1. Structure of Double-gate PBTTT OTFT

Fig. 1(a) illustrates the device structure of the DG OTFT. The device was fabricated by forming a top-gate insulator (TGI) and a top gate on a coplanar bottom-gate structure, and separate contact pads were implemented to apply bias voltages to each gate independently. Fig. 1(b) presents a cross-sectional view of the DG OTFT obtained by focused ion beam (FIB) analysis. Fig. 1(c) shows the architecture of the SG OTFT fabricated in a coplanar bottom-gate configuration, and Fig. 1(d) provides the corresponding FIB cross-sectional image. Fig. 1(e) summarizes the fabrication process flow for the DG OTFT.

Fig. 1. Device concept and physical characterization of the PBTTT OTFT. (a) 2D schematic image of the double-gate (DG) PBTTT OTFT, (b) FIB cross-sectional image of the DG OTFT, (c) 2D schematic image of single-gate (SG) PBTTT OTFT, (d) FIB cross-sectional image of the SG OTFT, and (e) fabrication process of DG OTFT devices.

1. Single-gate PBTTT TFT Characteristics and Long-term Reliability

To evaluate device stability under stress, the transfer and output characteristics of the SG PBTTT OTFTs were measured and used as a reference for comparison. Figs. 2(a) and 2(b) show the transfer and output curves measured under the flat condition, confirming that the fabricated SG OTFT operated with an on/off ratio of approximately 10⁵.

In addition, long-term stability was evaluated by monitoring the devices for 40 days under identical ambient conditions, with and without SU-8 passivation. Fig. 2(c) and 2(d) show the temporal change in subthreshold swing (SS) and field-effect mobility over the 40 days, respectively.

Fig. 2. Electrical characteristics of the single-gate (SG) PBTTT OTFT. (a) Transfer curves and (b) output curves of the SG OTFT measured under the flat condition. (c) Time-dependent trends in subthreshold swing and (d) field-effect mobility of the SG OTFT with and without SU-8 passivation.

For devices without SU-8 passivation, a rapid degradation in SS and mobility was observed, and the devices failed to operate after approximately 10 days. In contrast, passivated devices maintained their performance throughout the 40-day test. Based on these results, an SU-8 passivation layer was employed for subsequent experiments to ensure long-term operation.

2. Durability After Negative Gate-bias Stress (NGBS) and Bending Stress

To analyze device performance changes induced by NGBS, a continuous gate bias of -20 V was applied for 1000 s. Figs. 3(a) and 3(b) show the transfer and output characteristics of the SG OTFT after a 20 min recovery period following NGBS. After 1000s of NGBS, a V_th shift of −4.2 V was observed, and upon recovery, V_th shifted by 11.67 V, from −13.6 V to −1.93 V. Under the same NGBS condition, under the same stress conditions, the drain current decreased from 1.3 µA to −0.8 µA, and subsequently increased to approximately 1.6 µA after 20 minutes of recovery, exceeding its initial value. These results indicate that SG OTFTs undergo noticeable characteristic changes within a relatively short period of time. To examine changes in device characteristics under bending stress, bending tests of 100, 200, and 300 cycles were performed, and the transfer and output curves were measured at each step. Figs. 3(c) and 3(d) show the results for a device bent to a radius of 7.05 mm, which was selected as a representative fixed bending condition within the commonly used 5-10 mm range for flexible electronics and wearable devices ^{[24, 25]}. As the number of bending cycles increased, the on-current at V_G = −60 V decreased from 0.7 µA to 0.2 µA, and V_th increased from 13.34 V to 19.7 V. In addition, at V_G = −30 V, the drain current decreased from −0.4 µA to −0.1 µA. This degradation is attributed to a reduction in carrier mobility caused by the applied mechanical stress ^[16].

Fig. 3. Negative gate-bias stress (NGBS), recovery behavior, and bending durability of single-gate (SG) PBTTT OTFTs. (a) Transfer and (b) output characteristics of stress and recovery behavior of NGBS. (c) Transfer curves after the bending test (100 cycles step), and (d) output curves.

3. Double-gate PBTTT TFT Characteristics

For parameter extraction, the gate capacitance was determined according to the operating mode: bottom-gate capacitance (C_BG) for bottom-gate operation, top-gate capacitance (C_TG) for top-gate operation, and C_BG + C_TG for dual-gate operation in which both gates contribute simultaneously to channel modulation.

For DG devices, the electrical performance was compared by defining three operating modes depending on which gate terminal was biased: TG (V_BG = 0), BG (V_TG = 0), and DG (V_TG = V_BG) operation. Figs. 4(a) and 4(b) present the transfer and output characteristics measured under the three operations. Under DG operation, the on/off ratio was approximately 50× higher than that under BG operation and ∼10× higher than that under TG operation, and the field-effect mobility increased by ∼30× and ∼3×, respectively. The DG mode also exhibited the smallest SS, achieving 3.73 V/dec, compared with 7.41 V/dec and 7.02 V/dec for BG and TG operation, respectively. Consistently, the DG mode produced the highest drain current, whereas the BG mode yielded the lowest drain current.

Fig. 4. Double-gate (DG) PBTTT OTFT characteristics under flat and curved conditions and NGBS. (a) Transfer curves and (b) output curves of the DG OTFT measured under the flat condition. Transfer curves comparison of the DG OTFTs measured under (c) flat condition and (d) curved condition. (e) Transfer curve after NGBS. (f) Transfer curves and (g) output curves of top-gate bias variation with various bottom-gate bias (VTG: −30-0 V). (h) Transfer curves and (i) output curves of bottom-gate bias variation with various top-gate bias (VBG: −30-0 V).

Figs. 4(c) and 4(d) compare the performance of a device with W/L = 5000/75 µm under flat and curved conditions. Measurements were conducted using the same curved jig employed for the SG OTFT evaluation. Under the curved condition, the DG devices exhibited only limited variation in on-current and V_th compared with the flat condition. These results demonstrate that, compared with the SG structure, the DG OTFT provides improved reliability when operated as a flexible device. For direct comparison, the DG OTFTs were tested under the same NGBS conditions as the SG OTFTs, with a continuous gate bias of -20 V applied for 1000 s as a representative moderate bias-stress condition for stability evaluation. The −20 V bias was selected to represent a moderate electrical stress condition for comparing the SG and DG devices under the same bias-stress environment, rather than to determine the maximum tolerable electrical stress. The resulting transfer characteristics are shown in Fig. 4(e). In this case, V_th shifted from 0.8 V to −0.55 V, corresponding to a change of 1.35 V. Because the SG device exhibited a V_th shift of 7 V under the same stress, the DG device showed an approximately five-fold smaller V_th variation. This indicates that the DG structure provides improved stability under stress conditions compared with the SG structure.

A key advantage of the DG OTFT is its voltage-varying capability, whereby the operating characteristics can be modulated by independently biasing the two gates ^[17]. To examine this behavior, one gate was held at a fixed voltage while the other gate was swept. Figs. 4(f) and 4(g) show the transfer and output characteristics obtained under top-gate voltage-sweep conditions at several fixed V_BG, whereas Figs. 4(h) and 4(i) show the corresponding results under bottom-gate voltage-sweep conditions at several fixed V_TG. In both cases, systematic changes in on-current, drain current, and V_th were observed as a function of the fixed gate bias. These results suggest that DG OTFTs are promising for flexible analog-circuit applications, where tunable device characteristics under bending conditions are required. These results indicate that the DG structure offers additional electrical tunability of the drain current and threshold voltage ^[26].

The parameter differences between SG and DG operation were compared as a function of channel length. Fig. 5(a) compares the variation in V_th. Among the three operating modes, BG operation exhibited the highest V_th, whereas DG operation yielded the lowest V_th. Fig. 5(b) compares the field-effect mobility. Consistent with the V_th trend, the lowest mobility was observed under BG operation, while DG operation achieved approximately two-fold higher mobility than TG operation. Fig. 5(c) compares the on/off ratio, showing that BG operation resulted in the lowest on/off ratio, whereas DG operation provided the highest on/off ratio

Fig. 5. Parameter comparison of SG and DG OTFT with channel length variation, and threshold voltage shift after NGBS. (a) Threshold voltage, (b) mobility, (c) on/off ratio of top single-gate (TG), bottom single-gate (BG), and double-gate (DG). (d) NGBS-induced threshold voltage shift comparing SG and DG devices.

Fig. 5(d) compares the V_th shift as a function of NGBS time for SG and DG devices. For the SG device, a rapid negative shift of approximately −10 V within 300 s, followed by a further shift to −11 V. In contrast, the DG device exhibited a modest V_th shift of about 1.7 V up to 500 s, with no additional V_th change thereafter. These results confirm that DG operation delivers better performance in terms of V_th, mobility, and on/off ratio compared with TG and BG operation, and that the DG structure provides significantly improved stability under NGBS conditions relative to the SG structure. Such performance improvements may be attributed to the enhanced channel controllability achieved through the use of both top and bottom gates. In addition, simultaneous biasing of the top and bottom gates may help balance charge trapping in the top and bottom gate insulators, which could be attributed to the improved operational stability.