A Pathway to Enable Efficient Performance in Organic Field-Effect Transistors with Low-Cost, Scalable Contacts

The potential impact of organic semiconductors is vast, and for that impact to be felt they need to be incorporated into real-world applications such as OLEDs, sensors, and transistors. Many of these devices are held back by deficiencies in device design, where performance is limited by charge injection and transport across material interfaces. Having to contend with a nearly infinite selection of materials and geometries makes improvement a difficult task. To streamline the search through this vast parameter space, in this work we incorporated simulations into our experimental process in a feedback loop to explore the relationship between the injection barrier, trap density of states, dielectric capacitance, and device architecture in organic field-effect transistors (OFETs) in pursuit of a design window where select parameters become non-critical. As a result, we established a device configuration window where cost and complexity of processing can be greatly reduced without a steep reduction in performance. Our simulations revealed that the dielectric capacitance (<i>c_i</i>) plays an important role in determining the device mobility, and critically, has an effect that differs depending on the OFET structure. As such, increasing the dielectric capacitance yields a device that is more tolerant of trap states as long as the injection barrier is not very large; however, when an injection barrier is significant, lowering <i>c_i</i> yields devices that are able to function effectively with even large barriers (greater than 0.5 eV). Additionally, while for the case of high-quality contacts there is no notable difference, our results show that the staggered OFET geometry outperforms the coplanar geometry in the case of non-ideal contacts. To experimentally verify these trends, we fabricated staggered-structure OFETs with small and large injection barriers and with a range of values for <i>c_i</i>. For the case of a small injection barrier (nominally 0.0 eV), devices performed similarly when the dielectric capacitance was varied, as predicted by simulation. Mobility for the low-<i>c_i</i> devices based on indacenodithiophene-<i>co</i>-benzothiadiazole (C₁₆IDT-BT) was measured to be 1.6 ± 0.3 cm²/Vs (average obtained on over 100 devices), overlapping with the performance of the high-<i>c_i</i> devices at 1.5 ± 0.1 cm²/Vs. We then purposefully introduced an injection barrier of 0.4 eV and found that the mobility of the low-<i>c_i</i>devices was still 80% that of the small-barrier devices, compared to a 50% decrease for the high-<i>c_i</i> devices. Analyzing the contact and channel resistances and corresponding simulation data, it was discovered that the voltage drop across the source contact is more severe with a higher capacitance dielectric, leaving a reduced electric field strength in the channel and correspondingly lower device mobility. On the contrary, lowering the dielectric capacitance from 37 nF/cm² to 1.5 nF/cm²minimizes the voltage loss at the contacts, resulting in a channel mobility that is insensitive to injection barriers. We took advantage of this rule to improve the performance of carbon-based and organic contacts, which typically have the advantage of low-cost processability, but have work functions that introduce a large injection barrier. We selected a readily available graphite spray to pattern contacts that, while cheap and compatible with large-scale processing, come with the penalty of a nearly 1 eV injection barrier. By lowering the dielectric capacitance from 10 nF/cm² to 1.6 nF/cm², a one order of magnitude increase in mobility was observed. In summary, the low-<i>c_i</i> design offers a pathway to optimize charge injection and transport in OFETs while enabling the selection of cost-effective, scalable contact materials.