Impedance spectroscopy (IS) is a powerful tool for analyzing the electrical properties of devices, especially for capacitance-voltage analyses, which have become standard methods to investigate device characteristics. Recently, IS has also been extensively applied to organic devices, such as organic light-emitting diodes [1] - [5] , organic photovoltaic (OPV) cells [6] [7] and organic thin-film transistors [8] [9] , revealing the electrical properties including the carrier injection [1] [2] and charge transport properties [3] - [5] .

Along with IS measurements, other methods have been used to measure the mobility in highly resistive organic semiconductor films and other materials including the time-of-flight (TOF) method [10] [11] , the dark injection (DI) transient method [12] - [15] , a method to obtain the mobility from the I-V characteristics of the space charge limited current (SCLC), time-resolved electroluminescence (EL) spectroscopy [16] and the photoexcited time-resolved microwave conductivity technique [17] [18] . To accurately evaluate the transit time using the TOF and DI methods, a film that is substantially thicker than the light-emitting layer in an organic EL device is required. Because the electronic and optical properties of an organic semiconductor largely depend on the film thickness [19] , measurements should be performed on a film thickness which is equal to that in an actual device. In the SCLC method, measurements using the actual device structure are difficult and the electric-field dependence of the mobility cannot be determined. Time-resolved EL spectroscopy and photoexcited time-resolved microwave conductivity are methods used to evaluate the mobility under conditions that differ from the actual applied voltage. In contrast, IS measurements that enable the mobility to be measured in actual OPV devices and the electric-field dependency of the mobility can be evaluated. However, the photoconductivity in the low-vol- tage region, where an actual OPV device operates, has not been discussed. In all cases, the evaluation of the mobility under high-electric-field conditions, such as in the SCLC region, has been the main focus of discussion.

In this study, in-situ IS measurements of the photoconductivity of the carriers induced by photoirradiation in the layers in organic hetero-junction structured photovoltaic devices were performed. The IS measurements were carried out by changing the externally applied voltage and the intensity of the light irradiating the device. To establish a basic method, using the dependence of the frequency characteristics of the relaxation time on the light intensity, a change in the photo conductivity that originated from the carriers induced by photoirradiation in each layer was examined using a donor/acceptor hetero-structured device.

Figure 1(a) shows the configuration of the donor/acceptor heterostructured OPV device used in this study, the generation of an exciton accompanied by photoirradiation, and the process of generating holes and electrons at the interface between the donor and acceptor layers.

Figure 1(b) shows the dependency of the J-V characteristics on the light intensity. Under 1 sun of irradiance,

the device characteristics were: open circuit voltage, Voc = 0.48 V, space charge current density, Jsc = 3.4 mA cm⁻², fill factor, FF = 0.46, and the power conversion efficiency, PCE = 0.75%. The threshold voltage was determined to be about 0.5 V from the J-V characteristics (this is the voltage at which a large number of charges begin to be injected from the electrodes). The voltage range in which the device generates power as an OPV device, that is, the region where it operates, was 0 - 0.5 V. This voltage range was smaller than the bias voltage at which the mobility and conductivities are evaluated by using conventional methods. This demonstrates the importance and necessity to evaluate the conductivity and carrier mobility within the actual driving-voltage region of an OPV, which is the purpose of this study.

Figure 2 shows Bode modulus plots for each bias voltage under 1 sun of irradiance, that is, the characteristics of the imaginary part of the modulus (Im M) versus frequency.

A change in the resistance component was observed because M ~ iωZ, where ω is the angular frequency and i is the imaginary part.

There are two components, each with a relaxation time that responds to the frequency. The time constants for these two components did not change in the region from 0 to 0.4 V, which is equal to or below the threshold voltage. The conductivity within this voltage region depended on the frequency response.

From the formulas:

where τ is the relaxation time, and C = εS/d

where S is the area and d is the thickness of the layer

The peak frequency was derived from the following equation:

where σ is the conductivity, ρ is the electrical resistivity, ε is the permittivity.

In the Bode plots, there was a proportional relationship between f_peak and σ.

Figure 3(a) shows the dependency of the Bode plots on the light intensity at a bias voltage of 0.3 V. The peak frequencies of the two components shifted, which is in agreement with the changes in the photo-induced carriers that were generated according to the light intensity.

Figure 3(b) shows plots of the peak frequency versus light intensity. The changes in the slopes for the two components were the same, which indicates that the main change was a shift in the conductivity of each layer,

caused by holes and electrons being generated in equal amounts under light irradiation. The components of the relaxation time were identified from the dependency of the film thickness for each layer. The component in the low-frequency region was found to be CuPc and that in the high-frequency side was found to be C60. Given that the ratio of the peak frequency was a single digit, the ratio between the carrier mobility of the CuPc layer and the C60 layer was calculated to be a single digit using the formula:

where μ is the carrier mobility, n is the carrier density and e is the elementary charge.

Figure 4 shows the σ_AC values (AC electrical conductivity) at bias voltages from 0 to 0.5 V for an OPV device under 1 sun of irradiance. Using σ_AC = d/AZ, the conductivity of the entire device was determined to be 1.0 µS/m at 100 Hz. The carrier mobility µ was calculated to be on the order of 10⁻³ cm²/Vs using n = 10¹⁶ m⁻³. Therefore, because the mobility ratio between CuPc and C60 was only a single digit, the mobility of CuPc was calculated to be in the order of 10⁻⁴ cm²/Vs, and that of C60 was in the order of 10⁻³ cm²/Vs. These values agree well with the previously reported results evaluating the mobility of each material.

In-situ impedance spectroscopy (IS) observations of the photoconductivity of the carriers induced by photoirradiation in organic hetero-junction structured photovoltaic devices are presented. Analyzing the IS data showed the existence of two components with different relaxation times, which originated from the donor and acceptor

materials. The peak frequency of these components shifted as a function of the light intensity and they had the same slope. The conductivity originated from the holes and electrons that were photo-induced by photoirradiation was used to calculate the mobility ratio. The mobilities of the CuPc and C60 layers were calculated from the conductivities and were in the orders of 10⁻⁴ and 10⁻³ cm²/Vs, respectively. For the first time, in-situ measurements of the photoconductivity were performed by using an actual device in the low-voltage region in which OPV devices are operated. These results serve as a guideline for the future development and improvement of organic semiconductor devices based on multilayer structures.