Table of contents

Surface texturing is an imperative process to reduce the reflection of the incident light on solar cells, by enhancing sunlight diffusion into the silicon solar cells for photon generation. As a result, the current generation can be increased. In this study, the plasma texturing process with linear microwave plasma sources has been benchmarked with the industrial acidic iso-texturing process on 156×156 mm² multicrystalline substrates. By optimizing the plasma texturing parameters, the absolute solar cells efficiency can be increased by 4.9% for 150 µm thickness silicon substrate. The proposed process offers a significant advantage over the standard acidic iso-texturing without major modification in the existing industrial solar cells manufacturing sequence. In order to explain plasma-induced surface morphology changes, the Kardar–Parisi–Zhang (KPZ), Pétri–Brault, and Jason–Drotar models are used.

Electroluminescence (EL) under forward bias represents the total excess minority carrier density in cells. In contrast, EL under reverse bias can be detected as hot spots, which are closely related to harmful current shunt paths. In this study, we detected the shunt position using two kinds of EL. Additionally, we analyzed by the positions and origins of shunt sources using electron-beam-induced-current, lock-in thermography, and an electron-probe-micro analyser. We found two kinds of shunt and we detected a defect located in the depletion layer. We proposed shunt models in the depletion layer using the band model.

In this work we present a cell process for amorphous crystalline silicon heterojunction (SHJ) solar cells based on process steps well known in the photovoltaic industry. All amorphous silicon layers are deposited by plasma enhanced chemical vapor deposition (PECVD) in a one chamber direct plasma reactor working at a radio frequency of 13.56 MHz. The main focus of this work is to study the influence of p- and n-doped Czochralski (Cz) silicon base material with different surface morphology on the cell results of amorphous crystalline SHJ solar cells with intrinsic thin layers. Open circuit voltages V_oc of up to 700 mV are obtained on n-type Cz based SHJ cells (area 100 cm²) with rough surfaces. On p-type Cz based SHJ cells open circuit voltages were limited by the minority carrier bulk lifetime of the used base material.

In this work, we present the results of the replacement of silver screen printing on heterojunction crystalline silicon (c-Si) solar cells with a copper metallization scheme that has the potential to reduce the manufacturing cost while improving their performance. We report for the first time silver-free heterojunction c-Si solar cells on 6-in. wafers. The conversion efficiency reached is a record 22.1% for c-Si technology for this wafer size (V_oc = 729 mV, J_sc = 38.3 mA/cm², FF= 79.1%). The total power generated is more than 5 W for 1-sun illumination, which is a world record. Heat-damp reliability tests show comparable performance for mini-modules fabricated with copper metalized as for conventional silver screen printed heterojunction c-Si solar cells.

In producing the Si heterojunction interdigitated backcontact solar cells, we investigated the feasibility of applying amorphous Si emitter having considerable crystalline Si phase at the facing to transparent conducting oxide (TCO) layer. Prior to evaluating electrical property, we characterized material nature of hydrogenated microcrystalline p-type silicon (µc-p-Si:H) as crystallized fraction, surface morphology, bonding kinds in thin films and then surface passivation quality finally. The diode and interface contact characteristics were induced by the simple test device and then current–voltage (I–V) curve showed more linearity in µc/hydrogenated amorphous silicon (a-Si:H) emitter case. We fabricated heterojunction back contact (HBC) solar cells using p/n interdigitated structure and acquired the 23.4% efficiency in cell size with performance parameters as open-circuit voltage (V_oc) 723 mV, short-circuit current density (J_sc) 41.8 mA/cm², fill factor (FF) 0.774, in the cell size (at 2×2 cm²).

Modelling of solar cells today is general practice in research and widely-used in industry. Established modelling software is typically limited to one dimension and/or to small scales. Additionally, novel effects, like, e.g., the use of diffractive structures or luminescent materials, are not established. In this paper we discuss how the combination of different modelling techniques can be used to overcome these limitations. In this context two examples are presented. The first example concerns the combination of the open source simulation software PC1D with circuit modelling to investigate the effect of local shunts on the global characteristics of a silicon wafer solar cell. For the investigated example (4.5 cm² cell area) we find that a local point shunt reduces the solar cell efficiency by 4% relative. The second example concerns the modelling of diffractive gratings for thin silicon wafer solar cells. For this purpose, we use the rigorous coupled wave analysis to simulate Sentaurus technical computer-aided design (TCAD) is combined with the rigorous coupled wave analysis, a method to solve Maxwell's equations for periodic structures. Here we show that a grating can be used to improve the absorption in a thin silicon wafer solar cell considerably.

We present a numerical simulation study on the optimization of locally contacted rear surface passivated p-Si solar cells considering float-zone and Czochralski-grown bulk material, latter by taking different amounts of oxygen concentrations into account and, thus, varying the quality of the bulk material. The conversion efficiency potential is figured out by a broad variation of the nominal base resistivity, thickness and rear contact distance of the solar cell. To focus on the bulk and rear side recombination properties, the front side contact and emitter properties have been idealized to avoid recombination losses in this region. It turns out that high level injection effects playing a major role in the description of high resistivity bulk materials.

In this work we report a world record independently confirmed efficiency of 19.4% for a large area p-type Czochralski grown solar cell fabricated with a full area aluminium back surface field. This is achieved using the laser doped selective emitter solar cell technology on an industrial screen print production line with the addition of laser doping and light induced plating equipment. The use of a modified diffusion process is explored in which the emitter is diffused to a sheet resistance of 90 Ω/□ and subsequent etch back of the emitter to 120 Ω/□. This results in a lower surface concentration of phosphorus compared to that of emitters diffused directly to 120 Ω/□. This modified diffusion process subsequently reduces the conductivity of the surface in relation to that of the heavily diffused laser doped contacts and avoids parasitic plating, resulting an average absolute increase in efficiency of 0.4% compared to cells fabricated without an emitter etch back process.

Effects of surface temperature on high-rate etching of Si by narrow-gap microwave hydrogen plasma have been investigated. The etch rate strongly depended on the surface temperature. The optimum temperature for the etching at high rate was about 70 °C. With increasing the temperature higher than 70 °C, decrease in etch rate was observed, and activation energy for the reaction process was estimated to be about -1.8 kcal/mol. This value agreed well with the result of the previous studies using low-pressure plasma conditions. In addition, hydrogen concentration was also affected by the surface temperature. The reduction in hydrogen concentration near the surface was observed in the etched sample at higher temperature. From these results, reaction mechanism causing the decrease in etch rate with the temperature increase has been discussed in terms of both desorption and in-diffusion of hydrogen from the surface.

The computational fluid dynamics-based FLUENT program was employed to model the heat transfer and chemical reaction in a mono-silane Siemens reactor. The kinetic parameters for the 1-step overall reaction SiH₄→Si+ 2H₂, such as the pre-exponential factor, temperature coefficient, and activation energy, were carefully optimized to satisfy experimental data obtained from the 4-rod Siemens pilot reactor. Established models were successfully used to evaluate the effects of rod diameter, reaction temperature, and reactant gas flow rate on the deposition rate of silicon.

In this paper, we show that cell efficiency can be improved by simple preclean steps for silicon substrates. A very low reflectance of less than 1% is observed in the wavelength range from 450 to 700 nm for the silicon substrates precleaned using a NaOH solution followed by a gettering process. Lifetime measurements show that the silicon substrates treated by these preclean processes have a significantly improved minority carrier lifetime owing to a better surface condition than untreated silicon substrates. This result is further supported by the performance of the solar cells. A cell efficiency of 18.609% is achieved in the treated silicon substrates with a gain of 2.14% compared with the untreated silicon substrates.

The soldering quality of a Cu ribbon onto the Ag busbars of solar cells is one of the key factors affecting the cell-to-module loss and durability of a silicon photovoltaic module. In this study, we examined the effects of the surface chemistry and morphology of the screen-printed Ag busbars on the solder wettability and bonding uniformity of the Cu ribbon over the length of the busbar during the tabbing process. The surface of the as-fired Ag busbar was covered by a thin glass layer originating from the glass frit contained in the Ag paste. The presence of the thin glass layer significantly decreased the wettability of the solder, leading to the formation of voided regions at the solder/busbar interface. Although it had only a minor influence on solder wettability, increasing the roughness of the busbar surface resulted in the formation of more voided regions at the solder/busbar interface.

We sometimes observe poor internal quantum efficiency (IQE) in the long-wavelength region of p-type single-crystalline silicon (c-Si) solar cells having back-surface-field (BSF) layers made of p-type hydrogenated amorphous silicon (a-Si:H). We found that this poor IQE is caused by the appearance of a polycrystalline phase (PP) in a-Si:H. We showed that two-step growth of BSF structures with hydrogen plasma treatment (HPT) is effective for achieving a high performance of BSF with improved reproducibility. First, a-Si:H layers were prepared on c-Si substrates at a high growth rate to suppress the appearance of PP, and consequently, the defects at the a-Si:H/c-Si interface induced by the higher growth rate were effectively passivated by HPT. Finally, a-Si:H was deposited as a capping layer on the first a-Si:H layer after HPT. We will discuss the passivation mechanism of interface defects with HPT using the data of the hydrogen profile in the a-Si:H layers treated by hydrogen plasma.

To obtain a lower reflectance, we applied a maskless plasma texturing technique by reactive ion etching (RIE) to acid-textured multicrystalline silicon (mc-Si) wafer. RIE texturing produced a deep and narrow textured surface with an excellent low reflectance. Owing to plasma-induced damage, unless the RIE-textured surface is subjected to proper damage removal etching (DRE), it shows drops in open circuit voltage (V_oc) and fill factor (FF). RIE-textured samples with proper DRE showed sufficiently higher short circuit current (I_sc) than acid-textured samples without a drop in V_oc. In this study, we applied RIE texturing under optimized DRE condition to the selective emitter structure. In comparison with the acid-textured solar cells, RIE-textured solar cells have absolute gains in I_sc above 200 mA. We successfully fabricated a 6-in. mc-Si solar cell with a conversion efficiency exceeding 18% by applying selective emitter technology with RIE texturing.

Laser doping (LD) can help form emitters at low cost and in high throughput because it operates in air atmosphere and at room temperature. Moreover, LD makes high-efficiency structures in a very simple way by melting and diffusing impurities on selective areas. In this study, evaluation of the relationship between multicrystalline silicon (mc-Si) surface conditions and electric properties is carried out. The results show that the surface electric properties were improved and conversion efficiencies were increased. This was accomplished by changing the surface conditions from rough to smooth by surface etching and vacuum treatment before LD.

The effects of the low-temperature annealing on Zn-doped indium–tin-oxide (ITO) films such as the electrical, optical and structural properties were investigated. Zn-doped ITO films were fabricated by rf magnetron sputtering of ITO and Al-doped ZnO (AZO) targets on corning glass at room temperature. The content of Zn increased with increasing the power of AZO target. The carrier concentration of films shows the decreasing behaviour with increasing the content of Zn, due to a carrier compensation originating from the substitution of a doped Zn for an In or interstitial site. After the low-temperature annealing at 180 °C in vacuum, all films were slightly decreased a carrier concentration and increased the hall mobility because of the absorption of oxygen on the surface films. In addition, the average transmittance did not show a considerable change and had a high values over 80%. Especially, the Zn-doped ITO with atomic ratio of Zn/(In+Zn) of 6.8 at. % had the resistivity of 4×10^-4 Ω cm, the highest hall mobility of 41 cm² V^-1 s^-1, and the average transmittance of 82%.

Excellent surface passivation of boron emitters is demonstrated for industrial plasma-enhanced chemical vapor deposited (PECVD) SiO_x/AlO_x stacks. Emitter saturation current densities of 39 and 34 fA/cm², respectively, were achieved at 300 K on 80 Ω/sq boron emitters after activation by (i) a standard industrial firing process and (ii) a forming gas anneal followed by industrial firing. We find that the surface passivation by SiO_x/AlO_x stack can be effectively controlled by varying the SiO_x layer thickness. This stack is directly applicable to certain high-efficiency solar cell structures, by optimising the SiO_x thickness accordingly.

Owing to the volatilization of isopropanol (IPA), instability in the alkaline texturization of monocrystalline silicon has been a big problem for a long time. Many additives were adapted to replace IPA, such as high boiling point alcohols. In this experiment, as a new attempt, sodium lauryl sulfate (SDS), a type of anionic surfactant, was used as the additive in NaOH solution. The etching properties of silicon in 2 wt % NaOH/15–30 mg/L SDS solution were analyzed. To improve the wettability of silicon, two types of metal salt, NaCl and Na₂CO₃ with concentration from 2 to 15 wt %, were applied to the 2 wt % NaOH/15 mg/L SDS solution. The results showed that the effect of NaCl was better than that of Na₂CO₃. Finally, the role of the additive was discussed.

The mapping characterization for the structural and optoelectronic properties of textured SnO₂:F transparent conductive oxide (TCO) layers has been performed by spectroscopic ellipsometry (SE). From the SE analysis of the free carrier absorption in the SnO₂:F layer, the optical carrier concentration and mobility are extracted by using the Drude model. As a result, in the textured SnO₂:F substrate with a size of 9×9 cm², we have confirmed slight non-uniformities in the carrier concentration as well as the layer thickness. Moreover, in order to investigate the effect of the TCO inhomogeneity on hydrogenated amorphous silicon (a-Si:H) and microcrystalline silicon (µc-Si:H) tandem solar cells, we have simulated short-circuit current density (J_sc) in the a-Si:H and µc-Si:H layers by taking the TCO inhomogeneity into account using the optical admittance method. From this procedure, we have quantitatively estimated the variation of J_sc by the TCO inhomogeneity assuming a non-textured flat structure.

Kaneka has been proposing and developing the "advanced super-light trapping (ASLT)" structure for thin film silicon solar cells, which incorporates tailored light confinement structures in tandem thin film silicon solar cells to selectively enhance light trapping in top and bottom subcells. In this paper, we present the results of the development of intermediate reflectors with very low refractive index and the design and implementation of nanoimprinted substrates. We demonstrate that significant current gains are possible by reducing the refractive index of the interlayer beyond state-of-the-art values. In addition, we show that the transparent conductive oxide angular scattering properties correlate with the solar cell performance in the infrared part of the spectrum.

In this paper, we report newly developed high-haze glass/zinc oxide (ZnO) substrates with low resistivity by a combination of unique etched soda-lime glass and bilayered ZnO films with a Zn-rich (oxygen-poor) layer. The high mobility and low resistivity of bilayered ZnO films could be obtained with Zn-rich conditions. By depositing the ZnO films onto textured glass substrates, the obtained films exhibit an excellent light-scattering property, while their electrical property is still good. Furthermore, the bilayered ZnO films with a Zn-rich layer did not negatively affect the transparency of the films. Employing the bilayered ZnO films with a Zn-rich layer and an rms roughness of about 274 nm as the front transparent conductive oxide (TCO) in hydrogenated amorphous silicon (a-Si:H) solar cells, we improved the performance and quantum efficiency (QE) of the fabricated solar cells, particularly in the short-wavelength region without the deterioration of open-circuit voltage or fill factor. Thus, the developed glass/bilayered ZnO film with a Zn-rich layer is a new promising material since its resistivity is low while its light-scattering property is still high.

We have applied a triode electrode configuration in the plasma-enhanced chemical vapor deposition (PECVD) process to grow intrinsic hydrogenated amorphous silicon (a-Si:H) light absorbers for the fabrication of p–i–n junction solar cells. Although the deposition rate is lower (0.1–0.3 Å/s) than that of the conventional diode PECVD process, the light-soaking stability of the solar cell is markedly improved and less sensitive to the cell thickness due to the reduced Si–H₂ bond density in the a-Si:H i-layer. The a-Si:H single-junction solar cells exhibit low light-induced degradation of conversion efficiency (Δη/η_ini∼10%) in comparison with that of high-efficiency solar cells reported to date. By applying the improved a-Si:H layers as top-cell absorbers in a-Si:H/hydrogenated microcrystalline silicon (µc-Si:H) tandem device, the light-induced degradation can be reduced even further (Δη/η_ini\lesssim5%). As a result, we obtain confirmed stabilized efficiencies of 9.6 and 11.3% for a-Si:H single-junction and a-Si:H/µc-Si:H tandem solar cells, respectively.

We have investigated the fundamental optoelectronic properties of newly developed transparent conductive oxide (TCO) materials, e.g., titanium-doped indium oxide (InTiO). InTiO films, being deposited at 50 °C by the RF-magnetron-sputtering method followed by thermal annealing at 200 °C, show excellent optoelectronic properties for solar-cell application. We have demonstrated the improved photovoltaic performance of n–i–p microcrystalline-silicon (µc-Si:H) solar cells whose i layer is prepared at a high rate of 2.3 nm/s using a stacked structure of InTiO with aluminum-doped zinc oxide (AZO) as top (front) TCO layers.

The domain structure of BaSi₂ epitaxial films grown on vicinal Si(111) substrates has been studied in order to fabricate high-quality BaSi₂ crystals with large domains. The X-ray pole figure measurement shows that the BaSi₂ films grown on vicinal substrates as well as the on-axis substrate consist of three epitaxial variants which are equivalent in terms of 60° in-plane rotations, and that one of the variants is dominant in the film grown on the 2°-inclined substrate. The orientation maps produced by electron backscatter diffraction show that the domains with the b axis parallel to the miscut direction are larger than the others in the film grown on the 2°-inclined substrate, while the domain sizes of three variants are found similar in the films grown on the on-axis and 4°-inclined substrates. The possible origin of the large domain formation is discussed with the focus on the initial growth stage observed by atomic force microscopy. Nucleation from the step edge is proposed as the mechanism of the large-domain formation considering the lattice matching to the step edge, while nucleation is suggested to occur at the terrace edges on the 4°-inclined substrate.

In this study, a flattened light-scattering substrate (FLiSS) with a large refractive index contrast in plane is investigated as an approach for overcoming the detrimental effect of highly textured substrates for obtaining better light management in thin-film silicon solar cells. A FLiSS composed of a two-dimensional ZnO grating and a Ag reflector realize a high open circuit voltage in substrate-type microcrystalline silicon solar cells irrespective of the grating structure, with an enhanced infrared response. The optimum FLiSS structure is discussed with the help of numerical simulation.

Texturing the glass surface is a promising method for improving the light trapping properties of superstrate thin-film silicon solar cells, as it enables thinner absorber layers and, possibly, higher cell efficiencies. In this paper we present the optical and morphological properties of borosilicate glass superstrates textured with the aluminium induced texture (AIT) method. High haze values are achieved without any reduction in the total optical transmission of the glass sheets after the AIT process. Scanning electron microscope and atomic force microscope (AFM) measurements reveal a laterally uniform surface morphology of the AIT texture. We demonstrate that the surface roughness and thus the transmission haze can be controlled by adjusting the AIT process parameters. From the AFM images, we extract histograms of the local height and angle distributions of the texture. Samples with a wide angle distribution are shown to produce the highest optical haze. The results of this analysis provide a better understanding of the correlation between the AIT process parameters and the resulting surface morphology. This analysis is further extended to an amorphous silicon pin solar cell deposited onto the textured glass substrate.

We have successfully developed novel aluminum-doped zinc oxide (AZO-X) films with a high haze ratio by the combined use of an etched glass substrate and wet-etched AZO-X films. The effects of the use of an etched glass substrate and wet-chemical etching on the properties of AZO-X films were investigated. The texture size and rms roughness of these films largely increased with glass surface roughening. Post-treatment using wet chemical etching slightly increased the texture size and rms roughness. The etched glass approach has been found to be a promising method for achieving an AZO-coated glass substrate with a high haze ratio. Using high-haze ratio AZO-X films as the front transparent conductive oxide (TCO) layers in solar cells, we improved the quantum efficiency (QE) of these solar cells particularly in the long-wavelength region. Thus, the AZO-X films deposited on etched glass have a high potential for use as front TCO layers in silicon-based thin-film solar cells.

TiO₂ films were prepared by RF magnetron sputtering with process variations of substrate temperature and oxygen dilution ratio to investigate the effect of the optoelectronic properties of TiO₂ films on antireflection characteristics in solar cells. With an increase in substrate temperature from RT to 350 °C, the conductivity and refractive index of TiO₂ films increased. However, the absorption coefficient also increased. In the case of oxygen-diluted sputtering (0.5–1%), the conductivity and refractive index decreased with the increase in oxygen dilution. On the other hand, the absorption coefficient decreased simultaneously. To evaluate these optoelectronic property variations of TiO₂ films in terms of the antireflection effect in solar cells, amorphous silicon (a-Si:H)/amorphous silicon germanium (a-SiGe:H) tandem solar cells with various optoelectronic properties and thicknesses (20–40 nm) of TiO₂ films were fabricated. The TiO₂ film deposited at 350 °C and 0.5% oxygen dilution showed a high conductivity (∼10^-3 Ω^-1 cm^-1) and refractive index (∼2.56 at 550 nm). The fabricated tandem cell with the TiO₂ antireflection layer showed an efficiency of 11.22%, whereas the reference cell without TiO₂ exhibited an efficiency of 10.97%.

Wide-bandgap, high-quality p-type microcrystalline silicon carbide (p-µc-SiC) films have been prepared by radio frequency plasma-enhanced chemical vapor deposition (RF-PECVD) technique for use as window layers for single- and triple-junction thin-film silicon solar cells. We have found that the p-µc-SiC films have wider optical bandgaps and lower absorption spectra in the short-wavelength region than the conventional p-µc-Si films. The new p-type doping layer was applied as a window layer of a microcrystalline silicon (µc-Si:H) single-junction cell, and the thin-film solar cell with the new window layer showed higher open circuit voltage (V_oc) and conversion efficiency. In addition, the insertion of an optimized p/i buffer layer was essential for reducing atomic damage at the p/i interface and obtaining a higher conversion efficiency. The optimized p-µc-SiC layer and p/i buffer layer were adopted successfully as a new window layer for the bottom cell within the triple-junction cell structure.

The aluminium-doped zinc oxide (ZnO:Al) films grown by sputtering method were etched to improve the light scattering property. The high haze value (diffuse transmission to total transmission) of above 40% at 850 nm wavelength was obtained by the increase of etching time. But the resistance of film increased and a lot of pin holes were created due to the over etch for high haze. In order to solve these problems, the additional ZnO layer was deposited on etched ZnO:Al film without sacrifice of high haze. This method was able to compensate the deteriorated properties without the change of optical properties. Amorphous based silicon solar cells showed the improvement of photovoltaic performances by the additional deposition.

We have developed novel aluminum-doped zinc oxide films (AZO-X and AZO-HX films) with a high haze value using wet-chemical etching for various times after dc magnetron sputtering, and have investigated their electrical and optical properties, durability under high-humidity condition, and surface morphology. The AZO-X and AZO-HX films showed good balance between transmittance in the near-infrared area and durability under 85 °C–85%RH condition. These novel films also had a higher haze value after wet chemical etching than normal AZO films. The crater size and haze value of the AZO-HX film increased with increasing etching time in comparison with those of the AZO-X film. The haze value of the AZO-HX film was higher than that of the AZO-X film; their values are 90% at 550 nm and 60% at 800 nm. Furthermore, the AZO-HX film was applied in amorphous silicon (a-Si) single-type solar cells as the front electrode. The short-circuit current of the solar cell using the AZO-HX film was higher than that of the solar cell using the AZO-X film. As an optimization-based result, an efficiency as high as 10.2% was obtained, showing that the new AZO-HX film is a promising material for the front electrode of a-Si solar cells.

We prepared and applied a-SiO_x thin films to hydrogenated microcrystalline silicon solar cells (µc-Si:H) as a front antireflection layer (FAL) in order to reduce optical reflection loss. By inserting the optimized SiO_x FAL with a refractive index of ∼1.75 into the glass/ZnO interface a relative increase in short-circuit current density (J_sc) by 5% could be obtained which corresponded to an improved spectral response in the 550–950 nm wavelength regions. In addition, this optimized FAL did not deteriorate the properties of the ZnO layer because no significant changes in open-circuit voltage (V_oc) and fill factor (FF) were observed. As a result, the cell with an efficiency of as high as 8.28% (V_oc=0.495 V, J_sc=25.09 mA/cm², FF=0.667) could be obtained.

The flash lamp annealing (FLA) of electron-beam- (EB-) evaporated amorphous silicon (a-Si) films results in the formation of polycrystalline Si (poly-Si) films with at least a few µm long grains stretching along lateral crystallization directions. Unlike the case of using chemical-vapor-deposited (CVD) hydrogenated a-Si films as precursors, no peeling of Si films occurs even in the absence of Cr adhesion layers. Such a flash-lamp-induced crystallization occurs also in doped EB-evaporated a-Si films as in the case of undoped films. The p⁺/p^-/n⁺ stacked structure is sufficiently kept even after crystallization, although the profiles of dopants are slightly modified. This fact clearly indicates that the crystallization observed is not based on liquid-phase epitaxy (LPE) after the complete melting of the whole a-Si precursor during millisecond-order treatment but through LPE-based explosive crystallization (EC), self-catalytic lateral crystallization driven by the release of latent heat. The formation of poly-Si films with large grains and the sufficient preservation of dopant profiles would lead to the utilization of the poly-Si films formed for solar cell devices.

Hydrogenated amorphous silicon–germanium (a-SiGe:H) solar cells are fabricated with different thicknesses of the i/n graded layer and profiling shapes for appropriate band gap profiling. Comparison of the solar parameters between the U-shape profiling and the exponential shape (E-shape) profiling has been carried out at the same total thickness. In the U-shape profiling, as the thickness of the i/n graded layer increase, the fill factor (FF) and open circuit voltage (V_oc) of p–i–n single-junction a-SiGe:H solar cells increase, but the short circuit current (J_sc) of cells decreases. In the E-shape profiling, the J_sc of the a-SiGe:H cell is enhanced without significant losses in V_oc. For further analysis, a modified E-shape profiling is incorporated in a-Si:H/a-SiGe:H double-junction cells, which has resulted in the improvement of V_oc and FF of double-junction cells to 1.67 V and 0.753, respectively, without significant reduction in J_scSiGe^QE, 12.58 mA/cm².

The objective of this paper is to review current status and future prospect on CuInSe₂ (CIS)-based thin-film photovoltaic (PV) technology. In CIS-based thin-film PV technology, total-area cell efficiency in a small-area (i.e., smaller than 1 cm²) solar cell with top grids has been over 20%, while aperture-area efficiency in a large-area (i.e., larger than 800 cm² as definition) monolithic module is approaching to an 18% milestone. However, most of the companies with CIS-based thin-film PV technology still stay at a production research stage, except Solar Frontier K.K. In July, 2011, Solar Frontier has joined the gigawatt (GW) group by starting up their third facility with a 0.9-GW/year production capacity. They are keeping the closest position to pass a 16% module-efficiency border by transferring the developed technologies in the R&D and accelerating the preparation for the future based on the concept of a product life-cycle management.

Copper indium gallium diselenide (CIGSe) solar cells grown on glass substrates have reached an efficiency of 20.3%. Their industrial production is becoming increasingly relevant. While various deposition techniques for the fabrication of the absorber are used by different groups and corporations, molybdenum (Mo) has become the back contact material of choice. Oxidation of the bare Mo layer prior to absorber deposition is a phenomenon that is generally hard to control or to avoid. Since the incorporation of sodium (Na) into the absorber layer is commonly achieved by diffusion from a glass substrate through the Mo layer, oxidation of the back contact will influence the diffusion, and thus the availability of Na during the CIGSe growth process. In order to investigate this effect, Na containing glass substrates with Mo layers in different stages of oxidation have been prepared using a damp heat treatment. The samples were coated with CIGSe by physical vapor deposition in a multistage co-evaporation process. The CIGSe/Mo-interface is investigated by Raman spectroscopy and secondary ion mass spectroscopy, using a lift-off technique. The damp heat treatment led to the formation of an oxide layer (presumably MoO₂) and an increase of the sodium content in the grown absorber layers.

The efficiency of Cu(InGa)Se₂ (CIGS) solar cells with high Ga content fabricated by the three-stage method is lower than that with low Ga content in spite of a better matching solar spectrum. Secondary ion mass spectrometry (SIMS) measurement revealed that the band profile of CIGS films with high Ga content had a deep notch around 0.5 µm from the CdS/CIGS interface. In order to decrease the notch depth of the CIGS with high Ga content, the five-stage method was employed instead of the conventional three-stage method. As a result, we successfully obtained the efficiency of 14.9% using the CIGS absorber with an average band gap of 1.40 eV prepared by the five-stage method. Theoretical simulation revealed the effects of the notch location and depth on solar cell performance characteristics.

We formed Cu(In,Ga)(S,Se)₂ films using a three-stage process. The grain size of the films was reduced by increasing their S concentration. Judging from the phase diagrams of Cu–S and Cu–Se, a decrease in grain size with the increase in S concentration may be due to the higher the melting point of Cu₂S compared to that of Cu₂Se. The short circuit current density of the fabricated solar cells decreased significantly with an increase in S concentration. Secondary ion mass spectrometry showed an unsuitable compositional distribution of S and Se despite the constant S/Se flux ratio during growth. A maximum open circuit voltage of 0.83 V was obtained at a S/(S+Se) ratio of 0.80 in the fabricated solar cells in order to increase bandgap.

A modified three-stage method was introduced to deposit high-efficiency Ag(In,Ga)Se₂ (AIGS) solar cells. Cu(In,Ga)Se₂ (CIGS) and AIGS films were deposited by conventional and modified three-stage methods, respectively, to enable a comparison of the diffusion processes of Ag and Cu into (In,Ga)₂Se₃ films during the second stage. The diffusion coefficient of Ag is less than that of Cu, causing the easy segregation of Ag in the film. High-temperature annealing could improve the diffusion coefficient of Ag, consequently improving the solar cell performance. Air annealing of the AIGS solar cell can also improve the cell performance significantly, and finally an efficiency of 9.2% was achieved.

We have successfully prepared ZnCuInS₂ (Zn_2xCu_1-xIn_1-xS₂, ZCIS) thin films by spray pyrolysis deposition (SPD). The bandgap of the ZCIS thin film was widely controlled from 1.4 to 3.4 eV by substituting Zn for Cu and In of CuInS₂ (CIS). The resistivity of the ZCIS film was controlled by adjusting deposition temperature and composition ratio. ZCIS solar cells with a structure of glass/indium tin oxide (ITO)/TiO₂/In₂S₃/ZCIS/Au were fabricated. The cell with a bandgap of 1.8 eV showed an efficiency of 4.4%. However, the average V_oc is much lower than what is theoretically possible for absorbers with the bandgap. The secondary ion mass spectroscopy (SIMS) depth profile showed that a narrow bandgap layer, corresponding to a low Zn/(Zn+Cu+In) ratio, was formed at the interface between the buffer and the absorber by interdiffusion. The low V_oc is attributed to the existence of the narrow bandgap layer.

Cu(In,Ga)Se₂ (CIGSe) solar cells were fabricated independently by industrial scale co-evaporation in two separate production lines with the same nominal composition and thickness of the absorber film. Although the device properties were believed to be the same we observed substantial deviations of the respective values of the open circuit voltage (ΔV_OC = 40 mV) and of the fill factor (ΔFF= 4%), whereas the short circuit current was essentially the same. We performed fundamental device analysis, space charge and defect spectroscopy, transient photoluminescence as well as in-depth profiling of the chemical gradients of the absorber films. Using the results from the experiments we set up a simulation baseline which allowed us to conclude that the apparent deviations can be related to the presence of deep recombination centers with different concentration within the CIGSe absorber as well as to variations of the band gap grading.

New π-conjugated polymers containing difluorodioxocyclopentene-annelated thiophenes as an electron-accepting unit have been synthesized for application to p-type organic semiconducting materials in organic photovoltaics. The photophysical and electrochemical properties of these polymers were investigated by UV–vis absorption spectra and cyclic voltammetry measurements. The carrier mobility measurement of the copolymer on organic field-effect transistor devices revealed p-channel behavior. Bulk heterojunction photovoltaic devices made from blends of the polymer and [6,6]-phenyl-C₇₁-butyric acid methyl ester showed moderate photovoltaic characteristics with a power conversion efficiency of 0.77%.

High-efficiency cadmium-free Cu(In,Ga)Se₂ (CIGS) thin-film solar cells have been fabricated using a zinc compound buffer layer deposited by the chemical bath deposition (CBD) process. However, the zinc compound buffer layers such as ZnS(O,OH) are prone to plasma-induced damage during the subsequent ZnO sputtering process. A process causing less damage such as metal–organic chemical vapor deposition (MOCVD) is thus required for ZnO-based transparent conducting oxide (TCO) layers. In the present work, the boron-doped zinc oxide (ZnO:B) films were grown by MOCVD using diethyl zinc (DEZ), H₂O, and low-toxicity triethylboron (TEB). An UV-assisted MOCVD process was developed in order to reduce the resistivity of ZnO:B films. As a result, the resisitivity significantly decreased together with the increased electron mobility and carrier concentration. The comparison of performance was also carried out for the ZnS(O,OH)/CIGS solar cells with MOCVD-deposited ZnO:B and sputter-deposited ZnO:Al window layers.

CuInS₂ (CIS) solar cells with Zn(O,S) window layers deposited by co-sputtering were fabricated to optimize the conduction band offset (CBO) of Zn(O,S)/CIS. Zn(O,S) has an advantage in the control of the CBO because the conduction band minimum can be controlled by S/(S+O) compositional ratios. The efficiency of the CIS solar cells with the Zn(O,S) window layers increased with increasing S/(S+O) ratio from 0.19 to 0.50. However, the efficiency decreased at the S/(S+O) ratio of 0.59 owing to the decrease in the fill factor because of the double-diode-like behavior. The maximum efficiency was obtained at the S/(S+O) ratio of 0.50. Also, the CIS solar cell with standard bi-layer buffers of sputtered ZnO and chemical-bath-deposited CdS was fabricated and the performance was compared.

To quantitatively evaluate the substitution energies of Cd atom for Cu, Zn, or Sn atom in indium-free photovoltaic semiconductors Cu₂ZnSnS₄ (CZTS) and Cu₂ZnSnSe₄ (CZTSe), first-principles pseudopotential calculations using plane-wave basis functions were performed. The substitution energies of Cd atom in kesterite-type CZTS and CZTSe were calculated in consideration of the atomic chemical potentials of the constituent elements of Cu, Zn, Sn, and the doping atom of Cd. During the chemical bath deposition (CBD) of the CdS layer on the CZTS or CZTSe layer, Cu, Zn, and Cd atoms dissolved in the ammonia aqueous solution and formed [Cu(NH₃)₂]⁺, [Zn(NH₃)₄]²⁺, and [Cd(NH₃)₄]²⁺ complex ions. Therefore, the chemical potentials of Cu, Zn, and Cd atoms in [Cu(NH₃)₂]⁺, [Zn(NH₃)₄]²⁺, and [Cd(NH₃)₄]²⁺ complex ions were calculated. We found that the substitution energies of n-type Cd_Cu and charge-neutral Cd_Zn in CZTS and CZTSe are smaller than that of p-type Cd_Sn. The substitution energies of Cd_Cu in CZTS and CZTSe are smaller than that in chalcopyrite-type CuInSe₂ (CIS). However, the substitution energies of Cd_Cu, Cd_Zn, and Cd_Sn are positive values. The formation energy of charge-neutral Cd doping with the Cu vacancy (Cd_Cu + V_Cu) pair is a negative value and greatly smaller than those of donor-type Cd_Cu and neutral Cd_Zn in CZTS and CZTSe. These results indicate that the charge-neutral (Cd_Cu + V_Cu) vacancy pair is easily formed during the CBD of the CdS layer on the CZTS or CZTSe layer. A small amount of n-type Cd_Cu and neutral Cd_Zn would also be formed.

The effects of antimony (Sb) doping of the CdTe layer in the CdTe solar cells were investigated using Sb-doped CdTe powders as source materials for CdTe deposition by the close-spaced sublimation (CSS) method. Conversion efficiency increased with increasing Sb concentration below 1×10¹⁸ cm^-3, mainly owing to the improvement of the fill factor. Secondary ion microprobe mass spectrometry (SIMS) depth profile revealed that the Sb impurities at a concentration of approximately 1×10¹⁶ cm^-3 were incorporated into the CdTe layer when using the Sb-doped CdTe source of 1×10¹⁸ cm^-3. The observation of surface morphology showed that the grain sizes were improved by Sb addition. Therefore, the improved performance upon Sb addition to CdTe solar cells was probably due to the improvements in crystallinity, such as increased grain size.

Photoluminescence (PL), PL intensity mapping and time-resolved PL (TR-PL) studies have been applied to the Cu(In,Ga)Se₂ (CIGS) solar cell fabrication process. Measurements have been done just after the respective cell process for the preparation of the Al/ZnO:Al/ZnO/CdS/CIGS structure, in which CdS has been formed by the chemical-bath deposition (CBD) while undoped and Al-doped ZnO layers were deposited by RF magnetron sputtering. PL intensity does not change by depositions of CdS and undoped ZnO buffer layers. PL intensity decreases by the deposition of the ZnO:Al film due to the cell shunt at the edge. The electrical cell isolation by the mechanical scribing leads to the increase in PL intensity because of the formation of the hetero-junction under the open circuit condition. The decay curves of the as deposited CIGS film, CdS/CIGS and ZnO/CdS/CIGS are non-exponential and composed of dominant fast decay and weak slow decay components. After the ZnO:Al deposition, PL decay is represented by the single exponential curve with long decay time. They are discussed in terms of the junction formation. PL intensity mapping after cell processes has been correlated with the solar cell performance.

In this study, an optical simulation of Cu(In,Ga)Se₂ (CIGS) solar cells by the rigorous coupled-wave analysis (RCWA) method is carried out to investigate the effects of surface morphology on the light absorption and power conversion efficiencies. Various sub-wavelength grating (SWG) nanostructures of periodic ZnO:Al (AZO) on CIGS solar cells were discussed in detail. SWG nanostructures were used as efficient antireflection layers. From the simulation results, AZO structures with nipple arrays effectively suppress the Fresnel reflection compared with nanorod- and cone-shaped AZO structures. The optimized reflectance decreased from 8.44 to 3.02% and the efficiency increased from 14.92 to 16.11% accordingly. The remarkable enhancement in light harvesting is attributed to the gradient refractive index profile between the AZO nanostructures and air.

Wide-gap Cu(In_0.4,Ga_0.6)Se₂ solar cells with Zn(O,S) buffer layers deposited by atomic layer deposition (ALD) technique have been investigated. The band-gap energy (E_g) of the Zn(O,S) layer estimated by optical transmission and reflection measurements was varied from 3.2 to 3.6 eV. The solar cells with sulfur (S)-poor Zn(O,S) buffer layers showed a low open-circuit voltage (V_OC) owing to the cliff nature of the conduction band offset (CBO). In contrast, the solar cells with S-rich Zn(O,S) buffer layers showed a low short-circuit current density (J_SC) owing to the spike nature of CBO. Even if the CBO values were adequate, the best solar cell efficiencies were considerably low. These results suggest that the main cause for the low efficiencies is not interface recombination at the Zn(O,S)/Cu(In,Ga)Se₂ interface, but mainly bulk recombination in the Cu(In,Ga)Se₂ (CIGS) absorber layer.

Al-doped ZnO (AZO) was deposited on an undoped ZnO buffer layer at various deposition temperatures by RF-magnetron sputtering. We controlled the crystal c-axis orientation of AZO using the undoped ZnO buffer layer. With increasing deposition temperature of the undoped ZnO buffer layer, the alignment of the c-axis orientation of the AZO layer improved, and the sheet resistance of the AZO layer decreased. Comparing the electrical properties of the AZO layer with the c-axis orientation, the above correlation was confirmed. It is important to improve the alignment of the c-axis orientation for the improvement of the electrical properties of the AZO layer.

The influence of background pressure in multisource evaporation is not straightforward to explain, but it is clear that the amount of absorbed gases on the surface of the growing grains change with pressure. This may influence surface energies, and consequently the growth mechanism. We investigated the relation between the orientation of Cu(In,Ga)S₂ films and pressure during deposition. We varied the background pressure during the deposition of the precursor film (first stage) by varying the temperature of the sulphur source, by throttling the vacuum pump, or by introducing N₂ gas. We found that Cu(In,Ga)S₂ films prepared from In–Ga–S precursors show (112) orientation if the pressure during precursor deposition is less than 0.03 Pa when the pressure is mainly attributable to sulphur. When it is mainly due to N₂ gas, the pressure at which the orientation changes is 0.14 Pa. The orientation of the final film reflects the orientation of the (In,Ga)₂S₃ precursor immediately before entering the second stage.

Thin films of CuIn_1-xAl_xSe₂ (CIAS) were prepared by stacked elemental precursor layers in an inert ambient. The stacking sequence of precursor layers may affect the kinetics of phase formation. The soda lime glass (SLG)/Cu/Al/In/Se sample heated at 750 °C for 30 s with a temperature ramp rate of 15 °C/s may react to form single-phase CIAS thin films with a chemical composition fairly close to the predetermined value. Transmission electron microscopy (TEM) analysis of a selenized film prepared at 600 °C revealed the segregation of the content of Al to the substrate side and a relatively large variation of Al distributed in a lateral direction as compared with those of other elements. Changing the Cu and Al layer sequence may affect the reaction paths and lead to the formation of a mixture of two CIAS quaternary phases with different compositions.

We investigated the effect of H₂S annealing on Sb-doped Cu–In–S thin films prepared using a vacuum evaporation method. These CuInS₂ thin films were grown successfully when annealed above 350 °C in H₂S atmosphere. It was found that the thin films were closer to stoichiometry at lower annealing temperatures compared with the undoped thin films. The carrier concentrations, resistivities and mobilities of Sb-doped thin films annealed at 400 °C were approximately 1×10¹⁸ cm^-3, 30 Ω cm, and 1 cm² V^-1 s^-1, respectively.

The relationship between the selenization condition and the grain structure has been studied. A single-phase chalcopyrite Cu(In,Ga)Se₂ thin film with densely-packed grains and large grains have been prepared by the two-step selenization of the In/Cu–Ga bilayer precursor using diethylselenide (DESe). The formation of the In–Se compound in the early stage of the first-step selenization (T=350–450 °C) has been found to be important. The succeeding second-step selenization at high temperature of 540 °C led to the well developed (112) grain formation. The relationship between the selenization condition and the CIGS film structure is discussed with relation to the selenization mechanism.

Among many key parameters required to obtain a record-efficiency Cu(In,Ga)Se₂ (CIGS) cell, the band-gap of CIGS should have a double-graded profile in which the band-gap increases toward both of the back and front of the absorber. In an effort to obtain an increased Ga content near the junction area which will raise the band-gap energy of CIGS, a novel metal precursor layered with predetermined amount of Se was annealed in N₂ ambient. By inserting the Se layer in between metallic precursor layers, it was found that the front band-gap was increased due to the high Ga content by changing the direction of selenization reaction from inside to outside of metallic precursor. The proposed method is expected to provide a simple process for high quality CIGS photovoltaic absorber layer. The conversion efficiency of 6.80% with J_sc = 37.65 mA/cm², V_oc = 0.51 V, and FF= 35.4% in an active area of 0.48 cm² was achieved.

The typical crystal surfaces of CuInSe₂ (CISe) and related compounds were studied using density functional theory (DFT) methods. We evaluated energies and surface structures of (112), (), (110), (102), (100), (00), (001), and (00) surfaces on CISe. For CISe, the (112) surface had the lowest energy among these surfaces, and the (110) and (102) surfaces had slightly higher energies than the (112) surface. These surface atoms coordinate with the three surrounding atoms. We found that the (112) surface is most stable for CISe, and this result is consistent with the experimental results showing that the (112) surface is most frequently observed in polycrystalline CISe thin films. We also evaluated the surface energies of CuGaSe₂ (CGSe), CuAlSe₂ (CASe), CuInS₂ (CIS), CuInTe₂ (CIT), and AgInSe₂ (AISe). For CGSe, CIS, and CIT, the (112) surface had the lowest energy as in the case of CISe. However, for CASe and AISe, the (110) surface had the lowest energy.

Three-dimensional (3D) compound solar cells with the structure of <Au/CuInS₂/In_x(OH)_yS_z/porous TiO₂/compact TiO₂/florin-doped tin-oxide-coated glass plates> have been fabricated by spray pyrolysis deposition of CuInS₂ and chemical-bath deposition of In_x(OH)_yS_z for the light absorber and buffer layer, respectively. The effect of deposition and annealing conditions of In_x(OH)_yS_z on the photovoltaic properties of 3D CuInS₂ solar cells was investigated. In_x(OH)_yS_z annealed in air ambient showed a better cell performance than those annealed in nitrogen ambient and without annealing. The improvement of the performance of cells with In_x(OH)_yS_z buffer layers annealed in air ambient is due to the increase in oxide concentration in the buffer layers [confirmed by X-ray photoelectron spectroscopy (XPS) measurement]. Among cells with In_x(OH)_yS_z buffer layers deposited for 1, 1.5, 1.75, and 2 h, that with In_x(OH)_yS_z deposited for 1.75 h showed the best cell performance. The best cell performance was observed for In_x(OH)_yS_z deposited for 1.75 h with annealing at 300 °C for 30 min in air ambient, and cell parameters were 22 mA cm^-2 short-circuit photocurrent density, 0.41 V open-circuit voltage, 0.35 fill factor, and 3.2% conversion efficiency.

The effects of bismuth (Bi) incorporation into Cu(In_1-x,Ga_x)Se₂ (CIGS) thin films and solar cells have been investigated. 10–50-nm-thick Bi thin layers were deposited onto Mo-coated soda-lime glass (SLG) and SiO_x-coated SLG substrates by vacuum evaporation. CIGS thin films were then deposited by a three-stage process at substrate temperatures of 450–550 °C. The grain growth of CIGS thin films was enhanced, and the open-circuit voltage and hence the conversion efficiency was improved by the Bi incorporation when the SLG substrates were used. However, little effect was observed when the alkali barrier SiO_x layer was deposited on SLG substrates. As a result, we found that the Bi incorporation is beneficial for improving the cell performance when sodium exists simultaneously in CIGS layers.

The effects of antimony (Sb) doping into Cu(In_1-x,Ga_x)Se₂ (CIGS) thin films and solar cells have been investigated. 10–50-nm-thick Sb thin layers were deposited onto Mo-coated sodalime glass (SLG) and SiO_x-coated SLG substrates by vacuum evaporation. CIGS thin films were then deposited by a three-stage process at substrate temperatures of 450–550 °C. The grain growth of CIGS thin films was enhanced, and the open-circuit voltage and hence the conversion efficiency improved with the Sb doping when the SLG substrates were used. However, little or no effect was observed when the alkali barrier SiO_x layer was deposited on SLG substrates. As a result, we found that Sb doping is beneficial for improving the cell performance when sodium exists simultaneously in CIGS layers.

Cu₂ZnSnS₄ thin films were deposited by the spray pyrolysis method. Copper(II) acetate monohydrate, zinc(II) acetate dihydrate, tin(II) chloride dihydrate, sodium trihydrate, and pure sulfur powder were used as the starting materials of spray pyrolysis solutions, and N,N-dimethylformamide and monoethanolamine were used as the solvent and the stabilizer, respectively. The solution was coated on a Mo-coated soda lime glass substrate and after coating solutions, the films were annealed in a N₂ gas atmosphere at 360, 440, and 520 °C with or without placing two films face-to-face. It was found that the face-to-face annealing process prevented losses of S and Sn during the annealing process and the samples annealed with another precursor placed face-to-face at 440 and 520 °C showed narrow and large X-ray diffraction peaks and large grains in the surface images.

Cu₂ZnSnS₄ (CZTS) thin films were deposited onto Mo-coated and tin-doped indium oxide (ITO) coated glass substrates by using single step electrodeposition technique followed by annealing in N₂ + H₂S atmosphere. Subsequently, they were applied to the fabrication of thin film solar cells. Upon annealing, the amorphous nature of as-deposited precursor film changes into polycrystalline kesterite crystal structure with uniform and densely packed surface morphology. Energy dispersive X-ray spectroscopy (EDS) study reveals that the deposited thin films are nearly stoichiometric. Optical absorption study shows the band gap energy of as-deposited CZTS thin films is 2.7 eV whereas, after annealing, it is found to be 1.53 eV. The solar cell fabricated with CZTS absorber layer, showed the best conversion efficiency (η) 1.21% for 0.44 cm² with open-circuit voltage (V_oc) = 315 mV, short-circuit current density (J_sc) = 12.27 mA/cm² and fill factor (FF) = 0.31.

Cu-poor Cu_2(1-x)ZnSnSe₄ powders were prepared from elemental powders. The crystal structure of Cu-poor Cu_2(1-x)ZnSnSe₄ was examined by X-ray diffraction (XRD) and X-ray absorption fine structure (XAFS) analyses. Kesterite-type Cu_2(1-x)ZnSnSe₄ could be prepared in the range of 0≤x ≤0.0750. The lattice parameters were refined by the Rietveld analysis of X-ray diffraction data. The lattice constants a and c decreased with a decrease in the Cu/(Zn+Sn) ratio. However, there was little change in c/a value. On the other hand, the position of the Se atom (u parameters) changed considerably. The XAFS study showed that the local structure of Sn in Cu₂ZnSnSe₄ (CZTSe) changed with a decrease in Cu/(Zn+Sn) ratio and the local structural changes in Cu, Zn, or Se could not be clearly observed. These local structural changes around Sn are due to the disordering of Cu, Zn, and Sn atoms. The diffuse reflectance spectra showed that the band gap of Cu₂ZnSnSe₄ is 0.98 eV and that the band gaps do not depend on Cu/(Zn+Sn) ratio in the range of 0≤x ≤0.0750.

We prepared a Cu₂ZnSn(S_xSe_1-x)₄ (CZTSSe) solid solution from elemental powders. The CZTSSe solid solutions were synthesized by heating the elemental mixtures at 550 °C for 5 h in an N₂ gas. All CZTSSe materials were analyzed by Rietveld analysis using the kesterite structure with a space group of I (No. 82). Rietveld analysis showed that the lattice parameters, a and c, monotonically decreased with increasing S content. Moreover, the local structures of the Cu and S atoms were investigated by X-ray absorption near edge structure (XANES). Although the local structure of the S atom in CZTSSe hardly changed in relation to the S/Se ratio, we found that the surface of CZTSSe powders became slightly oxidized. On the other hand, Cu K edge XANES showed that the S/Se ratio could be easily determined from the XANES spectra. The band gap energies of the CZTSSe materials were determined by transmittance and diffuse-reflectance spectra. The transmittance spectra were recorded using CZTSSe films fabricated by a printing and high-pressure sintering (PHS) process. The band gap energy, E_g, monotonically increased from 1.05 eV for CZTSe to 1.51 eV for CZTS.

Copper–zinc–tin–chalcogenide (CZTSSe) with earth abundant elements has attracted increasing attention due to large absorption coefficient and band gap of ∼1.5 eV which is near the optimum band gap of single-junction photovoltaic devices. In this study, we used commercially available precursors to produce wurtzite Cu₂ZnSnS₄ nanocrystals by simple solvothermal synthesis. Different from the typical kesterite or stannite phases of CZTS, the nanocrystals synthesized in this study are in wurtzite phase with hexagonal crystal cell. The n-dodecanethiol was used to control the reactivity of metal ions, leading to the controlled size of CZTS nanoparticle by simply varying the reaction time. Furthermore, the as synthesized CZTS nanocrystals have novel wurtzite crystal structure. As a result, a red shift of absorption band edge between the CZTS nanoparticles with different size was obtained. Our study provides an extending method of CZTS nanocrystal ink preparation awaiting for further photovoltaic device application.

The valance band offset ΔE_v for the CdS/Cu₂ZnSnS₄ heterojunction is obtained on the basis of the first-principles pseudopotential method. Cu₂ZnSnS₄ is considered to crystallize in the kesterite structure. The total density of states and the local density of states for each atoms are calculated for the CdS(001)/Cu₂ZnSnS₄(001) supercell. There are two inequivalent interfaces between CdS and Cu₂ZnSnS₄ in the supercell, and different values of ΔE_v are obtained for them, i.e., ΔE_v1 = 1.1 eV and ΔE_v2 = 1.6 eV. The corresponding conduction band offsets ΔE_c are ΔE_c1 = 0.2 eV and ΔE_c2 = 0.7 eV with the conduction band minimum of Cu₂ZnSnS₄ higher than that of CdS.

This study demonstrates the effects of transition metal dichalcogenide, MoS₂ layer formation in between the copper–zinc–tin–sulphide (CZTS) absorber layer and Mo back contact from theoretical study and numerical modeling. The objective of this study is to elucidate the effects of n or p type MoS₂ on the overall CZTS solar cell performance. Energy band line-up of Mo/MoS₂/CZTS interface is analyzed to elucidate the interface properties. It is found out that p-MoS₂ layer in CZTS solar cell induces the same adventitious effect as p-MoSe₂ in CIGS solar cell. However, n-MoS₂ layer has detrimental effect on the CZTS solar cell by creating an additional back contact diode with p-CZTS layer and an ohmic contact with Mo layer. Thickness, bandgap energy and carrier concentration of n-MoS₂ all have been varied in the numerical simulation to observe its effects on the cell performance parameters. The results from numerical simulation show that MoS₂ layer as thin as 50 nm is sufficient enough to induce adverse effect on the solar cell performance. This could be caused by the increase in series resistance of the solar cell as n-type MoS₂ would inhibit hole current into Mo back contact due to the hole barrier between n-type MoS₂ and Mo back contact. The increase in MoS₂ bandgap and carrier concentration also results in detrimental effect to the performance of the cell mainly due to the possibility of electrons to drift towards the back contact and recombine.

We fabricated three-dimensional (3D)-structure solar cells with Cu₂ZnSnS₄ (CZTS). A CdS buffer layer was deposited around nc-TiO₂ by the chemical bath deposition (CBD) method. The CdS deposition time was varied from 5 to 30 min at 65 °C. CZTS absorber was deposited by spray pyrolysis deposition (SPD) onto the CdS buffer layer. Cu-, Zn-, Sn-, and S-containing solutions were used in SPD. The metal sources of copper(II) acetate monohydrate, zinc(II) acetate dehydrate, and tin(II) chloride dehydrate, and pure sulfur powder were desolved in N,N-dimethylformamide (DMF) as a solvent and mono-ethanolamine as a stabilizer. The solution-sprayed 3D-structure substrate was annealed in a nitrogen atmosphere at 250 °C for 30 min. A 3D-structure CZTS solar cell presented the best conversion efficiency η of 0.51% with a CdS buffer layer deposition time of 20 min. The dependence of each solar cell characteristic on CdS buffer layer deposition time was measured.

Cu₂SnS₃ (CTS) contains non-rare metals and it has suitable optical characteristics for the absorber layer of thin-film solar cells. In this study, CTS thin films were fabricated by sulfurizing Cu–Sn precursors deposited by co-electrodeposition. Solar cells with a structure glass/Mo/CTS/CdS/ZnO:Al/Al were fabricated from the films. The best cell had an efficiency of 2.84%. A relatively high conversion efficiency was obtained from films with Cu/Sn≤2.

Cu₂SnS₃ (CTS) has been reported to have various band gap energies in the range of 0.93–1.77 eV and an absorption coefficient of 1.0×10⁴ cm^-1. It consists of elements that are inexpensive due to their abundance in Earth's crust. Consequently, CTS is expected to be utilized in the absorber layers of thin-film solar cells. In this study, Cu/Sn stacked-layer thin-film precursors were deposited on glass and glass/Mo substrates by electron beam evaporation. CTS thin films were fabricated by sulfurizing the precursors at temperatures of 450–580 °C for 2 h in an atmosphere of N₂ and sulfur vapor. CTS films were estimated to have band gap energies of 0.96–1.00 eV by extrapolation. A solar cell fabricated using a CTS thin film sulfurized at 580 °C exhibited an open-circuit voltage of 211 mV, a short-circuit current of 28.0 mA/cm², a fill factor of 0.43, and a conversion efficiency of 2.54%.

Thin films of tin sulphide (SnS) were deposited onto glass substrates using the thermal evaporation method. The substrate temperature, T_s was varied in the range, 280–360 °C, keeping other growth parameters constant and the effects on the chemical and physical properties of the layers deposited were investigated. The layers were observed to consist of densely packed grains, up to 9.5 µm in diameter, and X-ray diffraction studies showed the layers had a strong preferred (040) orientation. The energy band gap, determined from optical studies was found to be in the range 1.30–1.34 eV, while the optical absorption coefficient was found to be >10⁴ cm^-1 for photons with energies greater than the energy bandgap. The electrical resistivity of the films decreased with an increase of substrate temperature. Heterojunction devices were made using chemical bath deposited cadmium sulphide (CdS) as the window layer and the minority carrier diffusion length in the SnS measured to be 0.23 µm using spectral response studies.

To achieve low cost solar cells, new fabrication processes should be developed for higher throughput and utilization rate. In this study, we focused on a sputtering method and found that a multinary compound ZnInS (II–III–VI) is suitable for this process. The ZnInS thin film deposited by sputtering had an n-type semiconductor characteristic. A AgInTe/ZnInS thin-film solar cell fabricated by the sputtering process in a layer structure of glass/Mo/AgInTe/ZnInS/AZO showed a conversion efficiency of over 1%, the origin of which was mainly the ZnInS layer. These results suggested that ZnInS is a strong candidate photovoltaic material for fabrication with the sputtering process.

SnO₂ thin films were successfully electrodeposited from an aqueous oxygen-bubbled tin sulfate solution and partnered with electrodeposited SnS thin films to fabricate SnS/SnO₂ heterojunction solar cell. The electrodeposited SnS/SnO₂ superstrate structure with 250-°C-annealed SnO₂ as a window layer exhibited an open circuit voltage of 40–90 mV and a short circuit current density of 1.5–9.7 mA/cm². The solar conversion efficiency was estimated to be in the order of 10^-2–10^-1%. The band discontinuities at the SnS/SnO₂ interface were evaluated by X-ray photoelectron spectroscopy. The valence band offset was determined to be approximately 1.85 eV. Using this value and the band gaps of individual layers, the conduction band minimum of SnO₂ is predicted to be higher than that of SnS by 0.65 eV.

Cu₂O thin films were deposited on indium–tin-oxide-coated glass from an aqueous solution containing CuSO₄, lactic acid and KOH by the galvanostatic electrochemical deposition at 40 °C with several different current densities. The photo-absorption of Cu₂O was increased and the conduction type was changed from weak p-type to clear p-type by raising the current value. Cu₂O(2)/Cu₂O(1)/ZnO three-layer heterojunctions were fabricated electrochemically by modulation of deposition current density of Cu₂O. The first Cu₂O layer Cu₂O(1) was deposited at a lower deposition current, and the second one Cu₂O(2) at a higher current. Under the optimized condition, the conversion efficiency of a Cu₂O(2)/Cu₂O(1)/ZnO solar cell was found to be higher than that of a Cu₂O(1)/ZnO solar cell.

BaCuSeF films were fabricated on glass substrates for solar cell application. The crystallographic orientation of the films depended on the thickness of the films. The BaCuSeF films with thicknesses of 0.35, 0.6, and 1.0 µm had 102 preferential orientation, and a thicker film of 1.3 µm was polycrystalline. All of the films showed average transmittance of >50% in the visible light region, and the determined band gap energy was 2.8 eV. All of the films showed p-type conductivity of more than 1 S cm^-1. The 0.35-µm BaCuSeF film showed the highest p-type conductivity of 19.2 S cm^-1. The BaCuSeF is applicable to electrodes for chalcopyrite-based thin film tandem solar cells.

Multijunction solar cells have evolved from their original development for space missions to displace silicon cells in high concentrating photovoltaic (CPV) systems. Today's three-junction lattice-matched production cells have efficiency of 39–39.5% under high concentration, and there appears to be little opportunity for further efficiency gain with this three-junction technology. Future generations of CPV cells will exploit more than three junctions, with metamorphic subcells, or both technical approaches to achieve efficiencies >45%. As new designs seek closer current matching and further spectral splitting, atmospheric variability will necessitate careful modeling to optimize energy output. These new cells will also be higher cost, which will favor higher CPV system concentration.

We propose a technologically feasible concept of a hot carrier (HC) solar cell (SC) which fulfills the electronic, optical, and to some extent the phononic criteria required. The energy selective process of HCs is implemented into the hot carrier absorber (HCA). Its electronic properties are investigated by a Monte-Carlo code which simulates random deviations of structure thickness and a normal distribution of random elastic electron (e^-) scattering. The structure can be grown epitaxially as a HC-SC test device.

We have revealed that a new scheme of solar cells, i.e., an intermediate-band-assisted hot-carrier solar cell (IB-HC-SC) using energy-selective contacts (ESCs) consisting of quantum wells (QWs) can be a practical solution to achieve significantly high conversion efficiency. There are three requisites unique to hot-carrier extraction for high conversion efficiency: (1) a long thermalization time of carriers in the absorber, (2) a narrow energy-selection width of the ESCs, and (3) a short equilibration time of carriers in the absorber. The use of an intermediate-band (IB) absorber relaxes the first requisite, because the two-step excitation via the IB dramatically suppresses entropy generation associated with hot-carrier extraction that is more remarkable at a shorter thermalization time. The suppression of the entropy generation allows us to use QW-ESCs to solve the issue related to the second requisite involved in practical ESCs consisting of size-distributed quantum dots. The new scheme provides limiting conversion efficiency of around 50% (0.1 sun)–65% (1000 sun) that are significantly higher than those of conventional IB solar cells, when the thermalization time of hot carriers is assumed to be 1 ns.

For a quantitative evaluation of the carrier transport dynamics of multiple quantum well (MQW) solar cells, carrier collection efficiency (CCE) was defined and its measurement procedure was proposed. CCE is essentially the quantum efficiency normalized to the saturation value at a reverse bias. It allows us to know whether substantial carriers are actually extracted at any bias voltage, and to uncover the bottleneck problems of carrier transport that emerge at the operation bias. The advantage of CCE analysis is that the dynamics of photoexcited carriers can be selectively examined independently of the diode characteristics of the devices if the effects of resistance are small enough. In the present study, GaAs p–i–n solar cells including various numbers of InGaAs/GaAsP MQWs with a band gap of 1.2 eV in the i-region of equal thickness were fabricated and characterized. Interfered carrier transport by increasing the well number was quantitatively and directly evaluated. With up to 30 periods of MQWs, the carriers generated, especially in the wells, were less likely to be collected than those generated in the top p-region at a moderate forward bias, but collection of both was found to be degraded severely with 40 periods of MQWs.

With the final goal of integrating III–V materials to silicon for tandem solar cells, the influence of the metal–organic vapor phase epitaxy (MOVPE) environment on the minority carrier properties of silicon wafers has been evaluated. These properties will essentially determine the photovoltaic performance of the bottom cell in a III–V-on-Si tandem solar cell device. A comparison of the base minority carrier lifetimes obtained for different thermal processes carried out in a MOVPE reactor on Czochralski silicon wafers has been carried out. The effect of the formation of the emitter by phosphorus diffusion has also been evaluated.

FluidReflex is a novel concentrator concept that has been designed to achieve a low cost concentration photovoltaic (CPV) system over 1000× by using reflective optics and a fluid dielectric. In this paper, results regarding the characterization of the concentrator Optical Transfer Function (OTF) are presented. The reflectance and transmittance of the different materials that the concentrator is composed of have been measured using a spectrometer. In addition, a method using component cells (isotypes) has been used to experimentally determine which of the subcells within the multijunction (MJ) solar cell is limiting the current when it is illuminated by the concentrator.

We study the capacitance–voltage characteristics of GaAs/AlGaAs coupled multiple quantum well (MQW) solar cells. It is found that the capacitance under illumination increases sharply if the bias is raised above -0.2 V, and gets maximum at a bias of about 0.2 V. This increment in capacitance by the illumination is ascribed to the reduction of the depletion layer thickness, caused by the spatially separated accumulation of photogenerated electrons and holes trapped in the MQW layer.

We have studied how current–voltage (I–V) characteristics of GaAs/AlGaAs multiple quantum well (MQW) solar cells under illumination depend on the wavelength of incoming photons at 6 K. Coupled MQW structures used in this study consist of 4-nm-thick GaAs and 3-nm-thick AlGaAs layers. It is found that I–V curves under illumination measured at 6 K exhibit distinct shapes that depend systematically on the energy E_ph of incoming photons; when E_ph is smaller than the bandgap E_g of AlGaAs, the photocurrent shows a single-step decrease with voltage, while for E_g < E_ph, it decreases in two or three steps. These behaviors are explained in terms of bias-voltage dependent changes in various dynamic processes of photogenerated carriers.

By high-accuracy in situ curvature measurement during the growth of InGaAs/GaAsP superlattice structures by metal organic vapor phase epitaxy, we have successfully observed the effect of thin GaAs insertion layers between InGaAs wells and GaAsP barriers on strain control. By analyzing curvature transients, we found that an inadequate gas-switching sequence induces the carry over of indium from the InGaAs layer to the overlying GaAs insertion layer. The resulting carry-over layer has an estimated thickness of 0.6 nm and adversely affects the average strain of the structure. Through consideration of the kinetics of surface atoms, it has been revealed that an optimized gas-switching sequence with a 1 s hydrogen purge after the growth of InGaAs wells is effective for preventing the carry over.

We discuss the influence of the barrier thickness of an InGaN/GaN multiple quantum well (MQW) structure on solar cell performance. As barrier thickness decreases, short-circuit current density increases and open-circuit voltage decreases. The open-circuit voltage is much lower than expected from the absorption edge because of the large leakage current and large ideality factor of diodes owning to the carrier tunneling through the barrier. An MQW with a 3-nm-thick barrier layer shows a much longer carrier lifetime than that with a 9-nm-thick barrier layer. This is one possible reason for a higher short-circuit current in solar cell with the 3-nm-thick barrier MQW structure than that with the 9-nm-thick barrier MQW.

An anti-soiling layer was coated on a poly(methyl methacrylate) (PMMA) substrate. The anti-soiling layer was prepared by coating the acrylic urethane capping layer, the inorganic/organic nano-graded intermediate layer, and the photocatalytic surface layer including modified WO₃ and partial hydrolyzed tetraethyl orthosilicate. The layers were coated by spin-coat method. The photocurrent from each subcell of the InGaP/InGaAs/Ge triple-junction solar cell can be calculated by multiplying the solar spectrum of AM1.5D (850 W/m²), the transmittance of the PMMA substrate, and the spectral response of the triple-junction solar cell. After 8 months exposure, the reduction rate of photocurrent of the sample without anti-soiling coat was 9.6%. On the other hand, that with anti-soiling coat could be suppressed to 3.3%.

Solar simulators for characterization of concentrator photovoltaic modules (CPV) are now commercially available, providing an alternative to outdoor performance characterization with clear advantages in terms of repeatability, availability, control of operating conditions, and cost. Nevertheless, optical concentration implies particular characteristics that introduce the need of new performance figures and characterization methods. CPV solar simulators have demonstrated characterization capabilities beyond simple current–voltage (I–V) curve acquisition. This paper summarizes other indoor performance tests developed at the Instituto de Energía Solar (IES-UPM) that have proven to be useful for the optical, mechanical, and spectral characterization and optimization of a CPV module.

Numerical simulation of optical absorption characteristics of gallium arsenide (GaAs) thin-film solar cells by the three-dimensional finite element method is presented, with emphasis on optimizing geometric parameters for nanowire and nanocone structures to maximize the ultimate photocurrent under AM1.5G illumination. The nanostructure-based GaAs thin-film solar cells have demonstrated a much higher photocurrent than the planar thin films owing to their much suppressed reflection and high light trapping capability. The nanowire structure achieves its highest ultimate photocurrent of 29.43 mA/cm² with a periodicity (P) of 300 nm and a wire diameter of 180 nm. In contrast, the nanocone array structure offers the best performance with an ultimate photocurrent of 32.14 mA/cm². The results obtained in this work provide useful guidelines for the design of high-efficiency nanostructure-based GaAs solar cells.

We fabricated a GaAs/AlGaAs quantum dot solar cell (QD-SC) in which GaAs QDs with an underlying quantum well (QW) layer were grown to enable the absorption of long-wavelength photons. Photoluminescence emission of these QDs with an underlying QW, referred to as QDW structure, is redshifted and significantly narrower than that of plain QDs. The extension of absorption wavelength is clearly observed in the spectral response of the samples while keeping zero-dimensional confinement as confirmed by calculation. Improvement in fill factor in the QDW sample compared with plain QDs is demonstrated.

The effects of background Zn doping on the performance of p–i–n GaAs solar cells with InGaAs/GaAsP multiple quantum wells (MQWs) in i-GaAs layer have been studied. The crystal growth was done by a planetary metalorganic vapor phase epitaxy (MOVPE) reactor. The background Zn doping, in an order of 10¹⁷ cm^-3, degraded the solar cell efficiency by modifying the energy band diagram in a way that obstructed carrier transports. It was shown by calculation that the carrier transports across the MQWs region suffered from decrease in built-in electric field in absorber layers, leading to an efficiency loss by radiative and nonradiative recombinations. Consequently, the external quantum efficiency and the current density of a Zn-contaminated MQW solar cell were terribly poor. Reactor baking at 850 °C for 20 min seems to remove Zn residues effectively without noticeable effects on the succeeding growth of MQW solar cells. The InGaAs/GaAsP MQWs fabricated in the thermally cleaned reactor have shown a potential to extend the absorption edge of GaAs solar cells and to improve the efficiency of multi-junction solar cells by current matching. Therefore, the growth of InGaAs/GaAsP MQWs by planetary MOVPE reactors requires a careful treatment regarding the background doping issue.

The effects of accumulating strain inside InGaAs/GaAsP multiple-quantum-well (MQW) solar cells were investigated and their correlation with in situ wafer curvature measurement was examined. The p–i–n GaAs solar cells, containing 20-period InGaAs/GaAsP MQWs in an i-GaAs layer, were fabricated by metalorganic vapor phase epitaxy. The strain inside MQWs was varied by changing In content in an InGaAs well, while maintaining other parameters. As evidenced by curvature transience, the excessive strain led to lattice relaxation, resulting in defects, dislocations, and poor crystal quality. Consequently, short circuit current density and open circuit voltage deteriorated, and solar cell performance degraded. The highest conversion efficiency was obtained in a strain-balanced MQW solar cell. InGaAs/GaAsP MQWs have a great potential for extending the absorption edge of GaAs cells and for enhancing the efficiency of III/V multijunction solar cells by current matching. Hence, the growth of InGaAs/GaAsP MQWs for photovoltaic application requires a strain monitoring system and careful control such that the accumulating strain is minimized.

First-principles calculations of GaAsN surface with low nitrogen (N) content grown by chemical beam epitaxy were performed to theoretically analyze the incorporation process of nitrogen and impurities at the atomic scale. As a result, stable surface structures of GaAsN(001) under hydrogen (H) atmosphere were determined. In these structures, N is suggested to readily substitute into surface sites, especially those that bond with H, compared with in the bulk. This indicates that N is incorporated into a thin film together with H. This may generate H-related defects, which may lead to the degradation of its electric properties. These defects are difficult to minimize by post-annealing processes. Therefore, the amount of H attached to the growth surface should be reduced in order to obtain high-quality crystals. The calculated surface phase diagram suggests that a condition in which the extent of the incorporation of H-related defects can be reduced exists.

Aiming at reducting in Joule energy loss of a photovoltaic cell under sunlight concentration, monolithic integration of GaAs cells has been realized, in which five subcells were connected in series and the total surface area of the cells occupied over 80% of the whole chip area. Using plasma etching with Cl₂, a sufficiently sharp mesa for device isolation was obtained. Insulation between etched mesa sidewalls and interconnect electrodes proved to be the most significant issues for the purpose of eliminating shunt resistance and securing a reasonable fill factor; the SiO₂ layer deposited by sputtering was much superior to polyimide as an insulator. The fabricated test device showed a short circuit current density of 20.7 mA/cm² and an open circuit voltage of 4.79 V, which were consistent with the values for a single subcell.

A composite film of poly(3,4-ethylene dioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) and titanium carbide nanoparticles (TiC-NPs) was deposited on an indium doped tin oxide–poly(ethylene naphthalate) (ITO–PEN) conducting plastic substrate by a doctor blade technique. This ITO–PEN substrate with the composite film was used as the flexible counter electrode (CE) for a plastic dye-sensitized solar cell (DSSC). Performances of the plastic DSSC with the platinum-free CEs containing PEDOT:PSS/TiC-NPs composite was investigated. A solar-to-electricity conversion efficiency (η) of 6.50% was achieved for the pertinent DSSC with PEDOT:PSS/TiC-NPs composite, using commercial N719 dye, which exhibited comparable performance to that of a cell with a sputtered-Pt film on its CE (6.84%). The homogeneous nature of the composite film PEDOT:PSS/TiC-NPs, the uniform distribution of TiC-NPs in its PEDOT:PSS matrix, and the large electrochemical surface area of the composite film are seen to be the factors for the best performance of the pertinent DSSC. Transmission electron microscopy (TEM) and X-ray diffraction (XRD) were used to characterize the TiC-NPs. Scanning electron microscopy (SEM), atomic force microscopy (AFM), and energy dispersive X-ray spectroscopy (EDX) were used to characterize the films. The high efficiency of the cell with PEDOT:PSS/TiC-NPs is explained by cyclic voltammetry (CV), electrochemical impedance spectroscopy (EIS), and incident photon-to-current conversion efficiency (IPCE) curves.

A novel Ru(II) porphyrin sensitizer, bis[4-(4-pyridylazo)-benzoic acid]tetrakis(3,5-di-tert-butyl-4-hydroxyphenyl)porphyrinato ruthenium(II), [RuTBP(azpyba)₂] derived from tetrakis(3,5-di-tert-butyl-4-hydroxyphenyl) Ru(II) porphyrin with azopyridine axial ligands have been synthesized and characterized by UV–vis, emission, nuclear magnetic resonance (NMR), differential pulse voltammetry, density functional theory (DFT) calculation, and subpicosecond time-resolved transient absorption. The RuTBP(azpyba)₂ sensitizer shows both the intra-porphyrin π–π^* absorption in the visible region and broad charge-transfer absorption from Ru²⁺ to the axial ligands in the near-IR region. The RuTBP(azpyba)₂-sensitized solar cell shows sensitivity in the near-IR region up to ∼1100 nm owing to the charge transfer absorption band. It was indicated that the excited electrons are injected to TiO₂ through the azopyridine axial ligand of RuTBP(azpyba)₂.

We investigate the effect of deuteration on geminate recombination in photoelectrochemical cells operating by interfacial charge transfer absorption bands. The trend in recombination in surface complexes of tetracyanoethylene (TCNE), 7,7,8,8-tetracyanoquinodimethane (TCNQ), and 11,11,12,12-tetracyanonaphtho-2,6-quinodimethane (TCNAQ) with TiO₂ is treated as internal conversion (IC) in the model compounds TCNX–O–Ti(OH)₃^-. The deuteration of TCNQ and TCNAQ significantly modifies the spectrum of vibronic coupling constants for many vibrational modes, but affects little the modes with the strongest contribution to IC. As a result, the overall effect on recombination/internal conversion is expected to be limited, slightly increasing its rate. We also consider the influence on the recombination of vibrational modes of the Ti(OH)₃ moiety which only crudely models the oxide surface. We conclude that even as the model is sensitive to the motions of Ti(OH)₃, the predicted trend in recombination in the series TCNE–TCNQ–TCNAQ holds under different treatments of Ti(OH)₃ modes.

Polythiophene derivatives with carboxylic acid groups directly attached to the thiophene rings were synthesized for polymer-sensitized solar cells (PSSCs). The polymer layer densely and uniformly covering the surface of the TiO₂ particle in the photoanode of the PSSCs allows excitons to diffuse to the polymer/TiO₂ interface efficiently. The PSSCs yielded a maximum incident photon-to-current conversion efficiency of about 80% at 480 nm and the open-circuit voltage (V_OC) values but also short-circuit current density (J_SC) values increased with decreasing hydrolysis ratio of the polymer.

A SnO₂ electrode is a promising candidate for a bottom electrode of dye-sensitized solar cells. One of the drawbacks is its low fill factor (FF). We clarified the cause of this low FF using our original hybrid cells consisting of carrier generation areas (a bottom layer consisting of TiO₂/a dye layer) and carrier transport areas (a top layer consisting of SnO₂ with and without a dye layer). A large decrease in FF was observed only when the SnO₂ charge transport areas were covered by photo excited dyes, leading to the conclusion that back electron transfer reaction from SnO₂ to oxidized dyes is a major route for the charge recombination. This was also confirmed by electron lifetime and dark current measurements.

The three-dimensional (3D) visualisation of the photovoltaic active layer in a dye-sensitised solar cell (DSSC) was newly achieved by employing an optical interference measurement system. The technique was used to measure with high precision the active layer thickness of a complete cell to evaluate the diffusion resistance of the electrode layer made of titania and the electrolyte layer containing tetracyanoborate ionic liquid. Additionally, the visualised image was compared with the local photovoltaic characteristics in a region with diameter 100 µm studied using an AC impedance technique as well as by transient photocurrent and photovoltage decay measurements to elucidate the charge transport properties of the ionic-liquid-based electrolyte.

We fabricated the organic photovoltaic (OPV) cells on the transparent and low resistive multilayer electrode of a few nm Ag layer embedded Al doped zinc oxide film (AZO/Ag/AZO) on a glass. AZO and Ag consisting of the continuous multilayer were fabricated using rf magnetron at room temperature successively. AZO (45 nm)/Ag (11 nm)/AZO (45 nm) electrode shows an electrical resistivity of 6.56×10^-5 Ω cm, optical transmittance of 89.3% at 550 nm and figure of merit (FOM) value of 4.94×10^-2 Ω^-1. An inverted structure OPV was fabricated on AZO/Ag/AZO transparent electrode in which TiO_x and PEDOT:PSS were used as an electron transport and hole transport layer respectively. The OPVs showed power conversion efficiency and fill factor as high as 2.07%, 0.51 respectively.

We have investigated organic thin-film solar cells from the microscopic viewpoint with light-induced electron spin resonance (ESR). The utilized cell structures are indium–tin oxide (ITO)/poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS)/regioregular poly(3-hexylthiophene) (P3HT):[6,6]-phenyl C₆₁-butyric acid methylester (PCBM)/Pd/LiF/Al. We performed simultaneous measurements of light-induced ESR and device characteristics using the same device. Under simulated solar irradiation, the light-induced ESR intensity of a solar cell increases, while the short-circuit current and open-circuit voltage of the cell concomitantly decrease with increasing duration of irradiation. We have successfully observed a clear correlation between the light-induced ESR intensity and the device characteristics. Trapping sites of photogenerated hole carriers (positive polarons) are ascribed to P3HT by the analysis of the ESR signals.

The electrical characteristics of silicon quantum dot superlattice (Si-QDSL) solar cells have been investigated using a two-dimensional (2D) device simulator, taking the quantum size effect into account. The average bandgap of a Si-QDSL embedded in an amorphous silicon oxycarbide (a-SiO_xC_y: E_g=2.5 eV) matrix increased from 1.21 to 1.75 eV with decreasing diameter of Si QDs from 10 to 3 nm owing to the quantum size effect. It should be noted that the V_oc of Si-QDSL solar cells also increased to 1.11 V with decreasing diameter of Si QDs to 3 nm. This result indicates that it is possible to enhance V_oc by the quantum size effect and that a Si-QDSL may be a promising material for all-silicon tandem solar cells.

Boron (B)- and antimony (Sb)-doped Si quantum dots (QDs) in Si₃N₄ films were fabricated using the co-sputtering method with a post-deposition anneal. The effect of B and Sb on Si QDs films was investigated in terms of structural, optical and electrical properties. It is found that a low dopant concentration induced negligible structural changes in the Si QD films. The PL intensity decreases with increasing B or Sb content. This could result from the non-radiative recombination processes attributed to defects associated with the dopants and Auger processes due to successful doping of Si QDs. For the B-doped sample the conductivity increases about 100 times, which could be attributed to an increase in carrier concentration. For the Sb-doped sample, a significant increase (six orders of magnitude) in conductivity suggests an effective Sb doping. The charge transport mechanism in the Sb-doped Si QD films matches well with the percolation-hopping model in low temperature region. Both B- and Sb-doped samples show thermally activated hopping conduction characteristics in the range of 220–320 K.

New ruthenium(II)–polypyridyl complexes 1a–1d (1a: X= Y= H, 1b: X= H, Y= Cl, 1c: X= H, Y= Me, 1d: X= COOH, Y= H) having 2,6-bis(4-carboxyquinolin-2-yl)pyridine derivatives as ligands were synthesized as sensitizers for dye-sensitized solar cells (DSCs) and their photophysical and photochemical properties were characterized. The photovoltaic performance of DSCs sensitized with 1a–1d was found to be different. The DSC sensitized with 1a exhibits higher efficiency and IPCE value than those sensitized with 1b–1d. The APCE value of the DSC sensitized with 1a was almost 80% in the 500–800 nm range. Therefore, the performance of the DSC sensitized with 1a might improve if the adsorption of 1a on the TiO₂ surface could be appropriately controlled by investigation of dye immersion condition or introduction of bulky substituents on the ligand in order to suppress dye aggregation.

A novel far-red sensitive and direct aromatic ring carboxy functionalized unsymmetrical squaraine dye based on pyrroloquinoline moiety has been synthesized for the application as sensitizer of dye sensitized solar cells. This dye bearing the characteristics of both of indole and quinoline moieties exhibits the photon harvesting up to 820 nm. The enhancement in the photon harvesting window by the pyrroloquinoline based squaraine dye was attributed to the presence of increased π-conjugation in its pyrroloquinoline donor part. This pyrroloquinoline bearing unsymmetrical squaraine dye exhibits a short circuit current density of 9.88 mA/cm², an open circuit voltage of 0.57 V and fill factor of 0.59 leading to the external power conversion efficiency of 3.33% under simulated solar irradiation of 100 mW/cm² at AM 1.5 condition.

We synthesized five donor–π-spacer–acceptor type organic photosensitizers bearing different types of electron-withdrawing groups (EWGs); trifluoromethyl, o-nitrophenyl, p-nitrophenyl, cyano, and carboxyl groups, on their acceptor part in the aim of observing an influence of the EWGs on spectral and photovoltaic properties from viewpoints of steric structure and π-conjugation. The EWG possessing smaller dihedral angle between the EWG and dye skeleton exhibited larger bathochromic shift in absorptions. Highly planer cyano group presented the most red-shifted absorption at 464 nm, and the highest conversion efficiency of 5.69% was obtained. In contrast, highly distorted o- and p-nitrophenyl groups exhibited blue-shifted absorption at 416 and 422, respectively; however, despite of resemble spectral properties, o- and p-nitrophenyl gave second best and the worst conversion efficiencies of 4.05 and 2.51%, respectively. By combination with computational chemistry, it was indicated that the configuration of the EWG and distance between TiO₂ surface and the EWG dominated electron injection efficiency.

In order to examine the effect of partial charges of dye molecules on charge recombination in dye-sensitized solar cells, we synthesized carbazole-based dyes with ether chains on a π-conjugated oligothiophene linker. The addition of negatively charged oxygen atoms on ether chains resulted in a decrease of electron lifetime, suggesting that partial charges on the dye increased I₃^- concentration at the surface of TiO₂.

The shifts in quasi-Fermi level (QFL) of TiO₂ films affected by additives are directly measured using three-electrode dye-sensitized solar cells (DSCs). The difference in QFL of TiO₂ under short circuit and open circuit are also analyzed. It is found that the QFL difference between the electrolyte side of the TiO₂ film and the equilibrium redox potential, as well as the QFL difference between the two sides of the TiO₂ film shift to higher potential owing to the addition of TBP to the electrolyte. The shift values are influenced by the cations present in the electrolyte. These directly determined changes in QFL in TiO₂ films both under open circuit and short circuit may provide valuable information for studying DSC operation mechanisms such as the dynamics of charge separation and charge recombination processes.

To improve the efficiency of dye-sensitized solar cells (DSCs) by controlling dye adsorption on TiO₂ surface, the effect of surface treatments on the properties of [NBu₄]₂[Ru(Htcterpy)(NCS)₃] (black dye; [NBu₄]: tetrabutylammonium cation; H₃tcterpy: 4,4',4''-tricarboxy-2,2':6',2''-terpyridine) on nanocrystalline TiO₂ films was investigated by analysis of the photovoltaic performance and the electron transport properties. Although the surface treatments do not affect on the condition band edge of TiO₂, the amount of dye on TiO₂ increases. The enhancement of dye adsorption by treatment of TiO₂ in HCl solution is more effective than that by dipping the dye solution containing deoxycholic acid (DCA) as additive. But the charge recombination between an electron in TiO₂ and I₃^- in the electrolyte can be reduced by the DCA treatment.

An indoline dye-coupled polyviologen was designed as a possible photoanode material for organic photovoltaic cells. A new donor–acceptor (D–A) coupled molecule (CVD131) was immobilized on a current collector by the reductive electropolymerization of its two cyanopyridinium moieties. The formed polyviologen could efficiently accept an excited electron from the indoline dye moiety.

We demonstrated the effects of Cu₂O doping at various weight ratios in TiO₂ photoanode films on the photovoltaic performance of dye-sensitized solar cells (DSSCs). The photovoltaic characteristics, namely, open-circuit voltage, short-circuit current, fill factor, and energy conversion efficiency, of DSSCs with 0.3 wt % Cu₂O doping in the photoanode are 690 mV, 4.55 mA/cm², 52.0%, and 1.96%, respectively. The photovoltaic parameters of the sample with 0.3 wt % Cu₂O doping are better than those of other Cu₂O-doped samples, which are inferior to undoped TiO₂-based DSSCs. The measured cell performance shows that the addition of Cu₂O to the TiO₂ photoanode in DSSCs results in the reduction of photovoltaic parameters. The degradation mechanism upon Cu₂O doping in the TiO₂ photoanode of DSSCs may be attributed to the variations in the valence numbers of copper and titanium ions during calcination and the reduction–oxidation reaction, resulting in the formation of oxygen vacancy–Ti⁺³ defects and related recombination centers.

The characteristic of TiO₂ passivation layers grown by plasma enhanced chemical vapor deposition as a function of its thickness on F-doped SnO₂ (FTO) electrode was investigated. The thickness of TiO₂ passivation layer was varied from 30 to 200 nm by controlling the deposition time. The electric resistance of the TiO₂ layers was depended on the thickness, so the optimized thickness in enhancing the connection and reducing the recombination of electrons on the surface of FTO electrode was determined. The dye sensitized solar cells fabricated with 40 nm thick TiO₂ passivation layer showed the maximum power conversion efficiency of 6.93%. It was due to the effective connection of mesoporous TiO₂ and FTO and the prevention of electron recombination from the FTO to electrolyte. The reduced resistance, enlarged electron diffusion length measured by the electrochemical impedance spectroscopy, intensity-modulated photocurrent spectroscopy and intensity-modulated photovoltage spectroscopy identified the connection and anti-recombination effect.

Multiwall carbon nanotubes (MWCNTs) and platinum (Pt) were coated on fluorine doped tin oxide (FTO) coated glass by a direct current electrophoretic method, and were used as dye-sensitized solar cell counter electrodes. Scanning electron microscopy (SEM) detected the deposition of carbon nanotubes on FTO-glass. The deposition of Pt nanoparticles on carbon nanotube surfaces was confirmed by transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS). Cell efficiencies of MWCNTs, Pt and MWCNTs/Pt dye-sensitized solar cells (DSSCs) were ∼1.41, ∼8.13, and ∼8.90%, respectively. The improvement of the composited MWCNTs/Pt cell efficiency is due to the presence of both catalysts (MWCNTs and Pt) enhancing counter electrode catalytic activity as observed by cyclic voltammetry (CV) and reducing charge-transfer resistance as observed by electrochemical impedance spectroscopy (EIS).

Dye-sensitized solar cell (DSC) module still lacks the long-term stability because of a volatile electrolyte. In this work, structural alternation of DSC parallel modules was proposed for high performance and stability. In the conventional module, the adhesion with the photo/counter electrodes was not stable with only one layer of sealant so that the stability was poor. On the other hand, buried grid module proposed in this work is possible to be stably sealed with only one layer of sealant. In this structure, grid electrodes are leveled with the same height of transparent conductive oxides (TCOs) because grid electrodes are inserted under furrows of etched TCOs. Consequently, the buried grid module had better long-term stability than conventional module. Additionally, the photo-current and efficiency were improved by the sonicated grid and the light scattering layer.

Soluble graphene oxide (GO) and plasma-reduced (pr-) GO were investigated using crystalline silicon (c-Si) (100)/GO/pr-GO hybrid junction solar cells. Their photovoltaic performances were compared with those of c-Si/GO/pristine conductive poly(ethylene dioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) heterojunction and c-Si/PEDOT:PSS:GO composite devices. The c-Si/GO/pr-GO and conductive PEDOT:PSS/Al heterojunction solar cells showed power conversion efficiencies of 6.5 and 8.2%, respectively, under illumination with AM 1.5 G 100 mW/cm² simulated solar light. A higher performance of 10.7% was achieved using the PEDOT:PSS:GO (12.5 wt %) composite device. These findings imply that soluble GO, pr-GO, and the PEDOT:PSS:GO composite are promising materials as hole transport and transparent conductive layers for c-Si/organic hybrid junction solar cells.

Three kinds of antimony compounds – Sb₆O₁₃, MgSb₂O₆ and ZnSb₂O₆ – were prepared in the form of nanocrystalline film and their photo-electrochemical properties were investigated. The preparation of Sb₆O₁₃ was based on thermolysis of a colloidal Sb₂O₅·4H₂O suspension. MgSb₂O₆ and ZnSb₂O₆ were prepared via low-temperature hydrothermal methods. All the compounds exhibited semiconducting properties applicable to dye-sensitized solar cell (DSSC). The energy band gaps were estimated to be 3.39 eV for Sb₆O₁₃, 3.60 eV for MgSb₂O₆, and 3.31 eV for ZnSb₂O₆, respectively. After sensitization with a conventional ruthenium-dye (N719), Sb₆O₁₃-based solar cell exhibited the highest open circuit voltage (V_oc = 0.76 V) whereas the V_oc values (0.44–0.46 V) of MgSb₂O₆ and ZnSb₂O₆ are relatively low. The V_oc values were proven to be related to the flat band potentials of the antimony compounds. The overall solar-to-electric energy conversion efficiencies were in the range of 0.7–1.0% under AM 1.5, 100 mW/cm² illumination.

Ge quantum dot (QD) layers inserted in an intrinsic region of a Si p–i–n diode cause additional photon absorption at longer wavelengths of the solar spectrum. We studied the mechanism of carrier extraction in Ge/Si QD solar cells using photocurrent, capacitance, and photoluminescence measurements. Our findings show that the photon absorption and carrier extraction in Ge/Si QD solar cells depend strongly on the thermal annealing process to form the p–i–n diode. Control of Ge–Si interdiffusion at the Ge/Si interface during thermal annealing is critical for the increase in the conversion efficiency of Ge/Si QD solar cells.

Here, we discuss the results related to improvement of electronic interactions and structural properties of hybrid organic/inorganic composites based on free-standing and surfactant-free silicon nanocrystals (Si-ncs). Performance of Si-ncs in bulk-heterojunction solar cells combined with a polythieno[3,4-b]thiophenebenzodithiophene (PTB7) is studied. Further we demonstrate that three dimensional surface engineering of Si-ncs by low-cost and room temperature DC atmospheric microplasma processing in ethanol considerably enhance the Si-ncs electronic interactions with polymers and enhance the overall external quantum efficiency conversion of bulk heterojunction solar cells without using any surfactant.

We have reported the fabrication of amorphous carbon nitride, a-CN_x, films with a graded band gap structure to improve their photoconductivity and their structural, optical, and electrical properties are studied. In this study, two different structured a-CN_x films with graded band gap structures were compared with single-layered a-CN_x films. One of the graded band gap structure films consists of four stacked layers with different band gaps (multilayered film), and the other film consists of one layer with a gradually changed band gap (graded-layered film). All of the a-CN_x films are prepared by the reactive radio frequency magnetron sputtering method using a graphite target in pure nitrogen gas. The ratio of photo- and dark-conductivity, σ_p/σ_d, of the multilayered film is quite low. In contrast, the σ_p/σ_d value of the graded-layered a-CN_x film is about 11 times larger than that of the single-layered film and 80 times larger than that of the multilayered film.

Bulk heterojunction solar cells were fabricated by blending semiconducting cadmium selenide nanoparticles with various capping ligands and regioregular poly(3-hexylthiophene). The effects of surface ligand modification of CdSe nanoparticles and thermal treatment of fabricated cells on the device performance were investigated. It was found that surface ligands of nanoparticles could affect the device performance by increasing the charge carrier separation at the nanoparticle/polymer interface by quenching photoluminescence. Thermal treatment of fabricated cell structure at 140 °C was found to be optimal for device performance, resulting in a maximum power conversion efficiency of 1.17% under AM1.5G simulated solar irradiation.

Antimony-doped tin oxide (SnO₂:Sb, ATO) films have been deposited on glass substrates using atmospheric pressure chemical vapor deposition (APCVD) method. The precursors are mixed with SnCl₄, SbCl₅, and O₂ to prepare the films. This study used synchrotron grazing incidence X-ray diffraction (GIXRD) to investigate the film microstructure. Our results show that the precursors of chlorine ions were involved in the doping mechanism, causing the microstructure of films to change slightly. The film has an average transmittance between 85.8 and 82.1% within a visible spectral range from 400 to 800 nm. The minimal resistivity was 6.1×10^-4 Ω cm after doping. The synchrotron GIXRD data show that the chlorine ions caused the lattice constant change. A possible mechanism was proposed to explain the enhancement in electrical property due to chlorine dopants.

The effect of a zinc oxide (ZnO) interfacial layer on the device performance of polymer solar cells based on poly(3-hexylthiophene) (P3HT) and [6,6]-phenyl-C₆₁-butyric acid methyl ester (PCBM) was studied. When a 15 nm ZnO layer was inserted between the active layer (P3HT:PCBM) and Al electrode, the power conversion efficiency (PCE) was increased by approximately 10% as compared to that without ZnO layer. From the systematic analysis of the current–voltage and impedance characteristics of devices with and without the ZnO layer at various temperatures and light intensities, we found that the device with the ZnO layer has a lower activation energy for carrier extraction and bimolecular recombination loss compared with that without the ZnO layer. Therefore, the enhanced efficiency of the device with ZnO is mainly attributed to the enhanced extraction efficiency and thereby increased short-circuit current density.

Electrospray-deposited poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) was investigated using several solvent solutions as a hole-transporting layer for spin-coated poly(3-hexylthiophene-2,5-diyl):[6,6]-phenyl-C₆₁ butyric acid methyl ester (P3HT:PCBM) mixture organic thin-film solar cells (OSCs). Among them, the uniform deposition of PEDOT:PSS films was realized using 70% N,N-dimethylformamide solvent solution. The P3HT:PCBM solar cells with electrospray-deposited PEDOT:PSS as a hole transport layer showed a relatively high performance with an open-circuit voltage V_oc of 0.47 V, a short-circuit current J_sc of 8.9 mA/cm², a fill factor FF of 0.48, and a power conversion efficiency of 2.0%, which were almost the same as those with a spin-coated PEDOT:PSS layer. These findings imply that the electrospray-deposited PEDOT:PSS is a possible material for use as a hole transport layer for conjugated polymer-based OSCs.

Photovoltaic (PV) devices using blend films of poly(methyl phenyl silane) (PMPS) and fullerene (C₆₀) are fabricated, and the effects of various technical parameters during device fabrication on the PV characteristics under AM 1.5G simulated solar light illumination of 100 mW/cm² intensity are investigated. The PV performance of the devices depends on the mixing ratio of PMPS:C₆₀, the kind of buffer layer, the thickness of the buffer layer, and the thermal annealing temperature of the devices. The devices optimized for the mixing ratio, kind of buffer layer, and thermal annealing temperature exhibit a short-circuit current density (J_SC) of 2.26 mA/cm², an open circuit voltage (V_OC) of 0.71 V, a fill factor (FF) of 0.35, and a power conversion efficiency (PCE) of 0.57%.

The energy level alignment of C₆₀/bathocuproine (BCP)/Ca interfaces as a function of BCP layer thickness has been studied by ultraviolet photoelectron spectroscopy. The results show that the energy level alignment is very sensitive to the BCP layer thickness. The energy levels of the lowest unoccupied molecular orbital of C₆₀ and BCP were found to be almost the same when the thickness of BCP interlayer is less than 1.6 nm. Such energy level alignment, which is favorable to the carrier transport, may be caused by an interaction between C₆₀ and Ca, where Ca passed through the BCP interlayer and diffused to the C₆₀ layer. The role and optimum design for BCP interlayer were discussed.

The interaction of hydroperoxyl radical (OOH) with a graphene surface has been investigated by means of density functional theory (DFT) method in order to elucidate the radical scavenge mechanism of graphene surface. The OOH radical is highly reactive and the radical plays an important part of materials chemistry. The DFT calculation showed that the OOH radical binds to the carbon atom of graphene surface and a strong C–O bond is formed. The binding energies were dependent on the cluster size and were distributed in the range 18–25 kcal/mol at the B3LYP/6-31G(d) level of theory. The potential energy curve plotted as a function of C–OOH bond distance showed that the OOH radical approaches to the carbon atom with an activation barrier (the barrier height is distributed in 20–25 kcal/mol). Also, it was found that structural change from sp² to sp³-like hybridization occurs by the approach of OOH.

The interaction between fluorinated ethylene carbonate denoted by EC(F) and a graphene surface was investigated using of density functional theory (DFT) method. The interaction system examined was a complex composed of graphene (consisting of 14 benzene rings) and one EC(F) molecule. Ten binding sites of EC(F) binding site on the surface and edge regions of the graphene, were identified as stable points. EC(F) bound to a hexagonal position corresponding to the central of benzene ring on the graphene surface and can also bind to the edge of the graphene. The EC(F) binding energies on the surface and edge sides were 0.5 and 2.8 kcal/mol, respectively. The activation barrier for the diffusion of EC(F) on the graphene surface was significantly low (less than 0.3 kcal/mol), indicating that EC(F) can move freely on the graphene surface.

Hole transport and other fundamental properties of defect fullerenes C₅₉ and C₆₉ were investigated using density functional theory calculations. C₅₉ and C₆₉ isomers without a four-membered ring and three neighboring five-membered rings are generally stable. Formation of a carbon vacancy in C₆₀ and C₇₀ slightly increases the highest occupied molecular orbital energy and greatly decreases the lowest unoccupied molecular orbital energy, so that the energy gap decreases by 1 eV. The reorganization energies of all defect fullerenes are larger than those of the original C₆₀ and C₇₀ because of the localization of injected carriers around the vacancy. The reorganization energy of defect fullerenes is closely related to relaxation of the C–C bond of unsaturated C atoms.

We demonstrate high-efficiency hybrid solar cells based on heterojunctions formed between n-type silicon nanowires (SiNWs) and p-type organic semiconductors fabricated using a simple solution-based approach. Two types of devices have been fabricated with different organic materials used, namely poly(3,4-ethylene dioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) and a small molecule, 2,2',7,7'-tetrakis(N,N-di-4-methoxyphenylamino)-9,9'-spirobifluorene (Spiro-OMeTAD). The cells are characterized and compared in terms of their physical characteristics and photovoltaic performance. Using SiNWs of the same length of 0.35 µm, it is found that the SiNWs/Spiro cells exhibit a power conversion efficiency of 10.3%, which is higher than the 7.7% of SiNWs/PEDOT cells. The results are interpreted in terms of the ability of the two organic semiconductors to fill the gaps between the SiNWs and the optical reflectance of the samples. The degradation of the SiNWs/Spiro cells is also studied and presented.

The uncertainty analysis of irradiance and temperature measurements in relation to the energy yield prediction of the photovoltaic (PV) modules are presented. A Monte Carlo simulation approach is demonstrated separately to propagate the monthly and annual measurement uncertainties of irradiance and temperature to annual energy yield prediction uncertainty for two commercially available PV modules. The annual irradiation uncertainty as measured with a thermopile pyranometer is calculated as ±1.56%. Uncertainty of the annual average of ambient temperature measurement is calculated as ±0.08 °C. Finally, the uncertainty in the energy yield estimation of the PV devices is determined as 2.8 and 15.5% for crystalline silicon (c-Si) and copper indium gallium (di)selenide (CIGS) modules, respectively.

It is useful to understand the effect of irradiance, temperature, and spectral irradiance on the photovoltaic cell and module performance, to estimate their current–voltage curves under various climate conditions for power rating and energy rating. In this paper, we present the accuracy of the three methods described at the IEC 60891 and investigate the applicability of these methods. Procedure 3 showed good agreements between the calculations and experiments within ±1%. We have also demonstrated that the accuracies of procedures 1 and 2 depend on the device structure. Although both methods can introduce larger deviations at lower irradiance level, these deviations are not negligible for energy yield prediction.

We have studied the power reduction rate and degradation modes of crystalline-silicon photovoltaic modules, which had been operated in an outdoor field for about 10 years. We measured the electric parameters of the modules before and after 10-year operation, and classified our photovoltaic modules into four groups that were manufactured by four manufacturers. Analyzing the age-related decreases in electric parameters of the modules, we found two dominant degradation modes by correlation analysis of the electric parameters. In addition, we found delamination mode degradation by visual inspection. We have recognized a degradation rate in terms of the statistical distribution of the module power reduction. The power reduction rate was found to be an average of 4.7% for ten years.

Photovoltaic modules with single-crystalline silicon cells and multi-crystalline silicon cells, which were exposed outside for over nine years from 1992, have been evaluated for their failures. Current–voltage characteristic, electroluminescence, and thermography (dark mode) measurement of the modules were carried out as well as those of each solar cell in the modules. Application of these measurements in combination has been shown to be beneficial for investing the failure positions and failure factors in the modules. A new method of analyzing the positions of interconnection failures in modules was also adopted in the present study that is by using electroluminescence measurement and by connecting wires to the interconnector of each individual solar cell.

Various photovoltaic (PV) technologies are commercially available today. The effect of solar spectrum on the performance of PV modules should be evaluated quantitatively in order to estimate the module performance with high accuracy and precision. Average photon energy (APE) has been frequently applied to evaluate the effect of solar spectrum. The purpose of this study is to enhance the precision and accuracy by introducing other indexes. In this study, we select solar spectra, the integrated spectral irradiance (ISI) and APE of which are equivalent to those of the standard AM1.5G spectrum. There is a slight difference in shape, although the shapes are approximately similar. We introduce one more index, which defines the spectral irradiance at the atmospheric window or the depth of the water absorption band. The introduction would further improve the accuracy and precision of the evaluation of the effect of solar spectrum.

Solar spectral models applicable to partly cloudy and overcast skies as well as cloudless skies are developed by the analyses of standing spectral observation data in Japan. To characterize solar spectral irradiance in detail by decomposing a global spectrum into a direct and diffuse spectrum, not only global spectra but also diffuse spectra have been observed. The average photon energies of the diffuse spectra were changed largely by the cloud cover in contrast to the little dependence in the direct spectra, which confirms the importance of separate observations. Spectral models describing the effects of clouds on the direct and diffuse spectra were derived; these models were demonstrated by regression analyses of the observations. Additionally, we have revealed a relationship associating Rayleigh scattering intensity in diffuse irradiance with the diffuse broadband irradiance and the solar altitude, which permits us to estimate the model parameter from the available parameters.

We used several analytical methods to identify the mechanism underlying the performance degradation in a photovoltaic (PV) module subjected to long-term (10 years) field exposure. Cloudy visual defects in this module were caused by delamination between the poly(ethylene vinyl acetate) (EVA) and antireflection coating films on the Si substrate. The delamination was considered to be caused by the formation of a segregation layer and oxidative degradation of EVA. Furthermore, it was found that sodium ions diffused from the superstrate glass into the EVA film and Si cell. We confirm that diffusion of sodium ions caused the degradation of Si cells and the superstrate glass of this module.

A control method to optimize the output power of a solar cell is necessary because the output of a solar cell strongly depends on solar radiation. We here proposed two output power control methods using the short-circuit current and open-circuit voltage of a solar panel. One of them used a current ratio and a voltage ratio (αβ control), and the other used a current ratio and a short-circuit current–electric power characteristic coefficient (αγ control). The usefulness of the αβ and the αγ control methods was evaluated. The results showed that the output power controlled by our proposed methods was close to the maximum output power of a solar panel.

The output energies of multicrystalline and amorphous silicon/microcrystalline silicon photovoltaic (PV) modules were estimated using the performance contour map as a function of module temperature and solar spectral irradiance distribution taking into consideration time-related degradation. Generally, the performance of outdoor PV modules degrades yearly. When degradation is not considered, the error of the estimated output energy tends to increase yearly. To estimate the precise output energy, the degradation rates of the performance of the PV modules was considered. The yearly degradation rates were estimated using the measurement data from the years 2005–2009 by linear approximation, and the performance contour map in 2010 was estimated. The estimated contour maps were similar to the actual ones, and the errors between the estimated and actual maps were decreased. The error between the estimated and actual energies was within 2.5% in the PV modules. The results indicated that the method of using the contour map taking into account the yearly degradation is useful for estimating the output energy of the PV modules after outdoor exposure.

The output energy estimation of photovoltaic (PV) modules was carried out using contour maps as a function of average photon energy (APE), which is an index of solar spectral irradiance and module temperature (T_mod) in our previous report. This method is useful for not only crystalline Si PV modules but also thin film Si PV modules, which have a narrow spectral response. However, this method requires spectrum measurement to obtain APE using a spectroradiometer, which is not widely installed in various locations. In this study, we used clearness index (Kt) and air mass (AM), which have a strong correlation with APE and T_mod. Kt and AM can be calculated easily from irradiance, time, and location. The comparison between the output energy estimation using Kt and AM and that using APE and T_mod was carried out. Results indicate that the accuracy of the output energy estimation using Kt and AM is as high as that using APE and T_mod. The proposed method using Kt and AM is useful for estimating the output energy of the PV modules.

The tropical regions will see an increasing share of solar photovoltaic (PV) applications. This paper therefore analyses the irradiance distribution in tropical Singapore and then experimentally determines the dependence of the PV efficiency on irradiance level, for nine different commercially available PV module technologies. For several of the PV module technologies, we also determine the dependence of the temperature coefficient of the modules' maximum power on the irradiance level. A full year of outdoor module testing data in Singapore show that the irradiance distribution has two energy peaks, one at around 400 W·m^-2 and the second at around 850 W·m^-2. Most PV technologies cannot fully convert the second peak due to the fact that, in Singapore, high light intensities are always associated with higher module temperatures, which in turn reduces the module efficiencies.

The reliability of photovoltaic (PV) modules is related to the ingress of moisture in some cases. We investigated the measurement method of moisture ingress into PV modules. In order to detect the moisture ingress route into the module, cobalt chloride (CoCl₂) paper was used. The change in the color of CoCl₂ paper is effective in detecting and quantifying moisture ingress. The results suggested that the main route of moisture ingress is along the back material and moisture gradually diffuses to the center of the cell. The rate of moisture ingress into the PV module depends on the water–vapor transmission rate (WVTR) of the back material. The amount of moisture estimated from a calibration curve is correlated to the amount of moisture calculated from the WVTR of the back material.

To accelerate the degradation with thermal fatigue in crystalline silicon photovoltaic modules, the modules were exposed to dry thermal stress with rapid thermal cycling, and module impedance was monitored in situ during this testing. The spikelike increase in module impedance at a temperature-alteration point was observed in the early stage of this rapid thermal cycling. The pattern of increase in module impedance proceeded step-by-step, from the early stage, via the double-spikelike pattern at two temperature-alteration points (the middle stage), and finally to the successive increases in module impedance in the high-temperature period (the late stage). The nondestructive analyses suggest that the interconnector failures without the defects of photovoltaic cells occurred. From these results, it is suggested that the pattern of increase in module impedance is related to the interconnection degradation of modules, and that the rapid thermal cycling with in situ monitoring of module impedance would be a useful procedure for the earlier detection of interconnection failures in photovoltaic modules.

See pdf for details

See pdf for details