Large Defect Density Research Articles

(Invited) Depth Analysis of Near Valence Band Maximum Defect States in Amorphous Oxide Semiconductors: Amorphous in-Ga-Zn-O

Since the thin film transistors (TFTs) using amorphous oxide semiconductor (AOS) was firstly reported in 2004[1], amorphous In–Ga–Zn–O (a-IGZO) have been studied intensively because they have large advantageous features such as large electron mobilities (larger than 10 cm2/Vs), small subthreshold voltage swing, and low temperature fabrication process [2,3]. A-IGZO with the nominal chemical composition of In:Ga:Zn=1:1:1 is now commercialized for liquid crystal displays (LCDs) for smart phone or high-resolution TV and large-size organic light-emitting diode displays, etc.. Although the a-IGZO TFTs are already commercialized, stability is still a critical issue if we use it for next generation display such as transparent or flexible. In particular, it is recognized that the negative shift of the threshold voltage (V th) is associated with the excitation of subgap photons from deep-energy defects just above the valence band maximum (called near-VBM defects), which was firstly found by hard X-ray photoemission spectroscopy (HAXPES) [4]. We recently reported that a-IGZO deposited at room temperature has large film-density-distribution in the depth direction [5]. In this study, we measured depth profile of near-VBM states including IGZO/substrate interface. It is important to understand the depth distribution also of defect density, because mobile electrons in field-effect device are induced at vicinity of substrate in case of bottom gate TFT structure. Although depth analysis was already reported for a-IGZO deposited by pulsed laser deposition (PLD), we report for a-IGZO deposited by sputtering method which is standardly used in mass-production process. As a result, we found that the depth profile is different between PLD and sputtering method. We used radio-frequency (RF) magnetron sputtering to deposit a-IGZO films on Si wafer with 150 nm SiO2 at room temperature. RF power, oxygen flow rate ratio (Ar:O2), deposition pressure were fixed at 70 W, 97:3 and 0.55 Pa, respectively. HAXPES measurements were carried out at BL15XU undulator beamline (hv = 6 keV) in SPring-8. Film thickness dependences were observed for 5 – 100 nm a-IGZO films. For extracting buried spectra, we used total reflection X-ray to modify the penetration depth [6]. From thickness dependence of VB spectra, we found that the near-VBM defect density increase with decreasing film thickness. It may suggest that a-IGZO at the vicinity of substrate has large defect density compared to bulk region. To separate the effect of surface and buried region, we extracted the buried spectra using total reflection of X-ray in HAXPES measurement and difference method. We successfully observed the depth profile in wide range from 0 to 100 nm and found that the defect density at the vicinity of substrate is several times higher than that at thick enough region. We will discuss more details including the relationship between the microstructure and near-VBM defects at the conference.

Effects of high-energy C ions irradiation on the deuterium retention behavior in V-5Cr-5Ti

Vanadium alloys are attractive candidate structural materials in future fusion reactors. To evaluate the effects of irradiation on the deuterium retention behavior in vanadium alloys, samples made of V-5Cr-5Ti were irradiated by 5.5 MeV carbon ions. Doppler broadening measurements of the positron annihilation radiation tests were carried out to investigate the defect properties in the irradiated samples. Then the irradiated and virgin samples were implanted with deuterium in an ECR (electron-cyclotron resonance) linear plasma device followed by thermal desorption spectroscopy experiments. It was found that the irradiation process introduced large density of vacancy-type defects and subsequently increased the deuterium retention in the V-5Cr-5Ti, which remains as a serious concern for the vanadium alloys in the application as the structural materials of the fusion blanket.

(Invited) AlGaN Nanowire Deep Ultraviolet Photonics

AlGaN-based deep ultraviolet (UV) light emitting diodes (LEDs) and lasers are important for a broad range of applications, including water purification, disinfection, surface treatment, and chemical and biochemical sensing. To date, the performance of AlGaN quantum well UV LEDs and laser diodes has been limited by the presence of large densities of defects and dislocations, large polarization field, and extremely inefficient p-type doping. In this context, we have developed nearly defect-free AlGaN nanowire heterostructures. We have achieved highly efficient p-type conduction of Mg-doped AlN and Al-rich AlGaN nanowires. With the use of AlGaN nanowires, we have further demonstrated, for the first time, electrically pumped semiconductor lasers operating in the UV-B and UV-C bands. The AlGaN nanowires are grown on Si substrate under N-rich conditions by plasma-assisted molecular beam epitaxy. Detailed structural characterization suggests the presence of strong compositional modulation. The Ga-rich AlGaN clusters have dimensions of 0.25 to 2 nm along the growth direction and of 2 to 10 nm along the lateral direction. The localized variations in the Ga-concentration are estimated to be ~10%, thereby providing strong quantum-confinement of charge carriers. By changing alloy composition, the emission can be widely tuned in the UV-C band from 207 nm to 280 nm. We have demonstrated AlN nanowire LEDs with a small turn on voltage of ~6 V, which is significantly smaller than previously reported AlN epilayers LEDs (> 20 V). This excellent electrical performance is largely ascribed to the significantly enhanced Mg-dopant incorporation in nanowire structures and the resulting efficient impurity band conduction. For ternary AlGaN nanowire UV LEDs, the device structure consists of n+-GaN/Al/p+-AlGaN tunnel junction, which can remove the resistive p-contact layer and improve hole transport. Compared to devices without tunnel junction, the output power is improved by 2 orders of magnitude. For the device operating at 274 nm, the maximum output power is > 20 mW. It is further envisioned that, with the use of AlGaN nanowires, UV-C LEDs with efficiency >70% can be potentially achieved. By further exploiting Anderson localization of light in such nearly defect-free AlGaN nanowires, we have demonstrated, for the first time, electrically injected lasers operating in the UV-B and UV-C bands. For lasers operating at 239 nm, the threshold current is ~0.3 mA at room-temperature. These studies provide a viable path to achieve high efficiency semiconductor laser diodes operating in the UV-C band that was not previously possible.

Performance Enhancement of Perovskite Solar Cells Induced by Lead Acetate as an Additive

The crystallinity of organic–inorganic hybrid lead halide perovskite active layers is critical to the performance of perovskite solar cells (PSCs). However, it is a challenge to control the crystallization process of perovskite films by solution process. Here, the fabrication of high‐efficiency PSCs by employing lead acetate (PbAc2) as an additive in the perovskite precursor is demonstrated for the first time. By adding a few percent of PbAc2 (molar ratio with respect to PbI2) into the precursors, the perovskite crystallization process is significantly retarded, leading to pinhole‐free perovskite films with large grains and low defect densities. As result, the power conversion efficiencies and stability of the PSCs are substantially improved. This work provides a convenient and unique approach for preparing high‐quality perovskite films that can be used in photovoltaics as well as other optoelectronic devices.

Germanium on Silicon Avalanche Photodiode

Silicon photonics is considered as one of the promising technologies for high-speed optical fiber communications. Among various silicon photonic devices, germanium on silicon avalanche photodiodes (Ge/Si APDs) have attracted tremendous attention due to their the properties of high performance and low cost. The sensitivity of 10Gb/s APD reached −29.5dBm at 1550 nm with the bit error rate of 1 × 10−12 in 2015, and 25 Gb/s APD reached a highest sensitivity ever reported at about −23.5dBm at 1310 nm in 2016. Furthermore, a linear 28Gbaud APD receiver is found to have about 5dB more sensitivity than PIN solution for both multimode fiber (MMF) and single-mode fiber (SMF). Even with large defect density from lattice mismatch between Ge and Si substrate layer, Ge/Si APD devices demonstrate good reliability and pass all required qualification tests according to GR-468. In this paper, the structure, materials, process, performances, and reliability of APDs will be reviewed.

Testing of time-of-flight neutron diffraction imaging using a high-resolution scintillator detector with wavelength-shifting fibers

In this study, time-of-flight (TOF) neutron diffraction images of a β-Sn single crystal and a coarse-grained polycrystalline steel plate were obtained using a high-resolution wavelength-shifting fiber based scintillator detector. An inhomogeneous intensity distribution of the 200 reflection from the β-Sn single crystal indicated a large defect density near the surface of the crystal. In addition, the diffraction spots from the individual grains of the steel plate were successfully observed separately in both the normal and width directions. This imaging technique has the potential to be used for evaluating the distribution of material characteristics in different regions inside bulk materials.

PtAl truncated octahedron nanocrystals for improved formic acid electrooxidation.

We successfully obtained truncated octahedron PtAl alloy nanocrystals (∼10 nm) via a facile one-pot strategy. The PtAl nanocrystals exhibited enhanced activity in formic acid electrooxidation compared to commercial Pt/C, which could be ascribed to the large density of defects on the surface as well as the synergy between Pt and Al.

Efficient and stable CH3NH3PbI3-x(SCN)x planar perovskite solar cells fabricated in ambient air with low-temperature process

Planar perovskite solar cells (PSCs) based on CH3NH3PbI3-x(SCN)x (SCN: thiocyanate) active layer and low-temperature processed TiO2 films are fabricated by a sequential two-step method in ambient air. Here, alkali thiocyanates (NaSCN, KSCN) are added into Pb(SCN)2 precursor to improve the microstructure of CH3NH3PbI3-x(SCN)x perovskite layers and performance of the as-prepared PSCs. At the optimum concentrations of alkali thiocyanates as additives, the as-prepared NaSCN-modified and KSCN-modified PSCs demonstrate the efficiencies of 16.59% and 15.63% respectively, being much higher than 12.73% of the reference PSCs without additives. This improvement is primarily ascribed to the enhanced electron transport, reduced recombination rates and much improved microstructures with large grain size and low defect density at grain boundaries. Importantly, it is revealed that the modified PSCs at the optimized concentrations of alkali thiocyanates additives exhibit remarkably improved stability than the reference PSCs against humid circumstance, and a continuous exposure to humid air without encapsulation over 45 days only records about 5% degradation of the efficiency. These findings provide a facile approach to fabricate efficient and stable PSCs by low processing temperature in ambient air, both of which are highly preferred for future practical applications of PSCs.

Spontaneous exchange bias formation driven byastructural phase transition in the antiferromagnetic material.

Most of the magnetic devices in advanced electronics rely on the exchange bias effect, a magnetic interaction that couples a ferromagnetic and an antiferromagnetic material, resulting in a unidirectional displacement of the ferromagnetic hysteresis loop by an amount called the 'exchange bias field'. Setting and optimizing exchange bias involves cooling through the Néel temperature of the antiferromagnetic material in the presence of a magnetic field. Here we demonstrate an alternative process for the generation of exchange bias. In IrMn/FeCo bilayers, a structural phase transition in the IrMn layer develops at room temperature, exchange biasing the FeCo layer as it propagates. Once the process is completed, the IrMn layer contains very large single-crystal grains, with a large density of structural defects within each grain, which are promoted by the FeCo layer. The magnetic characterization indicates that these structural defects in the antiferromagnetic layer are behind the resulting large value of the exchange bias field and its good thermal stability. This mechanism for establishing the exchange bias in such a system can contribute towards the clarification of fundamental aspects of this exchange interaction.

Vacancy‐Controlled Contact Friction in Graphene

AbstractDefects‐controlled friction in graphene is of technological importance in many applications, but the underlying mechanism remains a subject of debate. Here it is shown that, during the controlled oxidation in oxygen plasma and subsequent reduction induced by high‐energy photons, the contact friction in chemical vapor deposition grown graphene is dominantly determined by the vacancies formed instead of the bonding with add‐atoms. This effect is attributed to the vacancy‐enhanced out‐of‐plane deformation flexibility in graphene, which tends to produce large puckering of graphene sheet near the contact edge and thus increases the effective contact area. Modified graphene with large contact friction has a large density of defects, but remains a good electrical conductor, in which the carrier transport is strongly affected by quantum localization effects even at room temperature. It is also found that the oxidation process in graphene is substrate‐sensitive. Comparing to monolayer graphene on SiO2 substrate, the oxidation process progresses much faster when the substrate is SrTiO3, while bilayer graphene exhibits great oxidation resistance on both substrates. The collection of observations provides important information for tailoring the mechanical, electrical, and chemical properties of graphene through selected defects and substrates.

Characterization of Interface State Density of Ni/p-GaN Structures by Capacitance/Conductance-Voltage-Frequency Measurements**Supported by the Natural Science Foundation of Jiangxi Province under Grant No 20133ACB20005, the Key Program of National Natural Science Foundation of China under Grant No 41330318, the Key Program of Science and Technology Research of Ministry of Education under Grant No NRE1515, the Foundation of Training Academic and Technical

For the frequency range of 1 kHz–10 MHz, the interface state density of Ni contacts on p-GaN is studied using capacitance-voltage (C–V) and conductance-frequency-voltage (G–f–V) measurements at room temperature. To obtain the real capacitance and interface state density of the Ni/p-GaN structures, the effects of the series resistance () on high-frequency (5 MHz) capacitance values measured at a reverse and a forward bias are investigated. The mean interface state densities obtained from the – capacitance and the conductance method are and , respectively. Furthermore, the interface state density derived from the conductance method is higher than that reported from the Ni/n-GaN in the literature, which is ascribed to a poor crystal quality and to a large defect density of the Mg-doped p-GaN.

Capillary‐Bridge Mediated Assembly of Conjugated Polymer Arrays toward Organic Photodetectors

Large‐scale patterning of high‐quality organic semiconductors is crucial for the fabrication of optoelectronic devices with high efficiency and low cost. Yet, owing to the uncontrollable dewetting dynamics of organic liquid in conventional solution patterning techniques, large defect density of organic architectures is inevitable, which is detrimental to the device performance. To address this challenge, herein a capillary‐bridge‐mediated assembly technique is developed for regulating the dewetting process, yielding large‐scale 1D microstructure ordered arrays. The 1D arrays organic photodetectors exhibit a high optoelectronic performance of light on/off ratio exceeding 100, responsivity of 3.24 A W−1, detectivity of 3.20 × 1011 Jones and fast response speed, showing a great improvement compared with spin‐coated membrane devices. In addition, the significant enhancement of the device photodetection under the electronic field modulation is investigated by applying a back‐gate voltage and explained with the photocurrent predominating in the OFF state and the neglected thermocurrent and tunneling current promoting in the ON state of the phototransistor devices. The research offers a new insight for the facile fabrication of large‐scale integrated photodetectors and other organic devices based on patterned conjugated polymers.

Frequency and Area Dependence of High-K/Ge MOS Capacitors

This paper investigates the capacitance-voltage (C-V) measurements of Ge/Al2O3/ZrO2/TiN MOS capacitors at different frequencies for square devices with different side lengths (10 μm, 20 μm, 30 μm, 40 μm, 50 μm, 100 μm). The results show that the series resistance corrected capacitance (CC) for accumulation region is dependent on frequency, devices substrate and area. CC increases at a faster pace when a lower frequency (1KHz) is used than at a higher frequency (1MHz). Substantial decrease of CC per unit area is reported in Ge/High-K in comparison to Si/High-K devices when device area is increased. This decrease is attributed to the large non-uniform interface defect density and high leakage current associated with Ge.

Energetic mapping of oxide traps in MoS2 field-effect transistors

The performance of MoS2 transistors is strongly affected by charge trapping in oxide traps with very broad distributions of time constants. These defects degrade the mobility and additionally lead to the hysteresis of the gate transfer characteristics, which presents a crucial performance and reliability issue for these new technologies. Here we perform a detailed study of the hysteresis in double-gated MoS2 FETs and show that this issue is nothing else than a combination of threshold voltage shifts resulting from positive and negative bias-temperature instabilities. While these instabilities are well known from silicon devices, they are even more important in 2D devices given the considerably larger defect densities. Most importantly, the magnitudes of these threshold voltage shifts depend strongly on the density and energetic alignment of the active oxide traps. Based on this, we introduce the incremental hysteresis sweep method which allows for an accurate mapping of these defects and extract their energy distributions from simulations. By applying our method to analyze the impact of oxide traps situated in the Al2O3 top gate of several devices, we confirm its versatility. Since all 2D devices investigated so far suffer from a similar hysteresis behavior, the incremental hysteresis sweep method provides a unique and powerful way for the detailed characterization of their defect bands.

Thermally Stable MAPbI3 Perovskite Solar Cells with Efficiency of 19.19% and Area over 1 cm2 achieved by Additive Engineering.

Solution-processed perovskite (PSC) solar cells have achieved extremely high power conversion efficiencies (PCEs) over 20%, but practical application of this photovoltaic technology requires further advancements on both long-term stability and large-area device demonstration. Here, an additive-engineering strategy is developed to realize a facile and convenient fabrication method of large-area uniform perovskite films composed of large crystal size and low density of defects. The high crystalline quality of the perovskite is found to simultaneously enhance the PCE and the durability of PSCs. By using the simple and widely used methylammonium lead iodide (MAPbI3 ), a certified PCE of 19.19% is achieved for devices with an aperture area of 1.025 cm2 , and the high-performing devices can sustain over 80% of the initial PCE after 500 h of thermal aging at 85 °C, which are among the best results of MAPbI3 -based PSCs so far.

Effect of gate voltage polarity on the ionic liquid gating behavior of NdNiO3/NdGaO3 heterostructures

The effect of gate voltage polarity on the behavior of NdNiO3 epitaxial thin films during ionic liquid gating is studied using in situ synchrotron X-ray techniques. We show that while negative biases have no discernible effect on the structure or composition of the films, large positive gate voltages result in the injection of a large concentration of oxygen vacancies (∼3%) and pronounced lattice expansion (0.17%) in addition to a 1000-fold increase in sheet resistance at room temperature. Despite the creation of large defect densities, the heterostructures exhibit a largely reversible switching behavior when sufficient time is provided for the vacancies to migrate in and out of the thin film surface. The results confirm that electrostatic gating takes place at negative gate voltages for p-type complex oxides while positive voltages favor the electrochemical reduction of Ni3+. Switching between positive and negative gate voltages therefore involves a combination of electronic and ionic doping processes th...

Biomass-derived mesopore-dominant porous carbons with large specific surface area and high defect density as high performance electrode materials for Li-ion batteries and supercapacitors

Porous carbons with high specific surface area and high defect density have been prepared through direct carbonization of cattle bones without any additional activators and templates. Benefiting from self-activation induced by hydroxyapatites within the cattle bones, the high-defect porous carbons obtained at 1100°C (PC-1100) possess the high specific surface area (2096m2g−1), largest mesopore volume (1.829cm3g−1), a narrow mesopore size distribution centered at approximately 4.0nm and good electrical conductivity (5141Sm−1). Due to the synergistic effect of the defects and pores, PC-1100 as the anode for Li-ion battery exhibits a high reversible capacity of 1488mAhg−1 after 250 cycles at 1Ag−1 and 661mAhg−1 after 1500 cycles at 10Ag−1. Even at 30Ag−1, PC-1100 can still deliver a high reversible capacity of 281mAhg−1, showing superior lithium storage capability. Moreover, the symmetric supercapacitor based on the PC-1100 in neat EMIM-BF4 electrolyte delivers a high energy density of 109.9Whkg−1 at a power density of 4.4kWkg−1, and maintains an energy density of 65.0Whkg−1 even at an ultrahigh power density of 81.5kWkg−1, as well as a superior cycling performance (96.4% of the capacitance retention after 5000 cycles).

Frequency and Area Dependence of High-K/Ge Mos Capacitors

– Ge substrate devices in Complementary Metal Oxide Semiconductors (CMOS) technology have been extensively studied due to its high electron and hole mobility [1]. However, most of the Ge devices have some limitation as making a high-quality Ge/High-k gate stack is rather difficult compared to Si/High-k system. The purpose of this work is to further study the capacitance measurements of a MOS capacitor at different frequencies for square devices with different side lengths (10 μm, 20 μm, 30 μm, 40 μm, 50 μm, 100 μm). The series resistance corrected capacitance (Cc) for accumulation region increases on a faster pace when a lower frequency (1KHz) is used than at a higher frequency (1MHz) (Figure 1(a)) as a function of side length, a. Accumulation capacitance has a strong dependence on frequency in metal oxide semiconductor devices. Our devices follow the same trend as previous studies have reported [2]. On the other hand, our study shows substantial decrease of Cc per unit area in Ge/High-K in comparison to High-K/Si devices (Figure 1(b)). This decrease is attributed to the large non-uniform interface defect density [3], high leakage current [3] associated with Ge stacks and fringing field effect [4]. Capacitance at accumulation condition depends on interface capacitance (Cit) which changes with defects density. Defect density increases proportionally with area and High-K/Ge devices, being leakier, experience a higher decrease of Cc per unit area than High-K/Si devices. The impact of fringing field capacitance (CL) on parallel plate capacitance (CP) decreases when device area is increased (Figure 1(c)) and this contributes to further decrease in total capacitance Cc per unit area. This work explains the frequency dispersion of high-k MOS capacitance in accumulation region and the unit capacitance decrease when Ge substrate is used. The study also models the fringing field capacitance [5] and its effect on MOS devices. Figure 1 (a) Corrected capacitance (C c) measured at -1.5 V (accumulation) at 1 MHz and 1 KHz shows frequency dependence; (b) Unit Capacitance is plotted against area for Ge substrate and Si substrate sample for measurements at frequency of 1MHz. C s is substrate capacitance and C ox is oxide capacitance; (c) CL/CP as a function of area. CL and CP are calculated as per Yuan & Trick’s model [5]. [1] D. Kuzum, A. J. Pethe, T. K., K. C. S, IEEE T. Electron. Dev., vol. 56, p. 648, 2009. [2] P. R. Lee, J. R. Ruzyllo, Electrochemical Society, vol. 9, p.353-362, 2007. [3] Y. M. Ding, D. Misra, J. Vac. Sci. Technol. B, vol. 34, p. 021203 2016. [4] S. Maurya, B. R. Singh, Solid States Physics, AIP Conf. Proc. 1512, 742-743, 2013. [5] E. Barke, IEEE Transactions on Computer Aided Design, vol. 7, No. 2, 1988. Figure 1

Side-Group-Induced Polymorphism in Self-Assembled Monolayers: 3,5-Bis(trifluoromethyl)benzenethiolate Films on Au(111).

The structure and molecular organization of self-assembled monolayers (SAMs) depend on a complex interplay of intermolecular and molecule-substrate interactions, so that even a small change in molecular composition can result in noticeable changes in the SAM structure. Herein we show that decoration of the most basic aromatic SAM constituent, benzenethiol, with two trifluoromethyl groups leads to distinct polymorphism in the respective SAMs, in which the appearance of a specific structural phase or a combination of several different phases is dependent on the parameters of the preparation procedure. High-quality films with a single crystallographic phase and significantly large domains could only be prepared after a short immersion time (5 min) and an additional re-immersion of the sample in pure ethanol at an elevated temperature. A standard 24 h immersion at room temperature led to poorly defined films with a large defect density and only a small portion of the surface covered by well-ordered molecular domains.

Simultaneous Graphite Exfoliation and N Doping in Supercritical Ammonia.

We report the exfoliation of graphite and simultaneous N doping of graphene by two methods: supercritical ammonia treatment and liquid-phase exfoliation with NH4OH. While the supercritical ammonia allowed N doping at a level of 6.4 atom % in 2 h, the liquid-phase exfoliation with NH4OH allowed N doping at a level of 2.7 atom % in 6 h. The N doped graphene obtained via the supercritical ammonia route had few layers (<5) and showed large lateral flake size (∼8 μm) and low defect density (ID/IG < 0.6) in spite of their high level of N doping. This work is the first demonstration of supercritical ammonia as an exfoliation agent and N doping precursor for graphene. Notably, the N doped graphene showed electrocatalytic activity toward oxygen reduction reaction with high durability and good methanol tolerance compared to those of commercial Pt/C catalyst.