Little attention has been paid so far to physiological signals for emotion recognition compared to audio-visual emotion channels, such as facial expressions or speech. In this paper, we discuss the most important stages of a fully implemented emotion recognition system including data analysis and classification. For collecting physiological signals in different affective states, we used a music induction method which elicits natural emotional reactions from the subject. Four-channel biosensors are used to obtain electromyogram, electrocardiogram, skin conductivity and respiration changes. After calculating a sufficient amount of features from the raw signals, several feature selection/reduction methods are tested to extract a new feature set consisting of the most significant features for improving classification performance. Three well-known classifiers, linear discriminant function, k-nearest neighbour and multilayer perceptron, are then used to perform supervised classification.

Small‐molecule photovoltaic cells using diindenoperylene (DIP) as a new donor material in combination with the fullerene C60 as an electron acceptor are demonstrated. In addition to the successful application in planar and bulk heterojunction devices, a comprehensive analysis including structural studies, the determination of the energy level alignment and electrical transport investigations is given, stressing the correlation between growth conditions, film morphology, and device performance. Due to pronounced crystallinity and a large surface area of DIP films grown at elevated temperature, exceptionally high fill factors of almost 75% are achieved in planar heterojunction cells. Bulk heterojunctions exhibit large‐scale phase separation forming a bicontinuous network of both molecular species, which enables efficient exciton dissociation and charge carrier transport. The high ionization potential of DIP and the favorable energy level alignment with the fullerene C60 yield large open circuit voltages close to 1 V and comparable power conversion efficiencies of about 4% in both cell architectures.

In organic photovoltaic (PV) cells, the well‐established donor‐acceptor (D/A) concept enabling photo‐induced charge transfer between two partners with suitable energy level alignment has proven extremely successful. Nevertheless, the introduction of such a heterojunction is accompanied with additional energy losses as compared to an inorganic homojunction cell, owing to the presence of a charge‐transfer (CT) state at the D/A interface. Based on the principle of detailed balance, a modified Shockley‐Queisser theory is developed including the essential effects of interfacial CT states, that allows for a quantitative assessment of the thermodynamic efficiency limits of molecular D/A solar cells. Key parameters, apart from the optical gap of the absorber material, entering the model are the energy (ECT) and relative absorption strength (αCT) of the CT state. It is demonstrated how the open‐circuit voltage (VOC) and thus the power conversion efficiency are affected by different parameter values. Furthermore, it is shown that temperature dependent device characteristics can serve to determine the CT energy, and thus the upper limit of VOC for a given D/A combination, as well as to quantify non‐radiative recombination losses. The model is applied to diindenoperylene (DIP)‐based photovoltaic devices, with open‐circuit voltages between 0.9 and 1.4 V, depending on the partner, that have recently been reported.

We have used ultraviolet and inverse photoemission spectroscopy to determine the transport gaps (E t ) of C 60 and diindenoperylene (DIP), and the photovoltaic gap (E PVG ) of five prototypical donor/ acceptor interfaces used in organic photovoltaic cells (OPVCs). The transport gap of C 60 (2.5 6 0.1) eV and DIP (2.55 6 0.1) eV at the interface is the same as in pristine films. We find nearly the same energy loss of ca 0.5 eV for all material pairs when comparing the open circuit voltage measured for corresponding OPVCs and E PVG . V C 2012 American Institute of Physics.The energy level alignment at the donor/acceptor (D/A) heterojunction of an organic photovoltaic cell (OPVC) is decisive for its performance. In particular, the energy offset between the lowest unoccupied molecular orbital level of the acceptor [LUMO(A)] and the highest occupied molecular orbital level of the donor [HOMO(D)] sets an upper limit for the open circuit voltage V oc . 1-6 This has been expressed as e Á V oc ¼ HOMO(D) À LUMO(A) À D, where e is the elementary charge and D a loss term, which has been suggested to be related to the exciton binding energy 2 or radiative and nonradiative temperature dependent losses. 1,3,5 The HOMO(D)/ LUMO(A) offset is denoted in various ways in the literature, such as charge transfer gap, intermolecular gap, or donor/ acceptor gap and is often estimated by optical spectroscopy, 5,6 or electrical characterization, e.g., cyclic voltammetry, 6 reverse saturation current analysis, 2,7 or temperature dependent measurements of the open circuit voltage. 4,5 To avoid ambiguity, we use the term photovoltaic gap (E PVG ) (cf Fig. 1). To minimize energy losses during the photon harvesting process, it is desirable to maximize E PVG within the constraint of keeping the LUMO-LUMO (DE L ) and HOMO-HOMO level offsets (DE H ) sufficiently large to drive charge separation across the D/A junction. To quantify D for unraveling its physical origin, it is mandatory to have reliable E PVG values for comparison with corresponding V oc values. Unfortunately, simple models for estimating the energy levels at organicorganic interfaces are often invalid (e.g., vacuum level alignment 8,9 ), and more involved models have been brought forward. 10,11 For the time being, experimental determination of interface energetics is indispensable to understand the processes inside an OPVC based on reliable values of E PVG , but only few pertinent studies have been conducted to date. [12][13][14][15] The present study focuses on E PVG values at prototypical organic D/A pairs formed between four organic semiconductors [sexithiophene (6T), fullerene (C 60 ), diindenoperylene (DIP, chemical structure shown in Fig. 1), and poly(3-hexylthiophene) (P3HT)] determined using the combination of ultraviolet and inverse photoemission spectroscopy (UPS and IPES). The experiment for DIP/C 60 demonstrates that E PVG can be reliably inferred from measuring the offset between the D/A HOMO levels, once the acceptor's transport gap (E t ) is known and no changes of E...

We investigate different parameters influencing the occurrence of s-shaped current voltage (j-V) characteristics in planar heterojunction organic solar cells. It is shown how substrate modification, purity of the active organic material as well as variation of the top contact can affect the shape of the j-V curves. The studies are performed on vacuum-evaporated planar heterojunction solar cells with diindenoperylene (DIP) as electron donor and fullerene C 60 as acceptor. The focus is on the fill factor and forward current being the most direct indicators for s-shapes in j-V curves. We find that the main effect of substrate heating during film growth can be assigned to changes in energy barriers rather than to the modification of morphology and crystallinity, which is also influenced by elevated substrate temperatures. The decisive role of the barrier height between the anode work function and the HOMO (i.e., highest occupied molecular orbital) level of the donor is approved by comparing hole-injection layers with different work functions. By using donor materials of different purity we find a correlation between charge carrier mobilities and fill factors. Finally, it is demonstrated that an exciton blocking interlayer is essential to get high fill factors when aluminum is used as top contact, but is dispensable for samarium as cathode material. This finding can be ascribed to the protective effect of the interlayer from aluminum diffusion into the active semiconductor rather than to its role as exciton diffusion barrier. V