(Invited) Disentangling the Effects of Gated-Contacts on Transconductance in Organic FETs

Wednesday, 27 May 2015: 14:30
Conference Room 4L (Hilton Chicago)
E. Bittle (NIST, Wake Forest University), J. I. Basham (NIST, Penn State), T. N. Jackson (Penn State), O. D. Jurchescu (Wake Forest University), and D. J. Gundlach (NIST)
Recent reports of solution processed organic thin film transistors (TFTs) with exceedingly high field-effect mobility have spurred renewed interest in organic TFTs because of their potential to impact higher performance electronic applications than previously envisioned, while also stimulating considerable debate around the transport physics allowing such high charge carrier mobility in these intriguing material systems.  Most reports parameterizing TFT performance make use of DC electrical characterization methods and mathematical relationships that describe transistor current as a function of applied voltage for ideal devices.  However, the electrical characteristics reported for many of the highest mobility TFTs exhibit substantial non-ideal behavior.  Thus, a detailed investigation of organic transistors with such behavior is needed to discern the transistor channel properties from parasitic device effects that might affect the device operation and give rise to incorrect mobility estimates.  We report here on the use of impedance spectroscopy to disentangle the effects of contacts on device operation and extract the electrical properties of the channel.  Specifically, we look at the small signal AC impedance of rubrene single crystal transistors to investigate "ideal" low field electronic behavior.  The rubrene crystals are grown by vapor transport method and laminated onto prefabricated transistor test structures consisting of a heavily doped silicon substrate (gate electrode), thermally-grown silicon dioxide layer (gate dielectric), and photolithographically-defined metal electrodes (source and drain contacts).  The completed rubrene transistors have channel lengths ranging from 50 µm to 100 µm.  The impedance response is measured as a function of gate bias and frequency (varied from 20 Hz to 2 MHz).  By using a transmission line model to fit the transistor channel coupled with a parallel resistor-capacitor model of the contact impedance, we are able to observe the behavior of the transistor channel and contacts separately as a function of gate bias. We determine the low field mobility of the device independent of contact resistance and show that rapidly changing contact resistance governs the current flow at low gate voltage in DC current-voltage measurements.  This effect gives rise to large transconductance near threshold that reflects the “turn-on” of the gated-contacts rather than solely channel response making mobility extracted by using traditional DC current-voltage relationships describing only the channel inappropriate in this bias region.