for determining layer number involves analysis of relative Bragg-spot intensities in parallel-beam electron diffraction (PBED) patterns over a range of specimen tilt angles [3]. Such an approach has been explored via simulations for few and monolayer MoS₂ [4], while experimental methods aided by prior knowledge of layer number have been applied [5]. Accordingly, systematic experimental measurements employing PBED in TEM on MoS₂ specimens with initially unknown thicknesses, as directly compared to simulations, are needed to establish the practical feasibility, accuracy, and precision of such an approach.

Here, we combine PBED experiments in TEM with simulations to systematically study the robustness of the specimen tilt-angle approach for determination of layer number at the single-layer level in MoS₂. Bragg-spot intensities as a function of tilt angle were simulated through calculation of the associated crystal shape function of the reciprocal-space rods and the structure factors, following Mkhoyan and co-workers [4]. Few and monolayer specimens were prepared using mechanical exfoliation, with subsequent mounting onto holey SiN TEM membranes. Bragg-spot intensities for the {100} and {110} families of planes were mapped over a wide range of tilt angles and for different tilt axes (with respect to a particular in-plane crystallographic direction). Overall, agreement between experiments and simulations were good for an MoS₂ monolayer, indicating the method may be practically useful for identifying such specimens. However, deviations ranging from subtle to severe were observed with increasing layer number, indicating the approach becomes increasingly limited as sensitivity of Bragg-spot intensities and tilt angle to layer number decreases with increasing thickness. Potential sources of error and deviation from simulated results will be discussed [6].

[1] A. Splendiani, et al., Nano Lett. 10 (2010) 1271.
[2] A. Castellanos-Gomez, et al., Nano Lett. 13 (2013) 5361.
[3] J. C. Meyer, et al., Nature 446 (2007) 60.
[4] R. J. Wu, M. L. Odlyzko, and K. A. Mkhoyan, Ultramicroscopy 147 (2014) 8.
[5] J. Brivio, et al., Nano Lett. 11 (2011) 5148.
[6] This material is based upon work supported partially by the National Science Foundation under Grant No. DMR-1654318 and partially by the National Science Foundation through the University of Minnesota MRSEC under Award Number DMR-1420013. Partial support was also provided by the Arnold and Mabel Beckman Foundation in the form of a Beckman Young Investigator Award

We demonstrate, for the first time, wafer-scale graphene/MoS₂ heterostructures prepared by chemical vapor deposition (CVD) and their application in vertical transistors and logic gates. A CVD-grown bulk MoS₂ layer is utilized as the vertical channel, whereas CVD-grown monolayer graphene is used as the tunable-work-function electrode. The short vertical channel of the transistor is formed by sandwiching bulk MoS₂ between the bottom indium tin oxide (ITO, drain electrode) and the top graphene (source electrode). The electron injection barriers at the graphene-MoS₂ junction and ITO-MoS₂ junction are modulated effectively through variation of the Schottky barrier height and its effective barrier width, respectively, because of the work-function tunability of the graphene electrode. The resulting vertical transistor with the CVD-grown MoS₂/graphene heterostructure exhibits excellent electrical performances, including a high current density exceeding 7 A/cm², a subthreshold swing of 410 mV/dec, and a high on-off current ratio exceeding 10³. The large-area synthesis, transfer, and patterning processes of both semiconducting MoS₂ and metallic graphene facilitate construction of a wafer-scale array of transistors and logic gates such as NOT, NAND, and NOR.

Because of its two-dimensional structure and semiconducting properties, tin disulfide (SnS₂) is of interest for applications in electrochemical catalysis and sensing, as an electron transport layer for photovoltaics, and as an active material in photodetectors and thin film transistors. While the atomic and electronic structure of the basal planes of bulk and monolayer SnS₂ are well known, the same is not known for the edges, which could have a major influence on the performance of SnS₂ in the aforementioned applications. This paper reports on density functional theory (DFT) simulations and experimental measurements of the atomic and electronic structure of the edges of bulk and monolayer SnS₂ under different chemical conditions. We found that, with increasing sulfur coverage, the band gap of SnS₂ edges becomes smaller and there’s a transition from indirect to direct bandgap of bulk SnS₂ edges and form indirect to direct to indirect bandgap of monolayer SnS₂ edges. These results thereby determined the influence of chemical synthesis conditions on the electronic structure of the edges.

A few transition metal dichalcogenides are high work-function metallic compounds and could therefore be effective hole contacts in a variety of (opto)electronic devices. Nevertheless, when considering - for example - the field of photovoltaics, it is difficult to find reports of conductive contacts that are not elemental metals or transparent conductive oxides. Here, we incorporate the metallic compound TaS₂ into the device structure of emerging photoabsorbers (Cu₂BaSnS₄ and Cu₂SrSnS₄) and fabricate all-sulfide solar cells. Compared to reference cells built with a standard elemental metallic contact (Mo), the all-sulfide solar cells are more efficient by about 10% relative. We will discuss some of the properties of the TaS₂ hole contact, its stability, and the possible reasons for the efficiency improvement.
For the growth of Cu₂BaSnS₄ and Cu₂SrSnS₄ we propose an oxide precursor route involving thermal conversion of sputtered oxide films (Cu₂BaSnO₄ and Cu₂SrSnO₄) in an H₂S atmosphere at the same temperature normally used for the more common non-oxide precursors. Interestingly, Cu₂BaSnS₄ and Cu₂SrSnS₄ crystallize in a trigonal structure where Cu, Ba(Sr), and Sn have distinct coordination environments. This major structural difference from the well-studied tetrahedrally-coordinated kesterite Cu₂ZnSnS₄ implies that substitutional defect formation is in general unfavorable in Cu₂BaSnS₄ and Cu₂SrSnS₄. In fact, both compounds are found to have sharper absorption and emission edges than kesterite, and their room-temperature photoluminescence peak is well aligned to their band gap.
Finally, there is a long-standing dream of depositing sulfide semiconductors as wide band-gap absorbers on silicon solar cells, in order to realize highly efficient double-junction (tandem) cells. Progress towards this ambitious goal has been hampered by the relatively low efficiency of sulfide absorbers, and by integration issues with the Si bottom cell due to the relatively high temperatures needed to grow and crystallize many sulfides. Here we attempt to address the second issue by incorporating a very thin TiN diffusion barrier between the two sub-cells, which helps preserve the silicon cell during a sulfurization process at 550°C. Efficiencies up to 3.3% for Cu₂ZnSnS₄/Si tandems have been reached so far when incorporating the TiN barrier layer.

Photoelectrodeposition of Se-Te alloys using unstructured illumination spontaneously generates nanopatterned films with significant long-range order. The feature sizes, periodicities, anisotropies, and orientations of the nanoscale pattern can be selected by manipulating the input optical excitation. Isotropic morphologies consisting of ordered arrays of nanopores were generated using unpolarized and circularly-polarized illumination whereas linearly-polarized light resulted in highly-anisotropic lamellar-like morphologies with the in-plane orientation of the patterns controlled by the direction of the light polarization. Elliptical polarization inputs effected additional morphological complexity. The pattern periodicity was encoded by the illumination spectral profile. A single periodicity in single spatial direction was only generated even with the use of broadband and multimodal spectral profiles and the periodicity was found to be sensitive to all investigated tuning of such profiles. Structures with nonequal periodicities in the two orthogonal in-plane directions could also be generated and both periodicities could be independently controlled. The deposition process was assessed computationally using a two-step iterative model wherein a finite-difference time domain method was first used to calculate the spatially-varying levels of light absorption in the growing structure. This instructed a Monte Carlo method in which deposition was simulated to occur preferentially near areas of high localized light absorption. These computational results matched the experimentally observed patterns indicating that pattern development was directed by evolution of interface to maximize anisotropic light collection.

Constraining the physical size of solids can dramatically influence their electrical, optical, magnetic, thermal, and mechanical properties. Intrinsically low-dimensional materials, including van der Waals (vdW) bonded quasi-two-dimensional compounds (exemplified by graphite, hexagonal boron nitride, and transition metal dichalcogenides (TMD)) and quasi-one-dimensional vdW compounds (exemplified by transition-metal trichalcogenides (TMT)), are particularly intriguing, in that the bulk state already presents weakened inter-plane or inter-chain bonding, which leads to strong structural, electronic, and phononic anisotropy. Constraining the dimensions of these 2-D vdW materials down to “atomic thinness” can result in various degrees of additional size quantization with profound consequences. Therefore, it is a reasonable expectation that the 1-D vdW TMT materials would exhibit additional size quantization phenomena with novel and unexpected properties when isolated down to the few- and single-chain limit. Here we report the successful synthesis and structural characterization of HfTe₃ within the hollow cores of multiwall carbon nanotubes (MWCNT). The selectable inner diameter of the MWCNT constrains the transverse dimension of the encapsulated HfTe₃ crystal and thus, depending on the inner diameter of the nanotube, HfTe₃ specimens with many chains (~20), down to few chains (3 and 2), and even single isolated chains, are obtained. The MWCNT sheath simultaneously confines the chains, prevents oxidation in an air environment, and facilitates characterization via high resolution transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM). Together with complementary density functional theory (DFT) calculations, we find a coordinated interchain spiraling for triple and double chain HfTe₃ specimens. Additionally, HfTe₃ shows a structural transition via a trigonal prismatic rocking distortion to a new, unreported crystal phase, concomitant with a metal-insulator transition, as the number of chains is decreased below four.

Control of the doping in antimony selenide (Sb₂Se₃), both native and extrinsic, remains a key challenge for the material if it is to reach its full potential as a high-efficiency photovoltaic absorber. Carrier concentrations in Sb₂Se₃ are often low, with most reports of the material finding quasi-intrinsic or light p-type doping. Several studies have already shown the complicated defect chemistry of Sb₂Se₃, however, there is a paucity of studies into the deliberate doping of the material. In this study, we identify suitable doping strategies to achieve both n-type, and p-type material, first in single crystals and then apply this knowledge to create thin-film devices using a simple post-growth diffusion process. We investigate native defects and extrinsic doping in single-crystal Sb₂Se₃ through a combination of deep-level transient spectroscopy, density functional theory calculations, hot-probe, hall effect and secondary ion mass spectrometry techniques. We show that stoichiometric Sb₂Se₃ is an intrinsic semiconductor and identify tin and chlorine as effective extrinsic dopant species. We then discuss routes to achieve controlled doping profiles and high carrier concentrations in Sb₂Se₃ thin films. This study provides a greater understanding of doping in Sb₂Se₃, and offers a roadmap to high efficiency devices through extrinsic doping processes.

With the global demand for energy increasing year on year, diversification beyond current technologies and materials is crucial to meeting this demand by accessing more sustainable materials in a wider variety of device architectures and applications. In photovoltaics, while silicon is the dominant technology, its poor absorption puts upper bounds on device thinness, while established ‘thin-film’ materials such as CdTe allow strong absorption from nanometre-thin films and flexible devices but suffer from toxicity and low abundance of constituent elements.^1,2
Antimony Selenide, on the other hand, is a highly promising candidate chalcogenide photovoltaic absorber, possessing a very high absorption coefficient, near-ideal band gap and relatively abundant constituents. Solar cells utilising it as the absorber layer are nearing 10% in efficiency,³ and the pseudo-1D nature of the material, with van der Waals interactions, has been highlighted as a potential reason for benign grain boundaries.⁴ Our recent theoretical work, however, has found that despite a high dielectric constant and a ns² ‘lone-pair’ cation configuration, both characteristics that have been associated with the concept of defect tolerance, Sb₂Se₃ possesses multiple low formation energy intrinsic defects with mid-gap transition levels that could severely hinder future improvements in open circuit voltage.⁵
In this study, we discuss these hybrid density functional theory calculations on the intrinsic defects of Sb₂Se₃ with focus on examining how the amphoteric behaviour of selenium allows for such low formation antisite defects, which may pin the Fermi level. Further, we examine the carrier capture behaviour of these defects and which defect levels may be responsible for the trap states seen in experimental DLTS measurements. Finally, we have also performed calculations on numerous extrinsic dopants to assess possible routes to doping in Sb₂Se₃, including passivation of deep intrinsic levels, and, in collaboration with colleagues at the University of Liverpool, potential contaminants that may affect current and future devices. Through these results, our study examines the specific effects at play within Sb₂Se₃ but also explores the consequences on the applicability of ‘defect tolerance’ within post-transition metal chalcogenide materials.

(1) Haegel, N. M.; Margolis, R.; Buonassisi, T.; Feldman, D.; Froitzheim, A.; Garabedian, R.; Green, M.; Glunz, S.; Henning, H.; Holder, B.; Kaizuka, I.; Kroposki, B.; Matsubara, K.; Niki, S.; Sakurai, K.; Schindler, R. A.; Tumas, W.; Weber, E. R.; Wilson, G.; Woodhouse, M.; Kurtz, S. Science (80-. ). 2017, 356 (6334), 141.
(2) Peter, L. M. Philos. Trans. A. Math. Phys. Eng. Sci. 2011, 369, 1840.
(3) Li, Z.; Liang, X.; Li, G.; Liu, H.; Zhang, H.; Guo, J.; Chen, J.; Shen, K.; San, X.; Yu, W.; Schropp, R. E. I.; Mai, Y. Nat. Commun. 2019, 10, 125.
(4) Zhou, Y.; Wang, L.; Chen, S.; Qin, S.; Liu, X.; Chen, J.; Xue, D.-J.; Luo, M.; Cao, Y.; Cheng, Y.; Sargent, E. H.; Tang, J. Nat. Photonics 2015, 9 (6), 409.
(5) Savory, C. N.; Scanlon, D. O. J. Mater. Chem. A 2019, 7 (17), 10739.

The combination of oxide and heavier chalcogenide layers in thin film photovoltaics suffers limitations associated with either oxygen incorporation into the chalcogenide layer or a chemical incompatibility which results in dewetting issues and defect states at the interface. Here, we establish atomic layer deposition (ALD) as a tool to achieve three goals in this field. Firstly, it allows one to obtain highly pure Sb₂S₃ light absorber layer. Secondly, it provides high precision in the tuning of its thickness, demonstrated here between 30 and 160 nm. Thirdly, it is exploited to generate an ultra-thin adhesion layer of ZnS between 0.2 and 2.0 nm that simultaneously resolves dewetting and passivates defect states at the interface, thereby slowing down interfacial charge recombination. The materials stack is characterized by XPS, Raman, and SIMS, and its (photo-)physical properties are measured by UPS and transient absorption spectroscopy. The comparison of solar cells with and without oxide incorporation in Sb₂S₃, with an optimum ultra-thin ZnS interlayer below 1.0 nm, and with systematically varied Sb₂S₃ thickness provides a complete picture of the physical processes at work in devices with optimized power conversion efficiency beyond 5.0 %.

During the last decade, various chalcogenide materials (Cu(In,Ga)(S,Se)₂, CuGa(S,Se)₂ and Cu₂ZnSn(S,Se)₄) have been studied as photocathodes for solar water splitting because of their high optical absorption, suitable band position and tunable bandgap. However, concerns including Zn-Cu antisite disordering and element scarcity/toxicity (In, Ga) remain as obstacles for their further performance improvements and practical applications. Most recently, Cu₂BaSn(S,Se)₄ has attracted considerable attention as emerging chalcogenide materials with all earth-abundant elements and less tendency for antisite disordering compared to Cu₂ZnSn(S,Se)₄ while maintaining the desirable electronic and optical properties for solar energy conversion applications. Previously, studies of Cu₂BaSn(S,Se)₄ photocathodes focused on improvements of photocurrent density and onset potential. However, the stability of the Cu₂BaSn(S,Se)₄ was not fully understood and stability beyond 10 hours (a concern for its practical application) was not yet demonstrated. The current study investigates the electrochemical and photoelectrochemical stability of bare Cu₂BaSn(S,Se)₄ based photocathodes from both experimental and theoretical viewpoints. Additionally, long term stability of Cu₂BaSn(S,Se)₄ with overlayers and catalyst was examined.
The stability of the bare Cu₂BaSn(S,Se)₄ was found to significantly degraded at both positive and negative potentials but remains relatively stable in the middle potential range (e.g., 0 V/RHE to 0.4 V/RHE). Careful SEM, XPS, EDS and XRD measurements reveal different degradation mechanisms of Cu₂BaSn(S,Se)₄ electrodes at different potentials. For positive potential, an electrochemically induced mechanical delamination was observed to cause the device performance to degrade. When a more negative potential was applied (e.g., -0.4 V/RHE,), photoelectrochemical corrosion was found to be responsible for the photocurrent decrease. Self-reduction and oxidation potentials of Cu₂BaSn(S,Se)₄ was calculated thermodynamically and compared with the experimental results. Combining the theoretical and experimental results, the S/Se self-oxidation and Cu/Sn self-reduction were proposed to explain the degradation at positive and negative potentials respectively.
Long term stability of Cu₂BaSn(S,Se)₄ photocathodes with TiO₂/CdS overlayers and Pt catalyst was tested for more than 100 hours continuously with AM 1.5G solar simulator at 0 V/RHE. It was found that more than 50% of the photocurrent was retained after 100 hours of testing. The decrease of photocurrent was attributed to Pt catalyst delamination and verified by EDS and SEM images.
The present results elucidate the potential dependent stability and degradation mechanism of the Cu₂BaSn(S,Se)₄ system. Results of protected Cu₂BaSn(S,Se)₄ photocathodes highlight the possibility of practical Cu₂BaSn(S,Se)₄ solar water splitting devices with high efficiency (>10 mA/cm² at 0 V/RHE) low cost (solution-processed) and considerable stability (appropriate catalyst). The above results are also instructive for studying the stability of other existing chalcogenide materials and designing new chalcogenide materials with higher stability.

The design criteria for sustainable thin-film photovoltaic devices includes the chemical (e.g. abundance, toxicity, stability, scalability) and physical (e.g. band gap, absorption, doping density, contact behaviour) properties of the underlying materials. Many non-conventional inorganic materials are currently being investigated including oxides (e.g. Cu2O) and sulphides (e.g. SnS); however, none are close to reaching their theoretical potential as defined by the Shockley–Queisser limit [1].

I will discuss the latest advances in materials modelling [2] for the discovery of new materials for solar energy conversion. The role of predictive simulations can vary from high-throughput screening of candidate compounds, rigorous assessment of physical responses, to the optimisation of device architectures. Particular attention will be paid to what can learned from the high-performance of perovskite solar cells and the reduction of non-radiative electron-hole recombination that limits the performance of many new technologies. Examples will be taken from our exploration of kesterite (e.g. Cu2ZnSnS4), matlockite (PbFCl type), herzenbergite (SnS), and antimonselite (Sb2Se3) systems [3-5].

[1] "Emerging inorganic solar cell efficiency tables" J Phys Energy (2019); https://doi.org/10.1088/2515-7655/ab2338

[2] “Machine learning for molecular and materials science” Nature 559, 547 (2018); https://doi.org/10.1038/s41586-018-0337-2

[3] "Lone-pair effect on carrier capture in Cu2ZnSnS4 solar cells" J. Mat. Chem. A 7, 2686 (2019); https://doi.org/10.1039/C8TA10130B

[4] "Metastable cubic tin sulfide: A novel phonon-stable chiral semiconductor" APL Mater. 5, 036101 (2017); http://dx.doi.org/10.1063/1.4977868

[5] "Finding a junction partner for candidate solar cell absorbers enargite and bournonite from electronic band and lattice matching" J. Appl. Phys. 125, 055703 (2019); https://doi.org/10.1063/1.5079485

One promising approach to economical and sustainable fuel production is through photoelectrochemical (PEC) hydrogen production. However, the widescale deployment of existing state-of-the-art photoelectrodes has been hampered by issues with durability, efficiency and cost. Chalcopyrite-based tandem photoelectrodes are attractive candidates in that they offer a great flexibility in the choice of component materials necessary that drive the reactions for different device designs, offering ways to reduce the cost and to improve the performance. The synthesis of low-band gap chalcopyrites like CuInSe₂ and alloys with CuGaSe₂ (CIGSe) are known to exhibit a variety of Cu-poor phases (ordered-vacancy compounds or OVCs) that can greatly influence the resulting properties of fabricated devices, while the existence and properties of these phases in wider-band gap chalcopyrite compounds and alloys remains largely unexplored. Using hybrid functional calculations, we discuss the stability, optical and electrical properties of OVCs in a number of other chalcopyrite chemistries beyond CIGSe. We discuss the influence of such phases on the resulting absorption and band offsets that would result upon their formation, which have implications in device design. We additionally discuss experimental fingerprints that could be used to identify the existence of such phases in synthesized material.

This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 and funded by the HydroGEN Consortium within the Department of Energy Office of Energy Efficiency & Renewable Energy (EERE) and Fuel Cell Technologies Office.

Group-IV chalcogenides are a fascinating material system with rare properties that arise from the nature of their bonding. The rocksalt alloys of Pb_xSn_1-xSe have important mid-infrared optoelectronic properties, with certain compositions hosting topologically non-trivial states. We explore the epitaxial integration of PbSnSe alloys with zincblende III-V materials for control over these unique electronic properties via hybrid interfaces to a mature materials platform. Here, we use PbSe as a model material to explore how IV-VI materials nucleate and grow on III-V materials. The thin films in our study are synthesized by molecular beam epitaxy from compound IV-VI sources. We show how to independently tune the substrate surface chemistry and lattice mismatch using buffer layers of (In,Ga)(As,Sb) alloys; finding that both influence the growth behavior of PbSe. Characterization of the (001) rocksalt/zincblende interface by electron microscopy reveals a unique arrangement of atoms that mediates dissimilarity in atomic identity, crystal structure, and valence.
With these tools in hand, we demonstrate routes to create atomically sharp and uniform interfaces between these material systems, paving the way to the measurement of the band alignment. The growth mode and lattice-mismatch also results in the generation of dislocations at the interface of PbSnSe and III-V materials that propagate into the film. We do not fully understand the properties of these dislocations, their relationship to interface preparation, and their impact on electronic properties. To this end, we present results from atom probe tomography and electron microscopy to provide a detailed microstructural description of these features. We see evidence for strong deviation in composition at these defects. Finally, dislocations are not always harmful and here we show that the glide of these dislocations can be used to our advantage to grade across composition in PbSnSe in (111) oriented films to explore the growth of metastable cubic SnSe-rich phases. These growth studies are now guiding our designs of hybrid heterostructures to understand and control the unique properties of IV-VI materials.

Low dimensional chalcogenides become very attractive for electronic and optical applications, largely owing to their unique physical and chemical properties strongly augmented by low-dimensionality. The low dimensionality reduces dielectric screening and enhances Coloumb interaction, resulting in pronouced optoelectronic response and forming quasiparticles such as excitons and polaritons. Here we present our first-principles theoretical study of quasi-1D antimony chalcogenides, an emerging class of optoelectronic materials possessing intermediate bonding between weak van der Waals and strong covalent interactions. We show that this unique quasi-1D structure mediates its dielectric screening environment, giving rise to highly anistropic electronic and optical properties that are promising for low-cost photovoltaics and novel optoelectronics. We will further discuss the defect physics in antimony chalcogenides including both intrinsic and extrinsic point defects as well as their impact on relevant device performance. Our results provide valuable theoretical insights on the emerging antimony-chalcogenide based non-toxic and earth-abundant photovoltaics and open exciting avenues for optoelectronic applications of quasi-1D chalcogenides.

Metal chalcogenide semiconducting nanocrystals (NCs), such as CdSe, possess striking size-dependent optical properties and have been shown to have applications as both optoelectronic (i.e. solar cells and laser emitting diodes) and biomedical devices (i.e. biological markers) [1]. These types of materials typically consist of an inorganic core, which dictates most of the physical properties, and an organic layer on the surface, which helps dictate chemical stability. While much work has been performed towards understanding the fundamental physics of the core NC, the surface science of these materials remains relatively unexplored. Although it is believed that the surface of NCs are primarily responsible for chemical stability, recent reports have suggested that the surface layer does much more, including affecting photoluminescence (PL) quantum yields and PL lifetimes [2]. Quite remarkably, the surface layer even has been shown to convert nominally non-magnetic materials, such as CdSe, into “magnetic NCs” [3]. This result is curious, as the bulk state of these materials exhibit diamagnetic properties. These exciting results suggest that is may be possible to pair these induced magnetic properties with the inherent PL properties of these materials without introducing chemical dopants. The multi-functionalization of CdSe and ZnO semiconducting magnets may lead to better biomedical applications such as targeted drug delivery, targeted organelle extraction, and magnetic hyperthermia treatments.

To gain further insight on how the surface layer affects the magnetic properties, our current work focuses on altering the NC surface chemistry by controllably varying the surface ligand concentrations and by systematically varying the types of ligands (i.e. headgroup) on the surface of spherical CdSe and ZnO NCs. Our work reveals that CdSe NCs exhibit weak paramagnetic (PM) properties in the presence of an external magnetic field as previously reported [3]. Sequential ligand removal does not yield a proportional response to the magnetic properties, but in general the PM properties are enhanced with fewer surface ligands attached. We hypothesize that after the removal of ligands the surface of the NC restructures by forming Se-Se bonds. DFT calculations suggest that removal of a full surface layer can result in a Se-Se triplet state that is energetically favorable compared to the singlet state and may be the source of magnetism in this system [4].

Similarly, ZnO NCs become paramagnetic or ferromagnetic at the nanoscale and it is generally accepted that the mechanism behind the magnetic phenomenon in ZnO results from interstitial point defects near the surface of the NC [4]. It has been shown that by variation of both the size and type of surface ligands, ZnO NCs can exhibit differences in the magnetic responses. Preliminary results suggest that ZnO NCs on the order of 30 nm exhibit PM behavior and while 4 nm ZnO NCs exhibit ferromagnetic behavior. The size dependencies are suggestive of surface defects, as the aspect ratio increases as NCs decrease in size. Very little is known about how the magnetic properties are affected when altering the surface ligand concentration and while surface reconstruction could be the source of the magnetism, reconstruction alone may not be the only source of magnetic properties.

1 N. Sahu, N. Brahme, and R. Sharma, Luminescence 31, 1400 (2016)
2 E. Busby, N.C. Anderson, J.S. Owen, and M.Y. Sfeir, J. Phys. Chem. C 119, 27797 (2015)
3 R.W. Meulenberg, J.R.I. Lee, S.K. McCall, K.M. Hanif, D. Haskel, J.C. Lang, L.J. Terminello, and T. Van Buuren, J. Am. Chem. Soc. 131, 6888 (2009)
4 Giansante, C. & Infante, I. Surface Traps in Colloidal Quantum Dots: A Combined Experimental and Theoretical Perspective. J. Phys. Chem. Lett. 5209–5215 (2017). doi:10.1021/acs.jpclett.7b02193
5 M. V. Limaye, S.B. Singh, R. Das, P. Poddar, and S.K. Kulkarni, J. Solid State Chem. 184, 391 (2011)

The search for new thermoelectric materials requires optimizing a set of interrelated, complex, and often conflicting material properties. Amongst the many desirable features needed to achieve good thermoelectric performance in semiconductors, two critical ones are low thermal conductivity and high carrier concentrations (dopability). This presentation will highlight recent combined computational and experimental efforts to tailor and understand thermal and electronic properties of a chemically-diverse set of telluride-based diamond-like semiconductors (DLS) within the Cu₂II_BIVTe₄ (II_B = Zn, Cd, Hg, IV = Si, Ge, Sn) material space and other related compounds. We use combined computational first-principles methods, synthesis, and characterization to comprehensively assess this space for thermoelectric performance. To address thermal properties, first-principles modeling together with thermal transport measurements together suggest that substantial site disorder arising from the prevalence of II_B-Cu and Cu_IIB antisite defects gives rise to ultra-low thermal conductivities. The measured trend in the thermal conductivities corresponds well to the spectral mismatch of the Cu and group II_B species in the phonon spectrum, suggesting the key role of these antisite defects in scattering thermal carriers. To address dopability, we use density functional theory to investigate the intrinsic defect chemistry of several telluride DLS including Cu₂HgGeTe₄ and the related ordered oxygen-vacancy compound Hg₂GeTe₄, and predict achievable carrier concentrations. Experimentally, carrier density control has been demonstrated within the Cu₂HgGeTe₄-Hg₂GeTe₄ solid solution. Our first-principles analysis shows that Cu₂HgGeTe₄ can range from degenerate p-type to highly n-type under different thermodynamic environments. In agreement with high-temperature x-ray diffraction and resonant spectroscopy experiments, the predominant defects in the quaternary are found to be antisite defects with Cu and Hg. On the other hand, as Hg₂GeTe₄ does not contain Cu-related antisites, native defects exhibit relatively high defect formation energies, and Hg₂GeTe₄ is predicted to have an equilibrium Fermi energy near mid-gap for all growth environments and a wide dopability window. To overcome the small intrinsic carrier concentrations in Hg₂GeTe₄, we use first-principles to screen and recommend a set of extrinsic dopants that can be used to tune the carrier type from p-type to n-type.

Development of nanostructured materials has introduced revolutionary approaches for materials processing and electronic structure engineering. These materials can offer the advantages of crystalline inorganic solids combined with inexpensive solution-based device fabrication. Along these lines, semiconductor quantum dots are explored as the functional elements for printable electronics, light emitting devices, photodetectors, solar cells and lasers. All these applications require efficient coupling between individual nanostructured components. The advances in the surface chemistry of semiconducting nanostructures are poised to enable advances in additive manufacturing of semiconducting and multifunctional materials. Specifically, I will discuss inorganic linkers that permit electronic coupling between the nanocrystals and composition-matched molecular semiconductor "solders" that convert nanocrystals in the high quality inorganic semiconductors. I will also introduce a general chemical approach for photoresist-free, direct optical lithography of functional inorganic nanomaterials (DOLFIN). Examples of patterned materials include metals, semiconductors, oxides, and magnetic and rare earth compositions. No organic impurities are present in the patterned layers, which helps achieve good electronic and optical properties. The conductivity, carrier mobility, dielectric, and luminescence properties of optically patterned layers are on par with the properties of state-of-the-art solution-processed materials. The ability to directly pattern all-inorganic layers using a light exposure dose comparable to that of organic photoresists opens up new opportunities for thin-film device manufacturing.

Colloidal chalcogenide nanocrystals comprise an important class of materials with a plethora of uses in photovoltaic, thermoelectric, catalytic, phase-change memory and other applications. Traditionally, the properties of such chalcogenide colloids are tuned by accurate control over their size, exploiting quantum confinement phenomena. Multicomponent chalcogenide nanocrystals provide an extra tuning knob: composition-dependent effects (i.e., tunable ratio between constituent cations) are superimposed on nanoscale size dependences. Synthesis of multicomponent nanocrystals, however, remains a synthetic challenge, due to side reactions and difference in reactivity of starting elemental precursors.
In this talk, I will summarize our recent work, regarding ternary and quaternary chalcogenide nanomaterials. Taking example of I-III-VI group nanocrystals, I will highlight an amide-promoted synthetic approach that enables independent control over nanocrystal size and composition. The method is based on addition of amide superbase, which accelerates the nucleation rate of nanocrystals and, ultimately, diminishes the reactivity difference between M⁺, M²⁺, and M³⁺ ions. We will demonstrate a generality of amide-promoted synthetic approach for various chalcogenide nanocrystals and discuss the limitations of the method.
In the following, we will focus on quaternary Cu-Zn-In-Se nanocrystals, for which we develop a predictive model for the synthesis in broad range of Indium-rich compositions. We will discuss an interplay between size and composition effects on structural and optical properties of Cu-Zn-In-Se nanocrystals. We will detect optimal ternary and quaternary compositions with enhanced luminescence efficiencies, which can be associated with structural ordering of cations and cationic vacancies. These results can provide better understanding of structure-property relationships for multicomponent nanocrystals and ultimately guide the development of multicomponent nanomaterials with optimized optical and electronic properties.

The ability to control the assembly of inorganic nanocrystals within polymer hosts is an important requirement for the use of metal chalcogenide nanocrystals in optical, energy capture and electrical devices. A popular route to the forming polymer-inorganic nanoparticle mixtures is to synthesis the particles and then mix with the polymer. This has disadvantage that when removing the solvents typically used that the conformational entropic penalty of the polymer existing at the nanoparticle surface and the enthalpic interactions between polymer and particle cause the particles to become immiscible with the polymer.

A route that has been explored to tip the balance to better mixtures of polymers and particles is to synthesise the particles within the undissolved polymer, i.e. in situ. The polymers of interest are solution processed in the same solution as the precursors for the desired metal chalcogenide.^1,2 In particular, metal xanthates have proven to have the solubility in compatible processing solvents and also breakdown at relatively low temperatures, avoiding the decomposition of the polymer matrix. Our group has used this technique to form randomly ordered mixtures of polymer and lead chalcogenide nanocrystals where we reintroduced size control of the synthesised nanocrystals.^3–5

In this work we use electrospinning and electrospray techniques to form fibres containing highly aligned polymer chains and the required metal xanthate complexes. Upon heating the aligned polymers act as the environment the nanocrystals form within and direct the location of the resulting nanocrystals.

1 I. E. Uflyand and G. I. Dzhardimalieva, Thermolysis of Metal Chelates in Polymer Matrices. in Nanomaterials Preparation by Thermolysis of Metal Chelates, eds. I. E. Uflyand and G. I. Dzhardimalieva, Springer International Publishing, Cham, 2018, pp. 425–458.
2 T. Rath and G. Trimmel, Hybrid Mater., 2014, 1, 2299–3940.
3 E. A. Lewis, P. D. McNaughter, Z. Yin, Y. Chen, J. R. Brent, S. A. Saah, J. Raftery, J. A. M. Awudza, M. A. Malik, P. O’Brien and S. J. Haigh, Chem. Mater., 2015, 27, 2127–2136.
4 P. D. McNaughter, J. C. Bear, A. G. Mayes, I. P. Parkin and P. O’Brien, R. Soc. Open Sci., 2017, 4, 170383.
5 J. R. Brent, P. D. McNaughter and P. O’Brien, Chem. Commun., 2017, 53, 6428–6431.

Geology has the advantages of extreme temperatures, pressures, and cooling rates to achieve a dizzying array of diverse crystalline phases of metal chalcogenides. As solution phase chemists, we alternatively have access to diverse organochalcogenide precursors and chemistries with distinctive reactivities both in rate and mechanism. These precursors provide new methods to control crystalline phase in colloidal nanocrystal synthesis. At times, solution phase chemistry is even more powerful than geology as new “unnatural” polytypes of binary and ternary compounds not present in the geologic record can result. These new polytypes present an emerging class of materials with new optical, electronic, and chemical properties. Here we will present examples from our research along with a study of the mechanism of formation that will provide the basis for further discovery of new materials.
CuInS₂ has a band gap of 1.5eV and is a potential replacement for CdSe and CdS semiconductors in optoelectronic applications. While the favored thermodynamic structure is chalcopyrite, in some nanocrystal syntheses, a metastable wurtzite-like polymorph forms. We will show how the chalcogenide precursor reactivity influences the preferential formation of one of two binary intermediates, either pseudo-hexagonal Cu₂S or othorhombic In₂S₃. These binaries provides the anionic substructure which dictates the final phase of the ternary product after a second step of partial cation exchange.
With this mechanistic understanding, the class of hexagonal ternary copper sulfides has been expanded to now included the synthesis of wurtzite-like CuFeS_2, another rare, unnatural polymorph, instead of chalcopyrite. In this case, Cu₂S nanocrystals are intentionally prepared and used as a reagent. However, the exchange process itself can be disruptive, and there is a ligand dependence of the retention of structure even in cation exchanges at moderate temperatures.
The syntheses of hexagonal ternary copper sulfides are facilitated by the facile synthesis of pseudo-hexagonal Cu₂S, which is the thermodynamic phase. Similar intentional two step syntheses to hexagonal ternary copper selenides is hampered because the thermodynamic phase of Cu₂Se is cubic. Indeed, the hexagonal phase is not known to the geologic record, and it was synthesized only once before through cation exchange from CdS nanocrystals. We will present the first known direct synthesis of hexagonal Cu₂Se: a product that is facilitated by a didodecyldiselenide precursor. Similar syntheses that use dodcecylselenol or include phosphine precursors produce the thermodynamic cubic phase. Details of the relationship between precursors used and phase selection is revealed by in situ ⁷⁷Se NMR of nanocrystal syntheses.

PbS quantum dots are already commercialized in short-wave infrared (SWIR) photodetectors and are under development for single-photon sources, photovoltaics, and light-emission applications. Nevertheless, key aspects of their fundamental physics such as their excitonic fine structure, exciton-phonon coupling, and exciton dynamics are still not well understood. Using a popular synthesis from the literature,¹ we find that these PbS nanocrystals are coated in a previously unnoticed shell that is measurable in transmission electron micrographs.² The shell contains lead and chlorine, beyond the monoatomic chlorine termination previously proposed. Identifying this shell and its influence on the nanocrystals’ excitonic structure is a crucial first step toward understanding the physics of PbS quantum dots.

Earlier work has correlated the diameters of PbS nanocrystals to their excitonic absorption; however, we observe that PbS quantum dots synthesized in saturated dispersions of PbCl₂can deviate from the previous 1S_h-1S_e energy vs. diameter curve by 0.8 nm. In addition, their surface differs chemically from that of PbS quantum dots produced via other syntheses, which affects their excitonic absorption spectrum.

This result has important implications for understanding the growth mechanism of this reaction, the linewidth of these quantum dots’ photoluminescence, and electronic transport within films of these nanocrystals. Such fundamental knowledge is critical to further our understanding of PbS quantum dots and their development into optoelectronic devices.

1. Weidman, M. C., Beck, M. E., Hoffman, R. S., Prins, F., and Tisdale, W. A. Monodisperse, Air-Stable PbS Nanocrystals via Precursor Stoichiometry Control. ACS Nano, 8, 6363-6371 (2014).
2. Brittman, S., Colbert, A. E., Brintlinger, T. H., Cunningham, P. D., Stewart, M. H., Heuer, W. B., Stroud, R. M., Tischler, J. G., and Boercker, J. E. Effects of a Lead Chloride Shell on Lead Sulfide Quantum Dots. Journal of Physical Chemistry Letters, 10, 1914-1918 (2019).

Quantum dot light-emitting diodes (QLEDs) emitting at 410 nm were studied by time-resolved electroluminescence measurements. A novel transient device current overshoot after voltage turn-off was observed which is attributed to the accumulation and storage of charge carriers at the ligand-quantum dot interface. Short ligands produced short rise times and prevented current overshoots, whereas longer ligands caused a storage of charge carriers, a slower response and notable current overshoots. This overshoot can be attributed to the internal electric field between the injected, and stored and trapped electrons and holes upon switching-off of the external voltage. Thereby, trapped charges at the quantum dot/ligand interface recombine and produce an electroluminescence spike. Applying a dual step voltage pulse, prevents this overshoot, and instead a delayed luminescence is found. As the accumulated charge carriers are immobile and trapped in shallow states at the ligand/quantum dot interface, a reverse pulse was applied to fully deplete the emissive layer. With the transient overshoot disappearing after the device has been turned on and operated, a measure for degradation of QLEDs has been identified.
Upon using a 5nm thick Al interlayer under the hole transport layer, accumulated charge carriers can increase the conductance of the active layer during the forward voltage sweep. This introduces a memory window, a new device functionality, with ON/OFF states that can be reversed by applying a negative bias. The height of the memory window depends on the length of ligand passivating the chalcogenide quantum dot. During backward voltage sweeps, previously injected charge carriers recombine until the active layer is free of charge carriers and show a low conductance state. Voltage-capacitance measurements confirm the changes in charge carrier population within the active layer that depend on both the direction of the bias sweep as well as the ligands surrounding the quantum dots. A complete memory cycle with read, write, erase and re-read modes has been demonstrated.

While surfactant assisted synthesis is the most common technique used for the preparation of colloidal nanoparticles, researchers in the field often encounter serious irreproducibility issues. In this talk, we will present several examples for the role of ‘beneficial impurities’ in nanoparticle syntheses.[1-5] Specifically, controlled amounts of alkylammonium-alkylcarbamate surfactants were shown to strongly affect alkylamine-assisted synthesis of anisotropic colloidal ZnS nanoparticles.[4,5] Furthermore, a new cubic semiconducting binary phase (denoted as the π-phase) has been discovered in recent years in the form of colloidal nanoparticles. The first π-phase materials to be discovered were tin mono-sulfide, π-SnS [6,7] followed by tin mono-selenide, π-SnSe [8], and density functional theory calculations predict additional π-phase compound semiconductors.[9] Increasing experimental and computational evidence points out that ‘beneficial impurities’ can be used for achieving phase control in the tin mono-chalcogenide system.

References

1. F. Wang, R. Tang, W.E. Buhro, Nano Lett. 8 (2008) 3521.
2. J.E. Millstone, W. Wei, M.R. Jones, H. Yoo, C.A. Mirkin, Nano Lett. 8 (2008) 2526.
3. D.K. Smith, N.R. Miller, B.A. Korgel, Langmuir 25 (2009) 9518.
4. N. Belman, J.N. Israelachvili, Y. Li, C.R. Safinya, J. Bernstein and Y. Golan, Nano Lett. 9 (2009) 2088.
5. N. Belman, J.N. Israelachvili, Y. Li, C.R. Safinya, J. Bernstein and Y. Golan, J. Am. Chem. Soc. 131 (2009) 9107.
6. A. Rabkin, S. Samuha, R.E. Abutbul, V. Ezersky, L. Meshi, Y. Golan, Nano Lett. 15 (2015) 2174-2179.
7. R.E. Abutbul, E. Segev, L. Zeiri, V. Ezersky, G. Makov, Y. Golan, RSC Adv. 6 (2016) 5848-5855.
8. R.E. Abutbul, E. Segev, S. Samucha, L. Zeiri, V. Ezersky, G. Makov, Y. Golan, CrystEngComm 18 (2016) 1918-1923.
9. R. E. Abutbul, E. Segev, U. Argaman, G. Makov, and Y. Golan, Adv. Mater. 30 (2018) 1706285.

The theoretical maximum efficiency of a solar cell such as Shockley-Queisser limit has mainly focused on the radiative recombination as the limiting factor. However, in practical solar cells, the nonradiative recombination is unavoidable due to the formation of native point defects and often the dominant recombination mechanism [1]. We will predict the theoretical maximum efficiency of kesterite solar cells (Cu₂Zn(Ge,Sn)(S,Se)₄), taking into account both radiative recombination and non-radiative carrier recombination mediated by the native point defects. We calculate the thermal equilibrium concentrations of native defects and their capture cross-sections. we find that the sulfur vacancy (V_S), sulfur vacancy-donor complex and Sn antisites (Sn_Zn) produce deep levels and large capture cross-sections resulting in the low open-circuit voltage and the low solar-to-electricity conversion efficiency [2, 3]. The predicted upper limit is compared to the current generation of best performing kesterite solar cells. We suggest that the codoping and alloying may improve the efficiency of kesterite solar cells.

[1] Point defect engineering in thin-film solar cell. Nat. Rev. Mat. 3, 194 (2018); https://doi.org/10.1038/s41578-018-0026-7

[2] Identification of Killer Defects in Kesterite Thin-Film Solar Cells. ACS Energy Lett. 3, 496 (2018); https://doi.org/10.1021/acsenergylett.7b01313

[3] Lone-pair effect on carrier capture in Cu₂ZnSnS₄ solar cells. J. Mater. Chem. A 7, 2686 (2019); https://doi.org/10.1039/C8TA10130B

In recent years Cu-based chalcogenide Cu₂ZnSnS₄ (CZTS) has attracted a lot of attention as an earth-abudant alternative for CIGS solar cells. To date, the research in this field has been focused on solar cell applications. However, the efficiency of CZTS solar cells has not experienced a real improvement in the past five years. Therefore, research is beginning to explore new applications for this material such as sensors, water-splitting and charge-extraction layers. In this project, aluminium doping has been introduced to CZTS thin-films for the first time. CZTS films were deposited in atmospheric conditions via a low-cost and environmentally friendly solution-based method. The effects of up to 3 % Al doping in a range of different annealing temperatures from 350 °C to 550 °C without any sulfurization/selenization were examined. Our results show that Al has no significant effect for the samples heat-treated at 350 °C. On the other hand, we observed successful doping for the samples heat-treated above 450 °C. Our EDS, XRD and Raman results suggest that Al is replacing Sn sites in these samples. The Al doping caused a drop in resistivity while increasing the charge-carrier concentration, bandgap and grain sizes. The biggest reduction in resistivity was observed for 2% doped and heat-treated at 450 °C samples. In these samples the resistivity reduced from 13.34 Ω.cm to 0.42 Ω.cm while charge-carrier concentration increased from 7.85×10¹⁷ to 1.74×10²⁰ cm^-3. Such improvement in charge-carrier concentration is promising for using Al-doped CZTS thin-films as hole transport layers in Perovskite solar cells, while an increase in the bandgap is advantageous for water-splitting applications.

CZTS-based solar cell is one of the most attractive p-type solar absorbers due to its low cost and
earth-abundant elements. In addition, CZTS has a band gap of 1.5 eV and high absorption
coefficient (10⁴ cm^-1 ) required to be a near ideal solar cell absorber. We report the impact of
silver (Ag) doping on CZTS solar cell and how the dopant mitigated the defect density. We
deposited 10 to 20 nm thick Ag layer on top of the CZTS precursor before annealing in sulfur
environment for crystallization of the Ag-doped CZTS (ACZTS) film. Ag-doping has shown to
improve V_oc , but we observed it is important to control the Ag for optimal p-type conductivity.
Although V_oc improvement was achieved, Ag doping of CZTS increased series resistance due to
more defects in the crystalline structure. Our recent study showed that by doping a small amount
of Ag could improve V_oc up to 0.65 (V), compared to bare CZTS which was 0.605 (V). It
improved by 10% compared to the cell with no Ag ions. The formation AgS₂ which has a higher
formation energy compared to other defects is expected, which in turn may have it lowered the
number of Cu_Zn antisite defects. Moreover, Ag ion has large atomic radius improving grain sizes
of ACZTS and lowering GB effect. The highest efficiency based on Ag-doping of CZTS showed
6.4% in our lab compared to the CZTS cell with ~6% efficiency due to the gain in V_oc, Jsc also
improved due to lowering Cu_Zn antisite defects.

The future deployment of photovoltaics demands stable, abundant, non-toxic materials. Kesterite Cu₂ZnSn(S,Se)₄ (CZTSSe) has attracted attention due to its non-toxicity, low cost, earth-abundant elements and its band gap can be tuned easily. However, efficiency of CZTSSe (12.6%) is much lower than CIGS (22.9%), mainly due to a large V_OC deficit. Several factors have been suggested for this V_OC deficit, including large amount of defects, undesirably thick MoS₂ layer at the rear of the CZTSSe cell, coexistence of secondary phases and unfavorable band alignment at the CZTSSe/CdS heterojunction interface. The importance of the conduction band offset (CBO) at the absorber/buffer interface is well known. A cliff structure (CBO<0) can increase interface recombination, while a spike (CBO>0) results in blocking of the photocurrent and reduce fill factor. Commonly, it is assumed that a small cliff is present at the CZTSSe/CdS interface, which highlights the need to replace CdS with materials with a lower conduction band in order to avoid an unfavorable band alignment. Furthermore, this CdS-based solar cell is desirable to replace with a more environment-friendly highly effective n-type buffer layer.
In this work, we have demonstrated the non-toxic Zn(O,S) as an alternative buffer layer for CdS by chemical bath deposition (CBD) and atomic layer deposition (ALD) . One is chemical bath deposition (CBD) that is simple and well prepared. The other is atomic layer deposition (ALD) that can deposit uniformity for atomic level as well as the conformal coverage over large-scale areas. The Zn (O, S) buffer layer band gap (3.5eV) is wider than CdS (2.4eV) which can achieve higher short-circuit current density (Jsc) due to the enhanced transmission of the wavelength region between 350 nm and 550 nm.
Finally, we obtained 5.4% efficiency of CZTSSe solar cell with open circuit voltage (V_oc) of 440 mV, short-circuit current density (J_sc) of 25 mA/cm², and fill factor (FF) of 50.2% by chemical bath deposition (CBD) ZnOS. Subsequently, we got 9.77%(~ 10.75% in the cell effective area) efficiency with open circuit voltage (V_oc) of 460mV, short-circuit current density (J_sc) of 36.07 mA/cm², and fill factor (FF) of 58.06 % by atomic layer deposition (ALD) ZnOS, which is the highest reported efficiency of CZTSSe with a Cd-free buffer layer. Apart from this, we successfully established the number of precursor cycles using the ALD process to control the O/S ratio and thickness. It helps us adjust the appropriate conduction band offset (CBO) to avoiding an unfavorable band.
The morphology, elemental composition, and distribution of the absorber layers are being examined by X-ray diffraction (XRD), X-ray fluorescence spectrometry (XRF), scanning electron microscopy (SEM), Transmission electron microscope (TEM), and X-Ray Photoelectron spectroscopy (XPS).

Reference:
[1] V. Tunuguntla, L.C. Chen and K.H. Chen. et al., J. Mater. Chem. A., 2015,3, 15324-15330.
[2] Y.R. Lin, L.C. Chen and K.H. Chen. et al., Nano Energy, 2015, 16, 438.
[3] W.C. Chen, K.H. Chen and L.C. Chen. et al., Nano Energy, 2016, 30, 762-770.
[4] C.Y. Chen, K.H. Chen, and L. C. Chen. et al., Nano Energy, 2018, 51, 597-603.

Selenization is a process of supplying selenium (Se) in gas phase typically at the elecated temperature. The process has been applied for the formation of chalcogenide thin films for electronic, optical and thermoelectric devices. However, the commonly used method for selenization has limitations such as using toxic gas H₂Se gas, high temperature and high equipment cost due to the necessity of two zone furnace system. Therefore, a low-cost and facile method of selenization with a precise control of Se supply is needed.
Herein, we demonstrate one-step and one-zone selenization process of chalcogenide thin films by a new Se source which is a mixture of inert Al₂O₃ and Se. It was found that Al₂O₃ in the mixture plays a role as a controller of Se vaporization and therefore by adjusting ratios of Al₂O₃ vs. Se, we were able to control selenization vapor supply. And mechano-chemical synthesis of Cu₂GeS₃ particles and the transformation of a thin film of Cu₂GeS₃ to that of Cu₂GeSe₃ through selenization will be discussed. During our synthesis of Cu₂GeS₃, only elemental Cu, Ge and S were used without any solvent and the thin film of Cu₂GeS₃ was fabricated by the solution phase, ink printing method followed by the annealing under Se environment. When the selenization method was used for Cu₂GeS₃ semiconductor thin film, we found resulting selenized thin film has better morphology and electrical properties than that selenized with Se powder alone. This alternative method consumes much less Se than previous mehtods so that it can be considered environmentally less toxic and has a benefit of low-cost and facile process. And also, this process can broaden applicability by easily controlling the band gap of various chalcogenide materials and thin films.

Current world record CIGS solar cells are achieved by post-deposition treatment (PDT) of heavy alkali metal fluorides (such as KF, RbF and CsF). These alkali-PDT processes are predicted to form alkali-containing compounds such as K-In-Se and Rb-In-Se. The structure and chemistry have intensively been investigated so far. However, the properties of Cs-containing compounds formed by CsF-PDT have not yet been widely discussed. We thus investigated the influence of CsF-PDT on the formation of Cs-related compounds, and its beneficial and detrimental effects on solar cell performance.
CIGS thin films were deposited using three-stage deposition process on Mo-coated SLG substrates. CsF-PDT was performed at a substrate temperature of 350 ^oC for 2 minutes. The CIGS solar cells were fabricated with MgF₂(105nm)/Ni/Al (2-2.5μm)/ZnO: Al(300nm)/ZnO(100nm)/CBD-CdS/CIGS (2.5-3μm)/Mo (700nm)/SLG (1.8mm) structure. TEM and SAD analyses revealed that the Cs-related secondary compounds are present on the surface of CIGS thin film after CsF-PDT. XPS narrow scan spectra confirmed Cu depleted regions at the surface area with a few tens of nanometers, whereas SIMS and EDS-line scan showed that Cs was distributed throughout the CIGS absorber. We noticed that CsF-PDT modifies CIGS surface and formed Cs-related compounds such as Cs-In-Se or Cs-In-Ga-Se. These Cs-related compounds provide several beneficial electronic effects. Presence of Cs-related compounds at the surface of the CIGS thin film reduce interfacial recombination, thereby, improving open-circuit voltage and efficiency. Recombination mechanisms in the devices were characterized using temperature dependent current density-voltage measurements and time-resolved photoluminescence based on different Cu/([In+Ga]) and Ga/([In+Ga]) ratios. CsF-treated CIGS thin films with near-stoichiometric (Cu/([In+Ga]) yields higher efficiencies. Furthermore, we found the important role of Cs-related compounds in heat-light soaking (HLS) and heat-bias soaking (HBS).

CIGS solar cell is a high-efficiency cell with more than 22% efficiency, and can also be modularized in a large area. CIGS module efficiency is approaching the multicrystalline silicon module efficiency. For safety and supporting reasons, photovoltaic (PV) modules have to be grounded and PV modules are serially connected to generate high voltage and power. When several hundreds of volts induced to modules, degradation is occurred. This is called potential induced degradation. The term ‘potential induced degradation (PID)’ was first introduced by S. Pingel and coworkers in 2010 [1]. Efficiency deterioration by PID phenomenon has been reported in large capacity photovoltaic systems. In c-Si solar cells, decrease in shunt resistance by Na⁺ ion immigration is known as the cause of PID. However, the mechanism during PID in CIGS solar cells is not known yet. For commercialization of CIGS thin film solar cells, PID study in CIGS solar cell is important. Since PID is known to be common in modules, in this research, an experimental structure was designed to replicate PID in cell level. As an experimental method, the cell was degraded with 600 V bias at 65 °C in darkness. Light I-V, dark I-V and external quantum efficiency (EQE) were performed to investigate the degree of degradation. To image the position of PID by electroluminescence (EL) and lock in thermography (LIT) measurements were also performed. We tried to extract the mechanism of PID in CIGS solar cells through the change of element or doping concentration. Furthermore, recovery was confirmed with no bias at room temperature in darkness.

[1] Proceedings of the 35th IEEE PVSC 20–25 June 2010 pp. 2817-2822

Cu(In,Ga)Se₂ (CIGSe) has emerged as an attractive thin-film solar cell material owing to its high light absorption coefficient and tunable bandgap. In CIGSe processing and fabrication, the use of alkali treatments has been implemented extensively. Sodium treatment is considered a requirement for high efficiency CIGSe solar cell as it was found to improve the performance and growth of CIGSe absorber films. However, one of the more significant developments in recent years has been the discovery of the beneficial effects that potassium treatments have on CIGSe solar cells, particularly on the absorber surface. When applied, it was found that potassium induces copper and gallium depletion from the surface of CIGSe films. Moreover, a high-bandgap K-In-Se phase also forms on the surface of films subjected to potassium treatments. In practice, these two phenomena lead to an improved CIGSe device by mainly improving the quality of CIGSe/CdS p-n heterojunction.
Despite these significant beneficial effects, the use of potassium has been limited to expensive vacuum-based CIGSe processing. Here, we develop a route to apply potassium treatments to low-cost solution-processed CIGSe films grown from colloidal sulfide-based nanoparticle inks. By adding potassium through e-beam evaporation of KF prior to growth, we find that the grain growth of CIGSe is enhanced with potassium addition and that a larger-grained thicker coarsened film results compared to untreated selenized CIGSe film, similar to what is observed in sodium-treated films. We also observe via XPS that films treated with potassium show depletion of gallium from the surface, and that the amount of gallium depletion correlates with the amount of potassium added. Similar effects on film surface were also observed for films that have undergone potassium treatment through a soak in KCl solution prior to selenization. Furthermore, it was also observed through XPS that films treated with KF showed the presence of a different Se chemical state on the surface, which is attributed to a K-In-Se phase. Devices that were fabricated with potassium treatments were found to have enhanced optoelectronic performance which was mainly manifested in higher open-circuit voltage and higher fill factor, compared to films that are untreated or treated with sodium only.
In conclusion, we purpose a route to apply potassium treatments for solution-processed CIGSe devices. We also conduct a systematic study on the effects of potassium on CIGSe film growth, surface environment, and optoelectronic properties.

Tandem solar cells are a promising route to increase the efficiency of solar cells. They require semiconductor materials with a bandgap wider than that of silicon, e.g. GaAs, GaInP, CuInGaSe. However, due to vacuum processing of these materials, their price remains an order of magnitude higher than that of silicon. We propose to utilize chalcogenide based colloidal quantum dot (CQD) inks to achieve scalable and low-cost solution processing of the state-of-the-art photovoltaic materials. Films are prepared by thermally annealing and sintering the QDs, resulting in bulk direct bandgap semiconductors, which have been demonstrated to be highly efficient light absorbing layers. The key challenges in this field revolve around improving CQD packing and further reducing the defect densities. We will demonstrate that inorganic atomic ligands can eliminate the need for organic long chain ligands required for colloidal stability, and at the same time can be used as abundant passivating agents in the resulting films. We will demonstrate the proof-of-principle results based on lead-free chalcogenide based CQDs, primarily CdTe and CuInS.

The large and persistent photoconductivity displayed by some semiconductors provides a way to control magnetism with light, through illumination-control of free carrier concentration and thereby magnetic interaction in dilute magnetic semiconductors. CdS is a wide band-gap semiconductor that displays large and persistent photoconductivity, and is predicted to become magnetic when doped with certain non-magnetic dopants including Boron [1]. In this work, we experimentally test the prediction of magnetic CdS:B, and the hypothesis that magnetism can be controlled by photoconductivity. We make CdS:B nanoparticulate powders by co-precipitation [2]. We use X-ray diffraction and plasma optical emission spectroscopy to quantify boron doping, and we characterize the effect of doping on electronic structure using optical spectroscopy. We use nuclear magnetic resonance (NMR) and magnetometry to confirm the presence of magnetic B, and to study the position of B in the CdS lattice. Finally, we report the effect of above- and below-band gap illumination on the magnetism of B sites and the magnetic order of CdS.

[1] P. O. Bedolla, C. Gruber, P. Mohn, and J. Redinger, “p-electron magnetism in CdS doped with main group elements,” J. Phys. Condens. Matter, vol. 24, no. 47, p. 476002, Nov. 2012.
[2] A. Fakhri and R. Khakpour, “Synthesis and characterization of carbon or/and boron-doped CdS nanoparticles and investigation of optical and photoluminescence properties,” J. Lumin., vol. 160, pp. 233–237, Apr. 2015.

Degenerate n-type semiconductors such as SnO₂:F, In₂O₃:Sn or ZnO:Al, and of p-type, Cu_2-xS, Cu_2-xSe, Cu_2-xO of electrical conductivity superior to 10³ Ω ^{– 1} cm ^{– 1} and carrier concentrations of 20²¹ cm ^{– 3} make use of degeneracy of energy levels, the same way as metals do. They behave as conductors at wavelengths of electromagnetic radiation above 1500 nm, with high reflectance and low transmittance. In the case of thin films (150 – 200 nm) of copper chalcogenides, with optical bandgap of 1.3 to 2.5 eV, this leads to a bell-shaped transmittance curve, with near-zero transmittance at wavelengths below 500 nm and very low as well at wavelengths above 2000 nm. The overall optical transmittance of 25 – 30 % has a maximum at 550 – 650 nm, absorptance of 60 – 70 % and reflectance of 15 – 20 % for solar radiation. We deposited copper sulfide selenide thin films on the outer wall of on one side of cellular polycarbonate sheet by floating it on a chemical deposition solution of copper nitrate, thiourea, and sodium selenosulfate. This film, protected with food-safe adherent polyethylene sheet, is fixed at 2 cm from a layer of sliced fruit (strawberry) or fruit pulp (blackberry) placed under the sun. Due to the high solar absorptance of 60 %, the device acts as a 600 W/m² heater and raises the temperature of the fruit layer to 60 ^oC, which is optimum to conserve food value during drying and also to assure a viable drying rate. The cellular structure inhibits convectional heat loss to the ambient from the top (exposed) side of the sheet, without the copper chalcogenide coating. Further, the absence of UV radiation notably retains the antioxidant nutritional value of the product. This advantage is detected in free radical (DPPH) aided spectrophotometric characterization. The UV-inhibited drying also retains the appearance of the product. This approach is may be directly applicable in a rural setting, where a timely drying of fruit products during harvest reduces large-scale loss.

Can we build in “our” laboratory a functional photovoltaic prototype module, which is stable under the sun and is open to innovation to achieve technological viability? Yes. We present the prototype module, FTO/CdS/Sb₂S_xSe_3-x/C-Ag, which is stable, functional to light up 20 mW blue light emitting diode and open to innovation in any laboratory. It is built on commercial F-doped SnO₂ (FTO) substrate of 7.5 cm x 2.5 cm in area, with an 80 nm CdS thin film deposited in 1 h form a chemical bath at 80 ^oC. Subsequently four sequential depositions of Sb-S-Se are made on it in a total of 10 h at 80 ^oC from chemical bath containing potassium antimony tartrate, thioacetamide and sodium selenosulfate. This film is of 350 nm in thickness, and amorphous. The sun appears deep red when viewed through it. Seven graphite electrodes of 2 cm x 0.5 cm at a separation of 0.4 mm are placed on it by hand painting colloidal acrylic based graphite suspension. This structure now goes into an oven at 250 - 290 ^oC for 05 – 1 h, under a nitrogen flow. The Sb-S-Se film surrounding the cell is etched-off with NaOH solution, and the underlying CdS with dilute HCl. Using a SiC scribe the cells are isolated. The heating renders the graphite electrodes of sheet resistance 25 Ω and the underlying layer is transformed to a carbon-doped p-absorber. The Sb-S-Se film is crystallized, with the composition, Sb2S_0.9Se_2.1 with a bandgap of 1.42 eV, and electrical conductivity 10 ^{– 5} Ω ^{– 1} cm ^{– 1} under light. An estimated light generated current density, 28 mA/cm² under the sun should be available in a solar cell. The prototype module is completed by applying colloidal silver paint on the graphite electrode and interconnecting it to the silver-paint electrode on the FTO base of the neighboring cell. A commercial acrylic lacquer encapsulates the module. The prototype is of Voc 3.1 V, Jsc 22 mA, maximum power 30 mW. Innovation is possible with modification/ replacement of the CdS window, compositional (bandgap) variation along the thickness and heat-processing of the absorber film, and in the application of back contact. We also present results on module produced by combining chemical deposition and thermal evaporation in the production of the absorber film, whereby bandgap is modified along the absorber thickness.

An attractive class of semiconducting materials with non-toxic, low-cost and highly efficient characteristics have emerged as promising alternatives for applications in solar cell technologies. Novel materials or structures with properties that match the spectral distribution of the solar spectrum are therefore required to produce materials of high efficacy. One copper-based semiconductor metal chalcogenide, copper antimony sulphide (CAS) as thin films, is an excellent candidate for potential photovoltaic (PV) applications, since, in addition to the elements’ abundance, low toxicity and simple inexpensive processing protocols, bulk CAS has a strong absorption coefficient of 10⁵ cm^-1 and a band gap between 0.5 and 2.0 eV, matching well with the solar spectrum as well as excellent photostability. Interestingly, there are sparse reports of tuning the optoelectronic properties with metal dopants for the CAS material and few reports have been written for the synthesis of CAS thin films via the aerosol assisted chemical vapour deposition (AA-CVD) method. As-synthesized thin films of CAS microstructures have various morphologies and stoichiometries which are dependent on AA-CVD parameters of deposition temperature and time. Homogeneous films of deposition temperatures 500, 550 and 600 °C, synthesised at 30 minutes and 1 hour are non-stoichiometric with mixed morphologies and particle sizes range from 0.3 to 1 µm. Consequently, the optical properties indicate that as-synthesized CAS thin films absorb near the end of the ultra-violet region but mainly in the visible portion of the spectrum. Indirect transitions occur between 0.89 to 3.40 eV. With detailed elucidation from powder X-ray diffraction, Raman spectroscopy, scanning electron microscopy, atomic force microscopy, energy-dispersive X-ray spectroscopy, X-ray photoelectron spectroscopy, steady-state and time-dependent electronic spectroscopy we have been optimizing AA-CVD protocols to effectively control the materials properties, and if achieved, this will provide possible enhancement of photoconductivity and furthermore, PV efficiency of thin films.

Of orthorhombic crystalline structure, Sb₂S₃ and Sb₂Se₃ with optical bandgap (E_g) of 1.88 and 1.1 eV respectively can produce solar cell absorbers with E_g of 1.3 – 1.6 eV, which can offer short circuit current density of 25 – 35 mA/cm². Antimony sulfide selenide thin film solar cells reported from our group has conversion efficiency (η) of 5 – 6 % [1]; and from other groups, it had touched 9 %. To be commercially viable, this value should move toward 20 %. We present methodologies to improve material properties in antimony sulfide selenide thin films produced by thermal evaporation of chemical precipitate or commercial powders of the binary compounds and their mixtures. Chemical precipitates are produced in our laboratory from solution mixtures of potassium antimony tartrate, thioacetamide and sodium selenosulfate. To improve the crystalline grain diameter, SbCl₃(melting point, 94 ^oC) is added to the evaporation source mixture. To prevent Sb₂S_xSe₃ film becoming n-type, Se powder is added to the mixture. The material produced this way has the mobility-lifetime product improved by an order of magnitude, to 10 ^{– 6} cm² V^{– 1}. However, this value is still an order of magnitude inferior to that of CdTe. We also incorporated Te into the film at 3 – 9 atomic percentage, prompted by two basic features of Sb₂Te₃: an electron mobility (270 cm² V ^{– 1} s ^{– 1}) superior to that in the other two chalcogenides and a higher p-type electrical conductivity in it due to the low activation energy for Sb-vacancy creation. The E_g of the materials produced is 1.48 – 1.54 eV. All these modifications have led to the development of solar cells with 6.5 % efficiency at absorber thickness 350 – 400 nm and in cell area, 0.5 cm². In prototype modules of area 7 cm², it is 4.8%. These cells are encapsulated with commercial resin, for keeping them stable under the sun over many days of operation.
[1] Fabiola De Bray Sánchez, et al: Optimum chemical composition of antimony sulfide selenide for thin film solar cells, Appl. Surf. Sci., 454 (2018) 305-312. DOI: 10.1016/j.apsusc.2018.05.076.

One of the most attracting applications for emerging chalcogenide electronic materials is the use of low-dimensional/quasi-covalent compounds as absorbers for direct solar energy conversion into electricity. In this sense, 1D-Sb₂(S,Se)₃ is becoming a relevant thin film chalcogenide semiconductor. In particular, 1D-Sb₂(S,Se)₃ has shown remarkable improvements in the last few years, demonstrating solar cells in superstrate configuration with power conversion efficiencies higher than 7%. In fact, and similarly to CdTe, most of the devices reported in the literature so far have been prepared using this configuration, whiles the substrate configuration option is being marginally used. Additionally Sb₂(S,Se)₃ has shown a high degree of flexibility thanks to the relatively low temperatures required for the synthesis and crystallization of this compound (300-400 C), being fully compatible with polymeric, steel, ceramic and TCO/glass substrates. This versatility makes this compound very promising for ubiquitous applications such as building integrated photovoltaics, wearables, or autonomous IOT applications among others. Nevertheless, all these technological applications require the urgent development of efficient devices in substrate configuration.
In this work we present a systematic optimization of the synthesis of 1D-Sb₂Se₃ thin films onto conventional Glass/Mo back contact using standard substrate configuration. For the synthesis of 1D-Sb₂(S,Se)₃, an innovative sequential process based on reactive annealing under Se atmosphere of thermally evaporated Sb layers was developed. The study is centered in the analysis of the reactive thermal annealing conditions (temperature, time, pressure) on the compositional, structural and morphological properties of the 1D layers. By adjusting the annealing parameters, we obtain continuous layers with large and homogeneous grains, excellent crystalline quality, and 1D preferential orientation perpendicular to the substrate. We will show that the annealing temperature and Se quantity control in the layer, are the key parameters for controlling the morphology, as well as the structural, optical and electrical properties.
After a first optimization, we report a 1D-Sb₂(S,Se)₃/CdS heterojunction with a promising power conversion efficiency of 5.3% in substrate configuration, demonstrating a V_OC of 403 mV (the highest reported value for this configuration to the best of our knowledge).
Additionally, this study is complemented with a wide range of fundamental characterization techniques including photoluminescence, Raman, SEM, XRF, XRD, and with a complete analysis of the impact of the absorber stoichiometry under different regimes (Se-rich, Se-poor, Sb-rich and Sb-poor conditions). All this will be correlated with the optoelectronic characterization (JV, IQE, CV) of the solar cells. Finally, the main urgent challenges to develop Sb₂Se₃ type solar cells in substrate configuration will be presented and discussed.

Among SnS-CUB, SnS-ORT, SnSe-CUB and SnSe-ORT of cubic and orthorhombic crystalline structures, tin chalcogenides offer an interval of bandgap (E_g) of 1 – 1.74 eV of p-type conductivities owing to the relative abundance of Sn(II) vacancies in them, acting as acceptors. These convert to n-type semiconductors, SnS₂ or SnSe₂ of E_g 1.7 – 2.2 eV when heated at 350 – 420 ^oC in presence of sulfur or selenium vapor – in which donor centers created due to S – or Se vacancies dominate. We find that heating SnS-Se-CUB films at 350 ^oC leads to SnSe-ORT and when the heating is done in presence of Se, it forms n-type SnSe₂, both of conductivity 0.01 Ω ^{– 1} cm ^{– 1} allowing us to build thermoelectric couple with a Seebeck coefficient of 1 mV/K to explore further. In a solar cell structure of SnO₂:F/CdS/SnS-CUB:Ag/C-Ag, a conversion efficiency of 1.5% is achieved, opening up a path toward improving tin sulfide solar cells produced by chemical deposition via diffusion of Ag provided through ion-exchange reaction of SnS-CUB in a dilute AgNO₃ solution. An encouraging new result is that heating of SnS-CUB at 450 ^oC in presence of S-vapor produces phase-pure p-type Sn₂S₃ film with an E_g of 1 – 1.4 eV and a light generated current density for solar cells in excess of 30 mA/cm². Such conversion takes place via an intermediate formation of SnS₂ due the descending values of the enthalpy of formation of these materials: in kJ/mol, – 91 to – 99 (SnS-rock salt - cubic or ORT); – 131 (SnS₂); and – 230 (Sn₂S₃). With energy level diagram very close to that of SnS-ORT, and with an advantage of p-Sn₂S₃ to be phase-pure, free from detrimental n-SnS₂, here is an alternative tin chalcogenide for solar cells. A major attraction of Sn₂S₃ is that it may stabilize also an n-type semiconductor, as per theoretical model. With a low lattice thermal conductivity, it is also poised to be a prospective thermoelectric element.

Tin monosulfide (SnS) is expected as a light-absorber material for next-generation solar cells due to its advantageous properties: the optimal band gap (~1.0 eV), high light absorption coefficient just above the band gap, and non-toxic and earth-abundant constitutional elements. Since Sn vacancies are easy to form in SnS due to their low formation enthalpy and introduce holes, SnS naturally exhibits p-type conduction and inversion of the carrier type to n-type by doping is quite difficult [1,2]. Nevertheless, n-type SnS was reported in very limited case, that is, heavily Pb-doped SnS (Sn_1-xPb_xS, x>0.15), which would, however, cancel the advantage of non-toxicity of SnS [3]. SnS-based solar cells fabricated so far accordingly employed hetero-junction structures, for instance, p-SnS/n-CdS or p-SnS/n-Zn(O,S) [4]. These devices involving the hetero-junction exhibited low conversion efficiencies of ~5% at highest because of unfavorable band offset and/or defect states at the hetero-junction interface suffering from interface recombination [5]. In order to improve the conversion efficiency of the SnS-based solar cells, n-type SnS which enables p/n homo-junction is therefore indispensable.
In this study, we grew single crystals of Cl-doped SnS (i.e., SnS_1-xCl_x) using SnCl₂ as a flux, expecting that chlorine substituting sulfur of SnS would generate electrons. The mixture of Sn, S and SnCl₂ powders were sealed in the vacuum silica glass tube and heated at 520 °C for 24 h, and subsequently cooled down to 240 °C over a period of 36 h and then to room temperature over a period of 12 h. The remaining flux (SnCl₂) was washed away with ethanol and acetone.
Single crystals of SnS in the form of shining lamellate with a surface in perpendicular to the a-axis were obtained. Typical surface area and thickness of the crystals were from 10 to 20 mm² and from 28 to 39 μm, respectively. EPMA analysis indicated that Cl was homogeneously distributed in the single crystal with a concentration of 0.14 at% and segregation of Cl was not found.
Hall coefficient of the single crystal was negative, which indicates that the carrier type is successfully inversed to n-type by Cl doping. The Fermi level of the Cl-doped SnS single crystal determined by XPS was 0.1 eV below the conduction band minimum, whereas that of undoped p-type SnS is generally located at 0.2-0.3 eV above the valence band maximum [2,5], which clearly supports that Cl-doped one is n-type. Hall mobility of electrons of Cl-doped SnS single crystal measured in perpendicular to a-axis was 101 cm²V^-1s^-1, which is at the same order of magnitude as the Hall mobility of holes of undoped SnS single crystal (100-500 cm²V^-1s^-1 [6]).

[1] Z. Xiao et al., Appl. Phys. Lett., 106, 152103 (2015).
[2] A. Schneikart et al., J. Phys. D: Appl. Phys., 46, 305109 (2013).
[3] F. Y. Ran et al., Sci. Rep. 5, 10428 (2015).
[4] P. Sinsermsuksakl et al., Adv. Eneg. Mat., 4, 1400496 (2014).
[5] L. Sun et al., Appl. Phys. Lett. 103, 181904 (2013).
[6] L. A. Burton et al., Chem. Mat., 25, 4908 (2013).

Orthorhombic tin sulfide (SnS) has recently been emerged as a very promising absorber material for thin-film solar cells (TFSCs). It has an ideal optical band gap (~1.3 eV) and it comprises of relatively earth abundant constituents and non-toxicity. But till date, the highest efficiency obtained from the SnS-based solar cells is 4.36%, which is fairly low compared to its theoretical limit of ~32%. SnS is a non-cubic material unlike CIGS or CdTe, crystallizing in an orthorhombic structure (JCPDS No. 39-0354, a = 4.3291 Å, b = 11.1923 Å, c = 3.9838 Å). It easily leads to the formation of layered features. Therefore, controlling the morphology of the SnS absorber with dense and pinhole-free grains is crucial.
In this study, the influence of vapor transport deposition (VTD) conditions of tin sulfide (SnS), i.e., growth temperature and duration, on the formation of secondary phases, preferred orientation, and solar cell performance, was investigated. In the growth temperature effect experiment, the morphology is grew as plate form with increasing temperature and the secondary phase was found at low temperature. Also, it was confirmed that the film thickness wase increased linearly with duration and When the growth duration increases to 10 min, a dramatic improvement in the device performance is noted. Finally, fabricated SnS TFSCs (thin film solar cells) achieved near 4% efficiency (V_OC; 0.342 V, J_SC; 19.8 mA cm^-2, FF; 58.0%) at 600 ^oC growth temperature and 10 minutes duration. In addition, working stability under continuous illumination and damp-heat (85/85) conditions was investigated for the best cell in this study.

Germanium selenide, GeSe, is a member of the IV-VI family of chalcogenide semiconductors, including the champion thermoelectric SnSe, and has itself been studied in the context of many electronic applications, including thermoelectrics, photocatalysis, phase-change memory and photovoltaics.^1,2 Despite an increasing amount of research interest in recent years, some questions still remain over its fundamental properties, particularly its bulk band gap, recorded values of which range from 1.06 eV to over 1.5 eV with uncertain characterisation as indirect or direct.^3,4 Theoretical values using standard Density Functional Theory (DFT) also demonstrate similar variation, although the monolayer is consistently a direct gap semiconductor.⁵ Given the fundamental importance of the band gap to a material’s use within optoelectronic applications, clarification is necessary to establish the most optimal application for this material.
In this study, we use multiple levels of theory, from standard and hybrid DFT to quasiparticle self-consistent GW calculations in combination with Fourier-transform Infrared Spectroscopy (FTIR) measurements of single crystal and thin films of GeSe to characterise its band gap and provide further understanding into its suitability as a solar absorber. We also discuss how the balance between in-sheet covalent bonding and van der Waals interactions between layers affects the electronic properties of the material, including the influences that have previously lead to variations in recorded band gap. Finally, we examine its dynamical stability in comparison with Raman spectroscopy to examine the structural behaviour of the system, and how this can further influence the overall behaviour of GeSe. We believe these results will help influence future interest in GeSe as a solar absorber as well as further discussion surrounding the IV-VI family.

(1) Xue, D. J.; Liu, S. C.; Dai, C. M.; Chen, S.; He, C.; Zhao, L.; Hu, J. S.; Wan, L. J. J. Am. Chem. Soc. 2017, 139 (2), 958.
(2) Cagnoni, M.; Führen, D.; Wuttig, M. Adv. Mater. 2018, 30 (33), 1801787.
(3) Kannewurf, C. R.; Cashman, R. J. J. Phys. Chem. Solids 1961, 22 (C), 293.
(4) Elkorashy, A. M. Phys. Status Solidi 1986, 135 (2), 707.
(5) Xu, Y.; Zhang, H.; Shao, H.; Ni, G.; Li, J.; Lu, H.; Zhang, R.; Peng, B.; Zhu, Y.; Zhu, H.; Soukoulis, C. M. Phys. Rev. B 2017, 96 (24), 245421.

The photovoltaic properties of selenium have been discovered over 140 years and were utilized at that time to produce the first solid-state ‘thin-film’ solar cell. This discovery boosted modern research and technologies based on semiconductors, and lead to the implementation of selenium in many functional devices. Upon the rise of competitive semiconductors and mainly silicon, the usage in selenium slowly declined. In recent years, the photovoltaic community gains new interest in this simple elemental semiconductor as a possible absorber for extremely low-cost and highly scalable solar cells. For this application, selenium shows few desirable properties such as high absorption coefficient, intrinsic environmental stability and fabrication at low-temperature (below 200°C).
In this research, we first study the main parameters that are required to optimize the quality and efficiency of selenium-based solar cells, using modern tools and approach. We indicate the importance of careful annealing of the selenium film, very close to its melting temperature to obtain high-quality films. As selenium has an anisotropic structure, composed of 1D chains, we study how the alignment of the crystals in the thin film changes the device properties. Finally, we indicate the importance of light illumination (and excitation) of the film during the fabrication process to enhance and improve the crystallization of the film.
Following these results, we study how the selenium bandgap (1.7 eV), can be tuned to the optimal value for photovoltaic absorber (1.2-1.4 eV), by alloying selenium with the isomorphic low bandgap semiconductor tellurium. We found that the addition of tellurium causes a strong non-linear shift of the alloy’s valance band minimum energy, leading to a non-linear decrease of the bandgap. This non-linear shift enables to reach the desired bandgap by alloying a fairly small amount of tellurium (~10%). Photovoltaic devices based on the selenium-tellurium alloy indeed show improved current density and spectral response as a result of the optimized bandgap. The overall efficiency of the selenium-tellurium devices is still lower than the pure selenium devices, due to lower voltage and fill-factor, but we indicate the main causes to the lower properties and suggest some improvements to the device structure which should improve the device efficiency.

(1) Hadar, I.; Song, T. Bin; Ke, W.; Kanatzidis, M. G. Modern Processing and Insights on Selenium Solar Cells: The World’s First Photovoltaic Device. Adv. Energy Mater. 2019, 1802766.

The demand for high resolution and high-speed imaging devices has increased with the emergence of next-generation broadcasting systems such as full-featured 8K Super Hi-Vision with 33 million pixels and 120 Hz frame rate, which are respectively 16 and 4 times greater than those of current 2K Hi-Vision systems. However, with increasing number of pixels and frame rate, the sensitivity of the image sensor decreases. The sensitivity decreases because of a reduction in the amount of light received per pixel and per frame. To address this problem, we have been studying the stack-type complementary metal–oxide semiconductor (CMOS) image sensor overlaid with a poly-crystalline selenium (c-Se)/gallium oxide (Ga₂O₃) photodiode [1][2]. In this image sensor, high sensitivity can be realized by multiplying photogenerated charges using the avalanche phenomenon. c-Se has a high absorption coefficient in the entire visible light range compared to silicon (Si) that is mainly used to make conventional photodiodes. Therefore, incident light can be absorbed in a thinner film and charge multiplication can be induced using relatively low voltages compare to conventional Si photodiodes. Ga₂O₃ has a wide bandgap of 4.5 eV; the large energy barrier between the work function of indium tin oxide as an electrode and the valence band of Ga₂O₃ prevents hole injection from the electrode, thereby increasing the dark current.
To further reduce the operating voltage required for avalanche phenomenon in the c-Se/Ga₂O₃ photodiode, we investigated the effects of doping impurities into Ga₂O₃ and c-Se. First, Sn-doped Ga₂O₃ was fabricated via RF sputtering using a mixed target of tin oxide (SnO₂) and Ga₂O₃. By increasing the concentration of Sn that served as a donor, the photocurrent increased at a low voltage because the increase in Ga₂O₃ carrier concentration achieved via Sn doping caused the depletion layer to mainly spread into c-Se. Then, tellurium (Te)-doped c-Se was prepared by annealing amorphous Se (a-Se) formed on a thin Te nucleation layer. By increasing the carrier concentration of c-Se via the diffusion of Te that served as an acceptor, the avalanche phenomenon could be observed at a low voltage because a high electric field was applied to c-Se. By adding impurities to both Ga₂O₃ and c-Se, an avalanche multiplication factor of 10 could be obtained at an operating voltage of 16 V, which is lower than the operating voltage of non-doped photodiodes (21 V). We believe that the c-Se/Ga₂O₃ avalanche photodiode will lead to the development of high-sensitivity image sensors.
[1] S. Imura et al. IEDM Tech. Dig., (2014) 88
[2] S. Imura et al. Apl. Lett. 104, (2014) 242101

Imaging using x-ray radiation plays a significant role in advancing industry and fundamental research by utilizing the highly penetrating nature of x-rays to study the internal composition of materials. Image quality is fundamentally limited by the number image forming photons, which is governed by the detector quantum efficiency (QE). Conventional high spatial resolution scintillator-based indirect conversion detectors have poor QE due to thinning of the scintillator to minimize secondary optical scatter. This makes them non-ideal for emerging scientific imaging tasks such as phase-contrast x-ray imaging (e.g. for visualizing low density composite materials).

We have developed a direct conversion detector by integrating amorphous selenium (a-Se) photoconductor material and a CMOS readout circuit with 7.8 µm pixel pitch. Unlike the optical scatter in scintillators, the spread of absorbed energy from x-ray interactions in the photoconductor does not significantly degrade spatial resolution as the a-Se thickness is increased. The detector was vertically integrated by CMOS post-processing using thermal evaporation. An a-Se layer of 118 µm and top biasing electrode were directly deposited. After packaging, wire-bonding was performed below the glass transition temperature of a-Se to prevent crystallization. Finally, a high-voltage connection was provided from the package to the surface of the electrode. A biasing voltage of 500 V was applied to the electrode, resulting in an internal electric field of 4.2 V/µm.

The detector was characterized at 60 kV with 2 mm Al filtration. Using the slanted-edge technique the pre-sampling MTF of the detector was measured. The line-spread function had a full width at half-maximum of 8.7 µm and in the spatial frequency domain the MTF decreased to 30% at Nyquist frequency, ≈64 cycles/mm, i.e. equivalent to 7.8 µm spatial resolution. The theoretical upper bound on the DQE at zero spatial frequency, DQE(0), is equal to the detector QE of 0.49. The measured DQE(0) was determined to be approximately 0.42, a decrease due to the presence of noise and additional pathways of signal loss.

To the best of our knowledge, this characterization demonstrates the highest spatial resolution direct conversion detector reported for the hard x-ray regime. In addition, our results suggest that a-Se photoconductors offer a high detection efficiency alternative to scintillator technology for high spatial resolution x-ray imaging tasks such as phase contrast x-ray imaging.

Hexagonal selenium with a direct band gap has been developed for optoelectronic applications through the last 130 years. Most advances have been made using vacuum deposition or solution methods. Herein we demonstrate a simple two-stage melt processing (TSMP) method to incorporate selenium in printable triple mesoscopic solar cells under ambient conditions. While a simultaneously triggered polymerization and depolymerization between several types of Se chains happen during the melt processing, we successfully realize phase-pure hexagonal selenium inside the mesopores with high crystallinity. We found that the TSMP method has an obvious effect on the CB energy level, band gap and crystal phase of selenium by ultraviolet electron spectroscopy, ultraviolet-visible absorption spectroscopy and in situ X-ray diffraction. Compared with the single melting process, the power conversion efficiency of the printable mesoscopic device was increased eight times to 2% through TSMP. These findings provide a new strategy for the melting process to obtain a more efficient photovoltaic device.

Two-dimensional transition metal dichalcogenides (2D TMDs) have been brought into the limelight due to their unique optical, electronic, and mechanical properties, which enables them for a wide range of applications.[1,2] Defined by the chemical formula MQ₂, where M is a transition-metal and Q = S, Se, Te, each monolayer of a 2D TMD is composed of a metallic plane sandwiched by two Q planes, where the coordination around the metal atoms can be trigonal prismatic, octahedral, and distorted octahedral (named 2H, 1T, and 1T' phases, respectively).[3] Despite the great technological and scientific interest drawn by those compounds, the atom-level understanding of how the physicochemical properties of 2D TMDs evolve with system size is still far from satisfactory. Thus, we report ab initio density-functional theory investigations of 2D TMDs nanoflakes with diverse sizes and chemical compositions, namely (i) (MoSe₂)_n, with n = 15, 63, 108, 130, 154, and 192;[4] (ii) (MoQ₂)_n, Q = S, Se, Te with n = 1 – 16;[5] and (iii) (WQ₂)_n, Q = S, Se, Te, with n = 1 – 16, 36, 66, and 105.[6] We found that the 1T' phase has the lowest energy for small MoSe₂ nanoflakes, and a size-induced 1T' → 2H phase transition occurs, which is mainly due to the higher edge formation energy of the 2H phase. Furthermore, for the n = 1 – 16 size range for both MoQ₂ and WQ₂ nanoflakes we observed a transition in energetic preference from structures elongated in one dimension with almost-equilateral trusses composed metallic lattices (1D), to 1T' nanoflakes. The mechanism of stabilization of the 1T' nanoflakes in relation to the 1D geometries results from a combination of two stabilization effects, namely the edge charge modulation in the edges of the structures, and the distortion of the octahedral geometry throughout the core of 1T' nanoflakes. Both 1D and 1T' maintain the same Q-terminated edge configurations, which stabilizes both morphologies in relation to 2H stoichiometric monolayer fragments. Since the edge charge effects are more intense in the order Te < Se < S, the cluster size in which 1D → 1T' occurs increases in the order MS₂ < MSe₂ < MTe₂, M = Mo, W. Furthermore, for the WQ₂ nanoflakes we have established systematically that the most stable nanoflakes are composed of junctions between triangular building blocks. The geometric forms originated from those junctions are stabilized due to maximization of bonds formed within the core of the nanoflakes. Therefore, by extending the sizes of the nanoflakes we could extend our understanding to the interplay of nanoflake size and stability. As a whole, our investigations could establish that the mechanisms of stabilization of the group-6 TMDs nanoflakes can be summarized as follows: (a) conservation of the Q-terminated edge configuration; (b) geometry of the nanoflakes as junctions of triangular building blocks; (c) edge charge modulation effects in order to stabilize 1T' nanoflakes opposed to 1D stripes; and (d) octahedral distortions responsible for breaking the degeneracy of electronic levels. Therefore, we could establish the criteria for the proposition of stable MQ₂, M = Mo, W; Q = S, Se, Te nanoflakes, and our findings can help to explain independently obtained experimental results.[6]

[1] Chhowalla, M. et al. Nat. Rev. Mater. 2017, 2, 17033–17047.

[2] Besse, R. et al. Phys. Rev. B: Condens. Matter Mat. Phys. 2016, 93, 165205.

[3] Besse, R. et al. J. Phys. Chem. C 2018, 122, 20483–20488.

[4] Caturello, N. A. M. S. et al. J. Phys. Chem. C 2018, 122, 27059–27069.

[5] Da Silva, A. C. H. et al. Phys. Chem. Chem. Phys., submitted.

Novel, low-voltage, high-detectivity, solution processed, flexible near-infrared (NIR) photodetectors for opto-electronic applications were realized and their opto-electronic properties were investigated for the first time. This was achieved by synthesizing Ag₂Se nanoparticles (NPs) in aqueous solutions, and depositing highly crystalline Ag₂Se thin films at 150 °C with re-distributed Ag₂Se NPs in aqueous inks. The high conductivity and low trap concentration of the 150 °C annealed Ag₂Se films result from the Ag formed inside the films and the improved film quality, respectively. These factors are both critical for the realization of high-performance flexible photodetectors. The fabricated device exhibited a high detectivity of 7.14×10⁹ Jones (above 1×10⁹) at room temperature, delivering low power consumption. This detectivity is superior to those of reported low bandgap semiconductor systems, although the device had undergone 0.38% compressive and tensile strains. Moreover, the performance of the device is better than that of MoS₂-based phototransistors, black arsenic phosphorus field-effect transistors (FETs) or commercial thermistor bolometers at room temperature (D* ~10⁸ Jones) and exposed to mid-infrared (MIR) light.

Photon upconversion is a process by which two or more low-energy photons are absorbed and one higher-energy photon is emitted by a material. Materials that can achieve photon upconversion are desirable for many applications such as optoelectronic devices, drug delivery, and photovoltaics. A key advantage of using semiconductor nanoparticles for photon upconversion is their wide tunability in structure, which consequently affects their absorption and emission properties. We have synthesized colloidal quantum dot (QD) heterostructures for this purpose, in which two QDs with different bandgaps are separated by a wide-bandgap nanorod. While our CdSe(Te)/CdS/CdSe core/rod/emitter structures demonstrate near-infrared (NIR)-to-visible photon upconversion, their upconversion efficiencies show significant room for improvement. The performance of these structures can be enhanced through a better understanding of their underlying properties and the effect of these properties on upconversion efficiency. While we can further our understanding through theoretical modeling, we must also consider the constraints of available synthesis techniques. Synthesis parameters such as ion precursors, organic ligands, and temperature, among others, can be tuned within well-studied procedures to create structures with the desired composition, structure, and resulting optical behavior. For example, introducing a bandgap gradient along the nanorod through doping has been found to funnel carriers to the emitter, increasing both quantum yield and upconversion efficiency. We have therefore studied an “inverted” emitter/rod/core structure which allows this gradient to form more naturally during synthesis. We have also found that charge carrier separation can be improved by controlling the position of the core along the rod and by optimizing the length of the rod. This has led us to explore spherical core/shell/shell structures with varying shell thicknesses and gradients and to explore other material compositions such as PbS to further tune absorption and emission wavelengths. We present these synthesis methods, the resulting structures, and their impact on upconversion performance as characterized by photoluminescence and power dependence measurements. We describe how further upconversion performance enhancement can be achieved through improved structural control, which will allow for future incorporation into various photonic device applications.

Mid-wavelength infrared (IR) photons have high penetration through airborne obscurants such as fog and mists, rendering mid-IR photodetectors ideal for use in long-distance thermal sensing and enhanced night vision applications. Colloidal quantum dot-based mid-IR photodetectors present a promising path toward fabricating sensors and imagers at significantly reduced cost compared to state-of-the-art epitaxial grown quantum-well IR detectors. Herein, we report on recently discovered silver chalcogenide quantum dots that exhibit distinct optical absorption in the mid-IR. These colloidal quantum dots demonstrate a narrow bandgap metastable tetragonal phase, not available in bulk, and contain excess electrons, allowing intraband optical transitions between the first and the second conduction energy levels. In this talk, I will present the synthesis, characterization, and photoconductive photodetector device characteristics of this relatively benign colloidal IR material, in contrast to highly toxic Hg-based competing mid-IR devices.

Quantum dots composed of lead chalcogenides (PbX, X = S, Se, Te) have shown considerable promise for use in various applications ranging from photovoltaics to optoelectronics, including commercial implementations in displays and cell phones. However, there has been limited use with these nanoparticles in electronic devices due to the lack of coupling between multiple quantum dots. Consequently, there has been a significant amount of research focusing on the combination of self-assembly and directed attachment of colloidal quantum dots at fluid interfaces to form two-dimensional quantum dot superlattices. Currently, the mechanism for the formation of two-dimensional quantum dot superstructures is not fully understood and, not unrelatedly, materials that are produced experimentally contain defects that hinder the performance of the material. Various factors, such as inhomogeneous coupling, variation in quantum dot size and shape, and positional misalignment, create structural disorder that is detrimental to the overall charge transport in the chalcogenide material.

We are using ab initio Density Functional Theory (DFT) and Molecular Dynamics (MD) to study the atomic behavior behind the oriented attachment of multiple colloidal lead sulfide quantum dots. These methods serve as an important tool for elucidating the thermodynamic and kinetic factors that govern the interactions between the organic ligands bound to the surfaces of the quantum dots and the amine-based chemical trigger used to remove them from the surface and the epitaxial bridge formation between multiple colloidal quantum dots. We have studied reactions between the ligands and the chemical trigger in solution using the PBE functional in DFT.

While the DFT calculations provide an accurate picture of the interparticle interactions, we need to turn to a semi-empirical approach that will allow us to access orders-of-magnitude-larger system sizes. To do so, and to accurately model a system with multiple lead sulfide quantum dots, we use the Simple Molecular Reactive Force Field (SMRFF), which we have developed to model the reactive processes that give rise to PbS bond formation. The SMRFF potential is drawn from a combination of simple potentials (Morse/Tersoff, Lennard-Jones, and Coulombic). The parameters for each individual potential are obtained using a training set with multiple 32-atom geometries. For each geometry, the energy of the system is calculated using DFT and the SMRFF parameters are optimized to match the energy as closely as possible. With this fully parameterized SMRFF potential, we are then able to provide the first reactive force field simulation of epitaxial bridge formation between multiple PbS quantum dots. The ability to watch this process allows us to understand how misalignment and processing conditions, such as thermal annealing, affect the nature of the bridges including defect formation.

Epitaxially-fused colloidal quantum dot (QD) superlattices promise to combine the unique photophysics of QDs with the efficient band transport of bulk-semiconductors. PbS and PbSe QD layers have demonstrated multiple exciton generation in response to photo excitation by high energy photons. We would like to extend the multi-exciton generation to thick arrays of QDs, which is possible if the structural order of the film is improved to reduce electronic traps. Here we focus on fabrication of extended superlattice films with multi-layers of QDs and examine how the fabrication method affects the local order. Specifically, we form ordered arrays of QDs at a liquid-liquid interface and then reduce the spacing between QDs using in-situ ligand exchange. The resulting films are characterized using high-resolution scanning transition electron tomography (STET). We are able to resolve the position of each QD in a sample encompassing 1000’s of QDs. In addition of the particle positions, we also resolve the presence (or lack of) of epitaxial connections between neighboring QDs and the thickness of the neck formed between neighboring QDs. This morphological information is used to generate an electronic model. Our model shows that the energetic disorder caused by heterogeneous epitaxial bridging between QDs in a single domains is about one order of magnitude less important than the presence of disorder at grain boundaries between superlattice domains. This presentation will explore the link between structural and electronic order in these complex 3D samples.