Modeling and simulation of ZnO-based TFTs by dielectric engineering and temperature analysis for enhanced performance Authors: Pranaw Kumar, Syed Sadique Anwer Askari, Mukul Kumar Das Published: 2025 ABSTRACT AbstractThis work presents the modeling and simulation of thin-film transistors (TFTs) based on zinc oxide (ZnO). Strong agreement was found when the developed device model was validated against experimental data. Additionally, the analog performance of ZnO TFTs was examined to find out how the gate dielectric material, dielectric thickness, and operating temperature affect important electrical parameters such as drain current (ID), field-effect mobility (µFE), subthreshold slope (SS), threshold voltage (VTh), on/off current ratio (Ion/Ioff), drain conductance (gd), transconductance (gm), transconductance generation factor (TGF), and early voltage (VEA). With SS = 0.396 V/dec, VTh = 0.49 V, and µFE = 53.2 cm²/V·s, HfO₂ showed the best performance among the dielectrics under investigation. Improved characteristics were obtained by further optimization using a 120 nm HfO₂ layer: µFE = 54.2 cm²/V·s, VTh = 0.27 V, SS = 0.315 V/dec, and Ion/Ioff ≈ 10¹¹. According to temperature-dependent analysis, device performance deteriorates at higher temperatures (> 400 K) because of increased trap effects and Fermi-level shifts, which lower mobility and result in threshold voltage instability. The potential of ZnO-based TFTs for high-performance analog and optoelectronic applications is highlighted by their optimal performance at room temperature. These results, which enable the extraction of analog FoMs pertinent to circuit designers, are based on a device model calibrated to experimental data and a methodical sweep of dielectric material, dielectric thickness, and operating temperature.Keywords: Thin film transistor, Zinc oxide, High-k dielectric material, Gate oxide thickness, Thermal stabilitySubject terms: Engineering, Materials science, Nanoscience and technology, Physics Introduction Since thin film transistors are now the foundation of the thin film electronics industry, they have attracted a lot of attention from researchers in recent years. The TFT market is anticipated to grow at a CAGR of 17.34% between 2021 and 20261. LED or flat panel displays use TFT as a pixel switching element2–4. The production of large-area, high-resolution displays is the main goal of the display industry. As a result, TFT performance needs to be improved immediately. The TFT is a unique kind of transistor that has a supporting substrate on which a semiconductor, dielectric, and contact layer is deposited5. Advantages like uniformity, feasibility of low temperature processing and scalability makes metal oxide TFTs a better performer than TFTs using the hydrogenated amorphous silicon and poly silicon as channel layer6,7. Amorphous metal oxide has drawn remarkable attention of researchers for developing TFTs. Metal oxide-based semiconductor channel of thin film transistors (TFTs) has been studied for the last three decades8. Unique features like high mobility, superior material properties, ease fabrication, low process temperature make TFTs applicable for various useful applications9,10. Also, due to their wide band gap, these can be operated at high voltages11. Zinc oxide, due to its several noteworthy features used widely in applications like gas sensors, pyroelectric devices, transparent electrodes and surface acoustic devices12–14. The features like high exciton energy of 60meV, ferroelectric properties, wide direct band gap of 3.37 eV, n-type conductivity and piezoelectric effect make ZnO, very popular material for researchers as a semiconductor channel material15,16. Due to the advanced techniques gained in deposition methods, possibilities of growing ZnO thin films of excellent quality are very high. Though ZnO based TFTs have huge potential, there are few challenges too. First, the proper understanding of device physics is required for the TFTs based system. Second, extraction of threshold voltage in metal oxide TFTs is difficult due to disorder in the structure of semiconductor channel material. Third, most of the TFTs suffer from instability in threshold voltage. And lastly, in order to reduce the threshold voltage, suitable high-k dielectric material is needed. ZnO is very popular among those oxides, which at relatively low deposition temperature can be grown as a single crystalline material on substrates like plastics and glasses17. The wide band gap of ZnO makes possible for it to withstand visible light. The realization of TFTs with ZnO as an active layer is thus possible employing these key characteristics. Selecting good dielectric material in ZnO based TFTs has been a challenge for the researchers to analyze the performance parameters like threshold voltage, on-off current ratio, sub-threshold slope and carrier mobility. By using the atomic layer deposition (ALD) process, Carcia18 reported a mobility of 17.6 cm2/V-s for ZnO TFTs with varied gate insulators. A novel approach was investigated19, when TiO2 and Al2O3 gate insulators were superlatticed and deposited by ALD. SiOxNy gate insulators were deposited by RF sputtering on ZnO TFTs to lower the gate leakage current. It was noted that the carrier had a mobility of 70 cm2/V-s20. In ZnO-based TFTs, a researcher used Ta2O5 as the gate insulator and attained a high carrier mobility of 60.4 cm2/V-s21. This work reports a simulation study using Silvaco TCAD’s ATLAS tool22 to examine the suitability of ZnO as a channel layer in TFTs. Various analog parameters have been investigated in order to illustrate the use of ZnO-based TFT in analog applications. We also have reported the effect of different gate dielectric materials, gate dielectric thickness and operating temperature on the TFT device performance in this work. To the best of our knowledge, a unified TCAD analysis that concurrently varies dielectric material, dielectric thickness, and operating temperature for ZnO TFTs while evaluating analog FoMs (TGF, VEA) has not been systematically presented; prior works addressed subsets of these variables. Additionally, it has been observed that high density analog applications encounter difficulties in achieving improved device performance, such as low power operation. The TFTs can be successfully made to operate at low voltages with reduced subthreshold swing using a variety of techniques, such as changing the material, temperature variability, and geometric dimension scaling23–25. Many researchers view scaling the dimension of TFT as a potential approach to meet current device requirements. The dielectric thickness is likewise scaled by scaling. Downscaling has been shown to improve electrical parameters such as current ratio, field effect mobility and subthreshold slope, but at the expense of increasing tunneling gate leakage current, according to Kumar et al.26. In response, Vyas et al.27 proposed that high-k dielectric materials, such as Al2O3, HfO2, and others, could enhance TFTs’ electrical performance. High-k material has the same effect of scaling SiO2 without increasing the leakage current because it is physically thicker without being electrically thicker. The impact of gate dielectrics on oxide-TFT performance has been examined in previous experimental and simulation studies (e.g., S. Vyas et al.27, 2018; M. Labed et al.28, 2019; P. Kaur et al., 202229, taking reliability and high-κ integration into consideration. Nevertheless, there is a dearth of thorough research that concurrently investigates dielectric material, dielectric thickness, and operating temperature for ZnO-based TFTs, validates a calibrated trap/DOS model to experiment, and reports analog figures-of-merit like transconductance generation factor (TGF) and early voltage (VEA) across these conditions. By offering (i) a calibrated TCAD model verified by experiment, (ii) a methodical material/thickness/temperature sweep, (iii) the extraction of analog FoMs pertinent to circuit designers, and (iv) useful dielectric-thickness guidelines for optimal analog performance, this work addresses that gap. Temperature analysis is also necessary for many TFT applications, such as active-matrix liquid crystal displays (AMLCD), active-matrix organic light emitting diodes (AMOLED), biomedical devices, etc., since it indicates the device’s operating temperature range. The thermal factor kT/q sets a limit on the MOSFET’s switching speed, or SS value. The MOSFET’s minimum SS value is 60 mV/decade in an ideal scenario, but in a real-world scenario, it might be even lower. A thorough investigation is conducted on the TFT device in order to get the impact of thermal analysis. A detailed 2D simulation of a ZnO-based TFT and its analog analysis are carried out by adjusting various physical parameters while taking into account all of the aforementioned factors. It is evident that TFT based on ZnO material is a viable choice for upcoming analog devices. This work has made the following unique contributions: Gate dielectric material optimization: We have demonstrated the optimization of dielectric material at fixed thickness of 160 nm. The superior performer was HfO2. Gate dielectric material optimization: We have demonstrated the optimization of dielectric material at fixed thickness of 160 nm. The superior performer was HfO2. Dielectric thickness Impact: We have established that increase in dielectric thickness cause deterioration of TFT performance. On the other hand, too low thickness of dielectric material is also not acceptable. An optimized thickness of 120 nm of HfO2 dielectric material is reported. Dielectric thickness Impact: We have established that increase in dielectric thickness cause deterioration of TFT performance. On the other hand, too low thickness of dielectric material is also not acceptable. An optimized thickness of 120 nm of HfO2 dielectric material is reported. Thermal Impact: We have studied the impact of operating temperature on the performance parameter of TFT. It was shown that TFT shows superior performance at room temperature (300 K). Thermal Impact: We have studied the impact of operating temperature on the performance parameter of TFT. It was shown that TFT shows superior performance at room temperature (300 K). Validated Model: ZnO TFTs’ calibrated TCAD model was verified by our experimental transfer and output characteristics. Validated Model: ZnO TFTs’ calibrated TCAD model was verified by our experimental transfer and output characteristics. Thus, for the first time, simultaneous sweep of the operating temperature (300–500 K), dielectric thickness, and gate-dielectric material was done to study their impact on the performance parameter of TFTs. Thus, for the first time, simultaneous sweep of the operating temperature (300–500 K), dielectric thickness, and gate-dielectric material was done to study their impact on the performance parameter of TFTs. Analog figures-of-merit (gm, gd, TGF, and early voltage) are quantitatively extracted across the sweep to inform circuit-level implications. Analog figures-of-merit (gm, gd, TGF, and early voltage) are quantitatively extracted across the sweep to inform circuit-level implications. To explain deterioration in mobility/SS and other parameters, is calculated and its analysis is incorporated. To explain deterioration in mobility/SS and other parameters, is calculated and its analysis is incorporated. Design suggestions for high-performance analog and optoelectronic ZnO TFTs include dielectric selection and thickness. Design suggestions for high-performance analog and optoelectronic ZnO TFTs include dielectric selection and thickness. The following is how the work is organized: The TFT modeling and simulation approach, including dimensions, materials, and their properties, is covered in Device modeling and simulation Section. Design and theory Section comprises details of investigated thin film transistor. The simulation results, analysis, and discussion of the effects of various dielectric materials, as well as thickness and operating temperature of the dielectric material, on DC and analog parameters, are covered in Influence of high- κ dielectric material on the performance of ZnO TFTs: Section, Influence of thickness of dielectric on the analog performance of ZnO TFTs: Section and Influence of Temperature on the analog performance of ZnO TFTs: Section, respectively. Device modeling and simulation Physical models Silvaco Atlas, which solves equations relating the material’s microscopic parameters and the device’s geometry to its macroscopic performance using the finite difference method, is used to create the device simulation model. By considering the thickness of 200 μm along the z-direction (perpendicular to the x-y plane) in the equations, the entire device was simulated using a two-dimensional platform (x-y plane) while retaining its basic physical characteristics. Solving governing equations for thin amorphous oxide (ZnO) takes into account fundamental physical relationships for band tail states, carrier distribution, etc. Band tail state and carrier transport model When a positive gate bias is applied, accumulation in a bottom gate TFT creates a channel at the interface between the gate-oxide (such as HfO2) and the channel layer (ZnO). By including the drift-diffusion model, which essentially solves the continuity equation and Poisson’s equation consistently, it is now possible to simulate the transport carriers from source to drain through the channel. For carrier distribution, Fermi-Dirac distribution model has been considered. However, it’s crucial to solve these equations using the appropriate mobility model. The effective mobility model, as explained below, is taken into consideration in order to comprehend the carrier transport mechanisms in amorphous oxide channels and then incorporate them into the simulation model30,31. Mobility modeling in bulk material involves the following: (a) characterizing electron and hole mobility in relation to doping and lattice temperature, (b) describing the transition between the low-field mobility and saturated velocity regions, and (c) characterizing saturation velocity in relation to lattice temperature. Charge transport in these disordered systems has been described by mobility models such as the Multiple Trapping and Release (MTR) Model, Percolation Theory (Effective Medium Approaches), and Field-Dependent Mobility models. Carriers travel through extended states in the MTR model, but they are momentarily trapped in the bandgap’s localized tail states. Between trapping events, transport takes place via these extended states after carriers are thermally excited from trap states into the conduction/valence band. Mobility is highly influenced by temperature. According to percolation theory, carrier transport takes place in a disordered energy landscape via a network of “favorable” paths. Carriers can move above a critical energy level, also known as the mobility edge. Conduction is dominated by pathways with greater connectivity and energy continuity. The distribution of the energy barrier can affect non-linear field dependency. In fact, in a disordered amorphous material, the band gap contains a considerable number of randomly distributed defect states (tail states). In contrast to a-Si, the density of localized traps or band tail states in amorphous oxide semiconductors (ZnO) is substantially lower. The tail state carrier density is substantially less than the free carrier density at room temperature (300 K). Percolation thus outweighs trap limited conduction (TLC). The average height of the potential barrier (Φbarrier) and the average barrier width (Wbarrier) to distance (Dbarrier) ratio are the primary determinants of percolation conduction, as shown in Fig. 1. As a result, in the case of a TFT device, the conduction process depends on both the gate voltage and operating temperature. Interestingly, percolation takes over at high gate voltage32, while trap limited conduction predominates at low gate voltage33. Mechanism of multiple trap and release events and percolation for carrier transport. The ATLAS tool of the SILVACO TCAD program simulates the structural concept of TFT devices as well as systems of disordered materials. Based on the level of disorder in TFT structures, this technique allows us to define the energy distribution of defect states inside the band gap of semiconductor materials. This is crucial for the precise consideration of the electrical characteristics of materials like ZnO. A TFT module, in ATLAS simulates systems of disordered materials. The energy distribution of defect states in the band gaps of semiconductor materials can be defined. The disordered structure causes the continuous density of states (DOS) distribution to stretch from the valence band edge (Ev) to the conduction band edge (Ec). This also comprises Gaussian distributions of mid-gap states and band tail states with exponential decay. Hence, defect density models are therefore included. A continuous distribution of the sub-gap DOS, G(E), is provided here in order to properly model the device. It can be shown as follows22,34: This consists of exponentially acceptor like conduction band tail (CBT) and donor like valance band tail (VBT). It also includes GTA(E) and GTD(E) band tail states. A shallow acceptor [donor] like Gaussian band state that is caused by oxygen vacancies, GGA(E) [GGD(E)], is also making this composition. All the components of DOS are shown in Fig. 2 and are given as: Schematic density of states model used for the simulation of ZnO channel material. Carrier recombination model Drift alone (beyond threshold) and diffusion alone (for subthreshold operation) control the movement of electrons from source to drain procedures. However, these carriers undergo recombination-loss due to bulk native defects in ZnO. Shockley–Read–Hall (SRH) recombination governs the recombination mechanisms for the carriers resulting from defects, as explained from the equation below. Here, represents the difference between trap energy level and intrinsic fermi level. is the lattice temperature. and are electron and hole lifetime respectively. The trap assisted tunneling model also exhibits significant influence, according to Eq. 7, because it depends on certain field-dependent functions ( and ), and other factors which influence the SRH model. All the DOS input parameters with description and their respective values are given in Table 1. Table 1The density of States input parameters in ZnO34.ParametersDescriptionValueNTA (cm− 3eV− 1)Density of acceptor like states in conduction band1.12 × 1021NTD (cm− 3eV− 1)Density of donor like states in valance band4.0 × 1020WTA (eV)Characteristics decay energies of the CBT0.025WTD (eV)Characteristics decay energies of the VBT0.05NGA (cm− 3eV− 1)Maximum density of shallow acceptor like Gaussian states5.07 × 1017NGD (cm− 3eV− 1)Maximum density of shallow acceptor like Gaussian states1.5 × 1018WGD (eV)Characteristics deviation of shallow donor like Gaussian states0.1WGA (eV)Characteristics deviation of shallow donor like Gaussian states0.1EGD (eV)Energy peak for shallow donor like states0.4EGA (eV)Energy peak for shallow acceptor like states0.4 The density of States input parameters in ZnO34. In order to guarantee the accuracy of the simulation, pertinent and necessary physical models were included. Figure 3 compares the simulated and experimental transfer characteristics after the model was calibrated using experimental data35. Except for a slightly lower OFF current in the simulation, the results of the simulation closely match the experimental data. This disparity is probably caused by differences between the simulation’s assumed band tail parameters (conduction band edge intercept densities, characteristic decay energy, etc.) and the real values found in the ZnO film that was created35. Ids-Vgs curves for simulated and experimental35 ZnO TFT. Measurement-side parasitic in the manufactured device and insufficient modeling of contact and leakage phenomena in the simulation are probably the causes of the roughly two-order magnitude difference between the simulated and experimental off-current. Specifically, compared to an ideal, contact-limited simulation, Schottky barriers or large source/drain series resistance can suppress the measured off-current; on the other hand, unmodeled trap-assisted leakage, gate leakage through the dielectric, or surface leakage paths can increase experimental Ioff current. A large portion of the discrepancy may be resolved by adding series contact resistance and non-ideal contact physics (thermionic emission, thermionic-field emission, and tunnelling) to the device simulation. Therefore, we deduce that the observed two-order mismatch is caused by a combination of contact effects and trap-assisted leakage, with contact/Schottky effects being primarily responsible for suppressing the measured off current in our devices. Design and theory Device structure In this work, staggered bottom gate TFTs structures with ZnO as a channel material, as shown in Fig. 4, have been investigated. Aluminum has been taken as source, drain and gate contact material. It is because aluminum has low work function, high strength, low reactivity to ambient circumstances, high melting point and low contact resistance36. Many defects are created at the interface when ZnO comes in contact with Aluminum. Due to which, an increase in the surface doping density occurs near the contact. This reduces contact resistance. In this study, various gate oxide materials were used. Some important parameters of channel material (ZnO), used in the simulation, along with their source33,37,38 are provided in Table 2 for ready reference. Table 2 also shows other device material parameters and device structural parameters used in our simulation. The amorphous ZnO active layer is 0.04 μm thick, the HfO₂ insulating layer is 0.12 μm thick, and the substrate is glass. The TFT measured 30 μm in total length. The electrical contacts (source and drain), which are in charge of current flow, were placed on top of the device, while the Al-electrode material functioned as the gate. The channel length between the drain and source contacts is 10 μm, and both electrodes are 10 μm long. Schematic diagram of simulated ZnO based thin film transistors. Table 2Device structure and material parameters of ZnO based TFTs33,34,37,38.RegionParameter descriptionValueZnO(n-type channel)Effective DOS for free carriers [in CB] (NC)Effective DOS for trapped carriers (NT)Electron affinity (χ)Relative Dielectric constant (ε)Energy band gap at 300 K (Eg)Band mobility (electron) (µn)Band mobility (hole) (µp)Lifetime for electrons (τn)Lifetime for holes (τh)Doping concentration (Nd)ThicknessLength6.0 × 1019cm−3 eV− 12.0 × 1019cm−3 eV− 14.35 eV8.53.37 eV60 cm2/V-s0.4 cm2/V-s1 ns1 ns1015 cm− 340 nm10 μmDielectric MaterialsBand Gap, SiO2Permittivity, SiO2Band Gap, Si3N4Permittivity, Si3N4Band Gap, HfO2Permittivity, HfO2Band Gap, Al2O3Permittivity, Al2O3Band Gap, Ta2O5Permittivity, Ta2O5Band Gap, La2O3Permittivity, La2O3Thickness9.0 eV3.94.45 eV7.55.7 eV229.0 eV9.34.5 eV265.8 eV27Variable, 80/120/160/200/240 (in nm)Source and Drain contactsLength/thickness/widthSource-drain spacingWork function of Aluminium10/0.1/200 (in µm)10 μm4.08 eVGate contactsLength/ thickness/widthGate to source/drain overlap length30/0.1/200 (in µm)20 μm Device structure and material parameters of ZnO based TFTs33,34,37,38. ZnO (n-type channel) Effective DOS for free carriers [in CB] (NC) Effective DOS for trapped carriers (NT) Electron affinity (χ) Relative Dielectric constant (ε) Energy band gap at 300 K (Eg) Band mobility (electron) (µn) Band mobility (hole) (µp) Lifetime for electrons (τn) Lifetime for holes (τh) Doping concentration (Nd) Thickness Length 6.0 × 1019cm−3 eV− 1 2.0 × 1019cm−3 eV− 1 4.35 eV 8.5 3.37 eV 60 cm2/V-s 0.4 cm2/V-s 1 ns 1 ns 1015 cm− 3 40 nm 10 μm Band Gap, SiO2 Permittivity, SiO2 Band Gap, Si3N4 Permittivity, Si3N4 Band Gap, HfO2 Permittivity, HfO2 Band Gap, Al2O3 Permittivity, Al2O3 Band Gap, Ta2O5 Permittivity, Ta2O5 Band Gap, La2O3 Permittivity, La2O3 Thickness 9.0 eV 3.9 4.45 eV 7.5 5.7 eV 22 9.0 eV 9.3 4.5 eV 26 5.8 eV 27 Variable, 80/120/160/200/240 (in nm) Length/thickness/width Source-drain spacing Work function of Aluminium 10/0.1/200 (in µm) 10 μm 4.08 eV Length/ thickness/width Gate to source/drain overlap length 30/0.1/200 (in µm) 20 μm Theory TFTs rely on the alteration of a current that travels through the gate electrode between the source and drain. And, to create the field effect, the gate electrode alters the current flowing by capacitive injection of carriers adjacent to the dielectric semiconductor contact39,40. By applying appropriate value of gate voltage, the active layer or channel becomes conductive. The drain current rises linearly with rise in drain voltage. This linear increment of drain current continues till the condition, as per equation below4,41: Again, for the condition, , TFT device is stated to be in saturated condition, where the current is defined by Eqs. (8), This saturation drain current can be modelled by modulating VGS. This allows large number of carriers to get accumulated near at the interface of dielectric and semiconductor interface. Major performance parameters of TFTs are field effect mobility, threshold voltage, Ion/Ioff current ratio and subthreshold slope. The effective carrier mobility in the TFT channel, as determined by the transfer characteristics of the transistor, is known as field effect mobility. When a gate electric field is applied, it describes how effectively carriers, i.e., electrons and holes move through the channel. Carriers travel through the channel more quickly when there is high field effect mobility, which raises the current at the same gate voltage. It also shows the quality of the channel. Defects, scattering, interface states, crystallinity (which is crucial for oxide TFTs like ZnO and IGZO), and surface roughness all have a significant impact on this mobility. Field Effect Mobility, which is derived from the transistor’s current-voltage characteristics, is the mobility of charge carriers in the TFT channel when the gate electric field is present. The gate’s ability to modulate channel conductivity is directly reflected in this. In addition, the mobility seen in TFTs differs from the typical intrinsic bulk mobility of semiconductors. It is because of charge transfer is made possible by the narrow channel (thin film) in TFTs42. Field effect mobility in TFTs is determined by the following relation at low VDS43. The lowest gate-to-source voltage at which a conductive channel forms between the source and drain and permits a sizable drain current to flow is known as the threshold voltage () of a TFT44. Very few carriers are present in the channel below , indicating that the transistor is in the OFF state. Again, sufficient carriers are induced above , indicating the formation of a channel and a rapid increase in current. A low threshold voltage can be used for low-power operation and faster switching, while a high can be used for slower switching but better OFF-state leakage control. It is a crucial factor that controls TFT stability, power consumption, and switching behavior. One method for evaluating this parameter is to extrapolate it linearly from the IDS-VGS curve at very low VDS values. For digital switching applications (such as logic circuits and displays), a high current ratio allows the TFT to distinguish between ON and OFF states with clarity. Poor device performance is indicated by a low current ratio, which also suggests a high leakage current or a weak ON current. A TFT’s ability to transition between conducting (ON) and insulating (OFF) states is gauged by its current ratio (Ion/Ioff). Greater values indicate more energy-efficient devices, reduced leakage, and improved switching performance. Ion/Ioff current ratio is the ratio of ‘on’ current to the minimum measured current (off current). A higher ‘on’ current gives better driving capability. An ‘off’ current corresponds to leakage current. Hence, a higher Ion/Ioff ratio is preferred45,46. In the subthreshold (weak inversion) region of a TFT, the subthreshold slope (SS) is a parameter that indicates how well the gate voltage regulates the drain current. It is the amount of change in gate voltage needed to alter the drain current by a factor of 10, or one order of magnitude. Faster switching, improved gate control, and a sharper turn-on are all correlated with lower SS. A higher SS denotes poorer electrostatic control, more leakage, and slower switching. A value of SS = 60mV/dec is reported for optimal MOSFETs operating at room temperature (300 K). However, SS is typically much higher (100 mV/dec to several hundred mV/dec) for TFTs (ZnO, IGZO, a-Si, etc.) because of the following: trap states at the semiconductor/dielectric interface, bulk defects in the semiconductor, and poor dielectric quality. A TFT’s subthreshold slope indicates how abruptly it turns on. Better switching speed, less leakage, and more energy-efficient operation are all associated with lower SS. Generally, SS is the inverse of the slope of log (ID) at different gate voltage as per Eq. (11)47. Furthermore, analog properties like transconductance (gm), drain conductance (gd), transconductance generation factor (TGF) and early voltage (VEA) have also been studied. Transconductance (gm) is the rate at which the drain current (ID) changes in relation to the gate-to-source voltage (VGS) in a thin film transistor or any field-effect transistor when the drain-to-source voltage (VDS) remains constant. It indicates the degree to which the gate voltage regulates the channel current. A higher transconductance indicates better amplification capability because a small change in VGS results in a large change in drain current. The gain capability of TFTs in analog circuits is directly determined by their conductance, which is a measurement of the drain current’s sensitivity to gate voltage. gm is the derivative of drain current with respect to VGS at constant VDS28. The rate at which the drain current changes in relation to the VDS while maintaining a constant VGS is known as the drain conductance (gd) of a TFT. In the saturation region, gd quantifies the degree to which the drain current is dependent on the VDS. The ideal TFT would have zero drain conductance, or infinite output resistance, since the drain current should ideally be independent of VDS at saturation. However, in real TFTs, the drain current still rises with VDS, resulting in a finite gd, because of trap states, mobility degradation, and channel length modulation. Because it directly affects output resistance, gd, a crucial parameter for analog performance, measures how sensitive drain current is to VDS in a TFT’s saturation region. gd is determined by differentiating the drain current Eq. (8), with respect to VDS, keeping VGS constant48. The ratio of gm to drain current is known as the Transconductance Generation Factor (TGF) of a TFT. TGF indicates the transconductance generation efficiency of a TFT per unit of current consumption. For low-power analog/RF design, a high TGF indicates that the device can achieve strong gate control (high gm) while using less current. Poor efficiency is indicated by a low TGF, meaning that more current is required to produce the same gain. The gm /ID methodology is frequently used by designers in analog and radio frequency circuits to optimize power vs. performance trade-offs. Since TFTs (such as ZnO, IGZO, and HfO2-based) typically have lower mobility than crystalline MOSFETs, TGF aids in comparing efficiency across various materials and dielectrics. It can be expressed mathematically as follows (in V− 1)39: The extent that the drain current (ID) of a thin film transistor (TFT) depends on the VDS in the saturation region is described by the early voltage (VEA). Once the device reaches saturation, the drain current in a perfect MOSFET or TFT should stay constant (regardless of VDS). However, in actual devices, the current keeps rising slightly with VDS because of channel length modulation. The VEA is used to quantify this effect. The relationship below can be used to obtain it39: Influence of high- κ dielectric material on the performance of ZnO TFTs Since the last few decades, SiO2 has been the most widely used gate dielectric material. The size of the transistor can also be diminished by reducing the SiO2 gate dielectric thickness, which in turn raises the gate capacitances per unit area, thus enhanced the performance of the transistor. However, diminishing the SiO2 gate dielectric thickness beyond a certain limit, can lose its dependability due to a spike in tunneling gate leakage current, which may eventually result in higher power consumption and subsequently impair the transistor performance. To tackle this issue, researchers developed the idea of effective oxide thickness (EOT). EOT signifies increased physical gate dielectric thickness without an increase in electrical or effective gate dielectric thickness. This is easily accomplished by raising the gate dielectric’s dielectric constant, or ‘k’ which lowers gate leakage current while maintaining ideal gate capacitance. Considering SiO2 gate dielectric with dielectric constant ‘k’=3.9, and if we consider another gate dielectric material, ‘X’, with dielectric constant ‘k’=39, then the material ‘X’ can be 10 times thicker than SiO2 gate dielectric. As a result, high-k dielectric materials are physically thicker rather than electrically thicker, raising gate capacitance per unit area without commensurate leakage consequences4,27,41,49. In an optimal state, each carrier (or electron) is on the gate’s uphill side. The carriers have an approximate likelihood of landing on the downstream end of the gate despite really splashing over it when the gate dimension is incredibly thin. This is when gate leakage occurs, which is known as quantum mechanical tunneling. As mentioned, thinner gate cause tunneling current and thus results in high power consumption. By raising the gate dielectric’s physical thickness, this tunneling current can be decreased. The performance of the transistors is once again degraded when the electrical oxide thickness of SiO2 increases along with its physical thickness. As one of the strategies opted to enable even further miniaturization of the devices, high-k dielectrics have been utilized in semiconductor-based devices to substitute low k-SiO2 dielectric or any other dielectric50–52. Enhanced drain current and capacitance per unit area (Cox) of the device arise from larger gate dielectric thickness, boosting the device’s overall performance: Where is the dielectric constant of the oxide material. represents absolute permittivity. L, W, and stand for the dielectric material’s length, width, and thickness, respectively. This section has been focused on examining the effect of high-k dielectric gate material on the analogue performance of TFTs. The chosen gate dielectric materials are SiO2, Si3N4, Al2O3, HfO2, Ta2O5, and La2O3. Initially thickness of the gate dielectric material is kept 160 nm for simulating all the TFT structures. From the graph shown in Fig. 5, it is evident that TFT with La2O3 gate dielectric material shows comparatively better peak values than HfO2 gate oxide materials. But, as seen from Table 3, TFTs with Ta2O5 and La2O3 gate dielectric material have less mobility and lesser SS as compared to the TFTs with HfO2 gate dielectric material. Due to these problems associated with Ta2O5 and La2O3, we consider TFT with HfO2 gate dielectric a better performer. Different analog performance parameters for different gate dielectric material is plotted in Figs. 5, and 6. The VGS varied from − 2 V to 20 V, with a step size of 0.1 at room temperature of 300 K. At low drain voltage (VDS=0.1 V), Fig. 5 (a) and (b) depict the drain current in both linear and logarithmic scales, respectively. Maximum value of drain current (IDS) rises from 27µA to 240 µA, when gate dielectric material changes from SiO2 to HfO2. HfO2, Ta2O5, and La2O3 are examples of higher-κ dielectrics that increase gate capacitance without significantly raising leakage current, which results better transistor performance and a higher current at the same voltage. Conventional SiO2, which has a lower dielectric constant, shows poor performance. The Output drain current at fixed gate voltage (VGS) of 8 V for different VDS is plotted in Fig. 5 (c). Each curve exhibits current saturation, which denotes that the current first rises with VDS before plateauing. The dielectric material determines the current level at which saturation takes place. Rapid current increases at low VDS are a sign of improved conductivity and channel formation in high-κ dielectric materials. SiO2 rises very slowly, suggesting that this dielectric performs worse in devices. Higher saturation drain currents are made possible by materials with higher dielectric constants, which improves channel conduction and carrier mobility. Figure 5 (d) shows the drain conductance (gd) curve. Since the channel is in the linear/triode region, current is highly dependent on VDS, and at low VDS, gd is at its highest. The channel enters the saturation region, where drain currents should ideally not be dependent on VDS, as indicated by the sharp drop in gd that occurs as VDS increases. The gd is zero for all dielectrics at high VDS (> 8 V), indicating superior current saturation. In addition to having a higher drain current, high-κ dielectric materials have a larger gd at low VDS, which results in a lower output resistance. These work well for high transconductance applications (such as digital TFTs and switching), but they are not the best for high-gain analog. At low VGS (near the threshold voltage), transconductance (gm) increases dramatically, as seen in Fig. 6 (a). It peaks at higher VGS, after which, it either saturates or slightly declines. Although mobility degradation and series resistance effects, can lower gm at higher voltages, this shape is typical because gm depends on (VGS-VTh). When compared to SiO₂ and Si3N₄, high-κ dielectrics (La₂O₃, Ta₂O₅, and HfO₂) significantly increase gm. The industry is moving to high-κ dielectrics for TFTs and MOS devices. The gm in TFTs is greatly increased by high-κ dielectrics, which enhances analog performance. Early voltage (VEA) rises with VDS for all dielectrics, as shown in Fig. 6 (b). This indicates that the device gets closer to more optimal current saturation as VDS rises and the impact of channel-length modulation diminishes. At VDS = 20 V, Ta₂O₅ and HfO₂ exhibit the highest VEA values (up to 1200 V and 1050 V, respectively), indicating superior channel-length modulation suppression, more output resistance, and improved analog performance. The mid-range (up to 800 V) of Al₃O₃, Si₃N₄, and SiO₂ indicates moderate suppression of channel-length modulation, making their performance reasonable but not optimal. At low VGS (near threshold), transconductance generation factor (TGF) is extremely high for all dielectrics, as can be seen in Fig. 6 (c). TGF rapidly decreases as VGS rises, getting close to zero for VGS > 8 V. This demonstrates that while efficiency decreases at high overdrive voltage, it peaks at weak/moderate inversion. For low-power analog circuits, a device with a high TGF at low VGS can achieve a high gm with a low drain current. When low-voltage, low-power analog operation is needed, Ta2O₅ and HfO₂ work the best. (a) Drain current as a function of gate voltage in linear scale, (b) Drain current as a function of gate voltage in Log scale, (c) Output drain current as a function of drain voltage, and (d) drain conductance as a function of drain voltage, for different dielectric materials. Table 3Obtained parameters of TFT for different dielectric materials.Different oxide materials, for (Tox = 160 nm)VTh (in V)Mobility (in cm2/Vs)Current ratio (Ion/Ioff)SS (in V/Decade)SiO24.4937.01 × 10100.721Si3N42.1538.27 × 1090.611Al2O31.6640.56.3 × 1090.587HfO20.4953.21.5 × 10110.396Ta2O50.3548.62.5 × 10100.353La2O30.5141.01 × 10100.487 Obtained parameters of TFT for different dielectric materials. (a) Transconductance as a function of gate voltage, (b) Early voltage as a function of drain voltage, and (c) TGF as a function of gate voltage, for different dielectric materials. All the performance parameters like VTh, Ion/Ioff, SS and µFE are extracted from Fig. 5 (a) and (b) and is tabulated in Table 3. SiO₂ has the highest threshold voltage (VTh) (≈4.49 V), making it less appropriate for low-power circuits since it needs a high VGS to turn ON. Conversely, dielectrics with high κ, such as HfO₂, Ta₂O₅, and La₂O₃, exhibit very low VTh, making them ideal for low-voltage, low-power TFT operation. High-k dielectrics improve mobility because they have better gate control and less Coulomb scattering. Table 3 makes clear that, of the various oxide materials under study, HfO₂ has the highest mobility (53.2 cm²/V·s), while SiO₂ has the lowest (37.0 cm²/V·s). As previously mentioned, switching capability is determined by the current ratio. The best current ratio, 1.5 × 1011, is reported by HfO₂. In comparison to other dielectric materials, Si3N₄ and Al2O₃ exhibit a somewhat lower current ratio (~ 109). Additionally, moderate current ratio values are reported by SiO₂ and La2O₃. Sharper switching and improved subthreshold control are associated with lower SS. With the lowest SS (0.353 V/dec) and (0.396 V/dec), Ta₂O₅ and HfO2 exhibits extremely effective switching. SiO₂ has the slowest turn-on and the worst (0.721 V/dec). Subthreshold of TFT is calculated using (18)53: For oxide TFT like ZnO, depletion capacitance is small, hence, Where, . is the interface trap density. Despite having a slightly higher dielectric constant (εr = 27) than HfO₂ (εr = 22), La₂O₃ has a worse subthreshold slope: (a) For HfO₂, the interface trap density () is significantly lower54. It is well known that HfO₂ and ZnO form a more stable and cleaner interface. (b) Fermi level pinning is caused by fixed charges and dipoles within La2O₃. La2O₃ exhibits stronger Fermi-level pinning at the ZnO interface, more fixed charge, and more dipole-related charge fluctuations. By weakening gate control, this raises the subthreshold slope. HfO₂, offers a chemically stable interface with fewer hydroxyl groups, a lower defect density, and fewer fixed charges, which leads to a smaller and, as a result, a lower SS. Therefore, HfO₂’s cleaner interface and decreased trap-assisted charge modulation are responsible for its superior SS, even though it has a slightly lower permittivity. The best candidates for next-generation high-performance TFTs are high-κ dielectrics (HfO₂) which perform significantly better than traditional dielectrics (SiO₂, Si₃N₄) in terms of mobility, ON/OFF ratio, threshold voltage, and subthreshold slope. This analysis can be particularly helpful for the circuit designers working in the field of analog applications. Influence of thickness of dielectric on the analog performance of ZnO TFTs Hence, continuing our discussion in the last section, analog performance parameters are strongly affected by the gate dielectric materials. Replacing the low-κ dielectric material with high-κ material results with low leakage current, low power consumption and enhanced performance devices. Capacitance is inversely proportional to the thickness of gate dielectric materials. Thus, capacitance increases with the decrease in thickness of dielectric material. This induces large amount of charge carriers at the same output voltage, and thus cause increase in drain current. In order to study the impact of dielectric thickness, TFT with HfO2 as dielectric has been studied for five different thickness, 80, 120, 160, 200 and 240 nm. Drain current at different VGS is plotted both in linear and logarithmic scale in Fig. 7 (a) and (b). The VGS was varied from − 2 V to 20 V, with a step size of 0.1 at room temperature of 300 K, at fixed VDS of 0.1 V. Maximum reported value of drain current is 470 µA for gate dielectric thickness of 80 nm. But, since the threshold voltage obtained for the 80 nm thickness is of very low value, and thus making the device possibly, always in ON state, we don’t consider this thickness as optimum value for the analysis of analog performance parameters. Again, for the gate dielectric thickness of 120 nm, maximum reported current is almost 320 µA. This rise in drain current is due to the rise in capacitive coupling between gate and channel in case of low thickness of gate dielectric material51,52,55–57. The drain current variation with VGS for TFTs with varying dielectric thicknesses is displayed in Fig. 7 (a). After the threshold, drain current rises nearly linearly as VGS rises. The subthreshold region is indicated by the extremely low current at negative or nearly zero VGS. All curves exhibit strong gate modulation beyond the threshold. Stronger gate control and higher drain current are the results of a thinner dielectric’s large oxide capacitance. Conversely, a thicker dielectric result in less gate-channel coupling, which lowers the drain current and induced charge. At a fixed VGS of 8 V, Fig. 7 (c), displays the relationship between drain current and VDS. The ohmic region, or linear region, is demonstrated by the drain current increasing nearly linearly at low VDS. The saturation region of TFT operation is demonstrated by the drain current reaching saturation at higher VDS. By increasing gate capacitance and channel charge density, thinner dielectric layers raise drain current. Smaller drain currents result from weaker field-effect modulation caused by thicker oxides’ reduction of gate capacitance. Therefore, current drive capability is greatly increased by decreasing dielectric thickness. There is a trade-off, while thinner dielectrics increase performance, they may also cause reliability and leakage problems. Figure 6 (d) illustrates how gd changes with VDS. As the VDS rises, gd falls, and for all dielectric thicknesses, it nearly saturates to zero at about 8 V. Thinner dielectrics have higher gd values at low VDS. This is because thinner dielectrics offer more robust gate control over the channel, which raises conductance by increasing mobility and drain current. Also, gate-channel coupling is improved by a thinner dielectric, which lowers the threshold voltage and raises the gd. For high gd, a dielectric thickness of 120 nm is ideal because it improves gate control and increases current drive capability. Even though thin dielectrics enhance performance, extremely thin films may have decreased reliability and high leakage currents. As a result, stability and performance must be balanced in device design. gm rises rapidly at low VGS for all dielectric thicknesses, peaks, and then gradually falls or saturates as the VGS rises further, as can be seen in Fig. 8 (a). This behavior is typical because channel formation is strong and channel charge increases quickly at low VGS, while at higher VGS, incremental current gain is reduced by carrier scattering and mobility degradation. High-performance analog circuits are better suited for thin dielectrics (120 nm). Up to roughly 8 V drain bias, the VEA stays extremely low for all dielectric thicknesses, as displayed in Fig. 8 (b). After that, the VEA rises sharply as the VDS rises. For thinner dielectrics, the increase is substantially more pronounced. Better device performance for analog applications (amplifiers) is indicated by higher output resistance, which is a function of higher VEA. At low VGS, TGF is extremely high, as displayed in Fig. 8 (c). After that, it rapidly drops to almost zero for all dielectric thicknesses above about 5 V. TGF is extremely sensitive to VGS in the subthreshold/near-threshold region, as indicated by this steep decline. This decline could be explained by the transistor being in the subthreshold region at very low VGS, where even slight variations in VGS result in exponential increases in current and, consequently, high TGF. Additionally, the device enters strong inversion/saturation at higher VGS, where the relative gain factor decreases as the current rises linearly with VGS. Thin dielectrics are therefore better suited for applications that need low power consumption, strong subthreshold control, and abrupt switching. (a) Drain current as a function of gate voltage in linear scale, (b) Drain current as a function of gate voltage in Log scale, (c) Output drain current as a function of drain voltage, and (d) drain conductance as a function of drain voltage, of HfO2 dielectric material of different thickness. (a) Transconductance as a function of gate voltage, (b) Early voltage as a function of drain voltage, and (c) TGF as a function of gate voltage, of HfO2 dielectric material of different thickness. However, the other performance parameter of TFTs like VTh, Ion/Ioff, SS and µFE has been extracted from Fig. 7 (a) and (b) and are tabulated in Table 5. A steady increase in threshold voltage occurs from 0.01 V to 0.92 V. This is because a thicker dielectric lowers gate capacitance, which means a higher voltage is needed to create a channel. This also suggests that for low-power electronics, higher VTh at large thicknesses decreases low-voltage operation efficiency, which is undesirable. Field effect mobility first rises to a peak of 54.2 cm²/V-s (for thickness 120 nm) from 40.6 cm²/V-s (80 nm), after which it progressively falls to 38.5 cm²/V-s (240 nm). This could be explained by the fact that leakage and interface traps can inhibit mobility at thinner HfO₂ (80 nm). As mentioned earlier, . From this equation, we can write: Table 5Obtained parameters of TFT for different thickness of HfO2 dielectric materials.Different HfO2 Thickness (in nm)VTh (in Volts)Mobility (in cm2/Vs)Current Ratio (Ion/Ioff)SS (in V/Decade)800.0140.62.5 × 10100.2341200.2754.23.1 × 10110.3151600.4953.21.5 × 10110.3962000.7045.03.3 × 10100.4642400.9238.51.6 × 10100.512 Obtained parameters of TFT for different thickness of HfO2 dielectric materials. As HfO₂ thickness increases, there is a slight trend of decreasing extracted as can be seen in Table 4. This is anticipated since decreases with increasing thickness, which modifies the apparent calculated by the formula for a given SS. At 80 nm, the maximum occurs. This qualitatively supports our finding that the 80 nm device exhibits decreased mobility, which can be explained by more interface traps (and probably increased leakage) (Table 5). Table 4Table showing effective oxide capacitance and interface trap density for different dielectric thickness.Thickness of HfO2 Dielectric (in nm)Effective Oxide Capacitance () (in Fcm− 2)Interface Trap Density () (in cm− 2eV− 1)802.439 × 10− 74.45 × 10121201.623 × 10− 74.35 × 10121601.217 × 10− 74.3 × 10122009.739 × 10− 84.13 × 10122408.116 × 10− 83.85 × 1012 Table showing effective oxide capacitance and interface trap density for different dielectric thickness. The ideal thickness (120 nm) strikes a balance between low scattering and good dielectric quality. Again, a very thick dielectric (> 200 nm) decreases mobility, raises channel resistance, and decreases gate control. The optimal charge transport performance is thus achieved at 120 nm. From 2.5 × 1010 (80 nm) to 3.1 × 10¹¹ (120 nm), the current ratio greatly improves before declining at higher thicknesses (down to 1.6 × 10¹⁰ at 240 nm). Leakage from thin oxide (80 nm) may be the cause of this, as it lowers ION/IOFF. Strong gate control and reduced leakage are provided by the moderate thickness (120 nm). Again, a thickness of 200–240 nm weakens the gate coupling, resulting in a higher Ioff and a lower ratio. Therefore, once more, the ideal thickness for switching performance is 120 nm. With thickness, SS rises monotonically from 0.234 V/dec (80 nm) to 0.512 V/dec (240 nm). This is because a thicker dielectric reduces the channel’s electrostatic control, necessitating a greater gate voltage change to modulate current. Thus, at thinner oxides, particularly for 80–120 nm, low SS (nearer to ideal) is attained. The best TFT performance in terms of mobility, current ratio, and manageable SS is offered by the 120 nm HfO₂ dielectric thickness, which makes it ideal for real-world applications. Influence of temperature on the analog performance of ZnO TFTs In this section, effect of temperature on various performance parameter of TFTs, with HfO2 as dielectric and with thickness of 120 nm, is studied. There were two phases to the optimization of the HfO₂ gate dielectric. Initially, the dielectric’s physical characteristics, particularly its thickness, were optimized at a specific operating temperature. This method was chosen because the gate-oxide capacitance (), the electric-field coupling into the channel, and the TFT’s baseline electrical performance (µFE, VTh, SS, and Ion/Ioff) are all directly influenced by the dielectric thickness. Before introducing thermal variations, geometric and electrostatic factors are stabilized by determining the ideal thickness at a fixed temperature. It would be challenging to distinguish geometric effects (thickness-dependent ) from thermally induced material changes if the operating temperature were changed first, as this would concurrently alter microstructure, defect density, and interface chemistry. The second stage involved temperature-dependent simulations after the ideal thickness (120 nm HfO₂) was determined. This was required because, in the ZnO semiconductor, temperature mainly affects trap occupancy, interface states, and Fermi-level movement rather than the dielectric’s geometric parameters. The impact of thermal effects could be separated and assessed without interference from thickness-dependent changes in or field distribution by maintaining a constant thickness throughout temperature analysis. Consequently, a clean, decoupled assessment of the two effects and a more rigorous interpretation of the device physics result from optimizing the thickness first and then conducting temperature analysis. The effect of temperature variation (from 300 to 500 K, with a 50 K gap) on TFT’s analog performance is explained in this section. The covalent bond inside the body material’s lattice begins to break when the temperature rises above room temperature, producing a large number of electron and hole pairs (EHPs). The intrinsic carrier concentration of semiconductors () is directly proportional to the EHP generation rate56,57. According to , the increases exponentially with temperature. Additionally, the band profile of the device at both interfaces (S–C–I and D–C–I) is impacted by the variation in temperature. provides the relationship between energy band () and temperature, where α and β are the body material fitting parameters, (300) =3.37 eV for ZnO material, is the band gap at room temperature, kT is the energy term (product of T and Boltzmann constant), and is the impurity (donor/acceptor) concentration. Due to a shift in the fermi level with respect to the trap distribution within the energy gap, the transistor’s performance degrades with the increment in temperature57. Three factors are frequently held accountable for the degradation: (i) Electron/hole trapping at the junction of two layers of material (ii) the development of ionised oxygen vacancies, and (iii) the emergence of donor-like defects in ZnO. The first factor shifts the threshold voltages to peak values, whereas the second and third factors shifts the threshold voltage to the negative values58. Semiconductor oxide material are widely used in AMOLED, sensors, medical display circuit and many other applications. Operating temperature has a strong impact on the stability of circuit’s applications. For example, for the application based on medical circuits prefer to work in the range of 300–400 K. Thus, for such applications, temperature stability needs to be optimized and studied. At all temperatures, the drain current rises nearly linearly with gate voltage, indicating that it operates in the linear region as opposed to saturation, as shown in Fig. 9 (a) and (b). This behavior suggests that VGS has a significant influence on channel conductance. Mobility deterioration brought on by increased phonon scattering at higher temperature, is the cause of the current decrease with temperature. Higher field-effect mobility and improved on-state conduction are indicated by higher drain current at lower temperatures. Device performance in high-temperature settings may be impacted by suppressed drain current, which indicates decreased drive capability at higher temperatures. The output characteristics of a Thin-Film Transistor (TFT) are depicted in Fig. 9 (c), where drain current is plotted against VDS at various temperatures. Drain current rises nearly linearly with VDS at low VDS, suggesting that the linear/ohmic region is reached. The TFT’s saturation region is indicated by the drain current reaching saturation and becoming almost constant at higher VDS. This validates that the transistor is operating properly with clear channel pinch-off. Although the device operates steadily, high temperatures degrade performance, making it unsuitable for high-temperature applications. The gd of TFTs at various temperatures is plotted against the VDS in Fig. 9 (d). As VDS rises (due to saturation) and temperature rises (due to mobility degradation), gd falls. Better performance and stronger channel modulation are indicated by the highest gd at 300 K. Reduced gd at high temperatures is indicative of higher channel resistance and mobility suppression from phonon scattering. The curves maintain the same general shape even though the conductance decreases with increasing temperature, suggesting that TFT retains appropriate operating characteristics at all temperatures, but performance degrades. With increasing VGS, gm increases rapidly and reaches a maximum close to the threshold region, as displayed in Fig. 10 (a). At higher VGS, gm progressively falls and stabilizes. This behavior indicates that efficiency is reduced by mobility degradation after the VTh, where strong gate control is most effective. The overall gm is suppressed and the peak value shifts slightly lower at higher temperature. This results from phonon scattering at high temperatures, which weakens gate control by lowering carrier mobility. As temperature rises, VEA falls at all drain voltages, as depicted in Fig. 10 (b). This decrease results from increased scattering and mobility degradation at high temperatures, which exacerbates channel length modulation. The TFT has the highest VEA at 300 K, which indicates optimal analog performance. Performance deteriorates at 500 K because the VEA is much lower (40% lower than at 300 K). Higher intrinsic gain is correlated with a higher VEA. As a result, the TFT provides excellent analog gain at 300 K, which qualifies it for amplifier circuits. Analog gain reduces with increasing temperature, which restricts its applicability in hot conditions. One significant analog figure of merit is the TGF. It calculates the transistor’s efficiency in converting gate voltage to drain current in relation to power consumption. Better analog efficiency and gain are associated with higher TGF. TGF is at its highest close to the threshold voltage area, as shown in Fig. 10 (c). This demonstrates that TGF is strongest just above the threshold, where gate modulation to current conversion is most effective. TGF sharply declines as VGS rises after the peak. This suggests that at higher gate biases, where current increases more quickly than transconductance, analog efficiency is lost. In contrast to previous parameters, the 400 K curve displays the highest peak TGF, in this case. (a) Drain current as a function of gate voltage in linear scale, (b) Drain current as a function of gate voltage in Log scale, (c) Output drain current as a function of drain voltage, and (d) drain conductance as a function of drain voltage, of HfO2 dielectric material with thickness of 120 nm for different dielectric materials. (a) Transconductance as a function of gate voltage, (b) Early voltage as a function of drain voltage, and (c) TGF as a function of gate voltage, of HfO2 dielectric material with thickness of 120 nm for different dielectric materials. Higher temperatures yield larger TGF values than 300 K at very low gate voltages. TGF values, however, converge to zero at higher VGS for all temperatures. This implies that while high temperatures generally decrease drive capability (as shown in transfer/output plots), moderate temperatures may improve analog efficiency close to the threshold. All performance parameters of TFT with HfO2 dielectric and 120 nm thickness, have been extracted from Fig. 9 (a) and (b) respectively, and tabulated in Table 6. As the temperature rises, VTh shifts positively. This could be explained by the fact that higher temperatures cause more interface states and carrier scattering, which necessitate a stronger gate bias in order to invert the channel. At higher temperatures, the device requires a higher voltage to turn on, which marginally reduces switching efficiency. Mobility sharply declines as temperature rises. The reason for this is that at higher temperatures, phonon scattering becomes important. We separated two primary mechanisms to investigate the origin of the performance degradation observed above 400 K: (i) phonon (lattice) scattering, which decreases carrier mobility with increasing temperature, and (ii) trap-assisted processes (thermally activated detrapping and trap-assisted leakage), which both increase leakage and decrease mobile carrier density. We used both power-law and Arrhenius fits to analyze the temperature dependence of mobility and current in order to determine the dominant mechanism. Table 6Obtained parameters of TFT at different temperatures.Different Operating Temperature for HfO2 of Thickness of 120 nmVTh (in Volts)Mobility (in cm2/Vs)Current Ratio (Ion/Ioff)SS (in V/Decade)300 K0.2754.23.1 × 10110.315350 K0.3351.44.3 × 10100.314400 K0.3945.23.4 × 10100.312450 K0.4230.82.4 × 10100.310500 K0.4627.71.2 × 10100.309 Obtained parameters of TFT at different temperatures. A power-law dependence is produced by phonon scattering. Over 300–500 K, a log fit yields m=0.55 (very small). For oxide semiconductors, a phonon scattering exponent is usually (m=1 to 2). Therefore, m=0.55 shows that the rapid mobility collapse above 400 K cannot be explained by phonons alone. Again, hopping conduction or trap-dominated transport exhibits thermally activated behavior, with . For 400–500 K, excellent linearity (R2=0.995) is achieved. This robust activation behavior attests to the dominance of trap-related processes above 400 K. The subthreshold slope and gate leakage significantly increase with temperature in our devices, suggesting that trap activation and trap-assisted leakage dominate the degradation. The mobility decrease above 400 K clearly follows an Arrhenius trend with extracted activation energy tens–hundreds meV. Therefore, the data show that trap-assisted processes are the main cause of the observed performance loss above 400 K, even though phonon scattering contributes to mobility reduction at elevated temperature. Performance is deteriorated because of the direct decrease in drive current and transconductance caused by this decreased mobility. The current ratio drastically decreases as the temperature rises. This is because enhanced thermally activated carriers cause off-state leakage current to increase at higher temperatures, narrowing the ratio, reduced switching capacity, which makes it more difficult to distinguish between digital circuits’ ON and OFF states. The behavior of SS is almost constant with temperature. This is because the interface trap density and gate dielectric quality, which are temperature-stable in this HfO₂ device, are the main factors limiting SS. Sharp switching is maintained because the device maintains a good subthreshold slope even at high T. Sharp switching is maintained because the device maintains a good subthreshold slope even at high temperatures. At 300 K, the device performs at its best; at higher temperatures, switching efficiency and gain are decreased. This demonstrates that although the TFT has good gate control, it has thermal limitations. Conclusion The study examined the effects of temperature, thickness, and gate dielectric material on TFT performance using ATLAS TCAD simulations. The best dielectric was found to be HfO₂; thinner layers improve capacitance and current, but if they are reduced too much, they result in increased leakage and a low threshold voltage. With a thickness of 120 nm, HfO₂ demonstrated exceptional low-temperature performance, exhibiting a sharp SS (~ 0.31 V/decade), high mobility (~ 54 cm²/V-s), and a very high Ion/Ioff current ratio (~ 10¹¹). At the end, in order to examine the thermal stability and reliability, thermal analysis of the device was performed. These TFTs show improved performance for room temperature to moderate temperatures, but they perform less well for digital switching at high temperatures (> 400 K). The Ion/Ioff ratio decreased, mobility deteriorated, and leakage increased as the temperature rose, but SS stayed constant, indicating robust gate control. Future TFTs will need high-k dielectrics with a large bandgap, strong thermal stability, low trap density, stable interfaces, and good morphology. The current model does not fully account for interface state density variation with temperature and bias stress, despite the strong correlation between simulation and experimental data. There was no explicit modeling of effects like contact resistance, self-heating, and bias instability. Furthermore, the study ignores potential grain boundary scattering in polycrystalline ZnO films, which may affect mobility at higher temperatures, and assumes uniform dielectric properties. In order to more accurately forecast long-term device stability, future research may concentrate on: including bias stress and trap dynamics models. Also, investigating various high-k dielectrics or bilayer dielectrics (such as HfO₂/Al₂O₃ or HfO₂/Si₃N₄) in order to minimize interface traps and leakage. Looking into doping effects and bandgap engineering to enhance thermal stability and carrier transport even more. Investigating nanostructured ZnO channels for improved mobility and reduced defect density, such as nanorods or nanowires. Author contributions Pranaw Kumar : Conceptualization, Methodology, Simulation, Device Modelling, writing of original draft, Syed Sadique Anwer Askari contributed to formal analysis, validation, and review and editing of the manuscript. Mukul Kumar Das : Supervision, Conceptualization, Writing – review & editing. Funding Open access funding provided by Manipal University Jaipur. This work is funded by MHRD, Government of India, through the project, Establishment of Centre of Excellence in Renewable Energy, under FAST scheme (Sanc. No. F.No.5–6/2013-TS-VlI dated 4 August 2014) at the Indian Institute of Technology (Indian School of Mines), Dhanbad. Data availability The authors confirm that the data supporting the findings of this study are available within the article. The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Ethics approval This manuscript has not been submitted or published elsewhere. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Syed Sadique Anwer Askari, Email: syed.askari@jaipur.manipal.edu. Mukul Kumar Das, Email: mukulkdas@iitism.ac.in. References Associated Data Data Availability Statement The authors confirm that the data supporting the findings of this study are available within the article. The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.