Annotation of the results obtained in 2019
At the current stage of the project, work on experimental confirmation of the numerical modeling results obtained at the previous stage was carried out. An experimental setup was designed and built to measure the current of samples of organic field-effect phototransistors at different values of the gate and drain voltages in the dark and under illumination with optical radiation with different wavelengths and an inhomogeneous spatial distribution. In the basis of this setup was a probe station consisting of a movable platform on which samples of phototransistors are placed, and three probes, connected to the gate, drain and source electrodes. Last two probes and a movable platform are equipped with stepper motors that allow to position the platform in two horizontal coordinates and the probes in one vertical coordinate with an accuracy of 5 microns. The probe station is located in an airtight gas-filled glove box with an inert argon atmosphere, equipped with a gas purification system from oxygen and water vapor residues; residual oxygen concentration in the box was 2–4 ppm, water vapor concentration was less than 0.1 ppm. The probes were connected to sourcemeter instrument for measuring current-voltage characteristics, with an accuracy of 1 pcA in the range from 0 to 200 V, as well as reading current values in time steps down to 20 μs. An optical system consisting of a 10x objective (with the possibility of replacing it with a 100x objective), a beam splitter, and a CCD camera is placed above the platform with studied samples. Optical radiation was applied through beam splitter to study the photoelectric effect in the samples of phototransistors. Radiation was applied in two ways. The first method - radiation of a xenon lamp transmitted through a monochromator was transferred through an optical fiber and a system of lenses and mirrors, the wavelength could be set in the range from 350 to 900 nm with an accuracy of 1 nm; the entire sample (the entire channel of the phototransistor) was illuminated. The second method - a beam of laser radiation with a wavelength of 532 nm stretched along one coordinate was applied to the beam splitter; this method made it possible to illuminate a narrow band in the channel of a phototransistor up to 2 μm wide (using a 100x lens) to study a spatially localized photoelectric effect. To study the response times to switching on / off the incident radiation, a mechanical beam chopper was also installed in front of the beam splitter. Basing on the numerical modeling results, as well as various features of organic semiconductor materials and other factors, the following structure was chosen to create an ambipolar organic field-effect phototransistor. A doped silicon plate acts as a substrate and at the same time as a gate electrode, coated with a 200 nm thick layer of silicon oxide, which acts as a gate dielectric. On top of SiO2 another dielectric layer of a 30 nm thick PMMA polymer is placed. Due to the small thickness of the dielectric layers, a high value of the electric capacitance of about 1.7x10-4 F/m2 was achieved. Next follows the 50 nm thick active layer made of polycrystalline organic semiconducting conjugated oligomer. Totally a 7 different organic semiconductors were studied, among which the TMS-P4TP-TMS oligomer was selected. This oligomer has a low dielectric constant, small and comparable values of electron and hole mobilities of the order of 10-7 m2V–1s–1, threshold voltage values closest to zero, the smallest spread of these characteristics from sample to sample, and also the highest photocurrent among other oligomers. Above the active layer, drain and source electrodes are deposited with a special pattern, providing a channel width of 1 mm and various channel lengths from 10 to 150 microns. A calcium layer is used as the electron-injecting electrode, and MoO3 and Ag layers are used as the hole-injecting electrode. A method of laboratory samples of organic field-effect phototransistors fabrication was developed. The well-known method of manufacturing organic field-effect transistors was taken as a basis, and was refined taking into account the features of phototransistors necessary for studying a spatially localized photoelectric effect. The manufacturing process of the samples consisted of the following steps: silicon substrate preparation, active layer deposition, drain and source electrode deposition. Almost all stages of the fabrication and characterization of phototransistor samples were carried out in an inert atmosphere. The substrate preparation consisted of cleaning using ultrasonic bath and ultraviolet irradiation and applying an additional dielectric layer of PMMA polymer over a SiO2 layer. The active layer and the electrodes of the drain and source were deposited by vacuum thermal evaporation. In this case, when evaporation the electrodes, shadow masks were used, which form several samples on the same substrate with different channel lengths. The dependences of the photocurrent of phototransistors based on the TMS-P4TP-TMS oligomer were obtained when the entire channel was illuminated with radiation uniform in space, as well as radiation which intensity was not uniformly distributed along the channel, namely, when the incident radiation beam was a narrow rectangular 50 μm wide strip which illuminated part of the phototransistor channel (channel length was 140 μm). Such dependences were obtained for three different positions of the incident radiation beam in the channel of the transistor: near the hole-injecting electrode (left), in the center of the channel, and near the electron-injecting electrode (right). The obtained dependences of the normalized photocurrent Jph / Jdark on the gate voltage VG have the form of peaks which position is consistent with the spatial position of the incident beam on the phototransistor channel. When illuminating the channel region near the hole-injecting electrode, the peak shifts to the negative voltage region, which corresponds to the hole conductivity region of the transistor; illumination of the channel region near the electron-injecting electrode, results in the shift of the peak of the normalized photocurrent versus VG to the right to the region of positive voltages, which corresponds to the region of electron conductivity. However, in this case, the voltage scale for the dependences obtained in the experiment turns out to be increased, the peak widths are approximately 10 - 15 times larger than for the dependences obtained using the numerical model. This can be explained by the presence of additional (parasitic) resistances, in particular, at the contacts of the drain and source electrodes with the active layer, on which a voltage drop occurs, as a result of which the voltage in the transistor channel can be significantly lower than the applied voltage to the electrodes. In addition, it is seen that the experimental dependences of the normalized photocurrent on VG have an asymmetric shape with respect to the maximum. This is due to the inequality of electron and hole mobility. From the measured dependences of the drain current on time when the phototransistor is illuminated by incident radiation at a wavelength 532 nm interrupted with a frequency of ~ 2.5 Hz at various values of the gate voltage, the response times for switching the illumination on and off are determined. The response times for switching the lighting on and off at different VG in the range from –20 to 20 V coincide within the error range and equal 30 ± 15 ms. From the measured dependences of the current through the channel on time when switching the gate voltage from one value to another, the response time to a small VG voltage bounce from –11 to –10 V was 60 ± 26 ms, and to a large bounce of VG from –20 to –10 V the response time was 130 ± 40 ms. The response times of the phototransistor obtained in the experiment turned out to be significantly higher than the response times obtained by numerical modeling. This can be explained by large contact resistances, as well as the presence of additional parasitic electric capacitances of the drain and source electrodes. The numerical model of the phototransistor was finalized by taking into account parasitic resistances and capacitances of the drain and source electrodes. The model adequately describes the experimentally observed dependences of the normalized photocurrent on VG and relaxation times at source and drain resistances RS,D of 2.2 ± 0.7 x109 Ohms, and parasitic source and drain capacitances CSR,D of the order of ~ 10 pF. The spectra of external quantum efficiency (EQE) were measured in the wavelength range from 350 to 800 nm for an organic phototransistor based on TMS-P4TP-TMS, as well as for an organic photodiode with a vertical electrode structure based on the same organic semiconductor. It was found that the EQE of the photodiode approximately repeats in shape the absorption spectrum of the TMS-P4TP-TMS oligomer, while for the phototransistor the EQE does not decrease to zero at wavelengths longer than the edge of the absorption spectrum (~ 530 nm), but remains at the level of 3 - 6 %. This can be explained by the presence of traps at the interface between the active layer and the dielectric, which have energy levels within the band gap of the semiconductor. The results of the research are presented in the abstract of report at an international scientific conference and published in one scientific article in a journal included in the Web of Science, Scopus and RSCI databases.

 

Publications

---

Annotation of the results obtained in 2018
Within the framework of the project a one-dimensional stationary model based on a system of nonlinear differential equations has been developed, which is solved numerically after replacing the derivatives with finite differences. The model is a further development of a similar model that was used to describe a single-layer unipolar organic phototransistor [V.A. Trukhanov, at al., Proc. of SPIE 9942, 994210 (2016)]. In this project, the model was expanded by adding a more detailed description of the process of charge carriers photogeneration, as well as by taking into account the heterogeneous distribution of the incident light intensity along the x axis. The generation of free carriers under illumination occurs through the intermediate state of bound electron-hole pairs and depends on the electric field strength in the active layer. The basic equations of the model are the Poisson equation for the electric potential, the continuity equation for electron and hole current densities, and relations describing the drift and diffusion of the charge carriers. In this project, we studied the response of a phototransistor to incident radiation, which intensity distribution along the x coordinate had various forms, namely, constant, stepped, rectangular, and Gaussian. In this project, using the developed numerical model for the first time, the possibility of creating an organic phototransistor with high spatial resolution due to a spatially localized photoelectric effect has been shown [V.A. Trukhanov, JETP Letters, in press (2019)]. The distributions of potential, electric field strengths, and electron and hole concentrations in the transistor channel were calculated at various voltages at the drain VD and at the gate VG. The dependences of the electric field strength maximum position xm (which corresponds to the photosensitive region) on VG were calculated, and it was shown that the dependence xm (VG) is single-valued and monotonic. Then, for different distributions of the incident radiation intensity, the dependences of the normalized photocurrent Jph/Jdark on VG were calculated. Further, basing on the electric field maximum position xm dependence on VG, for these dependences the VG-scale was transformed into the x-scale, and it was shown that the dependences of the normalized photocurrent on VG reproduce spatial characteristics of non-uniform incident radiation with a high degree of accuracy, namely the spatial position and peak width. For the case of a Gaussian distribution along x of the incident radiation intensity, the similar results were obtained for both a small channel length of 1 μm and a larger channel length of 10 μm. The spatially localized photoeffect found in this work is nontrivial and requires further study, in particular, with the help of an experiment. Further, according to the work plan, a simple one-dimensional phototransistor model was expanded by taking into account the time dependence. For this, for all unknown functions (namely, the electric potential and electron and hole concentrations), the dependence on time t was added, and their total derivatives with respect to the coordinate were replaced by partial derivatives. The current continuity equations were also modified by adding partial time derivatives of charge carrier concentrations to them. The gate voltage VG and the bound e/h pairs generation rate G, which is proportional to the intensity of the incident radiation, are also now implied to be time-dependent. In the project, to study the speed of the phototransistor, the dependences of the photocurrent on time were calculated at instantaneous switching on or off of the incident radiation, as well as instantaneous switching of the voltage VG, that is, VG(t) and G (x, t) were described by step functions of t, and generation rate G (x, t), in addition, was described by the Gaussian function of the x coordinate. In the case of switching on and off the incident radiation, the normalized photocurrent Jph/Jdark versus time can be approximated by a biexponential dependence with characteristic relaxation times of 3 ns for a component with a large amplitude and 100 ns for a component with a small amplitude. The dependences of the normalized photocurrent of the phototransistor on time at switching the voltage VG from one value to another can be approximated by an exponential function with characteristic relaxation times of a few microseconds. Further, within the framework of the project, a search was carried out for model parameters describing the materials and structure of the phototransistor, which would provide high values of spatial resolution with acceptable values of external quantum efficiency, sensitivity, and speed. Since the optimal ratio of such characteristics as spatial resolution, external quantum efficiency, sensitivity and speed strongly depends on the specific practical application, therefore it was decided to calculate the dependencies of these characteristics on each of the parameters characterizing the materials and structure of the phototransistor. The spatial resolution of the phototransistor was characterized by the parameter wR - width at half maximum of the normalized photocurrent Jph/Jdark dependence on the voltage VG (VG determines the position of the photosensitive region along the x coordinate) when the phototransistor is illuminated with radiation which intensity is described by a rectangular function of x. An increase in the wR parameter means a deterioration in spatial resolution. The external quantum efficiency is equal to the ratio of the number of photogenerated charges to the number of incident photons and is therefore proportional to the photocurrent divided by the bound e/h pairs generation rate Jph/G0. The sensitivity of the phototransistor was determined as the ratio of the photocurrent to the dark current Jph/Jdark at the maximum point by VG. To study the response speed, the dependences of the photocurrent on time were calculated at instantaneous switching on of the phototransistor illumination at time t = 0, the intensity of which is described by the Gaussian function of the x coordinate, and characteristic relaxation times were calculated. A decrease in relaxation times corresponds to an improvement in speed. The dependences of the parameter wR, photocurrent, sensitivity, and relaxation times were calculated consequently on all parameters of the model — when one of the parameters was varied, the remaining parameters remained unchanged. Based on the obtained dependencies, conclusions are made about how various parameters affect the characteristics of the phototransistor. The change of only two parameters, namely the reduction of the channel length and the intensity of the incident radiation leads to a simultaneous improvement of all characteristics of the phototransistor. Changing the remaining parameters affects the characteristics ambiguously, for example, an increase in the dielectric constant of the semiconductor constituting the active layer leads to a significant increase in EQE and sensitivity, but it greatly reduces the speed and spatial resolution. A phototransistor with the parameters of its materials and structure, typical for organic field-effect transistors, but with lowered dielectric constant and increased electric capacity, can have a high spatial resolution, in particular, with an accuracy of tens of nanometers, reproduces the spatial position of a Gaussian or rectangular beam of incident radiation. Also, a phototransistor with such parameters has low, but acceptable from the point of view of measurement values of photocurrent, external quantum efficiency, sensitivity. In addition, a phototransistor with such parameters has a high speed with characteristic relaxation times less than a few microseconds. Therefore, the numerical values of the parameters used in this work can serve as the first approximation for development of the first experimental samples of phototransistors for studying the spatially localized photoelectric effect in them. In addition, within the framework of the project, calculations of two-dimensional and three-dimensional distributions of the electric field, electric potential, electron and hole concentrations in the bulk of the active layer of an organic field phototransistor using a technological computer-aided design (TCAD) system are carried out. A new type of phototransistor with two additional electrodes is also being investigated in which the presence of a spatially localized photoeffect in two coordinates is possible. A background was prepared for the experimental part of the project: a preliminary selection of organic semiconductors was carried out, they were purified, and samples of ambipolar organic field-effect transistors were fabricated. The results of the research are presented in the abstracts of three reports at two international scientific conferences and published in one article in the journal included in the Web of Science, Scopus and RISC databeses.

 

Publications

---