Device Length Research Articles

Does the shape of the copper intrauterine device play a role in expulsion? Results from the ongoing European Active Surveillance Study on LCS12

ObjectiveTo compare the incidence of expulsion for different copper intrauterine device (IUD) shapes. Study designWe conducted a secondary analysis of the ongoing, prospective, non-interventional European Active Surveillance Study on LCS12 (EURAS-LCS12). Users of newly inserted IUDs were recruited in 10 European countries via a network of approximately 1200 clinicians. We restricted the analysis to copper IUD users. In the main analysis, we classified copper IUDs by shape [Nova-T frame, Tatum-T frame, Multiload frame, frameless IUDs and intrauterine balls], without differentiation of size. We calculated the cumulative incidence, crude, and adjusted hazard ratios for expulsion. Covariates included in the adjusted analyses were age, BMI, parity, education, income, IUD user status, marital status, length of device, heavy menstrual bleeding, and clinician’s experience. ResultsWe included 26,381 copper IUD users from the EURAS-LCS12 dataset for this study. The most frequently used IUD shape was the Nova-T frame (14,724 [55.8%]) followed by the Tatum-T frame (4276 [16.2%]), frameless IUDs (3374 [12.8%]), Multiload frame (2962 [11.2%]), and intrauterine balls (IUBs) (1045 [4.0%]). Cox regression analysis regarding expulsions yielded an adjusted hazard ratio of 0.8 (95% CI, 0.7–1.0), 1.3 (95% CI, 1.0–1.8), 1.6 (95% CI, 1.2–2.1) and 3.6 (95% CI, 2.7–4.9) for Nova-T frame IUD, frameless IUDs, Multiload frame IUDs and IUBs versus Tatum-T frame IUD, respectively. ConclusionThe risk of expulsion following placement of a copper IUD is related to IUD shape, with Nova-T frame and Tatum-T frame IUDs demonstrating the lowest risk. ImplicationsOur finding of a higher risk of expulsion observed with Multiload frame, frameless, and intrauterine ball copper IUDs compared to Tatum-T frame and Nova-T frame devices during real world use has clinical importance. Clinicians may choose to use these data when counseling patients.

Ultralow 1/f noise in epigraphene devices

We report the lowest recorded levels of 1/f noise for graphene-based devices, at the level of SV/V2=SI/I2=4.4×10−16 (1/Hz), measured at f = 10 Hz (SV/V2=SI/I2 < 10−16 1/Hz for f > 100 Hz) in large-area epitaxial graphene on silicon carbide (epigraphene) Hall sensors. This performance is made possible through the combination of high material quality, low contact resistance achieved by edge contact fabrication process, homogeneous doping, and stable passivation of the graphene layer. Our study explores the nature of 1/f noise as a function of carrier density and device geometry and includes data from Hall sensors with device area range spanning over six orders of magnitude, with characteristic device length ranging from L = 1 μm to 1 mm. In optimized graphene Hall sensors, we demonstrate arrays to be a viable route to improve further the magnetic field detection: a simple parallel connection of two devices displays record-high magnetic field sensitivity at room temperature, with minimum detectable magnetic field levels down to Bmin = 9.5 nT/√Hz. The remarkable low levels of 1/f noise observed in epigraphene devices hold immense capacity for the design and fabrication of scalable epigraphene-based sensors with exceptional performance.

A high figure of merit of phonon-polariton waveguide modes with hbn/SiO2/graphene /hBN ribs waveguide in mid-infrared range

Natural hyperbolic materials can confine electromagnetic waves at the nanoscale. In this study, we propose a waveguide design that combines a high quality factor (FOM) with low loss, utilizing hexagonal boron nitride and graphene and gold substrate. The waveguide consists of a dielectric rib with a graphene layer sandwiched between two hBN ribs. Numerical simulations demonstrate the existence of two guided modes in the proposed waveguide within the second reststrahlen band (1360.0 cm−1<ω < 1609.8 cm−1) of hBN. These modes are formed by coupling the hyperbolic phonon polariton (HPhP) of two hBN rib in the middle dielectric rib and are subsequently modulated by a graphene layer. Interestingly, we observe variations in four transmission parameters, namely effective length, figure of merit, device length, and propagation loss of the guided modes, with respect to the operation frequency and gate voltage. By optimizing the waveguide's geometry parameters and dielectric permittivity, the modal properties were analyzed. Simulation results indicate that optimizing the waveguide size parameters enables us to achieve a high FOM of 4.0 × 107. The proposed waveguide design offers a promising approach for designing tunable mid infrared range waveguides on photonic chips, and this concept can be extended to other 2D materials and hyperbolic materials.

Tunable negative differential resistance behavior in the nitrogen-doped zigzag GeSe nanoribbon based single-gate field effect transistor

This article presents a theoretical study on the negative differential resistance (NDR) behavior of a bilaterally hydrogen-passivated Zigzag GeSe nanoribbon based single-gate field-effect transistor. It is focused on a 5-nm channel length device, and the study utilizes a simulation calculation that combines the density functional theory and non-equilibrium Green’s function. In this study, a nitrogen atom is introduced as a substitutional dopant to replace the Ge atom in the source and drain regions of the GeSe field-effect transistor to created special symmetry structure. The numerical results demonstrate that the current peak-to-valley ratio (PVR) of the device can be effectively controlled by applying a gate voltage and up to 105 with the current peak value of 0.3 nA at room temperature (300 K). The device configuration described in this study meets the requirements of real-world industrial applications. These findings highlight the potential of GeSe Zigzag nanoribbon for future electronic device applications at nanoscale.

Unraveling the crystallization kinetics of the Ge2Sb2Te5 phase change compound with a machine-learned interatomic potential

The phase change compound Ge2Sb2Te5 (GST225) is exploited in advanced non-volatile electronic memories and in neuromorphic devices which both rely on a fast and reversible transition between the crystalline and amorphous phases induced by Joule heating. The crystallization kinetics of GST225 is a key functional feature for the operation of these devices. We report here on the development of a machine-learned interatomic potential for GST225 that allowed us to perform large scale molecular dynamics simulations (over 10,000 atoms for over 100 ns) to uncover the details of the crystallization kinetics in a wide range of temperatures of interest for the programming of the devices. The potential is obtained by fitting with a deep neural network (NN) scheme a large quantum-mechanical database generated within density functional theory. The availability of a highly efficient and yet highly accurate NN potential opens the possibility to simulate phase change materials at the length and time scales of the real devices.

High‐Frequency fT and fmax in Organic Transistors: Performance and Perspective

AbstractIn state‐of‐the‐art organic transistors, the transit frequency fT is reported to be fT = 40 MHz for vertical organic transistors, where a voltage‐normalized fT corresponds to 0.36 MHz V−2. The reported highest fT for conventional organic transistors is fT = 160 MHz, where the voltage‐normalized fT is 0.1 MHz V−2. While the reported transit frequency fT of n‐MOSFETs for silicon technology when normalized by the applied biases exceeds 300 GHz V−2. The reported carrier mobility of organic semiconductors is over 100 cm2 V−1 s−1, which is almost an order of magnitude lower than that of silicon. However, the transit frequency fT of organic transistors lags far behind silicon technology by about six orders of magnitude at a comparable device length below 300 nm. In this work, a technology computer‐aided design (TCAD) simulator is adjusted to the experimental DC and AC characteristics of the 200 nm device length vertical organic permeable‐base transistors (OPBTs) based on C60 as semiconductor. The AC performance of the devices is investigated and quantitatively analyzed the influence and contribution of key design, structure and material parameters that determine fT and fmax. The feasible ways to enhance and optimize the highest achievable fT and fmax are identified to reach the and overcome the limitations to high‐frequency operation in organic transistors.

High-efficiency Ka-band frequency multiplier based on the nonlinear kinetic inductance in a superconducting microstrip

Local oscillator sources have become a major technological deficit approaching the terahertz frequencies. For applications, where a narrow linewidth is necessary, such as coherent remote sensing, the best solution to signal generation is frequency multiplication, where phase locking can be easily achieved using the fundamental source. We present an alternative to existing frequency multiplier technologies for cryogenic applications. The device is a superconducting waveguide that has kinetic inductance dependent on the rf current through the device. The nonlinear kinetic inductance is anomalous to a nonlinear optical medium, where the Kerr effect causes 3- or 4-wave mixing, as demonstrated in microwave parametric amplifiers utilizing the same nonlinear effect. In 4-wave mixing, three photons at the fundamental frequency are mixed to generate a single photon at the third harmonic. Dispersion engineering is used to achieve wideband phase matching and phase out harmonics higher than the third. The device length is designed to be the point where most photons are up-converted to the third harmonic. Simulations of these devices lead to expected efficiencies as high as 90%. We report results from a NbTiN frequency multiplier from 34.2 to 104 GHz with efficiency better than 50%.

Enhanced electron transport and optical properties of experimentally synthesized monolayer Si9C15: a comprehensive DFT study for nanoelectronics and photocatalytic applications.

Two-dimensional silicon-carbide (SixCy) materials stand out for their compatibility with current silicon-based technologies, offering unique advantages in nanoelectronics and photocatalysis. In this study, we employ density functional theory and nonequilibrium Green's function methods to investigate the electronic properties, electron transport characteristics, and optoelectronic qualities of experimentally synthesized monolayer Si9C15. Utilizing the modified deformation potential theory formula, we unveil Si9C15's significant directional anisotropy in electron mobility (706.42 cm2 V-1 s-1) compared to holes (432.84 cm2 V-1 s-1) in the a direction. The electrical transport calculations reveal that configurations with a 3 nm channel length demonstrate an ON state when biased, reaching a peak current of 150 nA. Moreover, this maximum current value escalates to 200 nA under tensile strain, marking an increase of approximately 100 times compared to the 5 nm channel, which remains in an OFF state. Si9C15 exhibits high light absorption coefficients (∼105 cm-1) and suitable band edge positions for water splitting at pH 0-7. Applying 1-5% tensile strain can tune the conduction band minimum and valence band maximum closer to the standard redox potentials, enhancing photocatalytic water splitting efficiency. Remarkably, under illumination at pH 0 and 7, Si9C15 can spontaneously catalyze water splitting, demonstrating its potential as a highly efficient photocatalyst. Our findings emphasize the importance of strain control and device length optimization for performance enhancement in nanoelectronics and renewable energy applications, positioning Si9C15 as a promising material for high-performance field-effect transistors and photocatalytic water splitting.

Handheld Biosensor System Based on a Gradient Grating Period Guided-Mode Resonance Device.

Handheld biosensors have attracted substantial attention for numerous applications, including disease diagnosis, drug dosage monitoring, and environmental sensing. This study presents a novel handheld biosensor based on a gradient grating period guided-mode resonance (GGP-GMR) sensor. Unlike conventional GMR sensors, the proposed sensor's grating period varies along the device length; hence, the resonant wavelength varies linearly along the device length. If a GGP-GMR sensor is illuminated with a narrow band of light at normal incidence, the light resonates and reflects at a specific period but transmits at other periods; this can be observed as a dark band by using a complementary metal oxide semiconductor (CMOS) underneath the sensor. The concentration of a target analyte can be determined by monitoring the shift of this dark band. We designed and fabricated a handheld device incorporating a light-emitting diode (LED) light source, the necessary optics, an optofluidic chip with an embedded GGP-GMR sensor, and a CMOS. LEDs with different beam angles and bandpass filters with different full width at half maximum values were investigated to optimize the dark band quality and improve the accuracy of the subsequent image analysis. Substrate materials with different refractive indices and waveguide thicknesses were also investigated to maximize the GGP-GMR sensor's figure of merit. Experiments were performed to validate the proposed handheld biosensor, which achieved a limit of detection (LOD) of 1.09 × 10-3 RIU for bulk solution measurement. The sensor's performance in the multiplexed detection of albumin and creatinine solutions at concentrations of 0-500 μg/mL and 0-10 mg/mL, respectively, was investigated; the corresponding LODs were 0.66 and 0.61 μg/mL.

Bubble and Dissolved Gas Distribution in Flow through Membraneless Water Electrolysis

Membraneless water electrolysis devices offer an alternative method for hydrogen production without the use of a membrane separator used in traditional alkaline (porous diaphragm) or PEM water electrolysis (polymer electrolyte membrane). This could offer lower capital cost devices (only requiring two electrodes and a cell body) with simple and scalable manufacturing. It could also result in lower operating costs if highly conductive liquid electrolytes are used. Membraneless devices come in a variety of forms including parallel plate, diverging flow through and buoyancy driven devices to name a few. However, the key concept between them is using fluid forces in the bulk electrolyte (controlled by device or porous electrode structure) to remove bubbles from the electrode surfaces and attempt to reduce cross-over of dissolved hydrogen and oxygen, as well as their two-phase flow [1]. Since there is no separator or membrane to restrict transport between the anode and cathode the fluid convection with laminar flows must be used to guide the products to the separate outlets with minimal cross-over and to minimise bubble accumulation on the electrode surfaces. Bubbles on the electrodes increase the activation overpotential by reducing the active electrochemical surface area and decrease the effective conductivity of the flowing electrolyte by inducing tortuous pathways in the electrode gap. In this study, we look at the single-phase and two-phase flow response of flow through membraneless electrolyser by coupling secondary current distribution (which includes the effects of ionic and activation overpotentials) with the volume of fluid method in OpenFOAM. The effects of increasing the length of the device and the electrolyte flow rate on the product cross-over and the dynamic potential of the devices are evaluated to give guidelines for process scale up from 1 to 10 cm. Furthermore, the effect of device porous electrode microstructure on the current density distribution and bubble accumulation is evaluated using custom microstructure reconstruction algorithms built into PMG 1.6 [2]. It was found that there is a balance in flow rate required for uniform flow over the electrodes which depends on the device length and the porous electrode properties. Ordered, graded or unstructured electrodes have different electrochemical performance, and this study highlights some electrode designs that could improve if future manufacturing efforts are increased. Depending on the current density distribution, dissolved gas accumulation and the wettability conditions of the porous electrodes, bubbles may emerge in the electrode gap.Figure 1 - Effect of the hydrogen and oxygen bubble accumulation (yellow) on the ionic current density distribution in a flow through membraneless water electrolyser

Electrodeless method for ultra-low mobility with carrier-resolution of nanochannel

As the channel lengths of electronic devices are scaled down to the nanometer range, the conventional methods to evaluate charge-carrier mobility approach a technical limit that is imposed by interfering effects of the electrode and forcing field. In this study, we demonstrate that electron spectroscopy provides additional (yet hidden) information on unipolar charge transport, which is free from conventional problems. We demonstrate that the estimated effective diffusion current through the target sample allows the measurement that is precise enough (10−4 cm2/V s) to obtain the mobility of electrons μelectron and holes μhole in nanolength organic channels. Using this method, we show how μelectron and μhole are correlated with the local structural order of poly(3-hexylthiophene) at the nanoscale. This method enables in situ charge-resolved observations of μelectron and μhole by eliminating the need for electrode and forcing field and will help to expand our understanding of charge conduction in nanoscale materials.

Foundry-Processed Compact and Broadband Adiabatic Optical Power Splitters with Strong Fabrication Tolerance

Optical power splitters play a crucial role as the fundamental building blocks for many integrated optical devices. They should have low losses, a broad bandwidth, and a high tolerance to fabrication errors. Adiabatic optical power splitters inherently possess these qualities while being compatible with foundry processes well suited for mass production. The long device lengths of adiabatic power splitters, however, are a limiting factor to achieve compact device sizes, which must be reduced. Here, we used a polynomial taper profile optimization algorithm to design 1 × 2 and 2 × 2 adiabatic power splitters with significantly shorter lengths than their adiabatic counterparts. The best-performing 1 × 2 and 2 × 2 power splitters had 20 μm and 16 μm coupling lengths, respectively. Our designs had minimum feature sizes ranging from 140 nm to 200 nm, and our measurements averaged across nine different chips showed excellent consistency in performance for devices with 180 nm and 200 nm minimum features. Both the 1 × 2 and 2 × 2 adiabatic optical power splitters had excess losses less than 0.7 dB over a 100 nm bandwidth, with a standard deviation lower than 0.3 dB. Furthermore, our measurements showed splitting ratios within 50 ± 3% over a 130 nm bandwidth. We also demonstrated the design of 1 × 2 power splitters with arbitrary splitting ratios, where splitting ratios ranging from 50:50 to 94:6 were achieved with standard deviations between 2% and 6%.

Enteroatmospheric fistulas in open abdomen in trauma associated with abdominal reintervention and VAC therapy.

Open abdomen is an alternative for the management of trauma patient, and negative pressure therapy of the wound using VAC® device is a genuine sort of treatment. Although the device poses technical advantages, risks are present and a critical complication is the enteroatmospheric fistula formation (EAF). To analize the role of negative pressure therapy length and VAC® device number of changes on the EAF formation in trauma patients udergoing open abdomen. Cut-off points were calculated using receiver operational characteristics curve. Values were compared with Student's t or Mann-Withney U tests, considering statistically significant p < 0.05. EAF were present in 39 cases (5.9%). A significant difference in the presence of EAF was present in patients with negative pressure therapy of the wound length ≥ 11.2 days (46.9 vs. 1.3%; relative risk [RR]: 3.67; 95% confidence interval [95% IC]: 2.4-6.68; p = 0.017) and when ≥ 2.6 VAC® device changes were performed (34.6 vs. 0.5%; RR: 6.92; 95% IC: 1.1-4.3; p < 0.001). At our institution, the practice of >3 VAC® device changes and length of therapy > 11 days should be carefully considered leading to reduce the risk of EAF formation.

Transcarotid thoracic endovascular aortic repair for rupture of the aortic isthmus due to high energy blunt trauma: A case report

Introduction: Blunt thoracic aortic injury (BTAI) is associated with very high mortality, and managing currently favors thoracic endovascular aneurysm repair (TEVAR) if feasible, especially BTAI due to high energy force, where patients usually suffer from polytrauma, which requires rapid discover and treat spontaneously. However, a remarkable group of patients is contraindicated for the transfemoral route- the established first-line approach in endovascular interventions of the aorta. Here, we report a successful case in which the carotid artery was used to perform TEVAR. Case presentation: A young female was transferred to our emergency department after injuries sustained by falling from the fifth floor. Her vital signs were stable on admission. Pan CT scan revealed multiple injuries: A grade III injury of the aortic isthmus, a grade III liver injury without active bleeding, and fracture of the sternum, multiple ribs, T12-L1 vertebral body, and pelvic ring. However, her hemodynamics became unstable, and we decided to perform an urgent endovascular repair. During preparations for TEVAR, we found that the diameter at the proximal landing zone was 17.1mm, so it cannot be accommodated by any of the available thoracic endograft devices. Under the emergency condition, we decide to perform TEVAR by abdominal aortic device and access via the carotid to suit the device's length. After inpatient rehabilitation, she was discharged without neurologic sequelae. Conclusion: Transcarotid TEVAR is feasible and could be considered as an alternative approach in some specific patients with BTAI.

Supercontinuum generation by saturated second-order nonlinear interactions

We propose a new approach to supercontinuum generation and carrier-envelope-offset detection based on saturated second-order nonlinear interactions in dispersion-engineered nanowaveguides. The technique developed here broadens the interacting harmonics by forming stable bifurcations of the pulse envelopes due to an interplay between phase-mismatch and pump depletion. We first present an intuitive heuristic model for spectral broadening by second-harmonic generation of femtosecond pulses and show that this model agrees well with experiments. Then, having established strong agreement between theory and experiment, we develop scaling laws that determine the energy required to generate an octave of bandwidth as a function of input pulse duration, device length, and input pulse chirp. These scaling laws suggest that future realization based on this approach could enable supercontinuum generation with orders of magnitude less energy than current state-of-the-art devices.

Inverse design of an ultra-compact and large-bandwidth bent subwavelength grating wavelength demultiplexer.

Inverse design has attracted significant attention as a method to improve device performance and compactness. In this research, we employed a combination of forward design and the inverse algorithm using particle swarm optimization (PSO) to design a bent ultra-compact 1310/1550nm broadband wavelength demultiplexer assisted by a subwavelength grating (SWG). Through the phase matching at 1550nm and the phase mismatch at 1310nm, we rapidly designed the width parameters of SWG in the forward direction. Then the PSO algorithm was used to optimize the SWG parameters in a certain range to achieve the best performance. Additionally, we introduced a new bent dimension significantly reducing the device length while maintaining low insertion loss (IL) and high extinction ratios (ERs). It has been verified that the length of the device is only 7.8µm, and it provides a high ER of 24dB at 1310nm and 27dB at 1550nm. The transmitted spectrum shows that the IL values at both wavelengths are below 0.1dB. Meanwhile, the 1dB bandwidth exceeds 150nm, effectively covering the entire O-band and C-band. This approach has been proven successful in enhancing performance and significantly reducing the device footprint.

Transcatheter closure of tubular PDA with amplatzer plug 4 in preterm infants weighing between 900 and 3,400 g: the pros and cons.

To our knowledge, no prior study has focused on the outcome of PDA occlusion using an Amplatzer™ Vascular Plug 4 (AP4) in ill preterm infants. This study aims to highlight the pros and cons of AP4 in this cohort. Between 2020 and 2022, 26 ill preterm infants underwent PDA closure in our centre. The median age, weight, and gestational age were 60 days (11-180 days), 1,900g (900-3,400 g), and 25 weeks (22-33 weeks), respectively. The indication of the intervention was hemodynamically significant PDA. A medical trial with non-steroid medication failed to close the ducts in all patients. Follow-up using echocardiography was done 24, 48, and 72 h after the intervention. Of 26 ducts, 21 were successfully closed with AP4. Five ducts shorter than 7 mm were unsuitable for AP4 and were closed with the Amplatzer Piccolo device. The median radiation time was 4 min (3-9 min). No early plug-related complications or deaths were documented. Plug-related jailing of the left pulmonary artery as a late complication was 9.5%, and LPA reintervention was required. All ducts were closed after 48 h. Implantation of the AP4 using a 4 F 0.38 guide wire-compatible catheter without inserting a long sheath makes the closure of tubular ducts with this device feasible and uncomplicated with a short intervention time. However, the limited sizes with fixed lengths of the AP4 make it unsuitable for ducts wider than 4.5 mm and shorter than the chosen device length, which can increase the risk of significant left pulmonary stenosis. A wide range of plug diameters and lengths is required to accommodate the large and short ducts.

Sensing characteristics of structural microfiber long-period gratings.

We present a detailed investigation into the sensing characteristics of a structural microfiber long-period grating (mLPG) sensor. By spirally winding a thinner microfiber to another thicker microfiber, periodic refractive index modulation is formed while the optical signal transmitted in the thicker microfiber is resonantly coupled out to the thinner microfiber, and then a 5-period four-port mLPG can be obtained with a device length of only ∼570 µm demonstrated a strong resonant dip of 25 dB. We studied the sensitivity characteristics of the four-port mLPG with surrounding strain, force, temperature and refractive index, and the obtained sensitivities were -6.4 pm/µɛ, -8418.6 nm/N, 7.62 pm/°C and 2122 nm/RIU, respectively. With the advantages of high refractive index sensitivity and wide wavelength tunable range, the four-port mLPG has great potential in applications such as tunable filters and biochemical sensor.

Virtual implantation of Derivo Embolization Devices using PreSize Neurovascular software: Accuracy evaluation and size selection comparison study.

Evaluating the deployed length of flow diverting stents (FDs) to select the optimal device size remains a challenging, yet crucial, task in aneurysm treatment. This study reports on the accuracy of PreSize Neurovascular (Oxford Heartbeat Ltd), a visualization and simulation software for FD intervention planning, in predicting FD deployed length, and on its impact on device size selection. Imaging data from consecutive patients treated with Derivo Embolization Device (Acandis GmbH) were collected from University Medical Center Hamburg-Eppendorf and retrospectively analyzed. Accuracy evaluation: prediction accuracy was calculated by comparing deployed FD lengths measured from imaging data and simulated by PreSize. Size selection comparison: two Interventional Neuroradiologists (INR1, INR2), blinded to the devices deployed, used PreSize to select the optimal device size (diameter and length). Their choices were compared against the deployed devices selected by conventional planning. Among 98 implanted devices, PreSize predicted deployed FD length with a mean accuracy of 94.54% (95% confidence interval [93.72%, 95.35%]). Among 98 aneurysm cases, PreSize-informed device lengths were significantly shorter (Wilcoxon signed-rank test, INR1: W = 394, P < .001, INR2: W = 305, P < .001) by 4.13 and 4.18 mm on average, and up to 20 and 25 mm, for INR1 and INR2, respectively, than the conventionally selected FDs. In 32% of cases, PreSize-informed devices resulted in fewer vessel bends covered by the FD while achieving sufficient aneurysm coverage. PreSize retrospectively predicted deployed FD lengths with high accuracy. Moreover, INRs in this study were more inclined to select shorter stent length in the simulation than they would have done conventionally.

Physical insight of thin AlGaN back barrier for millimeter-wave high voltage AlN/GaN on SiC HEMTs

In this work, physical mechanisms underlying carbon-doped buffer combined with an AlGaN back-barrier layer are investigated in state-of-the-art millimeter-wave AlN/GaN transistors. We have fabricated devices with and without the insertion of a thin AlGaN back-barrier layer with reduced carbon concentration to analyze the improvement resulting from this buffer architecture. More specifically, the impact of the Al mole fraction into the back-barrier, carbon doping in the buffer, and channel thickness on 100 nm gate length device performance has been studied. It appears that a 150 nm undoped GaN channel followed by a highly carbon-doped GaN buffer results in good electron confinement at the expense of a high current collapse. On the other hand, an Al mole fraction of 25% in the AlGaN back barrier layer coupled with a 150 nm undoped GaN channel provides excellent electron confinement, resulting not only in a low DIBL under high electric field but also low current collapse. Calibrated on experimental devices, TCAD simulations reveal that the electric field penetration inside the GaN buffer is prevented owing to a strong polarization from the back barrier when the Al-content is high enough. That is why, the electron confinement is superior for the 25% Al mole fraction in the back barrier along with reduced current collapse. As a result, careful engineering of the carbon concentration together with the undoped GaN channel thickness is crucial to achieve robust devices, which can, thus, deliver high device performance with superior voltage operation while using short gate lengths.