Search

We achieve large-area (1602 × 902 mm2) doping uniformity without layer peel-offs, based on a dual Na source, ie, partial Na out-diffusion from soda-lime glass and a homogeneously sputtered CuGa:NaF layer as an auxiliary source. We systematically investigate the optoelectronic, microstructural, and compositional characteristics of Cu(In,Ga)(Se,S)2 solar cells and analyze the underlying mechanism in detail. Na out-diffusion from soda-lime glass initially improves the cell performance according to the defect passivation and Na doping effects; further Na incorporation using the Na-doped layer enhances JSC, FF, and film conductivity, which is likely due to the enhanced cell homogeneity and the alleviation of carrier transport-limited characteristics. Excessive Na incorporation triggers the possible generation of layer peel-offs, which is closely related to the reduced adhesion force, void generation, decrease in S/(S + Se) ratio, Ga redistribution, bandgap reduction, and increase in the low-energy photoluminescence. These results indicate that the voids are created via Kirkendal mechanism based on the variability in atomic diffusion rates following compositional changes, resulting from the insufficiency in Na consumption or sulfurization. It is noted that an in-line codeposition technique enables realization of high-level Na doping as well as void-free interface state by suppressing the defect generation, which yields high-efficiency commercial-scale Cu(In,Ga)(Se,S)2 modules without layer peel-off problems.

Gapless interdigitated back contact (IBC) solar cells were fabricated with phosphorous back surface field on a boron emitter, using an ion implantation process. Boron emitter (boron ion implantation) is counter doped by the phosphorus back surface field (BSF) (phosphorus ion implantation) without gap. The gapless process step between the emitter and BSF was compared to existing IBC solar cell with gaps between emitters and BSFs obtained using diffusion processes. We optimized the doping process in the phosphorous BSF and boron emitter region, and the implied Voc and contact resistance relationship of the phosphorous and boron implantation dose in the counter doped region was analyzed. We confirmed the shunt resistance of the gapless IBC solar cells and the possibility of shunt behavior in gapless IBC solar cells. The highly doped counter doped BSF led to a controlled junction breakdown at high reverse bias voltages of around 7.5 V. After the doping region was optimized with the counter doped BSF and emitter, a large-area (5 inch pseudo square) gapless IBC solar cell with a power conversion efficiency of 22.9% was made.

Silicon oxynitride (SiON) could be used in combination with silicon nitride (SiN) to form a multilayer anti-reflection coating on the front side of selective emitter solar cells. In this study, these double anti-reflection layers were fabricated by a continuous deposition technique using the plasma enhanced chemical vapor deposition method. We attempted to determine whether this method is fast and cost effective and can achieve higher efficiency for solar cell manufacture. The results show that the short circuit current density for the double layer anti-reflection coating on selective emitter solar cells was higher by 0.5 mA/cm2 compared to the single layer coating owing to the improved optical reflectance. The incorporation of a SiN/SiON stack into the anti-reflection layer of the CZ selective emitter solar cells yields an energy conversion efficiency of 19.4%, which is higher than the efficiency (19.18%) for the reference solar cells with single layer SiN anti-reflection coating.

The illuminated current-voltage characteristics of Cu(In,Ga)(S,Se)2 (CIGSSe) thin film solar cells fabricated using two different buffer layer processes: chemical bath deposition (CBD) and atomic layer deposition (ALD) were investigated. The CIGSSe solar cell with the ALD buffer showed comparable conversion efficiency to the CIGSSe solar cell with CBD buffer but lower shunt resistance even though it showed lower point shunt defect density as measured in electroluminescence. The shunt paths were investigated in detail by capturing the high-resolution dark lock-in thermography images, resolving the shunt resistance contributions of the scribing patterns (P1, P3), and depth profiling of the constituent elements. It was found that the concentration of Na from the soda-lime glass substrate played a key role in controlling the shunt paths. In the ALD process, Na segregated at the surface of CIGSSe and contributed to the increase in the shunt current through P1 and P3, resulting in a reduction in the fill factor of the CIGSSe solar cells.

In this study, we systematically investigated the performance loss factors in Cu(InGa)(SSe)2 (CIGSSe) solar cells with various Cu/(Ga+In) (CGI) ratios, which were fabricated with sputtering and sequential selenization/sulfurization processes. The effects of the CGI ratio on the CIGSSe solar cells were determined by measuring the current–voltage (IV) curves and junction capacitance of the solar cells, as well as by performing X-ray analysis techniques on the cells. An increase in the defect density and decrease in the free carrier density were observed in samples with high Cu concentrations (CGI ratio >0.89), which resulted in a drastic decrease in the open-circuit voltage (VOC) and fill factor (FF). The temperature-dependent IV (IVT) and X-ray diffraction (XRD) results of the bulk characterization corresponded well with the capacitance measurements for all CGI ratios. The low crystal quality and short minority-carrier diffusion length at high CGI ratios resulted in a significant bulk recombination rate. A qualitative analysis of the interface characteristics was performed with IVT measurements, and the results showed that the recombination activation energy in the samples with high CGI ratios was lower than the bandgap (Eg), decreasing the VOC of these devices. For the samples with high CGI ratios, their inferior bulk and interface characteristics caused recombination to occur at the interface as well as in the bulk. However, the interface recombination rate was negligible for the samples with low CGI ratios (CGI ratio

This study evaluated the impact of partial shading on CuInxGa(1-x)Se2 (CIGS) photovoltaic (PV) modules equipped with bypass diodes. When the CIGS PV modules were partially shaded, they were subjected to partial reverse bias, leading to the formation of hotspots and a possible occurrence of junction damage. In a module with a cadmium sulfide buffer layer, hotspots and wormlike defects were formed. The hotspots were formed as soon as the modules were shaded; the hotspots caused permanent damage (wormlike defects) in the CIGS module. Specifically, the wormlike defects were caused by the window layer, leading to increased recombination and decay of the solar cell properties. However, a CIGS module with a zinc sulfide buffer layer did not exhibit the formation of hotspots or any visual damage. The reverse bias breakdown voltage of the CIGS PV module with the cadmium sulfide buffer layer was higher than that of the CIGS PV module with the zinc sulfide buffer layer. Copyright © 2016 John Wiley & Sons, Ltd.

We investigated the surface properties of a Mo back contact for large-area thin-film solar modules with high efficiency and good adhesion between Mo and the absorber layer. It was determined that the appropriate surface properties of Mo would improve the efficiency from 10% to above 15.0±0.21% and narrow the efficiency distribution in large-area modules. The Mo back contact was annealed at various temperatures between room temperature and 230 °C in air to control the amount of sodium diffusing from the soda-lime glass substrate during selenization and sulfurization, and to improve the uniformity of the unit cell. Before the heat treatment, the amount of sodium in the patterned area of the unit cell was more than 10 times of that in the central area of the cell. The patterned region with higher Na content had smaller grains than those in the central area with less Na, resulting in many peel-offs and shunting paths. The difference in sodium content was reduced after heat treatment. The optimized surface oxide of the Mo back contact had a thickness of around 3–5 nm and consisted of the MoO 3 phase. The grain boundary of Mo columnar structure near the surface consisted of the oxide layer.

Cu(In,Ga)(S,Se)2 (CIGSSe) solar cells with different S/(S+Se) (SSSe) ratios were fabricated using a sequential two-step sputtering and selenization/sulfurization process. The origins of the light soaking behavior in the CIGSSe absorber layer were studied based on the light soaking time (or dark storage time) dependence of the open circuit voltage (Voc) and built in potential (Vbi) as well as the temperature dependence of the current–voltage curve. In CIGSSe devices with a SSSe ratio >0.35, Voc and Vbi increase with increasing light soaking time and decrease with increasing dark storage time. In contrast, the Voc and Vbi of the CIGSSe devices with a SSSe ratio of approximately 0.25 showed slightly different behaviors, an increase with increasing light soaking time but little change with increasing dark storage time. The reasons for the different behavior in the CIGSSe devices with different SSSe ratios were attributed mainly to changes in the interface state by the SSSe ratio at the CIGSSe/buffer interface. The results suggest that the control of the SSSe ratio in the CIGSSe absorber fabricated by the proposed two-step process is an important key to determining the light soaking characteristics of the solar cell devices.

Cu 2 ZnSnS 4 (CZTS) thin films are considered as ideal absorption materials for next generation thin film solar cells due to their excellent optical and electrical properties as well as low cost. Two types of CZTS thin films were prepared via sputtering with Zn or ZnS targets (Cu/SnS/Zn and Cu/SnS/ZnS). After sulfurization, the microstructure and distribution of the elements and the electrical properties of the thin films obtained using the Zn and ZnS targets were compared. The CZTS thin films obtained from the Zn target exhibit a large grain, smooth surface and less ZnS accumulation between the CZTS and Mo, resulting in improved junction characteristics and electrical properties. The CZTS solar cells that were fabricated with the Zn target exhibit an improved conversion efficiency of 5.06% relative to that of CZTS solar cells using the ZnS target.

Cu(In,Ga)(Se,S)2 (CIGSS) modules with a frameless glass to glass (G2G) structure were successfully fabricated by replacing the currently used ethylene vinyl acetate (EVA) with new encapsulants such as polyolefins (POE), ionomer (IO) and liquid silicon (LSI), and their characteristics including optical, mechanical, and reliability properties were investigated. Most of the solid films for encapsulation except POE exhibit similar light transmittance in the visible-infrared wavelength range, while LSI shows the highest light transmittance with a difference of more than 2% in the overall wavelength range. The G2G structure encapsulated with IO exhibits superior values for the adhesion strength at which interface failure occurs (about 7.5 MPa) compared to other encapsulants. Water vapor transmittance ratio (WVTR) values of the encapsulants were evaluated to determine their suitability as water-resistant materials in terms of the CIGSS G2G structure. The IO encapsulant had the lowest WVTR value of about 1.75 g per m2 per day. Changes of cell-to-module (CTM) conversion ratio in the CIGSS G2G structure using various encapsulants were investigated using I–V measurements. The CIGSS G2G structure encapsulated by LSI shows the best CTM conversion ratio, since the module process occurs at the lowest temperature as compared with other encapsulants, therefore the module has a smaller thermal transient effect. Finally, the long-term reliability of the CIGSS modules based on various encapsulants was also evaluated by damp heat testing.

We have investigated the sulfurization process in a Cu(In,Ga)(Se,S)2 (CIGSS) absorber layer fabricated by a two-step sputter and selenization/sulfurization method in order to make an ideal double-graded band-gap profile and increase the open circuit voltage (Voc). The sulfurization process was controlled by temperature from 570 °C to 590 °C without changing H2S gas concentration and reaction time. Although the energy band-gap of the CIGSS absorber layer was increased with increasing sulfurization temperature, the Voc of the completed CIGSS device fabricated at 590 °C sulfurization temperature did not increase. In order to investigate this abnormal Voc behavior, the CIGSS absorber layer was measured by local electrical characterization utilizing Kelvin probe force microscopy, especially in terms of grain boundary potential and surface work function. Consequently, the abnormal Voc behavior was attributed to the degradation of grain boundary passivation by the strong sulfurization process. The optimum sulfurization temperature plays an important role in enhancement of grain boundary passivation. It was also verified that the Voc degradation in the CIGSS solar cell fabricated by the two-step method is more influenced by the grain boundary passivation quality in comparison with the slight non-uniformity of material composition among grains.

Passivated emitter and rear contact (PERC) is a promising technology owing to high efficiency can be achieved with p-type wafer and their easily applicable to existing lines. In case of using p-type mono wafer, 0.5–1% efficiency increase is expected with PERC technologies compared to existing Al BSF solar cells, while for multi-wafer solar cells it is 0.5–0.8%. We addressed the optimization of PERC solar cells using the Al paste. The paste was prepared from the aluminum–silicon alloy with eutectic composition to avoid the formation of voids that degrade the open-circuit voltage. The glass frit of the paste was changed to improve adhesion. Scanning electron microscopy revealed voids and local back surface field between the aluminum electrode and silicon base. We confirmed the conditions on the SiNx passivation layer for achieving higher efficiency and better adhesion for long-term stability. The cell characteristics were compared across cells containing different pastes. PERC solar cells with the Al/Si eutectic paste exhibited the efficiency of 19.6%.

Trigonal selenium (t-Se) nanowires and nanotubes were successfully prepared on a large scale via an environment-friendly synthetic process, in which no templates or surfactants were employed. These t-Se nanowires having a width of 70–100 nm and length up to tens of micrometers were synthesized in absolute ethanol at room temperature, while t-Se nanotubes with outer diameter ranging from 180 to 350 nm were obtained at 85 °C in water system. SEM and TEM analyses of the samples obtained at different stages indicated that the formation of these t-Se 1D nanostructures was governed by a “solid-solution-solid” growth process. The amorphous Se (a-Se) nanoparticles were initially generated and then would transform into crystal seeds for the subsequent growth of nanowires or nanotubes. Detailed experiments found that temperature and solvents as well as concentrations of starting materials were crucial to the formation of final morphology.

The effect of light scattering layers (400 nm, TiO particle) of 4 m thickness on the dye-sensitized solar cell has been investigated with a 12 m thickness of photo-anode (20 nm, TiO particle). Two different structures of scattering layers (separated and back) were applied to investigate the light transmitting behaviors and solar cell properties. The light transmittance and cell efficiency significantly improved with inserting scattering layers. The back scattering layer structure had more effective transmitting behavior, but separated scattering layer (center: 2 m, back: 2 m) structure (9.83% of efficiency) showing higher efficiency (0.6%), short circuit current density (0.26 mA/cm) and fill factor (0.02). The inserting separating two scattering layers improved the light harvesting, and relatively thin back scattering layer (2 m of thickness) minimized interruption of ion diffusion in liquid electrolyte.

This paper describes the formulation of nanocrystalline TiO 2 paste for dye-sensitized solar cell (DSSC) applications employing a series of TRITON™ X-based surfactants with an octyl phenol ethoxylate (OPE) backbone and various chain ends. Five different functional group-terminated TRITON™ X (or OPE) derivatives, [OPE5-OH], [OPE5-COOH], [OPE5-NH 2 ], [OPE5-Imidazole], and [OPE5-OPO(OH) 2 ], were prepared from OPE of 5 oxyethylene units ( n = 5). The carboxylic acid-terminated TRITON™ X (i.e. OPE5-COOH) was found to be the dispersant showing the lowest viscosity of [TiO 2 /ethyl cellulose/terpineol] paste. The photoanode film from the TiO 2 paste was built up on a conducting glass plate by a screen printing process followed by sintering heat treatments. Then, the TiO 2 film showed higher particle density in comparison with the control sample film prepared from the corresponding paste with no dispersant. The solar cell fabricated from the prepared TiO 2 film exhibited improved photocurrent characteristics under illumination.

Defect depth profiles of Cu (In1−x,Gax)(Se1−ySy)2 (CIGSS) were measured as functions of pulse width and voltage via deep-level transient spectroscopy (DLTS). Four defects were observed, i.e., electron traps of ~0.2 eV at 140 K (E1 trap) and 0.47 eV at 300 K (E2 trap) and hole traps of ~0.1 eV at 100 K (H1 trap) and ~0.4 eV at 250 K (H2 trap). The open circuit voltage (VOC) deteriorated when the trap densities of E2 were increased. The energy band diagrams of CIGSS were also obtained using Auger electron spectroscopy (AES), X-ray photoelectron spectroscopy (XPS), and DLTS data. These results showed that the valence band was lowered at higher S content. In addition, it was found that the E2 defect influenced the VOC and could be interpreted as an extended defect. Defect depth profile images provided clear insight into the identification of defect state and density as a function of depth around the space charge region.

We investigated the performance of dye-sensitized solar cells (DSSCs) by applying carbon nanotubes (CNTs) to the counter electrode using two different methods: screen printing and chemical vapor deposition. When the highly purified and aligned CNTs were used as the counter electrode, a conversion efficiency of DSSCs of over 10% was recorded. This CNT-deposited counter electrode shows a higher photoconversion efficiency than CNT-printed and Pt-coated electrodes under the same experimental conditions.

The microstructures and their relation to synthesis conditions of nanostructured γ-Ni-Fe alloys, synthesized by a mechano-chemical process, were investigated by using various techniques which include EDS, infrared and optical emission spectroscopy, XRD, TEM, BET and Laser particle analyzer as well as field emission SEM. The results indicate that nanostructured γ-Ni-xFe alloys with x ≈ 32, 46, 55, and 64wt% were successfully synthesized. The impurity contents in these γ-Ni-Fe alloys were very low. Typically, the microstructures of nano γ-Ni-46Fe and their variations with synthesis conditions were focused on. It was documented that γ-Ni-46Fe alloy with a minimum average grain size of 20nm and minimum average particle size ∼70nm can be obtained under specific synthesis conditions. Micro-pore analysis combined with BET and XRD experimental results indicated that the particles usually consisting of less than one hundred grains were actually small (sub-micrometer) polycrystals with no micro-pores within them. Experiments also showed that there were large hard-agglomerates which had an average size of ∼100nm in the nano γ-Ni-46Fe powders obtained at temperatures below 800°C. Above this temperature both the grain growth and the sintering process were significant. Moreover, it was revealed that upon annealing at temperatures below 800°C, the grains of nano γ-Ni-46Fe have an elongated shape with their long axis being in direction and with an aspect ratio of 1.37–1.25. Annealing at temperatures above 800°C caused their grains to change shape gradually to a disk-like form. In addition, the lattice of the nano γ-Ni-46Fe was found to be in a shrunken state with an average grain size of d (14) .

The sintering behavior of nanocrystalline (nc) γ-Ni-Fe powders was investigated by laser-photo-dilatometry. The sinterability of nc powders was found to depend crucially on the state of agglomeration. The compacted γ-Ni-Fe powders exhibited a bimodal pore distribution comprised of nanoscale intra-agglomerate pores and large, microscale inter-agglomerate pores. The results are discussed on the basis of the microstructural evolution during densification which was followed by optical microscopy and BET specific surface area measurements.

The present investigation has attempted to fabricate nanosized stainless steel powder of Fe-18Cr-8Ni(wt%) by mechano-chemical synthesis. The synthesis process was conducted by hydrogen reduction of a high energy ball milled mixture of Fe 2 O 3 -NiO-Cr(NO 3 ) 3 ·9H 2 O. In-situ alloying of Ni-Fe-Cr nanophase simultaneously occurring during reduction process is presumably responsible for the formation of nanophase stainless steel powders. The powder characteristics were examined by X-ray diffractometry and SEM observation. The kinetic phenomena in association with the oxide reduction and the alloying process were investigated by thermogravimetry, hygrometry and X-ray diffractometry, and discussed in relation to powder characteristics.

Beta-In(2)S(3) nanotubes were synthesized using an organic solution pyrolysis route. The shape of the beta-In(2)S(3) nanotubes was controlled from hexagonal nanoplates to nanotubes simply by changing the reaction time. The growth mechanism of the nanotubes was explained by oriented attachment. The beta-In(2)S(3) nanotubes had a diameter, wall thickness and length of 5.0 nm, 0.79 nm and >10 microm, respectively. The diameter of the beta-In(2)S(3) nanotubes was found to be dependent on the sulfur concentration.

Cu(In,Ga)(Se,S)2 (CIGSS) absorber layers were fabricated by using a modified two-stage sputter and a sequential selenization/sulfurization method, and the sulfurization process is changed from one-step to two-step. The two-step sulfurization was controlled with two different H2S gas concentrations during the sulfurization treatment. This two-step process yielded remarkable improvements in the efficiency (+0.7%), open circuit voltage (+14 mV), short circuit current (+0.23 mA/cm2), and fill factor (+0.21%) of a CIGSS device with 30 × 30 cm2 in size, owing to the good passivation at the grain boundary surface, uniform material composition among the grain boundaries, and modified depth profile of Ga and S. The deterioration of the P/N junction quality was prevented by the optimized S content in the CIGSS absorber layer. The effects of the passivation quality at the grain boundary surface, the material uniformity, the compositional depth profiles, the microstructure, and the electrical characteristics were exa...

We have investigated the charge collection condition around voids in copper indium gallium sulfur selenide (CIGSSe) solar cells fabricated by sputter and a sequential process of selenization/sulfurization. In this study, we found direct evidence of void passivation by using the junction electron beam induced current method, transmission electron microscopy, and energy dispersive X-ray spectroscopy. The high sulfur concentration at the void surface plays an important role in the performance enhancement of the device. The recombination around voids is effectively suppressed by field-assisted void passivation. Hence, the generated carriers are easily collected by the electrodes. Therefore, when the S/(S + Se) ratio at the void surface is over 8% at room temperature, the device performance degradation caused by the recombination at the voids is negligible at the CIGSSe layer.

After producing crystalline Si (c-Si) solar module, the output power of the c-Si module was decreased compared to the total amount of c-Si cells’ power due to the increase of series resistance and the light absorption of the cover glass and encapsulant. We define this power loss as cell to module (CTM) loss and the decreasing power rate as CTM conversion ratio (CR). The typical CTM CR in other groups has a value of about 94%. The CTM CR is typically decreased with increasing the efficiency of solar cell due to the increase of I2R loss. In this experiment, we have focused on the increase of CTM CR through detail analysis to each module component. As a result, the CTM CR in c-Si photovoltaic modules can be significantly increased by appropriately optimizing a module component. The maximum CTM CR in this experimental obtained about 97% using 19% efficiency cells, and it is 264.1 W power in 6 × 10 string module.

Solid state sintering behavior of nanocrystalline (35 nm) γ-Ni-Fe powders, which were sintered under hydrogen atmospheres at temperatures up to 1300 K, was investigated by laser-photo-dilatometry. The sinterability of the γ-Ni-Fe powders was found to depend crucially on the state of agglomeration and was generally limited to about 80 pct theoretical density. Metallographic observations revealed that the nanocrystalline powders showed a high degree of agglomeration which resulted in a difficult-to-sinter bimodal pore distribution. Although nanoscale intraagglomerate porosity was easily eliminated in the course of sintering, microscale interagglomerate porosity largely proved to resist densification. Experimentally observed shrinkage rates exhibit two densification maxima at the onset and the end of densification. No significant grain growth was observed during sintering.