Today, electronics are implemented on rigid substrates. However, many objects in daily-life are not rigid - they are bendable, stretchable and even foldable. Examples are paper, tapes, our body, our skin and textiles. Until today there is a big gap between electronics and bendable daily-life items. Concerning this matter, the DFG Priority Program FFlexCom aims at paving the way for a novel research area: Wireless communication systems fully integrated on an ultra-thin, bendable and flexible piece of plastic or paper. The Program encompasses 13 projects led by 25 professors. By flexibility we refer to mechanical flexibility, which can come in flavors of bendability, foldability and, stretchability. In the last years the speed of flexible devices has massively been improved. However, to enable functional flexible systems and operation frequencies up to the sub-GHz range, the speed of flexible devices must still be increased by several orders of magnitude requiring novel system and circuit architectures, component concepts, technologies and materials.

This paper presents analysis of the charge storage behavior in organic thin-film transistors (OTFTs) by means of admittance characterization, compact modeling, and 2-D device simulation. The measurements are performed for frequencies ranging from 100 Hz to 1 MHz and bias potentials from zero to -3 V on top-contact OTFTs that employ air-stable and high-mobility dinaphtho-thieno-thiophene as the organic semiconductor. It is demonstrated that the dependence of the intrinsic OTFT gate-source and gate-drain capacitances on the applied voltages agrees very well with Meyer's capacitance model. Furthermore, the impact of parasitic elements, including fringe current and contact impedance, is investigated. The parameters used for the simulation and modeling of all the dynamic characteristics correspond closely to those extracted from static measurements. Finally, the implications of the admittance measurements are also discussed relating to the OTFTs dynamic performance, particularly the cutoff frequency and the charge response time.

Analytical expressions for the gate-voltage dependence of the channel capacitance and the gate-to-contacts overlap capacitances in top-contact organic thin-film transistors (OTFTs) are derived and implemented in an organic compact capacitance model. The resulting modified model is verified by experimental data of transistors with constant mobility. The same model is analyzed by numerical simulations for OTFTs with a voltage-dependent mobility. The simulation results indicate that the quasistatic model describes well the simulated capacitances. In accumulation, the modeled values are slightly overestimated because of the generally accepted assumption of the charge-sheet model. It is also demonstrated that the quasistatic regime occurs at lower frequencies because of the reduced mobility at lower charge carrier concentrations.

Transport states in disordered organic semiconductors are widely accepted to be Gaussian distributed with a rather narrow variance (width) of the order of roughly 100–150 meV. This is a result of intensive analysis of experimental data and theoretical work using different model assumptions on hopping transport. For this transport mechanism, the dependencies of the carrier mobility on temperature, carrier density, and field are described by analytical approximations, which can be implemented in advanced device simulation programs. However, also the carrier density itself is determined by the Gaussian density of states (GDOS) and the energy integral over the product of the GDOS and the Fermi distribution function can only be calculated numerically. But meanwhile the published [G. Paasch and S. Scheinert, J. Appl. Phys. 107, 104501 (2010)] analytical approximation for the charge carrier density of organics with GDOS is implemented in the Sentaurus Device simulator. We present here results of detailed numerical simulations of organic field-effect transistors applying this model. It turns out that an important parameter is the distance of the GDOS maximum from the band edge, which is not directly accessible from experiment since the measured DOS shows a much larger band width (the narrow transport GDOS might be the tail). Only for a special case, depending on the variance, the deviations from a non-degenerate inorganic semiconductor are rather small. Furthermore, the results are affected by the properties of the source/drain contacts (Ohmic or Schottky type). Inclusion of both the calculation of the carrier concentration assuming the GDOS and the dependence of the mobility on this concentration shows that the influence of the latter is much larger. Nevertheless, the deviations from simulations for an inorganic semiconductor show that the carrier density for the GDOS should be taken into account in device simulations.

We have investigated the bias stress stability of poly(3-hexylthiophene) (P3HT)-based organic field-effect transistors (OFETs), prepared from chloroform or trichlorobenzene solutions on a thin SiO2 dielectric (bottom gate) with an additional self-assembled monolayer. Three different processes of gate oxide treatment were compared with regard to their reaction to prolonged constant gate bias. Furthermore, we analyzed the transconductance gm during gate bias stress, as well as the contact resistance Rc using the transmission line method and found that the bias stress has no influence on the contact resistance of P3HT-based OFETs, but with the addition that Rc strongly affects gm. The most stable characteristics were achieved for transistors with a dielectric coated with hexamethyldisilazane over 22 h, which exhibited a pronounced reduction of the threshold voltage shift and contact resistance in comparison with the other variants of gate dielectric treatment. Based on two-dimensional simulations, we demonstrate that an increased Rc in combination with a carrier concentration dependent, high mobility leads to a maximum in the transconductance curve, which can be prevented by optimizing the contact between source/drain and the channel region.

Carrier injection barriers determined by photoemission spectroscopy for organic/metal interfaces are widely accepted to determine the performance of organic field-effect transistors (OFET), which strongly depends on this interface at the source/drain contacts. This assumption is checked here in detail, and a more sophisticated connection is presented. According to the preparation process described in our recently published article [S. Scheinert, J. Appl. Phys. 111, 064502 (2012)], we prepared PCBM/Au and PCBM/Al samples to characterize the interface by photoemission and electrical measurements of PCBM based OFETs with bottom and top (TOC) contacts, respectively. The larger drain currents for TOC OFETs indicate the presence of Schottky contacts at source/drain for both metals. The hole injection barrier as determined by photoemission is 1.8 eV for both Al and Au. Therefore, the electron injection barriers are also the same. In contrast, the drain currents are orders of magnitude larger for the transistors with the Al contacts than for those with the Au contacts. We show that indeed the injection is determined by two other properties measured also by photoemission, the (reduced) work functions, and the interface dipoles, which have different sign for each contact material. In addition, we demonstrate by core-level and valence band photoemission that the deposition of gold as top contact onto PCBM results in the growth of small gold clusters. With increasing gold coverage, the clusters grow inside and begin to form a metallic, but not uniform, closed film onto PCBM.

The operation of organic devices as organic field-effect transistors (OFET) depends critically on the contact between the organic layer and the material for source/drain electrodes. Small barriers for carrier injection are required for efficient operation. In order to support the understanding of organic devices, photoemission spectroscopy has been used to determine the properties of metal/organic interfaces. Values for the hole injection barrier determined in the last decade by different groups are frequently of the order of 0.5–1 eV. It is not clear whether barrier lowering due to the image charge is sufficient to make contacts with such barriers efficient for carrier injection. Indeed, no results have been reported where the preparation of the samples for the photoemission study and for the devices are the same. Here we present results of such an investigation for OFETs with gold source/drain contacts. The measured hole barrier at the gold contact of 0.6 eV results from the Au work function of 4.6 eV. Taking into account the dependencies of the mobility on the carrier concentration and on the field for the Gaussian density of states (DOS) of disordered organics, measured OFET current characteristics cannot be described well with such contacts but rather for work functions of 4.7 eV or larger. Considering the method in determining the barrier from photoemission data and the Gaussian DOS of the hopping transport states, we present a quantitative connection between the barrier as determined from photoemission and the barrier as used in the device simulation.

are reported for the first time. The external quan tum efficiency peak values achieve up to 42% at ~ 350-400 nm and 26% at ~640 nm (PM-19:[60]-PCBM ). PACS. 73.61Ph Conducting polymers thin films - 84.60.Jt Photovoltaic conversion - 76.30.-v Organic materials EPR spectra 1 Introduction Solar cells based on conjugated polymers can be pro cessed from solution or dispersion which offers a v ery important technological potential for low-cost fabrication us ing high-volume processes like reel to reel technol ogies. One of the main reasons why the efficiency is still limited to 4-5% [1-5] with the standard system poly(3-hexylth iophene) (P3HT):[60]-PCBM in single polymer solar cells (on glass substrates) is the low absorbance of the two components of the photoactive layer in the range of the solar spe ctrum combined with the necessity of a thin film th ickness of this layer ( 2%) [10] mostly linked to problems wi th the charge transport. Here we report about the optical, electr ical and photovoltaic properties of a thieno[3.4- b]pyrazine phenylenevinylene thiophene (PM-18) and of thieno[3 .4-b]pyrazine phenylenevinylene copolymers (PM-19, PM-2 0) related to the homopolymeric standard poly(phenylen evinylene) (PPV) (MDMO-PPV). 2 Results and discussion The well-defined strictly alternating thieno[3.4- b]pyrazine-based copolymers were synthesized by Horn er polycondensation route [11-12]. Figure 1 shows the chemical structures of the photoactive donor materi als and the homopolymeric PPV reference (MDMO-PPV). PM-20 and P M-19 differ only in one side chain (2-ethylhexyloxy vs. n-octyloxy). In PM-18 one main chain PPV ring of the monomer unit is replaced by a 3.4-dihexylthiophene moiety. [6.6]-Phenyl-C

Active layers in organic devices prepared in solution based preparation routes are usually disordered. Their highest occupied molecular orbitals and lowest unoccupied molecular orbitals, or the valence and conduction band states, show an energetic distribution which can be approximated by a Gaussian density of states (DOS). The resulting dependency of the (electron and hole) mobility on temperature, carrier density, and field can be easily implemented into advanced device simulation programs. However, in addition the charge carrier density is needed as the integral over the DOS multiplied with the Fermi-Dirac distribution. We denote this normalized quantity as the Gauss-Fermi integral. Since it cannot be evaluated analytically, similarly as in the case of the Fermi-Dirac integral F(1/2), an analytical approximation is needed for efficient device simulation. In the present article, such an approximation is proposed with different expressions in the nondegenerate and degenerate regions with a continuous and differentiable transition between both regions. The approximation is also applicable to traps with a Gaussian DOS. (C) 2010 American Institute of Physics. [doi:10.1063/1.3374475]

The optical gaps of the low-bandgap PPVs (PM-20, PM-19, PM-18) are decreased down to 1.6-1.7 eV compared with that of MDMO-PPV (2.2 eV). The best lateral hole mobility was determined to be 2.1 x 10(-3) cm(2)/Vs (PM-18) in field effect transistors and exceeds that of MDMO-PPV (poly-[2-methoxy- 5-(3'.7'-dimethyloctyloxy)-1.4-phenylenevinylene], 8.5 x 10(-4) cm(2)/Vs). This allows to reduce the PCBM ([6.6]-phenyl-C-61(71)-butanoic acid methyl ester) content in solar cell devices down to 1:2 w/w giving a better eta AM1.5 than for MDMO-PPV:[60]-PCBM cells (PM-19:[60]-PCBM 2.32% on ITO-PET, 2.86% on ITO glass). The charge transfer to PCBM as acceptor occurs quite normally and shows an effective charge separation using light-induced spin resonance spectroscopy (LESR). The [70]-PCBM-center dot signals are shifted to lower field related to those of [60]-PCBM-center dot and overlap more with the polaron signal of PM-19. The LESR g-factor components of [70]-PCBM-center dot are reported for the first time. The external quantum efficiency peak values achieve up to 42% at similar to 350-400 nm and 26% at similar to 640 nm (PM-19:[60]-PCBM).

Low-cost fabrication of circuits with organic field-effect transistors (OFETs) as basic devices requires solution-based technologies. However, then the carrier mobility values can hardly exceed 0.01-0.1 cm(2) V-1 s(-1). For a cut-off frequency above 100 kHz to 1 MHz, relevant for broader applications, and an operation voltage below 10 V, the channel length of the transistors should be smaller than 1-10 mu m. Considering inevitable parasitic capacitances one must envisage submicrometer channel lengths. Demonstrations of different patterning have so far realized neither resolution nor alignment accuracy desired, and photolithography is too costly for the submicrometer regime. Here, an overview is given of short-channel OFETs based on the definition of the submicrometer structures by undercutting, and preparation of the devices using additional simple steps of well established microelectronics technology without any of the costly steps such as high temperature treatments, high resolution lithography, and ion implantation.

Analytical approximations for space-charge-limited currents (SCLCs) in systems with exponential or Gaussian trap distributions were widely used in analyzing organic diodes. The current follows a power law with a transition into the trap-free SCLC at high voltages and an Ohmic low voltage limit. The power coefficient γ is connected with either the decay constant or the variance of the distributions. Within these formulations, it is not possible to check the relevance of the numerous approximations needed to derive them. This concerns especially the relations of the contact work functions and of the layer thickness with the trap concentration, the position of the center of the trap distribution and its maximum value. Application of the analytical approximations to results of full numerical simulations allows one to set limits for the parameter ranges in which the approximations can be applied. In the case of the exponential distribution the analytical approximation is rather good for high trap concentrations and thicker layers. However, the simulations reveal a number of additional peculiarities. Such, the high voltage limit is usually not the trap-free SCLC but Ohmic and determined only by the anode barrier, the low voltage limit leads to a diodelike dependence with a large ideality factor and scaling with layer thickness and position of the trap distribution is extremely limited. In the case of the Gaussian trap distribution the simulations show indeed that the formula together with the connection between the power coefficient and the variance of the distribution fails completely. Thus, in principle, earlier analyzes of experimental data should be revised by using numerical simulations.

The current characteristics of organic field-effect transistors (OFET) often show a disadvantageous nonlinearity at low drain voltages. It has been shown recently [J. Appl. Phys. 102, 054509 (2007)] that in top contact (TOC) OFETs this effect can be caused by trap recharging if the contacts are of Schottky type. For bottom contact (BOC) OFETs, in spite of controversial discussions, Schottky contacts as origin of the nonlinearity are often stated. At first, it is shown here by a mixed mode simulation that for large ideality factors a Schottky contact only at drain leads to such a nonlinearity. However, with the same Schottky contacts at drain and source the effect is covered by the high resistance of the contact at source. Next, the different influences of Schottky contacts on BOC OFETs and TOC OFETs with varying overlap of the source/drain contacts with the gate are clarified. Further, it is demonstrated with detailed two-dimensional simulations that the combination of the presence of Schottky contacts with a field dependence of the mobility can cause the nonlinearity. For the mobility we use the field dependent Pool/Frenkel model, and the models of Limketai et al.[Phys. Rev. B 75, 113203 (2007)] and Pasveer/Coehoorn et al.[Phys. Rev. Lett. 94, 206601 (2005)], which depend in addition on the carrier concentration. Their influence on the device performance has been clarified by the simulations. Simulated profiles of concentrations and fields lead to the understanding of the mechanism causing the nonlinearity. This mechanism is especially effective for the Pasveer/Coehoorn model. The field dependence of the mobility is a consequence of the energetic distribution of the hopping states and can hardly be avoided in solution based deposition of the active polymer layer. A strategy to prevent the nonlinearity is therefore an optimization of the contact-polymer interface such that the contacts become Ohmic.

Gas sensors with small dimensions offer the advantage of electrical sensitivity modulation. However, their actual use is hindered by drift effects that exceed those of usual metal-oxide sensors. We analyzed possible causes and found the best agreement of experimental data with the model of internal dopant fluctuations. The dopants are oxygen vacancies exhibiting high drift-diffusion coefficients under the impact of electrical fields. Thus, the width parameters of space charge regions, which again control the sensor current, are undergoing slow changes. Moreover, the dopant distributions cause internal electrical fields that yield drift even after voltage switch-off. This behavior has been proven by simulations based on the literature values, using a converging combination of the classical electron drift-diffusion and Poisson equations with the Fokker-Planck solution for the dopants, which is of general relevance to other nonperfect semiconductor devices.

We report on the improved assembly and characterization of a small molecule organic field-effect transistor (OFET). Novel alpha,omega-dicyano substituted beta,beta'-dibutylquaterthiophene molecules (DCNDBQT) were synthesized and characterized by UV-Vis spectroscopy, differential scanning calorimetry, thermal gravimetric analysis and cyclic voltammetry. The ultra-thin organic film formation on TiO2 templates was effectively promoted through the specifically designed bifunctional self assembly molecules (SAM) 5-cyano-2-(butyl-4-phosphonic acid)-3-butylthiophene (CNBTPA). Excellent structural properties were found for up to 9 DCNDBQT molecule thick films prepared through UHV vacuum sublimation as investigated with UHV non-contact atomic force microscopy (nc-AFM) and X-ray diffraction. Both X-ray and nc-AFM data indicate that the DCNDBQT molecules form a well-ordered terraced structure exhibiting step heights of 1.5 nm to 2.0 nm layers. Hence, the DCNDBQTmolecules are linked to the functional SAM interface layer by H-bond inter-actions (see structure model) standing quasi perpendicular to the TiO2 template, and thus providing optimal orbital overlap neighbouring thiophene rings.The vacuum sublimated DCNDBQT molecules form a closed packed and dense molecular layer that was used to construct and operate a nanoscopic OFET-structure. The resulting field mobilities of 10(-5) cm(2) V-1 s(-1) reflect a high current density in our ultrathin but highly ordered structure.

Hysteresis effects do occur usually in polymer field‐effect devices. It has been suggested that trap recharging or mobile ions, or formation/dissociation of bipolarons (BP) in the accumulation layer can cause this effect. Here, at first a literature survey on the hysteresis in field‐effect transistors is given. Then examples of measured hysteresis in field‐effect transistors and in metal‐insulator‐semiconductor (MIS) capacitors are presented. The typical peculiarity is the dependency of the flat band voltage on the sweep direction of the gate voltage. Our recent numerical simulations on the trap recharging mechanism in MIS capacitors are continued and extended to transistors. Energetically distributed traps can lead to hysteresis. But the form of the hysteresis deviates from the observed one and extreme parameter values are needed. Thus it is more likely that trap recharging can modify a hysteresis caused by another mechanism. For discussing the bipolaron mechanism the equilibrium between polarons and doubly charged states of the polymer chains was analyzed anew. With a rate constant for the bipolaron formation determined recently by Salleo and Street relaxation times for formation/dissociation are determined. They indicate that these processes can cause the hysteresis. A further possible mechanism influencing the hysteresis is connected with complexes between polarons, bipolarons and mobile counter ions. The estimated rate constants indicate processes on the time scale of the measurements. Numerical simulations of these processes as well as of the role of mobile ions and their reactions are required for a full description of the hysteresis. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Space charge layers (SCLs) in metal-insulator-semiconductor (MIS) structures are critical for the operation of field-effect transistors (FETs). For many organic semiconductors, transport takes place as hopping in Gaussian or exponentially distributed states. However, existing theoretical descriptions of a SCL and advanced device simulation programs suppose a density of states other than a Gaussian or an exponential, employing often the nondegenerate limit for the concentrations. We present results of a simulation study for the MIS structure as the basic module of the FET and for a thin semiconducting layer on a metal substrate. The second system was extensively investigated by photoelectron spectroscopy to characterize the metal-organics interface occurring at the source/drain contact of FETs and as anode and cathode in organic light emitting diodes. For broader distributions, the densities deviate strongly from the nondegenerate limit which leads indeed in a MIS structure to a strong deviation of the dependence of the surface electric field (and, hence, the areal charge) on the surface potential. However, as one can control only the gate voltage directly, the dependency on this quantity determines device operation. For the variations of the layer thickness and gate insulator thickness, and doping in the wide range of interest, this dependency deviates only slightly from the nondegenerate approximation, essentially in the depletion region by a flatband voltage shift. In the accumulation region, which is determinative for FET operation, the remaining deviation can be removed almost perfectly by considering this flatband voltage shift. For the thin organic layer on a metal substrate, numerical simulations confirm the applicability of an analytical approximation for band bending and floating potential [G. Paasch et al., J. Appl. Phys. 93, 6084 (2003)] for the nondegenerate case and for the exponential distribution. Indeed, for small barriers at the interface, a band bending of up to the order of 100 meV can occur within the first 2 nm near the interface. In the interpretation of photoemission data such contribution will appear as part of the measured interface dipole.