Table of contents

This paper reviews the main research results related to PERC+ silicon solar cells. Compared to today's industry typical passivated emitter and rear cell (PERC) silicon solar cells with full-area rear aluminum layer, PERC+ solar cells apply an aluminum finger grid on the rear side and hence are able to absorb diffuse light from the rear side in addition to the direct sunlight which is absorbed from the front side. This bifaciality increases the energy yield of silicon solar modules by up to 25%. Since its first publication in 2015, the PERC+ cell concept has been rapidly adopted by several solar cell manufacturers due to the very similar process technology of bifacial PERC+ cells and main stream monofacial PERC cells. We summarize technological challenges, published PERC+ conversion efficiencies and PERC+ module technologies. First energy yield data of PERC+ field installations demonstrate the high energy yield potential of PERC+ solar cells.

In this study, we evaluated the effects of carbon and annealing process conditions on the oxygen precipitation in an n-type Czochralski (Cz) wafer for solar cells by infrared light scattering tomography (IR-LST). It was confirmed that precipitates grow larger and denser as carbon concentration increase. Thus, carbon promotes oxygen precipitation. We also evaluated the effect of oxygen precipitation on the minority carrier lifetime by photoluminescence (PL) imaging. It was confirmed that the interface between the precipitate and the Si matrix is a dominant recombination center since the surface area of the precipitate obtained by IR-LST measurement and the PL intensity show good correlation. In addition, it was also confirmed that carbon is involved in the supply of interstitial oxygen to precipitates through the formation and extinction of the thermal donor. We believe that it is important to understand and control the effect of carbon of controlling the oxygen precipitation behavior.

In the silicon wafers, interstitial oxygen and vacancies make oxygen precipitation in the cell process. The oxygen precipitates become combination centers for minority carriers, and it is resulted in decrease of minority carrier life-time (MCLT) and cell efficiency. In order to grow ingots with low oxygen concentration, we locally optimized hot-zone in the ingot grower. We have designed new hot-zone for commercial 8.3-in. diameter ingot growth using 24-in. quartz crucible and old grower. The average oxygen reduction is about 3 ppma at the top position of ingot, which is similar to the simulation result of oxygen distribution. Despite the inside temperature of grower rises, the concentration of oxygen was suppressed to 15 ppma or less as in a normal heater. As a result, it is possible to grow a low-oxygen ingot that can suppress the light induced degradation (LID), which can fundamentally help improve the efficiency of a high efficiency passivated emitter and rear cell (PERC).

We propose a way of replacing silicon dioxide (SiO₂) films in tunnel oxide passivated contact (TOPCon) solar cells by ultra-thin Si nitride (SiN_x) films. We deposit SiN_x films on n-type crystalline Si (c-Si) wafers by catalytic chemical vapor deposition (Cat-CVD), by which we avoid a plasma damage to the surface of c-Si. Thin (<5 nm) SiN_x films can be deposited with good controllability by tuning the deposition conditions. To improve the passivation quality of SiN_x films, hydrogen treatment was performed onto the SiN_x-coated c-Si surfaces. Their effective minority carrier lifetime (τ_eff) can be improved up to >1 ms by the hydrogen treatment for the samples containing SiN_x with a proper refractive indices and >10-nm-thick n-type amorphous Si (n-a-Si). The ultra-thin SiN_x films have sufficiently high passivation ability and have the required level for the passivation layers of rear-side contact in the TOPCon-like c-Si solar cells.

A high-performance light-trapping structure for Si was fabricated with an etching margin of only ∼1 µm using Ge islands grown by gas-source molecular beam epitaxy as etching masks. KOH solution containing isopropyl alcohol and HF + H₂O₂ + CH₃COOH mixed solution were used as etchants. The reflectance of the structure was shown to be comparable to that of a conventional pyramid texture, which requires a larger etching margin of ∼10 µm. In addition, a potential short-circuit current density (p-J_sc) of 42.3 mA/cm² was obtained for the sample after the deposition of indium tin oxide, which confirms that the light-trapping structure is applicable to crystalline Si solar cells with thinner wafers.

We investigated tunnel recombination junction (TRJ) structures using p-type nitrogen-doped cuprous oxide (Cu₂O:N) and n-type silicon-based thin films for perovskite/heterojunction crystalline silicon monolithic tandem solar cells. The TRJ structure using n-type hydrogenated amorphous silicon showed non-ohmic behavior. On the other hand, the TRJ structure using n-type hydrogenated microcrystalline silicon (µc-Si:H) showed ohmic behavior with sufficiently low contact resistance. This indicates that the p-type Cu₂O:N/n-type µc-Si:H junction is promising for the TRJ in perovskite/heterojunction crystalline silicon monolithic tandem solar cells.

We demonstrated the effectiveness of photoluminescence (PL) measurement at liquid N temperature after electron irradiation for the determination of the C concentration in P-doped n-type Czochralski-grown Si crystals. The disappearance of P-related lines simplifies the spectral analysis at 77 K, enabling us to estimate the C concentration from the G-line intensity ratio regardless of the difference in P concentration. The C concentration estimated by PL measurement at 77 K was in good agreement with those by measurement PL at 4.2 K and IR absorption. Unsusceptibility to the concentration of dopant impurities is a practical advantage of the PL measurement at 77 K over that at 4.2 K.

We confirmed the presence of non-bonded hydrogens (NBHs) in hydrogenated amorphous silicon (a-Si:H) films, using a combination of multiple techniques (Rutherford backscattering spectrometry/hydrogen forward scattering, Fourier-transform infrared spectroscopy-attenuated total reflection, and thermal desorption spectrometry). The hydrogen effusion profile of an a-Si:H film with large amounts of NBHs was analyzed in detail. We report the effect of NBHs on band structure and electrical conductivity, and we present additional considerations for previous data on number density of silicon, optical bandgap, and vacancy size distribution [J. Non-Cryst. Solids 447, 207 (2016)]. The effect of NBHs on the a-Si network is explained by the "dense restructuring model".

We describe a multi-diamond-wire saw for cutting monocrystalline silicon bricks into thin (120 µm) and thick (200 µm) wafers and label as fresh- and worn-wire sides. While almost no difference was found in the fracture stress of the thick (200 µm) wafers cut from either side, the thin (120 µm) wafers showed a lower fracture stress in those from the fresh-wire side compared to the worn-wire side. This is a remarkable result when wafers are sawn with conventional diamond wire. On the contrary, wafers sawn with improved diamond wire (100d-M6/12) showed a higher fracture stress compared to those cut with conventional diamond wire (100d-M8/16), for both the fresh- and worn-wire sides. Observing the subsurface areas of wafers by micro-Raman spectroscopy, we succeeded in quantifying the defective silicon fraction as the Raman crystallinity factor (Φ_c). We found that wafers having a higher fracture strength had a larger Φ_c.

This paper presents the efficiency distribution of an n-type bifacial solar cell. In this work, the solar cells were produced by our standard cell fabrication process, which is very similar to the industrial line process. The best cell efficiency was 20.4%, and the average efficiency was 20.06%. The cell efficiency ranged between 19.9 and 20.4% and showed a narrow efficiency distribution. This paper also presents the development of a boron selective emitter (p⁺/p⁺⁺) n-type bifacial solar cell. In this work, the cells were fabricated using the standard cell fabrication process based on tube-furnace thermal diffusion employing liquid sources: BBr₃ for the front-side boron emitter and POCl₃ for the rear-side phosphorus back-surface field (BSF). The p⁺/p⁺⁺ structure was formed by screen-printing resist masking and wet chemical etching technology. Both the front- and rear-side electrodes were obtained by using screen-printed contacts with H-patterns. The cell efficiency with the selective boron emitter was almost the same as that with the homogeneous emitter; however, the V_oc of the selective emitter solar cell was higher than that of the homogeneous emitter solar cell by 4.4 mV. Finally, we performed the recombination analysis of the completed cell.

The impact of Si wafer thickness on the photovoltaic performance of hydrogenated amorphous silicon/crystalline silicon (a-Si:H/c-Si) heterojunction solar cells was examined from the optical and electrical points of view. Optical characterization of c-Si wafers of various thicknesses showed that a realistic light-trapping scheme, i.e., pyramidally textured Si wafers with a dielectric antireflection coating and a back reflector, realizes an efficient quasi-Lambertian light absorption enhancement, even for very thin wafers. This indicates that high photocurrent densities are achievable by using the realistic light-trapping scheme, assuming that the parasitic absorption loss is minimized. The potentials of open-circuit voltage (V_OC) and the fill factor (FF) of thin c-Si cells were investigated using thin c-Si wafers passivated with intrinsic/doped amorphous silicon film stacks. It was experimentally confirmed that the implied V_OC increases steadily with decreasing wafer thickness down to 30 µm, while the implied FF weakly depends on the thickness. As a result of the trade-off between light absorption and implied V_OC, a high implied efficiency is expected for a wide range of wafer thicknesses. The V_OC increase by thinning the wafer was also experimentally confirmed in an a-Si:H/c-Si heterojunction cell with a thickness below 60 µm, resulting in a conversion efficiency of 21.0%.

Electrochemical processing can potentially be used for the cost-effective and large-scale fabrication of solar cells based on nanostructured Si thin films. Herein, we fabricated Si thin films for solar cell applications by electrodeposition in ionic liquids and performed a systematic investigation of film structure control and doping effects. Notably, relatively smooth and compact films were obtained by the modulation of the rest time during the electrodeposition under light irradiation, and the addition of AlCl₃ to the electrolyte allowed the electrodeposition of p-type Si thin films.

We report on our attempt to increase the carrier density in the p-type polycrystalline silicon (poly-Si) layers grown by aluminum-induced crystallization (AIC) employing a B-doped Si target. Hall measurement revealed that we could obtain a lower-resistance and heavily doped p-type continuous poly-Si thin layer formed by AIC using B-doped a-Si. According to our evaluation of tunnel oxide passivated contact (TOPCon) solar cells with AIC-grown poly-Si, the AIC-TOPCon solar cells fabricated at 570 °C using B-doped a-Si showed higher conversion efficiency of 13.5% than that of 12.8% when using nondoped a-Si. It is considered that the cell characteristics, particularly FF and V_oc, were improved owing to the lower series resistance and higher carrier density since both Al and B were successfully incorporated into the AIC-grown poly-Si.

To reduce optical loss and contact material cost in silicon heterojunction (SHJ) solar cells, copper plating has been considered as a suitable metallization technique. However, a plated copper contact on indium tin oxide (ITO) generally has low adhesion reliability. For this reason, suitable seed layer materials are required for adhesive copper plating. As a requirement of a suitable seed layer material, contact resistance between the seed and the ITO is also important, as well as the adhesion, since a high series resistance results in a low fill factor. In this work, we applied copper alloy (Cu–X) materials as a seed layer that was deposited by co-evaporating copper with other metals. The contact resistivity (ρ_c) values of the Cu–X seed layer and copper-plated contact with the Cu–X seed layer (Cu–X/Cu) on the ITO layer were evaluated by the transfer length method (TLM). Also, tape tests were carried out to check the adhesion of the Cu–X/Cu contact.

Impact of the ultralow-carbon concentration of below 1 × 10¹⁵ atoms/cm³ and thermal history on the bulk lifetime of phosphorus (P)-doped magnetic-field-applied Czochralski (MCZ) silicon was investigated. In order to accurately measure the long bulk lifetime, a direct-current photoconductive decay method was applied to a rectangular sample with a sectional area larger than 400 mm². The measurement of an ultra-long bulk lifetime of longer than 20 ms was demonstrated using the P-doped MCZ silicon with a carbon concentration of approximately 5 × 10¹⁴ atoms/cm³. Furthermore, the bulk lifetime of MCZ silicon with shorter exposure time at 300–600 °C was extremely short; we believe that the formation behavior of the carbon-related defects in the crystal growth process affect the bulk lifetime. Therefore, not only reducing the carbon concentration but also improving the thermal history at low temperature is effective for increasing the bulk lifetime of as-grown P-doped MCZ silicon.

In this work, we describe some aspects of the Hanergy silicon heterojunction (SHJ) solar cell design and its manufacturing-friendly process. Experimental results are reported mainly with regard to texturing, silicon-based thin film deposition, and transparent conductive oxide (TCO) coating optimization. A conversion efficiency of 22.83% with V_OC = 737.6 mV, J_SC = 37.97 mA/cm² and FF = 80.94% on a full-size M2 wafer has been certified by Fraunhofer ISE CalLab. Also demonstrated is a mean efficiency of 22.15% during a 45-day pseudo-pilot run. Two PV stations built with our bifacial 60-cell standard SHJ modules harvested 12.1 and 18.3% more solar energy in Chengdu than a single-sided n-type monocrystalline silicon reference PV station, mainly owing to improvements of the rear side.

In this article, we describe an n-type silicon solar cell with a conversion efficiency of 21.9% on a 156 mm pseudosquare Czochralski (Cz) wafer with a rear tunnel oxide-passivated structure and rear screen-printed electrodes. The thermal degradation of the rear structure during electrode firing was solved by adding a SiN_x film on the rear surface. Not only the SiN_x film but also the tunnel oxide structure was chemically etched using fire-through screen-printed electrodes, so that the electrodes contacted directly to the bulk silicon. A phosphorus-diffused layer appeared at the bulk silicon immediately below the tunnel oxide, and this shallowly and densely diffused layer reduced the contact resistivity between the electrodes and the bulk silicon layer. Degradation of rear passivation caused by etching on the tunnel oxide structure was improved by adding a diffused layer using POCl₃ before chemical oxidation. The diffused layer using POCl₃ was deep and had a low phosphorus concentration, thus the field effect worked successfully while Auger recombination was not enhanced.

We have developed high-quality surface passivation films for n-type crystalline silicon (c-Si) with an effective minority carrier lifetime (τ_eff) of more than 6 ms or a maximum surface recombination velocity (SRV_max) of less than 2 cm/s using two highly optically transparent silicon nitride (SiN_x) layers and an ultrathin SiO_x film. First, an ultrathin SiO_x film is unintentionally formed on the surface of c-Si by annealing c-Si in N₂ ambient at 350 °C. Then, on the SiO_x film, two SiN_x layers are sequentially formed at substrate temperatures (T_sub) of 100 and 250 °C by catalytic chemical vapor deposition (Cat-CVD), often called hot-wire CVD. With the combination of the ultrathin SiO_x film and the two SiN_x layers, we can obtain a novel passivation structure with high-quality surface passivation, high optical transparency, and high chemical resistance for subsequent processes.

The authors discuss industry related approaches at Fraunhofer ISE for bifacial p-type silicon solar cells, taking into account the well-known "passivated emitter and rear cell" (PERC), "passivated emitter and rear totally diffused" (PERT) and "passivated emitter and locally diffused" (PERL) architectures. In the case of PERC, challenges in terms of alignment, printability and the importance of bifaciality are addressed. In the case of PERT, a co-diffusion process is utilized to form the emitter and the back surface field simultaneously avoiding also critical shunts that can arise at the edges of such devices. For the PERL technology, the industrial feasible pPassDop approach is discussed. We report on front side energy conversion efficiencies for PERC of 21.4%, PERT of 20.5%, and PERL of 19.8%. Furthermore, bifaciality factors for PERC of 0.7, for PERT of 0.86, and for PERL of 0.89 are presented.

We investigated the distributions of interstitial oxygen (O_i) and substitutional carbon (C_s) in high-performance (HP) multicrystalline Si (mc-Si) and monocrystalline-like Si (mono-like Si) and compared them with those in conventional mc-Si, grown using the same furnace. The O_i concentration in mono-like Si grown using a Czochralski (Cz) silicon seed was the highest among the three crystals. On the other hand, the O_i and C_s concentrations in HP mc-Si grown using Siemens Si incubation seeds were the same as those in conventional mc-Si. Therefore, it is considered that O_i incorporated into the growing Si crystal originates not only from the quartz crucible wall but also from the seed. Additionally, O_i and C_s in HP mc-Si grown on the incubation seeds with adequately low O_i and C_s concentrations are distributed similarly to those in conventional mc-Si grown under the same conditions. We believe that it is important to consider the O_i and C_s concentrations in the feed stock materials both for the seed and whole ingots in the seed-casting method.

High-efficiency back-contact heterojunction crystalline Si (c-Si) solar cells with record-breaking conversion efficiencies of 26.7% for cells and 24.5% for modules are reported. The importance of thin-film Si solar cell technology for heterojunction c-Si solar cells with amorphous Si passivation layers in improving conversion efficiency and reducing production cost is demonstrated. Our attempts to reduce the production cost of a heterojunction c-Si solar cell by applying a SiO_x layer prepared by a plasma-enhanced CVD method are presented. The characteristics of heterojunction c-Si solar cells are clarified by comparing them with those of practical homojunction solar cells, and crucial targets for industrialization of back-contact heterojunction c-Si solar cells are discussed. Owing to the recent improvement of c-Si solar cells and perovskite solar cells, conversion efficiencies over 30% have become a realistic target by using a two-terminal tandem structure with a heterojunction c-Si solar cell and a perovskite solar cell.

To obtain high efficiency in crystalline silicon thin-film solar cells, it is necessary to develop a novel light-trapping structure that not only has high light absorptance and can be fabricated with a small etching margin but also has a small surface area for a small effect of surface recombination. In this study, optical properties of silicon nanowires (SiNWs) with submicron diameters were investigated by the finite-difference time-domain (FDTD) method in the infrared wavelength region, in which the absorption coefficient of crystalline Si is quite low and the increase in optical path length is very important. To verify the simulation results, SiNWs with submicron diameters were successfully fabricated by the metal-assisted chemical etching method using two types of etching masks. The optical properties and minority carrier lifetime of the obtained structures were measured, and the results suggest that SiNWs with submicron diameters such as ∼700 nm are effective for both light trapping and the suppression of surface recombination.

To reduce photovoltaic (PV) power generation costs, the reduction of PV module manufacturing costs, the improvement of cell and module efficiencies, and the long-term output power warranty of PV modules are necessary. Using the high-quality, low-cost crystal growth technology, we developed a high-quality casting process by the seed-cast method. Using a seed-cast wafer, an efficiency of 20.54% has been obtained with passivated emitter and rear cells (PERCs). The ohmic contact degradation of front electrodes and potential-induced degradation are the typical modes that significantly affect the module lifetime. Considering the field stresses induced by ultraviolet light, heat and humidity (H&H), and the electrical potential difference, sequentially combined stress tests of small modules and additional stress tests of field-aged modules are performed. The test results for the modules fabricated using our technology indicated that the modules have a sufficiently long lifetime of more than 30 years and are potential-induced degradation (PID)-free under these stresses in Japanese domestic environments.

We investigated the interface phenomena between the electrode by a low fire-through electrode paste and the silicon substrate in a crystalline silicon solar cell. Oblique polishing was used to evaluate the chemical bonding states in the depth direction. From the evaluation by X-ray photoelectron spectroscopy (XPS) and the hard X-ray photoelectron spectroscopy (HAXPES), we found changes in the chemical bonding states of Ag 3d, Si 2p, O 1s, and N 1s at the electrode–silicon interface. In addition, it was possible to confirm the reduction of Ag induced by the heat treatment at the interface. Thus, it became possible to evaluate the chemical bonding states at the interface between the electrode and the silicon substrate by adapting oblique polishing and XPS measurement.

Cu₂ZnSn(S,Se)₄ (CZTSSe) thin film photovoltaic absorber layers are fabricated by selenizing Cu₂ZnSnS₄ (CZTS) nanoparticle thin films in a selenium rich atmosphere. The selenium vapor pressure is controlled to optimize the morphology and quality of the CZTSSe thin film. The largest grains are formed at the highest selenium vapor pressure of 226 mbar. Integrating this photovoltaic absorber layer in a conventional thin film solar cell structure yields a champion short circuit current of 37.9 mA/cm² without an antireflection coating. This stems from an improved external quantum efficiency characteristic in the visible and near-infrared part of the solar spectrum. The physical basis of this improvement is qualitatively attributed to a substantial increase in the minority carrier diffusion length.

Polycrystalline CuGaSe₂ thin-films with various Cu-contents were deposited by three-stage co-evaporation method. Radiative transitions in the CuGaSe₂ were studied by micro-photoluminescence technique in relation to the Cu/Ga ratio in the films. In addition to a peak DAP1 around 1.60 eV, a peak DAP2 at relatively lower energy of 1.35 eV and a broadened deep-level emission DL roughly centered at 1.05 eV were observed in the photoluminescence (PL) spectra of the CuGaSe₂ films. Excitation-power dependence along with temperature dependence of the PL suggests that dominant recombination mechanism due to donor–acceptor pair recombination at lower temperature becomes impurity-to-band transition at higher temperature for all the three peaks. A deep donor-level, roughly 630 meV below the conduction band, which corresponds to DL-emission has been confirmed, and was investigated further in relation to the compositional variation in the CuGaSe₂ films.

A Cu₂ZnSnS₄ (CZTS) thin film was fabricated by co-evaporation of its constituents (Cu, Zn, Sn, and S) and subsequent annealing in S-flux using molecular beam epitaxy system. Two Sn crucible temperatures were considered, 1,010 and 1,020 °C. The Raman spectrum of the sample exhibited the main peaks at 336 cm⁻¹ attributed to CZTS and 472 cm⁻¹ attributed to Cu_2−xS. The annealing temperature that yielded CZTS with a lower amount of Cu_2−xS was in the range of 380–420 °C. The peaks of the Cu_2−xS phase were not observed in the sample fabricated at the Sn crucible temperature of 1,020 °C and annealed at 420 °C. The suppression of the Cu segregation near the surface, yielded a suppression of the Cu_2−xS phase, as revealed by the Raman spectra. The increase of the Sn ratio, achieved by increasing the crucible temperature, contributed to the suppression of the Cu_2−xS phase and improvement of the CZTS crystalline quality.

In this paper, zinc oxide (ZnO) nanorods (Nrods)–polymer based bulk hetero-junction solar cells were fabricated employing an eutectic gallium–indium (EGaIn) alloy negative electrode coated using a brush-painting method. Devices were fabricated using the structure; indium tin oxide (ITO)/poly(ethylenedioxythiophene) doped with poly(styrene sulfonic acid) (PEDOT:PSS)/ZnO-Nrods+polymer/electron transport layer (ETL)/EGaIn. The overall device was operated under environmental conditions and compared with a device fabricated using thermally evaporated-aluminum as a negative electrode. Adding the ZnO (ETL) between the active layer (ZnO-Nrods+polymer) and the cathodes (Al or EGaIn) improved the performances of the investigated devices. Power conversion efficiencies of devices with different cathode electrodes (Al or EGaIn) were comparable.

We employ simulation based approach for enhancing the efficiency of Cu₂ZnSnS₄ (CZTS) based solar cells. Initial benchmarking of simulation with the experimentally reported solar cell in literature is performed by incorporating a suitable defect model. We then explore the effects of (a) conduction band offset (CBO) at CZTS/CdS junction, (b) back surface field (BSF) due to an additional layer with higher carrier density, and (c) high work function back contact. Efficiency is observed to improve by about 70% upon optimizing the above three parameters. We also observe that utilizing BSF in the configuration can reduce the high work function requirement of the back contact. A work function of 5.2 eV (e.g., using Ni), a BSF layer (e.g., using SnS), and a CBO of 0.1 eV (e.g., using ZnS) constitute an optimal configuration.

We investigated the effect of post annealing following CdS deposition on Cu₂ZnSn(S,Se)₄ solar cells fabricated from nanoparticles. The fill factor and short-circuit current density of Cu₂ZnSn(S,Se)₄ solar cells increase by applying post annealing. Raman spectra showed that the crystallinity of CdS was improved. Electron-beam-induced current measurement revealed that the carrier transport near grain boundaries was improved by post annealing. The improvement of CdS/Cu₂ZnSn(S,Se)₄ interfacial properties and carrier transport near grain boundaries causes decreases in ideality factor and series resistance, leading to the improvement of fill factor.

We present a comprehensive study concerning the light exposure of bare Cu(In,Ga)Se₂ (CIGSe) layers in ambient air and its impact on the electrical properties of the solar cells. With time-resolved photoluminescence (TRPL) a degradation of the minority charge carrier lifetime after air-light exposure of CIGSe layers is observed. This degradation is related to a light induced modification of the CIGSe surface and persists upon completion of the solar cells. A strong reduction in solar cell performance from ∼16% down to ∼13% reflects the degradation effect of the solar cell's absorber. The deteriorated solar cell performance can be recovered by light soaking of the complete solar cell at increased temperatures of 85 °C.

To understand the effect of bandgap grading on carrier recombination for Cu(In,Ga)Se₂ (CIGS)-based solar cells in detail, samples with different bandgaps at the CIGS surface were fabricated by changing the Ga/(Ga + In) (GGI) ratio from 0.4 to 0 at the third stage of the conventional three-stage growth process. Optoelectronic characterizations, such as photoluminescence, temperature-dependent open-circuit voltage measurement and light-intensity-dependent current–voltage measurement, indicate that the photo-generated carriers move rapidly towards the location of the bandgap minimum, and the carrier recombination occurs mainly at this location. From simulation using a one-dimensional solar cell capacitance simulator (SCAPS-1D), a single-grade sample with the smallest bandgap on the surface of CIGS showed high recombination current at the surface, while the location of the maximum recombination current moved from the surface to the bulk for double-grade samples. This study suggests that controlling the bandgap grading is one way of suppressing recombination at the interface in CIGS-based solar cells.

Copper antimony disulfide (CuSbS₂) is a promising candidate for solar absorber material owing to its high photoabsorption property and earth-abundant constituent elements. In this study, we fabricated CuSbS₂ crystals of various nonstoichiometric compositions and investigated their optical and electric properties for their applications in photovoltaic devices. Band gap energies of CuSbS₂ crystals thus-obtained were almost constant (ca. 1.5 eV) irrespective of their compositions. Hall-resistivity measurements exhibited that a CuSbS₂ crystal with the compositional formula of 1.03/1.00/1.86 showed the best properties among the samples prepared in this study: the resistivity, hole concentration, and mobility of the sample were 1.4 Ω cm, 6.0 × 10¹⁶ cm⁻³, and 7.4 cm² V⁻¹ s⁻¹, respectively. Although further optimization would be expected, the obtained properties are suitable for an absorber of photovoltaics.

Cu₂ZnSnS₄ (CZTS) thin films are highly suitable as the light-absorbing layer in solar cells. In this study, Cu, Zn, Sn, and S are co-evaporated on the Mo/soda-lime glass (SLG) surface at 320 °C and then annealed in sulfur flux at temperatures of 320–450 °C within the same molecular beam epitaxy system (one-stop process). The Raman spectroscopy of the coated surfaces shows that CZTS is formed with a secondary Cu_2−xS phase, where the relative concentrations of the phases strongly depend on the annealing temperature. With an increase in the annealing temperature above around 400 °C, the fraction of CZTS decreases and that of Cu_2−xS increases. The depth profile analysis shows that Cu segregates to the surface and Cu_2−xS is formed on the film. The solar cells fabricated on these samples show that photovoltaic performance is limited by the surface Cu_2−xS. The observed external quantum efficiency behavior is consistent with the degradation of carriers collected in the solar cell formed on the CZTS film owing to the presence of the secondary Cu_2−xS phase.

Cu₂SnS₃ (CTS) p-type semiconductors are expected to be applied as a light absorption material for low-cost thin-film solar cells due to their advantageous physical properties. The influence of sodium addition to CTS thin films was investigated by comparing Na-free CTS and Na-doped CTS fabricated on alkali-free glass substrates. Grain growth for Na-free CTS, which has a Cu/Sn composition ratio of approximately 2.0, did not occur below 570 °C. In contrast, the addition of sodium to the CTS increased the grain sizes with an increase in the annealing temperature. Even with Na-free and Na-doped CTS, the grain sizes increased with a decrease in the Cu/Sn composition ratio. These results show that an excess of Sn combined with the presence of sodium accelerate the grain growth of CTS. Photovoltaic cells using the Na-doped CTS with a Cu/Sn ratio of 1.81 exhibited an open-circuit voltage of 242 mV, a short-circuit current density of 26.5 mA/cm², a fill factor of 0.523, and a conversion efficiency of 3.35%. The cells using CTS without sodium did not exhibit good photovoltaic characteristics due to the small grain sizes.

Analysis on the temperature-dependent Raman spectra measured between 27–450 °C range in relation to various Ge compositional ratios of Cu₂Sn_1−xGe_xS₃ thin-film alloys, which were prepared by closed-tube sulfurization technique, was carried out in detail. By XRD and Raman analyses, linear transformation of the lattice parameters for Cu₂Sn_1−xGe_xS₃ thin-film alloys with respect to the Ge composition was confirmed to be obeying the Vegard's law. By AC Hall effect measurement, reducing of carrier mobility was observed in the Cu₂Sn_1−xGe_xS₃ thin-film alloys with higher Ge content. By the temperature dependent Raman spectroscopy analysis, it is observed that the red shifting of Raman vibration energy due to the thermal expansion is resembled to that of the temperature dependency of energy bandgap of semiconductors. Anisotropic thermal dilatation behavior was observed in the Cu₂Sn_1−xGe_xS₃ thin-film alloy independent of its Ge composition.

To clarify the mechanism of efficiency enhancement effect by potassium (K) treatment for Cu(In_1−xGa_x)(Se_1−yS_y)₂ (CIGSSe) solar cells, we investigated the physical properties of deep and shallow defect levels of CIGSSe solar cells without or with potassium fluoride treatment via using steady state photo-capacitance and admittance spectroscopy. The results show that the defect density of deep levels (such as 0.8 and 0.7 eV above valance band) are similar, suggesting that the K treatment does not drastically change the deep defect properties. For the shallow levels, the activation energy of interface-related shallow defect decreased from 183 to 120 meV upon K treatment, suggesting the modification of interface. The modified interface may be caused by the diffusion of potassium to grain boundaries and the formation of Cu-depleted layer in the near surface region of CIGSSe, resulting in the prolonged minority carrier lifetime. A new level with activation energy of 250 meV was found after K treatment, and it contributed to the increase of carrier density. These features well explained the enhanced device performance of CIGSSe solar cells by K treatment in terms of defects.

Cu₂ZnSnS₄ (CZTS) is one of the most promising alternatives for photovoltaic (PV) applications, since it is made from earth abundant and low toxicity materials. To assess the potential of CZTS PV cells in the future global market, it is useful to investigate the sustainability by assessing the possible environmental impacts of large-scale production. Life cycle assessment (LCA) of CZTS absorber layers fabricated by vacuum and non-vacuum deposition techniques has been studied using information from laboratory-scale processes as well as large scale estimations. Copper indium gallium diselenide (CIGS) was selected as a reference vacuum-based process. The paper compares the environmental impacts of the vacuum and non-vacuum processes by considering the three main damage categories of climate change, human health and eco-toxicity using Sima Pro software. The results show that the CZTS films fabricated with both deposition techniques exhibit a significant potential benefit over the CIGS thin films. The toxicity to human health was found to depend critically on the compounds and components selected for preparing the precursors. The energy consumption for processing the CZTS thin films via the non-vacuum deposition technique is considerably less than that required for the other techniques.

(Cu_1−xLi_x)In(S_1−ySe_y)₂ (CLISSe) samples with 0.0 ≤ x ≤ 0.4 and y = 0.25, 0.50, and 0.75 were prepared by a mechanochemical process and sequential heating. The X-ray diffraction (XRD) peaks of the (Cu_1−xLi_x)In(S_1−ySe_y)₂ shifted to the lower angle side by the substitution of Li atoms for Cu atoms. The solid solution limit of Li for Cu in (Cu_1−xLi_x)In(S_1−ySe_y)₂ system widened with increasing Se content, y. The single-phase (Cu_1−xLi_x)In(S_1−ySe_y)₂ samples were obtained for the compositions with 0.0 ≤ x ≤ 0.05 for y = 0.25, 0.0 ≤ x ≤ 0.20 for y = 0.50, and 0.0 ≤ x ≤ 0.30 for y = 0.75. The crystallographic parameters such as the lattice constants a and c, c/a, and the atomic coordinate of a S/Se atom for (Cu_1−xLi_x)In(S_1−ySe_y)₂ were refined by Rietveld analysis using XRD data. Both the lattice constants a and c of the (Cu_1−xLi_x)In(S_1−ySe_y)₂ with a tetragonal chalcopyrite structure increased with increasing Li content, x. The band gaps of (Cu_1−xLi_x)In(S_1−ySe_y)₂ solid solutions widened with increasing Li content, x. To understand the band diagram of the solid solutions, the energy level of the valence band maximum (VBM) from the vacuum level was determined from the ionization energy measured by photoemission yield spectroscopy (PYS). The energy level of the conduction band minimum (CBM) was also determined by adding the band-gap energy to the VBM level. The VBM level of the (Cu_1−xLi_x)In(S_0.25Se_0.75)₂ solid solution decreased considerably with increasing Li content, x. The CBM level of the (Cu_1−xLi_x)In(S_1−ySe_y)₂ solid solution was approximately constant. The Li doping in CuIn(S,Se)₂ is useful for decreasing the VBM of the CuIn(S,Se)₂ absorber and increasing the band-gap energy of the CuIn(S,Se)₂ absorber without increasing the CBM level.

We demonstrated the potential of Fourier transform photocurrent spectroscopy (FTPS) in investigating the optical absorption and quality of a Cu₂ZnSn(S,Se)₄ (CZTSSe) absorber within the cell structure. We fabricated thin-film solar cell structures with a CZTSSe absorber, which was prepared by a nanoparticle coating technique and a postannealing process. The optical absorption spectra of CZTSSe absorbers prepared at different annealing temperatures were measured by FTPS. We successfully evaluated the optical band-gap energy from the spectra. An exponential tail was found in the spectra. Urbach energy increased with increasing annealing temperature. The deterioration of photovoltaic performance with increasing Urbach energy might be due to potential fluctuation rather than to structural randomness.

We study the interface between absorbing layers such as Cu(In,Ga)Se₂ (CIGS), Cu₂ZnSn(S,Se)₄ (CZTSSe), Cu₂SnS₃, and Mo back electrodes by first-principles calculation. The stabilities of the compounds formed at the interface were studied using density functional theory (DFT) methods. We evaluated the stabilities of MoSe₂, Mo₃Se₄, Mo₉Se₁₁, and Mo₁₅Se₁₉ in the Mo–Se system and MoS₂, Mo₂S₃, Mo₃S₄, and Mo₁₅S₁₉ in the Mo–S system from the estimated formation enthalpies, ΔH_f, using a DFT-D2 approach based on van der Waals (vdW) forces. For CIGS solar cells, we calculated the enthalpy changes of reactions between the CuInSe₂–Mo and CuInS₂–Mo systems. The enthalpy changes of reactions expressed by CuInSe₂ + 1/6Mo → 1/2Cu₂Se + 1/6In₆Se₇ + 1/6MoSe₂ and CuInS₂ + 1/6Mo → 1/2Cu₂S + 1/6In₆S₇ + 1/6MoS₂ had positive values of +28 and +41 kJ/mol, respectively. The enthalpy changes of others reactions of the CuInSe₂ (CISe)–Mo and CuInS₂ (CIS)–Mo systems had positive values, too. These results suggest that CISe and CIS do not easily decompose even when Mo coexists. On the other hand, for Cu₂ZnSnSe₄ (CZTSe) solar cells, the enthalpy change of the reaction expressed by Cu₂ZnSnSe₄ + 1/2Mo → Cu₂Se + ZnSe + SnSe + 1/2MoSe₂ had a negative value of −11 kJ/mol. The mixed-phase system of Cu₂Se + ZnSe + SnSe + 1/2MoSe₂ is more stable than the Cu₂ZnSnSe₄ + 1/2Mo system, indicating that CZTSe readily decomposes when Mo coexists. The thickness of the MoSe₂ layer at the interface between the CZTSe absorber and the Mo back contact tends to increase during the deposition of the CZTSe films. For Cu₂SnS₃ (CTS) solar cells, the enthalpy change of the reaction expressed by Cu₂SnS₃ + 1/2Mo → Cu₂S + SnS + 1/2MoS₂ had a small negative value of −4 kJ/mol. The mixed phase of the Cu₂S + SnS + 1/2MoS₂ system is slightly more stable than the Cu₂SnS₃ + 1/2Mo system. CTS decomposes slightly when Mo coexists. A thin MoS₂ layer is formed at the interface between the CTS absorber and the Mo back contact.

Aluminum zinc oxide (AZO) thin films were deposited by modified in-line direct current magnetron sputtering — so-called serial cosputtering — using two commercial rotatable targets. A metallic aluminum target (secondary target) was used to condition the surface of a rotatable ceramic aluminum-doped zinc oxide target (ZnO:Al₂O₃$99:1$ wt %) as the primary target from which the film was deposited. Consequently, the direct current power applied to the secondary target enabled the variation of doping concentration in the deposited film. An optical transmittance >80% with an improved resistivity (<10⁻³ Ω·cm) and a higher carrier concentration (3 × 10²⁰–6 × 10²⁰ cm⁻³) were demonstrated in 300-nm-thick AZO thin films. AZO films with 2.23 at. % aluminum concentration used as the transparent conductive oxide (TCO) layer in copper indium gallium diselenide (CIGS) solar cells showed the best results with an efficiency of up to 13.2% due to an improved fill factor. The results reveal that adjusting the aluminum concentration could be used to match the electrical and optical characteristics of the TCO layer to those of the solar cells.

For applications to polycrystalline thin-film tandem solar cells, we studied p-type conductive BaCuSF single layer and p-type BaCuSF and n-type In₂O₃:Sn (ITO) bilayer films. The BaCuSF films were prepared by pulsed laser deposition (PLD), and the ITO films were prepared by RF sputtering. The bilayer film showed ohmic current–voltage characteristic. A tunnel junction between these two layers was successfully fabricated, because p-type BaCuSF and n-type ITO layers had sufficiently high carrier concentrations. The BaCuSF/ITO bilayer films were employed as the back electrodes of CdS/CdTe solar cells. A CdTe solar cell with a 20-nm-thick BaCuSF/a 300-nm-thick ITO bilayer back contact showed a high conversion efficiency of 13.9% (V_OC = 818 mV, J_SC = 25.2 mA/cm², and FF = 0.675), which was higher than that of a CdTe solar cell with a BaCuSF single-layer back contact (11.1%). The efficiency is comparable to that of a CdTe solar cell with a SrCuSeF/ITO bilayer back contact (14.3%).

Cu₂(Ge_xSn_1−x)S₃ (CGTS) (0 ≤ x ≤ 1.0) powders were synthesized by mixing elemental powders and sequential heating in 5% H₂S gas atmosphere. The crystal structures of CGTS solid solution samples were refined by Rietveld refinement of the powder X-ray diffraction data. CGTS has a monoclinic crystal structure with a space group of Cc (No. 9), and the refined lattice parameters decreased linearly from a = 6.653(7) Å, b = 11.532(5) Å, and c = 6.656(2) Å of Cu₂SnS₃ (x = 0.0) to a = 6.416(3) Å, b = 11.303(2) Å, and c = 6.434(9) Å of Cu₂GeS₃ (x = 1.0) with increasing Ge, x. The band-gap energy (E_g) of the Cu₂(Ge_xSn_1−x)S₃ solid solution monotonically increased from 0.87 eV for Cu₂SnS₃ (x = 0.0) to 1.53 eV for Cu₂GeS₃ (x = 1.0) with increasing Ge, x. The energy level of the valence band maximum (VBM) from the vacuum level was determined from the ionization energy measured by photoelectron yield spectroscopy (PYS). The determined energy levels of VBM of Cu₂SnS₃ and Cu₂GeS₃, and their solid solution samples were shallower than those of Cu₂ZnSnS₄ and Cu₂ZnGeS₄, and their solid solution. The energy level of conduction band minimum (CBM) was calculated by adding E_g to VBM level. The energy levels of CBM of Cu₂SnS₃ and Cu₂GeS₃, and their solid solution samples were deeper than those of Cu₂ZnSnS₄ and Cu₂ZnGeS₄, and their solid solution. Even when the Ge content in Cu₂(Ge_xSn_1−x)S₃ solid solution samples increased, the VBM level did not change significantly but the CBM level increased markedly. We also determined the VBM levels of Cu₂SnSe₃, Cu₂GeSe₃, and Cu₂(Ge,Sn)Se₃ solid solution samples. The energy levels of the VBM of Cu₂SnSe₃ and Cu₂GeSe₃, and their solid solution samples were also shallower than those of Cu₂ZnSnSe₄ and Cu₂ZnGeSe₄, and their solid solution. On the basis of these results, we discuss the band diagrams of Cu₂(Ge,Sn)S₃ and Cu₂GeSe₃ solar cells with the following device structure: absorber layer/CdS buffer layer/ZnO layer.

Cu₂Zn(Ge_xSn_1−x)S₄ (CZGTS) samples were synthesized by a mechanochemical process and sequential heating. The phases in the obtained powders were analyzed by X-ray diffraction. The band-gap energies of the CZGTS samples were determined by the diffuse reflectance spectra of UV–vis–NIR spectroscopy. The band gap energy of the CZGTS system linearly increased from 1.49 eV for Cu₂ZnSnS₄ (x = 0.0) to 2.25 eV for Cu₂ZnGeS₄ (x = 1.0). Their energy levels of valence band maximum (VBM) from the vacuum level were estimated from the ionization energies measured by photoemission yield spectroscopy (PYS). The energy levels of conduction band minimum (CBM) were determined by addition of the band-gap energies to the VBM levels. The energy level of VBM of the CZGTS solid solution was almost constant. On the other hand, the CBM level of the CZGTS solid solution linearly increased from −3.96 eV for Cu₂ZnSnS₄ (x = 0.0) to −3.28 eV for Cu₂ZnGeS₄ (x = 1.0) with the increasing Ge content. For CZGTS solar cells with CdS buffer layer, unfavorable cliff-type conduction band offset was expected. We also synthesized Cu₂ZnSnSe₄, Cu₂ZnGeSe₄, and Cu₂Zn(Ge,Sn)Se₄ (CZGTSe) solid solution samples and determined their energy levels of VBM and CBM. For Cu₂Zn(Ge_xSn_1−x)Se₄ system with 0.3 ≦ x ≦ 1.0, similar cliff-type conduction band offset was is expected. However, desirable positive spike-type conduction band offset was expected for the Cu₂Zn(Ge_xSn_1−x)Se₄ solar cells with 0.0 ≦ x ≦ 0.2 and CdS buffer layer.

The output parameters of InGaP/GaAs/InGaAs inverted metamorphic triple-junction (IMM3J) space solar cells have been collected for about 3.5 years in orbit. These thin-film solar cells were mounted on the small scientific satellite "HISAKI" (project name SPRINT-A), which was launched on September 14, 2013. We developed the satellite component that analyzes the solar cell behavior to investigate next-generation instruments for electrical power systems of satellites. The degradation trends of the IMM3J solar cells confirmed that the conventional method based on the relative damage coefficient can be applied to predict the degradation of IMM3J solar cells that have a thin-film structure. This is the first flight demonstration of IMM3J solar cells, which confirms the superior radiation stability of IMM3J solar cells in space.

Because of the reflection on the module surface, the irradiance received by the solar cells constituting the PV module is less than that received by the module surface. In this situation, an antireflection coating is crucial in improving the performance of the PV module. In this study, we experimented with and simulated a middle concentrator photovoltaic (MCPV) module with a large-acceptance-angle lens. The module was coated with a silica-based coating, which has an antisoiling and an antireflection function. Using the silica-based coating, the optical efficiency of the optics of the MCPV module increased. As a result, under outdoor conditions, the short circuit current and the conversion efficiency of the MCPV module with the silica-based coating was increased by 3 and 2.9%, respectively, compared with the module without the coating. The short circuit current of the MCPV module with a 100-nm-thick coating achieved a 5.6% increase in the optical simulation.

In thin-film solar cells such as inverted metamorphic multijunction solar cells, a local shunt spot can cause thermal runaway because of low thermal conductivity along the in-plane direction of the junction. Since electrical performance can be greatly reduced by thermal runaway, an appropriate design of the solar cells is necessary to prevent this mechanism. However, quantitative analysis of the thermal runaway is difficult because its threshold is usually strongly affected by the testing conditions and the characteristics of the shunt spots. In this study, we proposed a method of analyzing the thermal runaway characteristics quantitatively. We intentionally induced a thermal runaway under a simulated space environment with an arbitrary artificial shunt spot by a laser beam. The thermal resistance of the shunt spots and the threshold temperature for the thermal runaway were estimated using electrical and thermal models. This method enables an optimized design of thin-film solar cells.

To investigate seasonal variations in solar cell characteristics, we measured the spectral irradiance in summer and winter and simultaneously examined the efficiency of a four-junction solar cell under real sunlight. Efficiencies of 38.3% at 38 suns in the summer and 40.3% at 64 suns in the winter were obtained. In the summer under 1 sun, a step-like current–voltage curve was captured, and the reason for this result as well as the cause of lower efficiency values in the summer are proposed herein. The thickness of each subcell required to improve the conversion efficiency was also optimized using numerical simulations.

The electrical properties of GaAs//indium tin oxide (ITO)/Si junctions fabricated by surface-activated bonding (SAB) are investigated with emphasis on their dependence on the temperature of postbonding annealing. The current–voltage (I–V) characteristics of n⁺-GaAs//ITO/p⁺-Si and n⁺-GaAs//ITO/n⁺-Si junctions without annealing are linear. Those of p⁺-GaAs//ITO/p⁺-Si and p⁺-GaAs//ITO/n⁺-Si junctions without annealing are nonlinear. Although the interface resistance of all the junctions increases with increasing annealing temperature, the resistances of the respective junctions after the annealing at 400 °C are still smaller than the series resistance of the actual SAB-based InGaP/GaAs//Si hybrid triple-junction cells (∼4 Ω cm²). The n⁺-GaAs//ITO/n⁺-Si junction reveals the lowest resistance among the investigated junctions after annealing. These results demonstrate that GaAs//ITO/Si junctions with an ITO intermediate layer could be effective for reducing series resistance in hybrid multijunction cells.

The elevated manufacturing cost of highly efficient III–V multijunction devices limits them to space and high-concentration terrestrial applications. Hydride vapor-phase epitaxy (HVPE), in contrast to metal–organic vapor-phase epitaxy, can reduce the manufacturing costs due to the use of cheaper group III metal sources, capability to grow crystals under a low arsenic overpressure, and its low installation cost. In this study, we characterized the abruptness of the heterointerface between the InGaP and GaAs layers in GaAs solar cells grown by HVPE. InGaP passivation layers, such as the back-surface field (BSF) and window layers, were introduced in order to enhance the performance of the GaAs solar cells. Owing to the reduction of the surface recombination, the devices fabricated with the incorporated window layer showed a significant improvement in the short-wavelength range of the external quantum efficiency response, compared with that obtained for the unpassivated cells. This provided a cell efficiency improvement from 9.25 to 20.75%. However, the introduction of the BSF layer degrades the cell efficiency to 17.92%, owing to the formation of an anomalous interlayer at the GaAs-on-InGaP heterointerface.

The growth of ternary InGaP alloys is often susceptible to atomic ordering, which can lead to an anomalous bandgap reduction as well as the formation of antiphase boundaries (APBs). In this study, we report on the effect of growth rate and growth temperature on the performance of lattice-matched In_0.52Ga_0.48P solar cells grown on GaAs(001) substrates miscut 2° toward (111)B by solid-source molecular beam epitaxy. The bandgap of the InGaP film was widened from 1.855 eV for a film grown at 480 °C and 1.0 µm/h to 1.880 eV for a film grown at 510 °C and 1.5 µm/h owing to the suppression of atomic ordering in the alloy. In terms of solar cell performance, the open-circuit voltage was improved from 1.243 to 1.311 V by increasing the growth temperature and the growth rate due to not only the widening E_g but also the improved electrical characteristics (due to the fewer APBs present). In addition, the decreased number of APBs improved the fill factor and short-circuit current density. Consequently, the highest η of 14.43% was obtained for an InGaP single-junction solar cell grown at 510 °C and 1.5 µm/h and that of 25.57% for a dual-junction InGaP/GaAs solar cell.

We discussed the outdoor operation of a fixed flat sub module with an InGaP/GaAs/InGaAs inverted metamorphic triple-junction solar cell (IMM module) for the first time in the world. The global solar spectrum distribution was assessed using the average photon energy (APE) and spectral matching ratio (SMR) indexes in this study. The conversion efficiency of the IMM module was more than 30% and was markedly affected by APE rather than by module temperature under real environmental conditions. $\text{SMR}_{\text{Bottom}}^{\text{Top}}$ markedly increased in the high APE region. In the case of APE of more than 1.60 eV, the photocurrent of the bottom subcell decreased owing to the blue-rich spectrum, which limited the short-circuit current. The absorption region of the bottom subcell overlapped with the water absorption region in the solar spectrum; therefore, environmental conditions such as water precipitation greatly affected the output of the current-matching inverted triple-junction solar cell.

The effects of wafer (surface) temperature on the surface morphologies of InP, InGaAsP, and InGaAs grown in a planetary metalorganic vapor phase epitaxy (MOVPE) reactor using tertiarybutylarsine (TBA) and tertiarybutylphosphine (TBP) as group V precursors were investigated. Compared with small horizontal reactors, InP growths require a relatively low temperature to prevent TBP predecomposition and parasitic gas-phase reactions. However, such a low temperature can prevent the surface migration of In adatoms and degrade surface morphology, especially when a substantial amount of Ga precursors is present. Hence, growths of InGaAs require a higher temperature than that of InP and InGaAsP. In this work, 1.04 eV In_0.79Ga_0.21As_0.46P_0.54 and 0.74 eV In_0.53Ga_0.47As single-junction cells as well as InGaAsP/InGaAs dual-junction solar cells have been successfully fabricated. The cell performances reveal that InGaAs layers grown at 590 °C exhibit great crystal quality, while 550-°C-grown InGaAsP layers contain a noticeable extent of defect density and lattice imperfection.

This paper discusses the opportunities and challenges of designing products using luminescent solar concentrator (LSC) photovoltaic (PV) technologies. The focus is on the integration of LSC PV technologies in PV modules, future products and buildings. It is shown that the typical material properties of LSCs — low cost, colorful, bendable, and transparency — offer a lot of design freedom. Two differently designed LSC PV modules with back contacted solar cells are presented including ray-tracing simulations and experimental results resulting from their prototypes. It is shown that the efficiency of a LSC PV module can be 5.8% with a maximum efficiency of 10%. Further the results of a design study which focused on product integration of LSC PV technologies are presented and discussed. In total 16 different and highly innovative conceptual designs resulted from this project, which were prototyped at scale to show their feasibility and integration features.

Deep-level defects were investigated and compared among three molecular beam epitaxy (MBE)-grown dilute nitride semiconductor GaInNAsSb solar cells, one of which was as-grown and the other two were annealed in a metal organic chemical vapor deposition (MOCVD) atmosphere (H₂ and AsH₃) or a nitrogen (N₂) atmosphere. A residual defect in the as-grown sample was firstly discovered and considered to be mainly responsible for the degradation in GaInNAsSb solar cells. Meanwhile, an N–H-related defect was confirmed, as well as the effects on background carrier concentration (BGCC) together with hydrogen incorporation. The N₂ annealing showed a significant effectiveness of suppressing the defects and improving the solar cell properties.

The power conversion efficiencies of single-junction III–V solar cells based on InGaP, GaAs, InGaAsP, and InGaAs for wireless power transmission were investigated by varying the wavelength and incident power of a laser diode. The GaAs solar cell exhibited the highest conversion efficiency of 52.7% at a laser wavelength of 824 nm and a laser beam irradiance of 4 suns. For each solar cell, the conversion efficiency peaked at an optimal incident laser wavelength and this effect was examined in detail. It was found that the conversion efficiency is maximal when the laser photon energy is 80–100 meV higher than the bandgap of the semiconductor material.

Fill factor (FF) is one of the important photovoltaic parameters, which is limited mainly by charge recombination in solar cells. It is still a challenging issue to realize a high FF with a thick active layer, which is required for the development of efficient bulk-heterojunction polymer solar cells. In this study, we discussed how recombination losses limit charge collection efficiency in polymer solar cells with low FF by transient optoelectronic techniques. As a result, we found that geminate recombination is almost voltage-independent, that bimolecular recombination is suppressed significantly under open-circuit condition, and that hole and electron mobilities are low and imbalanced. We therefore concluded that the low collection efficiency is caused by the bimolecular recombination of accumulated charges due to low and imbalanced mobilities.

The homeotropic alignment of non-peripherally hexyl-substituted phthalocyanine (C6PcH₂) molecules in thin films has been investigated. A spin-coated film C6PcH₂ was covered with a poly(vinyl phenol) layer, and a thermal treatment through a liquid crystalline phase was carried out. The molecular packing structure and induced homeotropic alignment in the thin films were clarified by the grazing incidence wide-angle X-ray scattering technique. The bulk heterojunction solar cell with the homeotropically aligned C6PcH₂ was fabricated, and the photovoltaic properties were discussed by taking the optical properties and carrier mobility in the active layer into account.

It is crucially required to understand charge traps in CH₃NH₃PbI₃ perovskites, which severely affect the charge recombination mechanism and hence have a critical impact on photovoltaic performance. Herein, we studied the photoexcited dynamics of CH₃NH₃PbI₃ perovskites with different grain sizes by measuring the intensity dependence of time-resolved photoluminescence (PL). As a result, we found that the PL decay for all the samples can be well analyzed using bimolecular radiative recombination and trap-assisted Shockely–Read–Hall recombination models. We discuss how the PL dynamics impact the photovoltaic performance of CH₃NH₃PbI₃ perovskite solar cells on the basis of the recombination analysis.

The organic molecule 2,7-dioctyl[1]benzothieno[3,2-b][1]benzothiophene (C₈-BTBT) was deposited on quartz glass, $(11\bar{2}0)$A-, (0001) C-, and $(1\bar{1}02)$R-single-crystal Al₂O₃ (sapphire), and (100)-, and (111)-single-crystal MgO substrates by vacuum thermal evaporation, and structural characterizations were carried out by X-ray diffraction analysis and atomic force microscopy (AFM) observation. The (001) out-of-plane orientation with a similar in-plane orientation was obtained irrespective of the substrate material and orientation, and its formation was governed by π–π-stacking-induced molecular ordering. The degree of orientation was reflected by the grain structure related to the substrate material. The growth model of the oriented C8-BTBT layer was speculated on the basis of experimental results.

TiO₂/CH₃NH₃PbI_3−xCl_x-based photovoltaic cells added with PbI₂ were fabricated and investigated. The fill factor and open-circuit voltage of the cell with 16 nm TiO₂ nanoparticles were increased by PbI₂ addition. For the cell, the short-circuit current density and open-circuit voltage were increased by using 23 nm TiO₂ nanoparticles in place of 16 nm TiO₂ nanoparticles. The external quantum efficiency increased, which would be due to the improved interface between the TiO₂ and perovskite layers. As a result, the conversion efficiency was improved by adding PbI₂ and using 23 nm TiO₂ nanoparticles. X-ray diffraction showed that the perovskite crystals changed from cubic to tetragonal, which resulted in the improved stability of the perovskite solar cell by PbI₂ addition.

The effect of NiO_x (0 < x) hole transport layer prepared by a radio-frequency sputtering method on the photovoltaic properties of planer-type CH₃NH₃PbI₃ perovskite solar cells (PVSCs) was investigated. The open circuit voltage of PVSC decreases with increasing applied power of the sputtering machine. The lack of hydroxyl groups on the surface of the metal oxides shifts the work function (WF) to higher energy levels. The X-ray photoelectron peaks of Ni 2p_3/2 at 855.6 eV and O 1s at 531.3 eV assigned to ONi(OH) decrease with the increasing power. Therefore, the decrease in the number of hydroxyl groups must have shifted the WF to higher energy levels. The shunt resistance of current–voltage curve and the internal quantum efficiency of the PVSCs is independent of NiO_x prepared at various powers. Assuming that the recombination effect can be neglected, the open circuit voltage (V_OC) decrease with increasing power is due to the shifted WF to higher energy levels.

The measurement of the optical absorption coefficient spectra of perovskite thin films of CH₃NH₃PbI₃ by resonant photothermal bending spectroscopy (resonant-PBS) and Fourier-transform photocurrent spectroscopy (FTPS) for the defect estimation of the material has been attempted. A fundamental absorption edge was observed in the photon energy ħω region of 1.5–1.6 eV by both techniques. In addition, extra absorption of about 10² cm⁻¹ was detected at ħω < 1.5 eV only in the spectra measured by resonant-PBS. This extra absorption seems to indicate the existence of localized states in the bandgap of CH₃NH₃PbI₃ films. The difference between resonant-PBS and FTPS profiles below the absorption edge is caused by the difference in measurement mechanism, that is, resonant-PBS utilizes a temperature rise of the sample by the nonradiative recombination of photoexcited carriers, while FTPS measures photocurrent arriving at electrodes by the electric field after the photoexcitation. These results seem to indicate that resonant-PBS is a useful tool for the defect estimation of CH₃NH₃PbI₃.

In this study, the effect of the aromatic N-heterocycle treatment on the performance of perovskite solar cells (PSCs) was examined. The device architecture in this study was as follows: fluorine-doped SnO₂-coated glass (FTO glass)/compact TiO₂/mesoporous TiO₂/perovskite (CH₃NH₃PbI₃)/2,2',7,7'-tetrakis(N,N-di-p-methoxyphenylamino)-9,9'-spirobifluorene (spiro-OMeTAD)/Au. A perovskite film was post-treated with three different aromatic N-heterocycles for the passivation of the interface. The overall power conversion efficiency (PCE) increased in the following order: triazine > pyrimidine > pyridine. An overall PCE of 16.0% (backward scan) was achieved for the triazine treatment under optimized conditions, although a large hysteresis was observed. Time-resolved photoluminescence (TRPL) measurements show that the triazine-treated perovskite film afforded a longer photoluminescence lifetime than the untreated perovskite film. These results indicate that triazine is the most effective for the enhancement of the cell performance among the three aromatic N-heterocycles.

We synthesized a pair of A–D–A–D–A (A = acceptor, D = donor unit) oligomers containing a benzothiadiazole (BT) unit, namely, DRCN5BT-HH and DRCN5BT-TT, which differs in the positions of alkyl side chains. Solution-processed bulk heterojunction solar cells consisting of PC₆₁BM and DRCN5BT-HH gave a high V_oc of 1.10 V and markedly improved power conversion efficiencies up to 3.26% after solvent vapor annealing (SVA). The effects of SVA on the morphology and microstructure of the active layer were investigated by atomic force microscopy (AFM) and grazing-incidence wide angle X-ray scattering (GIWAXS). For DRCN5BT-HH, crystallization, orientation, and molecular ordering were improved, and a fibrillar structure was formed with acicular aggregates of the donor. For DRCN5BT-TT, no marked structural changes were induced by SVA. We found that these morphological and microstructural changes induced by SVA correlate with the differences in the photovoltaic properties of both of the isomers.

CH₃NH₃PbI_3−xCl_x perovskite-based photovoltaic devices were fabricated by a spin-coating technique using hot airflow, and the effects of CuX (X = I, Br, or Cl) addition to perovskite precursor solutions on photovoltaic properties were investigated. Highly 100-oriented perovskite grains were formed using hot airflow during spin-coating. Compared with pristine perovskite CH₃NH₃PbI_3−xCl_x crystals, grain sizes of the perovskite crystals decreased with the addition of CuBr or CuI, and the coverages of perovskite layers increased, which improved photovoltaic properties. The perovskite solar cell with CuBr added showed the highest conversion efficiency in the present work.

We have been working on the fabrication of practical perovskite solar modules using an existing thin-film solar module patterning technology. Patterning processes were applied using laser scribing and mechanical scribing on a glass/F-doped SnO₂ (FTO)/compact TiO₂ (c-TiO₂)/meso-TiO₂/CH₃NH₃PbI₃/2,2',7,7'-tetrakis(N,N-di-p-methoxyphenylamine)-9,9'-spirobifluorene/Au structure. However, TiO₂ remained in the FTO–Au junction at the cell connection part: this appeared to create a series resistance that degraded the performance of the module. UV laser treatment was examined for application to the cell connection part to minimize this resistance. The results showed that the resistivity of the cell connection part, which was conventionally 0.26 Ω·cm², was reduced to 0.1 Ω·cm² by this laser treatment. We designed the solar module based on the value of a contact resistivity of 0.1 Ω·cm² and the current–voltage curve of a small single cell. As a result of prototyping solar cell modules (35 cells in series) with an aperture area of 354 cm² on a 203 × 203 mm² glass substrate, a module efficiency of 12.6% was achieved. There appears to be no serious impediment to the fabrication of the perovskite solar modules.

The structural stabilities of perovskite compounds were investigated by calculating the tolerance and octahedral factors, which are known as the empirical indexes of the stabilities of the perovskite structure. Elements of groups 2, 12, and 14, and transition metals were substituted in CH₃NH₃PbI₃ compounds, and the ranges of compositions with a perovskite structure were proposed. 540 compounds were surveyed in this work, 132 of which were used to satisfy the conditions of forming the perovskite structure. Of these 132 compounds, 83 were candidates for low-toxicity compounds.

The transport properties between PbSe quantum dots (100) facets with three different halide ligands (e.g., chloride, bromide, and iodide) and iodide ligand at (001), (110), (111) facets have been studied using density functional theory (DFT) and non-equilibrium Green's function (NEGF). Quantum conductance between iodide ligand attached surfaces has the highest value due to its extra ligand contribution near Fermi level. However, configurations with bromide and chloride ligands have much higher maximum quantum conductance due to their stronger coupling between surfaces with lower distances. The electrical fields formed between PbSe(110) surfaces attenuates the charge transport whereas between (111) surfaces it improves the transport near fermi level. In spite of the existence of electrical field and lower surface to surface distance, the zero-voltage bias transport between (001) surfaces is stronger. This is possibly understood from the slight increase in the DOS from the Pb s-orbital and Se p-orbital near Fermi level.

We synthesized Gd₃Ga₅O₁₂ garnet codoped with Er, Ni, and Nb to realize transparent ceramics of broadband-sensitive upconverters used for crystalline silicon solar cells. We investigated energy transfer between the Er³⁺ and Ni²⁺ dopants using time-resolved measurements of both Ni²⁺ and Er³⁺ emissions. Most of the Ni²⁺ dopants occupied the six-coordinated Ga sites in the synthesized powder samples. Under 1.1–1.45 µm excitation, energies absorbed by the six-coordinated Ni²⁺ sensitizers almost perfectly transferred to the Er³⁺ emitters, leading to upconversion (UC) emission of the Er³⁺ at 0.98 µm. Combined with the UC emission directly excited at 1.45–1.65 µm, a broad sensitivity range of 1.1–1.65 µm was realized. However, the internal quantum efficiency of the UC emission was lower than that of previously developed La(Ga_0.5Sc_0.5)O₃:Er,Ni,Nb, most likely owing to energy dissipation to a small amount of the Ni²⁺ dopants locating at the four-coordinated Ga sites. To improve the UC performance, the four-coordinated Ni²⁺ should be eliminated.

Thin-film solar cells fabricated using the epitaxial lift-off (ELO) process enable considerable cost reduction as well as light-trapping. A polyimide film was used as a support substrate for thin film devices to address the poor temperature tolerance of poly(ethylene terephthalate) films. Samples were stuck to polyimide films by an Au–Au bonding process. It was found that the insertion of a wetting layer enhanced the adhesion between a polyimide film and an Au layer. The achieved bonding was strong enough to carry out the ELO process, and no degradation was observed in cell performance. Thin-film solar cells with strain-balanced multiple quantum wells were also fabricated using the ELO process to benefit from the light-trapping effect. Quantum efficiency enhancement was observed in the long-wavelength range of 850–980 nm compared to the non-ELO-processed devices. The photo-absorption enhancement was due to the Fabry–Perot resonance between the surface and the rear Au electrodes.

In an ideal intermediate-band (IB) concept, the host semiconductor generates current by absorbing short-wavelength light and the IB is used to absorb long-wavelength light. Here, we investigate the impact of the host absorber thickness at the front side of the cells on the properties of InGaP-based InP quantum dot (QD) solar cells. We prepared the InGaP-based InP QD cells with the front i-InGaP layer and compared them with the cells without the front i-InGaP layer. The insertion of the front i-InGaP layer results in an increase in quantum efficiency in the visible region, indicating that the short-wavelength light absorbed at the InGaP host increases the short-circuit current density in the cell. In addition, a thick host layer leads to reduced quantum efficiency below the host bandgap energy, which indicates that the thermal escape from QDs is suppressed. These results indicate that the optimization of the host semiconductor thickness is critical for achieving the ideal operation of the IB concept in the QD solar cells.

We have proposed a light trapping concept for crystalline silicon photovoltaic (PV) cells used for power transmission from solar-pumped lasers (SPLs) emitting at 1064 nm. The underlying mechanism is multiple reflection between a multilayered angle-selective filter on the front surface and a diffuse reflector on the rear surface of the cell. For the stationary use and optical fiber connection of the SPL-PV combination in which the incident angles of the laser light to the PV cells (ϕ) change within 5–10°, double-cavity bandpass filters realize the required angle selection function. When the SPLs illuminate moving electric vehicles equipped with the PV cells at larger ϕ values, shortpass filters are appropriate. These specially designed angle-selective filters provide significant light trapping performance and consequently absorbances higher than the Yablonovitch limit up to ϕ = 45°. The fact that the present light trapping configuration requires no microfabrication processes is also a great advantage over two- and three-dimensional photonic crystal structures.

We have designed crystalline silicon photovoltaic (PV) cells used for power transmission from solar-pumped lasers (SPLs) emitting at 1064 nm. The practical light-trapping performance of the combination of a multilayered angle-selective filter on the front surface and a diffuse reflector on the rear surface was evaluated by ray-trace simulation, which was taken into account in device simulation. When the SPLs illuminate stationary PV cells or are connected using optical fibers and hence the incident angles to the PV cells are within 10°, a high laser-to-electricity conversion efficiency of around 50% would be feasible under 100 W/cm² laser illumination using the 50-µm-thick cells. For power transmission to moving objects such as electric vehicles, in which the incident angles change up to 30°, the efficiency of the 75-µm-thick cells is slightly lower because of the less significant light-trapping effect. To realize these high efficiencies, reduction of contact resistance and surface recombination velocity is required.

We have proven the light trapping concept for crystalline silicon (c-Si) photovoltaic (PV) cells used for power transmission from solar-pumped lasers (SPLs) emitting at 1064 nm. We fabricated a prototype of the PV cells in which four unique features were employed: (i) a 45-µm-thick c-Si wafer and an aperture of 100 µm diameter, (ii) a bandpass filter on the front surface and a diffuse reflector on the rear surface, (iii) a p⁺/p/n⁺ configuration illuminated from the p⁺-layer, and (iv) an Al/Ag double-layered electrode on the p⁺-front layer and a Ag electrode on the n⁺-rear layer for compatibility of ohmic contacts and high reflectance. Under 1064 nm laser illumination at 4.8 W/cm², the external quantum efficiency was as high as 0.92 despite the use of a thin c-Si wafer, resulting in a conversion efficiency of around 35%. No marked detrimental effect was observed under more intense illumination up to 21.3 W/cm².

We investigated the influence of the barrier layer height in silicon quantum dot superlattice (Si-QDSL) solar cells on the cell performance by two-dimensional device simulation taking into account the quantum size effect. Although at Si-QD diameters (d) of 5 and 10 nm better cell performance was obtained as the barrier height increased, in the case of d = 3 nm, the efficiency decreased as the barrier height increased. Further investigation showed that the band mismatch at the interface between Si-QDSL and the n-type layer degraded open-circuit voltage (V_oc) at d = 3 nm. After the optimization of the physical properties of the n-type layer, V_oc was improved to 1.43 V and the bandgap-V_oc (W_oc) offset became about 0.4 V, suggesting that the band mismatch is the significant factor for the V_oc loss. From these results, it was found that careful selection of n- or p-type materials is necessary when the strong quantum confinement effect occurs in Si-QDSL.

A light-trapping structure was fabricated on crystalline Si wafers. In this method, a Ge/Si(001) self-assembled dot structure was etched using KOH and a commercial alkaline solution including nanomask particles. Many nanoscale islands were formed by adding the nanomask particles. The enhancement of light absorption was successfully confirmed in samples with nanostructures after etching. Since the etching margin of a nanostructure was less than 2 µm, the nanostructure could be applied to an ultrathin Si substrate. To investigate the effect of surface passivation on the nanostructure, the nanostructure was passivated using hydrogenated amorphous Si thin films by plasma-enhanced chemical vapor deposition. By quasi-steady-state photoconductance measurement, the suitable implied open-circuit voltage and implied fill factor for heterojunction solar cells were obtained. From calculations based on these parameters, solar cells with nanostructures at a substrate thickness of 100 µm are expected to have a conversion efficiency of 23.21%.

We studied the dynamical properties of quantum dot solar cells in which GaSb quantum dots (QDs) were embedded in the middle part of an undoped layer sandwiched by n- and p-layers. Three kinds of GaSb QDs with GaAs or AlGaAs as a barrier material for QDs were prepared for comparison of their properties. The capacitance C was measured as a function of the frequency f in the ranges between 5 kHz and 3 MHz under different illumination conditions. A rise in capacitance in response to illumination was observed in all the samples, showing that photogenerated carriers (or holes) contribute to the dynamical properties of solar cell devices. Differences in capacitance change among the three samples are observed in magnitude and its frequency dependence. We discuss them in terms of the difference in dynamical processes, specifically the thermal escape rate of holes in QDs.

To analyze the dynamic behavior of Na ions in the p–n junction of crystalline Si photovoltaic (PV) cells during potential-induced degradation (PID), a bias voltage was applied between the electrodes of the testing PV modules. During the PID test with a negative high voltage stress (HVS), PID was suppressed by the simultaneous loading of a forward bias voltage, although that of a reverse bias voltage accelerated the progression of PID. However, in the recovery stage after the PID test, the opposite loading effects of the bias voltage with a positive HVS were found. In other words, the extent of recovery from PID was suppressed or enhanced by forward or reverse bias loading, respectively. Similar effects of bias loading were observed even under the recovery conditions without the positive HVS. The effect of Na ions within a depletion layer on the evolution of PID will be discussed for the interpretation of these results.

To ensure a longer lifetime for photovoltaic (PV) modules, many accelerated test standards are being developed or revised. Among the test standards, damp heat (DH) testing in the dark is widely employed to test the module durability. However, in the field, high temperature is usually accompanied by light irradiation. This difference can result in a degradation not observed in the field, especially in the case of Cu(In,Ga)(S,Se) (CIGS) PV modules. To explore suitable test procedures that simulate performance in the field appropriately, we investigated the effects of light irradiation and forward biasing, during DH testing and potential induced degradation (PID) testing of CIGS PV modules. In the case of the CIGS product we have tested, both light irradiation and forward biasing have suppressed degradation in DH as well as PID tests, and no degradation was found after light soaking. These results suggest the needs to add corresponding option(s) to the relevant test standards.

The solder joint degradation due to thermomechanical fatigue is investigated in this paper for photovoltaic (PV) mini-modules with ethylene vinyl acetate (EVA) of different viscoelastic properties. The mini-modules were laminated at different curing temperatures in order to obtain EVA encapsulation with different viscoelastic properties. The influence of viscoelasticity of EVA on the thermomechanical fatigue generated on solder joint is analyzed based on a two-dimensional (2D) finite-element model. Based on simulation of thermomechanical stresses accumulation, mini-modules with EVA cured at lower temperatures accumulated approximately 40% more stresses during the thermal cycle testing than mini-modules with optimal cured EVA. The tested mini-modules with EVA cured at lower temperature showed greater power degradation than the optimal cured mini-modules. An apparent increase in equivalent series resistance is the primary factor the power loss. A good correlation between the accumulated thermomechanical fatigue and the increase in equivalent series resistance is demonstrated with the tested samples.

In this paper, the current status of the power generation capacity of installed photovoltaic modules, which include monocrystalline Si, multicrystalline Si, thin-film multijunctions consisting of hydrogenated amorphous Si (a-Si:H) and hydrogenated microcrystalline Si, copper indium gallium selenide, thin-film single-junction a-Si:H, passivated emitter and rear cell (PERC) monocrystalline Si, Si heterojunction (SHJ), and interdigitated back contact (IBC) modules, at the National Institute of Advanced Industrial Science and Technology is reported. The degradation of PERC modules was confirmed by indoor measurement and a light soaking test. In the case of SHJ modules installed in 2016, the output power was unstable under the initial light irradiation conditions. In the case of IBC modules, potential-induced degradation was observed for one type within only 6 months of outdoor exposure, and for the other type, it was observed within four years of outdoor exposure. A formula for estimating the amount of power generation from the degradation rate was proposed.

Migration routes of Na ions towards the surface and into SiN_x films of Si cells during the potential-induced degradation (PID) test were analyzed by microscale measurements such as X-ray photoelectron spectroscopy, scanning electron microscopy, energy dispersive X-ray spectroscopy, and conductive atomic force microscopy. These measurements showed the appearance of high Na concentrations near the finger electrodes and at the top of texture structures of the SiN_x film surface. However, a high current conductivity of SiN_x films was observed at halfway between two finger electrodes and at the top of texture structures. These results suggest that focusing of electric fields originating from finger electrodes and the shape of texture structures affected the Na distributions and migration into the SiN_x films. The influence of the PID recovery test on the Na ion migration and SiN_x films is also discussed in the paper.

Quantitative evaluations of output power and the development of a technique for countermeasures against volcanic ash fall are required in a volcanic ash fall environment. In this study, in order to clarify the effect of volcanic ash fall on the output power of photovoltaic (PV) modules and the effects of the hydrophilic coating on PV module cover glass, we evaluated the output power characteristics of PV modules by an artificial test. The hydrophilic coating was carried out using an inorganic material and formed an antisoiling layer on the PV module surface. The output power characteristics of PV modules were measured using a solar simulator. The superiority of the hydrophilically coated cover glass was found between the setting angles of 40 to 60° and the most remarkable in the vicinity of 45°. Moreover, it was confirmed that the superiority of the hydrophilically coated cover glass increases with an increase in the amount of ash fall.

Two types of crystalline silicon (c-Si) photovoltaic (PV) modules have been tested in the cold-dry climate of the Gobi Desert of Mongolia, from 2002 to 2012, to verify the preliminary estimation of PV power generation. This study mainly focuses on the analysis of the long-term performance of the PV modules using indices such as energy yield, reference yield, performance ratio (PR), and characterization indices of the main electrical parameters as well as degradation rate. Average annual energy yields were 1880 h for Module 1 and 1789 h for Module 2. Thus, PRs were 0.85 and 0.80, while the average annual reference yield was 2223 h. On the other hand, degradation rates of Modules 1 and 2 were −1.28 and −0.86%/year, respectively. The electrical degradation observed in both modules was due to the loss of short-circuit current as open-circuit voltage reduction was not considered in this field test.

There is a growing need for the precise outdoor performance measurement of photovoltaic (PV) modules for low-cost onsite performance measurement, monitoring, and failure diagnosis. For the precise evaluation of a PV module, an accurate temperature measurement technique is required. It is necessary to measure the temperature of the solar cell in a module structure (junction temperature) because it determines the temperature characteristics of the PV module, rather than the temperature of the backsheet. In this study, a PV module with an internal thermocouple was fabricated. A thermocouple was inserted immediately below the solar cell so that it could be in direct contact with the cell, enabling an accurate temperature measurement. Moreover, the temperature of the solar cell in the PV module structure was predicted by heat flux calculation using the backsheet temperature, which can be measured easily. In this manner, the solar cell temperature was estimated accurately within an error of ±1 °C.

Production technologies of large-area, light-weight flexible thin film silicon solar cells and modules using 1-m-wide and 3,000-m-long heat-resistant plastic film substrates were developed. A unique monolithic device structure having through-hole contacts formed on a flexible plastic film was also developed. The unique solar cell device structure enables us to cut and connect flexible film cells in order to easily customize the voltage and current specifications of products. Durable solar cell and module designs and effective acceleration tests were established. In particular, an acceleration test that combines damp heat and a current injection test was established. Approximately 3,000 h of damp heat and current injection corresponds to 20 years of outdoor exposure in Japan. In this paper, we propose a pressure cooker and current injection test to achieve a much faster acceleration test. 300 h of pressure cooker and current injection test is approximately equivalent to 3,000 h of damp heat and current injection test and 20 years of outdoor exposure in Japan.

The temporal instability and spatial nonuniformity of solar irradiance, and their relationship are analyzed by field observations, and a method of detecting the events in which irradiance is stable and uniform is proposed for outdoor photovoltaic (PV) module performance characterizations. The relationships of instability and nonuniformity of irradiance with measurement conditions are discussed, e.g., a long measurement time worsens instability. When events are observed with repeated measurements under certain conditions and selected under the condition that the instability is low afterwards, the nonuniformity of the selected events is low. This means that for measuring and detecting uniform events well, two sensors are required, but they can also be detected with a single sensor by repeated measurements and selecting with instability. This suggests that repeated measurements may be a filtering method for finding stable and uniform events, which is required for high-accuracy outdoor PV module performance characterizations.

Solar irradiance intensity is usually reduced by cloud shading. However, an enhancement of solar irradiance has rarely been detected. For the detection of irradiance enhancement events, high-speed pyranometers are employed here. The enhancement is detected in almost one-third of days in a month on average in central Japan. From sky-camera observation and meteorological analysis, the enhancement occurs owing to the superimposition of the diffuse irradiance scattered at the edges of cumulus clouds near the sun on the direct irradiance passing through a gap in the clouds, which exist along meteorological fronts. Events of solar irradiance enhancement were detected in a five-month observation period and were analyzed statistically. Some of the events with the largest enhancement are amplified by more than 1.6, and their durations are from 30 s to 4 min. Statistical analyses showed events with either large amplifications or long durations but not events with both large amplifications and long durations.

Recently, outdoor photovoltaic (PV) module performance characterizations have been carried out mainly for testing installed PV modules. For high-accuracy outdoor measurement, solar irradiance with stability in time and uniformity in space are required. Short-period temporal instability and spatial nonuniformity of solar irradiance are observed using a series of PV module sensors (PVMSs). The speed and direction of cloud shadows moving on PVMSs are evaluated from the fluctuations of irradiance detected by several PVMSs. On the basis of cloud distribution, weather conditions are classified into four types, and the characteristics of solar irradiance fluctuations are examined for each type. The irradiance changes quickly under cumulous clouds. On the other hand, the irradiance becomes weak, but its short-period fluctuations are not evident under layered clouds. The amplitude of irradiance fluctuations reaches 70% of clear-day irradiance at noon in this observation. The spatial nonuniformity evaluated on the basis of the instantaneous difference in irradiance intensity between two PVMSs also reaches 1.8%/m of clear-day irradiance.

Light irradiation during a potential-induced degradation (PID) test for p-type crystalline Si photovoltaic modules was studied under various test conditions. PID progress was clearly delayed by light irradiation. Such effects by light irradiation strongly depended on the wavelength. Clear effects on PID delay were observed in the case of light irradiation with a UV component below approximately 400 nm. On the other hand, almost no effects were observed in the case of light irradiation without a UV component. Delay of PID progress was also observed even under high-humidity condition. It was also found that PID progresses in a partially shadowed region even under UV light irradiation, which means that a partial shadow easily brings about PID for PV modules exposed outdoors.

In this study, transient diffuse reflectance spectroscopy (TDRS) and electroluminescence (EL) analysis have been employed to clarify the carrier dynamics in potential-induced degradation (PID) in single-crystalline Si (sc-Si) solar cell. We first localized the PID-affected region in terms of degradation intensity on the modules on the basis of EL. The carrier dynamics in that region are then studied and clarified in terms of carrier lifetime, defect level, and photogenerated carrier density. Our results suggest that carrier relaxation in a fresh solar cell proceeds via band to band and/or shallow and deep donor–acceptor recombination. However, the dominant recombination in solar cells with PID is intercenter charge transfer via shallow-to-deep and/or deep–deep defects for which the carrier lifetime decreased drastically. Also, it is found that the carrier dynamics near the surface and bulk do not progress similarly as confirmed using 532- and 1064-nm-wavelength pumps.

The reliability and long-term durability of two bifacial photovoltaic modules, a glass–transparent backsheet (GB) module and a glass–glass (GG) module, were compared. The output degradations after UV, damp heat (DH) and thermal cycle tests for GB modules were similar to those for GG modules in the case of using ethylene vinyl acetate (EVA) encapsulant. In addition, GB modules using polyolefin encapsulant exhibited much higher reliability in a further extended DH test than GG modules using EVA encapsulant. It was also shown from an outdoor exposure test that the temperature of the GB module was clearly lower than that of the GG module, leading to a higher open-circuit voltage for the GB module. GB modules, even those using EVA encapsulant, had long-term durability similar to that of GG modules in addition to the advantages of a light weight, easy fabrication, and lower module temperature in a hot ambient.

A novel, nondestructive low-cost detection method for acetic acid distribution in a photovoltaic (PV) module during the damp heat (DH) test based on reflectance changes of tin film sensors is proposed and demonstrated. The sensor consists of a tin film evaporated on a glass substrate. Nineteen sensors and one gold film are laminated in the PV module, and the DH test was performed at 85 °C and 85% relative humidity for 7203 h. The time range of measurement can be controlled between 2000 to 6000 h by adjusting the tin film initial thickness from 70 to 160 nm.

The short-circuit current I_sc and spectral responsivity (SR) of crystalline silicon (c-Si) photovoltaic (PV) modules and cells are investigated at various device temperatures. The long-wavelength tail of SR shifts toward longer wavelengths as the temperature increases. Although the wavelength shift Δλ per temperature difference ΔT is dependent on the device structure and λ, the present results show that the shift in the quantum efficiency (QE), which is related to SR by QE ≈ SR × 1240/λ, can be approximated using the wavelength-independent constant of Δλ/ΔT ∼ 0.45 nm/K for various types of c-Si device. The results indicate that the temperature dependence of SR can be approximated using a simple relation for various types of c-Si PV device, which greatly improves the precision of spectral mismatch correction when SR at only one temperature is available.

The temperature dependences of monocrystalline silicon (mono-Si), interdigitated back-contact (IBC), passivated emitter and rear cell (PERC), silicon heterojunction (SHJ), and multicrystalline silicon (multi-Si) photovoltaic (PV) modules are estimated. Their characteristics such as temperature coefficients (TCs) for the maximum power (P_max), open-circuit voltage (V_oc), and efficiency are measured. IBC, PERC, SHJ, mono-Si, and multi-Si exhibit TCs at 25 °C of P_max and efficiency of −0.378, −0.404 to −0.373, −0.279, −0.432 to −0.415, and −0.516%/°C, respectively. IBC, PERC, SHJ, mono-Si, and multi-Si exhibit TCs at 25 °C of V_oc of −0.267, −0.269 to −0.263, −0.218, −0.312 to −0.306, and −0.334%/°C, respectively. Thus, IBC, PERC, and SHJ show weaker temperature dependences than mono-Si and multi-Si. The efficiencies of IBC, PERC, and SHJ at 25 °C are 19.00, 18.81 to 19.25, and 21.40%, whereas those of mono-Si and multi-Si are 15.84 to 17.26 and 15.54%, respectively. Thus, IBC, PERC, and SHJ are more efficient than mono-Si and multi-Si in the temperature range higher than 25 °C.

In this paper, we present the energy performance and degradation of multicrystalline silicon photovoltaic (PV) modules after a 30-year operation under hot and humid climatic conditions. A comprehensive analysis was conducted by visual inspection, electrical performance evaluation, and system final yield evaluation. Visual inspection revealed the yellowing and delamination of ethylene vinyl acetate (EVA), and cracks in the backsheet were found to be the most frequent defects. It is possible to perform a comparison between final and initial parameters. The average performance degradation over 30 years from 1986 to 2016 is 6.51%. The annual average degradation has been found to be only 0.22%/year. System output power and performance ratio (PR) were also analyzed, giving an annual PR of 69.72%. These results indicate that the actual lifetime of PV modules is significantly longer than the warranty periods of 25 years, and still with good performance and reliability.

Widespread implementation of building integrating photovoltaics requires aesthetically pleasant panels. Black surfaces are often visually appealing and, if based on a black photoabsorber, may lead to increased panel efficiency. Here, we demonstrate a combination of black silicon and black ribbons (or bus-bars), both fabricated with methods that are compatible with production lines of screen-printed, front contacted solar panels. Maskless, non-cryogenic reactive ion etch of silicon and subsequent coating with aluminum oxide for passivation purposes results in average reflectance of less than 1% in the UV–vis part of the solar spectrum, and in minority carrier lifetimes similar to those obtained on polished reference samples thanks to minimized surface damage, as confirmed by transmission electron microscopy. Inorganic blackening of bus-bars by a combination of electroplating and chemical etching results in reflectance values lower than 3.5% in the UV–vis, with no modifications to standard stringing equipment required.

With increasing installation of residential photovoltaic (PV) systems, distributed battery energy storage systems can be utilized for supply-demand balancing. Aggregators perform day-ahead scheduling for efficient energy management in some optimized method as ways. For optimization, however, criterions of values could change with the time or situation. In this study, we propose a method to investigate and compare various kinds of day-ahead charge/discharge schedulings based on different evaluation indices. First, we show that feasible schedulings are expressed as superpositions of reference schedulings obtained by simple strategies. Then, the method is applied to an aggregator with consumers who each has rooftop PVs and battery energy storage systems. A fairness index is defined as a correlation coefficient between battery usage and grid usage. Simulation results show a trade-off relationship between the total amount of charge/discharge of the batteries and the dispersion of power flows, and a relationship between the fairness index and the strategy of day-ahead scheduling.

A balancing group is an intermediate layer between the market and demand sides playing a key role in maintaining the demand and supply balance within the group after the massive penetration of photovoltaic (PV) systems. In this paper, the balancing group formulates an optimal plan to guarantee a highly efficient operation of conventional generators and the group's maximum output from PV generators simultaneously. The aggregator is provided a planned power flow (PPF) via a communication network by the balancing group in advance. Under these prerequisites, we propose a two-stage scheme for the demand side that consists of a day-ahead allocation algorithm of PPF and a real-time battery energy storage system (BESS) operation algorithm. The first stage of the scheme is almost the same as the case of the centralized effective BESS control, taking into account the maximum use of distributed BESSs. Owing to consumers' demand diversity, the second stage of the scheme refers to a distribution generated using past data to reduce wasteful operation. The proposed approach is verified via a numerical simulation using a measured dataset from Ota City, Japan.

White foam glass produced from waste glass as raw material is attractive from the viewpoints of waste recycling, white color with high reflectance, and weed control. It is considered to have potential for increasing the power output of bifacial photovoltaic modules, primarily owing to enhanced light reflectance from the ground and also reduced operation and maintenance cost through effective weed control. In this study, we investigated the effect of white foam glass spread in the photovoltaic installation site by measuring the albedo factor and observing the growth of weeds that are known to create shadows over a period of time. An effective increase in albedo was observed after spreading the white foam glass. The transmittance of the white foam glass was very low and the sunlight did not reach the ground below it. Consequently, plant growth and photosynthesis were obstructed and the effect of weed control was observed.

The purpose of this research is to study the drinking water production using a photovoltaic-thermoelectric hybrid system. The methodology used in this study was simulated by using a psychrometric chart and MATLAB/Simulink software to validate the operating conditions (temperature and humidity). We construct a portable drinking water system, consisting of a TEC-12706 array which is powered by a photovoltaic hybrid system. The system comprises a 50 W solar cell and an 80 AH battery. The thermoelectric module is mainly used to recover waste heat from the heat exchanger (hot side). On the basis of the principle of latent heat, molecules of water vapor can be converted into water droplets from the heat exchanger (cold side). The prototype has been tested at the Rajamangala University of Technology Lanna (Doi Saket campus), Chiang Mai, Thailand. The preliminary results showed that drinking water can be produced approximately 1 L per day at a temperature of 10–30 °C and a humidity of 60–80%. Hence, such a portable drinking water module may serve 1–2 persons per day under a flood crisis.

Accordingly to the COP21 Paris Agreement a net zero greenhouse gas emission energy system must be built no later than 2050. Such a fast power system transition will be very challenging for the conditions of Northeast Asia, a region with a large and fast growing power demand. Power system transition modelling was performed in order to check the technical feasibility of such a transition. The results of the simulation show that the transition can be accomplished and a 100% renewable energy system is both technically feasible and economically viable in Northeast Asia with average electricity generation cost of around 55 €/MWh. Solar photovoltaic (PV) will become the major energy source in Northeast Asia with a generation share of more than 70%; wind energy will contribute to 18% of the generation. Decarbonisation of the system can be achieved quite fast: by 2035 CO_2eq emissions in the power sector will decrease by 95 and 99% by 2045, respectively.